1 1 5 > z 5 s A\\EUNIVER% ^11 2 I 3 I g 5 lOS-ANGElfj^ If I | "^HHAlNll-3^ ^KHJTCHtf^ ^Bfi ^OF-CAIIFO/?^ ^01 I \\\EUNIVER% v^lOSANGElfj> c? i I I P ^lOS-ANGELfj-x = =3 S i $? "5- IK uj _* I dOS ANGELA JJMVSOV* 1 %ai\INW AfUNIVERJ/A ^vlOSANCM n gi^rg 1^- i ^UIBRARY rS "1 ^ 5 S i i ~5 SAE-UNIVER% ^clOS ANGEL ^ ^r i s .\FUNIVK/A ^Of-CAUFORfc METABOLISM AND GROWTH FROM BIETH TO PUBERTY FRANCIS G. BENEDICT AND FRITZ B. TALBOT PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON WASHINGTON, 1921 CARNEGIE INSTITUTION OF WASHINGTON PUBLICATION No. 302 PRESS OF THE NEW ERA PRINTING COMPANY LANCASTER, PA. W5 103 METABOLISM AND GEOWTH FROM BIRTH TO PUBERTY BY FRANCIS G. BENEDICT AND FRITZ B. TALBOT U4168.'} CONTENTS. PAGE Introduction 1 Basal metabolism 2 Previous studies of metabolism of children 4 Control experiments and basal metabolism 21 History and plan of research 25 Apparatus and experimental technique 28 Respiration apparatus 28 Experimental conditions 30 Discussion of results 32 Normality of cnildren studied 33 Standards for determining the normality of children 34 Earlier data selected for comparison with our data 36 Relationship between body-weight and age with boys 39 Relationship between height and age with boys 41 Relationship between body-weight and age with girls 42 Relationship between height and age with girls 43 General consideration of the ratios of body-weight to age and height to age with boys and girls 44 Relationship between height and body-weight with boys and girls 45 Growth 47 Anthropometric measurements as indices of growth 52 Physiological and anthropometrical significance of surface area 63 Relationship between surface area and body-weight, height, and age with boys 63 Relationship between surface area and body-weight, height, and age with girls 66 Normal, average, and ideal states of nutrition 69 Body-weight in relation to height as an index of state of nutrition 72 Pulse-rate 75 Methods of obtaining pulse-rate 77 Influence of age upon the pulse-rate 78 Average pulse-rate of children 81 Sex and minimum pulse-rate 83 Average pulse-rate with increasing age 85 Rectal temperature 85 Influence of food on metabolism 89 The element of novelty in measurements of metabolism 97 Metabolism as affected by growth 100 General methods of study 100 Metabolism during growth as shown by the individual child 101 Observations with subject. No. 145 101 Observations with 22 children during periods of 4 months to 3 years 115 Metabolism during growth as shown by groups of individual data 128 Method of grouping data 131 General trend of metabolism with increasing age 133 Total calories per 24 hours referred to age (boys) 133 Total calories per 24 hours referred to age (girls) 134 Total metabolism of children referred to age (earlier investigators) 136 General conclusions as to total metabolism and age in children 139 Total metabolism with increasing body-weight 139 Total calories per 24 hours referred to weight (boys) 140 Total calories per 24 hours referred to weight (girls) 141 Total metabolism of children referred to weight (earlier investigators) 142 iii IV CONTENTS. PAGE. Metabolism as affected by growth Continued. Metabolism during growth as shown by groups of individual data Continued. Metabolism per unit of body-weight referred to age 145 Calories per kilogram of body-weight per 24 hours referred to age (boys) ... 146 Calories per kilogram of body-weight per 24 hours referred to age (girls) . . . 149 Calories per kilogram of body-weight per 24 hours referred to age (earlier investigators) 150 Metabolism per unit of body-weight referred to weight 153 Calories per kilogram of body-weight per 24 hours referred to weight (earlier investigators) 157 Relationship between surface area of the body and metabolism 159 Total calories per 24 hours referred to actually measured body-surface area (boys) 161 Total calories per 24 hours referred to actually measured body-surface area (girls) 164 Calories per square meter of body-surface referred to body-surface 165 Comparison of the caloric output per square meter of body-surface with total body-weight 167 Comparison of heat production per square meter (measured), referred to body-weight, with earlier data (computed) 169 Age relations in the heat production per square meter of body-surface 173 Influence of sex and sexual change on metabolism 176 Metabolism in prepubescence 183 The prediction of the basal metabolism of youth 187 Predicted heat from total calories referred to weight (boys) 188 Predicted heat from total calories referred to surface (boys) 194 Comparison of the predicted metabolism of boys and men 197 Predicted heat from total calories referred to weight (girls) 198 Predicted heat from total calories referred to surface (girls) 203 Practical value of the prediction of basal metabolism 205 The 24-hour energy requirements of growing children 206 ILLUSTRATIONS. PAGE FIG. 1. Infant respiration apparatus as used at the Directory for Wet Nurses 30 2. Respiration apparatus as used for children at the New England Home for Little Wanderers 30 3. Relationship between body-weight and age with boys 40 4. Relationship between height and age with boys 41 5. Relationship between body-weight and age with girls 43 6. Relationship between height and age with girls 44 7. Relationship between height and body-weight with boys 45 8. Relationship between height and body-weight with girls 46 9. Relationship between body-surface and body-weight with boys 64 10. Relationship between body-surface and height with boys 65 11. Relationship between body-surface and age with boys 66 12. Relationship between body-surface and body-weight with girls 67 13. Relationship between body-surface and height with girls 67 14. Relationship between body-surface and age with girls 68 15. Body-weight, pulse-rate, and basal heat production per 24 hours, No. 145 . . 114 16. Body-weight, pulse-rate, and basal heat production per 24 hours, Nos. 139 and 171 124 17. Body-weight, pulse-rate, and basal heat production per 24 hours, Nos. 119, 122, 127, and 138 125 18. Body-weight, pulse-rate, and basal heat production per 24 hours, Nos. 113, 126, 131, and 142 126 19. Body-weight, pulse-rate, and basal heat production per 24 hours, Nos. 115, 123, 136, and 160 127 20. Body-weight, pulse-rate, and basal heat production per 24 hours, Nos. 148, 161, 172, and 173 129 21. Body-weight, pulse-rate, and basal heat production per 24 hours, Nos. 153, 155, 158, and 166 130 22. Basal heat production of boys per 24 hours referred to age 133 23. Basal heat production of girls per 24 hours referred to age 135 24. Basal heat production of boys per 24 hours referred to age (earlier investi- gators) 137 25. Basal heat production of girls per 24 hours referred to age (earlier investi- gators) 138 26. Basal heat production of boys per 24 hours referred to body-weight 140 27. Basal heat production of girls per 24 hours referred to body-weight 142 28. Basal heat production of boys per 24 hours referred to body-weight (earlier investigators) 143 29. Basal heat production of girls per 24 hours referred to body-weight (earlier investigators) 144 30. Basal heat production of boys per kilogram of body-weight per 24 hours referred to age 147 31. Basal heat production of girls per kilogram of body-weight per 24 hours referred to age 149 32. Basal heat production of boys per kilogram of body-weight per 24 hours referred to age (earlier investigators) ; 150 33. Basal heat production of boys per kilogram of body-weight per 24 hours referred to age (Sonde"n and Tigerstedt, C. Tigerstedt, and B. and T.) ... 152 34. Basal heat production of girls per kilogram of body-weight per 24 hours referred to age (Sonde"n and Tigerstedt, and B. and T.) 152 35. Basal heat production of boys per kilogram of body-weight per 24 hours referred to weight 154 v VI ILLUSTRATIONS. PAGE. FIG. 36. Basal heat production of girls per kilogram of body-weight per 24 hours referred to weight 155 37. Basal heat production of boys per kilogram of body-weight per 24 hours referred to weight (earlier investigators) 157 38. Basal heat production of boys per 24 hours referred to body-surface 161 39. Basal heat production of girls per 24 hours referred to body-surface 164 40. Basal heat production of boys per square meter of body-surface per 24 hours referred to surface 165 41. Basal heat production of girls per square meter of body-surface per 24 hours referred to surface 166 42. Basal heat production of boys per square meter of body-surface per 24 hours referred to body-weight 167 43. Basal heat production of girls per square meter of body-surface per 24 hours referred to body-weight 169 44. Basal heat production of boys per square meter of body-surface per 24 hours referred to body-weight (earlier investigators) 171 45. Basal heat production of boys per square meter of body-surface per 24 hours referred to age 174 46. Basal heat production of boys per square meter of body-surface per 24 hours referred to age (earlier investigators) 175 47. Basal heat production of girls per square meter of body-surface per 24 hours referred to age 175 48. Comparison of basal heat production of children and adults per 24 hours referred to body-weight . . . . * 179 49. Comparison of basal heat production of children and adults per 24 hours referred to body-surface 180 50. Comparison of basal heat production of children and adults per kilogram of body-weight per 24 hours referred to weight 181 51. Comparison of basal heat production of children and adults per square meter of body-surface per 24 hours referred to weight 181 52. Basal heat production of men per kilogram of body-weight per 24 hours referred to weight 182 53. Basal heat production of women per kilogram of body-weight per 24 hours referred to weight 182 54. Basal heat production of men per square meter of body-surface per 24 hours referred to body-weight 183 55. Basal heat production of women per square meter of body-surface per 24 hours referred to body-weight 183 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. INTRODUCTION. In the past decade the trend of thought in the physiology of growth has been towards a chemical analysis of the several growth factors. The embryonic animal (which with mammals receives nourishment from its mother through the placenta, and with other animals from the previously deposited food material in the egg) grows in accordance with the nutrients supplied. After birth various types of foods are brought to it by the mother or by other agencies. The selection of the diet best fitted, both in amount and in quality, to acquire growth has received a great deal of experimental attention. The importance of the mineral constituents and the nature of the proteins used in the construction of new tissue have been emphasized; particularly, the so-called " unidentified food accessory substances," which make for growth, have been exhaustively studied by Hopkins, 1 Osborne and Mendel, 2 and McCollum. 3 As a result of these extensive investigations of the subject, it is clear that a large number of factors, heretofore almost neglected in research, are absolutely essential for the proper growth of the immature animal. With a study of these essentials there has proceeded, although perhaps with less intensity, a study of certain physiological constants characteristic of the immature animal, particularly of the human animal. While the anthropologists have given us extensive measure- ments of the growth in that period of development in which growth is most marked, i. e., in the earlier years of life, relatively little is known regarding the fundamental basal metabolism during this period. The Nutrition Laboratory, in the belief that a careful survey should be made of the metabolism of all mankind from birth to old age, has been occupied for nearly a decade in the charting of this little-known field of human basal metabolism. The writers' special province has been that of infants and children. As an indication of the extent and thoroughness of the original program in its several subdivisions, we may call attention to our report of a research on the physiology of the new-born infant, in which over 100 new-born infants were studied, 1 Hopkins, Journ. Physiol., 1912, 44, p. 425. 5 Osborne and Mendel, Carnegie Inst. Wash. Pub. No. 156, 1911. 3 McCollum, The newer knowledge of nutrition. New York, 1918. 1 2 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. many a few minutes after their birth, and all during the first week of life. 1 An earlier report of a study of infants has also been published, with the primary object of indicating methods of study and certain physiological correlations, particularly pulse-rate and metabolism. 2 At the tune of this earlier report, the number of normal individuals obtainable was relatively few and a somewhat large number of atrophic infants, with a few pathological cases, were included to show what might be expected for variation in metabolism in the hospital ward. With the completion of the study of new-born infants, the next step was the study of the basal metabolism of children at various ages from one week to puberty. The results of this latest investigation we pur- pose presenting in this report. BASAL METABOLISM. Basal metabolism may be considered as the sum total of all the vital activities of the quiet organism in the post-absorptive condition, i. e., the minimum or maintenance metabolism unaffected by extrane- ous factors. This may be expressed in terms of heat produced or of gaseous exchange incidental to heat production (carbon-dioxide pro- duction and oxygen consumption). Using this basal metabolism as a standard, we may then measure definitely the effect of superimposed factors. In these studies of the basal metabolism from birth to puberty, it has been our aim to determine the metabolism at the different ages under identical conditions as far as possible, so that the results may be entirely comparable. To obtain comparable results with indi- viduals of varying ages, certain experimental conditions should exist. In the first place, in view of the pronounced influence of muscular activity upon basal metabolism, it is desirable that all subjects should have the same degree of muscular repose. An ideal condition would be complete muscular repose. The difficulty of securing such repose with young children is obvious, for infants of one or two years of age differ widely in muscular movements and temperament from an adolescent boy or girl. With adults and older children the muscular activity may be voluntarily so controlled that the increase in meta- bolism due to activity, even in restless periods, will be but 15 to 20 per cent. With crying children, with whom the activity is involuntary, the increment was found in our earlier studies to be at times over 200 per cent and on an average 65 per cent. 3 Second, it may be stated that, theoretically at least, the ingestion of any energy-producing food-material increases the metabolism by 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915. 2 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. For a briefer report, see Benedict and Talbot, Am. Journ. Diseases of Children, 1914, 8, p. 1. 3 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915, p. 112, table 17. INTRODUCTION. 3 stimulating the cells to greater activity. This stimulus is greatest with protein foods and least with fats. To secure comparable results, therefore, the ideal condition would be to study all subjects in the post-absorptive state, i. e., about 12 hours after the last meal, when the influence of the preceding diet had disappeared. With such differ- ences in dietetic habits, times of eating, and stomach capacity as exist between infants a year old and children 12 years old, the obtaining of ideally comparable conditions in this respect is likewise difficult. The third factor to be considered in determining the basal meta- bolism is that of sleep. While certain observations made in a number of laboratories imply that sleep, per se, is without profound effect upon the metabolism, yet in the light of our experience in this labor- atory during the past decade we are strongly inclined to think that this is an erroneous conception. Furthermore, the opportunities for error in the usual methods of determination of the metabolism during deep sleep are great. When breathing appliances, such as moutn- piece, nosepieces, or mask, are used in our experiments, the subjects have been for the most part required to keep awake on the general ground that with sleep the facial muscles relax, especially those about the mouth, and there is danger of leakage of air under these conditions. Practically the only method of experimenting which gives dependable results with a sleeping subject is the chamber method, i. e., with the subject asleep inside one of the several forms of respiration chamber. That sleep is a factor of great significance was clearly demonstrated in this laboratory with the subject of a fasting experiment. 1 Recent experiments, as yet unpublished, show that the metabolism of a sleeping subject differs appreciably from that of a subject awake, even though in both periods there is the greatest degree of muscular repose. A cursory examination of the data presented in the literature on the metabolism of children shows a wide diversity of results. The literature on the metabolism of infants, especially of the new-born, has already been reviewed briefly in the two publications giving the results of our earlier studies. 2 While it is unnecessary to refer here in detail to the work already cited, it seems desirable, before giving the results of our own observations, to supplement the previous review of the literature by a citation of the observations made by other workers with older children. 1 Benedict, Carnegie Inst. Wash. Pub. No. 203, 1915, pp. 343 et seq. 2 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, pp. 11 to 22, and Carnegie Inst. Wash. Pub. No. 233, 1915, pp. 12 to 38. PREVIOUS STUDIES OF METABOLISM OF CHILDREN. In considering the previous studies with children, one must differ- entiate sharply between measurements of metabolism which are possible only with respiration apparatus or calorimeters and the computation of the energy requirement for the day from a statistical study of the diets consumed by the infants or children observed. The researches on energy transformation of children may be divided into two classes: (1) those by the indirect method, in which a study is made of the respiratory exchange by means of some form of respira- tion chamber, and (2) those by direct calorimetry, in which the heat output is determined with a calorimeter. Beginning with the calorimetric researches of Richet, a number of French writers conducted researches on the metabolism of infants by direct calorimetry, several of these observations being made with calorimeters devised by d'Arsonval. These studies have already been summarized in one of our earlier reports, 1 but they have little present- day value, for the investigators disregarded the somewhat considerable withdrawal of heat from the body by the vaporization of water and gave no quantitative information regarding muscular activity. Fur- thermore, much of the French work was done with abnormal children, while our report deals exclusively with the physiology of normal youth. Calorimetric observations of the basal metabolism of children have also been made by Du Bois and associates, which will be cited in some detail later. The greater part of their computations and the deduc- tions therefrom are, however, based upon indirect calorimetry rather than upon direct calorimetric measurements. The observations of the basal metabolism of infants made by Howland with Lusk's calori- meter, 2 an abstract of which was given in our earlier report, are, so far as we know, the only successful studies of infants which have thus far been made by the direct method. Practically all of the observa- tions cited in the following pages, therefore, are those made by the indirect method. Andral and Gavarret, 1843. The first experiments on children, made by Andral and Gavarret, 3 although having mainly an historical interest, should certainly be considered in any careful analysis of the literature. Since at this early date the authors laid special stress upon 1 Detailed reference3 to the researches of these investigators are given in Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, pp. 11 to 14. Howland, Proc. Soc. Exp. Biol. and Med., 1911, 8, p. 63; Hoppe-Seyler's Zeitschr. f. Physiol. Chem., 1911, 74, p. 1; Trans. 15th Int. Congress on Hygiene and Demography, Washing- ton, 1913, 2, p. 438. Cited by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 22. 3 Andral and Gavarret, Ann. d. Chim. et d. Phys., 1843, ser. 3, 8, p. 129. 4 PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 5 the influence of puberty, the figures are, in the light of modern studies at this age, worthy of even greater consideration. Andral and Gavarret employed a copper mask attached to the subject's face and collected the expired air in large glass globes, this air being subsequently ana- lyzed. They report a large series of experiments (8 to 13 minutes long) with a great many individuals, some of them children. The experiments were all made in the early afternoon and at the same interval after food, and an attempt was made on the part of the experi- menters to have all subjects with the same degree of muscular activity, presumably in the sitting position. Although reported on the basis of carbon per hour, their data have been recalculated to the basis of carbon dioxide by Sonden and Tigerstedt 1 and are presented in table 1 as given by them, together with our computations of the heat produc- tion per 24 hours. TABLE 1. Metabolism of boys and girls (Andral and Gavarret). Carbon Heat Carbon dioxide (computed) Subjects. Age. per produced per Remarks regarding muscular system. hour. per 24 hour. 1 hours. 3 yrs. grams grams cols. Boys . . . 8 5.0 2 18.3 1,318 Average. 10 6.8 24.9 1,793 Well developed. 11 7.6 27.9 2,009 Do. 12 7.4 27.1 1,951 Average. 12 8.3 30.4 2,189 Very well developed. 14 8.2 30.1 2,167 Average. 15 8.7 31.9 2,297 Do. Girls... 10 6.0 22.0 1,584 Well developed. 1 11 6.2 22.8 1,642 Do. 1 Puberty not 13 6.3 23.1. 1,663 Mediocre. | established. 16* 7.1 26.1 1,879 Very well developed. J 15| 6.3 23.1 1,663 Average; puberty established. 1 Computed by Sonden and Tigerstedt, Skand. Arch. f. Physiol., 1895, 6, pp. 54 and 56. 2 Sonden and Tigerstedt give 16.7 gms. COz; this value corrected by us. 3 Heat computed from carbon-dioxide production, assuming 3 calories per gram of carbon dioxide. The authors conclude that with males there is a steady increase in the carbon-dioxide production between 8 and 30 years. At the time of puberty this increase suddenly becomes very large. Thus, the amount of carbon burned in one hour's time increases progressively from 5 grams in the case of a boy 8 years old to 8.7 grams in the case of a boy 15 years old, while with a young man of 16 years the amount was 10.2 grams. With well-developed males from 20 to 28 years of age the values for carbon increase gradually to 11.2 grams and 12.4 grams. With the females there is likewise an increase in carbon-dioxide production with increasing age, but at the time of puberty this increase Sonden and Tigerstedt, Skand. Arch. f. Physiol., 1895, 6, pp. 54 and 56. METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. suddenly ceases and the amount of carbon-dioxide produced remains stationary, being nearly what it was in childhood. Thus, with a girl 13 years old the carbon burned per hour was 6.3 grams; with a girl 15| years old, who had not yet reached puberty, the value found was 7.1 grams, while with another girl of the same age, who had reached puberty, the value was only 6.3 grams. Comparing the two sexes, Andral and Gavarret conclude that the male child burns more carbon than the female, on the average about 1 gram more of carbon per hour. The absence of weights and heights in the report lessens its value, but in view of the subsequent researches of Du Bois and certain data accumulated by us these early conclusions with regard to the influence of puberty are of special significance. Scharling, 1848. Publishing simultaneously with Andral and Gavar- ret, Scharling 1 reports experiments with a number of subjects, two of them children of the age-range we are considering in this report, one a boy of 9f years of age weighing 22 kg., and one a girl of 10 years weighing 23 kg. Scharling's apparatus was a respiration chamber, a brief description of which has been given in a previous report. 2 The values of special significance in connection with the discussion of basal metabolism are those on the boy and girl in table 2. TABLE 2. Basal metabolism of children (Scharling). Subject. Age. Body- weight. Carbon per hour in Danish "gran." Carbon dioxide per hour. 1 Carbon dioxide per kg. per 24 hours. Heat (com- puted) per 24 hours.* Heat (com- puted) per kg. per 24 hours. Boy Girl yrs. 9f 10 kilos. 22 23 76.2 74.8 65.5 75.1 grams. 17.2 16.9 15.0 17.2 grams. 18.7 18.4 15.7 17.9 cats. 1,238 1,217 1,080 1,238 cals. 56.3 55.3 47.0 53.8 1 In recalculating carbon to basis of carbon dioxide, estimated that 1 Danish "gran" (the measure used by Scharling) equals 0.0621 gram. 2 Heat computed from carbon-dioxide production, assuming 3 calories per gram of carbon dioxide. Forster, 1877. The next recorded observations on children are the studies made by Forster 3 in Munich in 1877. These are presented in a fragmentary manner in somewhat inaccessible publications. Certain of Forster's results were abstracted in our two previous publications, but as his studies have a special historical interest as being the first 1 Scharling, Ann. d. Chem. u. Pharm., 1843, 45, p. 214; reprinted in detail in Ann. d. Chim et d. Phys.,. 1843, ser. 3, 8, p. 478. 2 Benedict and Carpenter, Carnegie Inst. Wash. Pub. No. 261, 1918, p. 13. 'Forster, Amtl. Ber. d. 50 Versamml. deutsch. Naturf. u. Aerzte in Miinchen, 1877, p. 355; also v. Ziemssen's Handbuch der Hygiene, Leipsic, 1882, 1, p. 76. See also Magnus-Levy and Falk, Archiv f. Anat. u. Physiol., 1899, Suppbd., p. 356. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 7 made by the Pettenkofer and Voit method, and include observations on older children, it seems advisable to consider them somewhat more extensively here. In his recognition of the importance of a study of the metabolism of youth, Forster was undoubtedly stimulated by the wonderful observations of Pettenkofer and Voit. The statement made in our earlier publication, 1 that he used the large Pettenkofer- Voit respiration chamber, is obviously erroneous; we have been unable to find a description of the exact size of chamber that he used. In amplification of the records of Forster's investigations previously reported by us, the following data should be given: Forster 2 states that he studied a number of children varying in age from 14 days to 13 years, using a Pettenkofer- Voit respiration chamber and determining the carbon-dioxide excretion. The experiments with nursing infants, which continued from 3 to 5 hours each, were made in the intervals between the nursings. With the other children the experiments were made in the morning. The supper given the children the night before the experiment consisted of milk with a little bread; from 1| to 2 hours before the beginning of the experiment, they all received 100 grams of milk and a small piece of bread (about 50 grams). During the experiments the nursing infants slept the greater part of the time; the other children remained very quiet, for the most of the time sitting and with nothing special to do. In con- formity with the usage of a number of writers at about that period, Forster reported his results as grams of carbon-dioxide excreted per hour for each 10 kg. of body-weight. The results are given in table 3. TABLE 3. Carbon-dioxide production of children (Forster) Subject. Age. Carbon-dioxide production per 10 kg. per hour. Girl (nursing infant) 14 days grams. 9.0 Boys and girls 3 to 5 years 11.7 Do 6 to 7 years 11.7 Do 9 to 13 vears 8.9 These values, which are not far from 10 to 12 grams per hour for each 10 kg. of body-weight, Forster compares with those obtained by Pettenkofer and Voit for men calculated to the same basis of body- weight. The values for adults are, for the most part, one-half of those found with the children. Forster concluded that, since the children during the experiment were usually quiet and with approxi- mate hunger, there is a very much greater oxidation of nitrogen-free material with children than with adults under the same external 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 11. 'Forster, v. Ziemssen's Handbuch der Hygiene, Leipsic, 1882, 1, p. 76. 8 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. conditions. He pointed out that this is likewise true even during sleep, when the metabolism is lower, for Lewin's adult subject 1 gave off in sleep 3.5 grams of carbon dioxide per 10 kg. of body-weight per hour, while Forster's sleeping infant gave off 9 grams of carbon dioxide. Speck, 1889. Speck's observations on children really antedated the work of Forster, for, although his results were not published until 1889, 2 experiments on a 13-year-old girl, weighing 35 kg., 'were begun on September 2, 1871, and continued at intervals for two or three weeks. Speck points out that this child used less oxygen than a normal man, but when the results were computed on the basis of oxygen consumption per kilogram of body-weight the metabolism of the child appeared to be more intensive than that of the adult. All the experiments with this subject were made in the standing position and without food. Beginning September 5, 1877, six experiments in about two weeks were made with a 13-year-old boy, weighing 38 kg. During these observations the boy was without food and sat reading quietly. Both the total carbon-dioxide production and the oxygen consumption were less than with the average man, but much higher when computed on the basis of body- weight. Thus the total oxygen consumption was 239 c. c. per minute and 6.3 c. c. per kilogram of body-weight; the respiratory quotient was 0.824. Still another series of experiments was made by Speck August 22 to 29, 1881, the subject being a 10-year-old girl, weighing 25.4 kg. This girl, when studied 5 hours after breakfast and while she was sitting quietly reading, gave low values as compared to those obtained with a man, but on the basis of body-weight the results were much higher. The average total oxygen consumption was 172 c. c. per minute and 6.9 c. c. per kilogram of body-weight per minute; the respiratory quotient was 0.855. Since a normal man gave about 4 c. c. of oxygen per kilogram of body-weight per minute, he concluded that a small body absorbs relatively more oxygen than a larger one and that during the years of growth the oxygen absorption is greater than with adults; furthermore, that under like conditions males have a greater metab- olism than females. As Speck's experiments were made with the subject sitting (and in one instance standing) instead of with the subject lying, and in a number of instances with food in the stomach, the results are not strictly comparable with those of modern work. It is, however, sig- nificant that Speck's conclusions, based upon these physiological observations with persons of various ages, and particularly with him- self, have been unaltered by the results obtained by most modern writers. His conclusions regarding metabolism during youth are in accord with most subsequent work. iLewin, Zeitschr. f. Biol., 1881, 17, p. 71. * Speck, Physiologic des menschlichen Athmens, Leipsic, 1892, p. iv PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 9 Sonden and Tigerstedt, 1895. Concurrently with the description of the large respiration chamber constructed in Stockholm, Sonden and Tigerstedt reported an extensive series of observations of the resting metabolism of groups of individuals of various ages inside the respira- tion chamber. 1 The prime object of their study was to find the carbon- dioxide output of groups of individuals for use in determining the needs of school-houses and other public buildings for ventilation. Accordingly the investigators attempted to adjust the conditions of their experiments to comply so far as possible with the demands of the rooms and buildings under consideration. The subjects were always in the sitting position, usually after a breakfast, and not infrequently were eating small amounts of candy or apples. The experimental conditions were therefore not ideal for measurements of the basal metabolism. It is no adverse criticism of this wholly remarkable research to state frankly that the results are of no par- ticular value for comparison with later experiments carried out with the primary object of measuring basal metabolism. Notwithstanding this, probably no one research has contributed more to general in- formation as to the caloric requirements of human individuals at different ages than has this study of Sonden and Tigerstedt. The apparatus permitted the determination of the carbon-dioxide con- sumption only, but the investigators were able to compute the energy requirement from this with a considerable degree of accuracy, espe- cially when the food conditions existing prior to and during the experi- ments are considered. The experiments were made with subjects varying widely in age, but our own interest in the series at this time centers on the ages prior to puberty. Fortunately their research included an extensive series on children of both sexes, the studies usually being made on groups of six subjects. The boys ranged in age from 7 years and 314 days to 14 years and 199 days, and the girls from 7 years and 316 days to 14 years and 15 days. All of the children were taken from the schools in Stockholm and remained reasonably quiet during the observations, sitting in chairs and reading. In most of the experiments they ate apples and occasionally candy, but every effort was made by the experimenters to minimize the extraneous muscular activity. The results obtained with both boys and girls are given in table 4, but in considering the data it should be borne in mind that they are not primarily basal values. Sonden and Tigerstedt report their results as carbon dioxide excreted per 6 individuals and per half hour in grams. In presenting the data here we have converted them to calories per kilogram per 24 hours per individual, using the respiratory quotient 0.90 and the calorific value of carbon dioxide for this respira- tory quotient, i. e., 2.785 calories per gram of carbon dioxide. 1 Sonden and Tigerstedt, Skand. Arch. f. Physiol., 1895, 6, p. 1. 10 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 4. Minimum carbon-dioxide production and heat production of boys and girls (Sonden and Tigerstedt). Subjects. Average age. Average body-weight (without clothing). Carbon-dioxide production per indi- vidual per hour. Heat (computed) per kg. per 24 hrs. 1 yrs. days. MM, grams. cafe. Boys 7 314 20.1 20.7 68.8 9 217 27.5 28.0 68.1 10 192 30.2 30.7 67.9 11 143 31.6 30.3 64.1 12 173 34.1 30.7 60.2 13 313 44.5 43.0 64.6 14 199 45.3 39.3 58.0 Girls.... 7 316 21.8 22.3 68.4 9 334 26.6 19.7 49.5 11 57 31.0 24.0 51.7 12 68 36.2 25.7 47.5 13 53 39.5 25.0 42.3 14 15 44.3 27.0 40.7 1 In computing heat assumed respiratory quotient of 0.90. Usually the observations continued for periods of 4| hours. In order to approximate the minimum value for comparison purposes, we have tabulated only the absolute minimum found in the results. This was done on the theory that this minimum value represents the actual minimum metabolism during the period, for otherwise there would be an error in the experimental technique, which is improbable in view of the usual accuracy in experimentation of the Scandinavian investigators. 1 Recognizing the importance of obtaining data as nearly as possible during complete muscular repose, Sonden and Tigerstedt likewise made a few experiments with individuals who slept inside the respira- tion chamber, these including two experiments with boys. The first was made with a boy 11 years and 2 months old, weighing without clothing 32 kg. The experiment began at 6 p.m. The subject ate supper inside the chamber at 8 h 15 m p.m., took milk at 10 h 30 m p.m., and then went immediately to bed. The carbon-dioxide excretion was determined in 2-hour periods throughout the night. The mini- mum values were found between 2 and 6 a.m., the values for the two periods being 41 and 37 grams, respectively. The latter figure was the absolute minimum value for the carbon-dioxide production per 2 hours. The second experiment was made with a boy 12 years of age, weighing without clothing 38.3 kg. The experimental conditions were almost identical with those of the first experiment; his last meal prior to the experiment was at 4 p.m. He went to bed at 10 p.m. The minimum carbon-dioxide production was found at 2 a.m., i. e., 40 grams per 2 hours. The carbon-dioxide excretion in the period For the one apparent pronounced exception to the usual extraordinary accuracy of the Sond6n and Tigerstedt technique, see critique of von Willebrand's work, page 14 of this monograph. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 11 just preceding the minimum period was 41 grams and that in the period following was 42 grams. For computing the caloric output during this time, we are justified in using the respiratory quotient of 0.90 for the first experiment and probably 0.87 for the second experiment. On this basis the minimum caloric requirements for these two boys would be as given in table 5. These values may more properly be used for TABLE 5. Minimum metabolism of boys during sleep (Sonden and Tigerstedt). Age, 11 years 2 months. Age, 12 years. Remarks. Body-weight, without clothing, kilos. Carbon-dioxide production per hour grams. Heat, computed, per 24 hours, cals. 32.05 18.5 1,237 38.30 20.0 1,373 In computing heat, assumed a respiratory quotient of 0.90 for the 11-year old boy and of 0.87 for the 12-year old boy, as the latter had had some- what less food before retiring for the night. purposes of comparison with our own results, as they were obtained with the subject in the lying position, probably in deep sleep, which in considerable part may offset the influence of food. No night experi- ments during sleep were made with girls. Hellstrom, 1900. Employing presumably the large Stockholm respiration chamber, Hellstrom 1 studied the metabolism of a 9-year-old child weighing 23.2 kg. and found that the child had a heat production of 63 calories per kilogram per 24 hours or 1,499 calories per square meter per 24 hours. Since the original place of publication of Hell- strom's article has been inaccessible to us, we have had to rely upon scattered statements concerning this experiment by von Willebrand, 2 by R. Tigerstedt, 3 and more recently by C. Tigerstedt. 4 As we have no information at hand with regard to the activity of the child or the height and sex, the experiment is cited here simply as a matter of historical record. Rubner, 1902. Rubner published an interesting series of observa- tions on two brothers, one fat and the other thin; also a few experi- ments on two other boys. 5 As at that time the importance of muscular rest in measurements of basal metabolism had not been so greatly emphasized as at present, these children were allowed entire freedom of movement inside the respiration chamber. The two brothers remained in the chamber for 22 out of the 24 hours, the other two boys only 4 hours. As a result of the experimental technique Rubner 1 Hellstrom, Studier ofver mjolken sasom foda, Helsingfors, 1900, p. 132. 2 von Willebrand, Finska Lakaresallskapets Handlingar, 1907, 49, p. 421. R. Tigerstedt, Nagel's Handbuch d. Physiol., Braunschweig, 1909, 1, p. 476. * C. Tigerstedt, Ueber die Nahrungszufuhr des Menschen in ihrer Abhangigkeit von Alter, Geschlecht und Beruf, Helsingfors, 1915, p. 77; see also Skand. Arch. f. Physiol., 1916, 34, p. 151. t Rubner, Beitrage zur Ernahrung im Knabenalter, Berlin, 1902. 12 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. doubtless had data for some periods in which the subjects were asleep, but these were not published a fact to be regretted, for although the data were contaminated by the fact that food had been taken previ- ously, it is highly probable that the compensatory influence of sleep would in large part offset the stimulus due to the food. Since these results are not available, we are unable to use for comparison any of the data obtained by Rubner in this important research. Rubner compares the values found with these children with those reported by Sonden and Tigerstedt. In almost every instance he finds that the carbon-dioxide production per square meter per hour noted by Sonde'n and Tigerstedt was greater than that found by himself. Rubner does not believe these differences can be entirely accounted for by the restlessness of the subjects, although this must play a considerable role. We find ourselves in full accord with Rub- ner's severe criticism of the Sonden and Tigerstedt data, especially as to their use for comparison; we likewise believe that values obtained on this basis can not be used for a satisfactory demonstration of a material alteration in the basal metabolism of youth as compared with that in old age. It still remains a fact, however, that practically all of the criticisms raised by Rubner against the experiments of Sonde'n and Tigerstedt also apply to Rubner's own experiments. For instance, Rubner criticizes the fact that Sonde'n and Tigerstedt's subjects received food, but the process of digestion likewise had a part in his own experiments, although not so great as in those of Sonden and Tigerstedt. He furthermore takes exception to the fact that Sonden and Tigerstedt's subjects moved about in the chamber or were restless, but his subjects moved and indeed walked about in the chamber at tunes. Rubner's criticisms make it especially clear that only data obtained in the post-absorptive condition and in complete muscular repose are ideally suitable for comparing the metabolism of individuals of different ages. These conditions were not secured by either Sonden and Tigerstedt or by Rubner, although the values obtained by the Scandinavian investigators with the two boys studied throughout the night approximate very closely the basal metabolism, i. e., very nearly comply with basal conditions and prerequisites. Since Rubner wrote so late as 1902, it is somewhat surprising that no mention was made of the research of Magnus-Levy and Falk pub- lished in 1899, which met all of the objections raised by Rubner to the Sonde'n and Tigerstedt experiments. These experiments of Magnus-Levy and Falk were made, however, for the most part during relatively short periods, and Rubner strongly objects to respiration experiments made in short periods, an objection with which we can not agree. Rubner's chief conclusion is that the metabolism of nursing infants per square meter of body-surface is no larger than that of adults, although he does find an increased value for young boys. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 13 This increase he attributes wholly to differences in restlessness and activity and the greater intensity of development of the muscular system, an inherent restlessness which he thinks can not be controlled with young boys. If his results obtained during sleep had been reported, much more definite conclusions could have been drawn. Magnus-Levy and Folk, 1899. Magnus-Levy and Falk 1 presented in 1899 the results of the first systematic study of the basal metab- olism of normal individuals from childhood to old age. This research was carried out with all the precautions exacted by the Zuntz school as to quietness and accuracy of technique. The mouthpiece and the Zuntz-Geppert apparatus in general were used for the study of the basal metabolism of 11 boys and 9 girls, all 14 years old or younger. TABLE 6. Basal heat production of boys and girls (in lying position) (Magnus-Levy and Falk) Age, years. Body- weight (without cloth- ing). Height. Body- surface (height- weight chart). Heat (computed) Heat per 24 hrs. pre- dicted by multiple prediction (adult) formula. 1 Com- puted less pre- dicted. 1 Per- cent- age differ- ence. 1 Per 24 hours. Per kilo- gram per 24 hours. Per square meter per hour. Boys. 21 6 6 7 7 9 10 11 14 14 14 Girls. 6| 11 11 12 '12 12 13 14 kilos. 11.5 14.5 18.4 19.2 20.8 21.8 30.6 26.5 36.1 36.8 43.0 18.2 15.3 35.0 42.0 24.0 25.2 40.2 31.0 35.5 cm. 'iio 110 112 110 115 131 129 142 142 149 107 141 149 129 128 145? 138 143 sq. m. cols. 782 926 970 1,067 1,153 1,036 1,338 1,151 1,310 1,285 1,525 936 866 1,313 1,459 962 938 1,362 1,217 1,299 caZs. 68.0 63.9 52.7 55.6 55.4 47.5 43.7 43.4 36.3 34.9 35.5 51.4 56.6 37.5 34.7 40.1 37.2 33.9 39.3 36.6 cals. caZs. 776 829 844 856 881 1,075 1,002 1,179 1,188 1,309 cola. ' +150 +141 +223 +297 +155 +263 + 149 + 131 + 97 +216 +19.3 +17.0 +26.4 +34.7 + 17.6 +24.6 + 14.9 +11.1 + 8.2 +16.5 0.79 0.83 1.05 0.98 1.20 1.21 1.34 60.8 52.0 53.1 48.9 45.5 44.3 47.4 46.8 46.0 42.6 41.1 44.7 46.1 45.5 967 1,199 1,281 1,067 1,077 -101 + 114 + 178 -105 -139 -10.4 + 9.5 +13.9 - 9.8 -12.9 1.17 1.32 0.94 0.95 1.27 1.10 1.19 1,146 1,194 + 71 +105 + 6.2 + 8.8 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 238. At first thought it would seem difficult to make metabolism experi- ments with children, especially with a child as young as 1\ years, with an apparatus using a mouthpiece. Nevertheless the results show reasonably close agreement with each other. When it is re- membered that these investigations were carried out fully 20 years ago, it will be recognized that they represent a remarkable advance in 1 Magnus-Levy and Falk, Archiv f. Anat. u. Physiol., 1899, Suppbd., p. 314. 14 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. the knowledge of the basal requirements of young children over that shown in the earlier studies. Singularly enough, they represent nearly the only extensive investigation in this field for a period of over 20 years. The results obtained by Magnus-Levy and Falk are prac- tically the first data in the earlier literature which are available for comparison with modern work, although the latter requires even more stringent experimental conditions, particularly as to muscular repose and absence of food in the stomach. A recent biometric analysis 1 of the older subjects of Magnus-Levy and Falk shows that, on the whole, these subjects apparently had a somewhat higher metabolism than that of American subjects. In table 6 we have abstracted from the extensive material of Magnus-Levy and Falk the data dealing with children 14 years or younger, i. e., the ages covered by our own observations. On the assumption that the age relationship is linear, Harris and Benedict 1 point out that the boys, particularly the young boys, are characterized by measurably higher metabolism than would be com- puted from the multiple-prediction methods. These differences are by no means so pronounced with the girls. Further discussion of this point is made on page 197. Von Wilkbrand, 1907. Von Willebrand 2 made a number of experi- ments with boys 9 to 14 years old. As his results are published in a somewhat inaccessible journal, they are given in some detail in table 7 herewith. The apparatus used were the respiration chamber in Helsingfors and the Pettersson-Sonden gas-analysis apparatus. The body-surface area was calculated by the Meeh formula, the constant 12.165 being used for the two youngest boys and 12.847 for the two older. The experiments were usually 24 hours long and began in the morning. All three meals of the day were taken in the apparatus; the subject went to bed in the evening about 8 or 9 o'clock and rose in the morning about 6 o'clock. In some instances the subject slept for a short time during the day. In discussing his results, the author does not use the minimum values as reported in our table, but takes the average for the whole 24-hour period. He first discusses the normality of his children on the basis of weight, age, etc., but gives no heights for his subjects. As average weights, he uses 26 kg. for Veikko, 30.5 kg. for Viktor, 34.3 kg. for Julius, and 35.7 kg. for Silo. The average carbon-dioxide production per hour of these subjects is given as 17.9, 21.7, 20.2, and 18.6 grams, respectively. A comparison between the carbon-dioxide production of the subjects per hour while awake with that during sleep is given hi table 8. Von Willebrand reports that Sond&i and Tigerstedt found 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 233. * von Willebrand, Finska Lakaresallskapets Handlingar, 1907, 49, p. 417. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 15 TABLE 7. Metabolism of boys 9 to 14 years old (von WUlebrand). Subject, age, and body-weight. Carbon exhaled per 2 hours. Time of minimum carbon exhaled. Carbon-diox- ide production per 2 hours. 1 Max.' Min. Max. Min. grams. grams. grams. grams. ( 5.37 9 h 40 m p.m. to Il h 40 m p.m. |" ( 19.7 Veikko, 9 years: 25.9 kilos 14.96 1 3.25 \ 6.41 11 40 p.m. 1 40 a.m. 1 40 a.m. 3 40 a.m. I 54.9 11.9 1 23.5 [ 7.86 3 40 a.m. 5 40 a.m. J ( 28.8 f 6.25 7 30 p.m. 9 30 p.m. } f 22.9 Veikko, 9 years: 26.0 kilos 11.34 \ 2.75 9 30 p.m. 11 30 p.m. 41.6 J 10.1 1 6.26 11 30 p.m. 1 30 a.m. J 1 23.0 {6.05 9 35 p.m. 11 35 p.m. f 22.2 Veikko 9 years* 25 9 kilos 18.35 9.12 9.37 11 35 p.m. 1 35 a.m. 1 35 a.m. 3 35 a.m. 1 67.3 I 33.4 1 34.4 9.47 3 35 a.m. 5 35 a.m. J I 34.7 f 5.80 9 30 a.m. 11 30 a.m. f 21.3 Veikko, 9 yeara: 26.0 kilos 15.13 ! 5.04 1 6.32 3 30 p.m. 5 30 p.m. 9 30 p.m. 11 30 p.m. > 55.5 1 1 18.5 | 23.2 I 7.19 11 30 p.m. 1 30 a.m. J I 26.4 Viktor, 10 years: 30.0 kilos 16.18 f 9.34 1 9.20 6 15 p.m. 8 15 p.m. 12 15 a.m. 2 15 a.m. | 59.3 1 34.2 1 33.7 {9.71 11 30 a.m. 1 30 p.m. | (35.6 Viktor, 10 years: 30.3 kilos 16.04 4.13 9.48 5 30 p.m. 7 30 p.m. 11 30 p.m. 1 30 a.m. \ 58.8 15.1 34.8 9.68 1 30 a.m. 3 30 a.m. J 35.5 Viktor 10 years' 30 8 kilos 17.12 1 9.74 \ 9.40 7 30 p.m. 9 30 p.m. 9 30 p.m. 11 30 p.m. } 62.8 35.7 34.5 f 4.64 5 30 p.m. 7 30 p.m. f 17.0 Viktor, 10 years: 30.8 kilos 15.38 \ 7.25 11 30 p.m. 1 30 a.m. \ 56.41 \ 26.6 I 6.35 1 30 a.m. 3 30 a.m. j I 23.3 Julius, 13 years: 34.1 kilos 15.28 I 4.12 \ 5.14 1 30 a.m. 3 30 a.m. 3 30 a.m. 5 30 a.m. 56 - i lli 1 18.0 Julius 13 years 34 4 kilos 16.51 j 9.16 1 9.36 9 34 a.m. 11 30 a.m. 5 30 a.m. 7 30 a.m. J60.5 I 3.63 I 34.3 Julius, 13 years: 34.4 kilos 17.60 1 5.85 1 6.60 11 30 a.m. 1 30 p.m. 1 30 p.m. 3 30 p.m. 64.5 21.5 24.2 ' 8.99 9 30 p.m. 11 30 p.m. 33.0 9.58 11 30 p.m. 1 30 a.m. 35.1 Silo 14 years* 35.6 kilos 16.18 5.98 1 30 a.m. 3 30 a.m. 59.3 21.9 6.22 3 30 a.m. 5 30 a.m. 22.8 10.60 5 30 a.m. 7 30 a.m. 38.9 7.07 1 00 p.m. 3 00 p.m. ' 25.9 9.81 7 00 p.m. 9 00 p.m. 36.0 Silo 14 years- 35 1 kilos 13.49 7.39 9 00 p.m. 11 00 p.m. 49.5 27.1 7.98 11 00 p.m. 1 00 a.m. 29.3 8.14 1 00 a.m. 3 00 a.m. 29.8 f 5.77 7 00 p.m. 9 00 p.m. 21.2 Silo, 14 years: 36.5 kilos 13.44 \ 5.15 11 00 p.m. 1 00 a.m. 49.3 18.9 [ 5.45 1 00 a.m. 3 00 a.m. 20.0 1 In 9 cases the maximum carbon exhaled appeared in the morning hours; in 4 cases in the afternoon; in 1 case in the evening. J Computed from carbon exhaled, using the factor 11/3. the relation for the carbon-dioxide production between these two conditions to be 145 to 100. With regard to table 8, which is taken directly from von Willebrand's 1 table, it should be pointed out that von Willebrand, loc. cit., p. 462. 16 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. in computing the values for asleep the author did not select absolute minimum values, but rather took an average representing the entire time the subject was in bed. Thus he disregards completely the wholly impossible variations in the carbon-dioxide production from period to period. TABLE 8. Comparison of carbon-dioxide production of boys awake and asleep (von Willebrand). Subject. Age. Body- weight. Carbon-dioxide production per hour. Relation of awake to asleep. Awake. Asleep. Veikko years. 9 10 13 14 kilos. 25.9 30.8 34.1 36.5 grams. 24.8 23.6 25.8 21.1 grams. 11.7 15.4 13.5 11.1 212 : 100 154 : 100 191 : 100 190 : 100 Viktor Julius Silo Average 23.8 12.9 186 : 100 This whole research is very perplexing. The differences in the carbon dioxide found in the 2-hour periods in different experiments with the same individual, presumably when he is asleep, are extra- ordinary. Such differences are wholly outside of our experience in the Nutrition Laboratory or at Wesley an University. It appears as though there must be a gross experimental error, and yet Tigerstedt proves that the carbon-dioxide determinations in this apparatus ought to be accurate to within 0.76 gram carbon dioxide. 1 If an attempt were made to use the absolute minimum figures, one would find values for the carbon-dioxide production per 2 hours on the experimental days with Veikko of 11.9, 10.1, 22.2, and 18.5 grams; with Viktor, 33.8, 15.2, 34.5, and 17.0 grams; with Julius, 15.1, 33.6, and 21.5 grams; and with Silo, 21.9, 25.9, and 18.9 grams. These low values were not confined to any one part of the day, but varied widely as to the time they appeared. Niemann, 1911. Laying special emphasis upon the 24-hour metab- olism of a bottle-fed baby, and employing the small Pettenkofer- Voit chamber, Niemann, 2 in Heubner's clinic, studied a male child designated as "normal," but with a body-weight somewhat low for the age. The child was studied at the age of 3| months for 7 days, again at 5 months for 6 days, at 8 months for 6 days, and at 9 months for 17 days. Since the metabolism measurements were made for the entire 24 hours, and no subdivision can be made for the periods when asleep, values for basal metabolism are not obtainable. As an index of the total 24-hour caloric output of children, however, this experi- ment, along with others of the Heubner clinic and from the Kaiserin 1 Tigerstedt, Skand. Arch. f. Physiol., 1906, 18, p. 304. 2 Niemann, Jahrb. f. Kinderheilk., 1911, 74, pp. 22, 237, and 650. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 17 Auguste Victoria-Hails, are of special value, reported in table 9. Niemann's data are TABLE 9. Twenty-four hour heat production of male infant studied by Niemann. Age. Body- weight. Heat (computed) per 24 hours. Remarks. Per kilo. Per sq. m. mos. ^ 5 8 9 kilos. 5.12 5.90 5.57 5.98 cols. cats. 93.0 1,556 85.4 1,500 88.6 1,526 91.2 1,608 Computed by us, using the Lissauer constant for surface, which we subse- quently show is much more accurate than that of Meeh (see p. 60). Frank and Niemann, 1913. A very much undernourished male child, 3 months old, was studied by Frank and Niemann 1 in a respira- tion chamber after being fed for 6 weeks on breast milk and again after 4 weeks' feeding with cow's milk. The results have no special interest in a study of the normal respiratory exchange of normal children. Murlin and Hoobler, 1915. Employing an apparatus essentially that formerly described by us, Murlin and Hoobler 2 studied some hospital cases, of which several are considered by them as perfectly normal. 3 The values computed for the minimum heat production per square meter per 24 hours, using the more accurate Lissauer method of computing, are given in table 10. From these cases and 14 previ- TABLE 10. Minimum heat production of children per 24 hours (Murlin and Hoobler). Subject. Sex. Age. Body-weight (without clothing). Minimum heat production (computed) per 24 hours. Per kilo. Per sq. m. (Lissauer) . A. S W. L E. H E. N M. M C. M W. S M. M. M. M. M. F. M. mos. 2 2 2 3 51 loj 11-12 kilos. 5.7 4.4 4.7 4.1 6.5 9.4 9.4 cals. 50.0 58.2 63.0 61.5 61.4 61.8 60.5 colt, 863 936 1,035 960 1,115 1,266 1,239 ously published by us, they discuss extensively the metabolism in relation to several physiological factors. 1 Frank and Niemann, Charite-Annalen, 1913, 37, p. 94. 2 Murlin and Hoobler, Am. Journ. Diseases of Children, 1915, 9, p. 81. 3 One of these subjects, E. N., we would exclude from consideration because of underweight; likewise, we would question considering M. M., who, according to the authors, "no doubt had an incipient tuberculous infection." 18 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. Olin, 1915. In a study made by Olin 1 under Professor Robert Tigerstedt and published in 1915, the carbon production of boys from 9 to 19 years old was determined. The children were taken from an elementary school in Helsingfors and studied individually in 2-hour experiments, usually in the morning after a light breakfast. The subjects were not allowed to move about in the respiration chamber, but were required to sit quiet, reading. Although approximately 200 children were studied, Olin's final discussion of the results is based upon the data obtained with 162 subjects. These results, grouped according to age, are given in table 11, which shows not only the total carbon production per individual, but also the carbon production per kilogram of body-weight and per square meter of body-surface. The range in values for the carbon production on the basis of body-surface is also included in table 11, and also the percentage distribution according to age. Meeh's formula was employed in computing the body-surface, the constant 12.205 being used for boys under 13 years and 12.847 for those at and above that age. It will be seen that the TABLE 11. Carbon production of boys from 9 to 19 years of age (Olin). Carbon exhaled per 2 hours. Average Per square meter of body-surface. No. of sub- jects. Aver- age age. body- weight (without cloth- Body- surface (Meeh). Height. Per indi- vidual. Per kilo of body- Aver- Range. Percentage distribution. Less Great- ing). weight age. than er than 8.5 9.5 grams. grams. yrs. kilos. sq. TO. cm. grams. grams. grams. grams. p.ct. p.ct. 1 19 80.0 2.385 177 18.8 0.24 7.90 7.90 .'...., 7 18 65.6 2.086 176 17.6 0.27 8.48 8.31 to 8.65 1 57 9 17 55.4 1.864 170 16.2 0.29 8.66 8.52 8.80 33 22 18 16 59.2 1.948 171 17.6 0.30 9.05 8.89 9.21 50 33 19 15 52.9 1.805 165 16.8 0.31 9.13 8.97 9.29 21 42 22 14 49.6 1.726 160 15.8 0.32 9.13 8.97 9.29 41 36 26 13 43.1 1.573 154 14.4 0.34 9.22 9.08 9.36 20 46 27 12 38.1 1.396 148 13.4 0.36 9.68 9.56 9.80 15 51 14 11 36.1 1.327 143 13.4 0.37 10.14 9.92 10.36 7 71 15 10 31.5 1.217 140 11.8 0.38 9.79 9.63 9.95 40 4 9 35.9 1.299 139 11.6 0.32 8.60 carbon production per square meter of body-surface was much larger at 11 years of age than at any other age studied, and increased regularly as the age decreased from 18 to 11 years. In comparing her data with those of other investigators, Olin states that, as the experiments of Magnus-Levy and Falk were made with complete muscular rest and without food, these subjects were more quiet than hers, who were in 1 Olin, Finska Lakaresallskapets Handlingar, 1915, 57, p. 1434; also Skand. Archiv. f. Physiol., 1915, 34, p. 414. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 19 the sitting position rather than lying. She finds, also, that her values are much lower than those of Sonden and Tigerstedt. She concludes that the earlier statements that the metabolism is greater per square meter of body-surface with young individuals than with those of full growth are completely proved by her results, which show that the metabolism of boys .between 10 and 18 years of age is distinctly greater per square meter of body-surface than that of full-grown adults. The metabolism is, therefore, not dependent solely upon the body-surface, but also upon the age and growth. Helksen, 1915. Laying special emphasis upon the character of the diet, particularly the isodynamic relations between carbohydrates and fats, Hellesen, 1 working in the Kaiserin Auguste Victoria-Haus in Berlin, made two 3-day experiments 9 days apart, hi March 1911. As is usual with experiments made in that laboratory, the child was studied for 22 hours out of 24 hours and carbon dioxide and water- vapor alone were measured. As is customary with this type of experi- ment, the carbon-dioxide production for special periods of quiet was not observed. Hence no basal values can be given, and the data are of special interest only in indicating the total 24-hour heat output of a child of this age. From the average of the two experiments, each of three days, it was found that in the first period the child, a male, weighing 6,684 grams, had a heat production per 24 hours of 454.6 calories, per kilogram per 24 hours of 68.0 calories, and per square meter of body-surface of 1,245 calories. 2 In the second period the weight was 6,644 grams, the heat per 24 hours 492.4 calories, per kilogram per 24 hours 74.1 calories, and per square meter per 24 hours 1,353 calories. 2 Du Bois, 1916. Subsequent to the classical study of Magnus-Levy and Falk and the data obtained in the Helsingfors laboratory, the next extensive investigation of the basal metabolism of young boys was reported by Du Bois. 3 Using the respiration calorimeter, he studied the carbon-dioxide output, oxygen intake, and direct cal- orimetry of eight boys from 12 years 2 months to 13 years 11 months. For the purpose of obtaining further data regarding the relationship between body-surface and metabolism, the surface areas of the sub- jects were measured by the Du Bois linear formula. So far as the number of subjects is concerned, the experimental plan was very satisfactory, but only one experiment was made with each subject, this usually consisting of two successive 1-hour periods. Du Bois's own figures seem to indicate the futility of this method of experimenta- tion, for they show the following variations in the heat-production per square meter per hour for six of the subjects studied: J.D.D.B., 1 Hellesen, Nord. Med. Arkiv, 1915, 48, Nos. 14 and 18. 2 Computed by us, using the Lissauer formula. 3 Du Bois, Arch. Internal Med., 1916, 17, p. 887. 20 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. 12 per cent; Reg. F., 15 per cent; Harry B., 4 per cent; Raymond M., 7 per cent; Henry K., 13 per cent; and Leslie B., 25 per cent. Thus out of the eight boys whose values were used for basal averages, four show a difference of 12 or more per cent between the results obtained in the two periods. When there is a difference of but 3 or 4 per cent, an average may legitimately be used, but with variations so large as these, it is clear that one of the two periods may approach the basal metabolism and the other must be above it. Under these conditions the averaging of the values is questionable and tends to give a high value for the basal metabolism. Du Bois concludes that these boys show on the average a heat production per unit of surface area accord- ing to the Du Bois linear formula which is 25 per cent higher than the level for adults. In analyzing the figures, however, the wide variations between the results of the two periods should be borne in mind, for values differing 12 per cent or more can hardly give a satisfactory base-line. If the lower of the two values is selected to represent the base-line, this would reduce very perceptibly the average increase above the adult level. It may be questioned, therefore, whether even the lowest value represents the basal metabolism, and a second experiment on each boy to control these important findings would have been of inestimable value. Olmstead, Barr, and Du Bois, 1918. Recognizing the importance of studying the metabolism of young boys at about the age of puberty, Du Bois and his associates 1 repeated the series of experiments which were made with boys in 1916 to study the changes in the metabolism after two years of growth. In this series the wide variations between the two periods were eliminated, the only two subjects showing differ- ences of 12 per cent or more being J.D.D.B. and Reg. F. The former was not included in the calculation of the basal metabolism; the latter showed a variation of 17 per cent. Since the experimental conditions apparently approached the normal, it is probably legitimate in all cases except Reg. F. to average the results of the two periods, although it is still open to question whether duplicate experiments on separate days should not have been made. The authors note a pronounced decrease in metabolism in this second series of experiments as compared with the results of the first series, the boys when two years older showing a decrease in the metabolism of 13 per cent. If the minimum values obtained in the first series of experiments were used for the basal metabolism instead of the average of the two periods, this decrease would obviously be less. All the evidence seems to imply that the conditions under which the first experiments were made were abnormal in that the subjects had a much more irregular metabolism than would normally be expected. 1 Olmstead, Barr and Du Bois, Arch. Internal Med., 1918, 21, p. 621. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 21 The pulse-rates in all instances but one showed a pronounced decrease in the second series of experiments, L. B., for example, having a pulse- rate of 88 in the first experiment and when two years older a pulse-rate of 65. It seems almost impossible to ascribe this decrease in metab- olism to a difference in age of two years, and this, combined with the gross irregularities in metabolism in the two 1-hour periods in the first experiment, and pronounced alterations in the pulse-rates, makes it all the more desirable to have researches of this type include a sufficient number of observations with each subject to establish definitely the basal minimum metabolism for each individual. It is clear that the average values used by Du Bois for the first series of experiments are certainly not basal values. From the results of the second series of experiments, in which the agreement between the two periods is much more satisfactory, the authors consider that the metab- olism per hour according to the Du Bois linear formula of the boys is still higher than that of the average adult, the difference being about 11 per cent for boys 14 to 15 years old. CONTROL EXPERIMENTS AND BASAL METABOLISM. As will be seen from the foregoing abstracts of previous investi- gations, the influence of both muscular activity and food was dis- regarded in the earlier experimental work on metabolism; conse- quently the values obtained in these researches were usually not strictly basal. Most of this material was collected before the impor- tance of absolute muscular repose was sufficiently recognized by experimenters. Furthermore, ocular observations of the degree of repose were used in the earlier studies, and these are untrustworthy, since interpretations vary widely and accurate records are obtained only by the graphic method. While these faults in technique have, in later days, been corrected hi large part, greater uniformity in experimental conditions and standards should be the rule in all laboratories studying metabolism. As emphasized in a previous section (see page 2), a knowledge of the basal metabolism is very necessary, since it represents the basic requirements of the human body to which the requirements for all activities are added. According to modern conceptions of ideal studies of the basal metabolism, which require the subject to be in the post-absorptive condition (12 hours without food) and in com- plete muscular repose, the basal metabolism is a perfectly definite factor. Even if the values obtained are complicated by food, as in some of the earlier experiments in this study of the metabolism of children, yet the error can not be more than a maximum of 20 per cent for short periods and rarely as high as this. The criticism has been raised that the conditions outlined for obtaining the basal metabolism are not normal conditions, but this 22 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. is not true, since all humans lie quietly in bed some 8 or 10 hours each day. On the other hand, life in a respiration chamber is not normal, for the experimental periods of Sonde'n and Tigerstedt, and especially the 22-hour periods of Rubner, do not represent normal conditions in a child's daily existence, when he is living a more or less active and unrestricted life. This may be illustrated by comparing the daily requirement of Rubner's boys, i. e., about 1,700 calories, with the food intake of the boys in Gephart's study 1 at St. Paul's School in Concord, New Hampshire, which approximated 5,000 calories. An attempt to study the requirements of the day as a whole by a 22 to 24 hour sojourn in a respiration chamber would therefore be a great mistake and a return to the standards of 25 years ago. The only logical method for determining the child's energy require- ments for a day's existence is to obtain, first, the requirement for maintenance (the basal metabolism); second, the additional energy required for sitting in a chair reading or studying; third, the energy requirement for walking; and fourth, that for running and playing or at severe work. With adults we already have much information as to the energy requirements for the first three, and may later be able to determine the enormously variable factors for the fourth requirement. In criticizing the experiments of Sonde'n and Tigerstedt and others using a large respiration chamber, one should bear in mind the fact that they were not planned to study the so-called basal metabolism, for at that time the conception of basal metabolism was but imper- fectly outlined in the minds of the investigators. Furthermore, Sonde'n and Tigerstedt definitely claim that their results were not minimum or basal. Except when the data were determined during deep sleep, the basal metabolism for these subjects was not deter- mined, for young children as well as adults were studied under con- ditions of only comparative and not complete muscular repose and during the process of digestion. For a study of the general question as to whether or not there is an alteration in the metabolism per unit of weight or per unit of surface with individuals of different ages, it may be that the experimental procedure used by them would be justifiable, namely, a study of all subjects under presumably like (though certainly not basal) conditions. Similarly, Rubner's experiments were made to determine the total metabolism during the 24-hour period and thus could not be used for basal comparison. In Rubner's observations there was a con- siderable amount of sleep, which tended to compensate in part, at least, for the excess activity during the waking hours. Had Rubner's results been so published as to permit the computation of the values 1 Gephart, Boston Med. and Surg. Journ., 1917' 176, p. 17. PREVIOUS STUDIES OF THE METABOLISM OF CHILDREN. 23 obtained with the subject during sleep, his experiments would have contributed towards our knowledge of the basal metabolism. Magnus- Levy and Talk's results approach more nearly the modern idea of basal metabolism, for these experimenters insisted upon repose and upon absence of food. While it is assumed by Sonde"n and Tigerstedt and by other workers that the same degree of muscular activity can be approximated with children of different ages when complete muscular repose is not insisted upon, this still remains to be proved and may fairly be ques- tioned at this tune. It is quite possible that the relatively great differences in the activity of children, youths, and adults may account for the peculiar findings recorded by these several schools when an attempt is made to superimpose a definite amount of activity upon the modern standard of complete repose. From a purely scientific standpoint, clear-cut conditions are essential for comparative experiments, and if one wishes to study the differences hi metabolism due to differences in age or sex, and particularly for the various diseases, we must have uniform conditions. The basal condition presents especially good opportunities for such uniformity. Studies of this kind are particularly advantageous in clinical calor- imetry, for most of the subjects observed would be bed-ridden or hospital patients, spending a considerable part of their time under conditions of muscular repose. The experiments of Sonden and Tigerstedt served their special immediate purpose perfectly, namely, to supply knowledge regarding the carbon-dioxide production in school rooms and halls of a group of young children. Rubner's experiments on two boys served his special purpose, but neither series can be considered as more than studies of metabolism under special conditions. This likewise applies to the major part of the respiration calorimeter experiments made at Wesleyan University. Each series demonstrated its special highly important and fundamental point, but few of the experiments are useful for further comparison. The night experiments at Wesleyan University began at 1 a. m. and were ended at 7 a. m., while the subjects were still in bed, so that results were obtained almost invariably for six hours of repose. They are therefore of more general use, since the subject was asleep and quiet, with the influence of food at a minimum. It is the duty of twentieth-century experimenters to make experi- ments of more than passing value. Each experiment should contribute to our fundamental basal knowledge. Each year sees an increase in the significance of normal values; an effort should therefore be made to secure normal values which will be of service not simply in the current year but for a decade. There is no excuse for present-day controls which do not meet modern requirements. Hundreds of experiments with men and women in this laboratory have shown that 24 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. the post-absorptive condition and muscular repose (with graphic registration by means of a pneumograph and kymograph) are readily obtained. Conditions for normal pulse-records are thus ideal and should accompany each experiment. In other words, normal experi- ments should serve not only their immediate purpose, but should invariably contribute toward the sum of the knowledge of basal metabolism requirements of normal subjects. We wish to enter a plea for the same degree of intelligence in planning normal control experiments, as is shown in planning the original critical experiments in every series of metabolism experiments conducted in the future. If all workers in metabolism applied this critical analysis to their own work, the accumulation of basal material would be very rapid, and the work of the several laboratories would be much more strictly comparable than at present. While it is now practically agreed by all laboratories that ideal conditions are the post-absorptive state and complete muscular repose, the question of sleep is at present debatable ground. It is impossible to insure deep sleep, but it is possible as a rule to insure wakeful repose. HISTORY AND PLAN OF RESEARCH. Aside from several preliminary observations at the Nutrition Laboratory, in which only the carbon-dioxide output of infants was determined, the first studies made in this series on the gaseous meta- bolism of young children were carried out at the Massachusetts General Hospital. A respiration laboratory was established in that institution in January 1913, and observations by a member of the Nutrition Laboratory staff were made almost daily, except during the summer months, until June 1915. The infants first used were mostly from the Out-Patient Department, but it soon became evident that data regarding the normal metabolism of young children could not be obtained with these infants, for normal, healthy children are not to be found in a hospital. These earlier observations therefore included a considerable number of underweight children and a few abnormal cases. The results of the studies, which comprised data for children under two years of age, have been reported in detail in a monograph and elsewhere. 1 The later observations at the Massachusetts General Hospital were made with new-born infants from the Boston Lying-in Hospital, their age varying from 43 minutes to 8 days. The data for 105 new-born infants have been reported in a monograph and in several journal articles. 2 Subsequently the apparatus at the Massachusetts General Hospital was removed to the Directory for Wet-Nurses of the Boston Infants' Hospital, and studies of children, varying in age from two weeks to two years, were carried out. Conditions here were especially favor- able for the collection of normal data, as the children, mostly breast- fed, were the offspring of resident normal wet-nurses, and thus repre- sented an unusually good type of physical normality. Furthermore, the mothers were somewhat under control. As the inmates of this institution constitute a more or less floating population, a large number of babies were available for observation. It was also possible to follow the life-history of a number of the infants and make observations of their metabolism from time to time over a period of several years. Some of the children in these prolonged studies had been previously observed in the study of new-born infants. To obtain information as to the 24-hour energy requirement of young children, two 24-hour 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. See, also, Benedict and Talbot, Am. Journ. Diseases of Children, 1912, 4, p. 129; and 1914, 8, p. 1. 2 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915. See, also, Benedict and Talbot, Proc. Nat. Acad. Sci., 1915, 1, p. 600, and Talbot, Am. Journ. Diseases of Children, 1917, 13, p. 495. 25 26 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. experiments were carried out. The results of these two experiments have already been briefly reported. 1 To complete the accumulation of data on the normal metabolism of children from birth to puberty, it was necessary to obtain older children for further studies. Arrangements were therefore made, through the courtesy of Dr. Frederic H. Knight, the superintendent of the New England Home for Little Wanderers, to establish a respira- tion laboratory in that institution. Studies were begun in October 1917 and continued until July 1919. The observations included children of both sexes from 2 to 15 years of age. The obtaining of successful experimental periods was very much a question of the child's mental attitude. To secure the interested co- operation of the children a system of prizes was established. Each child volunteering as a subject for an experiment received a cake of sweet chocolate and a small sum of money, varying from 5 cents to 25 cents, according to the number of experimental periods possible to be carried out and the degree of muscular repose during the experiment. In the school-room it was made a privilege and an honor for a child to act as subject in these respiration experiments; the child reported at the end of an experiment to his teacher and schoolmates what prize he had won, and in this way interest and enthusiasm were maintained without difficulty. To obtain information regarding the actual weight and height of children of school age, for possible use hi establishing a normal stand- ard, supplementary data were collected regarding the weight and height of the pupils of a considerable number of private schools. The results of the observations made at the Directory for Wet- Nurses and the New England Home for Little Wanderers have not yet been reported in detail, 2 and form the subject of this monograph. This report and the two monographs previously referred to give a complete summary of the results obtained in the whole research on the metabolism of children from birth to puberty. It will be seen from the foregoing history of the research on the basal metabolism of children carried out by the Nutrition Laboratory during almost a decade that the general plan of study was to make observations on the respiratory exchange of a large number of normal children differing in sex, age, height, and weight. As the observa- tions were made with the children under conditions of muscular repose and in many instances with no food in the stomach, the basal or minimum metabolism was secured. By a comparison of the average values obtained, the influence of age, height, weight, and sex upon 1 Talbot, Am. Journ. Diseases of Children, 1917, 14, p. 25. 2 A preliminary report of the results was given by one of us in the Shattuck Lecture before the Massachusetts Medical Society, June 1919. See Benedict, Boston Med. and Surg. Journ., 1919, 181, p. 107; also Talbot, Am. Journ. Diseases of Children, 1919, 18, p. 229. HISTORY AND PLAN OF RESEARCH. 27 the heat production could be found, normal standards derived, and the basal energy requirements determined. A sufficiently large num- ber of children were studied to obtain reasonably complete data for all ages between birth and puberty. In addition to this general method of research, the problem of the energy requirements during the period of rapid growth in the first years of life was also studied by making observations on the same individual at intervals over a num- ber of years. In all, 23 children were thus studied for periods of a few months to three or four years. The data gathered in this research thus provide information not only as to the energy requirements of a large number of children, but on the energy requirements of the same child at varying ages. In our earlier publications 1 we have already acknowledged the cour- tesy extended to us by the Trustees of the Massachusetts General Hos- pital and the Boston Lying-in Hospital. We wish also to acknowledge at this point the courtesy of the Trustees of the Directory for Wet- Nurses and the Trustees of the New England Home for Little Wan- derers in allowing us to carry on these investigations hi their insti- tutions. Especially do we appreciate the hearty cooperation of the superintendent of the latter institution, Dr. Frederic H. Knight. Many of the observations were made possible through the individual interest of Miss Mary A. Slade, a teacher at the New England Home for Little Wanderers, who was very successful in arousing the interest of the children in the work. Finally, too much credit can not be given to the technical skill, devotion, and faithfulness of Miss Alice Johnson, Mrs. Dorothy A. Peabody, and Miss Inza A. Boles. i Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914; Carnegie Inst. Wash. Pub. No. 233, 1915. APPARATUS AND EXPERIMENTAL TECHNIQUE. The observations with young children ranging in age from a few minutes after birth to puberty included records of the height, body- weight (nude), pulse-rate, body temperature (rectal), muscular activity or repose, and measurements of the gaseous metabolism. From the latter, computations were made of the respiratory quotient. A general record was also kept of the condition of the child during the observations, including such data as drowsiness, sleep, crying, etc. Measurements were likewise made of the body-surface of a consider- able number of the children. RESPIRATION APPARATUS. The apparatus used for measuring the gaseous metabolism in the studies at the Massachusetts General Hospital has already been fully described in previous publications. 1 In this apparatus a small metal chamber, with a quickly removable cover made air-tight by a water- seal, is connected with a closed-circuit ventilating apparatus. A rotary pump in the ventilating circuit draws the air from the chamber, and forces it through absorbing vessels in which are absorbed the water and carbon dioxide given off by the subject. The air is then returned to the chamber after pure oxygen is introduced from a cylinder of the compressed gas to replace that used by the child. A spirometer in the ventilating circuit provides for expansion in the volume of air. The amount of carbon dioxide expired is measured by weighing the absorbing vessels; the amount of oxygen used is determined by metering the gas introduced into the ventilating circuit. Inside the metal chamber is a wire crib, with mattress, in which the child lies during the period of observation. Since only the basal or minimum metabolism was desired in these studies, it was necessary to have a record of the degree of muscular repose or activity of the child to be assured of comparable conditions. While ocular observations of the general condition and activities of the child can be made through the glass window in the cover, such obser- vations give no exact records to be used as a standard in comparisons of the metabolism measurements. The only method of obtaining reliable evidence of the presence or absence of muscular movement is by means of graphic records. Provision is made for such records in this apparatus by suspending one end of the wire crib by a spiral spring, with a small pneumograph parallel to it. By this means the 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 32; Am. Journ. Diseases of Children, 1914, 8, p. 21. 28 APPARATUS AND EXPERIMENTAL TECHNIQUE. 29 movements of the crib are transmitted through rubber tubing to a tambour and small pointer outside the chamber. The adjustment is so delicate that the slightest movement of the child changes the center of gravity and a record is made by the pointer on the smoked paper of a kymograph. Even the movements of the chest- wall due to res- piration are frequently indicated on the kymograph record. A straight line on the kymograph record thus shows that the child is absolutely quiet. The pulse-rate is obtained with a stethoscope connected with the observer outside the chamber by rubber tubing. Records were made at intervals by the attendant nurse and show that muscular activity instantly caused an increase in the pulse-rate. Accordingly, if periods with a low pulse-rate and no activity are selected for compari- son, we can be certain that the values represent the basal metabolism. This apparatus was employed at the Massachusetts General Hos- pital for about two years and was also used for the observations at the Directory for Wet-Nurses, the only change being that for the older children a larger chamber was constructed. The apparatus as used in actual work at the Directory for Wet-Nurses is shown in figure 1, with assistant in charge and nurse recording the pulse-rate. The end of the chamber may be seen at the right, with the tambour and kymo- graph for recording the muscular activity, and connections with the nurse who is counting the pulse-rate. The ventilating system is at the left, with the carbon-dioxide absorbers and spirometer on the upper shelf of the table, and the water-absorbers and blower on the lower shelf. The gas-meter for measuring the oxygen may be seen in the center. For the experiments at the New England Home for Little Wanderers a somewhat different form of chamber was used, i. e., a small clinical respiration chamber, described in detail by Benedict and Tompkins, 1 and illustrated in figure 2. This is substantially a duplicate of the apparatus used for studying the normal infants, with the exception that the chamber is much larger and hence of a somewhat different type of construction. In that used for the infants the cover shuts down like the lid of a trunk, while in the apparatus employed for the older children the cover is of a semi-cylindrical form and is suspended by two ropes connected with a counterpoise, thus providing for raising and lowering the cover. When the cover is lowered, it enters a narrow water-seal which makes it air-tight. A rectangular window in the cover provides for illumination. Figure 2 gives a general view of the respiration laboratory at the New England Home for Little Wanderers, with the respiration chamber at the right, and the venti- lating and absorbing system at the left ; the kymograph and tambour appear on a small table in the background. i Benedict and Tompkins, Boston Med. and Surg. Journ., 1916, 174, pp. 857, 898, and 939. 30 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. EXPERIMENTAL CONDITIONS. Usually the length of the observation was determined solely by the degree of repose of the subject. When the pulse-rate had fallen to a normal level and the kymograph record showed the child was abso- lutely quiet, the measurements of the metabolism were begun. If a normal pulse-rate and satisfactory kymograph record could not be obtained, the observations were discontinued for that day. The measurements of the metabolism were divided into periods of 20 or 30 minutes, the number of periods usually depending upon the con- ditions. Since to obtain ideal conditions for measuring the basal metabolism there must be no food in the stomach, the measurements were made in this way whenever possible. With the older children, voluntary muscular control and absence of food in the stomach could usually be secured. With young infants, however, the ideal conditions for studying basal metabolism could not be obtained, particularly as to the absence of food in the stomach. In the normal physiological state, the infant has more or less food in the stomach in process of digestion. The infant's natural protection against the lack of such food is restlessness, major activity, and crying. In consequence, when we attempted to secure the post-absorptive condition, we were almost invariably confronted with the fact that we no longer had a quiet infant, that is, one in muscular repose. Of the two factors affecting basal metabolism food and muscular activity the latter is so much the greater that our only alternative was to allow a minimum amount of food and thus secure muscular repose. This presence of food in the stomach during the observation con- taminates our comparable data, but was an experimental condition which, for the most part, it was impossible to eliminate with our youngest subjects. Schlossmann and Murschhauser 1 attempted to study the metabolism of an infant during prolonged hunger by giving water sweetened with saccharine and salted. The influence of the hunger was that commonly observed with adults, namely, a distinct increase hi acidosis. Experiments with adults have shown that acidosis stimulates metabolism, the acid unquestionably reacting upon the cells, stimulating them to greater activity. Accordingly, as set forth in an earlier publication, 2 it becomes a difficult matter to determine the exact points at which the post-absorptive condition begins and ends and hunger with stimulating acidosis begins. While we have in a number of our experiments been obliged to allow food in the stomach during the observation, on the other hand an examination of our data shows that in most of the observations with the younger children the subjects were asleep. In other words, at 1 Schlossmann and Murschhauser, Biochem. Zeitschr., 1913, 56, p. 355. 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 147. FIG. 1. Infant respiration apparatus as used at the Directory for Wet-Nurses. At the right are the respiration chamber, the tambour, and kymograph for recording the muscular activity, also nurse counting pulse-rate. At the left is the absorbing system, with the carbon- dioxide absorbers and spirometer on the upper shelf of the table, the water-absorbers and blower on the lower shelf. The gas-meter for measuring oxygen may be seen in center. FIG. 2. Respiration apparatus as used for children at the New England Home for Little Wanderers. The respiration chamber, with rounded cover and thermometers, is near the center of the picture, the absorber table at the left. On the upper shelf of the table may be seen the carbon- dioxide absorbers and the spirometer; on the lower shelf are the blower and the water- absorbers. Behind the table are the large oxygen cylinder and gas-meter; the tambour and kymograph stand on a small table in the rear. APPARATUS AND EXPERIMENTAL TECHNIQUE. 31 the age when it was most necessary to introduce into the metabolism measurements the stimulating factor of a small amount of food, we have the compensatory depressing influence of sleep. It is probable that these to a certain extent neutralize each other. At present, one may not exactly calculate the influence of sleep on the various children studied. Certainly, the ingestion of food on the one hand and the greater prevalence of sleep on the other tend towards a compensa- tion which makes the measurements of the metabolism of the younger children more nearly comparable with those made with the older subjects, who were in many instances awake. DISCUSSION OF RESULTS. The experimental evidence accumulated in connection with this research is especially complete with regard to certain physiological measurements, particularly pulse-rate and total metabolism, with incidental observations with regard to temperature. In the final analysis of these figures the writers regret that the question of time and expense made it impossible to record simultaneously with these measurements other important physiological factors, such as blood pressure, respiration-rate, alveolar air, and alveolar carbon dioxide, for the physiology of youth needs most exhaustive study and each series of observations increases greatly in value by being associated with other physiological measurements accurately determined. It is necessary, therefore, in this discussion to lay special stress upon the rather extensive series of anthropometric measurements and the measurements of pulse-rate, rectal temperature, and particularly basal metabolism. If the total energy metabolism is looked upon as the pooled energy for the day, involving all of the various vital processes which go to make up the physical and possibly intellectual activities of the day, it can be seen that its measurement, even if at the present time incapable of subtle analysis, is of great importance. Indeed, with the accumulation of data with regard to the total energy metab- olism of human individuals, it is becoming increasingly evident that this measurement is of great physiological if not indeed clinical im- portance. While the number of pathological factors, other than the febrile condition, that have thus far been noted to increase metabolism above normal are relatively few, it is highly probable that subsequent studies will show deviations from the average normal metabolism produced by disease or altered conditions of nutrition. The lowering of the normal basal metabolism results from a general state of undernutrition, and hence it is perfectly consistent to say that those subjects with lowered basal metabolism are distinctly below par. Where the health of the growing child is of as great significance as it is now apparently becoming, any index of vitality is valuable. Whatever restrictions or procedures for reducing metabolism may be justifiable with overfed adults, the consensus of opinion thus far points strongly towards the necessity for a surplus of food and a high nutritive plane for the best welfare of growing children. To throw some light upon the normal nutritive plane and to give an indication as to the general trend of the normal basal metabolism from birth through the period of early childhood are the main func- 32 NORMALITY OF CHILDREN STUDIED. 33 tions of our report. While this investigation was undertaken pri- marily with the idea of contributing to the abstract knowledge of the pure physiology of youth, it is gratifying to note that these results, as a natural consequence of the increased interest in the physiology of youth, have to-day a distinct practical value. In the discussion of our findings we shall consider, first, the normality of our children as shown by the relationships between body-weight, height, and age and by the anthropometric measurements indicating growth, particularly the body-surface area, with a view to establishing, if possible, the normal, average, and ideal states of nutrition. We will then consider the ideal physical proportions of children, and fixing these clearly in mind, will pass to a consideration of the physiological functions with special reference to pulse-rate, rectal temperature, and particularly the gaseous exchange and total energy metabolism. NORMALITY OF CHILDREN STUDIED. In all of our investigations thus far on the basal metabolism of humans of both sexes and of varying ages, one of the prerequisites has been that the subjects be individuals "presumably in good health." In this stage of our study, representing the metabolism of children from birth to puberty, it seems to us particularly necessary to scrutinize the normality of the children measured. Throughout the entire investigation we held firmly in mind the idea that we were primarily interested in physiological observations and should avoid, in so far as possible, any pathological or seemingly pathological cases. Conse- quently, there was more or less of a natural selection, which eliminated obviously pathological cases. In the course of our investigation, however, it so happened, due either to the scarcity of subjects at the time or to the special features of an individual case, that subjects not obviously pathological were studied, who on closer scrutiny would not be considered strictly normal. It was necessary to decide arbi- trarily, as we proceeded, what was and what was not physiological. No hard and fixed lines could necessarily be laid down; indeed, such lines do not exist. Ocular impressions of examining physicians, nurses, and attendants were relied upon to indicate whether or not the subject was normal. In all cases, however, there was a distinct mental reservation that before final publication the material would be thoroughly tested and only such as measured up to a reasonable estimate of normality would be included in the discussion on basal metabolism. The dominant thought in our consideration of the question of the normality of our children is to determine what degree of success attended our efforts to eliminate incipiently pathological material from our subjects and so enable us to state with accuracy that we are 34 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. dealing with the physiology of normal youth. In the first place, our own impressions of normality were substantiated by most careful physical examinations, in which always one and frequently two physi- cians passed critical judgment as to the normality of each child. Since this last analysis, however, must depend upon the personal judgment of the examiner, we have thought it would be highly desirable to check up the results of the physical examinations, so far as possible, by com- paring the cases passed upon as normal by the physicians with some recognized standards which would be completely free from the personal equation. To the uninitiated the personal impression made by a child is always of great weight in estimating the state of normality, and one instinctively says whether a child appears well-nourished or undernourished. But age, height, and weight, as well as girths, should be taken into consideration, and it is clear that personal im- pressions may very likely err in estimations of this kind, and therefore should be replaced, as far as possible, by accurately determined measurements. Furthermore, it is clear that sexual differences appear at an early age, so that boys and girls may not be classed together. At the completion of our experimental details, therefore, the selection of normal material presented a great deal of difficulty. We were instantly confronted with the question "What is normal?" While we firmly believe that all obviously abnormal children were excluded, it still remains an important point whether the general run of our children measure up to the idea of normality held by many writers. Heretofore the conception of normality has been in large part based upon the average, and consequently the averages selected from a large number of measurements of children of varying ages, weights, and heights have been used to indicate the average or normal growth for age, height, or weight of the child. This is not the place to define what is normal and what is average, but this point will be taken up in detail later on in this discussion. STANDARDS FOR DETERMINING THE NORMALITY OF CHILDREN. In considering the normality of children, special emphasis has always been laid upon the relationship between body-weight and age or between height and age, and only relatively recently between height and body-weight, irrespective of age. The age factor has always been the most important and the ratios of height to given age and body- weight to given age still remain the most commonly used criteria. At first sight this method of comparison would appear to be a relatively satisfactory one. The measurements of body-weight and of height are apparently very simple, and yet both are subject to considerable errors, particularly in the case of boys. Weight of clothing. For many obvious reasons, the weight of children above 5 years of age is almost invariably given with clothing. NORMALITY OF CHILDREN STUDIED. 35 Not infrequently the weights of children of all ages are given with clothing by various writers and an attempt is made (a very crude attempt) to correct for the clothing at the time of weighing. This method is inaccurate, since the weight of clothing varies with sex, with the seasons of the year, and with the social position of the child. Anyone can convince himself of the probable errors involved simply by emptying the pockets of a small boy. Scientifically accurate figures can therefore be obtained only when the subject is weighed naked. Bowditch, 1 Schmid-Monnard, 2 Vierordt, 3 Griffith, 4 and other in- vestigators have at various times reported the results of their studies of the weight of clothing worn by their subjects. In the measurements of our children we took the opportunity of studying to some extent the weight of clothing worn. Since, however, many of our children came from an institution where the clothing was more or less of uni- form style and amount, our estimates of weight of clothing are not of sufficient value to record, for they may have only a local significance, a fault in common with many other series. Another difficulty in using these two measurements of body-weight and height as standards for determining the normality of children is that usually they are taken on relatively few subjects, and for the establishment of an average or a normal value very considerable numbers of measurements of both height and body-weight at varying ages are essential. Height can and should be obtained on an indefinite number of subjects, weight also on a large number, although to be of value the weights should be nude weights. Unusual significance has been recently attached to the so-called "stem length," i. e., sitting height. 5 For greater accuracy, consideration should also be given to the amount of food eaten and the amoimt of water taken; the quantity of urine and feces unvoided are likewise of not inconsiderable im- portance. It is the present custom in physiological laboratories, where extreme accuracy is desired, to make sure, so far as possible, that food has not been taken into the stomach for several hours (usually 12) before the weight is measured, and that the bladder is emptied immediately before weighing and feces passed, if possible. It is obvious that these last refinements are with difficulty applicable to the measurement of a series of children. In our own investigations they were followed out in as many cases as possible. 1 Bowditch, Eighth Annual Report Mass. State Board Health, 1877, p. 275. 2 Schmid-Monnard, Jahrb. f. Kinderheilk., 1901, 53, p. 50. 3 Vierordt, Daten und Tabellen fur Mediziner, Jena, 1906, 3 Aufl., pp. 23-24. 4 Griffith, N. Y. Med. Journ., 1917, 106, p. 823. 6 Walker, Proc. Roy. Soc., B, 1915, 89, p. 157; Dreyer, Lancet, Aug. 9, 1919; and von Pirquet, System der Ernahrung, Berlin, 1917, p. 48. 36 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. EARLIER DATA SELECTED FOR COMPARISON WITH OUR DATA. Employing the commonly accepted standards for normality, i. e., the relationships between body-weight and age and between height and age, we have prepared four charts (figs. 3, 4, 5, and 6), on which we have placed several curves, comparing our measurements of weight and height referred to age for our laboratory children 1 with similar measurements derived from other sources. In selecting data from other sources for comparison we were influenced by the considera- tions outlined above relative to accuracy of measurements and number of subjects measured. Almost at the very outset it was clear to us, as has been pointed out by Holt, 2 that the difference in normal may to a great extent be one of racial characteristics and therefore that little, if any, consideration should be given to the average values of foreign writers, especially when these values deal with a very homo- geneous population. For a working basis, however, for this com- parison, we have chosen, first, as representing foreign children, the values from two typical foreign investigators, i. e., those of Quetelet 3 and Schmid-Monnard. 4 Fortunately the fundamental and classical investigations of Bowditch 5 and the more recent studies made in an effort to implant in the American mind the importance of conserving our youth have led to the accumulation of a considerable amount of data which may be stated to be fairly representative of the American people. The more recent data, particularly that of Crum 6 and Wood, 7 have been admirably collected by Gray 8 and have been chosen by us as best representative of American children. Finally, we have also made use of measurements secured on private-school children by Holt 9 in New York City and by ourselves in private schools in the vicinity of Boston and in eastern Massachusetts. The curves for all but our laboratory children and those of Holt were plotted from average values representing definite age-groups. Thus, the weights and heights from Quetelet, Schmid-Monnard, and Wood, as well as those of our private-school children, were averaged for each year, and those from Crum for each 6 months. The curves 1 We use the term "laboratory children" to indicate those children whose metabolism was directly studied by us in one or more of the several respiration apparatus, and to distinguish them from our group of private-school children. *Holt, Am. Journ. Diseases Children, 1918, 16, p. 359. Quetelet, Anthropometrie, Paris, 1871, pp. 177 and 346; also Sur 1'homme et le deVeloppement de ses facult&s, Paris, 1835, 2, p. 46. 4 Schmid-Monnard, Correspondenzbl. d. deutsch. Gesellsch. f. Anthropol., 1900, 31, p. 130. 6 Bowditch, Growth of children, Public Document, Mass. State Board Health, 1877. Cited by Holt, Am. Journ. Diseases Children, 1918, 16, p. 362. 6 Crum, Quarterly Pub. Am. Statistical Assn., Sept. 1916, n. s., No. 115, vol. xv, Boston, pp. 332336. 7 Wood, pjersonal communication to Gray. 8 Gray and Gray, Boston Med. Surg. Journ., 1917, 177, p. 894. 9 Holt, Am. Journ. Diseases Children, 1918, 16, p. 359. NORMALITY OF CHILDREN STUDIED. 37 representing Holt's data and our own laboratory data 1 were obtained by first plotting all the numerous measurements and then drawing through the plotted points smoothed curves representing the general trend. In each case these smoothed curves are really composite curves representing the judgments of five members of the laboratory staff, who were asked to sketch in on tracing paper their visualization of the general trend. While no difficulty was experienced in securing sufficient data with regard to foreign children from the classical researches of Vierordt and others, special consideration must be given the measurements of American children. One of the greatest difficulties in this connection is the fact that most of the body-weight measurements of American children, especially those above 5 years of age, include weight of clothing, and this necessitates a rather uncertain correction for clothing. In the data which we have selected, however, all measurements of body-weight, and indeed height, were taken without clothing, with one exception. Quetelet's measurements of weight involved weight of clothing, but fortunately he corrected his values for the weight of clothing worn by deducting one-eighteenth of the total weight for males and one twenty-fourth for females. Consequently we feel justified in using his values for comparison with our other data. Of special significance is our recalculation, averaging, and charting of a series of figures compiled from the article by Gray, 2 who has very carefully made most suitable selections and who cites only body- weights without clothing. For children from 6 months to 4 years of age Gray gives values from Crum 3 which are based upon 5,602 boys and 4,821 girls. For children from 5 years to 20 years of age he gives values received by him in a personal communication from Professor Wood, of the Life Extension Institute. 4 In reporting Crum's figures, Gray quotes him as inclined to think that the children measured by him may be regarded "as somewhat super-normal, or above average," since they were measured in connection with a "better baby contest," a movement prompted by the committee on public health and instruc- tion of the American Medical Association. The great majority of the 1 The data for the ages, weights, and heights of our laboratory children are given in tables 26, 27, and 28 (pp. 112, 116, and 120) in our subsequent discussion of basal metabolism. Addi- tional data for 8 boys and 4 girls, for whom basal metabolism measurements are not avail- able but for whom we have these physical measurements, are reported in tables 12 and 13 (pp. 54 and 58) in the discussion of anthropometric measurements. As will be explained in detail later, all the individual measurements secured by us for weight, height, and age were not used in plotting these charts, but the values were averaged to a certain extent. Thus, a child was considered a new individual with an increase in age of 6 months, or with an increase in weight of 1 kg. up to 10 kg. and of 10 per cent beyond 10 kg., and measure- ments obtained on a child on two or three successive days or on days relatively close together were often averaged as one value rather than being used separately. 2 Gray and Gray, Boston Med. Surg. Journ., 1917, 177, p. 894. 3 Crum, Quarterly Pub. Am. Statistical Assn., Sept., 1916, n. s., No. 115, vol. xv, Boston, pp. 332-336. See also Gray and Gray, loc. cit., p. 895, table 2. 4 Gray and Gray, loc. cit., p. 896, tables 3 and 4. 341683 38 METABOLISM AND GROWTH FROM BIRTH TO PltBERTY. children were of American-born parents, but were of different stocks, including German, Irish, Swedish, and some Italian. According to Crum's statement, the measurements were made upon " normal healthy children in various sections of the country" 1 and " should serve as a fair guide for many practical purposes without any additional refine- ments." 1 Professor Wood's measurements were made on several thousand boys and girls in the Horace Mann School, connected with Columbia University. The heights and weights were taken without clothing, but he adds the important information that the "weight of clothing ranges from about 3 pounds in five-year-old children to 6 or 7 pounds in older pupils, and is slightly greater for boys' than girls' clothing." 2 He considers that these measurements are fairly repre- sentative of healthy children in the United States. In quoting from Professor Wood, Gray very properly brings out that "a child who is short in stature for his age is apt to be under weight" 2 and that children of constant age but varying height should have different weights ; but for our purpose we have taken the average value of all the weights and all the heights at the several ages, as reported by Crum and Wood, and have included them in a combined curve on our charts. It is of considerable interest to note in the several charts that there is no striking break in the curve between 4 years of age, where Crum's data end, and 5 years of age, where Wood's begin, but that the curve is reasonably regular in character. To indicate the lack of data between 4 and 5 years of age this portion of the curve has been drawn in as a broken line. Holt's measurements were obtained on boys only, in the Browning School in New York City, a school which in his opinion represents one of the better grades of day schools in that city. In all, 1,774 observations were made "on about 350 different boys whose weights were taken semi-annually, without clothes, over a period of years, the average number of observations on each boy being five. This group of American boys, with but few exceptions, came from wealthy families and had had the advantage of good care and proper food all their lives." 3 Our own data for private-school children represent boys and girls probably of the same social class as those measured by Holt. Records of age, height, and nude weights were made on 886 boys from eight private schools and on 323 girls in two private schools, in the neighbor- hood of Boston and eastern Massachusetts. Private schools were selected because the children attending them were presumably living in the most ideal home and school surroundings and should be closer to the ideal American in physical development than those living in less favorable surroundings. 1 Gray and Gray, Boston Med. Surg. Journ., 1917, 177, p. 895. 2 Gray and Gray, loc. cit. , p. 897. 3 Holt, Am. Journ. Diseases Children, 1918, 16, pp. 360 and 362. NORMALITY OF CHILDREN STUDIED. 39 Our laboratory children, whose measurements, together with those of our private-school children, we will now compare with the earlier data we have selected, came from a different social and economic plane of life than our private-school children. This is important to bear in mind in the following comparisons, since it has been shown many times that the development of the child depends on his social surroundings as well as upon his physical well-being. In the several charts now to be considered, since we have but a few scattered obser- vations with our own boys beyond 13 years of age and with our own girls beyond 12 years of age, we have stopped the curves representing our measurements at 13 years for boys and at 12 years for girls. RELATIONSHIP BETWEEN BODY-WEIGHT AND AGE WITH BOYS. As an index of the state of nutrition of children, the relationship between body-weight and age is perhaps one of the oldest and earliest relationships which have been considered. This ratio of body-weight alone to age obviously does not take into account the very important factor of skeletal growth or length. Still, as it represents one of the earliest relationships utilized, and as it has been the basis of a large number of weight charts of various types, it requires special considera- tion. In figure 3, therefore, we have plotted curves for weight referred to age for boys, representing our own laboratory measurements, those in the two earlier European studies of Quetelet and Schmid-Monnard, the combined Crum-Wood data compiled by Gray, the material from Holt, and the values for the private-school boys measured under our direction. The relative positions of these curves are of very great significance. The curves of the two foreign investigators lie throughout practically their entire length measurably below the curve for our laboratory boys, which may undoubtedly in part be explained by the fact that they represent entirely different nationalities. Between themselves they agree reasonably well. Of singular significance is the fact that the combined curve of Crum and Wood is almost identical with our laboratory curve. It is very important to note that below 4 years of age our laboratory boys measure up essentially to the standard set by this curve, although, as has been pointed out before, Crum's children are considered by him to be somewhat supernormal. A number of our children at this age were, however, the offspring of resident wet-nurses. Holt's curve for private-school boys is measur- ably above the line for our laboratory boys and, indeed, above the continuation of the Crum-Wood curve after the thirteenth year, while the curve for our private-school boys lies even higher than that of Holt. In the case of these private-school children we have distinctly larger and heavier boys for the same age, as is unquestionably shown 40 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. by these two curves. It is important to bear in mind here that the Holt curve is derived from 1,774 measurements, while our private- school curve is based upon 886 measurements. Holt's values repre- sent a city school, attended by children of a social status somewhat better than average, who received better medical treatment and were unquestionably living in better hygienic surroundings than the average Kgs 70 65 60 55 50 45 40 35 30 25 .20 15 10 5 Yr WEIGHT REFERRED TO AGE. BOYS. / / 7 A -' lit / / / / $ 2 i/ 1 M / '/ / 4. V < <- ; '"' B. 3UIVM QUE1 -MON SO . & - Cl B. ANDT RUM-WOO: QUETELET -MONNARE SCHOOLS ) f AND SC T.PR HMIC VATE B Yrs. 2 4 6 8 10 12 14 16 18 20 FIG. 5. Relationship between body-weight and age with girls. superiority may be explained either by the fact that private-school life results in the development of excessively heavy children or that the children attending private schools represent a selected class enjoy- ing better economic conditions favoring growth. RELATIONSHIP BETWEEN HEIGHT AND AGE WITH GIRLS. Finally, we have to consider the relationship between height and age with girls. This ratio is shown in figure 6. In this figure the curve for our laboratory girls shows that on the whole their heights are measurably above both foreign standards up to about 8 years of age, but only slightly above the Crum-Wood standard. Beyond the age of 8 years our data agree very closely with both the Schmid- Monnard and Crum-Wood series, but are perceptibly below the private- school series. While, therefore, our girls exhibit a slight underweight above 8 or 9 years of age when compared with the American average, 44 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. Cms. HEIGHT REFERRED TO AGE- GIRLS- 170 ^- 150 / ' / S* ...--- 140 / // f 130 // 120 /x f 110 J, / 100 /I, / 90 / 7 80 / 70 / 60 50 40 1 C B.ANDT. RUM-WOOD QUETELET -MONNARD SCHOOLS AND SC T.PR HMIC VATE B Yrs. 2 4 68 10 12 14 16 18 20 FIG. 6. Relationship between height and age with girls. they do not exhibit a corresponding underheight. Hence it appears (although this is not the place for discussion) that our girls were slightly under weight on the basis not only of age but likewise of height. The private-school girls, as in all the other curves where the private- school children are charted, show a clear superiority in height for age. GENERAL CONSIDERATION OF THE RATIOS OF WEIGHT TO AGE AND HEIGHT TO AGE WITH BOYS AND GIRLS. From figures 3, 4, 5, and 6 it appears that the data for our laboratory children agree for the most part very closely with the average values found by Crum-Wood for these ratios, and that they are in general notably above the measurements of the two foreign observers, although in the case of our girls there is inferiority in weight after the age of 8 years. The private-school data, on the other hand, both Holt's and our own, indicate superiority not only in weight for age, but likewise in height for age throughout the entire age-range studied. This speaks for a distinctly better type of youth in our private schools. It is questionable whether this is due, first, to the growth actually made in the private schools during the course of time, or to the fact that the children entering private schools come from a better class of people, from a better social economic plane, and, particularly, receive NORMALITY OF CHILDREN STUDIED. 45 better medical care. Judged, therefore, by these old standards of height to age and weight to age, our laboratory children measure up to normal (average) in practically every instance, except in the case of the weights of our girls 8 years of age and above, but they do not measure up to the superior type of child found in the private schools in the environs of Boston. RELATIONSHIP BETWEEN HEIGHT AND BODY-WEIGHT WITH BOYS AND GIRLS. With adults, no one would think of considering the relationship between weight and age or height and age as of any particular sig- nificance, but, singularly enough, with children these two ratios, particularly that of weight to age, have long held the attention of physiologists to the exclusion of almost every other method of com- HEIGHT REFERRED TO BODY WEIGHT. BOYS. 160 150 140 130 120 110 100 90 80 70 60 50 40 2 .lilts}- X ^, ^ ^ ^ ^? ^ ~^ r ** - S s*~ / s- S . K' S *' , ?' B AND T.PR VATE B.ANDT. SCHOOLS J 7* / kgs.4 6 8 10 12 14 16 18 20 22. 24 2.6 28 30 32 34 36 38 40 42 FIG. 7. Relationship between height and body-weight with boys. parison. With adults, the comparison of height to weight has long been known to be a comparatively good index of the state of nutrition. If the individual is very tall and light in weight, he is obviously under- nourished; if he is very short and heavy in weight, he is obese. It seems much more logical to consider the relationship between height and weight of children than between weight and age or height and age. Consequently we have plotted on our chart (fig. 7) the values representing the height referred to weight for our laboratory boys. As stated on page 37, these values for weight and height may be found 46 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. in tables 27 and 28 (pages 116 and 120) and tables 12 and 13 (pages 54 and 58) ; they represent, in many instances, not individual measure- ments, but an average of several measurements, at a particular age or weight. On this chart we have arbitrarily sketched a curve showing the general trend of this relationship, the curve (as usual) representing the personal judgments of five different individuals. It is clear from this curve that, with the subjects we have studied, a boy with a height of 110 cm. would weigh approximately 20 kg; on the other hand, a boy weighing 30 kg. would have a height of approximately 134 cm. It is especially to be emphasized that this chart does not take age into consideration, but it more nearly indicates the typical proportions of our children. Since the analysis of the earlier charts demonstrated that our children usually, both for their height versus age and weight versus age, measure up very closely with the Crum-Wood standard, it can be seen that the curve here shown on figure 7 represents not far from the probable usual height-to-weight relationship of boys. 9 HEIGHT REFERRED TO BODY WEIGHT. GIRLS. 140 130 120 110 100 90 80 70 60 50 40 ,^- ;^-- - 1~ ^< ^ ...--' ^ f ^' ..<' ^ /? . K /. ^ ' S /' ' '/f V S'. '* B AND T. PR VATE B. AND1 SCHOOL * t / / kgs. 4 ( > { 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 4( FIG. 8. Relationship between height and body-weight with girls. Our extensive series of measurements of private-school boys enables us to plot on our chart (fig. 7) also the smoothed curve showing the general trend of the relationship between height and weight of these boys. It is of particular interest here to note that the curve showing the relationship between height and weight of private-school boys lies at all points above that showing the height-weight ratio for our labor- atory children. The conclusion is, therefore, that the private-school boys are lighter in weight for the same height than are our laboratory boys. This striking relationship will need special consideration. GROWTH. 47 If we consider the same character of data with our laboratory girls as with our boys, as plotted on figure 8, we see that the sketched line indicating the trend is in general form not unlike that obtaining for boys. Here again we have added the curve representing our private- school girls, 1 which shows, as in the case of boys, a line measurably higher at all points than the corresponding line for our laboratory girls. In other words, the private-school girls are on the average somewhat lighter in weight for the same height than are our girls. This particular phenomenon with private-school children attracts especial attention here, since in all of our earlier comparisons on the charts indicating the ratios of height to age and weight to age it- appears as if the private-school children enjoy a very marked superi- ority. They are taller for the same age and heavier for the same age. When, however, we compare the height to weight irrespective of age, we find that our laboratory children of both sexes are slightly heavier for the same height than are our private-school children. Thus the seeming superiority of the private-school children may to a large extent be questioned as not truly so great as at first sight appears. Our analysis has not been carried out far enough to prove whether or not there is a natural relationship between weight and height and children are heavier because they are taller. But when we consider the general configuration of children as a whole, we find that our labor- atory children, both boys and girls, are somewhat heavier for the same height than are the private-school children; on this basis, therefore, it would appear that our laboratory children of both sexes are some- what superior to the private-school children. From the critical examination of all these data it seems clear that our laboratory children, representative as they are of the institution rather than the select home, are on the whole fully up to the best American standards based upon large series of individuals. The striking superiority in height to age and weight to age of our private- school children is in part at the sacrifice of what is commonly considered the most advantageous relationship between height and weight. GROWTH. The factors determining the height and weight are so subtle, espe- cially during the years of adolescence, that they probably will require final analysis by the biometrician, but for this analysis to be effective there should be a very much larger mass of data at hand than we now possess. Under the circumstances, therefore, we can best secure an indication of the general trend of the relationships between these 1 This curve for our private-school girls should properly begin at 26 kg., as the data below 26 kg. are rather limited. Nevertheless, the curve is projected below 26 kg. to 16 kg., and for this section of the curve a different style of line has been used to show that the data are, strictly speaking, insufficient between 16 and 26 kg. 48 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. growth factors from our graphs and the sketched curves accompanying them. Almost immediately the fact is forced upon one that the re- lationships between height, weight, and age are by no means to be considered as primarily due to the age factor, but the general con- figuration of the body will indicate whether the child is too heavy or too light for his height. The relationships of height to weight are of most importance. This statement is not to be interpreted as meaning that the height-to-age ratio and the weight-to-age ratio should be ignored. Still, the erroneous ideas to which the consideration of the ratios of weight or height to age alone may lead one are well brought out in the discussion of our charts. Here it was shown that although the curves for the private-school children indicated apparently a great superiority over our laboratory children, a subsequent analysis of the relationship between height and weight alone proved that the superi- ority is by no means as great as would at first be implied, and, at least so far as the relation of height to weight is concerned, the private- school children are not superior to those of our laboratory series. The anomalous situation raised by an inspection of these charts leads at once to a consideration of the question of growth, for it is during the period from birth to puberty that we have the greatest changes in growth, both in height and in weight. Indeed, it has been stated that a young infant goes through physical changes in two or three months which it would take adults several years to equal. From the slopes of practically all of our curves it is very clear that the rate of growth is greatest during the first year of life and gradually diminishes as the child grows older. A study of a large number of measurements shows that the greatest gain in weight is made during the first 5 or 6 months of life, when a normal infant almost doubles its birth-weight. During the second 6 months of life an infant gains approximately the same amount of weight as during the first 5 or 6 months; but whereas the increase in weight during the first 5 or 6 months is 100 per cent, in the second 6 months of life it is only about 50 per cent. During the second year of life the increase in weight diminishes to approx- imately 25 per cent, and in the third year it is less than 20 per cent, The percentage increase in height becomes less and less with increasing age, somewhat similar to that noted for weight. Thus, the increase is about 21 per cent during the first 6 months of life, 14 per cent during the second 6 months, and correspondingly less as the child grows older. With these general principles of growth our curves conform in general, particularly the curve illustrating the weight-to-age ratio. It must be clearly recognized at the outset that growth should not be interpreted as meaning only addition of flesh. Growth means the skeletal growth as well, and it is only by establishing the proportion between the skeletal growth and the addition of flesh that the most intelligent consideration of growth can be made. Among the numer- GROWTH. 49 ous factors which affect growth, nationality, environment, and social status and the quantity and quality of the food are the most important. Environment and social status influence the quantity and quality of the food and also determine largely the medical care which a child receives. This latter factor in itself is becoming an increasingly important one in the element of growth, for defects in both anatomical and in dietetic conditions are recognized and corrected by early medi- cal attention. Influence of nationality on growth. The well-recognized influence of nationality on growth must be taken into account in the consideration of the normality of our children. Owing to the cosmopolitan character of the American people and the large influx of European blood, nation- ality undoubtedly plays an important role in the establishment of the American standard, but for comparison with our curves the use of standard growth-curves representing nationalities other than American is, strictly speaking, precluded. Certain nations are known to be tall and well-developed, while others are typically short. Thus, the Anglo-Saxon race and certain branches of the Chinese race are pro- nouncedly taller than other nationalities, while the Japanese race is characteristically short. A superficial comparison of the data from different countries reveals the fact that the growth-curves for different races of men are by no means identical; indeed, this may be said to be true even for similar races inhabiting different localities. Even in Japan there are localities where the people are much taller than the general run of the inhabitants, while in England it has been found that the average weight of infants, both male and female, in London is much higher than that of infants in Leeds and slightly higher than the average weight of infants throughout the whole of England. Con- sequently, in comparing the growth-curves of our children with other growth-curves, emphasis has been laid upon the American data, as American data alone may properly be used in this study of normality. Influence of environment and social status on growth. It has been shown by various authors that environment and social status play a part in the growth of children. Those children who live much of the time out of doors and whose families can afford to supply them with good and sufficient food obtain better growth than children of the same nationality and locality who have not the same opportunities. 1 As early as 1829, Villerme 2 concluded that stature is greater and growth sooner completed, all other things being equal, in proportion as the country is richer and the comforts of the inhabitants more general. Robertson 3 reports that "increasing unfavorability of environment results in a parallel increase of deficiency in weight and stature." 1 Burk, Am. Journ. Psychol., 1898, 9, p. 272. 2 Villermfi, Ann. d'Hyg. Pub. et de M6d. L6gale, Paris, 1829, 1, p. 351. 3 Robertson, Am. Journ. Physiol., 1916, 41, p. 553. 50 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. He also notes that "as the favorability of the environment decreases, the proportion of medical care extended to the children, as indicated by the percentage of removed adenoids, also decreases, while the degree of medical neglect, as indicated by the percentage of infected and unoperated adenoids, undergoes a parallel increase." Food and physical surroundings are not entirely responsible for under-growth. It has been shown that physical defects play a large part in preventing the normal growth and nourishment of children, even under good dietetic and hygienic conditions. Carious and infected teeth and enlarged and infected tonsils and adenoids are the most common physical defects responsible for under-growth and under-nourishment . Exercise. Ordinarily it is considered that the effect of exercise is almost immediately compensated by an increased food consumption, as commonly experienced in the ravenous appetities of vigorously exercising children, yet it may frequently happen that an underfed child automatically restricts his energy expended in work or play to a lower level, in order to provide for growth. In other words, a child that is undernourished can not supply the energy for vigorous exer- cise, since he needs this energy first for growth, and he will auto- matically cut down his activity and thus conserve energy. On the other hand, a child may be led, in the excitement of competition with his playmates, to excessive exercise. In this case the child (although furnished with a considerable amount of food) grows in height in other words, the skeletal growth continues but he does not gain in weight. The lack of growth in weight in this instance is due to excessive exercise and to the fact that extra food has not been added to compensate the increased muscular activity. If such a child is made to take regular rest periods during the day, he will commence to gain weight on a normal diet without an increase in the quantity of food. The value of rest in conserving the energy of children has been most tragically illustrated in the recent experiences of German mothers in the war, as reported by Leonard Hill: 1 "The mothers of Germany kept their war-starved children most of the day in bed, letting them get up at 11 a. m. and go to bed at 4 p. m. Thus they husbanded the national life, taught by scientific experience that growth will make good when the war is over and food supplies become ample." We thus see that the amount of food and exercise together play a very important part in the final resultant growth, and that while the quantity of food affects the general state of nutrition, it is not the sole cause for undernutrition. If the caloric intake is not sufficient to 1 Hill, The science of ventilation and open-air treatment, part i, Special Report Series No. 32, Medical Research Committee, London, 1919, p. 79. GROWTH. 51 cover the energy output due to play and activity, the child will auto- matically restrict his activity so that the limited amount of food fur- nished will provide first for growth, primarily stature. The relationship between food and height. Too much emphasis must not be laid on the caloric value of food only. It is becoming increas- ingly evident that certain unidentified factors in the food, the so-called "food accessories" or popularly termed "vitamines," play a very important role in skeletal growth and the subsequent addition of tissue. Practically all of our knowledge of the relationship between food and height, or skeletal growth, has been obtained from observations made by the American investigators, Osborne and Mendel, 1 and McCollum, 2 on white rats. These important studies have shown that the presence in the diet of the as yet unidentified "food-accessory substances," popularly termed vitamines, is necessary to normal skeletal growth. If these "food accessories" are absent, there is a stuntage of growth in general, but development is not particularly abnormal. Osborne and Mendel were able, by the removal of these "food-accessory sub- stances" from the diet, to defer the growth of white rats for a prolonged period. Subsequently, by supplying the proper foods and "food accessories," they caused these same white rats to attain normal growth without apparent impairment in condition, thus proving that the growth impulse or capacity for growth may be suppressed for a time and exercised later at a period far beyond the age at which growth usually ceases. Jackson and Stewart 3 found that although different organs in the body did not maintain their normal relative weight during inanition, after the proper food was given they resumed their natural growth and attained adequate size despite the early stunting. Retarded growth, therefore, does not necessarily mean that there can be no hope of attaining perfect adult form and function; on the contrary, when organic disease is absent, there is every reason to believe that there will be proper restitution on a correct diet. Food deficiencies may be of two distinct kinds. First, there may be an absence of the "food-accessory substances" which promote growth; second, there may be a caloric deficiency which manifests itself chiefly in the loss of previously stored body material or in a decrease in body storage or growth in weight during the growing period. With a diet insufficient in caloric content, but in which the "food- accessory substances" are still present, skeletal growth continues and, as Waters 4 has shown, even if the diet be very much reduced and much below that required for maintenance, there will still be a long-continued skeletal growth without corresponding gain in weight. Hence it seems as if in the long run, unless children are greatly undernourished 1 Osborne and Mendel, Carnegie Inst. Wash. Pub. No. 156, 1911. 2 McCollum, The newer knowledge of nutrition, New York, 1918. 3 Jackson and Stewart, Am. Journ. Diseases Children, 1919, 17, p. 329. 4 Waters, Proc. Soc. Promotion Agr. Sci., 1908, 29, p. 3. 52 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. and their diet is woefully deficient in the " food-accessory substances," they will have a reasonably normal skeletal growth with increasing age. It is at this point that we note the special significance of the time-honored early consideration of the relationship of height to age. If the height is not up to that ordinarily found for the age, then we may reasonably assume that the " food-accessory substances" in the diet are deficient, unless we are dealing with an abnormal class of individuals containing a large number of foreign population of normally short stature. If the height is equal to that normally found for the age but the weight is too low for the height, we may then look for a deficiency in the caloric intake that is, the calories ingested are not sufficient to take care of the heat output existing at the time, and while food may be ample for normal exercise, with excessive exercise there is deficiency in growth. Here we have the situation which is so popularly expressed by the statement that the boy "runs himself as thin as a rail." From the foregoing consideration of the several factors affecting growth, it can be seen that a study of our growth-curves on the whole that is, of our height-age, weight-age, and height-weight ratios is probably involved by the question of nationality, the question of environment and social status as reflected in the food, exercise, and medical care, the question of " food-accessory substances" in the diet, and the caloric intake. Some of these questions, certainly that of nationality and medical care, are extremely difficult to consider separately. From our curves, however, certain deductions may legitimately be drawn. In the first place, we have seen that, based on the relationship of height to age and weight to age, our laboratory children as a whole, both boys and girls, measure up to the best and most representative standard of American children, i. e., the data of Crum and Wood, but that in neither of these relationships do they measure up to the private-school children, either those studied by Holt or by ourselves. It appears, therefore, that the private-school children are taller and heavier for their age; but a subsequent con- sideration of the relationship between height and weight shows that on the whole they are a little thinner for the same height than are our children. ANTHROPOMETRIC MEASUREMENTS AS INDICES OF GROWTH. From a careful consideration of the various relationships between height, weight, and age we are convinced that the best ratio to indicate the normal state of nutrition is that of height to weight. Still, in a subject as important as this, it is desirable to make use of every conceivable measurement that may possibly contribute towards clari- fying the problem as to what is the normal state of nutrition. Many writers have used, in addition to height and weight, typical measure- GROWTH. 53 ments, such as the girth around the chest or the abdomen at the level of the umbilicus, the height of the subject when sitting, and other measurements. Pediatricians have not, however, accepted any of these generally. To show at a glance whether the individual is well nourished or poorly nourished we need some mathematical expression of relationship between either weight or height and one or more girths. In other words, some quantitative device is needed for con- firming the personal opinion, which is only too frequently based upon superficial inspection. For an entirely different purpose, i. e., primarily to compute the body-surface of our subjects by the extremely ingenious linear formula of Du Bois, 1 we made a large number of measurements of various parts of the body. Many of these are, it is true, not those conventionally recorded by anthropometers, but a number do give indications of growth and are of general as well as special value. Of the numerous measurements required by the Du Bois linear formula, we believe we are warranted in publishing only two, namely, the circumferences at the nipples and at the hips and buttocks. Consequently, in tables 12 and 13, in addition to the age, height, and weight of each child, we have given these two circumferences. 2 These are included hi our tables as much for the benefit of the future biometrician as for our own immediate use. The circumference at the umbilicus seemed to us too dependent upon accidental conditions (such as food in the stomach and gas in the intestines) to bring it into the same anthropometrical order of value as either of the other measurements, although obviously the chest measurement has its own inherent errors. Finally, we have recorded in the last three columns of these same tables the results of our study of body-surface measurements by two different methods. The surface areas by the Du Bois linear formula were obtained primarily to throw light upon the possible relationship between body-surface and basal heat-production. While we firmly believe that this relationship has been very much overestimated and its significance certainly grossly misunderstood, it still remains a fact that as a general index of growth the body-surface may, the- oretically at least, be of a somewhat higher degree of importance than the body-weight. 3 i Du Bois and Du Bois, Arch. Intern. Med., 1915, IS, p. 868. s The data given in tables 12 and 13 were not used in plotting our various charts showing the relationships between age, weight, height, and body-surface (except in the case of 8 boys and 4 girls especially noted in the tables). The values in these tables represent individual measurements in every case and not average values. The selection of data for each child was made on the basis solely of an increase in body-weight of 1 kilogram. The values, therefore, differ somewhat from those for similar body measurements given in tables 26, 27, and 28 (pages 112, 116, and 120) which represent chiefly average values. 3 For a consideration of the significance of the relationship between body-weight and the two- thirds power of the body-weight with other morphological measurements, see Dreyer and Ray, Phil. Trans., 1909-1910, 201, ser. B., p. 133; ibid., Phil. Trans., 1911, 202, ser. B. t p. 191; Dreyer, Ray, and Walker, Proc. Roy. Soc., 1912-1913, 86, ser. B., pp. 39 and 56. See also Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 168. 54 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 12. Body-girths of boys and comparison of body-surfaces, as measured by the Du Bois linear formula and as computed from the formula K\w 2 . Sub- ject No. Age. Body- weight (with- out cloth- ing). Height. Circumference. Body-surface. At nipples. At hips and but- tocks. Du Bois linear. K3w*. (') Per cent de- viation from Du Bois linear. 107 108 1.4 g 106 114 115 27 6 117 61 118 115 119 137 125 124 132 133 118 125 115 126 119 130 141 132 128 147 155 148 136 161 115 159 129 157 119 153 164 156 150 148 158 126 9 days kilos. 3.32 3.40 3.60 3.83 3.83 3.83 3.86 4.54 4.54 4.60 4.71 4.71 4.96 5.03 5.03 5.09 5.47 5.48 5.66 5.74 5.77 5.79 5.97 6.02 6.24 6.24 6.35 6.43 6.53 6.55 6.56 6.70 6.83 7.03 7.08 7.11 7.19 7.20 7.30 7.36 7.45 7.47 7.48 7.49 cm. 51.0 50.5 51.0 53.0 54.0 55.0 53.0 52.0 54.0 56.0 58.5 58.5 58.5 60.0 57.0 57.0 59.0 61.0 61.5 60.5 63.0 60.0 64.5 63.0? 61.5 64.5 61.5 63.0 68.0 66.0 66.5 67.5 65.0 66.5 62.0 64.0? 67.5 67.0 71.0 65.5 72.0 68.0 66.5 66.5 cm. *33 233 cm. sq. m. sq. m. 0.222 ll days .226 11| days .234 8 days *36 *34 34 .245 *33 30 .245 0.243 .245 .246 +0.82 8f days 8 days 274 1| mos 38 *38 36 35 37 *40 37 *40 40 39 39 39 38 40 40 33 *30 33 34 35 .266 .278 .271 .313 .274 .276 .280 .280 .290 .293 .293 .295 +3.01 '+".72 +3.32 -7.35 -2.66 2 mos 1$ mos 2| mos 1 mb. lj wks 4 J mos 2 mos. 1 wk 35 .301 2 mos. 1 wk. . lj mos. . . 38 37 36 39 36 40 38 .322 .312 .318 .325 .317 .327 .331 .310 .310 .317 .320 .321 .322 .329 .331 .344 .344 .350 .355 .361 .362 .362 .371 .378 .389 .391 .392 .395 .395 .399 .401 .404 .405 .405 .406 -3.73 - .64 - .31 -1.54 +1.26 -1.53 - .60 3 mos 3 mos. 1 wk. 3 mos. 1 wk 2 mos. l wka 4 mos 3 mos 4 mos. 1 wk 3| mos 2 mos. 3 wks 5 mos. 1 wk 6 mos. 3 J wks 5 mos. 1 wk 238 40 *39 42 39 42 40 44 41 41 "43 ' 4 y 43 *43 43 244 43 44 41 40 39 *41 40 38 42 40 40 43 42 .340 +1.18 -8.97 -5.74 -4.49 +1.69 -6.08 +1.07 +4.01 .390 .383 .379 .356 .395 .374 .374 4 mos. 1 wk. . . . 7j mos 7 mos. 3 wks. 7 mos 3 mos 7 mos. . . . 41 41 42 *38 41 44 45 .399 .386 .426 .394 .403 .400 - 1.00 +2.33 ' ^5.87' '+2.79 + .50 +1.50 6 mos. . 6 mos. 1 wk 9 mos. 7 mos. 6 mos. 7 mos. . . 7 mos. 6 mos 1 For values of K see table 14. 2 The measurements noted are not part of the regular series of Du Bois linear measurements. They are entered here on the assumption that they represent measurements similar to those taken in the use of the Du Bois linear formula. 8 The data for this subject were used in plotting the charts shown in figures 3 to 11 (pp. 40 to 66). 'Previously reported as O. C.; Benedict and Talbot, Am. Journ. Diseases of Children, 1914, 8, p. 1. GROWTH. 55 TABLE 12. Body-girths of boys and comparison of body-surfaces, as measured by the Du Bois linear formula and as computed from the formula K^w* Continued. Sub- ject No. Age. Body- weight (with- out cloth- ing). Height. Circumference. Body-surface. At nipples. At hips and but- tocks. DuBois linear. *,f Per cent de- viation from DuBois linear. 155 154 138 119 161 136 153 158 142 148 170 149 138 158 148 168 158 138 161 153 136 142 161 153 138 148 158 119 176 153 175 158 155 119 119 153 177 182 192 207 186 194 197 204 199 212 10 mos kilos. 7.56 7.91 7.94 8.08 8.09 8.19 8.44 8.48 8.78 8.87 9.20 9.33 9.53 9.55 9.76 9.95 10.0 10.1 10.1 10.2 10.6 10.7 10.9 11.2 11.3 11.3 11.7 11.7 12.2 12.5 12.5 12.7 12.7 12.7 13.6 13.9 14.6 15.4 18.8 18.9 19.3 19.7 19.9 19.9 20.2 20.5 cm. 71.5 66.0 68.0 70.5 70.5 71.0 68.5 70.5 66.0 72.0 74.0 75.0? 74.5 77.0 78.0 74.0 81.0 79.5 77.5 77.0 76.5 70.0 80.5 80.0 82.0 82.0 82.0 79.5 87.5 84.0 82.0 84.0 90.0 83.5 90.5 88.5 88.5 94.0 106.0 107.5 110.5 107.5 114.0 111.5 115.5 120.5 41 44 43 43 43 44 46 45 45 45 em. 43 45 46 45 45 41 45 45 49 45 sq. m. 0.412 .434 .414 .409 .419 .427 .435 .449 .442 .435 sq. m. 0.408 .421 .422 .427 .427 .431 .439 .441 .451 .454 .466 -0.97 -3.00 +1.93 +4.40 +1.91 + .94 + .92 -1.78 +2.04 +4.37 65 mos 4 mos. 3 wks. 7j mos. 8 mos. 3 wks. 11 mos. 3 wka 4 mos 8f mos 10 mos 55 mos .470 .476 .477 .484 490 +6.73 -1.24 +5.68 10 mos. 3$ wks 1 yr. 6 mos 46 47 47 *46 48 48 48 48 50 49 50 50 49 49 51 50 53 51 52 51 50 52 52 51 55 55 58 59 59 59 57 59 60 61 46 45 44 .446 .483 .458 12i mos. 9j mos. 44 46 49 47 49 50 49 50 49 47 51 52 48 51 49 53 50 54 54 52 52 55 57 58 58 60 58 60 57 58 .510 .508 .509 .505 .509 .530 .542 .553 .552 .521 .541 .541 .575 .570 .552 .562 .568 .609 .624 .611 .608 .649 .758 .804 .829 .820 .799 .820 .812 .882 .492 .495 .495 .498 .511 .514 .521 .530 .534 .534 .546 .546 .562 .571 .571 .577 .577 .577 .604 .613 .633 .671 .791 .794 .805 .817 .822 .822 .831 .839 -3.53 -2.56 -2.75 -1.39 + .39 -3.02 -3.87 -4.16 -3.26 +2.50 + .92 + .92 -2.26 + .18 +3.44 +2.67 +1.58 -5.25 -3.21 + .33 +4.11 +3.39 +4.35 -1.24 -2.90 - .37 +2.88 + .24 +2.34 -4.88 1 yr. 4 mos. 85 wks 1 yr. 2 mos 9 mos 8 mos. l wks 1 yr. 4 mos 2 yrs 1 yr. 10 mos. 3 wks 1 yr. 5 mos 2 yrs. 6 mos. 1 wk 1 yr. 1 mo. 3 wks. 2 yrs. 4 mos. 3| wks. . . . 2 yrs. 5 mos. 1 wk 2 yrs 10 mos 2 yrs. 6 mos 1 jr. 6 mos 2 yrs. 1 mo 3 yrs. 2 wks 2 yrs. 7 mos. 2 wks 4 yrs 5 yrs. 6 mos. 3 wks 7 yrs. 7 mos. 2 wks 4 yrs. 8 mos. 3 wks 5 yrs. 9 mos. 1 wk. 6 yrs. 9 mos. 3 wks 6 yrs. 10 mos. 2 wks. . . . 8 yrs. 1 mo 1 For values of K see table 14. 2 The measurements noted are not part of the regular series of Du Bois linear measurements. They are entered here on the assumption that they represent measurements similar to those taken in the use of the Du Bois linear formula. 3 The data for this subject were used in plotting the charts shown in figures 3 to 11 (pp. 40 to 66). 56 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 12. Body-girths of boys and comparison of body-surfaces, as measured by the Du Bois linear formula and as computed from the formula K^w 2 Continued. Sub- ject No. Age. Body- weight (with- out cloth- ing). Height. Circumference. Body-surface. At nipples. At hips and but- tocks. Du Bois linear. K3vfi. M Per cent de- viation from Du Bois linear. 187 215 205 2 213 201 2 200 2 208 193 2 231 2 216 201 2 226 218 222 209 224 202 223 217 218 242 211 232 228 244 253 235 229 247 249 241 236 245 240 256 252 237 246 243 259 255 254 260 250 258 5 yrs. 3 wks. kilos. 20.6 20.7 21.3 21.5 21.8 22.0 22.6 23.7 24.0 24.1 24.2 24.6 24.8 25.0 25.2 25.6 25.8 25.9 26.6 26.8 26.8 26.9 28.2 28.6 29.5 30.0 30.1 30.4 30.5 30.6 30.6 31.3 31.4 33.6 33.7 34.1 34.4 36.7 37.9 37.9 37.9 39.2 39.2 41.1 51.1 cm. 111.0 116.5 117.5 119.5 122.5 123.0 118.0 118.0 122.5 123.5 124.0 122.0 128.5 122.5 125.5 126.0 121.0 129.0 123.5 133.5 126.0 129.0 127.0 126.5 132.0 139.0 134.0 128.0 141.0 135.5 136.0 132.5 135.5 138.5 137.5 139.0 139.5 150.5 149.5 151.5 153.0 151.0 147.0 150.5 159.5 cm. 57 59 60 64 60 62 60 62 63 61 61 62 62 64 63 63 64 63 64 62 65 68 69 65 66 67 67 70 66 68 66 70 69 70 69 69 68 72 73 73 68 72 72 73 83 cm. 58 61 60 57 59 59 64 64 62 61 62 64 64 63 64 63 63 63 65 66 64 65 67 68 67 66 68 69 67 70 70 69 72 71 72 72 71 75 75 76 74 74 79 83 sq. m. 0.832 .838 .887 .872 .874 .876 .873 .904 .976 .976 .934 .926 .953 .942 1.009 1.025 .939 .991 .965 .988 .985 1.045 1.090 .998 1.082 1.089 1.084 1.057 1.153 1.183 1.100 1.103 1.166 1.237 1.224 1.245 1.217 1.256 1.256 1.341 1.303 1.335 1.303 1.368 1.493 sq. m. 0.842 .845 .861 .866 .874 .879 .895 .924 .932 .935 .937 .948 .953 .958 .968 .989 .999 1.004 1.025 1.030 .030 .032 .065 .076 .098 .110 .113 .120 .123 .125 .125 .142 .145 .197 .200 .209 .216 .270 .298 .298 .298 .327 1.327 +1.20 + .84 -2.93 - .69 0.00 + .34 +2.52 +2.21 -4.51 -4.20 + .32 +2.38 0.00 + 1.70 -4.06 -3.51 +6.39 + 1.31 +6.22 +4.25 +4.57 -1.24 -2.29 +7.82 + 1.48 + 1.93 +2.68 +5.96 -2.60 -4.90 +2.27 +3.54 -1.80 -3.23 -1.96 -2.89 -0.08 + 1.11 +3.34 -3.21 - .38 - .60 +1.84 7 yrs. 2 mos. 1 wk 8 yrs. 1 mo. 3? wks 6 yrs. 11 mos. 3? wks. . . 6 yrs. 11 mos. 3 wks. . . . 5 yrs. 7 mos. 3 wks 10 yrs. 3 mos. 3 wks 8 yrs. 4 mos. 3 5 wks. . . . 7 yrs. 3 mos. 2| wks. . . . 9 yrs. 5j mos 8 yrs. 63 mos 9 yrs. 3i wks 7 yrs. 10 mos. 1| wks. . . 9 yrs. 3 mos. 3 wks 7 yrs. 2 mos 9 yrs. 1 mo. 2 wks 8 yrs. 6 mos 9 yrs. 5? mos 11 yrs. 2 mos. 1| wks. . . . 8 yrs. 3 wks. . . 10 yrs. 4 mos. 9 yrs. 9 mos. 3 wks 11 yrs. 4 mos. 1? wks. . . . 12 yrs. 5 mos. 10 yrs. 7 mos. 1 wk 9 yrs. 10 mos. 3| wks. . . 11 yrs. 8 mos. 1 wk 11 yrs. 11 mos. 3 wks. . . . 11 yrs. lj mos 10 yrs. 8^ mos 11 yrs. 5j mos 1 1 yrs. 1 mo. 1 wk 12 yrs. 85 mos 12 yrs. 3 mos. 2 wks 10 yrs. 8 mos. 3| wks. . . . 11 yrs. 6 mos 11 yrs. 3 mos. 1$ wks. . . . 14 yrs. 1 mo 12 yrs. 8 mos 12 yrs. 7 mos. 3 wks 15 yrs. 1 wk 12 yrs. l mos 13 yrs. 8 mos 1 For values of K see table 14. 2 The data for this subject were used in plotting the charts shown In figures 3 to 11 (pp. 40 to 66). GROWTH. 57 With regard to the measurement of body-surface, we must assume that the Du Bois linear formula gives the actual areas very closely. A note of caution has been sounded, however, by Du Bois, 1 who specifically states that it does not seem advisable to use this formula for infants under 2 years of age until the factors have been tested by the measurements of other infants. In working out their formula, the Du Boises used a small cadaver of a child 21 months old, weighing 6.27 kg., with a height of 73.2 cm., measured about two hours after death. They state " the subject was small for her age and had suffered from rachitis; the epiphyses at the wrists were large and the thorax was pigeon-breasted, being narrow and very deep antero-posteriorly." 2 The error in measurement by the linear formula as compared with the measurement of the cast was 2.9 per cent. We are of the opinion that, in limiting the use of their formula, weight rather than age should be the criterion, and that in the case of children the limit should be not children under 2 years of age but children under 6.27 kg. in weight. With this modification of their limit we are hi full accord with the Du Boises. It seemed to us, however, that if we computed the surface areas of our children, both by the Du Bois linear formula and by some standard formula, such as the Lissauer formula, 3 we could compare the values obtained on these two bases for children below 6.27 kg. in weight, note the agreement between the two methods, and thus estimate approx- imately the probability of accuracy in the linear formula for weights below 6 kg. Exactly this procedure we have carried out. We have i carefully measured by the Du Bois linear formula the surface areas of 14 boys and 19 girls of weights under 6.27 kg. Comparing the areas thus measured with the areas computed by the formula proposed by Lissauer (10.3 %*), we found that with our boys there was a tendency for the area by the Lissauer formula to be slightly higher than that by the Du Bois linear formula; therefore, in estimating the body-surfaces of boys with weights of 6 kg. or less, we propose employing the formula K^tf, substituting the factor 10.0 for 10.3, in the belief that 10.0 is a more representative factor than the 10.3 used by Lissauer. This new factor of 10.0 gives results about 3 per cent lower than the Lissauer factor. Similarly, for estimating the body-surfaces of girls below 6 kg. in weight, the factor would be more nearly 10.1 on the average, rather than 10.3, that is, 1 per cent larger than the factor for boys. Accord- ingly, for computing surface areas of children with body-weights of 10 kg. or less, the factor would be not far from that proposed by Lissauer. This agreement between the new factors suggested by us (which were obtained by inspection of the surfaces computed by 1 Sawyer, Stone, and Du Bois, Arch. Intern. Med., 1916, 17, p. 855. 2 Sawyer, Stone, and Du Bois, loc. cit,, p. 856. 3 Lissauer, Jahrb. f. Kinderheilk., 1902, N. F., 58, p. 392. 58 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 13. Body-girths of girls and comparison of body-surfaces, as measured by the Du Bois linear formula and as computed from the formula K^w*. Sub- ject No. Age. Body- weight (with- out cloth- ing). Height. Circumference. Body-surface. At nipples. At hips and but- tocks. Du Bois linear. *$?. 0) Per cent de- viation from Du Bois linear. 49 26 111 113 2 110 109 123 12 131 121 48 113 120 127 35 122 139 145 152 48 131 151 "143 160 134 140 122 113 123 139 131 135 165 152 139 160 35 169 163 144 166 162 127 140 160 3 wks kilos. 2.84 3.56 3.57 3.65 3.70 3.75 3.78 3.85 4.20 4.34 4.80 4.81 4.88 4.94 5.03 5.07 5.17 5.19 5.22 5.49 5.54 5.55 5.64 5.73 5.90 6.00 6.02 6.03 6.04 6.09 6.11 6.17 6.20 6.24 6.46 7.00 7.05 7.17 7.49 7.63 7.80 7.85 8.00 8.04 8.11 8.12 cm. 49.0 50.0 53.0 53.0 53.0 51.0 51.5 53.5 53.0 55.5 58.0 56.0 59.5 58.0 57.5 58.5 58.5 63.0 62.0 60.5 61.0 60.0 60.0 59.0 62.5 64.0 60.0 62.5 65.0 63.0 65.0 62.0 63.0 63.0 65.0 67.0 65.5 64.5 67.5 63.0 62.0 68.5 69.5 67.0 68.0 68.5 cm. 30 cm. 227 sq. m. sg. m. 10 days 13 days 0.235 233 34 35 37 37 35 35 239 36 37 236 37 239 38 38 39 39 239 40 241 39 40 241 39 40 40 41 40 40 *39 42 41 43 41 241 41 243 47 45 244 46 46 44 29 31 33 34 236 31 w ks. 0.249 .244 .252 .239 .242 .244 .245 .248 .263 .269 .287 .287 .290 .293 .296 .298 .301 .302 .303 .314 .316 .316 .319 .323 .329 .333 .334 .335 .335 .338 .339 .342 .344 .346 .358 .388 .390 .394 .406 .411 .417 .419 .424 .425 .428 .428 -4.02 + 1.64 +6.75 -2.68 + 1.72 -3.22 -3.51 - .33 + .32 + 1.28 -2.12 -3.80 9 days . ... 11 days 11? days 2 mos 1 wk. 9 days 3 mos 233 36 35 236 39 35 36 38 233 39 233 37 37 41 40 39 38 40 40 ' '.298 ' .291 .311 .313 .304 .313 .312 .330 .342 .319 .347 .335 .339 .351 .336 2 mos 2 mos 2 mos. 3 j wks 5 wks 2 mos. 1 wk 4 mos 5 mos . ... 3 mos. 2 mos. 3 wks. 4 mos. 3 wks 3 mos. 3| wks 4 mos. 3J wks 6 mos. 1 wk 6 mos. 1 wk 7 mos. 3 wks 4 mos 8 mos. 3 wks 6 mos 7 mos +4.70 -3.46 0.00 - .29 -3.42 + 1.79 -5.98 -1.38 + 1.31 +4.00 -4.02 -2.34 +3.20 +2.66 - .70 +5.16 40 40 43 39 *46 46 *46 48 46 46 43 42 .368 .363 .383 .375 10 mos 4 mos 9j mos .423 .427 .406 .414 .431 .407 8 mos. 1 wk. . . 4 mos. 3 wks. 9 mos 8 mos 9| mos 1 jr. 2 mos 1 yr. 3| wks. 1 For values of K see table 14. 2 The measurements noted are not part of the regular series of Du Bois linear measurements. They are entered here on the assumption that they represent measurements similar to those taken in the use of the Du Bois linear formula. 3 The data for this subject were used in plotting the charts shown in figures 5 to 14 (pp. 43 to 68). GROWTH. 59 TABLE 13. Body-girths of girls and comparison of body-surfaces, as nwasured by the Du Beds linear formula and as computed from the formula K^w 2 Continued. T)_ J, r Circumference. Body-surface. GnK Jtsoay- weight Per DUD- ject N_ Age. (with- out Height At At hips and DuBois K-\/W*. cent de- viation o. cloth- nipples but- linear. (i) from ing). tocks. Du Bois linear. kilos. cm. cm. cm. sq. m. sq. m. 171 10 mos 8.18 73.5 44 46 0.415 0.430 +3.61 139 9 mos. 1 wk 8.26 70.0 46 48 .416 .433 +4.09 146 5 mos. 1 wk 8.30 68.5 44 50 .425 .435 +2.35 167 9 mos. 1 wk 8.52 69.0 45 48 .458 .442 -3.49 172 llj mos. 8.80 74.5 45 43 .456 .452 - .88 166 1 yr. 1 mo. 1 wk 8.97 71.5 45 47 .471 .458 -2.76 127 1 yr. 3 mos. 3 wks 9.02 73.0 48 48 .475 .460 -3.16 146 7 mos. 9.04 71.0 44 48 .426 .460 +7.98 145 10 mos 1 wk. 9.19 70.0 46 48 .464 .465 + .22 173 11^ mos 9.19 72.0 46 47 .467 .465 .43 171 1 yr. 2 mos. Ij wks 9.43 76.5 45 47 .465 .473 + 1.72 144 7 mos 9.70 65.5 49 56 .483 .482 - .21 139 1 yr. 2 mos. 3 wks 9.74 76.0 48 48 .473 .483 +2.11 122 1 yr. 4 mos 9.80 76.0 49 47 .495 .485 -2.02 172 1 yr. 1 mo. 1 wk 9.84 77.0 48 47 .503 .487 -3.18 122 1 yr. 6 mos. 3j wks 10.4 78.5 47 50 .522 .509 -2.49 145 11 mos. 3| wks. 10.5 73.5 49 49 .514 .513 .19 173 1 yr. 4 mos. 3 wks 10.5 78.0 47 50 .520 .513 -1.35 144 9 mos 10.6 67.5 50 57 ' .520 .517 .58 171 1 yr. 9j mos 10.8 85.5 49 47 .496 .526 +6.05 139 1 yr. 8? mos 10.9 82.0 49 50 .541 .530 -2.03 174 2 yrs. 2 wks 10.9 79.0 50 49 .536 .530 -1.12 166 1 yr. 8 mos. 3 wks 11.0 80.0 48 48 .518 .534 +3.09 172 1 yr. 5j mos 11.1 79.5 49 49 .551 .537 -2.54 173 1 yr. 10 mos. 3 wks 11.5 83.0 48 51 .536 .550 +2.61 166 2 yrs. 2 mos. 11.8 85.0 49 50 .568 .560 -1.41 145 1 yr. 2 mos 11.8 76.0 50 54 .562 .560 - .36 171 2 yrs. 3 mos 12.0 89.5 49 49 .573 .566 -1.22 139 2 yrs. 2| mos 12.3 88.0 50 51 .588 .575 -2.21 178 2 yrs. 10 mos. 3j wks. . . 12.3 92.0 50 49 .608 .575 -5.43 166 2 yrs. 9j mos 13.2 88.5 52 51 .599 .603 + .67 145 1 yr. 5 mos 13.4 80.0 53 59 .621 .609 -1.93 171 2 yrs. 10 mos. 1^ wks. . . 13.4 95.0 51 51 .603 .609 +1.00 139 2 yrs. 6 mos. 3 wks. .*. . . 13.5 92.5 55 54 .647 .612 -5.41 166 3 yrs. 4 mos. 2 wks 14.0 92.5 50 51 .604 .627 +3.81 171 3 yrs. 3 mos. 1 wk 14.2 100.0 51 50 .624 .633 + 1.44 179 3 yrs. 8 mos 14.5 98.5 53 54 .654 .642 -1.83 139 3 yrs. 7 mos. 1 wk 14.7 98.0 53 55 .655 .648 -1.07 145 1 yr. 9 mos. 3 wks 15.1 86.5 55 60 .668 .660 -1.20 190 5 yrs. 3| mos 15.2 103.5 53 54 .691 .663 -4.05 180 3 yrs. 10 mos. 3 wks. . . . 15.4 93.5 53 53 .691 .669 -3.18 183 4 yrs. 3 mos. 3 wks 15.7 97.5 53 56 .654 .677 +3.52 145 2 yrs. 4 mos 15.8 94.0 53 56 .664 .680 +2.41 184 4 yrs. 4 mos. 1 wk 16.2 103.0 55 56 .716 .692 -3.35 181 3 yrs. 11 mos 16.4 98.5 55 56 .701 .697 - .57 171 4 yrs. 2 mos. 1 wk 16.5 104.5 54 54 .681 .700 +2.79 195 6 yrs. 2j wks. ' 16.7 99.5 55 56 .666 .706 +6.01 145 2 yfs. 9 mos. 3 wks 16.7 96.5 55 58 .692 .706 +2.02 188 5 yrs. 1 mo. 1? wks 16.9 103.5 54 57 .736 .711 -3.40 145 3 yrs. 1 mo. 3 wks 17.5 98.5 55 58 .700 .728 +4.00 191 5 yrs. 5^ mos. 18.7 107.5 57 58 .748 .761 +1.74 1 For values of K see table 14. 60 METABOLISM AND GEOWTH FROM BIRTH TO PUBERTY. TABLE 13. Body-girths of girls and comparison of body-surfaces, as measured by the Du Bois linear formula and as computed from the formula K^lw 2 Continued. Sub- ject No. Age. Body- weight (with- out cloth- ing). Height. Circumference. Body-surface. At nipples. At hips and but- tocks. Du Bois linear. K-\lw z . n Per cent de- viation from Du Bois linear. 206 *185 196 206 189 220 203 219 225 210 227 221 198 214 239 230 233 238 234 248 233 251 257 239 kilos. 19.2 19.4 19.7 20.6 22.6 23.0 23.1 23.8 23.8 24.0 24.8 25.5 26.0 26.0 27.5 27.6 27.9 28.0 28.2 28.5 30.0 30.8 36.7 39.2 cm. 113.0 106.5 118.0 116.0 116.0 122.0 119.0 125.0 120.5 122.5 125.5 122.0 124.0 126.0 133.5 133.0 120.5 135.5 133.0 129.0 131.0 138.5 140.5 147.5 cm. 58 58 52 60 59 61 60 60 60 63 61 65 62 60 68 65 65 61 64 64 65 68 68 72 cm. 58 60 58 61 64 62 62 62 65 62 66 67 68 68 70 70 71 68 69 69 73 71 77 79 sq. m. 0.787 .775 .816 .835 .871 .933 .881 .890 .908 .940 .939 1.022 .966 .964 1.089 1.055 1.043 1.047 1.033 1.034 1.065 1.095 1.217 1.272 sq. m. 0.774 .780 .788 .825 .888 .898 .901 .919 .919 .924 .944 .962 .974 .974 1.012 1.014 1.022 1.024 1.029 1.036 1.072 1.090 1.226 1.281 -1.65 + .65 -3.43 -1.20 + 1.95 -3.75 +2.27 +3.26 + 1.21 -1.70 + .53 -5.87 + .83 + 1.04 -7.07 -3.89 -2.01 -2.20 - .39 + .19 + .66 - .46 + -74 + .71 9 yrs. 2 wka 8 yrs. 11 mos. 1 wk 9 yrs. 5 mos. 1 wk 8 yrs. 2 wks 9 yrs. 8 mos. 3 wks 9 yrs 6 yrs. 8 mos 8 yrs. 2 mos 11 yrs 10 yrs. 3 mos 9 yrs. 2j mos : . . 10 yrs. 9 mos. 3? wks. . . 10 yrs. 5 mos. 3j wks. . . 11 yrs. 10 mos. 3 wks. . . 10 yrs. 5^ mos 12 yrs. 2 mos. . . . 13 yrs. 3 mos. 3 wks 12 yrs. 3| wks 1 For values of K see table 14. 2 The data for this subject were used in plotting the ^charts 'shown in figures 5 to 14 (pp. 43 to 68). the Du Bois and Lissauer formulas) and the factor of 10.3 proposed by Lissauer rather strengthens us in our view that the Du Bois linear formula may be properly applied to children weighing 6.27 kg. or under. Based upon our comparison of surface areas by the Lissauer and Du Bois linear formulas, we propose a series of constants for use in the formula, the cube root of the square of the weight multiplied by a constant factor K, which will give the most probable surface areas of children. This constant, K, varies with children of different weights and there is a slight difference between boys and girls, as shown in table 14 herewith. In considering these constants, it will be noted that for both boys and girls up to about 10 or 15 kg. an average factor of 10.3 would not vary much more than 3 per cent from that found best fitted to the situation. Consequently, we must emphasize strongly here the accu- racy of the old Lissauer formula for computing the surface area of young children. While our own measurements of the body-surface GROWTH. 61 of children weighing 6 kg. or under would indicate that the factors for K are 10.0 and 10.1 for boys and girls, respectively, the Lissauer factor (10.3) is still in close agreement with our factors. This shows the complete futility of other constants formerly used, which were as high as 11.9, the constant of Meeh. 1 TABLE 14. Constants for computing surface area Boys. Girls. Body- weight (without clothing). Constant. Body-weight (without clothing). Constant. Up to 6 kg 10.0 Up to 6 kg 10.1 6 to 15 kg 10.6 6 to 10 kg 10.6 15 to 25 kg 11.2 10 to 20 kg 10.8 25 to 40 kg 11.5 20 to 40 kg 11.1 A close analysis of Lissauer's data has been given in earlier publica- tions from this Laboratory. 2 The constants obtained from the meas- urements of children by Lissauer range from 8.92 to 12.40; but Lis- sauer emphasizes the fact that 10 of his 12 children were very much under weight. The factor finally selected by Lissauer as representative of normal was a value determined on a normal infant, Si. While all of our own children were selected primarily from the standpoint of normality, and while the constants for children at this early age, accord- ing to our calculations, more nearly approach the factor 10.0 or 10.1, we still believe that the Lissauer factor is on the average very repre- sentative for children under 10 kg. in weight. Our factors indicate strongly that there is not a very great disproportion in the surface area with changes in body-weight under these conditions. In ex- plaining some of the aberrant types of metabolism frequently found with atrophic children, it has been the custom to lay considerable emphasis upon the fact that there is a profound disturbance in the relationship between body-weight and body-surface with under- weight children. 3 Based upon our normal measurements and a careful analysis of Lissauer 's measurements, we believe that our constants are the best factors for computing the body-surface from the body-weight and that a considerable degree of emaciation in all probability does not profoundly affect the constant. As an indication of the accuracy of these computations of surface area by the revised Lissauer formula, taking into account weight only and our several constants, we have recorded in tables 12 and 13 the areas as computed by the Du Bois linear formula and as computed using our several constants. It will be found that with a number of our children, especially at the lower weights, the Du Bois linear 1 Meeh, Zeitschr. f. Biol., 1879, 15, p. 425. 2 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 164; see also Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 143. 'Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 163. 62 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. measurements are lacking. We believe, however, from our com- parison of the areas by the two methods for the 14 boys and 19 girls under 6.27 kg. in weight, that for children at this lower weight-range the use of our constants gives the areas with a high degree of probability In the last column of the tables is indicated the percentage difference between the body-surface by the Du Bois linear formula as the base- line, and that computed by our constants. In examining the tables, it is important to note that the value of K changes with the sex and with the various weights, as indicated in table 14 given above. From tables 12 and 13 it can be seen that the elaborate series of Du Bois measurements are no longer necessary in computing the sur- face areas of children certainly not children approximating normal form but one can compute the area, using the factors as given in table 14. For the sake of convenience we have felt it desirable to give a table for boys and girls separately, showing the surface areas based upon our several factors for each increasing kilogram in weight from 3 kg. to 40 kg. From this table the interpolations for fractions of kilograms are readily obtained. While by no means certain of the importance of the surface area, particularly in relation to metabolism measurements, we still deem it desirable to have a tabular method for securing rapidly the body-surface of children based upon the admirable linear formula of Du Bois. These surface areas, as pre- sented in table 15, are primarily of significance as purely physical measurements. TABLE 15. Estimated body-surfaces for body-weights from 8 to 40 kilograms. Body- Body- Body- weight Surface 1 Surface 1 weight Surface 1 Surface 1 weight Surface 1 Surface 1 (with- esti- esti- (with- esti- esti- (with- esti- esti- out mated mated out mated mated out mated mated cloth- for boys. for girls. cloth- for boys. for girls. cloth- for boys. for girls. ing). ing). ing). kilos. sq. m. 89. m. kilos. sq. m. sq. m. kilos. sq. m. sq. m. 3 0.208 0.210 16 0.711 0.686 29 1.086 1.048 4 .252 .255 17 .740 .714 30 1.110 1.072 5 .292 .295 18 .769 .742 31 1.134 1.095 6 .330 .333 19 .797 .769 32 1.159 1.119 7 .388 .388 20 .825 .796 33 1.183 1.142 8 .424 .424 21 .853 .845 34 1.207 1.164 9 .459 .459 22 .879 .872 35 1.230 1.188 10 .492 .492 23 .906 .898 36 1.254 1.210 11 .524 .534 24 .932 .924 37 1.277 1.233 12 .556 .566 25 .958 .949 38 1.300 1.255 13 .586 .597 26 1.009 .974 39 1.323 1.277 14 .616 .627 27 1.035 .999 40 1.345 1.298 15 .645 .657 28 1.060 1.024 1 Surfaces estimated, using formula table 14. for values of K at the different weight-ranges, see GROWTH. 63 PHYSIOLOGICAL AND ANTHROPOMETRICAL SIGNIFICANCE OF SURFACE AREA. Studies of the body-surface measurements of children have been prompted primarily from an attempt to correlate the total heat pro- duction of children with some physical factor. The apparent dis- crepancy between the heat production per kilogram of body-weight and the total weight of the child at various ages early led to an effort to secure, if possible, a uniform basis for comparison of children and adults, i.e., individuals of greatly differing weights, and it was believed that the heat production per unit of body-surface was much more nearly constant than the heat production based either upon the body-weight or any other factor; in fact, so constant as to represent a "physiological law." The extensive series of body-surface measure- ments made by us were admittedly secured primarily in connection with a study of the possible relationship between heat production and body-weight on the one hand and heat production and body-surface on the other. A recent critical analysis of the so-called " body-surface law" appearing from this laboratory 1 vitiates, we believe, to a very large degree, the significance of this "law," particularly that part of it which recognizes a causal relationship between the surface area and the heat production. Our present position on this point can be no better set forth than by repetition of an opinion expressed six years ago 2 to the effect that we believe body-surface has no significance in connection with heat production, except that it represents a general morphological law of growth. We have, however, a series of carefully measured surface areas of children of varying ages. That these measurements are of direct physiological and anthropometrical importance is undoubtedly true. Thus, the marked changes appearing in the general configuration of the growing child as compared with the adult are familiar to all. Are these changes accompanied by disturbances in the general rela- tionship between surface area and body-weight, or surface area and height, or surface area and age? This question can be adequately answered only by graphic presentation of our data. RELATIONSHIP BETWEEN SURFACE AREA AND BODY-WEIGHT, HEIGHT, AND AGE WITH BOYS. Since these surface areas are all actually measured according to the Du Bois linear formula and are not computed from the body-weight, it seems perfectly justifiable to plot them against age, height, and body-weight, and this we have done in the six following charts for both boys and girls. 3 Thus, in figure 9 we have plotted the body- 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 129. z Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 168. See tables 26, 27, and 28 (pp. 112, 116, and 120) and tables 12 and 13 (pp. 54 and 58) for data plotted on these charts. It will be noted that in 20 instances with the very young boys and in 19 instances with the very young girls the surface areas have been computed from the Lissauer formula, since the Du Bois measurements were not made in these cases. We feel justified in plotting these values on our charts, however, since we have demonstrated the remarkably close agreement between the surface-area measurements obtained by the Lissauer formula and the Du Bois linear formula for young children. 64 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. surface referred to weight for our boy subjects. This curve, which represents the impressions of five members of the Laboratory staff, shows that hi general the relationship between the surface area and weight is almost a linear one, particularly after the 8-kg. weight. In- deed, it can be seen that a straight-line curve would at the higher Sq.m. BODY SURFACE REFERRED TO WEIGHT. i'Aj x^ ^* * * ^ *\ . x / '/^' X ^ i ..3 X x /? > x X X f ^ S / / ,A kgs.4 e 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 4 FIG. 9. Relationship between body-surface and body-weight with boys. weight-levels fit as well as any more complicated order of curve. The correlation between the body-surface and weight is strikingly shown hi this chart. The widest difference for a given weight is found at about the weight of 30 kg., namely, a difference of about 0.13 square meter or a maximum difference of about 12 per cent. So far as rela- tionship between body-weight and body-surface with boys is con- cerned, therefore, it is clear that there are no abnormal fluctuations in the curve at any point throughout youthful life, and if there are marked changes in the configuration of the body, these must involve weight as well as surface. In figure 10 we have plotted the measured body-surface against the height of our boys. Here, although the curve representing the general trend of the relationship is not a straight line, nevertheless it is reason- ably regular throughout the entire age-range studied. The deviations either side of the curve are somewhat greater than those found in the relationship between body-surface and weight, but there is nothing to GROWTH. 65 indicate an especially abnormal disturbance in the relationship between body-surface and height, although obviously there is not so close a correlation between height and body-surface as was found between weight and body-surface. Knowing the general correlation between weight and height, we may therefore infer that if there are gross changes in configuration with boys in this age-range, judged from the adult standpoint, they are not such as to disturb materially the relationship between surface area and weight or surface area and height. So-m. BODY SURFACE REFERRED TO HEIGHT. BOYS. / X X X- X . * y X jr' X ,> ' ^ ^ ^> ^ ^ ^ ^ ,'. X* #. s* *^ 1 ins. 60 70 80 90 1OO 1 1! 20 130 140 1! )0 16 FIG. 10. Relationship ^between body-surface and height with boys. Although from the well-known correlation between height, weight, and age of children of this age-range, one would a priori expect a reasonably close correlation between body-surface and age, we have plotted these factors in figure 11 for our boys. As has already been noted, there is very considerable regularity in the relationship between body-surface and height and body-surface and weight, but, contrary to our expectations, in the relationship between body-surface and age we note rather striking differences, and a straight-line curve will hardly suffice here to indicate the general relationship, particularly at the youngest ages. Beyond the age of 2 years the curve is reason- ably regular in form, but there are wide deviations either side at the higher age-ranges. We thus see that with boys there is a closer corre- lation between body-surface and weight or height than between body- surface and age. As evidence for the rather considerable changes in 66 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. surface area that may actually occur in children of the same age, these points are of interest, although it is clear that age itself is not an accurate index of surface area, any more than it is an accurate index of height or weight. Sq. m. BODY SURFACE REFERRED TO AGE. BOYS. 1.4 1.3 1.2 1.1 10 / . / .-. /. / '/. S" .9 .8 .6 .5 .4 .3 2 .. ^ . " . /. J/'. / /s / . '^ 'V $ t .1 FIG. 11. Relationship between body-surface and age with boys. RELATIONSHIP BETWEEN SURFACE AREA AND BODY-WEIGHT, HEIGHT, AND AGE WITH GIRLS. In figures 12, 13, and 14 we have plotted the relationships between body-surface and body-weight, body-surface and height, and body- surface and age for our girl subjects. The picture in all three instances is strikingly similar to that found for boys. The relationship between body-surface and weight is closely approximated by a straight line, although a slightly smoothed curve, as sketched on the chart, fits the situation somewhat better. No pronounced bend in this line is to be noted at any particular weight. The body-surface is not so closely correlated to height as to weight, as is seen by the somewhat wider scattering of the points about the curve sketched on figure 13. Thus, while girls of the same weight have essentially similar surface areas, girls of the same height may vary rather considerably in surface area. Perhaps the widest difference is noted with the two girls having heights of 85.5 cm. and 86.5 cm. Here the difference in height is but 1 cm., GROWTH. 67 and yet there is a difference of almost 0.2 sq. m. in surface area. Finally, in the relationship between surface area and age with girls, as exhibited in the scatter diagram in figure 14, we note very wide Sq.m 1.3 BODY SURFACE REFERRED TO WEIGHT. 2kgs.4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 FIG. 12. Relationship between body-surface and body-weight with girls. - BODY SURFACE REFERRED TO HEIGHT. GIRLS. . S / / . y? 2 / ^ '/ . ^ ^ <<* , ^ *f . vv" <^/ ' * >* ?" x Cms. 50 60 70 80 90 100 110 120 130 140 FIG. 13. Relationship between body-surface and height with girls. deviations from the general trend, quite similar to those found with boys, indicating again that the relationship between age and body- surface is by no means a close one. This irregularity in relationship 68 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. is in large part due, we believe, to the fact that children of different ages may have markedly different weights, while children of the same weight will have essentially the same surface area, and children of the same height nearly the same surface area. At no point in figures 12, 13, or 14 is there any noticeable deviation in the line that would indicate a profound disturbance In the relationship of body-surface to either height, weight, or age. Sq. m. BODY SURFACE REFERRED TO AGE. 1.2 1.1 1.0 .9 x / ""' /S' . '/>. X . .7 .6 .5 .4 3 / / . V-"- S ' *^x . / , -2 r .1 Yrs 1234 5 6 7 8 9 10 11 12 13 14 FIG. 14. Relationship between body-surface and age with girls. Although a further detailed analysis is out of place here, it is im- portant that these surface areas, as actually measured, should be permanently recorded. From these charts the reader may readily pick out the individuals deviating farthest from the general line show- ing the relationships between body-surface and age, weight, or height, and compare them with our list of subjects in tables 12, 13, 27, and 28, where all the anthropometric data are given. It was impossible to indicate the subjects' numbers on these charts without making the charts altogether too confusing. Owing to the earlier concept of an intimate causal relationship between body-surface and metabolism, it is important to use these carefully measured surface areas not only in reference to body-weight, height, and age, but likewise for comparison with the measured meta- bolism. This latter comparison is accorded treatment in the subse- quent chapters of this report. NORMAL, AVERAGE, AND IDEAL STATES OF NUTRITION. 69 NORMAL, AVERAGE, AND IDEAL STATES OF NUTRITION. It has long been recognized that body-weight alone referred to age may not be considered an ideal indication of the normality of the child's state of nutrition. Many investigators, in attempting to secure an index of nutritional state, have considered height as well as weight, both in reference to age, and others have added the girth of some part of the body or some power of the length. One of the most recent formulas is that suggested by Van der Loo, 1 who states that "as children grow taller they increase more proportionately in weight than in length, so that the weight divided by the square of the length gives a fairly good index for comparison of conditions in different children or in the same child at different tunes." For practical purposes it would be almost impossible to make direct Du Bois meas- urements and establish the relationship between surface area and any of the other physical factors as an index of normality of nutrition, although we have given tables (tables 14 and 15) whereby the surface area can be very closely approximated by computation based upon several constants. Normal height. From our earlier discussion it seems quite clear that we must consider the normal condition of the child from several bases. In the first place, what is the normal height of the child? With adults there is no such thing as normal height. The ranges of height with men and with women are very extended indeed. We form definite opinions as to whether a man is especially short or especially tall, but no one would care to state the normal height for man. The average height for a man is commonly given as 170 cm., but it is granted that there are very wide variations from this average height. With adults, then, the difference between normal and average is clearly recognized, but the average is taken as normal. Thus, the heights of a large group of individuals representing the general run of the population are measured and averaged, and this average value is considered as the normal value. We contend that this procedure is entirely erroneous when applied to children, although it is regularly employed and is a basis of most of the tables and charts in current use. With children there should be, or at least there is properly supposed to be, a reasonably definite height for an age. It has been stated earlier in our discussion that age, weight, and height are rather closely correlated with children, in contradistinction to the situation with adults. As children grow older they increase in height and weight approximately in the same degree. It has been repeatedly shown that racial characteristics appear very prominently in height. Our Ameri- can population is by no means of a pure strain, and a group of children 1 Van der Loo, Nederl. Tijdschr. v. Geneeak., Amst., 1919, 1, p. 447; cited in Journ. Am. Mod. Assoc., 1919, 72, p. 1403. 70 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. from a large public school, especially in our big cities, may represent a large percentage of foreign blood. Under these conditions an average figure may certainly be obtained, but we believe it is not justifiable to consider this average figure as normal. From the consideration of the charts in which our private-school data were plotted (figs. 4 and 6), it is clear that the private-school children on the whole are considerably taller for their age than are our laboratory children, or, indeed, the extensive series we quote from other writers. In other words, it seems evident that the conditions obtaining with the children of private schools in eastern Massachusetts produce a greater skeletal growth, as indicated by the height. On this basis, all of these private-school children are supernormal; in other words, they are certainly above the average, and the question immediately arises, "Is the average to be considered as normal? " If one objects to the values found with the private-school children as normal values, no exception can be taken to the expression " ideal value." Hence one should compare the height and weight of a child not with the average or the fictitious normal, but with the ideal, which is unquestionably represented more nearly by our data from private schools. We wish here especially to emphasize the difference between average, normal, and ideal. When a child is short for his age this instantly indicates one of two things. In the first place, the child may be the offspring of a race of people or of parents of normally short stature; secondly, there may be a serious deficiency in the growth-producing factors in the diet. This deficiency in growth- producing factors is to be sharply distinguished from the caloric con- tent of the diet, for it has been shown, with animals at least, that when they are maintained upon a diet of constant caloric value during the active period of growth, skeletal growth is made at the expense of the addition of tissue. We should no longer, then, compare the height of our children to the average and call this normal. The fact that a group of 800 private- school children may attain a height for age considerably above that of the average or so-called "normal" can be taken only as an index of the fact that this average represents children living under conditions which do not produce the best growth. In any educational campaign for the promotion of child welfare it is important to lay special stress upon those conditions favoring the largest skeletal growth. Conse- quently we believe that all previous charts indicating height for age are not ideal and represent simply a group of the population that has been stunted, in part at least, by abnormal living conditions and perhaps deficient dietary constituents. In laying down this thesis we are, of course, open to the criticisms that our private-school children were less contaminated by racial commingling and that the shorter statured people did not send their children to these private schools NORMAL, AVERAGE, AND IDEAL STATES OF NUTRITION. 71 in other words, that our private-school children represent the more purely typical American or Anglo-Saxon type. To a certain extent this is probably true, but we are not in a position to throw definite light upon this subject. We think it highly improbable, however, that this explanation completely accounts for the greater height of this group of children. For an estimate of the ideal height of children we believe, therefore, that one should rely not upon the so-called "normal" curve, but more nearly upon an ideal curve which is measur- ably higher than a normal or average commonly given. On this basis many analyses of the measurements of children which indicate that the children are above normal height simply mean that the normal level is arbitrarily adjusted at too low a point. Nutrition experts and pediatricians must hold this important relationship clearly in mind and not be content with the statement that a child is of average height when the possibilities of greater skeletal growth are presented by better living conditions, medical treatment, and general care. Normal body-weight. The emphasis laid upon the relationship between height and age is far overshadowed by that laid upon the relationship between body-weight and age. A child of a certain age is commonly supposed to have a certain weight, and if below this weight is considered an underweight child. As in the case of height, average weight is invariably taken as the normal weight. Normal tables or average tables have been prepared, and almost every writer combines several of the earlier series and obtains his own individual normal which he uses for his study. The very fact that this divergence and uncertainty exist in the minds of all students of the physiology of childhood shows that there has been an unwritten hesitation to accept as normal many of these values. We believe that with children certainly we should no longer consider the average as normal. With adults there are a large number of overweight individuals to compensate for the number of underweight individuals, so that the average value for body-weight represents a median line with approximately the same proportion of overweights as underweights. With children the situa- tion is quite the reverse. The number of overweight children, even using the erroneous term " normal" when applied to the average, are much fewer than the number of underweights. On standing in front of any of our public schools and noting the condition of the children running out at the end of a day's session, one may see at a glance that the obviously overweight children are very few indeed, while those who are obviously underweight usually pass by more rapidly than they can be counted. On this ground, therefore, to take an average value for children seems wholly erroneous. If a child is seemingly underweight for a given age, this may be due in part to his short stature possibly a racial characteristic 72 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. or may be due to a deficiency in the growth-promoting factors in the diet. In other words, underweight may be simply a concurrent factor with short stature, or, if the height is up to the average and the child is still noticeably underweight, this condition may be due distinctly to an insufficient caloric intake. This latter is the more probable and more common situation. If we refer again to our data for private- school children, we will recall that at all ages they were measurably heavier for their age than were the other normal series that we have reported, both our own laboratory series and the earlier standard series. As we pointed out, however, their greater weight is in large part due to their greater height. Still, the fact that outdoor environment, better medical attention, and probably better dietetic conditions have produced a larger and better conditioned child than the ordinary, especially in our public schools, is a factor that must not be overlooked. So-called "normal weight" is not normal, but is merely average. We believe that our ideal figures, as represented hi our curves for private-school children, more truly represent the normal and that pediatricians should strive for the higher weight for age as exhibited by our private-school children rather than for the average weight for age, although here again we clearly recognize the differences in nation- ality hi mixed groups, such as those being studied in any of the public schools, and the probably purer strain of nationality in our private schools. BODY-WEIGHT IN RELATION TO HEIGHT AS AN INDEX OF STATE OF NUTRITION. From the foregoing discussion it is obvious that an index of the state of nutrition based on the relationships of height to age and weight to age is subject to very considerable error, because although a child may be of normally short stature with an accompanying small body- weight, due to racial characteristics, on the basis of age he would be considered to be both underheight and underweight. If the short stature is due to racial characteristics and not to deficiency in the growth-promoting factors in the diet, the child may still be considered normal, indeed may be considered ideal. Before this condition can be established, however, a far greater study of the height-weight ratio of children of normally short parents should be made, and in con- sidering the average mixed population of American schools the element of racial characteristics must not be overlooked. Having shown that neither an average height for age nor an average weight for age is best suited for an index of nutritional state, since the height may be accompanied by varying weights and vice versa, it is clear that as an index of the best proportional distribution of flesh to skeleton the relationship between height and weight is most satisfactory. For a child of a given height a definite weight is pro- NORMAL, AVERAGE, AND IDEAL STATES OF NUTRITION. 73 ductive of a fullness of development and addition of flesh that may be termed ideal. When the child has too little flesh it is very obvious, and likewise when it has too much flesh. The problem then arises as to what is the best proportion between weight and height for children. Should children be somewhat light in build or distinctly overweight, as judged by the popular conception of underweight and overweight when applied to children? Referring again to our private-school data, we find that although these children are heavier and taller than other series of normal children at the same age, when the height and weight are compared they are on the whole somewhat thinner for their height than are our normal laboratory children selected for this study. On the basis only of weight referred to height, therefore, it would appear as if our laboratory children had somewhat the advantage over the group of private-school children, i. e., so far as proportion is concerned. It still remains a fact, however, that had our laboratory children been given the advantages of private-school children, namely, outdoor life, better medical care, operative treat- ment if needed, and better diet, particularly with regard to growth- promoting factors, the skeletal growth would probably have been greater than actually noted. The question is a serious one, then, as to whether we should con- sider a child of a certain age who has a large proportion of flesh for his height a better nourished child than one of the same age who is taller and at the same time heavier, but in whom the proportion between weight and height is not so great as with the shorter child. This question leads us to a consideration of the importance of the diet factors which play a role in growth. No one would seek for abnormal rapidity in the growth of children. In the normal development of the child growth proceeds with a considerable degree of regularity and, on the average, at a certain rate of rapidity. When children, however, are subjected to ideal outdoor life, with plenty of food and excellent medical care, they do grow in skeletal form, at least, as well as in total weight at a somewhat greater rate than otherwise. Is this desirable or not? Everything points to the desirability of this condi- tion, and yet on close analysis it is seen that these private-school children do not have the proportion of weight to height found with the group of laboratory children selected for our measurements. Which, therefore, of the two factors is the most important in the process of growth, height or weight? The striking difference between the private-school children and our laboratory children is the greater height and correspondingly greater weight of the former, although the weight is in all probability simply a natural concomitant of the height. The fact that the obviously ideal conditions of private-school life result in this increased growth would seem to be prima fade evidence of its desirability. On the other hand, we must consider for a moment 74 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. the relationship between weight and height which has resulted, with our laboratory children at least, in a better proportionment that is, these children are somewhat heavier for a given height than are the private-school children. The underweight child is a great care to nutritional experts, and so the greatest stress is laid upon the question of underweight, and appar- ently little, if any, attention is given to underheight. We have pointed out that underheight may be due to erroneous dietary condi- tions, although in many instances such conditions are perhaps entirely unsuspected. But the chief attention of all dietitians and pedi- atricians is given to the underweight of the child; hence, the stress laid upon the larger proportion of weight for height. The desirability of advocating this proportion is well substantiated by the importance ascribed to the relationship between weight and height in the best and recent studies of vital statistics. These statistics show clearly that longevity is better favored in youthful adults, particularly under 30 years of age, if there is a certain degree of overweight; that is, that those youths over the average weight usually have a somewhat better expectancy of life. Beyond the age of 35 years statistics show that a weight somewhat under the average insures a better life expectancy. If during the period of early adult age, longevity is favored by having the weight somewhat above the average, it seems a reasonable con- clusion that this same condition must be advantageous for children. Consequently we believe that during the entire period of growth the weight should, if possible, be somewhat over the average and should approach the ideal as indicated by the weight for ages of our private- school children. Indeed, it seems logical to assume that if the private- school children had been supplied with a larger amount of food, so that they could have put on more flesh and had a proportion of weight to height more nearly in accord with that found with our laboratory children, they would have presented an even more ideal picture. Apparently they were slightly underweight for their height, while our laboratory children, selected from by no means as good an environ- ment, showed a somewhat better proportion of weight to height better when judged on the basis that excess weight is advantageous during the period of growth. For these reasons we believe that all curves which represent a so-called normal, either for height or for weight, are drawn at too low a level, and instead of using the average for normal, as is commonly done, a value perceptibly higher than the average should be striven for in establishing any standards to represent the ideal rates of growth in height and weight for the various ages. To establish the normality of our laboratory children, then, we have the following proofs. These laboratory children represent a some- what better proportion of weight to height than the private-school children, represent a relationship of height to age and weight to age PULSE-RATE. 75 not quite so good as the children in private schools, but better than many of the earlier standards, and consequently may legitimately be regarded as of a degree of normality to satisfy present-day criteria. PULSE-RATE. One of the most striking indices of apparent changes in metabolic activity, induced either by muscular activity or by febrile conditions, is the pulse-rate. In our earlier treatment of the physiology of normal infants, 1 we laid special emphasis upon the importance of knowing the fluctuations in the activity as exhibited by the kymograph record of the movements of the crib and particularly upon the relationships between this curve for activity and both the pulse-rate and the metab- olism. Before the study of new-born infants, our observations on children were so scattered and represented so few normal subjects that we were unable to record normal pulse-rates for children of various ages. With the new-born infants, however, this was perfectly feasible, and in the report of that study, 2 data were recorded giving the average pulse-rate for the first 8 days after birth as 112 on the first day, and for the 7 subsequent days 114, 116, 116, 116, 122, 119, and 126, re- spectively. These average values were obtained from a considerable number of counts for different children. Those for the first day after birth represented 50 new-born infants, but on the later days the num- ber of subjects was less, particularly on the seventh and eighth days. In our report of the observations on the few normal subjects, made in the first study of the gaseous metabolism of infants, 1 we were primarily interested in such alterations in the pulse-rate of an indi- vidual infant as were due to changes in activity and not in the altera- tions due to changes in age. Accordingly, in this earlier study the period of observation did not exceed 30 to 45 days, except with a single infant. In the accumulation of our new data, however, special stress was laid upon the trend of the pulse-rate as the age increased. This could be studied advantageously in those series of observations in which the metabolism of the same child was studied over periods of 4 months or more, and in a few cases 3| years. Finally, with the older children, the unusually advantageous conditions under which the data were obtained make it seem desirable for us to record the pulse-rates and deduce therefrom average values which might be expected from children under quiet conditions. Even the earliest observers noted that the pulse-rate of infants was very difficult to obtain and varied under different circumstances. The great difficulties in securing accurate records can perhaps be no 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. 2 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915, table 19, p. 115. 76 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. better expressed than by quoting from the earliest report that we have found in English of observations on the pulse-rate of children. Publishing in 1694, Walter Harris of London stated: "But the Pulses of Children are naturally, or upon every little Alteration, do become so swift and frequent, that they always seem somewhat Feverish. Moreover, they are for the most part, so chagreen and froward, that not keeping their Wrest one moment in the same posture, do not suffer their Pulse to be touched. Lastly, there are so many things that do accelerate or otherways change their Pulses, that Sentiments taken thence should prove very uncertain, if not altogether false." 1 In the interesting book of Benjamin Waterhouse 2 we find a quo- tation from a paper read in 1768 at the Royal College of Physicians in London by the venerable Dr. Heberden: "The pulse of children under two years old should be felt when they are asleep; for their pulses are greatly quickened by every new sensation, and the occasions of these are perpetually happening to them while they are awake. The pulse then qf a healthy infant asleep on the day of its birth, is between 130 and 140 in one minute; and the mean rate for the first month is 120, for, during this time, the artery often beats as frequently as it does the first day, and I have never found it beat slower than 108. During the first year the limits may be fixed at 108 and 120. For the second year at 90 and 108. For the third year at 80 and 100. The same will very nearly serve for the fourth, fifth, and sixth years. In the seventh year the pulsations will be sometimes so few as 72, though generally more; and therefore, except only that they are more easily quickened by illness, or any other cause, they will differ but little from the healthy pulse of an adult, the range of which is from a little below 60 to a little above 80. It must be remembered, that the pulse becomes more frequent, by ten or twelve in a minute after a full meal." No further evidence as to the difficulties of making these physio- logical records is necessary. Perhaps the best confirmation of this evidence is the fact that so little is now known regarding the quiet resting pulse of children. On looking over the literature on the normal pulse-rate of children, it is at once obvious that very little interest has been taken in the subject and few accurate counts have been made which take into consideration all the factors which modify the normal rate of the heart. It has been the custom of practically all writers to report minimum and maximum pulse-rates and to follow what seems to us the very confusing and entirely irrational procedure of averaging these and reporting the result as the average pulse-rate. A summation of our data shows that any pulse-rate above the minimum is profoundly affected by the degree of activity; therefore, little, if any, value can be placed upon observations of pulse-rate other than those obtained with the child in repose. It is of importance to know 1 Harris, An exact enquiry into, and cure of the acute diseases of infants. London, 1694, p. 9. * Waterhouse, An essay concerning tussis convulsiva or whooping cough, with observations on the diseases of children. Boston, 1822. PULSE-KATE. 77 to what extent maximum pulse may develop during paroxysms of crying and with such activity as a child may exhibit when lying in bed, but for all normal purposes such records have but little, if any, scientific value. The younger the infant the greater is the difficulty of obtaining the pulse-rate. With older children the element of apprehension should not be entirely disregarded. If this apprehension is not present to any great degree, the special precautions necessary for small children will not be required for the older individuals. Our measurements were all made while the child was inside a hermetically sealed chamber, and the routine was invariable for all children studied. METHODS OF OBTAINING PULSE-RATE. The pulse-rate can be obtained either by palpation at the wrist or by direct auscultation from the heart. During infancy it is difficult to obtain an accurate pulse-rate from the wrist, because infants rarely remain quiet for more than a few seconds. Infants dislike to be forced to stay in one position, and when made to do so they usually struggle and cry. It is often impossible to get the pulse-rate from the wrist while the infant is asleep, for the slightest touch of the observer's hand wakens the child. The difficulties present during the waking hours are then accentuated by the fright which may result from the sudden awakening. Obtaining the pulse-rate by means of a stetho- scope held over the heart is also attended by many difficulties. Unless an infant is phlegmatic or becomes used to this procedure, he may squirm and cry and sometimes violently resist the application of the stethoscope. When a child or an infant resists, it is obviously im- possible for the observer both to hold the stethoscope on the chest- wall and to make an accurate record of the pulse-rate unless someone holds the infant. The most successful method of obtaining an accurate pulse-rate is by means of a small Bowles stethoscope fastened with adhesive plaster to the body-surface of the infant over the heart. A long rubber tube is run from the stethoscope under the clothing and out to the earpieces. The child can then take any position he desires without feeling that he is restricted and without realizing that the stethoscope has been applied. All of our own pulse-counts were obtained by this method. In full recognition of the difficulties attending a study of the pulse- rate, we have tabulated the results of our observations, taking into consideration only the minimum pulse-rate. Our studies contribute towards the solution of two main physiological problems: First, how does the pulse-rate vary with age; secondly, what is the average pulse-rate for children of various ages? Since the number of observa- tions made was considerable, the results offer a fair basis for answering these questions. 78 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. INFLUENCE OF AGE UPON THE PULSE-RATE. The pulse-rate at the end of fetal life is said to average between 135 and 140 per minute, but it is only with extra-uterine pulse-rate that our observations deal. In the study of new-born infants we were able in no case to make observations of the pulse-rate of the same child throughout the entire first week. The average pulse-rates for the first seven or eight days of life given in the report of that study must consequently be looked upon as showing only the general trend of the average pulse-rates of new-born children. In our first report we tabulated all of the available data for the pulse^rate for each of the first 8 days after birth, regardless of whether the measurement of the metabolism during these periods gave strictly minimum results or not. 1 Our conclusions were based upon these pulse data, but it was definitely stated that the values for the pulse- rate were those of infants during periods of approximately minimum heat production and that the metabolism during these periods could be considered as absolutely minimum in only a relatively few cases. In this publication it seems best to present the data for the pulse-rate of new-born infants for comparison purposes, using as a basis of selection, however, the periods in which the minimum heat production was obtained. The pulse-rates for these periods of minimum heat production are given in table 16. A comparison of the figures for TABLE 16. Pulse-rate during first 8 days after birth. Age. No. of subjects. Average pulse-rate. First day 50 112 Second day 25 110 Third day 19 109 Fourth day 20 116 Fifth day 14 113 Sixth day Seventh day . 7 4 118 114 Eighth day. . . . 95 pulse-rate given in the earlier report 2 with those in table 16 shows that this later method of selection lowered the daily averages on all but the first and fourth days, there being no change on these days; that is, when the effort is made to tabulate only pulse-rates with minimum heat production, in most cases this approximation to the minimum heat production is accompanied by a distinct reduction in the average pulse- rate. Since all of the subsequent material on pulse-rate in this report is based upon the values found during minimum muscular activity, for the final summation of pulse-rate values for children the averages given in table 16 will be used rather than those in the earlier report. 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915, table 19, p. 115. 'Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915. p. 116. PULSE-RATE. 79 TABLE 17. Basal pulse-rate of boys at different ages. No. of subject. 1 mo. H mo. 2 mo. 2J mo. 3 mo. 4 mo. 5 mo. 6 mo. 7 mo. 8 mo. 9 mo. 10 mo. 11 mo. 61 136 115 118 124 94 126 123 130 120 117 130 138 130 128 112 121 117 118 109 127 122 121 119 104 119 122 121 124 120 125 115 120 111 125.. 126.. 128.. 129.. 130.. 132.. 133 111 110 116 109 113 123 113 107 iis 118 134 114 115 119 136.. 137.. 138 129 116 119 114 132 133 124 129 116 141 142 122 115 119 108 103 107 117 101 109 110 117 104 116 147. 148. 149. 150. 153. 154.. . . 112 112 111 iio 127 115 126 120 iii 112 iis iis iis ii7 155.. . . 156.. . . 157 158 123 119 115 95 159 161 123 145 108 164.. 168.. 170.. 118 109 109 No. of sub- ject. 12 mo. 14 mo. 16 mo. 18 mo. 20 mo. 22 mo. 24 mo. 26 mo. 28 mo. 30 mo. 32 mo. 34 mo. 36 mo. 38 mo. 119.. 138. . 148.. 153. . 155. . 158.. 161 112 118 125 110 111 108 106 106 92 127 104 94 96 103 107 107 101 100 110 107 93 86 93 96 83 98 83 "75" 135 123 126 112 116 107 84 87 175 94 85 176 For the first 5 days after birth the number of infants is sufficiently large to justify a study of the general trend of the pulse-rate during this time. On the last 3 days the number is obviously too few, but the results are included here to complete the data. From this table it can be seen that there is a slight tendency for the pulse-rate to rise and remain at the higher level after the first 3 days, with a tendency to decrease successively on the first 3 days. All of these values are measurably less than that reported for the last hours of fetal life. 80 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 18. Basal puke-rate of girls at different ages. No. of subject. 9th day. llth day. 12th day. 2 wks. 3 wks. 1 mo. li mo. 2 mo. 2| mo. 3 mo. 4 mo. 5 mo. 6 mo. 2 94 97 95 112 35 134 125 48 127 159 139 49 115 126 110 118 125 113 146 134 132 134 120 134 120 116 121 116 123 132 120 123 122 125 138 135 123 130 125 122 121 127 117 131 121 127 121 127 134 106 135 103 110 139 131 126 140 113 118 111 121 145 117 146 121 132 151 126 152 121 No. of subject 7 mo. 8 mo. 9 mo. 10 mo. 11 mo. 12 mo. 14 mo. 16 mo. 18 mo. 20 mo. 22 mo. 24 mo. 122 138 132 123 113 115 127 125 115 113 114 109 100 131 120 122 139 128 128 123 133 117 122 110 103 145 117 110 108 109 98 103 88 83 146 122 160 123 126 125 115 116 162 119 163 116 165 101 166 111 128 126 117 105 100 106 88 167 127 171 132 165 136 119 123 116 114 172 123 124 118 119 106 173 121 120 111 103 174 104 No. of subject 26 mo. 28 mo. 30 mo. 32 mo. 34 mo. 36 mo. 38 mo. 40 mo. 42 mo. 44 mo. 46 mo. 4 yr. 139 110 100 93 91 80 82 85 145 73 70 76 74 71 70 166 98 86 96 90 90 76 171 111 101 91 90 178 88 180 91 181 88 183 78 184 93 In the present report we are more particularly interested in the pulse-rates of children over a week old. To eliminate completely the PULSE-RATE. 81 possibilities of sexual differentiation, a subject which will be discussed somewhat later in this report, we have tabulated the values for boys and girls separately, those for boys appearing in table 17 and those for girls in table 18. Unlike the average pulse-rates for new-born infants shown in table 16, the pulse data presented in tables 17 and 18 do not represent the pulse-rates obtained in the periods of minimum heat production. The basis of selection here used was the degree of muscu- lar repose, the pulse-rates tabulated being those obtained in periods of minimum activity, as indicated by the tracings on the kymograph, without reference to the heat production. The pulse-rates of boys and girls as presented in these tables offer an opportunity for studying the basal pulse-rate of the same child at varying ages. This is par- ticularly true of boys up to the age of 38 months and of girls up to 4 years. One of the longest series for boys is that with No. 119. The irregu- larities in the figures reported for this subject show clearly that there is very little evidence of a definitely established trend until the child is somewhat over a year old, but that during the second year there is a distinct tendency for the pulse-rate to decrease. Another long series is represented by No. 158. Here again considerable irregularity is noted previous to the age of 14 months ; thereafter there is a reason- ably uniform decrease. These irregularities during the first 14 months are to be observed with practically all of the boys. During the second year of life, however, the values indicate a definite tendency toward a generally lowering pulse-rate. This finding with boys is likewise noticeable with girls, although the irregularity in the first year is by no means so striking. Indeed, with a number of subjects there is a reasonable degree of decrease in the pulse-rate subsequent to the age of 4 or 5 months. It is characteristic of all these children, however, that after the first year and a half the decrease is reasonably well established. AVERAGE PULSE-RATE OF CHILDREN. The second important factor to which our data contribute is the average pulse-rate of children of the same age, both as to its absolute value and as to the deviation therefrom which can be expected for children in repose. With boys, one of the most extensive series numerically that we have for any age represented in table 17 is that for the group 7 months old, in which the basal pulse-rate for the 13 boys ranged from 107 (No. 126) to 127 (No. 154). The average for this group is found to be 117. It is clear, therefore, that it is only with a great deal of reserve that one may speak of the pulse-rate of a boy of 7 months as being 117. Even wider differences are observed with the 4 boys of 20 months, the lowest being 92 and the highest 127, with an average of 111. 82 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. From the foregoing discussion of the considerable variation to be found with an individual in the earlier years, one would naturally expect similar large differences between groups of older children of the same age. The considerable amount of data obtained for the first four years of a child's life have been supplemented by records for older children, which are given in table 19 for boys and girls 5 to 13 years TABLE 19. Basal pulse-rate of individual children 5 to 13 years of age. B( >ys. Gi rls. 6 yrg. 7 yrs. 8 yrs. 9 yrs. 10 yrs. 11 yrs. 12 yrs. 13 yrs. 5 yrs. 7 yrs. 8 yrs. 9 yrs. 10 yrs. 12 yrs. 90 72 86 99 74 84 83 85 79 79 94 84 71 73 86 78 80 81 89 89 82 76 69 76 67 68 67 66 65 74 68 71 80 114 79 82 90 76 74 71 76 86 69 83 97 83 85 88 84 85 69 80 74 76 78 81 78 79 71 77 72 67 69 of age. Almost without exception, however, the averages in this table represent individual children, the data for each child, except in one case, being obtained at only one age. Here again we find con- siderable differences. Thus, for the 7 boys 7 years of age, the lowest is 74 and the highest 99. With the 5 boys 9 years of age, the lowest record is 71 and the highest 86, while the 9 boys 11 years of age show a maximum difference in their basal pulse-rates of 22 beats. With girls, one of the largest groups at the earlier ages is that for 6 months, but the pulse-rates for these 8 girls range only from 118 to 132, with an average of 124. This approach to uniformity, which is much closer than that noted for boys, does not by any means hold for all ages, since at 16 and 18 months differences amounting to 40 and 32 beats, respectively, are observed. The largest group for the older ages represented in table 19, that for 9 years, has values for pulse-rate for the 6 girls ranging from 71 to 97. The extraordinarily high value of 114 for one of the 5-year-old girls may have some special explanation which is at present unknown to us. Bearing in mind the irregularities seen in the careful examination of the data for these several age-groups, we may average these pulse data and attempt to portray the general trend of the minimum or basal pulse-rate of boys and girls from birth to 13 years of age. In so doing we have left out of the averaging all age-ranges represented by less than three individuals. These values are brought together in table 20. The average values shown for boys indicate a reasonably constant pulse-rate for the first 14 months of life, ranging from 113 to 125, if PULSE-RATE. 83 TABLE 20. Comparison of average minimum pulse-rates of boys and girls. Age. Boys. Girls. Age. Boys. Girls. No. of sub- jects. Average mini- mum pulse- rate. No. of sub- jects. Average mini- mum pulse- rate. No. of sub- jects. Average mini- mum pulse- rate. No. of sub- jects. Average mini- mum pulse- rate. Iday . 2 days . 3 days . 4 days . 5 days . 6 days . 11 days. 29 13 8 11 11 113 108 105 117 114 21 12 11 9 3 5 3 3 3 5 6 6 8 8 5 4 5 5 3 7 8 110 112 112 114 111 118 110 132 138 129 129 119 119 124 121 121 116 124 118 124 121 16 mos 8 7 3 4 4 3 3 114 115 105 103 97 94 89 87 84 79 88 87 91 18 mos. 20 mos. 22 mos. 24 mos. 26 mos. 28 mos. 30 mos. 34 mos. 38 mos. 40 mos. 46 mos. 5 4 3 4 107 111 106 94 5 3 98 91 l| mos. . 2 mos. . . 2? mos. . 3 mos. . . 4 mos.. . 5 mos. . . 6 mos. . . 7 mos. . . 8 mos. . . 9 mos. . . 10 mos. . . 11 mos. . . 12 mos. . . 14 mos. . . 5 6 5 7 9 8 11 13 8 7 7 3 3 117 125 117 116 124 119 115 117 117 114 113 122 125 3 3 4 3 3 4 4yrs. . 6yrs. 7yrs. 8yrs. 9yrs. lOyrs. 11 yrs. . 12yrs. . 13 yrs. . 3 7 4 5 3 9 6 4 83 83 80 78 83 75 69 73 3 3 6 4 74 77 85 80 3 76 we exclude the first 5 days, and from 105 to 125 if these earlier values are included. Thereafter the picture is a gradual decrease, persisting throughout the second and much of the third year. The data between 3 and 6 years are lacking. During this period there has been a very considerable fall, the tendency to a decrease continuing subsequent to 7 years. The lowest value is 69 at the age of 12 years. With girls, the averages show that after the first 11 days there is a distinct ten- dency for a rise in pulse-rate, the return to the rate of the first week not taking place until shortly after the end of the first year. There is then a continued decrease, the lowest record being 74 at the age of 7 years. SEX AND MINIMUM PULSE-RATE. The intimate relation between pulse-rate and metabolism, shown with the same individual in the large majority of observations hi this laboratory, makes a careful examination of pulse-rate with respect to sex of special importance. The analysis of data for the basal metab- olism of men and women has shown that women as a class have a lower metabolism than men, not only per individual, but per unit of weight. On the other hand, the pulse-rate of women as a class was shown to be higher than that of men, as measurements for 90 women and 121 men in our series gave an average of 68.67 for women and 84 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. 61.26 for men. 1 These differences are substantiated by the fact that with men three different groups of 28, 116, and 50 men showed average pulse-rates of 62.5, 61.3, and 61.3, respectively. Two groups of women, one of 68 and the other of 22, showed pulse-rates of 69.1 and 67.3, respectively. In other words, it seems thoroughly established that the women as a class have a pulse-rate somewhat higher than men, in spite of the fact that their metabolism is distinctly lower. This sup- plies very clear evidence that while pulse-rate and heat production may be closely correlated in the same individual, in groups of indi- viduals the pulse-rate may vary enormously and " average" pulse-rate may have little, if any, connection with " average" heat production. Since the pulse-rates of men and women show a difference, it becomes extremely important, in studying our groups of children, to note at what point, if any, there is a definite change in the pulse-rate, and further comparisons of values for males and females will be of special interest in this connection. Such comparison may be made from the pulse-rate data for boys and girls in table 20, which gives an oppor- tunity of noting the differentiation, if any, due to sex. To this end, wherever values for both boys and girls are recorded at the same age, the higher of the two values has been italicized. Thus, for children 1 day old, 29 boys gave an average minimum pulse-rate of 113, while 21 girls had a pulse-rate of 110. On the next day the conditions for very nearly the same number of boys as girls were reversed, the girls showing a pulse-rate 4 beats higher than the boys. Pursuing this method of analysis for the entire group of data in table 20, and passing over those ages for which records are available for only one of the two sexes, we find that at 11 age-periods the boys have a higher pulse-rate than the girls of the same age, while at 15 age-periods the girls have a higher pulse-rate than boys of like age. On this basis, therefore, it would appear that the pulse-rate of the girls was, on the whole, somewhat higher than that of boys. The italicized figures in the table show no great regularity in the appear- ance of these high values with either sex. The most consistent record is that from 1| months to 10 months, the only ages at which the girls are not higher being that of 4 months, and again of 5 months, when the average pulse-rate for both sexes is the same. After 10 months the italicized figures indicate but little regularity as to sex. On the whole, the picture can not be said to speak pronouncedly for a higher pulse-rate with girls than with boys. In making this general conclusion, however, it is important to note that the data under con- sideration are at best somewhat meager, although they may be relied upon as far as they go. But many observations of the minimum resting pulse-rate of boys and girls are necessary before final conclusions 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 66. RECTAL TEMPERATURE. 85 can be drawn, for obviously the pulse-rate at other ages should be studied and supplementary data should be obtained for the higher ages included in our observations. AVERAGE PULSE-RATE WITH INCREASING AGE. From the foregoing analysis it can be seen that there is no striking difference between the pulse-rates of girls and boys. Accordingly, as a tentative measure of the average minimum resting pulse-rate of children of both sexes, we have taken the values given in table 21 as approximately normal values. In view of the special conditions under which these pulse-rates were obtained, namely, complete muscu- lar repose and with the subject entirely unconscious of the records, it is seen that we have here true physiological values uncontaminated by activity. TABLE 21. Approximate normal minimum values for pulse-rates of children during complete muscular repose. End of year. Av. min. pulse-rate. End of year. Av. min. pulse-rate. First Second 122 100 Seventh Eighth 78 78 Third 89 Ninth 82 Fourth. . 87 Tenth 81 Fifth .... 91 Twelfth 72 RECTAL TEMPERATURE. Any physiological study of body temperature, to be of true scientific value, must deal with* temperatures taken deep in the body trunk. The extraneous factors of exercise, mouth-breathing, and the effect of previously taken foods so greatly vitiate all measurements of the buccal temperature that they have little, if any, value except for demon- strating the absence of fever. It has been the custom of many clini- cians to take the temperature of young children in the axilla or in the groin. These records, aside from likewise showing the presence or absence of fever, have no physiological value, for it has been found that even when these cavities are well-closed a very considerable period of time is required to raise their temperature to that approximating the interior of the body. In our study of certain physiological factors during growth, the rectal temperature was measured primarily to demonstrate the absence of any febrile condition, since the presence of fever would of course preclude further observations with the child in this abnormal state. Every reasonable effort was made to have these measurements meet the exactions of scientific accuracy, but as ordinary clinical mercury thermometers were used, even though well-tested, and there 86 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. were changes in the personnel of the assistants from time to time, as well as possibilities of difference in the depth of insertion of the thermometer bulb, we may not look upon these observations as a refined physiological study of changes in body temperature of children. It is to be regretted that a study of the diurnal rhythm of the rectal temperature of children of various ages could not have been included in the research, but this was impracticable. To use occasional obser- vations of rectal temperature as a basis for determining the physio- logical temperature of children is open to serious question, since the well-known influences of activity, ingestion of food, and sleep make such measurements liable to gross variability. When one considers that the normal range in temperature is but 2 or 3 F., it will be seen that if a study of the body temperature is to be made with strict scientific accuracy, a much more sensitive measurement should be used than that employed in this research. But as the measurements were all made with the child inside a chamber, lying quietly on a comfortable bed, we believe that although the method was admittedly defective, the conditions were essentially comparable and the values may legitimately be used for drawing conclusions. The rectal temperature was usually recorded, at least with the younger children, just before and just after each respiration experiment. Several hundred observations of the body temperature of these children were thus obtained. It appeared unnecessary to give detailed publi- cation of all these records, and we have therefore averaged the two observations for each experiment and grouped them in table 22, accord- ing to sex, as general averages for certain average ages. In these averages the relatively few records of 100.5 F. or over have not been included. The number of observations entering into each average is also shown hi the table. If measurements were made with but a single subject for any particular age, the values were not included in this comparison. From these average values an indication of the general trend of the body temperature of children during the period of growth may be obtained. The values for boys range from 97.1 F., the very low value for the two boys averaging 5 years of age, to a maximum of 99.5 F. in the group of six boys for If years. During the first month or two of life the temperatures for boys are somewhat low, but with evidence of a tendency to rise thereafter, the maximum continuing for approximately two or three years. Subsequently the figures incline to run below rather than above 99 F., and after 10 years all values are 98.8 F. or under. Special attention has already been called to the extra- ordinarily low value of 97.1 F. for the two boys 5 years of age, which must not be looked upon as characteristic of that age. This table must not be considered as indicating the minimum tem- peratures at these various ages. According to previous experimenting RECTAL TEMPERATURE. TABLE 22. Average rectal temperature of boys and girls ttf different ages. 87 Boys. Girls. No. of subjects. No. of meas- ure- ments. Average age. Aver- age rectal tem- pera- ture. 1 No. of subjects. No. of meas- ure- ments. Average age. Aver- age rectal tem- pera- ture. 1 44 34 88 68 f hr. to 24 hrs 1 to 2 days > F. 298.2 2 98.7 30 31 60 62 f hr. to 24 hrs 1 to 2 days F. 2 98.3 2 98 8 31 62 2 to 3 days. . 2 98.8 29 58 2 to 3 days 2 99 27 54 3 to 4 days 2 98.5 24 48 2 98 7 22 43 4 to 5 days 2 98 5 18 36 *98 5 9 17 5 to 6 days 2 98 7 13 26 5 to 6 days *98 5 10 20 6 to 7 day 2 98 5 6 12 6 to 7 days *98 6 6 12 7 to 8 days 2 98.8 3 6 7 to 8 days 2 98.7 7 6 22 19 8 to 20 days 98.7 98 7 8 4 30 16 8 days to 3 wks 98.5 98 9 8 50 2 mos 99.0 6 22 2 mos 99.1 8 8 30 36 3 mos 4 mos 99.1 99.0 7 7 27 38 3 mos 4 mos 98.9 98.8 8 11 30 40 5 mos 6 mos 99.1 99.0 9 8 32 35 5 mos 6 mos 98.9 98.9 13 47 7 mos 99.0 6 24 7 mos 98.9 6 20 8 mos 990 4 10 8 mos 98.9 8 32 9 mos. . . 988 8 26 9 mos 99.3 4 18 10 mos. 99 1 5 14 10 mos 99.1 5 18 1 yr. 99 3 5 16 99.5 6 58 1 yrs 99 5 8 89 1 yr. . 99.4 5 23 2 yrs 99 3 8 74 Ij yrs.. . . 99.5 6 26 2| yrs 99.0 6 46 2 yrs 99.4 3 16 3 yrs 99 1 4 30 2j yrs 99.3 2 2 97 1 5 30 3 yrs. 99.2 2 7 4 21 5|yrs 7 yrs 98.4 99.1 5 4 19 13 3^ yrs 4 yrs 99.2 98.9 5 15 8 yrs 98.6 4 7 4| yrs 98.4 3 7 8| yrs 99 2 3 4 51 y re 98.1 2 4 99 1 2 4 65 yrs 98.5 " 3 6 9 yrs. 99.1 3 6 8 yrs 98.8 2 6 10J yrs. 98 6 5 15 9 yrs 98.4 2 10 11 yrs. 98 6 2 4 Qi yrs 98.8 3 10 Ill yrs 98 3 2 5 10| yrs 98.8 2 6 12| yrs 98 7 2 5 11 yrs 98.8 x 3 7 12 yrs 98.6 1 Any reading of 100.5 F. or over was excluded from these averages. 2 Data obtained from table 9, Carnegie Inst. Wash. Pub. No. 233, 1915, pp. 46-79. in this laboratory, the minimum temperature was found some time during the early morning hours, when the subjects were in deep sleep and the stomachs practically empty. As an indication of the influence of age, however, we believe that these values are not without sig- nificance. It is important to call attention here to the fact that the period of highest average rectal temperature, as shown by these more or less heterogeneous data for boys, is not far from 10 months to 2 years. As will be seen later, this corresponds to the period of relatively highest metabolism. While the evidence is by no means beyond criticism, it is of significance at least to point out this coincidence 88 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. between the maximum metabolism and the high average rectal tem- perature. The values obtained with girls are somewhat better distributed for the different ages than those with the boys. These range from a minimum of 98.3 F. for girls on the first day of life, to a maximum of 99.5 F. at 11 months and If years of age. Here again we find a rela- tively low temperature in the first few weeks of life with a subsequent rise and a high period from 11 months to 2 years of age. Thereafter the temperatures with girls begin to fall and subsequent to 3^ years are below 99 F. In this respect, therefore, the picture exhibited by the average temperature values for girls confirms in a striking manner that noted for boys. Here, also, the maximum level in the body tem- perature, i. e., from 11 months to 2 years, corresponds to the maximum level for the metabolism, which will be subsequently noted. No pro- nounced sex differentiation in the values for boys and girls is apparent from these two groups of data. While it seemed inadvisable to publish the individual records of the body temperature, the data have been carefully compared to note the differences, if any, between the records at the beginning and end of the experiment. The influence of muscular activity and, in some cases, that of food obtain in these measurements. For the experiments the child was placed in the chamber and instructed to remain quiet. No observations were made until this quiet condition had been secured. During the experiment of an hour or more, and with conditions of reasonably complete muscular repose and frequently sleep, there would normally be a definite fall in temperature. On the other hand, the experiments were at times continued until the child awoke or became restless; this activity would make it necessary to discontinue the experiment, since the main object was a study of the basal metabolism. Under these conditions, the activity at the end of the experiment would tend to increase the body temperature and the second record would thus be higher. A study of all the experiments in which records were obtained at the beginning and end shows that in the 189 experiments with boys used in the mini mum-metabolism table (see table 27, page 116), there was no rise or fall in 19 experiments, an average rise of 0.55 F. in 52 cases, and an average fall of 0.71 F. in 118 cases. Thus, with the boys, 10 per cent of these 189 experiments showed no change in body temperature, 27.5 per cent showed a rise, and 62.5 per cent a fall, the range in variation being from +0.55 F. to -0.71 F. In the 214 experiments with girls used in the mmimum-metabolism table (see table 28, page 120), there was no change in temperature in 24 experi- ments, an average rise of 0.49 F. hi 51 experiments, and an average fall of 0.71 F. in 139 experiments. In these experiments with girls, therefore, there was no change in 11.2 per cent, a rise in 23.8 per INFLUENCE OF FOOD ON METABOLISM. 89 cent, and a fall in 65 per cent, with a range in average variation from +0.49 F. to -0.71 F. While in this comparison only those experiments were used which were included in the minimum metabolism tables, but little if any difference in the results was found when all the measurements were compared. It may be safely concluded, therefore, that there was a tendency in general for the temperature to fall during the experiments and that this was independent of sex. The body temperatures were usually below 100 F. and, if 100 F. at the beginning, generally fell below at the end. This tendency to a fall was undoubtedly due to the contrast between the quiet and muscular relaxation in the experi- mental period and the pre-experimental activity. This interpretation of the change is made the more probable by the fact that, from an inspection of the records showing a rise at the end of the experiment, it appears that in the majority of cases the increase hi body tempera- ture was accompanied by a considerable degree of activity in the last period of the experiment. In the majority of experiments with children older than 2 years, the temperature measurement was made only at the end of the experi- ment. Accordingly, this comparison of the measurements before and after the experiment applies more particularly to children 2 years old or younger, although some 20 boys and 5 girls are included hi the data compared who were over 2 years of age. In conclusion, it may be said that these observations on the body temperature of young children indicate, on the whole, that during the first two years of life there is a definite tendency for the rectal temperature to increase slightly, with a maximum at about 2 years of age. Thereafter the body temperature slowly falls, with no perceptible general trend apparent from 5 to 13 years of age. These variations hi temperature show strikingly the desirability of more extensive studies of the diurnal range by means of some more sensitive measurement, such as a bolometric or thermo- electric method, with continuous or semi-continuous registration. INFLUENCE OF FOOD ON METABOLISM. As outlined in a previous section (see page 30), with the youngest children it was not possible to obtain ideal conditions for measuring the basal metabolism, i. e., complete muscular repose and with no food in the alimentary tract, or the "post-absorptive" condition. With adults the post-absorptive condition is usually not secured until 12 hours after an ordinary meal and the maximum increase following a protein-rich meal may be as high as 40 to 50 per cent for a short tune, with its effect possibly continued even longer than 12 hours. With the children observed in this study, excessively large meals were not the rule, the last meal before the observations with the respiration apparatus being purposely considerably reduced. 90 METABOLISM AND GKOWTH FROM BIRTH TO PUBERTY. A careful control of the muscular activity was secured through graphic records. With the youngest children, and especially the infants, however, it was not possible to obtain the required degree of muscular repose when there was no food in the stomach, because the want of food caused restlessness and frequently crying. In recognition of this difficulty, it was necessary to compromise by supplying as small an amount of food as would produce comfort and consequent muscular repose. This amount of food, even though small, inevitably influenced the metabolism. When older children were studied, it was possible to postpone the observations after a meal for a longer period of time, even for 4 or 5 hours. Accordingly, in the subsequent analysis of the metabolism data for the children of various ages, it must be remembered that as the age of the children increases the influence of the ingestion of food decreases correspondingly. The basal metabolism of children under 2 years of age can thus be com- pared with that of older children only on the distinct understanding that the values for the basal metabolism for the younger children are higher than they normally would be, owing to the influence of food. The quantitative measurement of the influence of the ingestion of food upon the metabolism of infants should be given special study, such as has been done for adults in a previous publication from this laboratory. 1 Certain more or less fragmentary evidence has, however, been accumulated in the present research with children, in part by design and in part by accident. With several of the children, a pro- longed series of observations was made after food had been taken, some of these continuing 9 or 10 hours without interruption. The results of these observations give some information as to the possible influence of food. Here again we find a difficulty in interpretation in that the somewhat subtle influence of food is profoundly affected by a change in the activity; hence only periods of comparable muscular activity, or preferably muscular repose, can legitimately be used to determine the influence of food upon the metabolism. That is, it must be assured that the increase in the metabolism after food is not due to muscular activity before assuming it is due to the stimulus of food. The data contributing to a study of this problem are brought to- gether in table 23, in the order of increasing age. Unfortunately, long observations during which the minimum metabolism after feeding has been measured were made only with relatively young children, the oldest being but 13| months old. This table gives not only the age and weight of the individual children, but also the energy content of the food, the time of feeding, the interval between the food and the end of the period of observation, the heat produced on the basis of 24 hours, the pulse-rate, and the relative activity. The time when presumably basal metabolism was reached is indicated by an italiciza- tion of the lowest value for heat. 1 Benedict and Carpenter, Carnegie Inst. Wash. Pub. No. 261, 1918. INFLUENCE OF FOOD ON METABOLISM. 91 TABLE 23. Heat production of infants after food. Subject No., age, and weight. Date. En- ergy of food Time of feeding. Inter- val be- tween food and end of period Heat produced | per 24 hours. 1 Pulse-rate. Rela- tive ac- tivity.* 1916 cals. h. m. cals. 127 (2| mos.; 5.03 kilos) Apr. 20-21 60 6 h 48 m to 7 h 07 m p.m. 1 21 287 117 I 1 56 301 119 II 2 32 286 118 I 3 7 287 116 I 3 44 299 112 III(?) 4 20 281 114! II 4 57 255 115 III(?) 5 31 267 115 II Apr. 21 90 1 12 to 1 21 a.m. 1 11 291 121) II 90 6 42 to 6 50 a.m. 1 41 336 117 III(?) 100 9 28 to 9 42 a.m. 1 18 280 115 II(?) 60 12 32 to 12 47 p.m. 40 322 119 III(?) 115 (4| mos.; 5.86 kilos) Mar. 25 80 5 40 to 548 p.m. 1 28 348 137 II 2 10 335 131 I 2 49 328 125 II 3 32 312 125 Mar. 25-26 110 9 54 to 10 35 p.m. 1 9 362 137 1 44 312 128 2 27 342 126 I 3 11 296 124 I 119 (6 mos.; 7.21 kilos) Apr. 15-16 80 7 08 to 7 14 p.m. 1 8 403 107 1 58 397 115 2 39 397 118 3 27 391 115 III(?) 4 41 392 117 Ill 5 18 373 122 II 5 41 369 122 III(?) Apr. 16 210 2 06 to 2 26 a.m. 1 23 422 119 I 1 50 477 129 II 2 35 457 132 H(?) 3 2 437 125 K?) 140 (6 mos.; 6.38 kilos) Apr. 16-17 70 6 24 to 6 38 p.m. 1 24 380 113 I 1 56 356 112 I 2 30 405 114 I 2 58 390 114 II 5 3 356 109 II 5 48 353 115 III 6 13 337 109 IK?) Apr. 17 170 1 38 to 1 58 a.m. 1 364 119 II 1 31 400 120 II 2 7 359 112 II 3 10 373 108 II 123 (6imos.; 6.09 kilos) .... Apr. 17-18 60 6 45 to 7 04 p.m. 1 11 334 118 K?) 1 42 359 124 I 2 9 378 124 II 2 35 361 125 I 3 7 363 123 II 3 39 349 121 I 4 5 353 123 II The lowest value, presumably the basal value, is italicized in each series of observations The designations for the activity are given the following values: I, very quiet, probably asleep; II, slight movements, few in number; III, some activity, but generally quiet. The method of estimating the activity is described by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 136. 92 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 23. Heat production of infants after food Continued. Subject No., age and weight. Date. En- ergy of food. Time of feeding. Inter- val be- tween food and end of period. IHeat produced per 24 hours. 1 Pulse-rate. Rela- tive ac- tivity. 1 1916 cals. h.m. cals. 123 (e^mos.; 6.09 kilos) (Continued.) Apr. 17-18 60 6 h 45 m to 7 h 04 m p.m. 4 35 5 13 6 9 330 335 333 118 116 119 I II II 7 12 328 118 III 8 1 322 120 I 9 8 812 1241 II(?) 131 (6imos.; 5.76 kilos) Apr. 19-20 50 6 46 to 6 51 p.m. 1 8 1 43 331 354 120, I 129 I 2 13 354 12l! I 2 45 371 130| II 2 53 329 127 I 3 46 340 128 H( ? ) 4 16 315 129 II 4 47 318 130 I 5 59 342 130 I 7 8 328 126 I 9 7 326 141 I(?) 9 36 391 |143 I Apr. 20 100 5 56 to 6 02 a.m 59 373 J137 II 1 48 392 131 II 2 17 372 118 I 90 8 59 to 9 07 a.m 1 27 299 120i II 2 25 328 !ll9 I 3 21 337 118 I 100 3 51 to 3 57 p.m 1 59 325 114 I 3 6 323 122 II 139 (6jmos.; 6.11 kilos). ... Feb. 19 130 5 35 to 5 44 p.m 1 59 350 122 II 2 50 363 123 I 3 42 328 125 I Feb. 19-20 80 9 35 to 9 55 p.m 1 13 393 129 II(?) 2 4 371 |129 I 2 52 354 126 II 3 41 342 125 I 4 30 366 125 II 5 50 332 |127 III 6 22 325 |124 I 7 19 345 125: II(?) 136(7imos.; 9.08 kilos) ... Mar. 11-12 50 5 48 to 6 08 p.m 1 31 500 106 II 2 16 3 10 505 109 II 471 106 1 II 4 4 463 104 II(?) 4 57 461 103 IK?) 5 51 472 106 UK?) 6 30 481 106 II Mar. 12 230 12 45 to 1 07 a.m 1 15 501 109 I 1 51 520 IllO III 148(8|mos.; 8.87 kilos) ... Feb. 26 170 4 55 to 5 10 p.m 2 28 493 108 I 3 13 475 107J I 3 58 481 106] II 1 The lowest value, presumably the basal value, is italicized in each series of observations. 1 The designations for the activity are given the following values: I, very quiet, probably asleep; II, slight movements, few in number; III, some activity, but generally quiet. The method of estimating the activity is described by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 136. INFLUENCE OF FOOD ON METABOLISM. 93 TABLE 23. Heat production of infants after food Continued. Inter- -. Subject No., age, and weight. Date. En- ergy of food. Time of feeding. val be- tween food and end of [eat produce >er 24 hours Pulse-rate. Rela- tive ac- tivity.* period. w ~ 1916 cols. h. m. cols. 184 (8| mos. ; 887 kilos) .... Feb. 26 170 4 h 55 m to 5 h 10 m p.m. 4 47 467 108| II 6 13 470 114] III Feb. 26-27 130 11 30 toll 42 p.m. 1 24 511 il!7 I 2 14 506 !l!8 I 2 52 516 113 II 3 34 478 114 I 4 5 522 112 II 4 46 479 115 II 5 21 502 117 I 171 (13 mos.; 8.70 kilos) . . . Mar. 4-5 f310 1170 4 00 to 4 10p.m.} 5 45 to 6 00 p.m. ) 2 27 629 136 II 2 57 608 134 I 3 29 647 133 II 4 13 616 132 II(?) 4 53 602 131 II 5 44 584 128 III 6 45 608 130 IK?) 8 37 579 132 I 9 11 645 129 II 10 4 570 126 I 10 51 554 129 I 11 17 555 131 II 1 The lowest value, presumably the basal value, is italicized in each series of observations. 1 The designations for the activity are given the following values: I, very quiet, probably asleep; II, slight movements, few in number; III, some activity, but generally quiet. The method of estimating the activity is described by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 136. The basal metabolism of a child, like that of an adult, is not neces- sarily a fixed factor; hence it is unjustifiable to use a basal value obtained on one day for comparison with the metabolism after food on another day to determine the effect of the food. The question of the selection of a suitable basal value for experiments in such studies has already been exhaustively discussed in a previous publication. 1 It was there pointed out that only basal values obtained on the same day were legitimate for comparison with the metabolism after food, and also that the metabolism after one or two days of fast was not a true basal value; also that for ideal comparisons the basal value without food should first be obtained and the metabolism with the superimposed factor of food studied immediately thereafter. The most extensive series of observations after food were made with the three children Nos. 123, 131, and 171. Of these, the series with the oldest child No. 171, on March 4-5, 1916, was the longest con- tinued, observations being made for over 11 hours after the food had 1 Benedict and Carpenter, Carnegie Inst. Wash. Pub. No. 261, 1918, p. 47. 94 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. been taken. Here, 2| hours after food, the metabolism was 629 calories, rising an hour later to 647 calories, thereafter falling continu- ously until two minimum periods of 545 and 555 calories are found at 9 and 11 hours, respectively, after the food had been taken. If we consider the absolute minimum of 545 calories as basal, then the 647 calories, or the maximum heat production, represents an increase of 102 calories, or about 20 per cent, due to the influence of food. But this increase is relatively soon after feeding. The pulse-rate likewise shows a distinct tendency to fall off after the first rise due to the taking of food. The average value of 550 calories, which may be taken as the basal value, was determined, roughly speaking, 10 hours after the last meal. An examination of the chart for No. 171 (see fig. 16, page 124) shows that at the age of 13| months there is a decided break in the line for the total calories, indicating a fairly low value at that time. The period of observation is considerably longer than that for most of the children in the series. It is therefore quite likely that, had the experi- ment been shorter, the point would more closely correspond to the general trend of the curve. But great stress should not be laid upon individual variations from day to day or from period to period. On April 17-18, 1916, a long experiment was made with the infant No. 123, in which the child took food containing but 60 calories of energy. The metabolism was thereafter measured continuously for over 9 hours. While the absolute minimum value of 312 calories was found in the last period, the values for the last 2 hours indicate relative constancy. The maximum of 378 calories, which occurred 2 hours and 9 minutes after food, is equivalent to an increase of 66 calories above the minimum, or, roughly speaking, 20 per cent. The pulse-rate shows a general tendency to decrease after the maximum metabolism is reached, although the minimum pulse-rate is not coincidental with the minimum metabolism. The series of observations with No. 131 continued for 9| hours after food. The amount of food taken was insignificant in quantity and energy content, and the minimum value actually appeared about 4^ hours afterwards. The maximum value (391 calories) was found 9^ hours after food, or some 5 hours after the minimum value appeared. The great increase in the pulse-rate is wholly inexplicable, for the kymograph records indicate activities of but I and II, and the subject was presumably asleep. We have no explanation for these unusual figures. With the other children, shorter observations were the rule after food. With No. 127, on April 20-21, 1916, five feedings were given, the observations after the first feeding continuing without break for 5 hours and 31 minutes. The absolute minimum of 255 calories was found about 5 hours after food. As the maximum value, which was INFLUENCE OF FOOD ON METABOLISM. 95 obtained 1 hour 41 minutes after the third feeding, was 336 calories, there was a maximum rise of 81 calories, or approximately 30 per cent. Singularly enough, there was no marked change in the pulse-rate, the maximum increase being but a few beats. The series of observations with No. 115 after the two feedings agree very satisfactorily, although the question of the actual basal value is in doubt, since this must be taken as but 3 hours and 11 minutes after the second feeding. The total increase on this basis is, therefore, 66 calories, or about 22 per cent above basal. The observations with the other children were approximately 3 to 6 hours in length and usually indicate the minimum at the end of the period of observation, with maximum increases shortly after food was taken, these corresponding to not far from 20 to 30 per cent with most of the children. Of special interest with No. 119 is the fact that in the second feeding the energy in the food was almost three times that in the first feeding, and under these conditions a distinctly higher metabolism was noted. If we accept the basal value as 369 calories, which was obtained 5 hours and 41 minutes after the feeding of 80 calories in breast milk, we find an increase of 108 calories 1 hour and 50 minutes after the second feeding of breast milk with an energy con- tent of 210 calories. This increase after the second feeding corre- sponds to about 29 per cent, as compared with the maximum increase of 34 calories, or 9 per cent, 1 hour and 8 minutes after the first feeding of a very much smaller amount of breast milk. With No. 140, who was the same age as No. 119, the energy content of the milk in the first feeding was 70 calories, while that of the second feeding was 170 calories. Here the increase due to the larger amount of milk was very slight, if any. It is also worthy of note that No. 136, after a second feeding, showed a maximum heat production no greater than that found after the first feeding, although the second feed- ing had an energy content four or five tunes larger than that of the first. From an inspection of the data in table 23 it is clear that in most instances the absolute minimum values were not found before 8 to 9 hours after the last food. In nearly every one of the long series of observations the minimum values occur at or near the end. An exception to this is No. 131 in the observations on April 19-20, when approximately the minimum was found 4| hours after the food. The energy content of the food in this case was, however, but 50 calories. Numerically the same conditions hold with regard to the observations on March 11-12, 1916, with No. 136. It is probable, therefore, that with as small an amount of food as 2 or 3 ounces of breast milk, with an energy content of approximately 50 calories, the metabolism is nearly basal at the end of 4 hours. With larger amounts of food the stimulating effect may persist for 9 or 10 hours. 96 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. It has been currently believed that the influence of food is con- siderably smaller with children than with adults and that the protein utilized for growth does not participate in the stimulating action formerly termed "specific dynamic action." Frankly, we were misled in some of our earlier observations by this belief, but the evidence here presented certainly gives no indication of a material lessening in the influence of food. Indeed, the general conclusion to be drawn from these data is that the influence of food upon the metabolism of infants is wholly comparable to that noted with adults. This renders all the more potent our contention that basal metabolism measure- ments with young children are not ideal in the commonly accepted definition of the term basal as applied to adults, since they are not made under conditions which preclude the stimulating action of food. A rough assumption may be made that the basal metabolism as measured in our series of experiments is, in practically every instance, from 8 to 15 per cent higher than the true basal. This fact must be clearly recognized in the subsequent analysis of the general trend of the metabolism for children of varying ages. In the light of the variations shown in table 23, however, the direct application to our data of this percentage correction is not justifiable. Consequently, until further observations are made upon the quantitative effects of food on the metabolism of children, one can only emphasize the fact that the younger the child is, the greater is the deviation from basal towards a higher measurement of the metabolism. It is a matter of regret that the influence of food was not studied with older children, but since it has been made clear that the basal values are too high, especially with the younger children, it is not impossible that these high values may, when corrected, equalize the values for heat production per square meter and bring them more in accordance with those found with adults. On the other hand, it would appear that if a correction were applied for children 6 months and under, it would only further distort the comparison curves for the metabolism. (See discussion of these curves in a subsequent section, page 176.) Final treatment of these curves, i. e., in com- parisons of children ranging in age from birth to puberty, and par- ticularly in comparisons of children with adults, must be subject to the possibility of a correction to obtain the true basal. On the one hand, there will be a distinct objection to establishing a so-called correction on relatively few long experiments; on the other hand, there will be a distinct objection fco comparing true basal values with the basal values of youth which are too high as a result of previous food. Except in cases of great weakness and probably disease, no metabolism values for youth may be considered as under basal, the tendency always being for such values to be above rather than below true basal. ELEMENT OF NOVELTY IN MEASUREMENTS OF METABOLISM. 97 Aside from the physical difficulties met with in attempting to secure measurements of the metabolism of children under true basal condi- tions, we have to deal with the possibility of an incipient acidosis. With adults, complete fasting soon produces a reduction in the store of glycogen and shortly thereafter acidosis sets in. The earlier experiments of Schlossmann and Murschhauser 1 seemed to show that this acidosis appeared much earlier with children than with adults. Experience in this laboratory, both with a fasting man 2 and with diabetics, 3 indicates strongly that the metabolism is noticeably stimu- lated by acidosis. As has been shown in other publications, it becomes a serious problem to decide upon the exact point when the stimulating effect of food ceases and the stimulus of acidosis begins. Even with adults this point has never been sharply denned, although from the general trend of metabolism during long experiments it appears that a minimum is reached not far from 10 to 12 hours after the last food is taken, unless the last meal was particularly rich in protein. Since one may designate the measured metabolism of these younger children only as approximating basal, comparisons between these values and true basal values for adults must be made with extreme caution. THE ELEMENT OF NOVELTY IN MEASUREMENTS OF METABOLISM. Throughout our entire series of experiments we were frequently con- fronted with the question as to how much of a role the element of novelty played in the determination of metabolism. Naturally, when the children went into the respiration chamber for the first experi- ment, their attitude toward the apparatus was somewhat different from that in the second experiment. While every effort was made to accustom the children to seeing the apparatus in working order and to assure them that there was nothing distressing or uncomfortable in the experiments, the novelty of being placed inside the respiration chamber might conceivably affect the basal metabolism; hence values secured on the first day might be looked upon as aberrant. For a strict comparison of the results obtained under the two con- ditions, it is important in the first place that the two days be reason- ably close together, so as to eliminate the factors of age and weight. Secondly, the periods for comparison must be of the lowest order of activity; that is, the child should preferably be in complete repose in both periods. After a careful scrutiny of our data, we have selected a number of experiments with children of both sexes and varying ages, which give suitable data for such comparison. These are brought together in table 24, which shows the heat production on the 24-hour 1 Schlossmann and Murschhauser, Biochem. Zeitschr., 1913, 56, p. 396. 2 Benedict, Carnegie Inst. Wash. Pub. No. 203, 1915, p. 365. 3 Benedict and Joslin, Carnegie Inst. Wash. Pub. No. 176, 1912, pp. 125 and 134. 98 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 24. Comparison of metabolism and pulse^rate in first and second experiments with children. Heat (computed) Pulse- per 24 hrs. rate. Relative Subject No. Age. Date. No of activity. Amt. Diff. beats. Diff. Boys: yrs. mos. cols. cals. 192 5 7 May 21, 1919 866 90 11,1 May 22, 1919 962 + 96 89 - 1 I, II 194 5 9 Mar. 19, 1918 876 85 1,1 Mar. 22, 1918 830 - 46 87 + 2 I, I 204 7 2 Mar. 29, 1919 999 81 1,11 Apr. 1, 1919 908 - 91 86 + 5 1,1 209 7 10 Mar. 1, 1918 1,094 83 I, I, II Mar. 18, 1918 1,096 + 2 74 - 9 I, I, II 228 9 10 Feb. 5, 1919 1,241 85 I, II, I Feb. 6, 1919 1,215 - 26 75 -10 11,11 235 10 7 Feb. 19, 1919 1,063 93 1,1, I Feb. 20, 1919 1,056 - 7 84 - 9 I, I, I 242 11 Nov. 25, 1918 1,215 87 II II Nov. 27, 1918 1,186 - 29 88 + 1 I II 246 11 6 Feb. 6, 1919 1,320 70 I II Feb. 7, 1919 1,314 - 6 64 - 6 ,1 250 12 2 June 24, 1918 1,355 71 I, II II June 27, 1918 1,211 -144 61 -10 I, II II 256 12 9 Feb. 20, 1918 1,365 87 I II Feb. 23, 1918 1,290 - 75 72 -15 II II 258 13 8 Apr. 17, 1919 1,529 66 I.I Apr. 18, 1919 1,504 - 25 62 - 4 I, II 259 14 1 May 6, 1918 1,260 78 I, I May 8, 1918 1,220 - 40 66 -12 1,1 260.....'. 15 . May 1, 1919 1,417 74 I, I, II May 6, 1919 1,431 + 14 71 - 3 1,1,1 Average: 1st day 1,200 81 2d day 1,171 - 29 75 - 6 Girls: 178 2 11 Mar. 19, 1918 572 84 I, I, I, I Mar. 21, 1918 574 + 2 91 + 7 1,1,1 189 5 3 Feb. 20, 1919 875 80 I, I Feb. 21, 1919 831 - 44 78 - 2 II, I 220 9 1 Apr. 11, 1918 1,031 101 II, II Apr. 12, 1918 996 - 35 97 - 4 11,11 225 9 5 Jan. 22, 1919 954 88 1,1,1 Jan. 23, 1919 957 + 3 82 - 6 I, I, II 238 10 10 Apr. 26, 1919 987 89 III May 3, 1919 1,103 +116 81 - 8 1,1,1 251 12 2 Mar. 8, 1919 1,027 79 I, I, I Mar. 11, 1919 l!o42 + 15 76 - 3 1,1, I Average: 1st day . . 908 87 2d day 917 + 9 84 3 Gen. average: 1st day . . 1,108 83 2dday... l!o91 - 17 78 - 5 basis for the first and second days of observation, the pulse-rate, and (to indicate the degree of repose) the activity as estimated from the kymograph records. To indicate any variations due to sex, the data have been grouped separately for boys and girls. ELEMENT OF NOVELTY IN MEASUREMENTS OF METABOLISM. 99 With the majority of the subjects the observations were made on succeeding days, and with few exceptions not more than three days apart. These exceptions were Nos. 209, 238, and 260, with whom the observations were made 17, 7, and 5 days apart, respectively. The agreement in the values for the heat production for the two days is, in most cases, reasonably close. The results for the second day are, in many instances, lower than those for the first day, but the reverse is likewise true, so that, taking the plus and minus signs into considera- tion, one may state that the general average difference between the first and second day for all of the children is but 17 calories. Bearing in mind the difficulty of estimating relative activity from day to day, especially at the lower grades of activity, we find that the records of the relative activity in the last column of table 24 indicate that the average activity on the two days was equal. Of special significance, however, is the pulse-rate, both when com- pared for the first and second days of experimenting and particularly when compared with the value for the total heat production. In the majority of instances the pulse-rate is lower on the second day than on the first day, the average of the values for all of the children on the second day being 78 beats per minute as compared with 83 beats on the first day. In many cases the pulse-rate is notably lower on the second day. The influence of "training," so to speak, is thus definitely towards a decidedly lower pulse-rate. A few instances of a higher pulse-rate on the second day occur, these being found with three boys, Nos. 194, 204, and 242, and with one girl, No. 178. The largest change is with the boy No. 256, whose pulse-rate fell from 87 to 72, or a total change of 15 beats. Changes of 9 or 10 beats are noted in several cases, especially with the boys. With these larger pulse-rate changes, there is in most cases a corresponding change in the metabol- ism, a fall in the pulse-rate being accompanied by a downward tendency in the metabolism. The changes are, however, by no means uniform, and the reverse is not infrequently found, namely, a lowering in pulse- rate with a slight increase in metabolism. This lack of harmony between pulse-rate and metabolism, w^hich is more particularly evident with the girls, is sufficient to show that with children in the growing period there apparently is much less correlation between pulse-rate and metabolism with the same child than is found either with younger children or with adults. When the data are examined with reference to their grouping as to sex, some difference is shown in the values. It might be expected that the boys would give results for the second day but little, if any, lower than for the first day, as they would naturally be less apprehensive and sensitive to outside influences than the girls and even enjoy the novelty of the experience. The data, however, show that with the boys there was almost invariably a fall on the second day, but three 100 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. cases out of the thirteen giving an increase. The pulse-rate also shows a fall in all but three cases, but no one of these three was simul- taneous with the rise in the metabolism. Unfortunately, the data for the girls are somewhat meager as compared with those for the boys. With four of the six girls the metabolism rose on the second day, but with two of the girls the rise was insignificant. The pulse-rate almost invariably fell slightly. Neither the boys nor the girls show a differ- ence in the comparison of results due to age. The slight changes noted between the two sexes must be considered primarily of a psycho- logical and only secondarily of a metabolic nature. The values in table 24 thus demonstrate that the element of novelty plays no appreciable role in the basal metabolism. This finding is in full accord with the recent series of observations by Hendry, Carpenter, and Emmes 1 on a group of medical students, and makes it seem all the more probable that the small amount of preliminary training considered absolutely essential for metabolism measurements may be reduced to a minimum. Apparently, with complete muscular repose, the influence on the pulse-rate of previous experience in the respiration chamber is, with children, of minor significance. METABOLISM AS AFFECTED BY GROWTH. GENERAL METHODS OF STUDY. An ideal study of the metabolism of children from birth to puberty would be the continuous measurement of the same child at frequent intervals throughout this period of life. All past experience has shown that the physiological activities of a child can by no means be represented by a straight line or by a regular curve function, for there are gross irregularities which are inexplicably and inherently a part of physiological life. Accordingly it is necessary to base deductions not upon the analysis of the metabolism of one child alone, but upon the metabolism of several children. Hence we felt it our duty to obtain measurements for a number of boys and girls during as long a period of life as they were available. The children of wet-nurses could be more or less controlled from birth, but as each year passed they became more widely scattered, and consequently it was increas- ingly difficult to secure them for observation. But even under these conditions we were able to hold and to study a representative number of these children. In all, studies were made with 23 children over varying periods of tune. In no instance, however, could we approx- imate the ideal of intermittent measurements from birth to puberty, for after three or four years of observation it became impossible to keep in touch with the children. 1 Hendry, Carpenter, and Emmes, Boston Med. and Surg. Journ., 1919, 181, pp. 285, 334, and METABOLISM AS AFFECTED BY GROWTH. 101 While such a method of intermittent study may still be considered as the ideal method, since it not only supplies an index of the individual variations for the same child from age-period to age-period, but also variations for different children at the same age and weight, it became necessary for us, in the absence of "ideal" conditions, to rely upon the accumulation of a considerable mass of evidence for not only the 23 children who were studied more or less continuously, but likewise a large number of isolated observations on other children. This col- lection of more or less isolated data, when combined with the semi- continuous observations, makes possible the charting of a considerable mass of material regarding the metabolism of children in the period from birth to puberty, thus supplying an excellent index of the general trend of the metabolism of a child. By this method of procedure it is perfectly legitimate to consider values obtained with the same individual with changes in age and weight of some magnitude as representative of a new child at the indi- cated age and weight. The basal metabolism of an adult, as measured from month to month and from year to year in the decade from 20 to 30 years of age, does not vary greatly as a result of a change in age, and there is usually in this time no material change in weight or height. With children, on the other hand, the rapid growth and changes in stature and weight make it wholly illogical to average the values found for any child during a considerable range in age and to consider that the result represents the true average value for that child. In other words, a child of 1 year is one individual, but a child of 1| years is still another individual. Accordingly, in grouping the children for general consideration, a definite method of selection was followed in determining the degree of change in either weight or age which would make it desirable to consider the child as a new individual. (See page 131.) In analyzing our experimental data, we naturally turn first to the picture presented by the results for individual children with whom observations were continued for several months or years. Following this treatment of the data, we may properly proceed to summary tables and charts giving the results for all of the children studied, including not only those data representing long periods of time, but also the shorter isolated determinations of the basal metabolism made upon a large number of children of different ages who were usually observed on but one or two closely following days. METABOLISM DURING GROWTH AS SHOWN BY THE INDIVIDUAL CHILD. OBSERVATIONS WITH SUBJECT No. 145. Of the 23 subjects that were studied over relatively long periods of time, No. 145 (a girl) has been selected for detailed treatment to indi- cate the method of procedure which was followed with all of the chil- 102 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. dren. The data for this child are given in table 25. For complete record it has been necessary to include data for the body-weight, height, age, the estimated energy of food taken prior to the experiment, the length of time elapsing between the taking of food and the beginning of the first observation, the carbon dioxide produced per hour in the several successive experimental periods, the respiratory quotient, the average pulse-rate for each period, the relative activity as estimated from the kymograph records, notes as to whether the subject was asleep or awake, the total heat production (computed) per 24 hours, and likewise the heat produced per kilogram of body-weight per 24 hours and per square meter of body-surface per 24 hours. The relative activity is expressed on an arbitrary scale ranging from the greatest degree of repose (I) to the greatest activity (VI). The body- surface was determined by the Du Bois linear formula and direct measurements made upon the subject for practically every observation. As has already been pointed out, while ideally the observations should have been made without food in the stomach, this is an abnormal condition for young children. It will be noted that as the child grew older the length of time between the feeding and the beginning of the first experimental period lengthened, until it was finally between 4 and 5 hours long. As indicated in the outline of the description of the experimental procedure, the carbon-dioxide production was directly measured for each experimental period. These periods were approximately 30 minutes in length. As it was impracticable to determine the oxygen consumption in periods as short as this, due chiefly to the difficulty in obtaining accurate measurements of the average temperature of the air inside the respiration chamber, we determined only the total consumption of oxygen for the entire sojourn of the child inside the chamber. Thus for four experimental periods, each of 30 minutes, the oxygen consumption would be determined only for the full 2-hour period of the experiment, and this total amount used in the computa- tion of the respiratory quotient. In most of the studies there was a preliminary period in which various adjustments were made and in which the oxygen consumption was not determined. During the preliminary periods the amount of carbon dioxide produced was almost invariably larger than in later periods, this being due to the activity of the child, who was frequently awake at this time. The carbon-dioxide production for these periods is likewise recorded in table 25, but the measurements of the oxygen consumption, which were used for computing the respiratory quotients, were made only in the main periods of the observation, i. e., those represented by the values inclosed in brackets. It occasionally hap- pened that, even during the preliminary period, the carbon-dioxide production was at a minimum. Thus, on April 2, 1917, the carbon- METABOLISM AS AFFECTED BY GROWTH. 103 TABLE 25. Results of observations on the gaseous exchange of No. 145 (girl). S H Heat (computed) a .9 . I, I per 24 hours. 5 g Date. 3 1 11 jl o 1 ! ii f 1 Remarks. 11 "8 o g 1 ! 1 3 | a- a I .s-s '3 ,0 3 'a 3 *l "1 5 a P : a' 1 l *I " a 1916 \cals. h. m. 0m. cals. cals. cals. June 15 5 mos. ; 62 80 1 30 4.78 i (337 64 1 1,109 in i cm.; 5.27 4.71 \ 0.84 1 332 63 ! 1,092 117 i Probably kilos. 4.94 I 348 66 ! 1,145 116 i asleep. 4.46' J [315 60 ! 1,036 119 i June 16 5.22 kilos. 80 30 7.9l! , 580 111 1 1,908 126 IV 4.38 321 61 1,056 117 i 4.87 357 68 1,174 116 ii 4.34 0.80 318 61 1,046 118 i Do. 4.64 ; 340 65 1,118 117 ii 4.93 361 69 1,188 118 in Nov. 22 10 mos.; 160 1 8.70 608 66 1,310 117 n Asleep. 70 cm.; 9.19 kilos. 8.86 } 6.78 1 n OK 8.01 ' 85 619 474 560 67 52 61 1,334 1,022 1,207 113 117 112 in i in 1 Probably J asleep. 8.23 J 575 63 1,239 120 VI Dec. 4 9.50 kilos. 130 30 7.64 511 54 1,099 102 II | 7.94 } 7 ' 99 090 (531 534 56 56 1,142 1,148 111 108 II III [Asleep. 7.55 U<9 505 53 1,086 111 II J 11.23 J 751 79 1,615 126 V Dec. 22 11 mos.; 150 8.26 543 53 1,042 110 I Do. 72.5 cm.; 7.90 519 51 996 114 II Do. 10.2 kilos. 7.80 8.90 7.73 0.92 513 585 508 51 58 50 985 1,123 975 111 114 109 I II I } Probably J asleep. 9.22 606 60 1,163 113 V 1917 Jan. 11 lyr.; 150 1 7.08 477 46 928 103 I Asleep. 73.5 cm.; 8.88 599 57 1,165 113 II Probably 10.5 kilos. asleep. 7.32 0.89 494 47 961 106 II Asleep. 8.49 573 55 ! 1,115 113 II Awake, then asleep. Jan. 31 10.9 kilos. 260 1 8.80 .... 583 53 I 1,094 106 IV 8.27 f548 50 I 1,028 105 II 1 8.40 \ 0.91 { 557 51 1,045 111 I [ Asleep. 8.74 J [579 53 1,086 106 II J 10.14 672 62 1.261J 119 VI Feb. 19 2 lyr. Imo.; 470 30 9.22 606 54 1,104 112 I ) 76.0 cm.; 11.3 kilos. 8.59 9.09 } 0.92 f564 \597 50 1.027J 113 53 ! 1,087 118 I I \ Do. 13.86 911 81 1,659 136 V Mar. 15 1 yr. 2 mos. ; 240 4 30 10.40 701 59 1,247 118 VI 76.0 cm.; 7.68 1 (518 44 922 105 II } 11.8 kilos. 8.53 1 n CQ 575 49 1,023 110 II \ Do. 8.73 ' 89 589 50 1,048 104 II J 10.84J J 731 62 1,301 V 1 Includes 120 cals. taken 4 hours before the observation. 2 Rectal temperature at beginning of observation was 99.8 F.; at end, 100.5 F. 3 Includes 100 cals. taken 3j hours before the observation. 4 Includes 100 cals. taken about 4 hours before the observation. 104 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 25. Results of observations on the gaseous exchange of No. 145 (girl) Continued. 5 W) Heat (computed) st fj 3 11 si per 24 hrs. i 2 i T) 3 5| 1 ""t o 1 ^ s" I r3 i Date. P ii || 1 3 1 H P, c3 1 Remarks. ij fw Jo JS || i .g M |l ? |D 1 a < 1 Q (2 * *a 1917 cals. h. m. gms. caZs. cals. cals. Mar. 27 12.1 kilos. 320 1 9.00 .... 612 51 1,048 103 ii Asleep. 9.00 (612! 51 1,048 101 ii Do. 8.95 I f| QQ I 610 50 1,045 105 i 1 Probably 8.83 f U.OO 1 601 50 1,029 108 i / asleep. 9.46 J [644 53 1,103 108 in Apr. 2* 12.0 kilos. 140 8.73 589 49 1,026 109 i 9.09 613 51 1,068 112 ii [ Asleep. 9.55 0.89 644 54 1,122 113 ii J 9.97 672 56 1,171 121 IV Apr. 20 1 yr. 3 mos. ; 2403 1 8.61 '591 47 998 105 ii \ Probably 78.0 cm.; 8.66 87 594 47 1,003 101 ii / asleep. 12.5 kilos. 8.43 579 46 978 97 in Asleep. 9.49 651 52 1,100 105 IV May 15 1 yr. 4 mos. ; 78.5 cm.; 320< 30 9.24 8.10 } 618 49 (541 43 1,035 906 100 92 IV II \ Probably 12.6 kilos. 8.65 [ 0.90 \ 578 46 968 97 III / asleep. 9.58 J [640 51 1,072 100 V May 18 12.5 kilos. 310* 10.05 660 53 1,065 106 V 9.74 640 51 1,032 100 IV 8.25 0.92 542 43 874 99 I Probably asleep. 8.61 566 45 913 97 II Asleep. 10.14 666 53 1,074 108 VI June 20 1 yr. 5 mos.; 80 cm.; 640 30 10.51 9.30 I 709! 53 f 6271 47 1,142 1,010 108 103 III I \ Probably 13.4 kilos. 9.79 \ 0.98 \ 660 49 1,063 110 I / asleep. 13.59 J [917 69 1,4771 119 V June 21 13.4 kilos. ( 7 ) ( 7 ) 9.92 646 48 1,040 ! 102 IV 8.67 565 42 910 98 II } 9.29 0.93 605 45 974 105 I [ Asleep. 9.69 631 47 1,016 95 I Nov. 6 15.1 kilos. 170 30 9.93 653 43 978 84 I Do. 9.03 593 39 888 89 I Do. 9.96 0.92 654 43 979 91 I Probably asleep. 11.71 769 51 1,151 100 V Nov. 15 1 yr. 10 310 1 30 12.05 109 VI mos.; 86.5 9.45 87 rv Asleep, cm.; 15.2 sobbing. kilos. 1 Includes 100 cals. taken about 3^ hours before the observation. 1 Rectal temperature at beginning of observation waa 100.5 F. ; at end, 99.8 F. Bad diarrhea on two preceding days; cutting teeth. J Includes 100 cals. taken about 4 hours before the observation. 4 Includes 100 cals. taken 4 hours before the observation. 6 Includes 170 cals. taken 4 hours before the observation. ' Includes 230 cals. taken 4 hours before the observation. 7 Energy of food not known. Probably at least 150 cals. were taken just before the observation. 8 Includes 140 cals. taken about 5^ hours before the observation. METABOLISM AS AFFECTED BY GROWTH. 105 TABLE 25. Results of observations on the gaseous exchange of No. 145 (girl) Continued. 5 H Heat (computed) 1| 1 "3 a 8.2 1 per 24 hours. 1 5 ^ Date. 1| ll || H"* 3 b | || I 1 1 Remarks. .S "3 so o r? r* I* |- ll .1 1 1 d-1 | 1 .83 < 1 j! 3 a- H 1 II 1 1 1918 cals. h. m. gms. cals. cals. cals. Jan. 18 2 yrs.; 170 5 12.74 111 VI 90.0 cm.; 12.48 93 IV 15.4 kilos. Jan. 19 15.3 kilos. 120 5 8.40 582 38 841 80 I Asleep. 9.34 0.86 647 42 935 85 I Do. 11.24 779 51 1,126 103 III Mar. 5 2 yrs. 2 90 5 30 7.99 539 35 797 72 I Do. mos.; 91.0 9.20 0.89 620 41 917 74 II Do. cm.; 15.2 10.30 695 46 1,028 85 III Asleep, kilos. but waked May 16 2 yrs. 4 170 4 30 8.98 ] f628 40 946 69 I ] mos.; 94.0 8.55 I f\ OK 1 598 38 901 67 II Asleep. cm.; 15.8 9.88 f U.ou | 691 44 1,041 70 I J kilos. 13.46 J [941 60 1,417 96 VI May 22 15.7 kilos. 200 4 13.63 971 62 1,458 93 VI 8.16 f581 37 872 69 I Do. 8.70 0.83 I 620 40 931 71 I Do. 8.94 [637 41 956 73 III Probably Nov. 5 16.7 kilos. 210 4 13.22 969 58 1,400 100 VI eep. 9.92 } f727 43 1,051 78 I Asleep. 9.10 } 0.80 \ 667 40 964 76 II Do. 11.60 j [851 51 1,230 78 III Awake, then asleep. Nov. 13 2 yrs. 10 280 4 14.38 1024 62 14,95 93 VI mos.; 96.5 9.18 } [654 40 955 76 I Asleep. cm.; 16.4 7.98 \ 0.83 \ 568 35 829 71 I Do. kilos. 10.60 J [755 46 1,102 83 IV 1919 Jan. 16 3 yrs.; 200 3 30 10.89 761 45 1,119 93 V 97.5cm.; 9.26 ^ 647 38 952 75 I Do. 17.0 kilos. 12.28 > 0.85 858 51 1,262 77 IV Jan. 17 16.8 kilos. 220 4 30 10.02 ' 735 44 1,081 71 I Do. 9.56 701 42 1,031 76 III Asleep, 0.80 but rest- less. 9.36 686 41 1,009 73 I Asleep. 11.74 861 51 1,266 86 V Mar. 10 17.5 kilos. 234 4 13.27 927 53 1,324 94 VI 9.86 9.58 | 0.85 /689 \670 39 38 984 957 71 71 I III Do. Asleep, then awake. Mar. 17 3 yrs. 2 254 4 9.96 /716 41 1,036 72 II Asleep. mos.; 98.5 8.94 ' \643 37 931 67 I Do. cm.; 17.5 14.50 1043 60 1,509 97 VI kilos. June 13 3 yrs. 5 319 4 30 11.50 789 45 1,126 88 III mos.; 101.5 9.28 1 /637 36 909 72 I Do. cm.; 17.5 9.94 > 0.87 \682 39 973 72 I Do. kilos. June 14 17.4 kilos. 249 4 12.09 924 53 1,300 88 V 10.38 9.38 }0.76 (793 \717 46 41 1,115 66 1,008 70 I I Do. Do. 106 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. dioxide production in the preliminary period was 8.73 grams, 1 while the lowest during the subsequent periods was but 9.09 grams, even when the child was asleep. But this case is exceptional. In the first observation with No. 145, that of June 15, 1916, the carbon dioxide produced in the first three periods varied only from 4.71 to 4.94 grams per hour. This is an astonishingly good agreement between periods and is in large part explained by the fact that the relative activity (indicated on an arbitrary scale) was I in all three of the periods and the child was probably asleep the entire time. Since the carbon dioxide values agreed so well, as a natural consequence the values for the computed heat for these three periods likewise agree well with each other. The carbon-dioxide production and heat for the fourth period were somewhat lower than for the other periods and thus represent the absolute minimum as determined for this day. Before further consideration of the data in table 25, the possible errors entering into the measurements for any given period should be pointed out. For example, the measurement of the carbon-dioxide production for the individual periods is based upon the increase in weight of a set of absorbing bottles containing soda-lime and sulphuric acid. This requires the weighing of the bottles at the beginning and the end of each period. While all such weighings are verified by a second observer, an error in either one of these weights will obviously affect the value for the carbon-dioxide production for that period. Such a method does not admit of duplicate determinations for any period a palpable defect. While this is true of practically all modern methods in which a respiration chamber is used, i. e., either the total amount of carbon dioxide is measured only by one absorbing train or, if duplicate gas-analyses are made of samples taken from an air-current, but one main meter or spirometer measures the air-current, still, in considering the carbon-dioxide production for any individual period, such possibility of error must be borne in mind. On June 15 the minimum heat production of 315 calories per 24 hours found in the fourth period is reasonably well confirmed by the second period with 332 calories per 24 hours, which is about 5 per cent higher. An average of these two figures would therefore give the most probable basal value for the child on this day. Since, however, we have an actually determined value of 315 calories, it is clear that, aside from technical errors, this heat production must be the basal or minimum heat production. In general, it may be said that the minimum basal metabolism is that actually observed, barring tech- nical errors. We are then specially interested in evidence to prove the probability of a low value. Specifically, is 315 calories the real minimum value, or shall we tacitly admit the probability of technical 1 In table 25 it seems desirable for purposes of comparison to represent the carbon-dioxide production on the basis of one hour rather than for the actual time of observation. METABOLISM AS AFFECTED BY GROWTH. 107 errors and average the two lowest periods? Obviously, if 315 calories had been the result of duplicate simultaneous determinations, its validity could hardly be questioned. In an earlier communication 1 we emphasized strongly the relation- ship observable between the pulse-rate and the total metabolism. The pulse-rate for the four periods in this experiment varied only between 111 and 119. In the earlier observations just referred to, special stress was laid upon the countings of the pulse-rate to secure a suf- ficiently large number of counts to use for comparison with the metab- olism which was simultaneously determined. In the present studies the records of the pulse-rate were necessarily subordinated in many instances to other technique. The average of 111 for the pulse-rate in the first period does not, therefore, necessarily represent the same number of counts as does the average of 119 for the fourth period. Effort was made, however, to secure a sufficient number of counts to give a representative average. In this observation the lowest carbon- dioxide production per hour (that of the fourth period) occurred when the pulse-rate was actually the highest. Judging from these data alone, the correlation between pulse-rate and metabolism is by no means positive and gives no further evidence of the validity of the value for the metabolism in this fourth period. A second observation of 6 periods was secured on the next day (June 16), The high carbon-dioxide production for the preliminary period of 50 minutes (which is here given as 7.91 grams per hour) is readily accounted for by the activity in this period represented by the arbi- trary designation IV. An accentuation of the pulse-rate is also found in this period, with a rate of 126 beats. The agreement in the carbon- dioxide production for the several periods of the main experiment is reasonably close, the values ranging from 4.34 to 4.93 grams per hour. The two lowest values, 4.38 and 4.34 grams, were coincidental with periods in which the activity was but I. For this day it is obvious that 4.34 grams of carbon dioxide, or from 318 to 321 calories, may justifiably be considered as the minimum. This amply confirms the single low value of 315 calories noted for the fourth period of the day before. While, therefore, the value for the last period on June 15 gave a true measure of the basal metabolism on this day, it has far greater weight from its substantiation by the values found in the two periods on June 16. There was an interval of 5 months between the observation of June 15 and the next on November 22, during which there was a consider- able increase in the length and weight of the child. The higher values for the carbon dioxide are thus easily explained by these increases. On November 22 we have but one period with the lowest grade of 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914, p. 130. 108 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. muscular activity. In this period the carbon-dioxide production was 6.78 grams per hour, corresponding to a heat production of 474 calories per 24 hours or 52 calories per kilogram of body-weight. It is unfortu- nate that no other period confirms this low value and that no test was made until December 4, nearly 2 weeks later. On this latter date no values are found so low as the minimum for November 22, although one period gives 505 calories per 24 hours or 53 calories per kilogram of body-weight 1 as compared with 52 calories observed on November 22. It is not unreasonable to consider that these two periods verify each other. In analyses of the period values in these experiments, it is evident that verification by similar figures on another day or duplicate periods on the same day are highly desirable for establishing the true minimum metabolism. But this is by no means possible in every day's experi- mentation. Uncontrollable activities of III or more immediately rule out estimates for minimum values. Frankly, the pulse-rate does not give invariably so close a correla- tion with the activity and the metabolism as we had expected. Thus, with the low period on November 22, the pulse-rate was 117, while hi the most active period it was but 120. The respiratory quotients in the first four observations with this child range from 0.80 to 0.90 and are fully in line with the quotients which might be expected from the character of the diet. No abnormalities in the respiratory quo- tients are to be seen in the entire series, save that the low value of 0.76 in the last observation on June 14, 1919, is unusual. An examination of the figures for the carbon-dioxide production as the observations progressed, giving due regard to the' variations in activity, shows that the carbon-dioxide production during periods of niinimum activity gradually increased as the age, height, and weight increased. With this child, at least, the minimum or basal metab- olism for any given day is usually marked sharply. The only experimental factor which varies in this series of observa- tions, other than that of activity, is the food, as the amount taken, the character of the food, and the time between feeding and the first observation differ more or less on the several days. Theoretically the stimulus to cellular activity and to the metabolism increases with the increase in the amount of food and with the shortening of the interval between the taking of food and the beginning of the observations. Practically, when one attempts to correlate the amounts of food given, its character, and the length of time prior to the observation, it is only with difficulty that clear evidence can be found of the influence of food upon the basal metabolism. At first sight it would seem that this point is illustrated by the data for January 31, 1917, and those 1 When the body-weight changes without material change in age, comparisons are best made on the basis of equal units of weight. METABOLISM AS AFFECTED BY GROWTH. 109 three weeks later (February 19). In one instance food containing 260 calories was taken immediately before the observations began; in the other, food containing 470 calories was given 30 minutes before the observations. The carbon-dioxide production and total heat pro- duction on February 19 are noticeably greater than those for January 31, but owing to a coincidental increase in body-weight, the heat production per kilogram of body-weight and that per square meter of body-surface are practically unaltered. The pulse-rate for February 19 is distinctly higher on the average than on January 31, although in both observations the subject is reported as being asleep, with, if anything, the lesser activity on the second day. A complicating circumstance here is the fact that on February 19 there was a slight rise in rectal temperature, the record at the beginning of the observations being 99.8 F. and that at the end 100.5 F. The possible influence of food and the possible influence of a slightly febrile state are thus commingled and no sharp deductions can be made. The influence of a slightly febrile temperature seems to be more marked upon the pulse-rate than upon the metabolism itself. On the two days when slightly febrile temperatures were noted, namely, on February 19, and April 2, 1917, the pulse-rates were noticeably higher than on the days immediately before or after. A corresponding increase in the metabolism is not always definitely noted. Still, the very fact that the pulse-rate was so obviously disturbed by this slight rise in temperature makes it all the more important to avoid the use of observations when the body temperature is even slightly febrile, especially in comparison experiments. Perhaps one of the most pronounced instances of the influence of food is that shown by comparing the tests of May 18 and June 20, 1917. In the second observation it would appear as if the influence of food was clearly present, although the minimum value of 47 calories per kilogram of body-weight is but 10 per cent above the mmimum of May 18. After the age of 2 years, the food was almost invariably taken 4 to 5 hours prior to the observations and hence may be con- sidered as of comparatively little effect. After that age, the calories in the food represent not far from one-third to one-fifth of the daily requirement. Experience with adults would imply that at the end of 4 hours the peak of the stimulating effect of food would have been considerably passed. In none of these observations, therefore, have we the ideal condition of the post-absorptive state. The fact that in most cases the observations were made with the child asleep, and that in many instances, especially in the later years, the food had been given 4 or 5 hours before, undoubtedly minimizes and compen- sates the influence of the ingestion of food. Still, in any comparison of the metabolism of children at varying ages, it must constantly be 110 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. borne in mind that food is almost always taken just prior to a test with children under 2 years of age and that 4 or 5 hours may elapse between the taking of food and the beginning of the metabolism meas- urements for the older children. On examining the estimates for relative activity in the last column of table 25, one finds that the observations rarely show days when no periods with an activity of either I or II are available. In only two instances (November 15, 1917, and January 18, 1918) does the factor of persistent restlessness appear. The data for these two days are included in the table, however, to give a complete record of the obser- vations with this child. Furthermore, the values for the carbon- dioxide production and the pulse-rate for these active periods are not without interest. Almost invariably the high activities (IV, and particularly V and VI) are accompanied by a very great increase in the metabolism and like- wise an increase in the pulse-rate. The periods with activities V and VI usually correspond to the maximum heat production. But absolute reliance must not be placed upon these estimates of activity from the kymograph records, for not infrequently anomalous figures appear. Thus, on November 22, 1916, the last period has an activity of VI, with a total metabolism of 575 calories, which is exceeded by the first two periods of the day, and yet activities II and III are re- corded for these periods. As a rule, however, the metabolism is approximately proportional to the activity, and this factor is a valu- able index of the metabolism of a subject. The pulse-rate is a like valuable index, but taken by itself it is at times, especially with youth, very unreliable. The carbon-dioxide production, the pulse-rate, and relative activity taken together make the selection of minimum periods rarely a matter of great difficulty. When these are confirmed by sustaining figures on days immediately preceding or following, the base-line becomes even more definitely fixed. Occasionally low and seemingly aberrant figures appear in table 25. Thus, in the first experimental period on March 15, 1917, but 7.68 grams of carbon dioxide were measured, an amount corresponding to but 44 calories per kilogram of body-weight. This value is con- siderably lower than those found for the next three or four days of observation; yet on April 20, 1917, a heat production of 46 calories per kilogram is noted, and it is probable that 7.68 grams of carbon dioxide or 44 calories per kilogram of body-weight represents the true minimum value for March 15. While the observations with the child No. 145 were, with one exception, larger in number than those with any of the other children, they are given here in detail primarily to indicate the method of study and analysis of the results. Since we are dealing with an organism that is continually increasing in age, length, and weight, a comparison METABOLISM AS AFFECTED BY GROWTH. Ill of the metabolism values from day to day is best made by some graphic form of representation, such as a chart. This comparison has been made in figure 15, in which, to avoid confusion from a multi- plicity of points, the most important (19 in all) have been selected for plotting. In selecting the points for a chart to show the changes hi the metab- olism of an individual as the age increases, values for single days could not advantageously be used. They were accordingly not infrequently averaged, most of the points in this chart being made up of one or more days. Since a selection of data was necessary, which is always undesirable in the preparation of scientific reports, we have given in table 26 a summary of the actual minimum values and averages used in figure 15, to show exactly how these points were derived. Many of these values were likewise used on the general charts or group summaries (see figures 23 to 47, pages 135 to 175), those not so used being indicated by an asterisk. Thus, this child (No. 145) appears on the general charts nine times, or as nine different indi- viduals, since at all of these nine points there was sufficient differ- ence in either age or weight to meet our requirements for the designation of children as new individuals. We believe such points may properly be included in any general chart in which values for a number of children are represented. For all of the individual charts (see figures 15 to 21, pages 114 to 130), which were prepared primarily to indicate the basal metabolism at different ages of the same child, well-substantiated figures were invariably sought. In some instances it may be a little difficult for the reader to see why certain values are not selected or included in an average. For instance, the values obtained with the child No. 145 on June 21, 1917, appear in figure 15, while those for June 20, 1917 (see table 25), do not, although one might naturally expect that an average for these two days would be used. Since, as poined out on page 106, the possibility of an analytical error is always present, it can be seen that the value of 565 calories for the first period on June 21 is not substantiated, except that it agrees essentially with the low values on May 18. The minimum value on June 20 (627 calories) is noticeably higher than that found in two periods on June 21. Conse- quently, it was necessary here to make a selection of material, and it seemed logical to choose for charting an average of the three periods on June 21, i. e., 600 calories, rather than to use the unsupported minimum. On March 5, 1918, a very low value of 539 calories is noted in the first period, with 620 calories in the following period. Here again we have an unsupported minimum figure which it seems unwise to include, as there was a change in age of only 4 months, i. e., January to March. The value of 620 calories, although fitting well into the general curve, 112 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 26. Minimum heat production of No. 145 (girl). 1 Date of experiment. Age. Body- weight (with- out cloth- ing). Height. Body- surface (Du Bois linear for- mula) . Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 1916. kilos. 5.27 5.22 cm. sq. m. 119 118 cals. 315 318 cals. 60 61 cals. 1,036 1,046 June 16. ... Average. . Nov. 22 Dec. 4 Dec. 22 1917. Jan. 11 Jan. 31 Feb. 19 Mar. 15 Mar. 27.... Apr 20 5 mos 10 mos. 1 wk *10j mos. 5.25 62.0 0.304 119 317 61 1,041 9.19 9.50 10.2 10.5 10.9 11.3 11.8 12.1 70.0 70.9 72.5 73.9 73.2 76.1 76.0 75.8 .464 .465 .521 .514 .533 .549 .562 .584 117 111 109 103 105 113 105 108 474 505 508 477 548 564 518 601 52 53 50 46 50 50 44 50 1,022 1,086 975 928 1,028 1,027 922 1,029 *llmos. 3 wks *1 yr. 5 mo 1 yr. 2 mos *1 yr. 2 mos. 1| wks. 105 101 97 591 594 579 47 47 46 998 1,003 978 Average. . May 15. . . *1 yr. 3 mos 12.5 77.9 .592 101 588 47 993 12.6 .597 .620 92 99 541 542 43 43 906 874 May 18. . . . 12.5 Average. . June 21 Average. . Nov. 6 Average. . 1918. Jan. 19 *1 yr. 4 mos 12.6 79.0 .609 96 542 43 890 98 105 95 565 605 631 42 45 47 910 974 1,016 1 yr. 5 mos 13.4 80.0 .621 99 600 45 967 84 89 91 653 593 654 633 43 39 43 978 888 979 1 yr. 9 mos. 3 wks. . *2yre 15.1 86.5 .668 88 42 948 15.3 90.0 .692 80 582 38 841 May 16 May 22. . . . 15.8 15.7 .664 .666 67 69 598 581 590 38 37 901 872 Average. . Nov. 5 Nov. 13 2 yrs. 4 mos 15.8 94.0 .665 68 38 887 16.7 16.4 .692 .685 76 71 667 568 40 35 964 829 Average. . 1919. Jan. 16 2 yrs. 10 mos 16.6 96.5 .689 74 618 38 897 .680 .680 75 73 647 686 38 41 952 1,009 Jan. 17 Average. . Mar. 10 Mar. 17 *3 yrs 16.9 97.4 .680 74 667 40 981 17.5 17.5 .700 .691 .696 71 67 670 643 38 37 957 931 Average. . June 13 .... 3 yrs. 2 mos 17.5 98.5 69 657 38 944 72 72 637 682 36 39 909 973 Average. . *3 yrs. 5 mos 17.5 101.3 .701 72 660 38 941 1 Normal, breast-fed; adenoids removed at 1 yr.; developed to very large normal child. NOTE. The data indicated by asterisks (*) were not used on the general metabolism charts (figs. 23 to 47, pages 135 to 175) or on the anthropometric charts (figs. 5, 6, 8, 12, 13, and 14, pages 43 to 68) but were used, along with the other data in this table, on the individual chart for this child (fig. 15, p. 114). METABOLISM AS AFFECTED BY GROWTH. 113 is likewise not supported, although this is probably the true value for the basal metabolism. These selections seemed desirable in preparing the individual chart for No. 145, and it is for the purpose of showing investigators the basis for these and similar selections that the detailed figures for this child are given in table 25. If there is disagreement with our selection of the values shown in table 26 and used in the chart, other data can be chosen, but according to our experience, the selected figures are the most probable and best experimentally substantiated values for the basal metabolism of this child at the various age-levels. In the analysis of the figures in table 25, little attention can profit- ably be paid to the values of heat per kilogram of body-weight and per square meter of body-surface, other than to note that, in general, there is a disposition for the heat per kilogram of body-weight to decrease as the child grows older. An examination of the values for heat per square meter of body-surface shows very considerable vari- ations. Those variations ascribable to activity (for which note indices in the last column of the table) should be completely disregarded in any careful analysis. Consequently, the detailed data do not lend themselves to a comparison of values from day to day, from month to month, or from year to year. It may be noted roughly that with high activity there is almost invariably a high heat production per square meter of body-surface, but since we are interested primarily in the basal metabolism of children, i. e., the metabolism measured when activity is not present, those periods with high activity can have value only as indicating the possible maxima when the child is ex- tremely restless. The distribution and range of the basal values can much better be studied by means of the chart. The chart in figure 15 indicates the changes in body-weight, pulse- rate, total calories per 24 hours, calories per kilogram of body-weight, and calories per square meter of body-surface, from the age of 5 months to 41 months. Considering first the body-weight, since this, like the other children studied, was a normal child, we find the curve quite in line with that normally expected, namely, a definite progressive increase throughout the entire period, the most rapid period of change being from the fifth to the twenty-first month. The total calories, al- though showing considerable differences from tune to tune, range from 317 to 667 calories; in general the curve is reasonably parallel to the curve for body-weight. In other words, as the child regularly increased in weight, height, and age, the total calories per 24 hours likewise regularly increased. With so rapidly changing an organism as a growing child, compari- sons should be made at two ages or weights on some basis other than total calories. The two bases most commonly used by physiologists have been those corresponding to a unit of weight, i. e., the calories 114 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. per kilogram of body-weight, and to a unit of surface, i. e., the calories per square meter of body-surface. These have both been charted, the former showing a distinct tendency for a downward trend from 61 calories at 5 months to an approximate level at 38 calories between the twenty-fourth and forty-first months. It thus appears that per kilogram of body-weight the heat production of the infant is consid- erably greater than that of a child from 2 to 3 years of age. No.145(Fj 18 21 24 27 30 33 36 39 42 FIG. 15. Body-weight, pulse-rate, and basal heat production per 24 hours (No. 145). When the calories per square meter are charted (and it must be recalled that we have on this chart only values representing the minimum metabolism at the different ages) we find variations ranging from 1,086 to 841 calories. Although the curve is extremely irregular, there is a tendency downwards, with a slight rebound after 24 months. As this curve stands, it is apparently not so regular as that for the calories per kilogram of body-weight, and taken by itself gives very little evidence of a physiological law correlating the energy output and the body-surface of the child. It should, furthermore, be recalled that (in the case of this child, at least) the body-surface was not METABOLISM AS AFFECTED BY GROWTH. 115 computed by the erroneous formula of Meeh, nor was it even com- puted up to 10 kg. by the more exact formula of Lissauer, but it was actually determined from an elaborate series of Du Bois body- surface measurements. Even with these most favorable conditions for comparing the surface of the body with the heat production, we find very great irregularity instead of the constancy hopefully pre- dicted by earlier writers. The individual blocks indicating the height of the pulse-rate show a tendency for this factor to decrease with increasing age, a low level being reached at not far from 27 months, with but slight fluctuations thereafter. Were the observations of this series the only ones available for drawing final conclusions, the evidence accumulated in tables 25 and 26 and in figure 15 would warrant more refined analysis, with a more extensive series of deductions. The hint of great individual irregularities shown in all the curves except that of body-weight, and particularly the curve for calories per square meter of body-surface, would alone make us somewhat cautious in drawing general con- clusions. Since 22 other subjects were studied (though not so ex- tensively, as a rule, as in this particular case), it seems desirable first to consider giving the data individually by means of charts. These charts include children at ages ranging from below 5 months to more than 41 months, and of both sexes. Generalized conclusions may then be drawn from the whole series in addition to those tentatively drawn from the picture presented by the data for No. 145. OBSERVATIONS WITH 22 CHILDREN DURING PERIODS OF 4 MONTHS TO 3 YEARS. As the detailed results for the other 22 subjects studied over rela- tively long periods would require considerable space for tabular presentation, it did not seem desirable to give the data in full, and the basal metabolism of these children is therefore shown by means of the curves in figures 16 to 21. Tables 27 and 28 give the data used in the charts for boys and girls, respectively. These tables were pre- pared primarily to show the data used in plotting the general group charts (figures 22 to 47, pages 133 to 175), but additional data em- ployed in plotting the individual charts (figures 16 to 21) were like- wise included. As in table 26, the data not used in the general charts are indicated by an asterisk, but all of the data in these tables appear in the charts in figures 16 to 21 for the individual children. Two of these twenty-two children, Nos. 139 and 171, were studied during a period of time as long as that represented by the chart for No. 145, i. e., approximately 3 years. The others were studied for the most part during periods of a few months or a year or two. Since the charts for Nos. 139 and 171 are strictly comparable with that for No. 145, at least so far as period of study and elaborateness of measure- 116 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 27. Minimum heat production of boys at different ages, [Children normal unless otherwise stated. The data indicated by asterisks (*) were not used on the general metabolism charts for boys (figs. 22 to 45, pages 133 to 174) or on the anthropo- metric charts (figs. 3, 4, 7, 9, 10, and 11, pages 40 to 66), but were used, along with the other data in this table on the individual charts for these children (figs. 15 to 21, pages 114 to 130).] Subject No. Age. Body- weight (with- out cloth- ing). Height. Body- surface (Du Bois linear for- mula). Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 1 6 (F.R.) i 27 106 107 108 1 112 (M.D.) 114 115 117 118 119 ' 61 U24 (L.L.) *125 8 days kilos. 4.54 3.86 3.83 3.38 3.40 3.99 3.83 3.83 3.91 4.17 cm. 52.0 53.0 53.0 51.0 50.5 54.0 55.0 sq. m. 20.283 .253 .252 .232 .233 .260 * .252 .243 124 129 108 124 135 127 125 118 127 135 125 129 124 129 124 106 120 108 124 121 126 118 118 98 122 114 118 118 122 125 107 107 108 95 88 94 138 119 117 cals. 191 198 163 163 202 196 214 186 200 215 229 255 296 368 350 315 389 227 232 339 268 253 378 362 369 428 438 463 506 684 653 650 664 641 687 717 233 269 285 cals. 42 51 43 48 59 49 56 49 51 52 52 54 51 61 54 51 57 50 49 56 54 50 63 55 51 58 56 57 59 58 53 53 53 51 55 53 51 52 53 cals. 675 783 647 703 867 756 849 765 791 799 881 941 876 1,061 992 921 1,040 830 835 1,003 856 832 1,142 1,034 949 1,044 1,082 ,132 ,174 1,278 ,126 ,085 ,116 1,080 1459 1,149 818 878 891 83 days 8f days 10 days 11| days 17 days 1 mo 1 mo * lmos * 1 mo. 3 wks... . * 2 mos. . . . 4.38 4.71 5.86 6.01 6.45 6.23 6.83 4.54 4.71 6.08 4.96 5.10 6.00 6.56 2| mos 4j mos. 58.5 63.0 .271 .338 * 5 mos. 3 wks * 6 mos. 2 wks 65!6 55.0 58.5 65.0 58.5 64.5 .374 .273 .278 .338 .313 .331 7 mos. 3 wks 1 mo. 3 wks if mos 4 mos 1 mo. if wks * 1 mo. 3 wks 4 mos * 4 mos. 3f wks 6 mos 7.21 7.44 67.5 .389 * 65 mos * 7 mos 7.82 * 7f mos 8.08 8.51 11.8 12.4 12.4 12.6 12.6 12.5 13.6 4.60 5.13 5.39 7 mos. 3f wks. 70.5 79.5 .431 .535 1 yr. 1 mo. 3f wks. . . * 1 yr. 4 mos. * 1 yr. 4 mos. If wks. . * 1 yr. 6 mos * 1 yr. 7 mos. 1 wk. . . . 1 yr. 8 mos. &5.5 ' 90.5 56.0 57.0 58.5 .593 .624 2 .285 2 .306 .320 2 yrs. 1 mo 2 mos 2s mos. . . . 2f mos. . . . 1 Previously reported under initials given by Benedict and Talbot, Am. Journ. Diseases of Children, 1914, 8, p. 1. Data for boy designated by initials M. D. also reported by Bene- dict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. Results for Nos. 6, 27, and 61 reported under the same subject number by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915. 2 Lissauer surface. 1 Certainly breast-fed. Probably many others were, but we are uncertain. METABOLISM AS AFFECTED BY GROWTH. 117 TABLE 27. Minimum heat production of boys at different ages Continued. Subject No. Age. Body- weight (with- out cloth- ing). Height. Body- surface (Du Bois linear for- mula) . Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 1126 128 '129 (E.F.) '130 (A.S.) U32 1133 136 137 (R.E.) 138 141 6142 147 148 149 '(H.T.) '150 153 * 2 mos. 1^ wks 2 mos. 3 wks * 3 mos kilos. 5.79 5.90 6.12 7.49 8.24 6.32 7.07 6.02 6.18 5.81 6.70 8.19 8.93 9.08 10.6 11.2 cm. 60.5 sq. m. 0.333 117 105 105 109 103 117 111 113 119 132 117 126 120 103 121 112 114 132 128 111 103 106 120 98 103 94 128 110 101 116 102 112 106 111 113 111 123 109 108 111 115 105 101 105 107 109 cols. 276 267 290 370 413 296 311 305 318 329 365 445 454 461 588 622 324 414 473 577 590 592 656 645 642 658 382 372 340 456 437 496 367 347 391 379 411 412 467 512 537 559 420 428 378 396 cols. 48 45 47 49 50 47 44 51 52 57 55 54 51 51 55 55 64 52 55 60 59 57 60 57 57 56 58 41 37 43 41 45 57 53 59 57 59 56 53 56 56 50 45 59 52 50 cols. 844 802 838 925 979 841 828 888 923 1,006 980 1,042 1,002 1,034 1,155 1,229 1,070 1,007 1,121 1,274 1,161 1,128 1,247 1,166 1,166 1,163 1,058 793 702 874 825 902 987 916 1,000 1,024 1,087 1,046 1,074 1,145 1,173 1,073 912 1,108 976 966 6 mos 66.5 69.5 61.5 62.0 63.0? 64.0 62.5 67.5 71.0 73.5 76.5 .400 .422 * .352 * .380 * .341 .344 .327 .373 .427 .453 .509 * .303 * .411 7j mos 3 mos 3 mos 3 mos 3^ mos. 3 mos. 2j wks 6 mos 7 mos * 7 mos. 1 wk 9 mos *10| mos 4^ mos 5.04 7.98 8.56 9.54 10.1 10.3 10.9 11.3 11.3 11.8 6.55 9.14 9.21 10.7 10.6 10.9 6.50 6.55 6.66 6.66 6.92 7.39 8.87 9 15 60.0 68.0 74^5 79.0 so.5 4 mos. . * 6 mos. 10 mos. 3j wks .453 .508 .526 1 yr. 4 mos. 3| wks. . * 1 yr. 6 mos. 2 wks. . . 1 yr. 7 mos. 3 wks. . . * 1 yr. 9| mos * 1 yr. 10 mos. 3 wks. . 1 yr. 11 mos. 3 wks.. 5 mos. 82.0 62.0 67.5 .566 * .361 .470 5 mos. * 6 mos. 8 mos. 1 wks * 9 mos. 1 wk *10J mos 5 mos. 1 wks 5 mos. 1 wk * 5^ mos * 6 mos * 6 mos. l wks 7 mos. 70.0 63^6 66.0 .522 .373 .379 68.0 72.0 .394 .435 9.71 11.3 9.33 7.23 7.22 7.87 78.0 82.0 75.0 72.0 67.0 .458 .521 * .457 * .386 .388 5| mos 6 mos 6 mos. 1 1 wks * 1\ mos 1 Certainly breast-fed. Probably many others were, but we are uncertain. 8 Lissauer surface. 8 Previously reported by Benedict and Talbot under initials given, Am. Journ. Diseases of Chil- dren, 1914, 8, p. 1. Data for boys designated by initials E. F., A. S M R. E., and H. T. also reported by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. * Slow to walk because of weak musculature; flesh flabby. "Flesh slightly flabby; breast-fed. 'Above average weight. 7 Shortly recovered from cervical adenitis. 'Adenoids removed at 2 yrs.; backward mental development; breast-fed. 118 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 27. Minimum heat production of boys at different ages Continued. Subject No. Age. Body- weight (with- out cloth- ing)- Height. Body- surface (Du Bois linear for- mula) . Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 153 (cont.) '154 155 156 *157 (P.W.) 158 U69 161 '164 *(R.L.) 168 *(R.S.) "170 (E.G.) 175 176 177 182 186 187 8 mos. 3 wks * 9^ mos kilos. 8.44 8.56 8.77 10.2 10.3 10.8 11.2 12.5 12 9 cm. 68.6 sq. m. 0.435 106 107 124 99 105 105 106 98 103 91 94 126 . 117 110 117 107 90 126 120 120 125 101 91 131 124 116 108 114 105 83 88 85 79 84 83 75 116 121 144 124 115 108 106 83 100 82 76 68 cals. 453 484 547 553 606 648 657 667 664 668 769 452 407 413 486 711 665 457 439 383 425 377 405 518 535 586 572 612 541 588 650 622 623 671 720 719 413 429 473 503 455 531 479 587 720 649 800 716 791 cals. 54 57 62 54 59 60 59 54 51 49 56 57 62 62 64 56 50 63 62 51 57 51 54 61 61 61 60 62 55 59 55 51 49 53 55 56 59 64 58 50 59 53 51 47 58 45 52 37 38 cals. 1,041 1,078 1,159 1,095 1,139 ,161 ,188 ,152 ,139 ,117 1,263 1,041 1,070 1,076 1,180 1,252 1,118 1,092 1,147 950 1,057 936 983 1,154 1,168 1,213 ,197 ,195 ,067 ,153 ,201 ,134 ,109 ,192 ,274 ,255 1,104 1,086 1,129 988 1,140 1,115 1,046 1,057 1,245 1,067 1,244 883 951 * lyr. l|wks 1 yr. 6 mos 77.6 .505 * 1 yr. 7$ mos * 1 yr. 10 mos 2 yrs 80.0 84.0 .553 .579 2 yrs. 5 mos * 2 yrs 6^ mos 2 yrs. 9 mos. 3 wks. . * 3 yrs 2 wks 13.7 13 7 87.0 .598 6 J mos 7 mos 7.91 6.54 6.69 7.56 12.7 13.3 7.32 7.11 7.48 7.48 66.0 68.0 .434 .380 * 7^ mos 10 mos 2 yrs. 6 mos 71.5 90.0 93.5 65.5 64.0? 66.5 .412 .568 .595 .419 3 .381 .403 2 yrs. 10 moa 7 mos 7 mos 7 mos. . . * 7 mos. * 8 mos. 7.41 * 9 mos. 7.46 11 mos. 3 wks 8.48 876 70.5 .449 1 yr. 6 mos 9.55 9.50 9.94 9.99 10.0 11.7 12.1 12.7 12.7 13.1 77.0 .483 .510 .541 .562 * 1 yr. 7 mos * 1 yr. 8 mos. 3 wks. . . * 1 yr. 10 mos. 3 wks. . 2 yrs. 1 mo 2 yrs. 6 mos. 1 wk. . . * 2 yrs. 8 mos. 1 wk.. . 2 yrs. 10 mos * 2 yrs. 10 mos. 1 wk. . * 3 yrs. 1 mo. . . . 81.0 82.0 8416 * 3 yrs. 1 mo. 1 wk 7^ mos 1\ mos 12.8 7.03 6.70 8.09 10.1 7.58 9.94 9.37 12.6 12.3 14.6 15.5 19.3 20.6 66.5 67.5 70.5 77.5 71.0 74.0 74.0 82.0 87.5 88.5 94.0 110.5 111.0 .374 .395 .419 .509 3 .398 a .476 3 .458 .555 .578 .608 .643 .811 .832 9 mos. . . 1 yr. 2 mos 8 mos. . 9i mos. . 10 mos. . 2 yrs. 5 mos. 1 wk. . . 2 yrs. 5 mos. 1 wk.. . 2 yrs. 7^ mos. 4 yrs 4 yrs. 9 mos. 1 wk. . . 5 yrs. 3 wks. 1 Certainly breast-fed. Probably many others were, but we are uncertain. * Previously reported by Benedict and Talbot under initials given, Am. Journ. Diseases of Chil- dred, 1914, 8, p. 1. Data for boys designated by initials P. W., R. L., and E. G. also reported by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. 3 Lissauer surface. 4 At 17 mos. had diarrhea; lost weight, then gained; had frequent colds; normal development. 6 Approximately normal. 8 Overnourished. METABOLISM AS AFFECTED BY GROWTH. 119 TABLE 27. Minimum heat production of boys at different ages Continued. Subject No. Age. Body- weight (with- out cloth- ing). Height. Body- surface (Du Bois linear for- mula). Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. kilos. cm. 83. m. cals. cals. cals. 192 5 yrs. 6 mos. 3 wks.. 18.8 106.0 0.758 90 866 47 1,143 1193 5 yrs. 7 mos. 3j wks. . 23.7 118.0 .903 70 855 36 947 194 5 yrs. 9 mos. 1 wk. . . 19.8 107.5 .820 86 813 41 991 197 6 yrs. 9 mos. 3 wks.. 19.9 114.0 .795 99 925 47 1,163 U99 6 yrs. 10| mos 20.2 115.5 .812 72 843 42 1,038 201 7 yrs. 3^ wks 21.1 122.5 .865 83 864 41 999 7 yrs. 3 mos. 2j wks. . 24.4 124.0 .928 75 1,021 42 1,100 202 7 yrs. 2 mos 25.2 121.0 .940 74 944 37 1,004 204 7 yrs. 2 mos 19.9 111.5 .820 73 838 42 1,022 2205 7 yrs. 2 mos. 1 wk. . . 21.3 117.5 .887 78 899 42 1,014 '209 7 yrs. 11 mos 25.1 125.5 1.013 74 1,057 42 1,043 211 8 yrs. 1 mo. 1 wk. . . . 26.8 129.0 1.072 78 1,033 39 964 212 8 yrs. 1 mo. . . . 21.1 120.5 .888 69 785 37 884 <215 8 yrs. 2j mos. 20.8 116.5 .845 94 984 48 1,165 217 8 yrs. 6 mos 26.7 123.5 .971 80 1,097 42 1,129 6 218 8 yrs. 7 mos. 24.7 128.5 .956 65 932 38 975 9 yrs. 5^ mos 26.8 133.5 .988 78 1,054 39 1,067 222 9 yrs. 3 wks 25.0 122.5 .942 71 1,038 42 1,102 '223 9 yrs. 1 mo. 2 wks. . . 25.9 129.0 .991 86 1,011 39 1,020 2224 9 yrs. 3 mos. 3? wks. . 25.4 126.0 1.027 77 959 38 934 ^228 9 yrs. 9 mos. 3 wks. . . 28.5 126.5 1.003 79 1,209 43 1,205 229 9 yrs. 11 mos. . . 30.1 128.0 1.054 76 1,092 36 1,036 232 10 yrs. 4 mos 28.1 127.0 1.087 80 1,037 37 954 235 10 yrs. 7 mos. 1 wk. . . 30.3 134.0 1.088 85 1,015 34 933 236 10 yrs. 8 mos. 3| wks. . 31.0 132.5 1.084 76 1,147 37 1,060 237 10 yrs. 9 mos 33.6 139.5 1.213 73 1,192 36 983 240 11 yrs. 1 mo. lj wks.. 33.8 138.5 1.244 1,230 36 989 241 11 yrs. 1^ mos 30.6 136.0 1.100 '"77 1,086 36 987 242 11 yrs. 2 mos. lj wks. . 26.8 126.0 .985 80 1,117 42 1,134 243 11 yrs. 3 mos. 1 wks.. 37.9 149.5 1.256 69 1,282 34 1,021 244 11 yrs. 4 mos 29.5 132.0 1.083 63 1,039 36 960 "245 11 yrs. 5j mos.. 31.7 135.5 1.165 61 1,213 38 1,041 246 11 yrs. 6 mos 36.9 150.5 1.245 67 1,283 35 1,031 2247 11 yrs. 8 mos. 1 wk.. . 30.5 141.0 1.155 68 1,023 34 886 2249 11 yrs. 11 mos. 3 wks. . 29.9 135.5 1.187 63 1,087 37 915 250 12 yrs. 1 mo. 2? wks. . 41.0 150.5 1.370 61 1,211 30 884 252 12 yrs. 3 mos. 2 wks.. 34.1 139.0 1.245 66 1,167 34 937 253 12 yrs. 7 mos. 3 wks. . 30.4 140.0 1.085 73 1,163 38 1,072 254 12 yrs. 7 mos. 3 wks.. 39.0 151.0 1.333 66 1,163 30 872 "255 12 yrs. 8 mos 37.9 153.0 1.303 67 1,096 29 841 Z 256 12 yrs. 8 mos. 2 wks. . 32.8 137.5 1.221 72 1,246 38 1,020 12258 13 yrs. 8 mos 50.8 159.5 1.494 63 1,481 29 991 "259 14 yrs. 1 mo 38.2 151.5 1.335 61 1,200 32 899 "260 15 yrs. 1 wk 39.0 147.0 1.300 72 1,401 36 1,079 1 Tall, over-nourished. 2 Slightly undernourished. 3 Tall. 4 Somewhat thin. 5 Tall, thin. 6 Overnourished. 7 Fairly well developed and nourished ; slightly under par. Below par mentally. Very muscular. Stupid, overnourished. 1 Tall, gawky, slightly undernourished. 2 Overnourished; tall; puberty. 3 Thin, slightly undernourished. 'Puberty (?). 120 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 28. Minimum heat production of girls at different ages. [Children normal unless otherwise stated. The data indicated by asterisks (*) were not used on the general metabolism charts for girls (figs. 23 to 47, pages 135 to 175) or on the anthropo- metric charts (figs. 5, 6, 8, 12, 13, and 14, pages 43 to 68) but were used, along with the other data in this table, on the individual charts for these children (figs. 15 to 21, pages 114 to 130).] Subject No. Age. Body- weight (with- out cloth- ing). Height. Body- surface (Du Bois linear for- mula). Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. i 2 i 12 i 26 1 35 i 48 49 109 U10 (E.P.) Ill <113 116 KA.C.) <120 KB.D.) 122 123 127 10 days kilos. 3.73 4.20 3.56 4.42 5.07 7.17 4.81 5.54 2.68 3.86 3.71 3.57 3.65 3.99 440 cm. 53.0 53.0 50.0 54.0 58.5 64.5 56.0 61.0 48.5 51.5 51.0 53.0 53.0 sq. m. =0.248 .268 .239 .278 .304 .384 .293 .323 .199 .253 .247 * .240 .249 96 133 110 117 134 125 127 139 116 114 122 147 141 135 129 134 119 133 - 120 115 120 126 122 126 135 136 136 127 129 131 129 117 124 114 112 114 115 127 cats. 152 199 185 198 223 329 211 388 139 200 182 200 173 178 211 207 217 253 289 324 351 163 274 257 299 329 597 653 235 251 280 253 312 325 334 410 255 468 cafe. 41 47 52 45 44 46 44 70 52 52 49 56 47 45 48 45 44 51 53 54 54 55 57 50 53 55 59 63 61 57 63 55 51 54 53 61 51 58 cols. 611 743 774 712 734 857 720 1,201 698 791 735 833 695 695 754 745 728 861 871 967 975 759 927 830 911 948 1,161 1,251 963 923 1,098 891 920 962 957 1,188 876 1,125 9 days 10 days 8 days 1 mo. 1 wk 4 mos 1 mo. 1 wk 2 mos. 3 wks 11 days . . . 12^ days. . 13 days . . 13 days 3f wks. * 5 wks. * 1| mos 1 mo. 3 wks * 2 mos 4.55 4.88 4.98 5.54 6.04 6.49 2.99 4.90 5.15 5.62 6.03 10.1 10.4 3.85 4.37 4.44 57.5 .278 4 mos * 4 mos. 3? wks 5 mos. 3 wks 63.0 .332 66.5 .360 * .214 * .297 .309 1^ mos 2 mos 58.0 58.5 2 mos. 1 wk * 3 mos.. . . 3 mos. 3| wks 1 yr. 6 mos 62.5 78.5 53.5 .347 .514 .244 * 1 yr. 6 mos. 3 wks. . 2 mos. 1 wk 2j mos. 2 mos. 3| wks 3 mos. 1 wk 4.59 6 mos. 1 wk 6 mos. 3 wks 7 mos. 1 wk 8 mos. 3 wks 2 mos. 3? wks 9^ mos 6.09 6.06 6.34 6.75 5.03 8.11 63.0 65.6 57.5 67.0 .339 .345 .291 .416 1 Previously reported under initials given by Benedict and Talbot, Am. Journ. Diseases of Children, 1914, 8, p. 1. Data for girl designated by initials A. C. also reported by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. Results for Nos. 2, 12, 26, 35, 48, and 49 reported under-the same subject number by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915. 1 Lissauer surface. * Data as previously published incorrect. 4 Certainly breast-fed. Probably many others were, but we are uncertain. 6 Slightly undernourished. 8 During second year had colds; continued to develop normally; breast-fed. METABOLISM AS AFFECTED BY GROWTH. 121 TABLE 28. Minimum heat-production of girls at different ages Continued. Subject No. Age. Body- weight (with- out cloth- ing). Height. Body- surface (Du Bois linear for- mula). Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 127 (cont.) 131 134 (L. R.B.) *135 (M.C.) <139 140 '144 6. 6 146 H51 152 160 *11 mos. If wks kilos. 7.98 cm. sq. m. Ill 115 108 111 114 93 111 122 132 118 119 120 119 121 124 106 103 136 121 129 119 124 135 129 124 131 113 128 116 109 99 110 101 100 93 85 78 80 84 85 115 121 112 119 119 126 117 123 126 cols. 453 500 486 485 549 474 235 279 277 311 335 314 340 351 353 331 333 315 280 339 314 325 406 419 484 528 491 516 511 531 525 590 549 607 579 655 608 601 656 624 334 353 445 360 419 355 357 417 444 cdU. 57 61 56 57 61 53 54 60 57 59 60 55 58 59 58 55 54 61 53 59 52 53 58 51 54 55 50 51 51 50 47 48 43 45 44 47 43 41 46 43 66 45 42 43 46 63 55 71 70 cols. 1,051 1,131 1,063 1,025 1,138 1,019 933 1,045 945 1,003 1,074 928 1,005 1,057 1,057 973 967 1,006 864 1,000 952 926 1,060 997 1,094 1,110 989 1,024 987 986 948 1,000 895 936 930 1,026 953 918 1,019 937 1,047 835 856 847 984 1,086 997 1,219 1,251 * 1 yr. If wks * 1 yr. 1 mo 8.28 8.62 * 1 yr. 2f mos 8.56 9.06 8.89 4.34 4.64 4.86 5.27 5 55 1 yr. 4 mos 73.0 0.482 3 mos. 55.5 68.6 .252 .310 * 3 mos. 1 wk * 3 mos. 3 wks 4 mos 1$ wks. * 5 mos * 6 mos. 1 wk 5.76 5.87 5.97 6.08 5.99 6.17 5.19 5.27 5.75 * 6 mos. 3f wks * 7 mos 7 mos. 2f wks 4 moa 62.6 64.0 63.0 63.0 .333 .340 .347 .313 4 mos 4f mos * 6 mos. 5.99 6 mos. 1 wk 7 mos 6.11 7.00 8.29 9 03 65.0 67.0 70.0 .351 .383 .420 9 mos. 1 wk *10f mos 1 yr. 2 mos. 3 wks. . . * 1 yr. 4 mos * 1 yr. 5 mos * 1 yr. 6f mos 1 yr. 8f mos * 1 yr. lOf mos 2 yrs. 2f mos. 9.67 9.81 10.2 10.0 10.8 11.1 12.3 12.8 13.6 13.4 14.0 14.3 76.0 82.6 .476 .538 88.0 92!5 .590 .648 * 2 yrs. 5 mos 2 yrs. 6 mos. 3 wks. . * 2 yrs. 8 mos. If wks. . 3 yrs. 2 mos. 3 wks. . 96.0 .638 14.7 * 3 yrs. 7 mos. 3 wks. . 3 yrs. 9 mos. 3 wks.. 4 mos. 3 wks 5 mos 14.4 14.7 6.02 7.91 10.6 8.30 9.04 5.64 6.52 5.90 6 31 99!6 60.0 62.0 67.5 68.0 71.0 60.0 65.5 62.5 '.666 ' .319 .423 .520 .425 .426 .327 .360 .342 9 mos 5 mos. 1 wk 7 mos 6 mos 6 mos 7f mos 1 Slight facial eczema. 2 Previously reported by Benedict and Talbot, Am. Journ. Diseases of Children, 1914, 8, p. 1. Data for girls designated by initials L. R. B., and M. C. also reported by Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 201, 1914. 3 Lissauer surface. 4 At 3 yrs. had measles; no other illness except colds; breast-fed. * Certainly breast-fed. Probably many others were, but we are uncertain. See table 26, p. 112, for data for No. 145. 122 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 28. Minimum heat production of girls at different ages Continued. Subject No. Age. Body- weight (with- out cloth- ing). Height Body- surface (Du Bois linear for- mula) . Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 160 (cont.) 162 163 165 *166 167 *171 <172 '173 kilos. 7.05 7.63 8.12 8.11 8.41 8.15 8.00 7.63 6.24 7.92 7.98 8.62 9.21 9.45 11.1 11.4 11.6 11.8 cm. 65.5 sq. TO. 0.375 126 114 113 117 119 111 119 116 93 116 119 126 109 105 96 105 84 97 91 95 cals. 492 522 522 537 530 505 413 375 338 505 482 557 559 570 597 625 603 621 655 692 710 741 686 686 522 502 557 545 606 613 612 664 643 638 735 649 666 657 718 713 568 608 686 677 697 647 712 600 573 552 cals. 70 68 64 66 63 62 52 49 54 64 60 65 61 60 54 55 52 53 55 53 53 54 50 49 61 61 66 63 64 65 63 66 61 58 60 53 55 46 44 44 65 65 70 68 68 62 64 65 61 56 cals. 1,312 1,286 1,283 1,313 1,283 1,268 1,002 938 914 1,238 1,100 1,255 1,152 1,171 1,120 1,143 1,075 1,093 1,128 1,155 1,239 1,260 1,179 1,138 1,140 1,210 1,289 1,310 1,303 1,330 1,330 1,425 1,302 1,236 1,269 1,106 1,164 1,053 1,054 1,059 1,246 1,277 1,350 1,317 1,377 1,242 1,292 1,300 1,170 1,131 * 1 yr 1 yr. 3j wka 68.5 .407 * 1 yr. 1? mos * 1 yr 2 mos 8 mos 8 mos. 1 wk 8 mos. 3 wks 9 mos. 1 wk *10 mos. 3 wks * lyr. 1^ wks 1 yr. 2| mos * yr. 4 mos yr. 8 mos. 3^ wks. . * yr. 10 mos * yr. 11 mos. 3 wks.. * 2 yrs. 2 mos 69.5 63.0 63.0 68.5 i .412 i .400 .370 .408 74.5 .485 80.0 .533 2 yrs. 4 mos. 12.0 13.2 13.3 88.0 88.5 .581 .599 2 yrs. 9^ mos * 2 yrs. 11 mos. * 3 yrs. 1^ mos. 13.7 88 79 126 123 140 129 141 124 110 123 107 114 109 102 99 93 85 89 116 121 122 123 116 111 116 126 113 119 * 3 yrs. 2 mos. 13.7 14.0 8.52 8.18 8.40 8.70 9.43 9.50 9.75 10.1 10.6 11.0 12.2 12.3 12.1 14.2 16.5 16.2 8.80 9.30 92.5 69.0 73.5 76.5 .603 .458 .415 .465 9 mos. 1 wk 10 mos *10$mos * 1 yr. 1 J mos 1 yr. 2 mos. 1| wks. . * 1 yr. 3^ mos * 1 yr. 4 mos. 3 wks. . * 1 yr. 5 mos. 1 wk. . . . 1 yr. 9| mos 85.5 .494 * 1 yr. 11 mos 2 yrs. 3 mos. 1 wk. . . * 2 yrs. 4 mos. 1 wk. . . * 2 yrs. 5 mos 3 yrs. 3 mos. 1 wk. . . 4 yrs. 2 mos. 1 wk. . . * 4 yrs. 3 mos. 2 wks. . 11 J mos 89.5 .579 100.0 104.5 .624 .681 74.5 .456 * 1 yr. 1^ wks 1 yr. 1 mo. l wks. . . * 1 yr. 2 mos. 1 wk. . . 9.84 10.0 77.0 .508 * 1 yr. 3 mos 10.2 * 1 yr. 4 mos 1 yr. 5 J mos 10.4 11.1 9.22 9.46 9.98 ' 80.0 ' 72.0 .551 .461 11^ mos * 1 yr. 1 mo * 1 yr. 2 mos 1 Lissauer surface. 2 At 1 yr. 4 mos. had measles; no other illness except colds; breast-fed. 3 At 1 yr. 1 mo. had chicken pox; no other illness except colds; developed normally; breast-fed. 4 Adenoids removed when 1 yr. old. s During second year had colds. METABOLISM AS AFFECTED BY GROWTH. 123 TABLE 28. Minimum heat-production of girls at different ages Continued. Subject No. Age. Body- weight (with- out cloth- ing). Height Body- surface (Du Bois linear for- mula). Aver- age pulse- rate. Heat (computed) per 24 hours. Total. Per kilo. Per sq. m. 173 (cont.) 174 178 '179 180 181 2183 184 188 189 190 "191 <195 '196 198 2203 "206 210 "214 219 3220 221 225 '227 7 230 8233 234 238 239 "248 "251 257 * 1 yr. 3 mos 1 yr. 5 mos * 1 yr. 6| mos 2 yrs. 1 mo. ... kilos. 9.81 10.6 11.0 11.0 12.4 14.5 15.0 16.4 15.7 16.2 16.6 22.7 15.2 18.7 16.4 19.7 26.4 23.1 19.2 20.8 23.9 26.0 26.0 23.7 22.5 26.1 24.0 24.8 27.9 29.8 28.2 28.0 27.4 39.2 28.8 30.9 37.1 cm. "78.0' sq. m. 0.522 109 109 97 102 88 76 92 88 77 92 110 79 81 85 91 73 74 74 71 86 79 65 81 97 76 79 82 77 68 74 90 76 72 75 78 74 cols. 554 614 622 604 543 560 640 771 673 715 782 829 637 790 792 747 918 849 752 863 894 1,002 930 880 977 902 924 919 999 896 923 944 984 1,500 1,062 1,012 1,318 cols. 56 58 57 55 44 39 43 47 43 44 47 37 42 42 49 38 35 37 39 42 38 39 36 37 43 35 38 38 36 30 33 34 36 38 37 33 36 cols. 1,140 1,176 1,154 1,112 894 856 926 1,100 1,028 999 1,075 951 922 1,056 1,191 915 926 964 956 1,031 961 1,014 959 954 1,042 883 1,020 983 944 844 894 902 909 1,179 1,026 920 1,079 79.0 92.0 98.5 93.5 98.5 97.5 103.0 103.5 116.0 103.5 107.5 99.5 118.0 124.5 119.0 113.0 116.0 122.5 123.5 126.0 125.0 122.0 122.0 120.5 125.5 133.0 131.0 133.0 135.5 133.5 147.5 129.0 138.5 140.5 .543 .608 .654 .691 .701 .654 .716 .728 .872 .691 .748 .665 .816 .991 .881 .787 .837 .930 .988 .970 .922 .938 1.021 .906 .935 1.058 1.062 1.033 1.047 1.083 1.272 1.035 1.100 1.222 2 yrs. 11 mos. 3 yrs. 8 mos. 3 yrs. 10 mos. 3 wks. . 4 yrs. 3 mos. 3 wks.. 4 yrs. 4 mos. 1 wk. . . 5 yrs. 1^ mos 5 yrs. 3 mos. 1 wk. . . 5 yrs. 3| mos 5 yrs. 5| mos 6 yrs. 3 wks 6 yrs. 5? mos. 6 yrs. 9 mos. . 7 yrs. 1 mo. 2 wks.. . 7 yrs. 4 mos 8 yrs. 2 wks 8 yrs. 5 mos. 3 wks. . 8 yrs. 2 mos 8 yrs. 11 mos. 1 wk. . 9 yrs. y mo. . . 9 yrs. 3i wks 9 yrs. 5 mos. 1 wk. . . 9 yrs. 9 mos 10 yrs. 3 mos 10 yrs. 5 mos. 2| wks. . 10 yrs. 5 mos. 3j wks. . 10 yrs. 9 mos. 3 wks. . 11 yrs 12 yrs. 1 mo 11 yrs. 10 mos. 3 wks. . 12 yrs. 2 mos 13 yrs. 3 mos. 3 wks.. 1 Flesh slightly flabby. 2 Overnourished. 3 Slightly undernourished. 4 Well developed and nourished, with chronic endocarditis. 6 Tall, thin. 6 Weak ankles, the result of old infantile paralysis. 7 Thin. 8 Overnourished; puberty. 9 Puberty established at 12 yrs. 1 mo. 10 Fat, healthy, rosy-cheeked. 11 Puberty beginning. ments are concerned, it is of interest to consider them somewhat in detail. It is important to note that the scale upon which the charts for these children are drawn is somewhat different from that used for No. 145. (Compare figs. 15 and 16.) 124 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. The individual points in these charts are so numerous as to produce curves that are by no means smoothed. The general trend of the body-weight curves of both Nos. 139 and 171 is that of a progressive increase, as is common with a normal infant. The total calories per 24 hours fully substantiate the general line noted for No. 145, namely, a rather rapid rise until about 2 years of age, after which there is a tendency for a clearly defined rise, although at a somewhat slower rate. No.171(F.) No.139(F.)r FIG. 16. Body-weight, pulse-rate, and basal heat production per 24 hours (Nos. 171 and 139). While emphasis must again be laid upon the inherent errors in the measurement of individual periods, an attempt has been made to select only such points as were fully substantiated by the results for another period on the same day or on the day immediately preceding or following. Consequently in these charts the individual points can for the most part be considered as truly representative of the metabolic plane at the time of measurement. A high point in the total calories is never based upon a single experimental period. This is emphasized to bring out the fact that with children the regularity in the heat production from day to day is not perfect on any basis, and rather considerable fluctuations may normally be expected to obtain, even in periods with complete muscular repose and (though not with a true post-absorptive condition) at least with the influence of food very considerably minimized. The calories per kilogram of body-weight for these two children (Nos. 139 and 171) show a fall (as the weight and age increase) quite in conformity with that noted with No. 145. The calories per square meter of body-surface for both the children show a maximum occurring from about 1 year 4 months to 1^ years, and a tendency towards a fall thereafter. With No. 171, for example, the maximum value is 1,425 METABOLISM AS AFFECTED BY GROWTH. 125 calories per square meter of body-surface and the minimum 1,053 calories. With No. 139 a greater uniformity is observed, the range being from 1,110 to 864 calories. The curve for No. 171 is sharply distinguished from the curve for No. 145, in that it shows an early -No.122(F.)- No.119(M.)_ No.127(F.)r 60 1100 TaW 7 -No.138(M.)- A ER SQ FIG. 17. Body-weight, pulse-rate, and basal heat production per 24 hours (Nos. 119, 122, 127, and 138). period of low values for the calories per square meter of body-surface which does not appear in the chart for No. 145. The pulse-rate, high at about 1 year, gradually falls off with both Nos. 139 and 171, and fully confirms the observations drawn from the chart for No. 145. 126 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. Intimate analysis of the other 20 charts is hardly necessary. Since, however, the charts thus far analyzed show no special values in the earlier months of life, at least the physiological relationships at this period should be pointed out. Attention is called to the charts for No. 113(F.)- 950 900 | k 51' 450 850 48 400 45 350 300 X X / P C E A R L |ra / V . V CALOR E o S . / x* OTAL ( ALORIE 5 N/ ^ ^ '''BOG Y. WEIG ^ L . V A y / Cals per kilo. 46 / S/ ir 750500' 43 \ss- SQ E M. CALOF IES PEI KILO^ / 700450 40 \ / / /X 2 X 11 400 37 \ / / TOTAL CALOR IES Pulse 10 350100 N / -| x - ,-^- X W B E^ FIQ. 18. Body-weight, pulse-rate, and basal heat production per 24 hours (Nos. 113, 126, 131, and 142). those children with whom observations were made at an early age, such as Nos. 113, 115, and 119. (See figures 17, 18, and 19, pages 125 to 127.) In these charts the weight-curve and total calories present the usual features, namely, progressively increasing weight and progressively increasing total calories. The increase in the values for the calories per kilogram of body-weight and per square METABOLISM AS AFFECTED BY GROWTH. 127 meter of body-surface with an increase in age is worthy of consider- ation. With No. 115, for example, the heat per kilogram of body- weight rises from 49 calories at the age of 1 month to 61 calories at the age of 5 months, the whole trend of the curve being upwards. When measured on the basis of body-surface, the heat is as low as 765 calories at 1 month and rises to 1,061 calories at the end of 5 months. No.115(M.)- No.136(M.)- No.160(F.) CALORIES PER SQ. M. FIG. 19. Body-weight, pulse-rate, and basal heat productionjper 24 hours (Nos. 115, 123, 136, and 160). An examination of other charts, such as that for No. 113 (fig. 18) and those for the younger children, shows similar general trends, namely, low heat values on the bases of body-weight and body-surface, with a gradual increase as the age advances from 7 months to 1 year. Of special significance is the fact that these wide variations in the 128 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. heat production per square meter of body-surface are found with the very young children. This is in conformity with variations observed by us in our studies of new-born children, in which it was noted that the heat production per square meter of body-surface decreased at times to 459 calories. 1 When our early analysis of the figures for the heat production of new-borns was made, our use of the Lissauer formula for computing the body-surface was obviously open to some criticism. Subsequently, as shown in another section of this report, it was found that the Lissauer formula gives measurements agreeing admirably with those of Du Bois for children weighing up to 10 kg. During this early period from birth to 6 or 7 months, the body-weights are for the most part under 10 kg. Hence we have every reason to believe that our estimates of the heat production per square meter of body-surface are as close as can possibly be made in the present state of physiological science. The pronounced individuality of the children studied in these longer series, as evidenced by the fluctuations in the smoothed curve, are altogether too great to permit any consideration of a normally pro- gressing, increasing metabolism. It is thus difficult to draw general deductions from this extensive series of charts. Those permissible are, first, the normally increasing body-weight common to all normal children; second, the reasonably close paralleling of the total calories with the body-weight curve, namely, an increase in total calories with an increase in body-weight; third, the low calories found on the bases of body-weight and body-surface shortly after birth, increasing to a maximum not far from one to two years of age, with a tendency for a definite decrease thereafter. Finally, the pulse-rate is noticeably highest in the period from birth to 1 or 1| years, with a tendency to fall thereafter. No perfect picture of the general physiological trend can possibly be made from a visualization of these several groups of data for the individual children. Final recourse must therefore be made to our main study of a large number of children of different ages, weights, lengths, and body-surfaces, so as to plot all values on large charts and thus study the general trend. METABOLISM DURING GROWTH AS SHOWN BY GROUPS OF INDIVIDUAL DATA. In the preceding section it has been made clear that, with a given individual, there is no smoothed curve for metabolism like that, for example, obtained for body-weight with a fasting dog 2 or, indeed, with a fasting man. Daily basal metabolism is subject to very considerable fluctuations, which are nowhere more strikingly shown than in the charts for individual children in figures 15 to 21. Plotting all the 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915, pp. 96 and 100. 2 Benedict, Carnegie Inst. Wash. Pub. No. 203, 1915, pp. 77 and 75. METABOLISM AS AFFECTED BY GROWTH. 129 points on the charts in one scatter-diagram and sketching a curve indicating the general trend would obviously smooth the individual differences, not only for the same child from day to day, but between the children studied. This we have not done, but recognizing the value of grouping all observations in a series of charts so as to present the general trend of metabolism of a relatively large number of chil- No.172(F.) No.173(F.) FIG. 20. Body-weight, pulse-rate, and basal heat production per 24 hours (Nos. 148, 161, 172, and 173). dren, and thus visualize the influence of sex, age, weight, and surface on metabolism, we have gathered together not only the data from the 23 individual charts in figures 15 to 21, but also the isolated observa- tions on a large number of children, mostly of the higher ages, and plotted these values in several scatter- diagrams. In these diagrams the caloric output is referred respectively to age, weight, and surface area in a number of ways. 130 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. In studying these diagrams, we may ask: Is the metabolism of a child of given height, age, and weight the same as that of another child of the same height, age, and weight? Are there individual differences in metabolism? What is the influence per se of height, weight, age, and area on metabolism? What are the sex differences? -No.155(M.) No.158(M.)- FIG. 21. Body-weight, pulse-rate, and basal heat production per 24 hours (Nos. 153, 155, 158, and 166). All these problems may only adequately be studied by an intelligent comparison of extensive series of metabolism measurements. The individual curves show the general relationships between growth and metabolism, but they give information only obscurely and indirectly which may be used for a comparison of one child with another on the basis of age, height, and weight, and throw no light on sex differences. METABOLISM AS AFFECTED BY GROWTH. 131 When children are considered as a class, the gross differences in age, weight, and stature make it extremely difficult to find two children of exactly the same age, weight, and height. Certainly any series of metabolism measurements would have to be greatly extended and include a large number of subjects to secure two individuals who were strictly comparable in these respects. The time and expense required for such measurements would prohibit any attempt to make sufficient studies with a large number of individuals for the establishment of probable standards for the many combinations of the four variables sex, age, weight, and height. This likewise holds true for adults; and yet it is very important, physiologically at least, to have some conception of differences in metabolism with different individuals. METHOD OF GROUPING DATA. Any plan for the comparative study of the metabolism of children involves one or more forms of classification. Following the custom of physiologists, we have charted the values first on the basis of age, then of body-weight, and finally of body-surface. Since with adults it has been clearly shown that there is a sexual differentiation, it seems desir- able to consider the boys and girls separately, even though a critical analysis of the data for new-born babies 1 showed no sexual differen- tiation during the first week of life. A comparison of the various individual charts (figs. 15 to 21), in which the sex was indicated, or even a superficial inspection of these general charts (figs. 22 to 47) for boys and girls, gives but little, if any, suggestion of a sexual differ- entiation. Such grouping is of value, however, for use in a more thorough comparison of all the data in several large plots (see figs. 48 to 51, pages 179 to 181), by which the differences in the results due to sex may be discerned. As previously stated on page 101, during the period of growth represented by the observations in this study there are such rapid changes in weight and age that it has been considered perfectly legiti- mate, when appreciable variations in these two factors occur, to regard the child as a different individual and to plot the values accordingly in these group charts. In determining the points on the charts in figures 22 to 47, each child was considered a new individual after (1) an increase in weight of 1 kg. for children weighing less than 10 kg. ; (2) an increase in weight of 10 per cent for children over 10 kg. ; (3) an increase in age of 6 months. Change in height per se was not considered. While the classification for weight and age was not strictly adhered to, there was but little deviation from the rule. So far as possible, breaks in the continuity of evidence were avoided, and if an observation was made at the end of a series, it was not necessarily Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915; Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919. 132 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. excluded, even if the full change in weight or age had not been reached. It frequently happened that the full change in weight occurred con- siderably inside of 6 months, and there were also instances when a change in age of 6 months was not accompanied by a full change in weight. By this method, certain children, notably No. 145, appear a number of times upon the charts or scatter-diagrams. For instance, in figure 22, the actual number of boys represented is 88, while the number of points plotted is 129. These latter in reality correspond to an equal number of children, since the values found on the same boy are characterized by an appreciable difference in age or weight. A biometrical analysis of the data in this series of observations has not been attempted, but on each of the group charts an effort has been made to indicate the apparent general trend of the metabolism on the basis selected for comparison by laying on arbitrarily a smoothed curve. It must be emphasized here that these curves do not represent mathematically determined trends, but are simply sketched from observation of the general distribution of the points. The curves were prepared in the manner previously referred to (see page 37), i. e., five members of the Laboratory staff, accustomed to plots and curves, each drew on a separate sheet of tracing-paper a curve which appeared to him as the most probable. These curves were then com- bined and the line reproduced on the respective charts represents the average of these five plots. While this procedure is admittedly un- mathematical, it serves at least to indicate the general trend of the metabolism. All of the precautions cited in our discussion of normality entered into the selection of the individuals and points plotted in these group charts. In a preliminary communication published elsewhere 1 re- garding this study of children during the period of growth, a series of charts was given similar to those in figures 22, 23, 26, 27, 30, 31, 35, 36, 42, 43, 45, and 47. The earlier charts differed only in the number of points included, which was somewhat greater than in the present series, as a more rigid exclusion of material was made previous to the final printing here. A comparison of the two sets of smoothed curves brings out the interesting fact that those obtained in the preliminary charting of the values, which were prepared in much the same manner as the later curves, do not differ by a measurable amount from the curves subsequently sketched for the final series. In other words, the eliminations made exclusively on the basis of a more critical exam- ination of the protocols and histories for evidence of the normality of the children have resulted in the removal of an approximately equal number of points above and below the line, so that the position of the line itself is not materially changed. This fact is of considerable 1 Benedict, Boston Med. and Surg. Journ., 1919, 181, p. 107. METABOLISM AS AFFECTED BY GROWTH. 133 practical as well as physiological interest as indicating that, in an attempt to secure a more nearly perfect measure of normality, the evidence as to the general trend of the metabolism has not been affected. One can thus, even at this point, make the deduction that with the general population as studied, the deviations above or below a central tendency are such as to balance, and that the influence upon metabolism of slight deviations from physical normality is negligible. GENERAL TREND OF METABOLISM WITH INCREASING AGE. In this series of group charts, the first comparisons have been made for the total metabolism as referred to age. TOTAL CALORIES PER 24 HOURS REFERRED TO AGE (BOYS). The total calories per 24 hours referred to age for the boys have been charted in figure 22. An inspection of this chart shows a rapid rise in the total metabolism during the first year of life and from the Cals. TOTAL CALORIES REFERRED TO AGE. BOYS. 1500 1350 1200 1050 900 750 600 450 300 150 ^ * * __,.-- . . ^> <* . . ^ ^ . * ^~ ' * ^** ^ * ^ i t- tn. 1 2 3 4 3 6 7 10 n i 2 13 14 15 1 FIG. 22. Basal heat production of boys per 24 hours referred to age. Point inclosed in square signifies puberty established. first to the thirteenth year a somewhat slower but steady increase. This general trend seems to be continued beyond the thirteenth year; but only three points are available for comparison beyond this age. The interpretation of the sketched curve is beset with a number of difficulties. In the first place, the arbitrary laying-down of a smoothed curve on a plot of this character gives too much weight to the possi- bility of a constancy in metabolism. That this constancy is not actually present is clearly shown by the deviations from the line all 134 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. along the curve, particularly after 8 months of age. It would appear as if the metabolism during the first 8 months followed with singular accuracy the direction of the curve. At this point we find the second great difficulty in the proper interpretation of this type of curve, namely, the percentage deviations. For example, at the age of 11 years, a deviation of half a square either side of the curve corresponds to 75 calories, or, with a basal metabolism of 1,125 calories, a difference of not far from 7 per cent. At the age of 6 months a like deviation has exactly the same numerical value as at 11 years, i. e., 75 calories, but on the percentage basis this variation at 6 months represents an error of about 20 per cent, since the basal metabolism is considerably less. Hence the seemingly close grouping of points about the general line in the earlier years is only apparent and does not necessarily indicate a greater uniformity in the metabolism. This particular phase must be borne in mind for all of the charts, since it is the common custom of physiologists to consider deviations in metab- olism either side of a so-called normal on the percentage basis and these charts can not be so used. The chief usefulness of the chart in figure 22 is to indicate the tendency for the metabolism to increase rapidly during the first year of Me and to rise steadily, though not so rapidly, during the remainder of youth. It should be pointed out at this juncture that two of the three values beyond the age of 13 years lie above the projected line. With one of these boys (13 years and 8 months old) signs of puberty were very clearly present. 1 The other two showed no signs of puberty. These facts are emphasized, since subsequent discussion of the metabolism as influenced by puberty is necessary, owing to the great stress laid upon this point throughout the literature, beginning with the earlier studies of Andral and Gavarret. Our data do not permit the dis- cussion of the influence of puberty upon the metabolism, as the obser- vations did not extend to this point, but the accumulation of experi- mental material along this line is now in progress at the Nutrition Laboratory. Neither is this the time to consider the possibility of predicting the metabolism of various ages by referring to the general curve or central tendency for the basal metabolism, except as we may lay down the general principle that as the age increases the metabolism increases, and with a reasonable degree of uniformity. TOTAL CALORIES PER 24 HOURS REFERRED TO AGE (GIRLS). It is the general opinion that boys as a rule are much more active physically than girls, are less controllable, and can less easily acquire a condition of repose. This has a bearing upon any analysis of the 1 The values for children showing evidences of puberty are represented on the charts by en- closing the point in a square. See discussion of these values on page 183. METABOLISM AS AFFECTED BY GROWTH. 135 metabolism figures for boys and girls as to sexual differentiation. By design, our observations included a relatively large number of girls, and therefore provide sufficient points for special chart treatment. The total metabolism of the girls referred to age is depicted in figure 23. The total number of points is essentially the same as for boys, and for the most part end prior to puberty. Beyond the age of 11 years, the values are much scattered and but four points are available. One of these points, that for a girl 12 years and 1 month old, is the same child as that indicated by the point for 11 years. At both ages the points are specially designated for convenience in comparison of the metabolism before and after the establishment of puberty. This difference in metabolism is discussed in a later section. (See page 183.) TOTAL CALORIES REFERRED TO AGE. GIRLS. 1350 1200 1050 900 750 600 450 300 150 O V ' ^ .- / ^ ^. ' :^ ^ ^ - &~~ f ^ . i /. V * ' '/ { re.. 1 2 3 4 5 \ * \ 3 11 1 2 1 3 14 FIG. 23. Basal heat production of girls per 24 hours referred to age. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 12 years 1 month (puberty). The sharp rise in the total metabolism in the first year of growth is shown in figure 23, as well as the general steady increase thereafter. The smoothed curve does not fit the points quite so satisfactorily as with the boys, since between the ages of 2 and 4 years there is clearly an alteration in the general trend, which is not noticeable with the boys. This requires a slight alteration in the direction of the line, but again it must be remembered that this line is purely hypothetical and may not be looked upon as indicating a definite regularity in metabolism, but only the general trend. While only a superficial inspection can be given these charts for boys and girls, there seems to be a general tendency for the points to group themselves somewhat more closely about the general line with the boys than with the girls. 136 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TOTAL METABOLISM OP CHILDREN REFERRED TO AGE (EARLIER INVESTIGATORS). Since our study was made primarily to secure the metabolism under conditions giving the basal metabolism, i. e., with the values unaffected by muscular activity and with but little and preferably no influence of food, it is important in comparing our results with those of earlier writers to include of the latter only those obtained under conditions approximating complete muscular repose. As shown in a considera- tion of the previous literature on this subject (pages 4 to 21), relatively few of the earlier values meet these conditions. For ex- ample, all the studies of Andral and Gavarret 1 were made with the children in the sitting position and the data obtained (see table 1, page 5) indicate a total metabolism per 24 hours much higher at the low ages than that found by us with either boys or girls. Such values of the earlier studies as are suitable for comparison have, how- ever, been charted, together with the lines showing the general trends noted on our several charts. The values for the boy and girl studied by Scharling 2 can advan- tageously be plotted, but the data for Forster's 3 children are averaged in such a way that it would be difficult to apply them on our charts. For example, he finds no material change in the carbon-dioxide pro- duction per 10 kg. with children throughout the period from 3 to 7 years of age, namely, 11.7 grams per 10 kg. per hour, or approximately 3.50 calories per kilogram per hour. Since no data as to the body- weight are given, it is impossible to plot his values on our charts, but those given for children from 3 to 7 years of age correspond to about 84 calories per kilogram per 24 hours, which is far in excess of values noted by us for any age. This high metabolism is made the subject of special discussion by Forster. Speck's experiments, 4 which were made after the ingestion of food and with the child in the sitting or standing position, are also un- suitable for comparison with our data. The results of Sonde"n and Tigerstedt's extensive series 5 for the most part can not be employed here, with the exception of the data obtained with two subjects asleep. Thus with one boy, 11 years 2 months old, and with a body-weight of 32 kg., the total heat was 1,237 calories per 24 hours, a value quite in conformity with some points noted by us. Another boy, 12 years of age, showed a somewhat higher heat output of 1,373 calories. 1 Andral and Gavarret, Ann. d. Chim. et d. Phys., 1843, ser. 3, 8, p. 129. 2 Scharling, Ann. d. Chem. u. Pharm., 1843, 45, p. 214; reprinted in detail in Ann. d. Chim. et d. Phys., 1843, s6r. 3, 8, p. 478. * Forster, Amtl. Ber. d. 50 Versamml. deutsch. Naturf. u. Aerate in Munchen, 1877, p. 355; also v. Ziemssen's Handbuch der Hygiene, Leipsic, 1882, 1, p. 76. See, also, Magnus-Levy and Falk, Archiv f. Anat. u. Physiol., 1899, Suppbd., p. 356. 4 Speck, Physiologie des menschlichen Athmens, Leipsic, 1892, p. 217. 8 Sonden and Tigerstedt, Skand. Archiv f. Physiol., 1895, 6, p. 1. METABOLISM AS AFFECTED BY GROWTH. 137 Rubner's classic experiments 1 unfortunately give no values which are suitable for comparison. The data of Magnus-Levy and Falk 2 for both boys and girls are especially suitable for use and have all been plotted on our charts. Owing to the extensive criticisms applicable to the work of von Willebrand, 3 we do not feel justified in employing any of his data for comparison purposes, although it is not improbable that certain scattered minimum values are normal. Olin's experi- ments 4 were likewise made under conditions that can not be looked upon as basal. Finally, values have been plotted for such of Murlin and Hoobler's 5 children as were strictly normal, and also the results obtained in the extended series of Du Bois 6 and his collaborators are plotted, although criticism has been raised of these latter as to the selection of minimum periods and the degree of muscular repose. TOTAL CALORIES REFERRED TO AGE. 1650 1500 1350 1200 1050 900 750 600 450 300 150 1 =' o o ^ f " ^ ^ ? ^ ^ ^ ^ * . ^ ^ ^ *** / s / rf-S 5CHA ,OND UAGr> 3LINC EN A US-L 5 MDTI EVY / GERS kND f TEDT ALK *-* /IURLIN AND HO DBLE \ t x- .. o. DU BOIS (1916) Q-DU BOIS 0918) fre. 1 J 3456 789 10 11 1 FIG. 24. Basal heat production of boys per 24 hours referred to age (earlier investigators). Comparison for boys. The points representing the children studied by us are so numerous as to make it unwise to reproduce them on the same chart with those obtained by former investigators. Conse- quently, points for the results found by the earlier workers have been plotted and the line corresponding to the general trend shown in our studies as laid on the chart in figure 22 has been placed for comparison on the chart in figure 24, which gives the results for boys. The 1 Rubner, Beitrage zur Ernahrung im Knabenalter, Berlin, 1902. 2 Magnus-Levy and Falk, Archiv f. Anat. u. Physiol., 1899, Suppbd., p. 314. 3 von Willebrand, Finska Lakaresallskapets Handlingar, 1907, 49, p. 417. 4 Olin, Finska Lakaresallskapets Handlingar, 1915, 57, p. 1434; also Skand. Archiv f. Physiol., 1915, 34, p. 414. 6 Murlin and Hoobler, Am. Journ. Diseases of Children, 1915, 9, p. 81. 6 Du Bois, Arch. Internal Med., 1916, 17, p. 887. Olmstead, Barr and Du Bois, Arch. Internal Med., 1918, 21, p. 621. 138 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. striking feature of this chart is that above the age of 1 year practically all of the earlier observations lie above the line, the exceptions being two boys studied by Du Bois and one boy observed by Magnus-Levy and Falk. Aside from these three cases, the general trend of metab- olism in practically all the other studies was at a noticeably higher level than that found by us. A number of Magnus-Levy's observa- tions lie very close to our line; this would seem to confute the general conception suggested by Harris and Benedict 1 that the observations of Magnus-Levy possibly indicate a racial difference in metabolism. From a critical analysis of the earlier researches it would seem probable that the conditions for basal metabolism, particularly with respect to muscular activity, were by no means so rigidly adhered to in the earlier observations as in ours, and we believe that no evidence exists thus far to suggest that the differences in the results may not be entirely explained by a difference in activity, without the necessity of implying a racial difference in the metabolic level. Certainly, with Du Bois's data, we are dealing with American material. Cals. TOTAL CALORIES REFERRED TO AGE. GIRLS. 10OU 1350 1200 1050 900 750 600 450 300 150 ^ ^ -'- . ^ ' . -" - *"' ^ ^ -^ ,- " / / / rf-5 CHAF ?LINC I A- MURLIN AND HOOBLER - MAGNUS-LEVY AND FALK 'rs. 1 2 3 4 { > 6 7 8 9 1< ) 1 12 13 14 FIG. 25. Basal heat production of girls per 24 hours referred to age (earlier investigators). Comparison for girls. While very few measurements of the metab- olism of girls have been made, for the sake of consistency we have plotted in figure 25 the few observations we have been able to find in the literature and have laid our curve for girls upon the same chart. We find here two values of Magnus-Levy below our line; the other values lie above it. The two low values of Magnus-Levy for 12-year- old girls are of special interest, since, in the analysis of Magnus-Levy's data made by Harris and Benedict, 2 these two girls had a predicted 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 235. 1 Ibid., p. 236. METABOLISM AS AFFECTED BY GROWTH. 139 metabolism calculated with the American normal (multiple prediction) formula for man that agrees remarkably well with that found, while all the other values are much higher. With girls, therefore, as with boys, the results of the earlier observations tend, in general, to have a higher level of basal metabolism than was found in our series. GENERAL CONCLUSIONS AS TO TOTAL METABOLISM AND AGE IN CHILDREN. Our observations indicate a continually increasing metabolism from birth to 13 years of age with both boys and girls. A slight deviation in the general trend, as shown by the hypothetical smoothed curves laid on these charts, suggests that the trend for girls is slightly different from that of boys, especially about the age of 2 to 4 years. No other sexual differentiation can at this point of the analysis be observed. Reference to all the available earlier observations suitable for comparison as basal measurements shows that for boys, save in rare instances, the observations always lie considerably above our line representing the general trend. The scattered observations with girls indicate substantially the same situation. From a careful analysis of all the earlier experiments, we believe that the results of our observa- tions more nearly approach the true basal values than those of the previous investigations; hence, in lieu of further data, they must be looked upon as the closest estimates of the true basal metabolism of youth that have thus far been obtained. TOTAL METABOLISM WITH INCREASING BOOT-WEIGHT. While physiological observations with children are commonly referred to age, it is particularly unfortunate that the total metabolism should be thus referred, for experience with adults has shown us that a number of physiological factors play an important role in the total metabolism, among these being body-weight, stature, and age. Of these, body-weight has by far the greatest influence, much greater than that of age; consequently we should more properly refer the total number of calories per day to the body-weight rather than to the age of the child. In general, the larger the child is, one would a priori expect a greater metabolism. Furthermore, since we have found that the metabolism increases with age, and since age and body- weight, especially during the period of growth, go more or less hand in hand, we should expect changes in metabolism to be in reasonable conformity with changes in weight. Our curves representing the relationship between total calories and age may simply be an ex- pression of the fact that as children grow older they likewise grow heavier, and the larger organism has a larger heat production. Physio- logically, therefore, the better method of comparison is on the basis of weight rather than of age. 140 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TOTAL CALORIES PER 24 HOURS REFERRED TO WEIGHT (BOYS). In figure 26 we have plotted for boys the total calories per 24 hours referred to body-weight. The general sweep of the curve and the dispersion of the points about a central sketched line is not unlike that for boys for the total calories referred to age. (See fig. 22, page 133.) In fact, a superficial inspection would imply that of the two charts the points are even more closely grouped about the central line in figure 26. In such comparison it should be noted that the scales in the two charts are somewhat different, since each vertical division in figure 22 corresponds to 150 calories, while in figure 26 it corresponds to but 100 calories. Cals. TOT At CALORIES REFERRED TO WEIGHT. BOYS. 100U 1400 1300 1200 MOO 1000 sS l&K^ ^^ ^ ^ >- ^ : ^ ' . ' 900 X ^ , 800 >^ ^r 700 X ^^ 600 ..; ^ 500 . ^ ^ 400 v ', 300 /' : 200 /' 100 /' 1 Jkgs.4 6 8 10 12 14 16 18 20 2 2 24 26 28 30 32 34 36 38 40 42 FIG. 26. Basal heat production of boys per 24 hours referred to body-weight. Point inclosed in square signifies puberty established. The general close agreement between the different points and the sketched central curve is rather striking, although at times the devi- ations amount to 150 or more calories. Thus, at 19 kg. one point varies considerably over 100 calories from the line, corresponding to a deviation of 15 per cent. Another child, weighing 38 kg., is 180 calories below the general line, while a third child, weighing 9 kg., is 125 calories, or 25 per cent, below the central line. Similar instances of values above the line may likewise be noted. While the child designated as No. 258, with a weight of 50.8 kg., seems on the chart METABOLISM AS AFFECTED BY GROWTH. 141 to have a metabolism differing widely from the central tendency, it must be remembered that if the chart were extended to the 50 kg. range, the metabolism would be shown to be much nearer the central line, though still above it. On the whole, the values lie reasonably close either side of the line; indeed, so close that the use of this curve to predict the normal metabolism will be subsequently discussed. (See page 188.) Attention should be called to the evidence in figure 26 of differences in metabolism, even with children of the same weight. In some instances, i. e., with children weighing 38 or 39 kg., we find a range of from 1,100 to 1,400 calories, or a difference of 27 per cent; and at 21 kg. we have one value of 785 calories and another at 985 calories, a difference of 200 calories, or 25 per cent. At about 11 kg. we have a range from 450 to 660 calories, approximately 200 calories, or about 45 per cent. Thus we see that with certain individuals our extremes may be fairly wide. On the other hand, the general grouping of the points around the central line is suggestive of a clearly defined trend. Since the general picture is so similar to that of the chart for calories referred to age, we apparently have here another expression of the fact that as the children increase in age they increase in weight, and that the age chart and the weight chart are more or less inseparable, since weight is inevitably correlated with age and heat production increases both with age and weight. It would be undesirable, how- ever, even to imply that age and weight are of equal or, indeed, of comparable importance hi determining basal metabolism. It has been shown with adults that there is a definite influence of age, with, on the average, an actual decrease in the daily heat production with men of about 7.15 calories per year and with women a decrease of 2.29 calories per year. 1 Still, this same analysis indicates that the influence of weight far exceeds that of age; hence we must conclude that with children the changes in metabolism noted with different ages are due primarily not to the age element, but to the fact that age changes concurrently with weight. TOTAL CALORIES PER 24 HOURS REFERRED TO WEIGHT (GIRLS). The values for the girls included in this study are charted hi figure 27. The line representing the general trend of metabolism gives a picture of rather rapidly increasing metabolism until the weight of 10 or 12 kg., with a tendency thereafter for the metabolism to increase at a slower rate, which is still present when the weight limit of the chart (39 kg.) is reached. Practically, this chart should have ended at the weight of 31 kg., there being but two values retained beyond this weight. The widest deviations of individuals from the general line are, for the most part, well within 100 calories, which, for the higher-weight 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 115. 142 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. ranges, represents a relatively small percentage deviation. A com- parison ,,of the percentage deviations for these children with those for adults has special interest, since with adults it was noted that the scatter of the individual points from the central line was very considerable. In consideration of the rapidly changing body-mass with children, the compact arrangement of the points in these charts has, however, a somewhat greater significance. With pronounced TOTAL CALORIES REFERRED TO WEIGHT. GIRLS. 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 . -*' ,.-"' f .. ,,--''" ^ * ^ ^ ^ kgs. 46 8 10 12 14 1< > u 20 22 24 26 28 30 32 34 36 38 40 FIG. 27. Basal heat production of girls per 24 hours referred to body-weight. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 39.2 kg. (puberty). alterations in area, stature, and weight, such a stringent conformity w the central tendency may not be expected as with well-developed adults. So close are these points, on the whole, to the general curves for the boys and girls that the possibility is considered later of using the two curves for predicting the metabolism of children whose basal heat output is unknown. (See page 205.) TOTAL METABOLISM OF CHILDREN REFERRED TO "WEIGHT (EARLIER INVESTIGATORS). The special advantages of referring total metabolism measurements to weight rather than to age observed with the children in this research make it likewise important to inspect the values reported by other METABOLISM AS AFFECTED BY GROWTH. 143 investigators on this basis. We have consequently plotted in figure 28 the values for boys obtained by other investigators. For purposes of comparison we have laid on the smoothed curve from the chart in figure 26, representing the general trend of the total metabolism of boys as found by us. With the exception of one child under 6 kg. of weight, all of the observations lie above the smoothed line. Since our curve has been carried only to 42 kg., five of Du Bois's TOTAL CALORIES REFERRED TO WEIGHT. 1600 ,1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 2 , 9 1 t . " ^ "^ ^ "" ^ ^ , ^ ^ . ^ ^ ^ ^" * / ,/ / / / / / / /' rf-S CHA 3LINC a-ft rfURL NAN DHO )BLE * / x-SONDEN AND TIGERSTEDT - MAGNUS-LEVY AND FALK o-DU 8015(1916) -DU BOIS0918) | [ kg.4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 5 FIG. 28. Basal heat production of boys per 24 hours referred to body-weight (earlier investigators) . boy scouts and one of Magnus-Levy and Talk's children are outside of this range. If the curve were projected, one of these values would lie approximately on the line, another somewhat below it, and four considerably above it. The high values noted on the basis of weight for the children studied by Magnus-Levy and Falk and by Du Bois have been criticized, the Magnus-Levy and Falk values by Harris and Benedict 1 and the Du Bois values hi the historical section of this monograph. (See page 19.) The fact that practically all the values lie above the general trend found by us would imply again that our curve represents more nearly the true basal value than any curve that can be drawn from the earlier work, when the requirements for basal conditions were apparently not so rigidly adhered to. 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 235. 144 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. Although but few measurements with girls are available for this comparison, they have been plotted in figure 29 and the corresponding curve derived from our observations on girls has been applied. While the points all lie above our curve, they are for the most part somewhat closer to the line than was found with the corresponding values for boys. Almost no data are available for girls at the lower weights. Cals. TOTAL CALORIES REFERRED TO WEIGHT. GIRLS- 15OO 1400 1300 1200 1100 1000 900 800 . , ^ x x ^ ^ ' ^ ^ ^ ^ ^ ^ 700 600 500 400 300 POO ^X ^ S ^ / \ <^ : ? >s ^..^^ 10 11 12 13 14 15 16 FIG. 30. Basal heat production of boys per kilogram of body-weight per 24 hours referred to age. Point inclosed in square signifies puberty established. The values for the earlier months of life, as shown in the chart in figure 30, indicate a clear tendency for a lower metabolism per kilo- gram of body-weight than at the end of the first year, thus justifying the upward trend of the curve. Subsequent to 5 years, the values are definitely lower than those in the first 3 years of life. Although there is great irregularity in the dispersion of the points, it was thought best to represent the trend after 6 years by a straight line, for the irregularity noted in the distribution of the points is perhaps no greater than that found with adults. A most careful analysis of the material for adults indicated that the straight-line equation gave as close a representation of the changes in metabolism per kilo- gram of body-weight with changing years as could be expressed by a curve of a higher order. Still, the laying on of this curve must be looked upon only as an empirical representation of a general trend and not as a mathematically established average. Of special significance in this chart, therefore, is first the extra- ordinary dispersion of the points about the smoothed curve. What- ever degree of regularity has been heretofore assumed by physiologists 148 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. in the heat production per kilogram of body-weight, the chart certainly shows with children a great diversity with age changes. The figures indicate a somewhat lower general metabolism per kilogram of body- weight during the first 6 months, with the highest metabolism per kilogram throughout the age-range of this study to be at 1 or 2 years of age. While the values for the period from 3 to 5 years are but few in number, they indicate a rather rapid fall; after the age of 5 years the decline is definite, though not so marked. The small number of individuals studied over 13 years of age hardly justifies a continuation of this curve beyond that period. The heat production per kilogram of body-weight has frequently been considered as somewhat of a physiological constant, but we find on this chart a range in values extending from 29 to 64 calories per kilogram, in other words, a variation of over 100 per cent. Even during the first year of life there is a range of from 41 to 64 calories. After the fifth year the range is from 29 to 48 calories. While the smoothed curve laid on this chart does imply a slight general trend, there is nothing here approximating mathematical constancy, and certainly nothing that can be considered as a physiological law estab- lishing a relationship between the heat production per kilogram of body-weight and the age. The charts comparing total calories do not permit a comparison of individuals at different weights or different ages; hence this method is of value, since it supplies some suggestion as to the relative intensity of metabolism at different ages. On the assumption, erroneous though it is, that each kilogram of body-substance has the same heat-producing capacity, one can conclude that at the age of 1 to 2 years there is a greater relative intensity of metabolism than at any other period of life up to 13 years, and that the young organism per unit of mass produces a larger amount of heat than does the older. It would appear from this chart that with children we are dealing with physiological entities and not with crystalline structures, each having its mathematically established planes. Although the method of expressing the metabolism on the basis of per kilogram of body- weight permits a very gross comparison of different individuals, the entire absence of uniformity and the wide scatter of the points about the central tendency show that such comparison can have but very slight individual mathematical significance. Between the ages of 5 and 13 years, although the points are widely scattered, it would appear as if a straight line represented the trend as well as any other form of curve. This is of significance as being preliminary to the straight-line tendency exhibited with male adults, and this curve therefore brings out primarily the high metabolism per unit of mass noted with boys at about the age of 1 to 2 years. METABOLISM AS AFFECTED BY GROWTH. 149 CALORIES PER KILOGRAM OF BODY-WEIGHT PER 24 HOURS REFERRED TO AGE (oiBLS). Our observations on girls at different ages, being practically as numerous as those of boys, allow us to study the caloric output per kilogram of body-weight as referred to age. In figure 31 we have plotted the individual points for all of our girls on this basis. The striking scatter of all these points makes it extremely difficult to lay on anything in the nature of a smoothed curve that could be considered justifiable. From the concensus of opinion of five observers, a curve has been sketched which indicates nothing more than a general trend. Cals. CALORIES PER KILO. REFERRED TO AGE. GIRLS. . .. . - ^ I * X \ . ' 9 9 . . V * >vv - <^ , 7"" -^ " J --~- "-^^^ . "** - . v Yrs. 1 \ 3 4 5 6 7 8 9 10 11 12 13 14 FIG. 31. Basal heat production of girls per kilogram of body-weight per 24 hours referred to age. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 12 years 1 month (puberty). During the first six months on the average there is a somewhat lower metabolism per unit of weight than appears later, with the highest values occurring at about one year. The general form of the curve is not unlike that in figure 30 for boys. Although it is hardly the place to emphasize a sexual differentiation, it is worth while indicating here that with boys the highest values lie at 64 calories, while with girls there are six of the individual points which lie higher than 64 calories, practically all of them being inside the first year. When it is remembered that the element of weight is in large part removed by this comparison, one is impressed by the wide variations to be found in the calories per kilogram of body-weight of children of various ages. For example, at or about the age of 1 year we find variations ranging from 42 to 70 calories, and it is only after the age 150 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. of 7 years that the individual points tend to lie reasonably close to the general smoothed curve. From the general picture presented by the chart for girls, it can be seen that the conclusion drawn from that for boys still holds, namely, that the calories per kilogram of body-weight exhibit such wide devia- tions from a general mean as to make it impossible to conceive of anything approximating a physiological law associating the heat production per kilogram of body-weight with the age. The scattering of the points makes the laying-on of this smoothed curve a distinct violation of mathematics and open to severe criticism. We wish again to emphasize that the curves on all of these charts are not to be interpreted on the basis of mathematical accuracy, but simply to indicate the central tendency and general trend. CALORIES PER KILOGRAM OF BODY-WEIGHT PER 24 HOURS REFERRED TO AGE (EARLIER INVESTIGATORS) . Since the measurement of the metabolism per unit of mass plays such an important role with the earlier writers, the comparatively few individual values found by other investigators with boys are plotted in figure 32. On this chart we have likewise laid our general line Gals. CALORIES PER KILO. REFERRED TO AGE. 1 \ ' \ X o ^^-v \ ^..^^ T> s^ Q ' ^ \ | rf- SCHARLING x-SONDEN AND T GERS kND F TED" ALK o- I a- I A- n U BC )U BC rfURL )IS(19 )IS (1? N Afs 16) 18) DHO OBLE *> R *> D :!==. -I MGf> US-L :VY / fn. 1 234 56 7 8 9 10 11 12 13 14 15 16 FIG. 32. Basal heat production of boys per kilogram of body-weight per 24 hours referred to age (earlier investigators). from figure 30. With the single exception of one value at the age of 2 months, no other point in the entire series lies below our general line up to the age of 13 years, but for the most part they lie very con- siderably above our line. Beyond 13 years two points obtained by METABOLISM AS AFFECTED BY GROWTH. 151 Du Bois with boy scouts in his second series of experiments indicate values a little below our projected general trend. This comparison, in which the influence of variations in weight is in part eliminated, still shows that practically all of the values obtained in the earlier work are on a higher metabolic level than ours. We are thus forced to the conclusion that much of the earlier work was un- wittingly affected by muscular activity to such an extent that it fails to meet modern requirements for basal metabolism measurements. Still, the general trend of the earlier work is not widely different from that observed in our research, for a smoothed curve passed through the points on the chart hi figure 32 would be of approximately the same order, although at a higher level than that shown by our results. This suggests that the differences in the various series of observations are due wholly to differences in the degree of approximation to basal conditions. The earlier work does not, however, bring out the dis- tinctly lower metabolism in the first few months of life which is so clearly shown by the general averages of our more numerous data. The relatively few observations of earlier writers on girls make it seemingly unnecessary to burden this report with a reproduction of an additional chart comparing the values per kilogram of body-weight referred to age for these investigators. It is sufficient to state that all of the earlier values lie above our smoothed curve, with the excep- tion of two of Magnus-Levy and Falk's girls at 11 and 12 years, both of which lie but little below. The points for the Magnus-Levy and Falk girls, however, are grouped about the curve with a distribution not dissimilar to that for our own values. In general, all the girls previously measured who were 10 years of age and over show a reason- ably close agreement with our smoothed curve on the basis of calories per kilogram of body-weight. The importance of studying the metabolism during youth was clearly emphasized in the classical studies of Sonde"n and Tigerstedt, 1 but owing to the absence of conditions prerequisite for the determina- tion of basal metabolism, the values are not strictly comparable and only two could be used in the chart in figure 32. We have, however, selected minimum carbon-dioxide periods and computed the minimum calories per kilogram referred to age; these are plotted in figure 33 for boys. Upon the same chart we have laid our curve for basal metabolism of boys from 7 to 16 years and likewise included the values recently cited by Carl Tigerstedt 2 for boys from 9 to 14 years of age. This last citation is preceded by a careful statement as to the im- portance of muscular rest, and these values have accordingly been 1 Sonden and Tigerstedt, Skand. Archiv f. Physi<51., 1895, 6, p. 1. 5 Tigerstedt, Carl, Ueber die Nahrungszufuhr des Menschen in ihrer Abhangigkeit von Alter, Geschlecht und Beruf, Helsingfors, 1915. See also Skand. Archiv f. Physiol., 1916, 34, p. 151. 152 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. selected with this in mind. Similar results are shown in figure 34 for girls. Cals 72 68 64 60 56 52 48 44 40 36 32 28 CALORIES PER KILO. REFERRED TO AGE. BOYS. \ v \ ^ x \ \ \ \ SONDEN'AND V TIGERSTEDT \ ^- \ V X \ \ \ V TIGE RSTE DT ^ s^ ^ ^ \ V ^ >^ ^^ ^ \ " \ B. ANC T. 10 11 12 13 14 16 17 18 19 20 .21 22 23 FIG. 33. Basal heat production of boys per kilogram of body-weight per 24 hours referred to age (Sonde"n and Tigerstedt, C. Tigerstedt, and Benedict and Talbot). The two crosses represent the values found by Sonden and Tigerstedt with boys during sleep. Cals. 72 68 64 60 56 52 48 44 40 36 32 28 CALORIES PER KILO. REFERRED TO AGE. GIRLS. \ \ ' \ \ ^ \ \ SONDEN AND \TIGERSTEDT ^ >**, ~-~-~. B.A1 ^DT. \ J ^ \ --. ^ \ yrs. i 10 11 12 13 14 15 16 17 18 FIG. 34. Basal heat production of girls per kilogram of body-weight per 24 hours referred to age (Sonden and Tigerstedt, and Benedict and Talbot). The striking feature of these charts is the great difference between the values reported by Sond6n and Tigerstedt and those in our METABOLISM AS AFFECTED BY GROWTH. 153 smoothed curves. Of importance, however, is the fact that the values obtained by Sonden and Tigerstedt with two sleeping boys 11 and 12 years old, respectively, and indicated in figure 33 by small crosses, lie close to our line. This explains not only why our values are lower than the other values found by Sonden and Tigerstedt, as well as the composite values derived by Carl Tigerstedt from the observations of Hellstrom, Rubner, von Willebrand, and Sonde"n and Tigerstedt, but likewise, we believe, why they are lower than the results of the observations of Magnus-Levy and Falk, Du Bois, and others. When the experimental conditions under which the early investigations were made more nearly approach basal requirements the values are found to be more in line with our smoothed curve. As will be seen later (page 209), while we must disregard in large measure the earlier work as a standard for basal metabolism, these results have a great practical value for estimating the probable 24-hour total daily requirements of the growing, active child. The two series thus supplement each other perfectly. METABOLISM PER UNIT OF BODY-WEIGHT REFERRED TO WEIGHT. In referring the metabolism of children to age, undue stress is laid upon the age element; from the earlier analysis of the metabolism of adults of different ages, we have every reason to believe that, while the age factor is by no means to be ignored, it does not hi any way compare with the weight factor. With youth, gross differences in metabolism are noted with variation in age, but these differences may in large part be ascribed to the concomitant weight changes, since a child changing in age is likewise changing in weight. Theoretically, at least, a more logical comparison of the metabolism of different children is not upon the basis of age, but upon weight. The weight element is in part removed by computing the calories per kilogram of body-weight. Even then, strictly speaking, the comparison still is best made with children of various weights rather than of various ages. From the analysis of the charts in figures 22 to 32, it is seen that in general the pictures of the metabolic changes for the various weights are not unlike those for age, and hence we are prepared to find the curves for the calories per kilogram of body-weight referred to body-weight somewhat similar to those in which these values are referred to age. Our values for boys have been plotted in figure 35 and a smoothed curve sketched to indicate approximately the general trend. The very wide scatter of the points, particularly below 18 kg. in weight, is worthy of special notice and is fully in accordance with variations noted in the age charts. It seems reasonably clear that at the weights under 6 or 7 kg. there still is a tendency for the metabolism to be somewhat lower per kilogram than a little later. Hence we feel that 154 METABOLISM AND GROWTH PROM BIRTH TO PUBERTY. the rise in the sketched curve is perfectly justifiable. The maximum occurs at about 7 or 8 kg., this approximating the average weight for children of one year. Thence the curve decreases with a reasonable degree of regularity. Not until 24 kg. and over is reached do the points lie sufficiently close to the central line to give a clear idea of physiological regularity in the metabolism per kilogram of body- weight referred to the total weight. Cate. CALORIES PER KILO. REFERRED TO WEIGHT. . . . . 1 . ^ < .' /' \ \ . / * . \ \ / \ X ' \ . "Sv ^ v r~> -^ v \ t / ^. ~\. ^ ^- . -\ ^~^ 3S&& kas. 4 e 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 FIG. 35. Basal heat production of boys per kilogram of body-weight per 24 hours referred to weight. Point inclosed in square signifies puberty established. The suggestion of sexual differentiation between boys and girls, even at the early ages, noted in the calories per kilogram at the age of one year, makes it desirable to consider by itself the metabolism per kilogram of the girls at different weights. These values are plotted on the chart in figure 36, together with our smoothed curve indicating the general trend of metabolism. The most pronounced feature of this chart is the wide scatter of the individual points from the smoothed curve, this being even more evident for the girls than for the boys. Still, it is reasonably clear that the general trend shows a rise between 3 and 8 kg., where the maximum is found, with a clearly defined decrease thereafter to about 26 kg. Subsequent to that point the line appears to be reasonably level, but the number of points available beyond 30 kg. is so few as hardly to justify discussion. Up to 18 kg. there is such a wide dispersion of points about the smoothed curve that no conclusion may be drawn regarding a physiological law, even METABOLISM AS AFFECTED BY GROWTH. 155 though the general trend seems to be established with reasonable clearness. The correlation between the calories per kilogram and the body-weight is not, therefore, a very striking one. While it is suf- ficient to indicate a low metabolism at the early weights, with a rise to a maximum at about 8 kg., so far as the individual is concerned one can not predict at the weight range of 6 to 8 kg. whether the metabolism will be 43 or 70 calories per kilogram. After 18 kg. there seems to be a reasonable degree of compactness in the grouping of the individual points. Cals. CALORIES PER KILO. REFERRED TO WEIGHT. 69 66 63 60 57 54 51 48 45 42 39 36 33 30 \ .' . . . . / s^~ ^ \* '/ ' . \ / \ . . \ ~ " . \ X x >> ^ . ^ ^ . _. * _ - kgs. 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 4( FIG. 36. Basal heat production of girls per kilogram of body-weight per 24 hours referred to weight. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 39.2 kg. (puberty). At the same weights, very great differences in metabolism are noted in both figures 35 and 36, even with the unit here employed, namely, the calories per kilogram of body-weight. It is, however, clearly established from both curves that the metabolic activity per unit of weight is very much greater at the lower weights than at the higher. This is in full conformity with the experience with adults, the heavy individuals having a lower heat production per kilogram of body- weight than the light men and women. A popular interpretation of 156 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. this phenomenon with adults has been that with the heavy individuals there is a large proportion of inactive fat, which does not materially contribute to the heat production. This explanation, while reasonably clear for adults, does not, we believe, hold true with children, for on this basis we should expect the children with the higher weights to have a much larger proportion of inactive body-fat than the smaller younger children, and in consequence their heat production per kilo- gram of body-weight would be lower. As a matter of fact, physical examination has shown that young children usually have a much larger proportion of fat than older children, so that the phenomenon exhibited in figures 35 and 36 is exactly contrary to what would be expected if one considered solely the proportion of fat in the body. At this point a note of caution must be sounded. In our discussion of the conditions laid down by us as the basal requirements, we have stressed considerably the fact that with young children one of the pre- requisites for basal measurements can not be satisfactorily met, i. e., the post-absorptive condition. With adults, measurements are made approximately 12 hours after the last meal, when, it has been experi- mentally demonstrated, the stimulus of the previously ingested food has practically disappeared. Our observations show that the younger the children, the more difficult it is to secure long periods without food. Furthermore, with young children, particularly, it is difficult to deter- mine exactly at what hour the stimulus of the previously ingested food ceases and the point at which the stimulus of the ever-occurring incipient acidosis begins. Since the influence of the previously ingested food is in inverse proportion to the age of the children in- cluded in this study, one must bear in mind that the high portion shown in both curves in the two figures 35 and 36 is undoubtedly in part influenced by the previous ingestion of food. We believe, how- ever, that if a correction were possible for the influence of food, on the percentage basis this part would still lie somewhat above that for the older and heavier children. Furthermore, we have every reason to believe that the values observed at the earlier weights (from 3 to 5 kg.) were fully as much affected by the previous food as those between 6 and 8 kg., and there is clear evidence of a somewhat lower metabolism at these weights. While, therefore, the previous ingestion of food unquestionably raises somewhat the level of the curve showing the general trend of metabolism per kilogram of body- weight, probably no theoretical correction could remove all of this difference, especially as we have likewise a somewhat compensatory effect due to the fact that many of our younger children were in deep sleep, a condition that we believe definitely lowers the metabolism. So while our children could not be compared on the basis of a " post-absorptive condition" with adults, for the most part they were asleep and had a somewhat lower metabolism due to the specific effect of sleep than did the adults, who were for the most part awake. METABOLISM AS AFFECTED BY GROWTH. 157 CALORIES PER KILOGRAM OF BODY-WEIGHT PER 24 HOURS REFERRED TO WEIGHT (EARLIER INVESTIGATORS) . The general trend of the calories per kilogram of body-weight referred to weight may be compared with the individual points noted by other observers. This has been done for boys in figure 37 for the weight ranges up to 43 kg. To avoid enlarging this chart unduly, it was necessary to omit a number of values obtained by Du Bois and published in the second paper on his studies with boy scouts. 1 The Cals. CALORIES PER KILO. REFERRED TO WEIGHT. BOYS. . - - /" x . / \ / \ \ I \ X ' . s I \. \ ^ \ Vj, " US-L iVY / a- 1 /1URL N AN DHO R "~--~ ^*+ ~~~~^ kgt.4 6 8 10 12 14 16 18 20 22 24 26 28 3O 32 34 36 38 40 42 44 FIG. 37. Basal heat production of boys per kilogram of body-weight per 24 hours referred to weight (earlier investigators). TABLE 29. Values outside weight range in figure 37 (Du Bois). Heat per Subject. Weight. kilogram per 24 hours. kilos. cals. R M 49.1 26.6 R F 48.6 29.3 H B 49.1 31.9 H. K 49.3 32.4 four boys who were outside of the weight-range studied by us gave the values shown in table 29. Aside from the values in table 29 and the infant of 5.5 kg. shown on the chart, all the points lie very considerably above our smoothed curve. If an approximate line were to be laid through these earlier values, it would follow with reasonable regularity at a higher level 1 Olmstead, Barr and Du Bois, Arch. Internal Med., 1918, 21, p. 621. 158 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. that laid down from our observations. It therefore seems clear, from all observations, that the metabolism per unit of body-mass is noticeably higher at 6 to 8 kg. of weight than at any other time of life. Our curve indicates that the metabolism is lower from 3 to 5 kg. than from 6 to 8 kg., but this is not shown by the earlier work. The gradual fall from 8 to 42 kg., at which point our observations end, is also apparent in all the observations. From the comparison of the calories per kilogram of body-weight in all the series, the general picture is thus essentially the same, and the newer data simply extend and confirm the earlier observations. The striking feature of this com- parison is that the new work shows a distinctly lower level all along the line and brings out the lower metabolism during the first few months of life. The scarcity of earlier material available for comparison does not justify publishing here a chart for girls, but such a comparison has been made. Two of the values obtained by Magnus-Levy and Falk for girls weighing between 40 and 42 kg. lie but little above our curve. In general, the points for girls up to 40 kg., though above our line, lie fairly close to it, and are not more widely scattered than our own observations. The number of points is so small, however, that they give no suggestion of a general trend, but, so far as they go, they are in fair conformity with our curve and indicate a higher metabolism at the lower weights. Notwithstanding the objections raised in earlier paragraphs regard- ing the probable influence of food, the evidence in all the observations demonstrates that there is a profound physiological difference in the metabolism of children weighing 6 to 8 kg. from that obtaining at any other period of life. If correction could be made for the composition of the body, it would appear that (per unit of weight of body-material other than fat) the metabolism would be even greater at the early weights and that undoubtedly with these weights (6 to 8 kg.) there is greater metabolic intensity per unit of active protoplasmic tissue than at any other point in the life of youth. In an earlier report 1 we emphasized the extraordinarily low heat production of new-born infants, particularly on the first day of birth, attributing this in part to the temperature changes and perhaps weak condition of the organism after the birth and bath. It would seem as though this lower metabolism is characteristic of very young children and that the metabolism per unit of mass gradually increases until about the age of 1 year, or a weight of 6 to 8 kg., and thereafter steadily declines throughout the period of youth. 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915, pp. 103 and 118. METABOLISM AS AFFECTED BY GROWTH. 159 RELATIONSHIP BETWEEN SURFACE AREA OF THE BODY AND METABOLISM. For decades the surface area of the body has by many physiologists been considered to have an intimate (in fact, a determining) relation- ship with the heat production. An extensive critique 1 of the body- surface law makes it unnecessary here to do more than to summarize in the following manner : Height and weight have independent influences upon metabolism; body-surface, with its rather close relationship to weight, likewise has an apparent relationship to metabolism. Since body-surface represents more nearly a general morphological law of growth than body-weight does, the relationships between accurately measured body-surface and metabolism are frequently much closer than between body-weight and metabolism. Biometric analysis has shown, how- ever, that certainly with the older methods of estimating body-surface, namely, the Meeh formula with its several constants, 2 body-weight and body-surface are equally closely correlated with heat production. When more exact methods for estimating body-area are used, par- ticularly the linear formula and the resultant height-weight chart of Du Bois, 3 the correlation between area and metabolism is slightly better than that between weight and metabolism, particularly if the method employing regression equations suggested by Harris and Benedict 4 be employed. The earlier estimates of body-surface area (from the Meeh formula) are so erroneous and the factors have such large coefficients of variation that at best they are only rough approx- imates, and the modern physiologist may well disregard completely all consideration of body-surface as calculated from the Meeh formula. The freeing of physiology from the cumbersome, wholly erroneous method of Meeh is due to the admirable work of D. and E. F. Du Bois, 5 who, by an extensive series of painstaking measurements of surface and casts from the surface of the body, have established a method of estimating the surface area of the body with a very considerable degree of accuracy. This method agrees perfectly with an entirely different method of measurement based upon a photographic procedure. 6 It' should be stated, however, that the photographic method could not have been developed without the work of the Du Boises. As a result of the critique of the body-surface law presented by Harris and Bene- dict, we believe that the accurate measurements of body-surface made possible by Du Bois may legitimately be used in a manner heretofore never practicable in metabolism experiments, provided that they are considered as physical measurements and with no erroneous concep- 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 129. z Meeh, Zeitschr. f. Biol., 1879, 15, p. 425. Du Bois and Du Bois, Arch. Intern. Med., 1916, 17, p. 863. < Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 188. 6 Du Bois and Du Bois, loc. cit. 6 Benedict, Am. Journ. Physiol., 1916, 41, p. 275. 160 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. tions as to the existence of a causal relationship between surface-area and heat elimination. Nearly all of our records of the surface areas of the children in this research, especially those for children above 1 year of age, 1 are based upon actual measurements of the surface-area by the Du Bois linear formula, and hence represent true physical measurements rather than computations from body-weight, which so long supplied the only basis for body-surface estimates. With our children, therefore, this accurately determined physical measurement may legitimately be employed exactly as we used the body-weight; if we so chose, we could likewise use the stature. Since the preponderance of evidence is slightly in favor of the correlation between body-surface accurately measured and basal metabolism on the one hand, and body-weight and basal metabolism on the other, a comparison of the surface-area and total metabolism is of physiological interest. For practical use it is highly desirable to determine a normal trend of basal metabolism referred to some simply measured factor, such as weight, for the purpose of predicting the heat production of a subject whose weight is known but whose metabolism has not been measured. The multitudinous measurements involved in the Du Bois linear for- mula may therefore, in many instances, rule out the possibilities of comparing the measured surface-area and the total metabolism, or using the measured surface-area as a unit for estimating basal meta- bolism, as has so long been attempted from either body-weight or from the surface-area as computed by the Meeh formula. In discussing our values on the basis of body-surface, it should be emphasized again that body-surface must be looked upon simply as a physical measurement approximating perhaps more closely the general morphological law of growth than does body-weight, and hence by this very fact, perhaps, giving a somewhat better idea of the relationship between the mass of active protoplasmic tissue and heat production than would the weight alone. We believe there is no causal relationship between body-surface area and heat production. All of our experimental evidence, not only for children but for adults under various conditions of nutrition, implies that the production of heat in the body is not determined by the loss. Even if it were granted, for the sake of argument, that the reverse is true, the physical and physiological factors influencing the heat loss from the surface of the human body are so different at different parts of the body as to pre- clude any generalization that equal areas result in equal heat loss. With this explanation clearly in mind, we may proceed to an analysis of the data obtained in this study, using the measured body-surface area as the unit of reference. 1 For 20 boys and 19 girls, nearly all of them very young, the body-surfaces were computed by the Lissauer formula. The data given for these children do not therefore represent actual measurements. (See tables 27 and 28, pp. 116 and 120.) METABOLISM AS AFFECTED BY GROWTH. 161 TOTAL CALORIES PER 24 HOURS REFERRED TO ACTUALLY MEASURED BODY-SURFACE AREA (BOYS.) Employing the body-surface areas actually measured by the method of Du Bois, we have plotted in figure 38 the total heat per 24 hours and the actually measured body-surface areas for all of our boys. As with the other charts, a smoothed curve has been laid on, approximat- Cals. 1700 1600 1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 .2 TOTAL CALORIES REFER RED TO SURFACE. BOYS. t 3 s ' o X ''' ./ [ : /x . xi . ' / "< / r 2 f ^ V V: o-Dl sou BOIJ >d9ie > (19K ) y v a- IX. 9 / ' sq.m..3 .4 .J . .7 .8 .9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 FIG. 38. Basal heat production of boys per 24 hours referred to body-surface. Point inclosed in square signifies puberty established. ing as closely as possible the central tendency. In plotting these values we have followed the scale adopted by Harris and Benedict 1 of considering one-tenth of a square meter and 100 calories heat production as of equal value. In contradistinction to the results obtained on adults, 2 the tendency of the metabolism is not in a straight line, but in a distinct curve, a curve strikingly similar in character to that found for the boys when 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 157, diagram 26. 2 Harris and Benedict, loc. cit. 162 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. the total calories were referred to weight in figure 26 (page 140). Since there is a close relationship between surface area and weight, this is not surprising, but it is important to note the scatter of the various points about the hypothetical smoothed curve. If the heat production is more closely proportional to the surface area than it is to the body-weight, we should logically expect a closer grouping of the points around the curve. A superficial inspection of the curves in figures 38 and 26 does not indicate that the scatter varies much in the two curves. If anything, it would appear that the points lie closer to the curve in the body-weight chart (fig. 26) than in the body-surface chart (fig. 38). The question of absolute scale values for ordinates and abscissae enters here, however, and since subse- quently both of these charts and smoothed curves are put to special use in an attempt to predict the heat production of the various children, further discussion at this point is unnecessary. The smoothed curve indicates a decidedly rapid increase in metab- olism with increasing surface area up to about 0.6 square meter. Thereafter, although there is an increase in heat with larger surface, it is somewhat less in degree, as indicated by the sketched curve. Considering the variations in total calories for boys of like surface, we find some striking differences. For example, with the surface of 1.3 square meters, one boy has a heat production of 1,096 calories and another of 1,401 calories, a difference of 28 per cent. Again, at about 0.80 to 0.85 square meter, we have variations from 716 to 984 calories, a difference of 268 calories, or about 37 per cent; while at approximately 0.53 square meter we have a range of 456 to 684 calories, a difference of 228 calories, or 50 per cent. These percentage differences are essentially of the same order as that noted in discussing the chart showing the relationship between weight and total calories (fig. 26, page 140). Indeed, by reference to the discussion on page 141, it will be seen that the differences there were, if anything, somewhat less. So far as this picture thus far goes, the evidence is slightly in favor of a greater degree of regularity in the relationship between heat and weight than between heat and body- surface, even though we are now referring to body-surface accurately measured and not approximately computed. A more rigid test of this, however, will be made when we come to consider in a subsequent section the utilization of the general smoothed curves in figures 26 and 38 as a basis for predicting the heat production of unknown subjects. What is of special physiological significance, however, is that the general picture presented by the chart in figure 38 indicates that the heat production referred to surface area is almost identical with that in which reference is made to body-weight. Extremely few children studied by earlier investigators can be compared to our measurements. They are confined exclusively to the METABOLISM AS AFFECTED BY GROWTH. 163 series of boys studied by Du Bois and his collaborators, whose body- surface areas were actually measured. If any attempt were made to use the values found by Magnus-Levy and Falk, it wonld be neces- sary to calculate the probable surface area based only upon the mea- surements given by Magnus-Levy for the weight and height of the children studied. Our analysis of the anthropometric data obtained in our studies shows us that from the weight alone we can compute reasonably close values for practically all children whose weight and sex are given, 1 and confirm them by inspection from some of the data actually obtained in our own series. Still, we do not wish to confuse the comparisons made in figure 38 by introducing computed surface areas, but do emphasize the fact that we are dealing here with a rela- tively recent physical measurement of children, namely, the surface- area by the Du Bois method. The values for Du Bois's 1916 and 1918 series for boy scouts have, for convenience, been included in figure 38, but it should be stated that these points were not laid thereon until after our smoothed curve was prepared. Though the range for our own values for surfaces, ages, or weights do not justify extensive comparison with the 1918 series of boy scouts studied by Du Bois, we have plotted the latter in figure 38 for purely comparative purposes. It is seen that in almost every case the values found by Du Bois in both series of experiments lie above our smoothed curve. While a few of these lie fairly close to the curve, most of them Ire considerably above it, indicating a metabolism of boys with these measured surfaces noticeably higher than that found by us. Although it is probable that physiologists as a rule will not hi the future make the Du Bois measurements a part of their regular records, such measurements are strongly recommended. At present the whole question as to the best index of physical character to correlate with measured metabolism is still in abeyance. As pointed out earlier (see page 53), the Du Bois measurements have a specific value entirely aside from that connected with the computation of the body-surface, in that they give typical girths and lengths which are of importance for indicating the several stages of growth. We strongly urge all pediatricians to include these measurements in their records and in detail. With the accumulation of a mass of data on this subject, further comparisons may be made, with important deductions to be drawn therefrom. Now that the Du Boises have given us an accurate physical measurement of surface-area, and as this measurement more probably approximates the true growth-changes than the measure- ment of mere weight, the significance of surface-area measurements per se should not be lost sight of by any worker hi metabolism. 1 See tables 12 to 15, pages 54 to 62. 164 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TOTAL CALORIES PER 24 HOURS REFERRED TO ACTUALLY MEASURED BODY-SURFACE AREA (GIRLS) . The extended series of Du Bois measurements were likewise made for most of the girls in this study, thus again permitting the comparison of the metabolism with actually measured surfaces. These data are brought together hi figure 39, upon which we have laid a curve showing the general trend. There is clear evidence of a general increase in 15OO 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 .2 / / '/ ' S' '/ /: . '/' /'! * t /' . '/ ' _/ r sq.m.3 4 .5 .6 .7 .8 .9 10 11 I 2 li FIG. 39. Basal heat production of girls per 24 hours referred to body-surface. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 1.272 sq. m. (puberty). metabolism with increasing body-surface, and the curve is not unlike that found for boys in figure 38. It also bears a striking resemblance to that found when the heat production was referred to body-weight with both boys and girls (figs. 26 and 27). The variation of the points above or below the central line is apparently not unlike that with the corresponding curve for body-weight (see fig. 27), and they scatter with reasonable regularity about a fairly straight line. Espe- cially divergent results are found at 0.52 square meter, and the two points showing a surface area above 1.2 square meters seemingly lie rather high. Evidently further observations are necessary with the METABOLISM AS AFFECTED BY GROWTH. 165 larger surface-areas to project this line properly. Hence we have purposely ended it at an area of 1.1 square meters. The main conclusions to be drawn from this chart, therefore, are that there is a distinct tendency for the metabolism to be increased as the surface-area increases and after the lowest areas it is a reason- ably straight-line function, so far as the general drift is concerned, up to areas of 1.1 square meters. Individual points, especially at the lower areas, lie so far from the general line that it would be difficult to conceive of any physiological regularity that would suggest an intimate relationship between surface-area and the total calories. Since none of the observations of the basal metabolism of girls in the literature were accompanied by surface-area measurements, no points for earlier values are placed on this chart and no textual dis- cussion is deemed desirable, as we are dealing here with measured metabolism on the one hand and measured body-surface on the other. CALORIES PER SQUARE METER OF BODY-SURFACE REFERRED TO BODY-SURFACE. From the charts comparing total heat with measured surface-area, it is clear that the larger areas give the larger total heat production and that there is a tendency above the lower areas to a straight-line Cals. 1300 CALORIES PER SO. M. REFERRED TO SURFACE. BOYS. 1200 1100 1000 900 800 700 .2sq.m..3 1.1 1.2 1.3 1.4 1.5 FIG. 40. Basal heat production of boys per square meter of body-surface per 24 hours referred to surface. Point inclosed in square signifies puberty established. relationship, yet the comparison between the heat production per square meter of surface w r ith the total surface shows these relation- ships much more clearly. Accordingly we have plotted the calories per square meter of body-surface referred to the surface, in figure 40 for the boys and in figure 41 for the girls. In figure 40 it is perfectly obvious that a straight-line curve or any- thing approximating a straight line will not correspond with the general trend. The sketched curve shows a rise in the calories per square meter with increasing area up to about 0.5 square meter; 166 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. thereafter there is a tendency for a slow but reasonably regular fall with increasing areas. The scatter of the points about this central line is very wide. If we take tentatively the line corresponding to 1,000 calories per square meter as running through approximately the center of the curve, the deviations either side of this line exceed 10 per cent in a large number of cases; for the children with the smaller body-surface they exceed plus or minus 25 to 30 per cent. The chart is important, however, in indicating substantially the earlier findings on the body-weight charts, in that with small surface-areas Cal8 - CALORIES PER SQ. M. REFERRED TO SURFACE. GIRLS. 1400 1300 1200 11OO 1000 900 800 700 600 2. . - ' ^ -s * / / ^< 9 , . . '. I ' *'. * ./ r - -. * ' / * . : .* A / sq.m.3 .4 .5 .6 .7 .8 .9 1.0 1.1 1.2 1. Fia. 41. Basal heat production of girls per square meter of body-surface per 24 hours referred to surface. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 1.272 sq. m. (puberty). there is a greater intensity of metabolism per unit of surface, not forgetting the intimate relationship between surface and weight, and that this does not depend upon a causal relationship between surface- area and heat-loss. While this chart is not readily compared with other charts on the weight and age bases, it is clear that the scatter of the points is so great as to rule out any conception of a physiological law indicating relationship between the heat production per unit of surface and the actual surface as measured on the child. In considering the calories per unit of surface-area referred to total area with girls (fig. 41), the characteristic wide scatter of points noted with boys is likewise here observed. While the line corresponding to 1,000 calories would roughly fit into the general middle of the curve, the deviations either side of this point range from to nearly 40 per cent for the children of smaller body-surface but from 0.7 to 1.1 sq. m., the points group themselves somewhat better about the central line than they do with boys. The specific high metabolism with the smaller children and the absence of any regularity in the chart as a whole which would suggest a physiological law should be emphasized. METABOLISM AS AFFECTED BY GROWTH. '167 We believe that this is the first time that actually measured surfaces have been so extensively applied in comparisons with the simul- taneously measured total calories. So far as the total calories are concerned, the general trend with both boys and girls is not unlike that for the total heat referred to weight. So far as the calories per square meter referred to area are concerned, the deviations from the central line are too great to permit of any theorizing as to the existence of a strict physiological relationship between heat production and surface area. COMPARISON OF CALORIC OUTPUT PER SQUARE METER OF BODY-SURFACE WITH TOTAL BODY- WEIGHT. In the discussion of the charts dealing with body-surface thus far, we have seen that with increasing body-surface there is an increased total heat production, but when the calories per square meter of body- surface are referred to the surface, the general trend shows an increase in the metabolism to an area of 0.5 sq. m., and thereafter a definite CALORIES PER SO. M. REFERRED TO WEIGHT. BOYS. 1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 600 2 . . / /^~ ~\. ^s, / \ \, . 7 ' 1 -. . ' / L -t-~^ jl . 1 ^ 50 ZSSr . __,_ /' I I 1 1 ' kgs.4 3 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 4 FIG. 42. Basal heat production of boys per square meter of body-surface per 24 hours referred to body-weight. Point inclosed in square signifies puberty established. decrease. These results throw absolutely no light upon the possi- bility of a relationship between area and heat production, for the surface may be simply another approximate expression of increasing body-mass, and particularly the mass of organic protoplasmic tissue 168 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. participating in heat production. Accordingly a comparison of the heat production per square meter of body-surface with children of different weights has a special interest. This comparison is made for the boys in figure 42. The wide scatter of points on this chart makes the drawing of any smoothed curve very problematical. About the only clear feature of the plot is that a straight-line curve will not represent the general trend, for there is decided evidence of a gradual increase in the heat production per square meter of body-surface up to about 12 kg. in weight, with a decrease thereafter. This increase, of course, has special interest for comparison with the rise in the intensity of meta- bolism in the first part of the curves shown on several of our earlier charts. This increase thus seems to be clearly established, whatever basis of comparison is employed. It is commonly believed that the calories per square meter are constant for all sizes and weights of individuals. The surface areas here shown are not computed or estimated, but are carefully measured ; yet on this chart we find variations in the calories per square meter ranging from 647 calories at a weight slightly under 4 kg. to a maxi- mum of 1,278 calories at a weight of about 12 kg. This great diversity occurs with the lower weights, but even after 12 kg., taking for example a point at approximately 28 to 30 kg., we have ranges from 886 to 1,205 calories. The sketched curve is not unlike that shown when the calories per square meter are compared with the actually measured surface-area (fig. 40), but it is impossible to consider such a wide scatter of obser- vations as indicating more than a general trend. While, as would be expected, a considerable number of points lie within 10 per cent, especially at the higher weight values, the large number of points lying outside these wide limits certainly does not justify the effort to stress this relationship as evidence of a physiological law. The calories per 24 hours per square meter of body-surface referred to weight, as plotted for the girls in figure 43, show such a wide diversity in the distribution of the individual points that only a complicated form of curve, which has obviously not been mathematically deter- mined, will serve to indicate the general trend. Characteristics of this sketched curve are the rapid increase in the values up to 10 kg. in weight, and a tendency for the values to decrease thereafter. Con- trary to the results of the observations on boys, there is with girls a clear tendency in the relatively few observations we have to indicate a flattening-out of the curve after 22 kg. Still, extreme caution is necessary in considering any data above 30 kg. as of real significance in indicating the general trend, since we have only three points beyond this weight. The two extremely high values between 36 and 40 kg. are the identical observations which showed abnormally high values METABOLISM AS AFFECTED BY GROWTH. 169 on the body-surface chart in figure 39. Further evidence at these weights is absolutely necessary for intelligent discussion. The accumu- lation of such data is now in progress at the Nutrition Laboratory. CALORIES PER SQ. M. REFERRED TO WEIGHT. 1300 . 1250 1?00 115O * 1100 .^ 1050 ' / / ^ \ 1000 / ^ X 950 900 '/ /T . r~ ---~. -~. ~-r^t ii * = flfiO / . 800 / 7RO ./ 700 / ~* 650 /' 600 . ;k S ..< ( $ t 1 D 1 2 1 * 1 5 1 i 2 2 2 2 4 2 6 2 8 3 3 2 3 4 3 6 3 8 4 FIG. 43. Basal heat production of girls per square meter of body-surface per 24 hours referred to body-weight. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 39.2 kg. (puberty). Although the tendency is for the highest metabolism per square meter of body-surface to be noted at about 10 kg., some of the lowest values found in the entire series are also obtained about this time. A large personal element in basal metabolism is therefore apparent, certainly with the younger children. The general form of the curve is somewhat different from that on the same basis shown for the boys, but this is not the place for a special discussion of sexual differentiation. COMPARISON OF HEAT PRODUCTION PER SQUARE METER (MEASURED), REFERRED TO BODY- WEIGHT, WITH EARLIER DATA (COMPUTED). As brought out in preceding sections, certain difficulties occur in comparing the results of our study, in which the basal metabolism was directly determined and the surface-area actually measured, with those of earlier investigators, who, aside from Du Bois, were unable to measure the body-surface. Since, however, data were obtained in the 170 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. earlier investigations for the total metabolism and the body-weight, and occasionally for height and age, we may compute the surface areas for the children studied, probably with considerable exactness. In computing the body-surfaces of children whose metabolism is given in the earlier literature, we have been guided in large part (1) by the clearly established legality of the Lissauer formula for chil- dren up to 10 kg. in weight; (2) by the reasonably satisfactory agree- ment of the Du Bois height-weight chart down to 20 kg. with body- surfaces actually measured. For the children studied by Murlin and Hoobler, the body-surface was computed by means of the Lissauer surface formula. For the majority of those studied by Magnus-Levy and Falk, the surface areas were derived from the height-weight chart of Du Bois. Since no heights were given for Scharling's children, the two boys of Sonde"n and Tigerstedt, and a few of Magnus-Levy and Falk's children, the surface areas were computed from the body-weights by means of the formula K ^5* and our new values for the factor K. (See tables 14 and 15, pages 61 and 62.) The use of these computed surface areas is permissible in computing the calories per square meter of body-surface; but it hardly seems that such a computed area can be used in a comparison of the calories per square meter of body-surface and the body-surface itself, and this has not been attempted. In referring the calories per square meter to weight and to age, the units of measurement for comparison are perfectly definite and hence proper to use. To avoid any misunderstanding, we call attention again to the fact that for some of the very young children (20 boys and 19 girls) in our own series of observations, the body-surfaces were not measured, but were likewise computed. 1 For these children the Lissauer formula was used, as it was not possible to employ the height-weight chart, since the body-weight was less than 20 kg., which is the lower limit of this chart. As our own results show that the Lissauer formula agrees remarkably well with the measured surfaces for other children weighing less than 10 kg., we feel justified in including these in our tables with the actually measured surfaces. If we disregard the age of the child measured by Sawyer, Stone, and Du Bois, 2 and consider only weight, it seems quite reasonable to argue that since the Du Bois measurements applied to a child weighing a little over 6 kg. and the Lissauer formula and the linear measure- ments agree up to 10 kg., we can logically employ the Du Bois linear formula in measuring the body-surface of children weighing somewhat less than 6.3 kg. Du Bois specifically questions the use of the linear formula for body-surface calculations for children under 2 years of age. It would appear, however, that the agreement between the Lissauer 1 See tables 27 and 28, pp. 116 and 120. 2 Sawyer, Stone, and Du Bois, Arch. Intern. Med., 1916, 17, p. 855. METABOLISM AS AFFECTED BY GROWTH. 171 formula and the Du Bois surface formula is not a purely accidental one, but represents in all probability a reasonably close physiological relationship. Accordingly, while we must recognize that the Du Bois linear formula may not, according to its authors, be used at present for children under 2 years of age, we have frankly disregarded this and have employed it for these younger children in computing the body- surfaces from the actual measurements. Until it has been disproved that the linear formula does not apply for children so young as this, we believe it is justifiable to use. it in this way. If body-surface still continues to attract the attention of physiologists as much as it has in the past decades, a complete verification of the linear formula for children under 2 years of age may naturally be expected inside of a few years. CALORIES PER SO- M. REFERRED TO WEIGHT. 1350 1300 1250 1200 1150 11 00 1050 1000 950 900 850 800 750 700 650 600 2 * . . / /^ ^ \ / \ . g c * / ^ " . . 1 =. ! / t . ' . 1 , /. y / / cr-5 CHA >OND /AGN =iLINC EN A US-L ND T EVY / GER ND 5TED "ALK A-I rfURL N AN DHO OBLE R / x-5 -t r o-DU BOIS (1916) o-DU BOIS (1918) kgs.4 < FIG. 44. Basal heat production of boys per square meter of body-surface per 24 hours referred to body-weight (earlier investigators). In the comparison of our metabolism data with earlier studies, in which special attention is called to the total calories per square meter per 24 hours referred to body-weight, it is necessary to bear strictly in mind that the surface-areas of the children studied by the earlier investigators were all computed, except those for the Du Bois boy scouts. We have charted the values for boys on the usual scale in figure 44. The sketched curve, which shows the general tendency 172 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. found in our own results, has also been laid on the chart, although here again it must be emphasized that this curve was drawn with great difficulty and simply shows the general trend in a diagram, with very wide dispersion of the various points. While comparison with figure 42 shows that few of these points lie outside of the field of our observations, yet we should note that, with the exception of one boy weighing about 6 kg., the points for the children earlier studied lie above our sketched curve. Furthermore, if a smoothed curve were laid through the earlier observations, it would have approximately the same general trend as our own curve, although at a higher level. While our own data do not warrant the extension of our smoothed curve beyond 42 kg., the data for certain heavier boys should here be recorded. In the 1918 series of Du Bois, the values for four boys, all weighing about 49 kg., are reported and are given in table 30. With all four boys the surfaces were actually measured. TABLE 30. Values outside weight range in figure 44 COt* Bois) . TABLE 31. Values outside heat range in figure 44 (Magnus-Levy and Falk) . Cals. per Kilos. sq. m. per 24 hrs. 49.1 1,054 49.1 1,037 49.3 1,039 48.6 938 Method. Kilos. Cals. per sq. m. per 24 hrs. Height-weight chart 20.8 1,459 K^lw 1 11.5 1,448 *** 14.5 1,468 Magnus-Levy and Falk report the heat production of three boys, each with such a high value that it would be necessary to extend our chart beyond permissible limits to include them, although they weighed less than 21 kg. Table 31 shows the data for these three boys. It is clear that the points for 6 of the 7 boys would lie very con- siderably above our line if projection beyond the 42-kg. level were permissible. Since we believe that the projection would in all prob- ability fall somewhat lower rather than continue on a level, all of the 7 boys would be above the projected line, although the value for the boy weighing 48.6 kg. studied by Du Bois would fall practically on the line. With girls the number of earlier observations is extremely few, hardly justifying special chart treatment. The values have been plotted, but are not reproduced here. All of the values are con- siderably above our smoothed curve. The earlier data for girls are altogether too sparse to warrant even an attempt to lay on a smoothed curve, so no comparisons with our general trend are possible. METABOLISM AS AFFECTED BY GROWTH. 173 Thus, with girls, as with boys, practically all of the earlier work shows a metabolism so much higher than that found by us as to lead us to suspect that strict maintenance of muscular repose was not insisted upon in these earlier observations. Our own results obvi- ously more nearly approach basal, though still not uncomplicated by a possible effect of previously eaten food. AGE RELATIONS IN THE HEAT PRODUCTION PER SQUARE METER OF BODY-SURFACE. Thus far the consideration of the heat loss per square meter of body- surface for our subjects has been confined to the two bases, weight and measured body-surface. In view of the small and clearly established age relationship in the metabolism of adults, it is important to note whether age has a special effect upon the metabolism of rapidly growing children. As previously pointed out, the differentiating of the age effect is difficult, owing to the fact that age changes are concurrent with weight, stature, and surface changes. Still, for purposes of special discussion in subsequent chapters, it is advisable to consider the heat per square meter referred to age exactly as we have studied the total calories and the calories per kilogram of body-weight referred to age. This is particularly true when it is recalled that, beginning with the days of Andral and Gavarret, special emphasis has been placed upon the influence upon metabolism of approaching puberty; hence in our curves it is desirable to note the ages of the various individuals and to find the general trend of metabolism at these ages, independent of weight or surface, save as the surface area is partly compensated by computing the heat per square meter exactly as weight has been partly compensated in previous comparisons by computing the heat per kilogram of body-weight. Figure 45 gives the values for boys for heat per square meter of body-surface referred to age. Recalling that age changes are in the main concurrent with weight and surface changes, it can be seen that the general trend of this curve is somewhat similar to that for the cal- ories per square meter when referred to weight (fig. 42) and when referred to body-surface (fig. 40), exhibiting an increase during the early years up to an age of about 2 years, with a tendency towards a straight-line decrease thereafter. The scatter of individual points makes it difficult to lay on a smoothed curve. We do not defend this use of a straight-line curve and can only take refuge in our oft-repeated statement that the line must be understood to indicate only a trend and may not be referred to with any mathematical exactness. As in previous charts, one of the boys is marked as showing un- mistakable signs of puberty, this point lying above our smoothed curve. Attention will be again called to this fact in subsequent discussion of the influence of the prepubertal stage upon metabolism. 174 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. The chief importance of the curve in figure 45 is to show the pro- nounced metabolism of early youth up to 1 or 2 years. As we have already stated, there can be no great preponderance in the proportion of active protoplasmic tissue in a child 1 or 2 years of age. In all probability we have here a specifically high metabolism at this age, which is certainly worthy of further investigation, with, if possible, a minimizing or elimination of the disturbing factor of stimulus from food (see page 30). Cals. CALORIES PER SQ. M. REFERRED TO AGE. 67J 12 13 14 15 16 FIG. 45. Basal heat production of boys per square meter of body-surface per 24 hours referred to age. Point inclosed in square signifies puberty established. For the studies on boys made by earlier investigators, a plot is made in figure 46 and our smoothed curve included in the chart. The usual experience is here repeated, namely, that the earlier values, with but one exception, lie above the smoothed curve; in most in- stances, the points are very much above the line. Especially worthy of note are the high values recorded by Du Bois between the ages of 12 and 14 years. These have brought once more into active discussion the influence of the prepubertal stage upon basal metabolism. Our observations on girls are all plotted in figure 47, upon which is laid a somewhat complicated smoothed curve, this being the result of an attempt to represent the general trend in spite of the great irregularity and wide dispersion of individual points. The most clearly established feature of the curve is the prevalence of low values at the early ages, rapidly ascending until about 1 or 2 years, with a tendency towards a subsequent decrease. With two of the girls, puberty was definitely established and with one other it was beginning. Of the METABOLISM AS AFFECTED BY GROWTH. 175 two former, one point lies below the curve and the other considerably above it. The point for the third girl (12 years and 2 months) lies slightly below the smoothed curve. With one of these girls, measure- ments were made prior to and subsequent to puberty. At the age of CALORIES PER SQ. M. REFERRED TO AGE. rf-SCHARLING .SONDEN AND TIGERSTEDT -MAGNUS-LEVY AND FALK ^.-MURLIN AND HOOBLER o-DU BOIS(1916) D-DU BOIS0918) FIG. 46. Basal heat production of boys per square meter of body-surface per 24 hours referred to age (earlier investigators). CALORIES PER SO- M. REFERRED TO AGE. GIRLS. FIG. 47. Basal heat production of girls per square meter of body-surface per 24 hours referred to age. Points inclosed in squares signify puberty established. For No. 239 compare point inclosed in diamond (prepubescence) with point inclosed in square at 12 years 1 month (puberty). 176 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. 11 years, prior to puberty, the heat per square meter was 909 calories. Somewhat over a year later, when the child was 12 years and 1 month old, the heat per square meter had increased to 1,179 calories. This will likewise receive special consideration subsequently. Bearing in mind that age, weight, and surface area changes are so closely interwoven, it is impossible in any of these age charts to emphasize specifically an age influence per se, other than to draw attention to the high point at the age of 1 or 2 years which seems unquestionably to be a specific age influence. The extremely few girl subjects in the earlier literature make a comparison with our work impracticable. Such few observations as are on record that can be taken as approximating basal (chiefly those of Magnus-Levy and Falk) would all lie considerably above our general smoothed curve. Special chart representation for these scattered observations hardly seems necessary. INFLUENCE OF SEX AND SEXUAL CHANGE ON METABOLISM. The wide differences in activity and dietetic habits of boys and girls, commonly observed in every household, early led to a belief in a striking difference between the metabolism of children of the two sexes. Unfortunately, many of the earlier comparisons disregarded body-weight and considered age only. While age-changes with both boys and girls are closely followed by weight-changes, for a strict comparison it is obvious that one may not compare a 12-year- old boy weighing 38 or 40 kg. with a 12-year-old girl weighing 30 kg. Even the early experiments of Andral and Gavarret 1 are used by the authors as the basis for considerable discussion of the differences between boys and girls. As the authors did not report the body- weights of the children, we can not recompute the data on the better basis of energy per kilogram of body-weight. They conclude that with boys and men there is a steady increase in the carbon-dioxide production from 8 to 30 years, and that between these ages the carbon- dioxide production is greater in amount than that of girls and women of similar ages. Furthermore, they believe that the sexual difference is most pronounced in the adult period (16 to 40 years), the exhalation of carbon dioxide by man during this period being about twice as much as that of woman. The work of Andral and Gavarret is admir- ably presented by Sonden and Tigerstedt 2 in connection with the discussion of their own researches. As the results of Andral and Gavarret or of Sonde"n and Tigerstedt may not be looked upon as basal in character, the comparison is probably justifiable, since in all 1 Andral and Gavarret, Ann. d. Chim. et d. Phys., 1843, s6r. 3, 8, p. 129. 2 Sonden and Tigerstedt, Skand. Arch. f. Physiol., 1895, 6, pp. 54 and 56. INFLUENCE OF SEX ON METABOLISM. 177 cases the subjects were fairly quiet, although modern conditions for basal measurements were not obtained. That the question of body-weight was seriously considered by Andral and Gavarret is shown by a portion of their discussion, but they discard this method of comparison as irrational, maintaining that while a woman 25 years old weighs much more than a child 10 to 14 years old, she produces no more carbon dioxide. They further contend that at the menopause the body-weight of women does not necessarily increase, and they have shown that the exhaled carbon dioxide at that time continually increases. It is a matter of regret to modern writers that, although the body-weights of the subjects were known to these investigators, they did not publish them. Special emphasis was laid by Andral and Gavarret upon the influence of puberty on metabolism. They maintain that with boys the carbon dioxide exhaled increases suddenly to a large amount at the time of puberty, while with girls, on the contrary, the carbon dioxide excretion suddenly ceases to increase at this period and remains nearly stationary until the menopause. It then suddenly increases in a very remarkable manner, and finally, as with man, decreases in proportion to the approach to extreme old age. Speck 1 studied only three children within the age-range observed by us (a girl of 10 years, a girl of 13 years, and a boy of 13 years), and his conclusions were largely based upon experiments made with older individuals; hence they have little immediate significance in our discussion. Most of the discussion as to the difference between the sexes in the classic paper of Magnus-Levy and Falk 2 pertains to the ages beyond those of childhood. According to their results, the total metabolism of boys during the years of puberty did not exceed that of adult normal men. With some of the boys the metabolism was less and with some approximately the same as that found with adult men. They men- tion particularly one boy, accustomed to metabolism experiments, who was studied when he was 16 years old and weighed 57.5 kg., and again 6 years later, when he was 10 kg. heavier. During the period of establishment of puberty the oxygen consumption was 235.6 c. c. per minute and the carbon-dioxide production 192.2 c. c. When puberty had been fully established, the oxygen consumption was 231.3 c. c. and the carbon-dioxide production 200.2 c. c. per minute. In other words, approximately the same values were found at both times. In the later stages of the period of pubertal change, therefore, the total metabolism corresponded to that found after puberty had been fully established. Since there was an increase in weight, the metabolism per kilogram of body-weight obviously decreased, i. e., 1 Speck, Physiologie des menschlichen Athmens, Leipsic, 1892, p. 217. 2 Magnus-Levy and Falk, Archiv f. Anat. u. Physiol., 1899, Suppbd., p. 314. 178 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. the oxygen consumption from 4.10 c. c. to 3.43 c. c. per minute per kilogram. Their conclusions are as follows: The gaseous exchange of children per unit of weight is greater than with adults, being larger the younger and lighter the child; this does not apply to the first year of life. Per unit of body-surface (Meeh) , the metabolism of children is much greater than that of adults, but during the first year of life it is prob- ably somewhat lower than during later child life. The metabolism of females is not actually less than that of males; certainly with adults there is no difference. In this respect they do not agree with Sonden and Tigerstedt, who believe that the meta- bolism of women is less than that of men. With the younger children the gaseous exchange per kilogram of body-weight for girls is somewhat less than with boys; with larger children the gaseous exchange is about the same with boys and girls. In general, Magnus-Levy and Falk conclude that the metabolism of women in middle life is approximately the same per kilogram of body- weight as that of adult men of the same age and weight; with children and elderly people, the metabolism of females is slightly less than that of males (about 5 to 10 per cent). Finally, we should refer to the conclusions of Sonde"n and Tigerstedt, 1 although these were not founded upon basal metabolism measure- ments. These investigators maintain that in general the carbon- dioxide production of boys, on the basis of both weight and surface, is considerably greater than that of girls of the same age and body- weight, and that the carbon dioxide production of girls, both per kilo- gram of body-weight and per square meter of body-surface as computed by the Meeh formula, is to the boys as 100 is to 141. This finding, as Sonde"n and Tigerstedt point out, was earlier suggested by Scharling 2 and Speck, 3 as well as Andral and Gavarret, 4 although the experiments of Scharling and Speck were so few as to make their deductions little more than speculation. In considering the several charts and diagrams for the measurements made upon our boys and girls, we have occasionally hinted at small but obvious sexual differences in the general form of the curves. Still, from a casual inspection of the individual curves, it would be almost impossible to assert the presence of a pronounced sexual differentiation. For the special purpose of noting sex differences, if they exist, comparison should be made upon the same chart. In figure 48 the total calories are referred to body-weight for both boys and girls, the two curves being taken directly from figures 26 and 27, 1 Sonden and Tigerstedt, Skand. Archiv f. Physiol., 1895, 6, p. 95. Scharling, Ann. d. Chem. u. Pharm., 1843, 45, p. 214; reprinted in detail in Ann. d. Chim. et d. Phys., 1843, ser. 3, 8, p. 478. 8 Speck, Physiologic des menschlichen Athmens, Leipsic, 1892. 4 Andral and Gavarret, Ann. d. Chim. et d. Phys., 1843, ser. 3, 8, p. 129. INFLUENCE OF SEX ON METABOLISM. 179 respectively, and here superimposed. In this comparison of total calories to weight, we find that the absence of a real sexual difference shown with new-born infants 1 persists until about the weight of 11 kg., and that subsequently there is a tendency for the boys to have a somewhat higher metabolism than girls of the same weight, the line for boys being perceptibly above that for girls. Cals S200 2000 1800 1600 1400 1200 1000 800 600 400 200 TOTAL CALORIES REFERRED TO WEIGHT. '^ - ---" ^"' ^ ^=~ *. . :== ., . _~ - ^ -"' - ' __~- _ ' ^f' ^ ,' s/^ S ^> GIR BO LS AND WOMEN /S AND MEN j /\ / > 2kgs. 10 14 18 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 FIG. 48. Comparison of basal heat production of children and adults per 24 hours referred to body-weight. Attention should here be called to the fact that in the preliminary presentation of this material, 2 a somewhat different relationship between the curves for boys and girls was noted in that the curve for girls rose above that for boys at the weight of 35 kg. Subsequent revision and elision has justified the caution pronounced at that time to the effect that the number of individuals measured at the higher weights were, with both sexes, too few to justify sweeping conclusions. As here shown in the latest and more complete analysis of our material, we have clear evidence of a sexual differentiation in basal metabolism exhibited above 11 kg., in which the boys show persistently a somewhat higher metabolism than girls of the same weight. Since a comparison of the metabolism of youth and adults is of general interest, the trends of the metabolism on the basis of total calories referred to weight are also shown in figure 48 for men and women. These curves will be given special discussion later. A comparison may also be made of the two curves for boys and girls in figures 38 and 39, in which the total heat was referred to actual measurements of the body-surface by the Du Bois method. This 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915; Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 219. 2 Benedict, Boston Med. and Surg. Journ., 1919, 181, p. 107. 180 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. comparison has been made in figure 49. As has been pointed out, the general trend of the two curves is alike and the two sexes remain at essentially the same metabolism until the area is 0.48 sq. m. From this point the line for the boys rises above that for the girls, and there is no evidence of a tendency for the two lines to cross later. Figures 48 and 49 thus give clear evidence of a sexual differentiation between boys and girls, with the boys on the whole showing a higher meta- bolism after the body-weight and body-surface have reached 11 kg. and 0.48 sq. m., respectively. 2200 2000 1800 1600 1400 1200 1000 800 600 400 200 TOTAL CALORIES REFERRED TO SURFACE. / x X i / x /x X X ^ S ^ j // / GIR LS Al rS Atv >JD W D Mf DMEr IN - 7 BO^ * .2sq.m..4 .6 10 12 14 1.6 1.8 2.0 2.2 FIG. 49. Comparison of basal heat production of children and adults per 24 hours referred to body-surface. The comparison of the total calories to measured surface is, however, a somewhat novel procedure, and physiologists are more accustomed to comparisons of the calories per kilogram of body-weight and the calories per square meter of body-surface referred to weight. Since both these bases of measurement should show a sexual differentiation, if such exist, we have prepared charts giving these comparisons. (See figs. 50 and 51.) The curves for boys and girls in these two charts were taken directly from figures 35, 36, 42, and 43, and are here simply superimposed to bring out the sexual differentiation, which is essentially that noted with the weight curves. To make an approximate comparison of the heat production per kilogram of body-weight and per square meter of body-surface between youth and adults, we have laid on these charts lines representing grossly the trend of metabolism with women and men with increasing INFLUENCE OF SEX ON METABOLISM. 181 weight, as was done in figure 48 for the total calories referred to weight. It is highly important that these curves should not be interpreted as giving a sharp picture of the actual metabolic changes with increasing weight. Accordingly, since these curves were drawn from plots representing all of our adult measurements, we publish at CALORIES PER KILO. REFERRED TO WEIGHT. ^^ // \ \ \ GIR B0\ LS AND W 'S AND ME OME N x ^ ^ ^~~--~. ., > ~- ~~-^ ---.^ *" "^ , , - = , " . . r== . : ' ._ FIG. 50. Comparison of basal heat production of children and adults per kilogram of body-weight per 24 hours referred to weight. this point the charts from which these adult curves were derived. (See figs. 52, 53, 54, and 55.) The laying of a smoothed curve on these charts is, owing to the scatter of the points, extremely difficult. While the straight line which is at a constant level hi the case of men when the heat per square meter is referred to weight (fig. 54) would seem to be an admission of the constancy of heat production per Cals. 1200 1100 1000 900 800 700 600 t CALORIES PER SQ. M. REFERRED TO WEIGHT. < ^ i GIRLS AND WOMEN BOYS AND MEN / \ "^s s^. / / "*--. - . =L L__ 1 7 1 kgs.6 10 14 18 22 26 30 34 38 42 46 50 54 58 62 66 7O 7 4 78 82 86 90 9- FIG. 51. Comparison of basal heat production of children and adults per square meter of body-surface per 24 hours referred to weight. square meter of body-surface with men, we believe that no one in- specting this chart, with its wide scatter of individual points, can conclude that this line, which represents trend only, can be logically looked upon as a demonstration of the general thesis that the heat production per square meter of body-surface with man is constant. With women it would seem as if a slight slant to the line more closely 182 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. fitted the general trend, but here again, owing to the wide scatter of the points, there is no evidence of regularity in the relationship between heat production and surface area. With youth, even with the wide scatter of the individual points in figures 42 and 43, the lines denoting the general trend show pronounced deviation from a uniformly hori- zontal level. CALORIES PER KILO. REFERRED TO WEIGHT. MEN. 46kgs.50 54 58 62 66 70 74 78 82 86 90 94 FIG. 52. Basal heat production of men per kilogram of body-weight per 24 hours referred to weight. Several important points should be emphasized in connection with the comparisons of the metabolism of youths and adults on all bases of comparison as indicated in figures 48, 50, and 51. One is the quite remarkable coincidence of the extension of the line for boys with that laid down for men, an agreement that is somewhat less striking in the case of the girls and women. The data for the higher-weight boys and lower-weight men, and particularly the higher-weight girls and lower-weight women, are still altogether too few to consider this part of the curves as clearly established. At the moment of writing, Cals. 36 CALORIES PER KILO. REFERRED TO WEIGHT. 30 =, 24 12 34kgs.38 42 46 50 54 58 62 66 70 74 78 82 86 90 94 FIG. 53. Basal heat production of women per kilogram of body-weight per 24 hours referred to weight. further experimental data are being obtained at the Nutrition Labor- atory for ages between 12 and 20 years for both sexes. Finally, empha- sis must again be laid upon the fact that all of these lines represent at best only general trends, particularly with adults. 1 From a consideration of all of the charts in which the metabolism curves for boys and men on the one hand and girls and women on the other are compared, it is evident that the metabolism of boys and men is for practically the entire period of life perceptibly and con- 1 See page 132 for description of method of sketching curves. INFLUENCE OF SEX ON METABOLISM. 183 sistently higher than that for girls and women. While new data on the uncertain period between the weights of 30 and 45 kg. may some- what modify these general curves, nevertheless it appears clearly established that males have on the whole a higher metabolism than CALORIES PER SQ. M. REFERRED TO WEIGHT. MEN. 1050 950 850 750 4 1 : L. :. \ -.'. ..-.; ' . .' " *kgs.48 52 56 60 64 68 72 76 80 84 88 9J Fio. 54. Basal heat production of men per square meter of body-surface per 24 hours referred to body-weight. females. When we compare the calories per kilogram referred to weight and the calories per square meter referred to weight, it can be seen that after a weight of about 14 kg. the differences between the two sexes remain almost uniformly constant throughout the entire weight-range. CALORIES PER SQ. M. REFERRED TO WEIGHT WOMEN. 650 36kgs.40 44 48 52 56 60 64 68 72 76 80 84 88 92 96 FIG. 55. Basal heat production of women per square meter of body-surface per 24 hours referred to body-weight. METABOLISM IN PREPUBESCENCE. Any discussion of the sexual differences between boys and girls would be incomplete without reference to the important transforma- tion in sexual life taking place at the time of puberty, a transformation that, in the minds of a number of writers, has been clearly apparent in the course of the general total metabolism. Even the early studies of Andral and Gavarret 1 gave this factor serious attention, but un- fortunately their data are too meager to be of much value at this tune. In 1916 there appeared the first of two remarkable papers by Dr. E. F. Du Bois, 2 of New York, in which the metabolism at the period of development of boys immediately preceding puberty was studied. The importance of the problem under consideration can be no better set forth than by quoting the initial paragraph of this paper. Andral and Gavarret, Ann. d. Chim. et d. Phys., 1843, ser. 3, 8, p. 129. Du Bois, Arch. Intern. Med., 1916, 17, p. 887. 184 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. "In the period of development of boys, the years immediately preceding puberty are of especial interest. By this time the figure has lost most of its childish characteristics and the mind has reached a point of great in- telligence. Although the individual has scarcely passed the half-way mark in the years of growth, and has only attained half his future weight, yet he resembles the adult much more than he resembles the infant. At this stage the sex glands have not yet begun the rapid development of puberty with its profound effect on the whole organism. Curiously enough there is a sudden increase in the rate of growth which takes place at this time. In fact, we may consider boys in the period of prepubescence as individuals of adult form but of small size, growing rapidly, and as yet scarcely influenced by the internal secretions of the sex glands. The study of their respiratory exchanges may throw light on many problems." While we unfortunately have to differ with Du Bois as to these experiments meeting the strict requirements for basal metabolism, no paper has called more attention to the possibilities of changes in metabolism in youth as compared to the adult period than has this. We feel that the very high values found by Du Bois must have been due in large part to muscular activity; consequently, a comparison is of interest between his results and those of our observations with boys between 12 and 13 years old, which were obtained under conditions more closely approximating basal. On reference to figure 22 (page 133), we find records of 6 boys between the ages of 12 and 13 years. Of these, 2 lie above the pro- jected line indicating the general trend of the metabolism, and 4 below. From an examination of this line alone, one can see no increase in the general trend of metabolism peculiar to this age; but since there may be the reasonable objection that our boys were of abnormal weight for their age, we may note the effect of computing the calories per square meter. An examination of figure 45 shows us that of these 6 boys, 2 are again above the line representing the general trend and 4 below, with no indication of a supernormal metabolism. While the greater part of our observations were made with children in the period of prepubescence, a few observations were made with boys and girls after puberty was established. These are indicated on the several charts by a special designation, i. e., by a square surrounding the point. In figure 22 only one boy is so indicated, the point lying somewhat above the general trend. In figure 45, in which the calories per square meter of body-surface are referred to age, this point like- wise lies above the line for the general trend. From this one observa- tion, therefore, one might infer that there was a slight tendency for an increase in metabolism after the onset of puberty rather than prior to it; but obviously no special consideration should be given to a single observation, especially as the determined value is not much above the general trend. The influence of prepubescence upon the metabolism of girls is like- wise of special physiological interest. We note in figure 23, which INFLUENCE OF SEX ON METABOLISM. 185 gives the total calories referred to age, that within the age limits of approximately 10 to 12 years three points are specially designated to show puberty, one very considerably above the general line and two below. With two of these girls, puberty w^as well established, one at 10| years and the other at 12 years and 1 month, while with the third child, at 12 years and 2 months, puberty was beginning. It was possible to make observations with one of these girls (No. 239) at 11 years, before puberty began. At the earlier age (specially desig- nated by inclosing the point in a diamond) the metabolism was 984 calories, while 1 year and 1 month later, after puberty was established, it was 1,500 calories. Only in this one case is there apparent evidence of a striking effect of puberty itself upon metabolism. An examination of the rest of the curve shows no indication of a pronounced increase in metabolism during the prepubescent age. The evidence with this one girl in figure 23 is, however, very decep- tive, since on reference to figure 27, in which the total calories are referred to weight, it is seen that in the year and one month inter- vening between the two observations, this subject had gained in weight practically 12 kg. As a matter of fact, she had also increased 14 cm. in height. Accordingly, comparison may not be made directly on the basis of total calories referred to age without taking into con- sideration some of the physical factors. With the striking increase in weight, the first obvious correction would be to consider the calories per kilogram of body-weight with this subject at the two ages. This has been done in figure 31, and here we note that the striking differ- ences have practically disappeared, namely: at the age of 11 years, the heat is 36 calories per kilogram of body-weight; at 12 years and 1 month, 38 calories. Precisely the same relationship holds true when we compare calories per kilogram of body-weight at different weights, as shown in figure 36. Consequently, a portion at least of the striking difference in total metabolism noted with this child before and after the establishment of puberty must be ascribed to the pronounced alteration in body-weight. Having in mind the older conception of the significance of the body- surface area, we should also consider the caloric output per square meter of body-surface with this child. On reference to figure 47, where the values are compared on this basis, we find that after the age of 10 years there is a wide scatter of individual points. The one girl at the age of 10| years with puberty established shows a value considerably below the general line. Another at 12 years and 2 months, with puberty just beginning, is likewise slightly below the line. The child with values before and after establishment of puberty (No. 239) is represented at the later age, 12 years 1 month, when puberty was fully established, by the highest point on the chart. The heat production per square meter per 24 hours for this girl at 11 years 186 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. was 909 calories, while at 12 years and 1 month it had risen to 1,179 calories, or an increase of nearly 30 per cent. The general picture is therefore essentially the same as that noted in figure 23 for the total calories referred to age. When the caloric output per square meter of body-surface is referred to body-weight, as in figure 43, a wide difference in the heat production per square meter at these two weights is exhibited. While, therefore, the great difference in total heat production noted on the chart for total calories referred to age (fig. 23) is in large part removed by refer- ence to heat production per kilogram of body-weight (fig. 31), there still remains a very striking difference between the prepubertal and pubertal stage when the heat production per square meter is con- sidered. Precisely the same order of differences is to be observed with this girl when the total calories are referred to the surface and the calories per square meter are referred to surface in figures 39 and 41, respectively. What little evidence, if any, can be drawn from these charts for boys and girls as to the influence of prepubescence and the establish- ment of puberty on metabolism may be summed up in the statement that prior to puberty there is no tendency for a change in the general trend of the basal metabolism. The establishment of puberty in at least one girl resulted in a relatively high metabolism, which was made clear by measurements both prior to and subsequent to the establish- ment of puberty. Sonde"n and Tigerstedt 1 and Olin 2 report no increase per kilogram or per square meter following puberty. Since with our boys and girls we did not find the increase in metab- olism which Du Bois found, it becomes necessary for us to disagree with his findings for boys 12 and 13 years old, and we believe our con- clusion is justified, that the prepubescent period is without significant effect upon the metabolism of the boy or girl. The evidence regarding the effect of puberty fully established is sufficiently divergent in the existing researches to warrant much further study on this important point. Du Bois's evidence indicates strongly a decrease in metabolism following the prepubescent stage. The very meager evidence we possess indicates a tendency to an increase, although admittedly this is largely based upon the measurements of one girl, which were made before and after the establishment of puberty. Had not the question of the influence of puberty upon the metabolism been raised by earlier investigators, we should not feel that we were in any way justified on the basis of this one experiment in discussing the question. We believe that the chief point to be raised here is that further studies with children of this age are imperative. Such investigation is now in progress at the Nutrition Laboratory. 1 Sonddn and Tigerstedt, Skand. Archiv f. Physiol., 1895, 6, p. 75. 2 Olin, Skand. Archiv f. Physiol., 1915, 34, p. 432. PREDICTION OF THE BASAL METABOLISM OF YOUTH. 187 THE PREDICTION OF THE BASAL METABOLISM OF YOUTH.i In the last analysis, one of the most important factors in a metab- olism study is the potentiality of drawing from the data a method of predicting the unknown metabolism of a subject. The establish- ment of a normal, when based upon unvarying laws of either physics or chemistry, results in a standardization of values that makes possible the immediate estimation of the probable resultant of any two or more physical factors. In physiology the normal variation is so great as absolutely to preclude a mathematically established standard without deviations therefrom. On the assumption that the children in this study were normal in the commonly accepted use of that word, we do find, however, that the total metabolism follows a reasonably uni- form curve in some of the relationships studied. While nearly all of our numerous charts show a wide scatter of points, certain charts are characterized by a fairly close grouping of points around the hypothetical line indicating the general trend. These are specially the charts showing the total calories referred to body-weight for both boys and girls (figs. 26 and 27, pages 140 and 142) and the total calories referred to actually measured surfaces for both boys and girls (figs. 38 and 39, pages 161 and 164). From these curves it would appear that it is not impossible to predict with reasonable accuracy the probable basal metabolism of a child when either the weight or the measured body-surface is known. The body- weight is very readily and frequently obtained, but the Du Bois method of measuring the body-surface, while very specific and readily acquired after a little careful consideration of the directions, never- theless requires such an extensive series of measurements as to make its general use impractical. In general, the grouping of the individual points about the curves in figures 26, 27, 38, and 39 is often so seemingly compact as to suggest that the curve sufficiently represents the general trend for it to serve as a foundation for the prediction of the basal metabolism of unknown subjects. Still, on closer inspection, the scatter of individual points is noted to be considerable. One deceptive feature of these curves, making an intelligent comparison with curves for adults difficult, is the fact that the deviation from the general line varies greatly as to its percentage value according to the weight of the individual or the size of the surface area; consequently it is only with percentage relationships that one may properly deal. From an inspection of the curves for calories referred to both surface and weight, it is difficult to estimate with the eye as to which basis, i. e. } surface or weight, would give the better method for prediction, although in the earlier consideration of these curves we noted that >A preliminary discussion of this point has recently appeared: Benedict, Proc. Nat. Acad. Sci., 1920, 6, p. 7. 188 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. there seemed to be a somewhat wider dispersion of points about the curves for surface than about the curves for weight. For an exact analysis of, (1) the suitability of these curves for prediction, and (2) of the relative merits of surface and weight curves for prediction, mathe- matical treatment of each child on the two bases of weight and surface referred to total calories is essential; consequently we have prepared two tables for the boys and girls, respectively, giving the essential physical data for the individual children, and the values for both the measured heat and the heat predicted on the two bases of weight and surface. Since we are testing the accuracy of our general curves for the prediction of heat, the differences between the predicted and actual heat are given in these tables for both methods of prediction on the basis of the predicted less the actual, so that if the prediction is lower than the actual the difference will be indicated by a minus sign. The percentage differences are also given, since percentage values can alone be used for the comparison of children of various sizes. In addition to the data for the predicted heat on the bases of weight and surface, we have employed a third method of prediction for the boys, in which use was made of the multiple prediction formula of Harris and Benedict 1 as proposed by them for the prediction of the most probable total heat production of men. Exactly the same treatment with regard to the differences between predicted and actual, as well as the percentage differences, is accorded this method of prediction. Thus we have in the tables the actual total heat production per 24 hours of each child and then (by the three methods of prediction) the most probable heat production of a child of similar weight or (in the case of surface) of similar surface, taken from the corresponding curve. The differences of the actual from the smoothed curve values are then noted and recorded and the percentage values found. Since with the first two methods of prediction, i. e., those from curves, the smoothed curve is nothing more than an attempt to equalize all the inequalities in the values, theoretically there should be practically the same number of plus and minus differences in the whole series, and in any event the sum of the plus or minus differences should practically equal zero. PREDICTED HEAT FROM TOTAL CALORIES REFERRED TO WEIGHT (BOYS). The predicted heat for boys, based upon the curve for total calories referred to weight (see fig. 26) is given in table 32 and may first be considered. A superficial inspection of the table shows an approxi- mately equal distribution of plus and minus signs. While this w r ould 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919. In this monograph the differ- ences were computed by deducting the predicted values from the actual, instead of the actual values from the predicted as in the present report. PREDICTION OF THE BASAL METABOLISM OF YOUTH. 189 tend to indicate that the differences between individuals were in the main eliminated in drawing the curve, special consideration should be given to the numerical and percentage differences of the predicted less the actual heat. The numerical differences are expressed in column g of table 32 and vary from to as high as 177 calories. Since there was a very wide variation in the heat actually measured, i. e., from a minimum of 163 calories with two of the youngest boys to as high as 1,401 calories with the oldest boy, the percentage differences alone are of value for comparative purposes. The average percentage differ- ence for the whole series of boys, as given in table 33, indicates that the total heat can be predicted from the curve for calories referred to weight with a deviation of db7.4 per cent. Although an average error of this magnitude may be considered as large, it must be remem- bered that the data for growing children are extraordinarily few. Hence, if we are to take advantage of this series of observations, the hint given by the general trend of the line in figure 26 may on the whole be regarded, provided that too much emphasis is not placed upon this method of prediction. On inspection of the percentage deviations in table 32, we find that the differences of the predicted from the actual are rather considerable in some instances, there being deviations with 7 boys of 20 per cent or more. In all of these seven instances the weight of the boys is 10.7 kg. or below, suggesting that the error of prediction with small children is much greater than it is with the larger -children. Appar- ently some weight not far from 10 kg. represents an approximate dividing-line between a reasonably close prediction and a much more gross prediction. If we use 10 kg. as an arbitrary dividing-line, and calculate the average deviation of the predicted heat from the measured heat for children weighing 10 kg. and above, we find this average deviation to be 6.3 per cent. (See table 33.) For the boys under 10 kg., we find the average percentage deviation to be 8.7 per cent. Thus, what is obvious to the eye, both with regard to the dispersion of points and to the general magnitude of the percentage differences in column h of table 32, is amply verified by a calculation of the average percentage deviation of the predicted values from the actually observed values. That some other weight than 10 kg. may not be slightly better, mathematically, for a division-line is not disproved, but for practical purposes this division seems to be all that the calculations justify. To test the theory that some other weight might be better for division purposes, the percentage differences have been computed for boys weighing 15 kg. and above. On this basis we find that the prediction error is slightly less that is, it now becomes 5.8 instead of 6.3 per cent. Since, however, we purpose predicting the heat, not from a formula, but by drawing points from this curve, the lessening of the 190 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. !!? 5 M ^ Wooix/)] 3 .S3 3 - -3- IQ^J 03 III .OOCOO5Tt ee o eo S ' " - s g s g s s 7 asa 1 PREDICTION OF THE BASAL METABOLISM OP YOUTH. 193 i +++ i ++-H + i + i i ++ i i i i i i + i i + i ++ i i 7 i i i i i ++ i ++ I + I i i )i+i i + i++ii i i i i ++++ *52S2j^SSJ52512 CO '~'' :c>c>l/ 5 c C>"5'^ < ^O(NCOCOCOOT}O5t^iCl^CO s s $5 & s s s s g s s g.s o.l|il.|S||lss2 SIISISS ' +++ ' ++++ i + i i ++ i i + i 7+7++ i + i ++ i ii i + i i + i ++ i +++ i ++++ i + i i ++ i 7+ i 7+7++ i + i ++ i i i+i+i I++7++ I + + + I ++ I + I + I I + + + I + I I + I + + + + I ++ I I I -U I + I I ++ I + + i +++7++ i + i + i i +++ i + i i +7++++ i ++ i i i -H i + i i ++7++ !P SSSI^S^SS 1 I co . : IH co ^H ; co OO^rfroO^Ori-n^cQTJXTS .. caoooaodoooooooo^oo f "iifllillBlIIllItll 194 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. percentage deviation simply means that when the body-weight is above 15 kg., the error of prediction is slightly less than when it is between 15 and 10 kg., and considerably less than with boys under 10kg. TABLE 33. Comparison of the actual basal metabolism of boys with the metabolism predicted (a) from body-weight, 1 (b) from body-surf ace f and (c) from the adult masculine normal (multiple prediction) standard. 3 (Average values. 4 ) Deviation Basis Number Actual Predicted Predicted of Group. of of heat heat less predicted prediction. boys. per 24 hrs. per 24 hrs. actual. from actual. cats. cols. cats. p.ct. All boys Weight 128 652 657 43 7.4 Below 10 kg Do 60 364 368 30 8.7 Above 10 kg. Do 68 907 911 54 6.3 All boys Surface 128 652 654 46 7.5 Below 0.45 sq. m. Do. . . 52 344 342 25 7.7 Above 0.45 sq. m. Do. 76 863 867 59 7.3 Above 10 kg Multiple predic- tion standard 68 907 901 56 6.3 1 See figure 26, p. 140, and table 36, p. 206. * See figure 38, p. 161. Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 227. 4 Averages obtained from data in table 32, pp. 190 to 193. A comparison of this method of prediction may be made with the results of recent attempts by the Nutrition Laboratory to predict the metabolism of male adults. By means of a multiple-prediction formula derived from the analysis of the basal metabolism of 136 men, a group of 31 college students was tested on this basis and the results com- pared with the results of actual measurements. 1 The predictions were, on the average, within 5.3 per cent, a value percepibly better than the 6.3 per cent noted from the predictions with the boys. 2 This comparison is, however, not quite fair, since these college students were nearly all 20 to 26 years of age, and were unquestionably more homogeneous than a group of children ranging in age from a few months to 15 years. We may still feel, therefore, that on the whole the pre- diction of the metabolism of children from the curve in figure 26 is not greatly inferior to that for the best existing method of prediction of the values for adults. PREDICTED HEAT FROM TOTAL CALORIES REFERRED TO SURFACE (BOYS). In table 32 the values have also been incorporated for the heat pre- dicted from the curve for total calories referred to surface as accurately measured by the Du Bois method. (See fig. 38, page 161.) Giving 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 234. 2 The fact that Harris and Benedict computed these differences by deducting the predicted from the actual heat values does not affect the percentage in this case. PREDICTION OF THE BASAL METABOLISM OF YOUTH. 195 our attention first to the actual differences as shown in column j, we find that they range from to as high as 191 calories; but realizing the very great differences in the actual heat production, we again note that percentage differences alone can be considered for com- parative purposes. Exactly the same reasoning with regard to the effect of the curve in smoothing out plus and minus differences in the individual points obtains here, namely, that we would expect to find essentially the same number of plus as minus values, and a super- ficial inspection of the table bears this out. On the basis of weight, 7 instances were noted in which the per- centage difference was 20 per cent or more, while the heat based on the surface gives but 5 instances. All but one of these are with children whose body-weight is 10.7 kg. or below. This observation has two important features: First, it confirms the view previously expressed that the error of prediction is largest with the children of small weight. Second, it implies that, so far as extreme errors are concerned, the prediction by surface is somewhat better than the prediction by weight. It is, however, only after a consideration of all the data that one can draw final conclusions. With all the series of boys the average per- centage deviation of predicted from the actual is 7.5 per cent. (See table 33.) Thus it appears that while the number of cases of extreme error was less when the measured surface was used in pre- dicting the total calories, on the average this method of prediction gives slightly greater error than when the calories are referred to weight. Our general impression that the error of prediction is greater with children with smaller surface is sufficiently substantiated to justify our making a division of the values and considering boys with a surface area below 0.45 square meter in one group and those above this area in another. This division will be not unlike the comparison on the basis of weight, since a body-weight of 10 kg. corresponds approxi- mately to a surface area of 0.45 square meter. The results for this division are also given in table 33 and show that with the boys below 0.45 square meter of surface, the error of prediction is 7.7 per cent, while the boys above this area have a slightly better percentage, namely, 7.3 per cent. It thus appears that with the smaller boys the error of prediction is somewhat greater by weight than by surface, i. e., 8.7 per cent as compared with 7.7 per cent. This fact, taken alone, would imply a greater correlation between body-surface area and heat production than between body-weight and heat production. On the other hand, with the boys weighing 10 kg. and above and with a surface area of 0.45 square meter and above, we find that the error of prediction is 1 per cent greater when surface area is used that is, 7.3 per cent for prediction by surface as compared with 6.3 per cent by weight. With boys with a body-weight of 10 kg. or above, 196 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. therefore, more satisfactory results may be obtained in predicting the metabolism from a curve in which the calories are referred to weight than from a curve in which the calories are referred to surface. If we employ another line of division, as was done in making com- parisons on the weight basis, and use a surface area of 0.65 square meter as the dividing-line, we find that the average percentage error for boys with this area or above is somewhat lower, being 6.8 per cent, as compared with 7.3 per cent with the division at 0.45 square meter. It should be borne in mind that the curves used for predicting the metabolism were arbitrarily laid on these charts, and represent simply the resultant personal impressions of five skilled workers with metab- olism curves. They are not based upon a mathematical analysis of the points as a whole, and such an analysis has not been attempted. Still, we believe that the curves have a sufficient degree of accuracy to show that there is a definite superiority in the prediction from the curve for calories versus weight to that from the curve of calories versus surface. It thus becomes of physiological interest to recall that, as a result of metabolism measurements upon adults, it is commonly believed that the relationship between weight and metabolism is by no means so satisfactory as the relationship between surface area and metabolism. The comparison of the predicted and measured heat values for these boys appears to show that the prediction of the metab- olism from the weight is not only as good as that from surface, even though we are dealing here with actually measured surfaces, but that it is actually a somewhat better method of prediction for children over 10 kg. Hence these curves and predictions establish the fact that the correlation between weight and basal metabolism is, at least with boys over 10 kg. in weight, of a higher order than the cor- relation between measured body-surface and metabolism. In the analysis of the basal metabolism data for 136 men, the average deviation without regard to sign of the predicted from the observed values was in the case of men, when calculated from the body-weight by equations, 97.6 calories. 1 Since the average heat pro- duction of these men was 1,632 calories, 1 this gives an average per- centage deviation of essentially 6 per cent. With boys from 10 to 41 kg., we find from the curve of total calories referred to weight a deviation of 6.3 per cent. It is of significance that the marvelous changes in muscle, bone, and fat all changes due to the rapid period of growth through which our study of boys progresses should not alter appreciably the percentage error of prediction from the general curve so as to make it any greater than that with men. In other words, in spite of the great changes due to growth, the average error is hardly greater for boys than with adults, with whom growth altera- tions have practically ceased. 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, pp. 180 and 182. PREDICTION OF THE BASAL METABOLISM OF YOUTH. 197 COMPARISON OF THE PREDICTED METABOLISM OF BOYS AND MEN. This discussion thus far has been based upon methods of prediction and, indeed, comparison between youths and adults regarding but one physical factor at a tune, i. e., weight or surface. Since with adults it has been shown that weight, stature, and age have each an independent influence on basal metabolism, the comparison of youth with adults will not be complete unless the simultaneous use of these factors is made. In an earlier report, 1 the metabolism of groups of boys studied by Magnus-Levy and Falk and by Du Bois was compared to the com- puted metabolism of boys by means of the prediction formula derived from the analysis of 136 men, which took into consideration weight, stature, and age changes. It was found that the values for the actually measured metabolism of these boys was invariably very much higher than those computed by the multiple-prediction formula, which assumed that their metabolism was the same as that of adults of like sex, height, weight, and age. The Du Bois boys, in particular, showed an actual measured metabolism very much greater than that predicted from the formula for men proposed by Harris and Benedict. This was in line with Du Bois's interpretation of his own results on the basis of the heat production per square meter. In view of the calculated results obtained for boys by Harris and Benedict, it seems desirable to compute the metabolism of the boys in our study with the multiple- prediction formula for adults, for while the curves in figures 50 and 51 show a perceptibly higher heat production per kilogram of body-weight and per square meter of body-surface for boys than with men, it will be important to note if the resultant effect of the three varying factors, age, weight, and height, as included in the prediction by the planar equations, is at all in conformity with the trend in metabolism noted with adults. We have, therefore, computed the metabolism of all of these boys, using the multiple-prediction formula for men, 2 i. e., heat equals 66.4730 + 13.7516w + 5.0033s - 6.7550a. These val- ues have been recorded in table 32, together with the differences between the predicted and actual, both numerical and percentage. Although special emphasis has been laid in the foregoing discussion upon boys with body-weights of 10 kg. and over, those with smaller body-weights are likewise included. With the very young children, it will be seen that by the multiple-prediction method the error is practically +100 per cent. The metabolism as predicted is found to be too high in every instance until the boy No. 138, with an age of 10 months and 3| weeks, is studied, when a reversal in sign is found. Subsequently, the predicted metabolism seems to be for the most part not far from that actually measured, showing plus and minus values in about equal numbers. 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, pp. 237 and 238. 2 Ibid., p. 227. 198 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. A more careful analysis may be made by comparing the actual heat produced per 24 hours, on an average, for boys weighing 10 kg. and above, with the predicted metabolism calculated from the multiple- prediction formula. These average values are shown in table 33, in which we find that the average predicted metabolism is within 56 calories, or a percentage deviation of predicted from actual of 6.3 per cent. We thus have here a prediction of the metabolism of young boys, based upon the multiple-prediction standard of adults, which shows for the boys weighing 10 kg. or over a remarkable uniformity with that observed from the curve for calories referred to weight. While Harris and Benedict, in putting forth their multiple-prediction methods, specifically stated that its extension to youth under 16 years of age was problematical, it appears that this formula applies to males of all ages of approximately 1 year and over. Considering the rapid changes in weight and stature as compared with the change in age which takes place in this short period from 1 to 13 years, it is an astonishing agree- ment with the masculine adult prediction standards. Below the age of 1 year it is clear that the formula does not hold true, and the curves of either weight referred to metabolism or of surface referred to meta- bolism must be relied upon for prediction. The multiple-prediction equations which involve factors for age, weight, and height do not, however, better the situation, sh>ce taking all these factors into consideration gives us a prediction with an error identical with that when the values obtained from the curve in figure 26 for weight alone are considered, i. e., 6.3 per cent. It should be remembered in this connection that, during this period of growth, the factors of age, weight, and stature are intimately corre- lated, very much more so than is the case with adults. Adults weigh- ing 70 kg. are much more likely to vary in stature than boys of 30 kg. Similarly, adults weighing 70 kg. may vary in age from early youth to old age, while the variation in age of boys weighing 30 kg. will be very much less. It is thus probable that the body-weights of boys automatically include the variations in age and stature. For practical purposes, therefore, and until the metabolism of children is given biometric analysis, the prediction of the total basal metabolism of boys 10 kg. and above may be made with a reasonable degree of accuracy directly from the curve based upon body-weight. PREDICTED HEAT FROM TOTAL CALORIES REFERRED TO WEIGHT (GIRLS). While the charts that have been discussed thus far give little indi- cation of sexual differentiation in the accuracy or ease of prediction of the metabolism of girls as compared with boys, it has seemed desir- able to consider the boys and girls separately in regard to the prediction PREDICTION OF THE BASAL METABOLISM OF YOUTH. 199 TABLE 34. Comparison of the actual basal metabolism of girls with that predicted: (a) by the curve for "Total calories referred to weight"; 1 and (6) by the. curve for "Total calories referred to surface." 2 Sub- ject No. (a) Age. Body-weight (without ^ clothing). () +i ,J 3 w Body-surface (Du Bois linear ^ formula). (e) 1 3 ! Is 1 < Heat per 24 hrs. predicted from curve for "Total calories referred to weight." 1 Heat per 24 hrs. predicted from curve for "Total calories referred to surface." 2 (/) 1 *o I (0) |f .Sb 11 (h) II jl 0) i 1 (/) 8^ si 11 (As) lisr II S3 '*"" 49 116 26 111 113 110 2 123 109 12 131 35 113 48 120 127 35 122 139 145 131 48 113 151 160 134 140 122 131 123 139 135 165 113 152 123 139 160 35 163 144 166 11 days l mos 10 days 13 days 3 wks 13 days kilos. 2.68 2.99 3.56 3.57 3.65 3.71 3.73 3.85 3.86 4.20 4.34 4.42 4.55 4.81 4.90 5.03 5.07 5.15 5.19 5.25 5.27 5.54 5.54 5.64 5.90 5.99 6.02 6.03 6.08 6.09 6.11 6.17 6.24 6.49 6.52 6.75 7.00 7.05 7.17 7.63 7.91 7.92 cm. 48.5 50.0 53.0 53.0 51.0 53.0 53.5 51.5 53.0 55.5 54.0 57.5 56.0 58.0 57.5 58.5 58.5 63.0 62.0 58.0 61.0 63.0 60.0 62.5 64.0 60.0 62.5 62.0 63.0 65.0 63.0 63.0 66.5 65.5 65.0 67.0 65.5 64.5 63.0 62.0 68.5 sq. m. .20 .21 .24 .24 .25 .25 .25 .24 .25 .27 .25 .28 .28 .29 .30 .29 .30 .31 .31 .30 .31 .32 .33 .33 .34 .34 .32 .35 .33 .34 .35 .35 .37 .36 .36 .35 .38 .38 .38 .40 .42 .41 cafe. 139 163 185 200 173 182 152 235 200 199 235 198 207 211 274 255 223 257 315 317 311 388 289 355 417 331 334 329 353 312 325 333 338 351 357 410 406 492 329 375 353 505 cola. 124 149 189 190 196 200 201 210 210 233 242 247 256 273 279 287 290 295 297 301 303 320 320 327 344 349 351 352 354 355 356 359 363 377 379 391 405 408 414 440 455 456 cala. - 15 - 14 + 4 - 10 + 23 + 18 + 49 - 25 + 10 + 34 + 7 + 49 + 49 + 62 + 5 + 32 + 67 + 38 - 18 - 16 - 8 - 68 + 31 - 28 - 73 + 18 + 17 + 23 + 1 + 43 + 31 + 26 + 25 + 26 + 22 - 19 - 1 - 84 + 85 + 65 + 102 -49 -10.8 - 8.6 + 2.2 - 5.0 + 13.3 + 9.9 +32.2 -10.6 + 5.0 +17.1 + 3.0 +24.7 +23.7 +29.4 + 1.8 + 12.5 +30.0 + 14.8 - 5.7 - 5.0 - 2.6 -17.5 +10.7 - 7.9 -17.5 + 5.4 + 5.1 + 7.0 + 0.3 +13.8 + 9.5 + 7.8 + 7.4 + 7.4 + 6.2 - 4.6 - 0.2 -17.1 +25.8 + 17.3 +28.9 cats. 130 150 187 188 201 198 200 194 207 229 205 243 243 265 271 262 281 288 293 281 289 307 320 313 334 331 302 341 321 330 346 341 373 359 359 338 391 380 393 415 441 424 cols. - 9 - 13 + 2 - 12 + 28 + 16 + 48 - 41 + 7 + 30 - 30 + 45 + 36 + 54 - 3 + 7 + 58 + 31 - 22 - 36 - 22 - 81 + 31 - 42 - 83 - 32 + 12 - 32 + 18 + 21 + 8 + 35 + 8 + 2 - 72 - 15 -112 + 64 + 40 + 88 - 81 - 6.5 - 8.0 + 1.1 - 6.0 + 16.2 + 8.8 +31.6 -17.4 + 3.5 +15.1 -12.8 +22.7 + 17.4 +25.6 - 1.1 + 2.7 +26.0 + 12.1 - 7.0 -11.4 - 7.1 -20.9 +10.7 -11.8 -19.9 0.0 - 9.6 + 3.6 - 9.1 + 5.8 + 6.5 + 2.4 + 10.4 + 2.3 + 0.6 -17.6 - 3.7 -22.8 + 19.5 + 10.7 +24.9 -16.0 10 days 2 mos. 1 wk. 12 days 9 days 3 mos. 8 days 1 mo. 3^ wks. 1 mo. 1 wk 2 mos 2 mos. 3| wks 1 mo. 1 wk 2 mos. 1 wk 4^ mos 5 mos 4 mos. 1? wks 2 mos. 3 wks 4 mos. . 6 mos. 7 mos. 4 mos. 4 mos. 3 wks 3 mos. 3| wks 7 mos. 2j wks 6 mos. 1 wk 6 mos. 1 wk 4 mos 8 mos. 3 wks 5 mos. 3 wks 6 mos 8 mos. 3 wks 7 mos 10 mos 4 mos 8 mos. 1 wk 5 mos 9 mos. 1 wk 1 See figure 27, p. 142, and table 36, p. 206. * See figure 39, p. 164. 200 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. TABLE 34. Comparison of the actual basal metabolism of girls with that predicted: (a) by the curve for "Total calories referred to weight"; 1 and (6) by the curve for "Total calories referred to surface" 2 Continued. Sub- ject No. (a) Age. Body-weight (without -5- clothing) . to 4 3 Body-surface (Du Bois linear ^~- formula). Actual heat per 24 hrs. Heat per 24 hrs. predicted from curve for "Total calories referred to weight." 1 Heat per 24 hrs. predicted from curve for "Total calories referred to surface." 2 (/) (f) it all 83 & ( i"5T Q> If |x g &o m O PH g CO 1 (/) - 5 lo- (*) 1! Is II 162 127 160 171 139 146 167 172 146 127 145 166 173 171 139 172 122 145 144 173 171 13 9 174 172 16 6 145 16 6 171 13 9 178 166 145 13 9 13 9 166 171 179 139 180 145 190 183 145 184 181 kilos. 8.00 8.11 8.12 8.18 8.29 8.30 8.52 8.80 9.04 cm. 69.5 67.0 68.5 73.5 70.0 68.0 69.0 74.5 71.0 73.0 70.0 74.5 72.0 76.5 76.0 77.0 78.5 72.5 67.5 78.0 85.5 82.0 79.0 80.0 80.0 76.0 88.0 89.5 88.0 92.0 88.5 80.0 92.5 96.0 92.5 100.0 98.5 99.0 93.5 86.5 103.5 97.5 94.0 103.0 98.5 sq. TO. 0.41 .42 .41 .42 .42 .43 .46 .46 .43 .48 .46 .49 .46 .47 .48 .51 .51 .52 .52 .52 .49 .54 .54 .55 .53 .56 .58 .58 .59 .61 .60 .62 .65 .64 .60 .62 .65 .67 .69 .67 .69 .65 .67 .72 .70 cals. 413 468 522 502 419 360 522 568 419 549 474 559 606 528 686 597 508 445 614 643 531 604 712 597 518 655 735 590 543 692 600 607 655 686 657 560 624 640 633 637 673 590 715 771 cals. 460 464 465 467 472 472 481 492 502 502 508 508 509 517 527 534 544 548 564 564 564 572 580 583 583 604 610 616 619 622 645 650 655 665 665 670 678 683 690 692 694 704 706 715 720 cals. + 47 - 4 - 57 - 35 + 53 + 112 - 41 - 76 + 83 - 47 + 34 -51 - 91 - 89 - 1 -152 - 53 + 40 + 119 - 50 - 79 + 41 - 24 -129 - 14 + 86 - 45 -119 + 29 + 79 - 47 + 50 + 48 + 10 - 21 + 13 +118 + 59 + 50 + 59 + 57 + 31 +116 o - 51 + 11.4 - 0.9 -10.9 - 7.0 + 12.6 +31.1 - 7.9 -13.4 + 19.8 - 8.6 + 7.2 - 9.1 -15.2 -14.7 - 0.2 -22.2 - 8.9 + 7.9 +26.7 - 8.1 -12.3 + 7.7 - 4.0 -18.1 - 2.3 +16.6 - 6.9 -16.2 + 4.9 +14.5 - 6.8 + 8.3 + 7.9 + 1.5 - 3.1 + 2.0 +21.1 + 9.5 + 7.8 + 9.3 + 8.9 + 4.6 +19.7 + 0.0 -6.6 cals. 429 433 423 432 438 444 482 479 445 509 489 513 485 490 502 538 544 551 550 552 523 568 573 581 563 592 611 609 620 637 629 649 673 664 633 652 679 689 712 691 712 679 689 733 721 cals. + 16 - 35 - 99 - 70 + 19 + 84 - 40 - 89 + 26 - 40 + 15 - 46 -115 -116 - 26 -148 - 53 + 43 +105 - 62 -120 + 37 - 31 -131 - 34 + 74 - 44 -126 + 30 + 94 - 63 + 49 + 66 + 9 - 53 - 5 +119 + 65 + 72 + 58 + 75 + 6 + 99 + 18 - 50 + 3.9 - 7.5 -19.0 -13.9 + 4.5 +23.3 - 7.7 -15.7 + 6.2 - 7.3 + 3.2 - 8.2 -19.2 -19.1 - 4.9 -21.6 - 8.9 + 8.5 +23.6 -10.1 -18.7 + 7.0 - 5.1 -18.4 - 5.7 +14.3 - 6.7 -17.1 + 5.1 +17.3 - 9.1 + 8.2 + 10.9 + 1-4 - 7.7 - 0.8 +21.3 + 10.4 +11.3 + 9.2 + 11.8 + 0.9 +16.8 + 2.5 - 6.5 9^ mos. 1 yr 3^ wks. 9 mos. 1 wk 9 mos. 1 wk. ........ ll mos 9.06 9.19 9.21 9.22 9.43 9.67 9.84 10.1 10.2 10.6 10.6 10.6 10.8 11.0 11.1 11.1 11.8 12.0 12.2 12.3 12.4 13.2 13.4 13.6 14.0 14.0 14.2 14.5 14.7 15.0 15.1 15.2 15.7 15.8 16.2 16.4 10 mos. 1 wk ll mos 1 yr. 2 mos. If wks. . . 1 yr. 2 mos. 3 wks. . . . 1 yr. 1 mo. 1 wks. . . . 9 mos yr. 5 mos 1 yr. 9| mos yr. 8| mos 2 yrs. 1 mo 1 yr. 5 mos yr. 8 mos. 3 J wks. . . 1 yr. 2 mos 2 yrs. 4 mos 2 yrs. 3 mos. 1 wk. . . . 2 yrs 2^ mos 2 yrs. 9^ mos 1 yr. 5 mos 2 yra. 6 mos. 3 wks. . . 3 yrs. 2 mos. 3 wks. . . 3 yrs. 4j mos 3 yrs. 3 mos. 1 wk. . . . 3 yrs. 8 mos 3 yrs. 9 mos. 3 wks. . . 3 yrs. 10 mos. 3 wks. . 1 yr. 9 mos. 3 wks. . . . 5 yrs. 3 j mos 4 yrs. 3 mos. 3 wks. . . 2 yrs. 4 mos 4 yrs. 4 mos. 1 wk. . . . 3 yrs. 11 mos i See figure 27, p. 142, and table 36, p. 206. * See figure 39, p. 164. PEEDICTION OF THE BASAL METABOLISM OF YOUTH. 201 TABLE 34. Comparison of the actual basal metabolism of girls with that predicted- (a) by the curve for "Total calories referred to weight"? and (6) by the curve for "Total calories referred to surface z Continued. (a) (fe) to (<0 CO Heat per 24 hrs. Heat per 24 hrs. predicted from predicted from | curve for "Total curve for "Total 4* 2* calories referred calories referred J a to weight." 1 to surface." 2 Sub- vB-'Sf m s (/) (0) (h) (0 0) (*) ject Age. ^ 5 Q ~ a , ^ No. 11 .SP ^^ 2 49 ffi^ S -S. 03 "^-- ! n I 1 1 1 Ji ""* JfX 1 IL if |x & 1 1 11 Is 1 n |f kilos. cm. sg. 0t. cals. cals. cals. cola. cola. 195 i 6 yrs. 3 wks 16.4 99.5 0.67 792 720 - 72 - 9.1 689 -103 -13.0 171 4 yrs. 2 mos. 1 wk. . . . 16.5 104.5 .68 718 723 + 5 + 0.7 703 - 15 - 2.1 145 2 yrs. 10 mos 16.6 96.5 .69 618 725 +107| +17.3! 710 + 92: +14.9 188 i 5 yrs. 1? mos 16.6 103.5 .73 782 725 - 57| - 7.3 742 - 40] - 5.1 145 ' 3 yrs. 2 mos. . . . 17.5 98.51 -70 657 748:+ 91+13.9 716 + 59 - 1 - on 191 ; 5 yrs. 5^ mos. 18.7 107.5 .7fi 790 774 16 2.0' 758i 32 n^ o.v 4 1 206 7 yrs. 4 mos. 19.2 113.0 -79 752! 7Rft' 4_ aal -i- A A\ "yon _i_ ue ~ rt.J + e i 196 6 yrs. 5| mos 19.7 118.0 .82 747 798 + 51 + 6.8 813 + 66 O. J + 8.8 206 i 8 yrs. 2 mos 20.8 116.0 .84 863 QOc! QQ A. a. san 33 3.8 220 9 yrs. mo 22.5 122.0 .94 977 868 ! 109 11 9J 01 n _ R7 6.9 189 5 yrs. 3 mos. 1 wk. . . . 22.7 116.0 .87 829 873; + 44 + 5.3: 858 + 29 + 3.5 203 7 yrs. 1 mo. 2 wks. . . . 23.1 119.0 .88 849 882 + 33 + 3.9! 865 + 16! + 1.9 219 8 yrs. 11 mos. 1 wk.. . 23.7 125.0 .92 880; 894 + 14 + 1.6 898 + 18 + 2.0 210 8 yrs. 2 wks 23.9 122.5 .93 894 898 + 4 + 0.4 904 + 10j+ 1.1 225 9 yrs. 5 mos. 1 wk. . . . 24.0 120.51 .91 924 900 - 24 - 2.6 885 - 39 - 4.2 227 9 vrs. 9 mos. . . . 24.8 1 25.5 .94 919 924 + 5 4- n s' onl_ 11 _ 19 214 8 yrs. 2 mos. 26.0 126.0 .97 930! 950 + 20 + 2.2J 936i + 6' + 0.6 210 8 yrs. 5 mos. 3 wks. . . 26.0 123.5 .99 1,002 950 - 52 - 5^2 950 - 52 - 5^2 221 9 yrs. 3| wks. 26.1 122.0 1.02 902 953 + 51 + 5.7 9791+ 77 + 8.5 198 6 yrs. 9 mos 26.4 124.5 .99 918 960 + 42 4- 4. ! QM! -i- a 4- a a 239 11 yrs. 27.4 I33.fi 1.08 984 985; + 1 + 0.1 1 flS-ll 4- ftl 4. ?; 9 230 10 yrs. 3 mos 27.9 133.0 1.06 999 998J - 1 1*012 4- 13 4- 1.3 238 10 yrs. 9 mos. 3| wks. . 28.0 135.5 1.05 944' 1,000 + 56 + 5.9| lio02] + 58 + 6.1 234 10 yrs. 5 mos. 3| wks. . 28.2 133.0. 1.03 923 1,004 + 81 + 8.8 990; + 67 + 7.3 248 11 yrs. 10 mos. 3 wks. . 28.8 129.0, 1.04 1,062 1,016 - 46 - 4.3 992 - 70 - 6.6 233 10 yrs. 5 mos. 2j wks. . 29.8 i 131.0 1.06 896 1,040 +144 +16.1 1.016J +120 +13.4 251 12 yrs. 2 mos 30.9 138.5 1.10 1,012 1,068 + 56 + 5.5 1,050 + 38 + 3.8 1 See figure 27, p. 142, and table 36, p. 206. 2 See figure 39, p. 164. value of the curves for total calories referred to weight and total calories referred to measured surface. The usual scatter of points about these curves is such as to indicate a reasonable regularity and suggest the possibilities of their use for prediction purposes. As we have seen with boys, exact deductions from the curves themselves with regard to the accuracy of prediction are difficult and only a mathematical consideration of the individual differences between the predicted and actual heat of individual subjects can give us a clear conception as to their accuracy. Consequently, a table has been pre- pared for girls (table 34) in which are given the values for heat as 202 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. actually measured and as predicted from the curve in figure 27 (total calories referred to weight) and likewise the heat predicted from the curve in figure 39 (total calories referred to surface). For each of these series of values, the differences between the predicted and actual and the percentage differences are given. Considering first the prediction of metabolism from the curve of total calories referred. to weight, we find the plus and minus signs are fairly equal in number, as would be expected from the way the curve was laid on. The magnitude of the differences ranges from to 152 calories, but since the children vary in weight from 2.7 to 30.9 kg., percentage differences alone have value for comparison purposes. In 11 instances the error of prediction is 20 per cent or more; but 10 out of these 11 girls weigh 10.6 kg. or less, thereby confirming the ob- servation made with boys that the error is smaller with the heavier children. TABLE 35. Comparison of the actual basal metabolism of girls with the metabolism predicted: (a) from body-weight, 1 and (b) from body-surface. 2 (Average values. 3 ) Group. Basis of prediction. No. of girls. Actual heat per 24 hrs. Predicted heat per 24 hrs. Predicted less actual. Deviation of predicted from actual. All girls Weight 114 cals. 545 cals. 554 cals. 46 p. ct. 97 Below 10 kg Above 10 kg All girls Do. Do. Surface 58 56 114 354 743 545 357 757 544 41 52 48 11.8 7.5 98 Below 0.45 sq. m. . . . Above 0.45 sq. m. . . . Do. Do. 49 65 315 718 312 718 36 57 11.6 8.5 1 See figure 27, p. 142, and table 36, p. 206 8 Averages obtained from data in table 34, pp. 199 to 201. See figure 39 p. 164. The summaries showing the average differences between predicted and actual and the percentage differences for girls are given in table 35. The average error of prediction for all the girls on this basis is there given as 9.7 per cent. Since we note that the errors are greater with the earlier weights, the girls may be separated into those weighing above or below 10 kg. We find that those below 10 kg. have a devia- tion of 11.8 per cent and above 10 kg. 7.5 per cent, thus furnishing a mathematical demonstration of our conclusion from the inspection of the figures in column h of table 34 that the accuracy of the prediction increases with the increase in weight. If a division is made at 15 kg., as was done for the boys, the percentage deviation for the girls with this or a higher body-weight is 6.1 per cent. A special reason for separating the girls at 10 kg. is based upon the fact that we found that there was no sexual differentiation in the metabolism of boys and girls up to the weight of 10 kg., so on this PREDICTION OF THE BASAL METABOLISM OF YOUTH. 203 basis alone this point seems to be a justifiable place for separation. From a comparison of the figures for the prediction of metabolism from weight for boys in table 33, it will be seen that in every instance the prediction for boys is materially better than that for girls. Thus, with body-weights of less than 10 kg., the error for the boys in the prediction of the metabolism was 8.7 per cent, and for the girls 11.8 per cent. Above 10 kg. the metabolism of the boys was pre- dicted with an error of 6.3 per cent and the girls 7.5 per cent. For 15 kg. or above, it was 5.8 per cent for the boys and 6.1 per cent for the girls. It should be emphasized that these differences are not due to an error in the curve (which is a smoothed curve repre- senting all our measurements), but to the more pronounced physio- logical differentiation with children of the lower weights. This is particularly the case with children in the first months of life, a con- siderable number of whom appear on our charts. PREDICTED HEAT FROM TOTAL CALORIES REFERRED TO SURFACE (GIRLS). While our analysis of the prediction of the metabolism from the surface for boys gives little reason to expect a better method of pre- diction for girls with this criterion, it is necessary to analyze carefully the data in table 34 for the heat predicted on this basis. The actual differences range from to 148 calories, an extremely wide range; but again reference must be made primarily to the per- centage differences, indicated in column k, for purposes of comparison. With 11 subjects the prediction has an error of 20 per cent or more, but 10 of these are girls with a body-weight of 10.6 kg. or below, corresponding to approximately 0.50 square meter of body-surface. Thus, on the basis of surface as well as weight, the larger children apparently exhibit less errors of prediction by this method of analysis. When the actual averages are compared (see table 35), we find that the average deviation of the predicted from the actual measurement is 9.8 per cent, this being the largest general deviation by our methods of prediction, considering the boys as a whole and the girls as a whole. This error in prediction is also slightly higher than that found on the basis of weight and confirms the deduction made from the values for boys, i. e., that the prediction made from surface is on the whole slightly inferior to that from weight. Realizing that the smaller children, i. e., those with the smaller body- surface, show the greater proportion of large deviations, it remains to be seen whether or not any particular area will give the better predic- tion. As with boys, a body-surface area of 0.45 square meter, which corresponds to a body-weight of approximately 10 kg., has been selected as the dividing-line. According to table 35, the percentage deviation for the girls with a body-surface area below 0.45 square 204 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. meter is 11. 6 per cent, while for those above it is 8.5 per cent, showing a measurably greater error with the smaller children. If, again, we use for the arbitrary dividing-line, the surface area of 0.65 square meter, the percentage deviation for the girls of this surface area or over is 6.8 per cent, as compared with 8.5 per cent on the basis of 0.45 square meter. It is thus seen from the data in table 35 that the error of prediction with girls by surface is slightly better with the smaller children, i. e., those below 10 kg. and below 0.45 square meter, than is the prediction from weight, but with the larger girls the prediction by weight is considerably better than by surface, i. e., 7.5 per cent against 8.5 per cent. It will be noted that in table 34 we have omitted figures regard- ing the prediction of the metabolism of girls by the adult multiple- prediction formula of Harris and Benedict. The formula for men, which is used to predict the metabolism for boys, is given on page 197 of this report. The formula for women differs considerably from that for men in that the first term is large, being 655. 0955. l The other terms for weight and stature are both positive, that for age alone being negative. Consequently, since the ages are all small with girls, with a maximum of 15 years, the only negative term would at most corre- spond to not far from 70 calories. As this formula deals only with those organisms having a heat prediction of not less than 700 or 800 calories, it is obviously impossible to apply the formula to the heat production of young girls. While we might have used the formula for men for this computation and thus studied the relative differences in metabolism between boys and girls, it was believed to be unnecessary, as our several charts point out very clearly the sexual differentiation. The application of the formula for women to girls is fraught with considerable danger. Throughout the biometrical treatment of the basal metabolism of women and men by Harris and Benedict, it was brought out repeatedly that no method at present available makes it possible to predict the metabolism of women with any approach to the accuracy of the prediction for men. Although in the biometric study referred to approximately 100 women were included, they were by no means so harmonious in physical type as were the men, the correlation of stature and weight being on the whole about half that for men. Unfortunately, the data with which this correlation can be compared are very few, comprising generally the Cambridge students, both male and female, studied by Pearson. 2 With these students a much more homogeneous group was found, since the correlation between weight and height with the female students was considerably better than that found for the male students. With the men included 1 Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 227. 2 Pearson, Proc. Roy. Soc. Lorid.. 1899, 66, p. 26. PREDICTION OF THE BASAL METABOLISM OF YOUTH. 205 in the Harris and Benedict study, the partial correlation between weight and stature corrected for constant age was for 136 men 0.5772, while for 103 women it was but 0.2995. 1 This alone shows that with the women the correlation between height and weight was considerably lower than with the men, thus indicating greater irregularity in physical configuration and departure from an average. PRACTICAL VALUE OF THE PREDICTION OF BASAL METABOLISM. Prediction factors are of importance in that it is highly desirable, particularly in clinical work, to have some conception of the general trend of the metabolism of children of a definite age or weight. We believe that we have sufficiently emphasized in all of our charts that the scatter or dispersal of the individual points is so great as to necessi- tate great caution in considering that any particular normal child may indicate a fixed normal metabolism, to which all other normal children of the same age, sex, height, and weight should conform. This is far from the case. But if due cognizance is taken of the probable deviations from the general trend, the curves for total calories referred to body-weight and total calories referred to body-surface may be considered as distinctly helpful in indicating whether or not there is great diversity from the general trend with children having any particular configuration or any particular illness, or subsisting upon any dietetic regime or nutri- tional plane, or with any physiological change affecting the group. As will be seen, the emphasis in this discussion is laid upon the group. It is clear from the most cursory examination of our several charts that the individual child may vary greatly from the average, and hence it is only with extreme caution that deviation may be interpreted as being of pathological or abnormal nature with any individual child. If special conditions of diet, life, or pathological development are to be studied, it must be clearly established that the special conditions result in the deviation from the general trend in a considerable number of instances or a group. With these cautions in mind, we may suggest that the curves given in figures 26 and 27 may be used directly for predicting the most probable basal metab- olism of a quiet, resting boy or girl. To assist clinicians in rapid estimations of these predictions, we have prepared a table (table 36) giving the most probable heat production for each half kilogram of body-weight for both boys and girls. If the weight of the boy or girl is known, the physician may thus read directly from table 36 the predicted basal metabolism for a child of this weight and compare it with the heat actually measured to note if the measured values are aberrant in any way. Owing to the effort and time required i Harris and Benedict, Carnegie Inst. Wash. Pub. No. 279, 1919, p. 59, table 12. 206 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. for making the calculations of the body-surface by the Du Bois method, it seems unwise to tabulate the changes in metabolism for each tenth of a square meter. These can always be obtained by direct inspection of the curves in figures 38 and 39 in case surface-area measurements are made. TABLE 36. Basal heat production of boys and girls per 24 hours predicted from body-weight. Body- weight (with- Predicted heat. Body- weight (with- Predicted heat. Body- weight (with- Predicted heat. Body- weight (with- Predicted heat. out cloth- ing). Boys. Girls. out cloth- ing). Boys. Girls. out cloth- ing). Boys. Girls. out cloth- ing). Boys. Girls. kilos. cals. cals. kilos. cola. cals. kilos. cals. cals. kilos. cals. cals. 2.5 115 110 11.5 607 595 20.5 873 818 29.5 1,103 1,032 3.0 150 150 12.0 625 610 21.0 885 830 30.0 1,115 1,045 3.5 180 185 12.5 643 625 21.5 898 842 30.5 1,127 1,058 4.0 210 220 13.0 660 640 22.0 910 855 31.0 1,140 1,070 4.5 240 253 13.5 678 652 22.5 925 867 31.5 ,150 1,080 5.0 270 285 14.0 695 665 23.0 940 880 32.0 ,160 1,090 5.5 300 318 14.5 710 678 23.5 953 890 32.5 ,170 6.0 330 350 15.0 725 690 24.0 965 900 33.0 ,180 6.5 360 377 15.5 740 700 24.5 978 915 33.5 ,190 7.0 390 405 16.0 755 710 25.0 990 930 34.0 ,200 7.5 418 432 16.5 768 722 25.5 1,005 940 34.5 ,210 8.0 445 460 17.0 780 735 26.0 1,020 950 35.0 ,220 8.5 470 480 17.5 793 747 26.5 1,033 962 35.5 1,230 9.0 495 500 18.0 805 760 27.0 1,045 975 36.0 1,240 9.5 520 520 18.5 818 770 27.5 1,058 987 36.5 1,248 10.0 545 540 19.0 830 780 28.0 1,070 1,000 37.0 1,255 10.5 568 560 19.5 845 793 28.5 1,080 1,010 37.5 1,265 11.0 590 580 20.0 860 805 29.0 1,090 1,020 38.0 1,275 24-HOUR ENERGY REQUIREMENTS OF GROWING CHILDREN. With all of these children, the only activity possible was the move- ment of the arms and legs while the child lay on a bed or cot inside the respiration chamber. There could be no running about and no external muscular work performed. An earlier discussion 1 of the possibilities of an increased metabolism due to the muscular activity of an infant in the lying position has shown that on the average a maximum increase in metabolism of 65 per cent (in exceptional cases of over 200 per cent) may be obtained as a result of vigorous kicking and convulsive crying on the part of a new-born infant. While our studies were primarily for an investigation of the basal metabolism, and consequently the greatest degree of repose was sought by every possible means, we incidentally secured a considerable amount of data with regard to muscular activity. Certain of the children were restless, so much so that the observations had to be stopped. With very young infants there was at tunes crying, but with all these condi- tions the children were lying in bed. Aside from the gross records on 1 Benedict and Talbot, Carnegie Inst. Wash. Pub. No. 233, 1915, p. 111. 24-HOUR ENERGY REQUIREMENTS. 207 the kymograph, to which we have arbitrarily assigned values of I to VI, the statistics of these periods of activity can have little quantitative value to other observers, and accordingly they are not published here. All of the results for these periods have been gathered together, how- ever, and certain deductions seem justified by an inspection of the data. In the first place, the very great increases noted with new-born children may occasionally be obtained with children up to the age of 9 months, but the average of 65 per cent found with new-born babies does not continue after the first few weeks of life. Subsequently, until the age of 6 months, the average increase due to activity may be as high as 40 per cent. As the age of the child increases and voluntary muscular control becomes more possible, the increments noted by us decrease in value, until after 7 years of age an increase due to activity of over 30 per cent of the basal value is rarely found. This obtains for girls as well as for boys. Two important points stand out as a result of this inspection of data : First, that the younger the infant the greater is the percentage increase in metabolism during lying on account of crying and active movement of the body, arms, and legs; and second, that with the older children voluntary control has so increased as greatly to reduce the maximum activity. These values have special significance in that they throw light upon the practical use of the basal value in the computation of the total 24-hour requirement of the child. From the fact that in our experi- ments children after 7 years of age rarely had a heat production more than 30 per cent above the basal in periods of restlessness or activity, we feel justified in concluding that the basal value is almost synony- mous with the heat production of children over 7 years of age while in bed, save for the stimulating effect of food. Since children are for the most part anywhere from 8 to 12 hours of the day in bed, this forms an important quota of the total 24-hour requirement. Further than this with the older children we may not go. The great differences in muscular activity of the boy or girl at play, differences which are fully attested by recent studies of the food consumption in boys' schools, 1 shows that the actual caloric output during the waking period and when the children are actively at play is very much greater than the basal output, i. e., that obtained when the child is in bed resting quietly. With babies, whose time for the most part is spent in bed, some- what more definite information is at hand, as two 24-hour experiments have been made which have already been reported. 2 With one infant, E. L., 2 months and 3 weeks old, and having a body-weight at the tune of observation of 5.03 kg., the basal heat-output per 24 1 Gephart, Boston Med. and Surg. Journ., 1917, 176, p. 17. 2 Talbot, Am. Journ. Diseases of Children, 1917, 14, p. 25. 208 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. hours was found by computation of results in minimum periods to be 285 calories. The child was 22 hours and 31 minutes inside the respira- tion chamber. The total heat-output, calculated on the basis of 24 hours, was 372 calories, or an increase of 87 calories (approximately 30 per cent) over basal. The second child, E. S., weighed at the time of the observation 5.76 kg. and was 6 months and 1 week old. She was for 23 hours and 10 minutes inside the respiration chamber, giving a total heat-output per 24 hours of 404 calories. Computation of the basal metabolism by the selection of quiet periods gave 338 calories, the increase over basal being 66 calories or approximately 20 per cent. From these two experiments, therefore, it would appear that al- though it is possible for young children to have percentage increases over the basal for relatively short periods of 60 or 70 per cent, or even more, the total 24-hour heat production on the average may be taken as not far from 25 per cent above the basal. These two isolated points at the ages of 2 months and 3 weeks and of 6 months and 1 week can give only a general hint as to the increase in the 24-hour demand over basal. When children are able to leave the crib or cot and actively exercise by creeping, crawling, and walking around, especially older children when running and playing, the increase above basal becomes very great, and our own observations can obviously contribute in no wise to an estimate of the 24-hour demand on this basis. The proper estimation of the food needs of growing children can probably never be completely and satisfactorily made from gaseous metabolism experiments. With children, with whom the problem of growth plays so active a r61e, one must supply energy not only for maintenance and for juvenile physical activity, but likewise energy for growth. Even if respiration-chamber experiments were made in which an accurate measurement was made of the entire heat-output for 24 hours of a number of individual children engaging in the usual 24-hour day activity, there still would be the growth factor to be allowed for. In any complex of this nature, one is not justified in saying that, since one of the three factors is difficult of estimation, the others can not be satisfactorily determined. If possible, it is important to determine the basal needs and also the need for extra physical activity. So far as the basal value is concerned, our experiments, we believe, are reasonably conclusive. So far as the extra needs are concerned, our data supply little if anything of value. The maximum activities noted with children lying in the respiration chamber can, with children over 1 or 2 years of age, have little meaning. Children below this age are, for the most part, lying either in the bed or crib, and conse- quently these periods of maximum activity can well correspond to those occurring in the life of an ordinary young infant. 24-HOUR ENERGY REQUIREMENTS. 209 Few of our experiments have included the entire 24 hours, and we cite only two. Accordingly it is necessary for us to consider observa- tions made in other laboratories. In practically all the experiments in which 24-hour periods were possible, few if any were available for comparison with our basal measurements, but they are of prime import- ance for estimating the probable total 24-hour heat-output for very young children. We have collected from the literature a number of respiration-chamber experiments with children ranging in age from 1| hours to 9 months. These were practically all carried out with the Pettenkofer-Voit type of apparatus and for the most part were made in Berlin. The data for these experiments are given in table 37, in which is recorded the heat-output of these children computed on the basis of per kilogram per 24 hours. This corresponds very closely, in all probability, to the actual heat production, but does not corre- spond to the energy needs, since the amount required for growth is not included. Finally, we have recorded the values for the basal heat per kilogram per 24 hours, as predicted from the body-weight curve TABLE 37. Heat production per kilogram per 24 hours of normal boys studied by other investigators. Actual Heat heat per kilo. Duration (com- per Investigator. of obser- Age of child. Body- weight. puted) per 24 hrs. predicted Condition as to activity. vation. kilo. from per body- 24 hrs. weight. 1 kilos. cafe. cols. f 12 hrs. . ca. 1 J hrs. . . 3.1 55 50 Absolutely Birk and Edelstein* I 18.6 hrs. ca. 1 day . . . 3.0 55 50 quiet.' Not stated. [22.2 hrs. ca. 2 days. . 3.0? 47 50 Do. Rubner and Heub- ner 4 6 days. . 9 wks. 5.2 67 54 Generally quiet. Niemann* 6 days. . 3| mos 5.1 93 54 Generally quiet; cried some. Do. 6 days. . 5 mos. 5.9 85 55 Quiet. Do. 6 days. . 8 mos. 5.6 89 55 Do. Do. 17 days. . 9 mos. 6.0 91 55 Do. Hellesen* 3 days. . 5$ mos. . . 6.7 68 56 Do. Do 3 days. . 6 mos. . . . 6.6 74 55 Do. Rubner and Heub- ner 8 6 days. . 7$ mos. . . . 7.6 78 58 Generally very quiet. Do.. . . 3 days. . 6j mos. . . . 9.8 68 55 Restless. Do 1 day... 5| mos. . . . 9.6 68 55 Do. i See table 36, p. 206. * Birk and Edelstein, Monatsschr. f. Kinderheilk., 1910-11, 9, p. 505. 4 Rubner and Heubner, Zeitschr. f. Biol., 1898, 36, p. 1. "Niemann, Jahrb. f. Kinderheilk., 1911, 74, p. 22. 6 Child had had light case of grippe and indigestion previously. ' Hellesen, Nord. Med. Arkiv, 1915, 48, Nos. 14 and 18. Rubner and Heubner, Zeitschr. f. Biol., 1899, 38, p. 315. ' Girl. 10 Rubner and Heubner, Zeitschr. f. exp. Pathol. u. Therapie, 1904-05, 1, p. 1 'Without food. 210 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. in figure 26 (see page 140) for weights similar to those noted with the children studied by these investigators. In practically every instance the measured heat, as computed on the 24-hour basis, is greater than the predicted heat. The one exception is the value found on the third day of the observation of Birk and Edelstein, when the measured heat was 47 calories per kilogram as compared with our prediction of 50 calories. In all the other instances, values very much higher than the predicted heat are found. This would be expected, however, as the predicted heat represents only the basal output, while the measured heat includes for the whole 24 hours not only the basal output, but the heat due to activity when it exists, and particularly that due to the stimulus from food. The predicted heat for the children with body-weights of 3 to 10 kg. ranges from 50 to 58 calories per kilogram of body-weight. Using 55 calories as an approximate average, it can be seen that the measured heat is very considerably higher in most cases, although the values of Birk and Edelstein are only 5 per cent higher, on the average, than the basal. With Niemann's infants, the measured heat is nearly 70 per cent higher in some cases. Even when the protocols indicate that the child is quiet, we find great increases above the basal. It is clear, therefore, that the ordinary 24-hour life of a young child results in a heat production not far from 30 to 40 per cent above basal, a value somewhat in excess of that noted in our two experiments previously cited. 1 Beyond the weight of 10 kg., or the age of 9 or 10 months, very little has been done aside from the classic research of Sonden and Tigerstedt 2 and that of Rubner. 3 Professor Carl Tigerstedt, 4 in his monograph on the food intake of man, has collected the observations of Hellstrom, Rubner, von Willebrand, and Sonde"n and R. Tiger- stedt, with boys ranging from 9 to 14 years of age. Although he specifically states that the activity was to a high degree reduced, no claim is made for basal values and the calculated energy output is consequently upon a "moderately quiet" basis. The data thus col- lected are reproduced here in table 38. From these data, Tigerstedt concludes that the caloric needs, computed on the basis of per kilo- gram of body-weight, decrease with increasing age. He also finds the same result when the computations are made on the basis of body- surface area, although this is not shown with as great regularity as in the case of weight. As heretofore stated, these observations were made with children on a moderately quiet basis that is, the children 1 Talbot, Am. Journ. Diseases of Children, 1917, 14, p. 25. * Sond6n and Tigerstedt, Skand. Archiv f. Physioi., 1895, 6, p. 1. * Rubner, Beitrage zur Ernahrung im Knabenalter, Berlin, 1902. 4 Tigerstedt, Carl, Ueber die Nahrungszufuhr dea Menschen in ihrer Abhangigkeit von Alter, Geschlecht und Beruf, Helsingfors, 1915. Also published in Skand. Archiv f. Physioi., 1916, 34, p. 151. 24-HOUR ENERGY REQUIREMENTS. 211 were inside a respiration chamber with the activity considerably restricted. TABLE 38. Twenty-four hour energy requirements of boys. Age. Net energy. Authority. Per kg. Per sq. m. yrs. 9 10 11 12 13 14 cals. 56 49 50 48 37 34 cals. 1,325 1,233 1,356 1,277 946 862 Tigerstedt, Carl, Ueber die Nahr- ungszufuhr des Menschen in ihrer Abhangigkeit von Alter, Geschlecht und Beruf, Helsingfors, 1915, p. 78. Also published in Skand. Archiv f. Physiol., 1916, 34, p. 151. Thus far we have been able to consider children at two distinct levels of activity: (1) basal, i. e., with complete muscular repose; (2) with activity restricted by the confines of a respiration chamber. With most children many hours in the day are spent in the school-room. The classic research of Sonden and Tigerstedt has supplied us with information as to the energy requirements of children under these conditions of modern activity; their computations were made by Carl Tigerstedt 1 on the basis of calories per kilogram of body-weight, assuming that 1 gram of carbon dioxide corresponds to 3 calories. Although these values are in no sense basal, they represent data obtained with an approximately uniform muscular activity for the various ages. The results for the age-period within the range of our study are given in table 39, and show clearly a decrease in metabolism per kilogram of body-weight from 7.9 years to 15.5 years of age. The values all lie considerably higher than not only the basal, but also the values for 24-hour chamber experiments given in table 38, thus showing on the whole a somewhat greater degree of activity for the school children. These latter values more nearly approximate the true 24- TABLE 39. Twenty-four hour energy requirements of children (Sond&n and Tigerstedt). Boys. Girls. Authority. Age. Heat (computed) per kilo, per 24 hrs. Age. Heat (computed) per kilo, per 24 hrs. yrs. 7.9 9.6 10.5 11.4 12.5 13.8 14.5 15.5 cals. 70 75 69 66 62 62 60 50 yrs. 7.9 9.9 11.2 12.2 13.2 14.0 15.1 15.6 cals. 70 53 52 46 43 41 35 40 Tigerstedt, Carl, Ueber die Nahrungszufuhr des Mensch- en in ihrer Abhangigkeit von Alter, Geschlecht und Beruf, Helsingfors, 1915, p. 79. See also Skand. Arch. f. Physiol., 1916, 34, p. 151. Tigerstedt, Carl, loc. cit. 212 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. hour requirement than those previously discussed, although even here the children were not studied during their outdoor or extra-chamber activity. In this connection a study may be made of the curves in figures 33 and 34 (page 152), in which comparisons are made between the basal metabolism values of our observations and the values obtained in the chamber experiments of foreign investigators just discussed. The curves representing our values for boys from 7 to 16 years and for girls from 7 to 14 years are obviously projected for the sake of comparison, as material for the later ages is lacking. The other two curves on the boys' chart correspond to the values computed by Carl Tigerstedt from the earlier chamber experiments (table 38) and the extensive series of observations by Sond6n and Tigerstedt on school children ranging in age from 8 to 23 years. On the assumption that the school children would have a varying degree of activity which would be measured with a fair degree of accuracy by the carbon- dioxide production per 2 hours, we selected for the plotting of the Sonden and Tigerstedt curve the more probable minimum carbon- dioxide values at each age, discarding those that were obviously increased by activity. (See table 4, page 10.) Even with this method of eliminating the excessively high carbon-dioxide production, it can be seen that the heat-output obtained with school children was very much higher than either the values found in the 24-hour chamber experiments included in table 38 or (more especially) the basal values obtained by us. Of special significance are the two experiments of Sonde*n and Tigerstedt with two sleeping boys, one 11 and the other 12 years of age. Their results, indicated by small crosses, lie very near our basal values. As an approximate figure, one may state that the caloric output of school children in the school-room would be, with boys, approximately 75 per cent above basal (see figure 33) ; with girls the increment would be more nearly 50 per cent (see figure 34). Although not sufficiently accurate for use in computing the entire 24-hour needs of growing children, these figures are of distinct value in interpreting the food needs of young children and the sum total of the day's activities of a growing child. It is a tribute to the foresight and skill of the Scandi- navian investigators to realize that now, after more than 25 years, these experiments still remain of definite practical value. To sum up, we have in these two measurements, i. e., the basal value and the value for school children, two steps in the important computation of the 24-hour requirement. The basal requirement is substantially the energy requirement during sojourn in bed, this period with children varying from 8 to 10 or even 12 hours. The school value will represent the requirement for 5 to 6 hours of the day. It is thus possible to compute the total energy output for 16 to 18 24-HOUR ENERGY REQUIREMENTS. 213 out of the 24 hours, though hi the remaining 6 hours the child may be occupied in such vigorous muscular exercise as to require a very large correction of these figures before the ultimate 24-hour computation may even be approximated. To explain the extraordinary needs of growing children solely upon the basis of activity is somewhat difficult. The activities are, it is true, enormous. The food consumption is proportionately great. The deposition of tissue must be provided for from the food intake, and this, in turn, augmented above the true needs for simple physical activity. In all probability a factor by no means to be neglected is the stimulus to metabolism resulting from the ingestion of food. As has been shown in the report of an earlier research 1 on mixed diets, especially when large amounts of food are taken, approximately 6 per cent of the total caloric intake is eliminated as extra heat, which has been technically termed the "cost of digestion." The final computation of the total 24-hour food-needs or heat-output of a growing active child will require considerable research. The heat- output of children at play is entirely a matter of speculation. The determination of such heat production is by no means a technical impossibility; indeed, the large respiration chamber at the Nutrition Laboratory is designedly constructed for the measurement of exactly this type of group activities, and it is hoped that information on these points may ultimately be secured. In any event, it is clear that the caloric needs of growing children are very much greater than they are commonly supposed to be. The lesson to be drawn from our observa- tions on private-school children (see page 72) is that outdoor life and physical activity contribute towards the development of a larger indi- vidual, so far as height, i. e., skeletal growth, is concerned, with like- wise a greater weight with children of the same age. But it is probable that even these children, with superior surroundings and presumably better medical examination, care, and dietetic supply, may advan- tageously be supplied with larger amounts of food than they at present take. One could infer, therefore, from these observations, that, aside from the possibilities of digestive derangements, it would be impossible to supply the growing child with an excessive amount of food. Every effort may legitimately be expended to secure a maximum skeletal growth and the development of children above so-called average weight. We believe that our investigation shows clearly that the average weight for children is distinctly below the ideal or physio- logically desirable weight. 1 Benedict and Carpenter, Carnegie Inst. Wash. Pub. No. 261, 1918, p. 341. ' -' * * cJ? ^ ""^ ?? s 1 A 000 376 352 ^ & O; T 1 O i o I I S i S I 1 I ^ $ ^ AfclOSANGR&a g |/^V-^^ i I Vt . > g ^