National Cholesterol Education Program Report of the Expert Panel on Population Strategies for Blood Cholesterol Reduction U.S. DEPARTMENT OF HEALTH & HUMAN SERVICES Public Health Service » National Institutes of Health + National Heart, Lung, and Blood Institute (oP National Cholesterol Education Program NIH Publication No. 90-3046 November 1990 Population Panel Report DedicationTo JoserH STOKES III, M.D. Dr. Joseph Stokes was a member of this panel until his untimely death in June 1989. It was his characteristically clear-sighted vision of the need for population strategies to lower blood cholesterol that led to the formation of the panel, and his contributions to the formulation and execution of the panel's work were extraordinary. Dr. Stokes played a pivotal role in shaping the development of the National Cholesterol Educa- tion Program (NCEP) as a leading member of the program's Coordinating Committee. His insight, scholarship, and com- mitment to public health were vitally important in ensuring the early success of the NCEP. Dr. Stokes’ leadership and effective advocacy on behalf of the NCEP mirrored his dedication and monumental contributions to the prevention of cardiovascular disease in the United States. We dedicate this report to the memory of Dr. Joseph Stokes with respect and affection. Population Panel Report ii National Cholesterol Education Program Expert Panel on Population Strategies for Blood Cholesterol Reduction Richard A. Carleton, M.D.—Chairman of the Panel Professor of Medicine, Brown University Physician in Chief, Memorial Hospital of Rhode Island Pawtucket, RI Johanna Dwyer, D.Sc., R.D. Director Frances Stern Nutrition Center New England Medical Center Hospitals Boston, MA Laurence Finberg, M.D. Professor and Chairman, Children's Medical Center of Brooklyn S.U.N.Y. Health Science Center at Brooklyn Brooklyn, NY June Flora, Ph.D. Assistant Professor Institute for Communication Research Stanford University Stanford, CA DeWitt S. Goodman, M.D. Professor of Medicine Columbia University College of Physicians and Surgeons New York, NY Scott M. Grundy, M.D., Ph.D. Director, Center for Human Nutrition University of Texas Southwestern Medical Center at Dallas Dallas, TX Stephen Havas, M.D., M.P.H., M.S. Associate Professor University of Maryland School of Medicine Department of Epidemiology and Preventive Medicine Baltimore, MD Gertrude T. Hunter, M.D. Professor Community Health and Family Practice Howard University College of Medicine Washington, DC David Kritchevsky, Ph.D. Associate Director Institute Professor Wistar Institute Philadelphia, PA Ronald M. Lauer, M.D. Professor of Pediatrics and Preventive Medicine Director, Division of Pediatric Cardiology University of Iowa Hospitals and Clinics Iowa City, IA Population Panel Report Russell V. Luepker, M.D. Professor Division of Epidemiology School of Public Health University of Minnesota Minneapolis, MN Amelie G. Ramirez, M.P.H. Assistant Director for Administration and Community Health Promotion South Texas Health Research Center University of Texas Health Science Center at San Antonio San Antonio, TX Linda Van Horn, Ph.D., R.D. Associate Professor Northwestern University Medical School Department of Community Health and Preventive Medicine Chicago, IL William B. Stason, M.D. Director, Northeast Health Services Research Program VA Medical Center West Roxbury, MA Joseph Stokes III, M.D.* *Deceased Ex-Officio Members James I. Cleeman, M.D.—Executive Director of the Panel Coordinator, National Cholesterol Education Program Office of Prevention, Education, and Control National Heart, Lung, and Blood Institute Bethesda, MD Nancy Ernst, M.S., R.D. Nutrition Coordinator National Heart, Lung, and Blood Institute Bethesda, MD William Friedewald, M.D. Associate Director for Disease Prevention National Institutes of Health Bethesda, MD Basil Rifkind, M.D. Chief, Lipid Metabolism - Atherogenesis Branch Division of Heart and Vascular Diseases National Heart, Lung, and Blood Institute Bethesda, MD Consultant Karen A. Donato, M.S., R.D. Coordinator, Professional and Patient Nutrition Education Office of Prevention, Education, and Control National Heart, Lung, and Blood Institute Bethesda, MD Staff Marguerite A. Evans, M.S., R.D. University Research Corporation Bethesda, MD Laurie A. Quint-Adler, M.P.H., R.D. University Research Corporation Bethesda, MD iii Population Panel Report Table of Contents I. OVERVIEW AND SUMMARY II. BACKGROUND AND INTRODUCTION A. B. C. Coronary Heart Disease Rationale for Both a Population Approach and a High-Risk Approach to Blood Cholesterol Guidelines From Other Authoritative Sources 1. International 2. United States II. CONCLUSIONS OF THE POPULATION PANEL IV. RECOMMENDATIONS A. S— TT 0m mg Nw Nutrient Intake 1. Rationale for Nutrient Intake Recommendations Eating Patterns Healthy Children and Adolescents Special Groups Health Professionals The Food Industry Mass Media . Government Educational Systems Measurement of Blood Cholesterol Research and Surveillance Page 12 15 17 17 18 20 24 24 25 26 27 27 28 29 30 Vv Population Panel Report Table of Contents (Continued) V. SCIENTIFIC EVIDENCE FOR RECOMMENDATIONS AFFECTING THE GENERAL PUBLIC \% A. Diet Patterns, Blood Cholesterol, and Health 1. oS ak WN Introduction Clinical, Laboratory, and Metabolic Research Epidemiologic Research Observations in Special Populations Intervention Studies Other Risk Factors for CHD Low Blood Cholesterol and Risk Eating Patterns in the United States 1. Bro Ww Current Dietary Patterns in the United States Fat in the American Diet Change Required in American Diet To Achieve Various Targets Dietary Safety and Adequacy Special Concerns Involving Women What Influences Eating Patterns? 1. 2. 1. Overview Factors That Influence Eating Patterns and Approaches to Changing Them . Can Intervention Modify Dietary Behavior? The Multiple Levels at Which Change is Possible: Individual, Network, Organizational, and Community Individual-Level Interventions Network-Level Interventions Page 33 77 82 Population Panel Report Table of Contents (Continued) 4. Organization-Level Interventions 5. Community-Level Interventions 6. Conclusions 7. Summary ~ VI. ETHNIC, CULTURAL, AND MINORITY CHARACTERISTICS THAT INFLUENCE DIET AND HEALTH A. Introduction 1. Dynamics of Culture Change 2. Health Belief Systems 3. Acculturation 4. Poverty B. Characteristics of Important Population Subgroups 1. Blacks Hispanics Asian/Pacific Islanders Native Americans LI SE Summary VII. SCREENING THE POPULATION FOR BLOOD CHOLESTEROL LEVEL VIII. GLOSSARY IX. BIBLIOGRAPHY Page 83 91 92 92 93 93 93 93 94 94 96 96 97 98 99 100 101 103 105 vii = wey tam ——— oo SAREE RAIS Lr hra - DE = AM nm = ah Sle. Sa. a tie = an Do EE SL SSD - . a ate SL. I a == Population Panel Report List of Figures Relationship Between Serum Cholesterol Level and CHD Death Rate (MRFIT Screenees) Cholesterol Distribution in the U.S. Population (Males/ Females Combined, Ages 20-74, NHANES II 1976-80), and Potential Changes in the Distribution The Seven Countries Study: Total Blood Cholesterol and Percentage of CHD Deaths in Men Originally Free of Cardiovascular Disease Eating Pattern Determinants A Multilayered View of Cholesterol-Lowering Strategies: Levels of Intervention and Corresponding Targets of Change List of Tables Recommendations of the Adult Treatment Panel of the National Cholesterol Education Program for Classification of Patients CHD Risk Factors Other Than LDL-Cholesterol Dietary Recommendations Related to Blood Cholesterol Levels by United States Organizations Since 1983 Guidelines for Referral of People Based on Their Blood Cholesterol Level Mean Daily Intake of Food Energy, Nutrients, and Food Components for Men, Women, and Young Children from the Continuing Survey of Food Intakes by Individuals (CSFII), 1985 Sociodemographic Characteristics of Population Subgroups in the United States Page 10-11 43 70 76 13 30 68 95 1X Population Panel Report I. Overview and Summary Coronary heart disease (CHD) is a major pub- lic health problem in the United States. Despite substantial success in reducing premature deaths from CHD in the past two decades, this disease continues to kill more than 500,000 Americans annually. About 1,250,000 Americans suffer myocardial infarctions each year, and millions more have angina pectoris. In addition, signifi- cant degrees of asymptomatic CHD are very common in our population. In addition to its impact on the Nations health, CHD costs the U.S. economy over $50 billion annually. Coronary heart disease is the result of athero- sclerosis, in which deposits of cholesterol and other lipids, along with cellular reactions, thicken ar- tery walls. This process gradually reduces the caliber of the artery and restricts blood flow. In- adequate blood flow may cause injury to or death of tissue beyond the site of reduced flow; in the coronary arteries, this leads to myocardial infarc- tion or sudden death. Many factors influence not only whether a person will develop CHD but also how rapidly atherosclerosis progresses. Genetic predisposi- tion, gender, and advancing age are recognized factors over which we have little control. High blood cholesterol, cigarette smoking, high blood pressure, excessive body weight, and long-term physical inactivity are also key risk factors over which we have considerable control. Control of each of these factors is important in the preven- tion of CHD. People with diabetes may also be able to avoid or delay vascular disease by control- ling the other risk factors. High blood cholesterol levels clearly play a causal role in CHD. This conclusion is based on experimental laboratory work, extensive clinical and pathological research, and numerous epidemi- ologic studies over the past several decades. The magnitude of the problem posed by elevated blood cholesterol levels is very clear. High CHD rates occur among people with high blood cholesterol P levels of 240 mg/dL (6.21 mmol/L) or above. However, an even larger number of cases occurs in Americans with blood cholesterol levels below 240 mg/dL. It is important to note that the aver- age cholesterol level for the adult U.S. population is about 210 mg/dL. Furthermore, approximately 55 percent of adult Americans have cholesterol levels at or above 200 mg/dL. For these reasons, two kinds of strategies are needed: patient-based strategies, which seek to help those with the highest blood cholesterol lev- els, and population-based strategies, which seek to reach all Americans. The population approach aims both to lower the blood cholesterol level of individuals and to reduce the average cholesterol level throughout the population. When both approaches are used, the effects are synergistic. Two important types of research provide ground for optimism that CHD morbidity can be reduced, and that the trend of reduced CHD death rates will continue or accelerate. First, studies consistently indicate that reducing blood choles- terol can reduce the likelihood of developing or dying from CHD. Dietary factors in individuals and in entire populations have important effects on blood cholesterol levels (although genetic makeup, expressed, in part, as a low level of HDL- cholesterol or as a high level of LDL-cholesterol, also plays a major role). Second, this report clearly indicates that many effective strategies exist for helping Americans develop more healthful eat- ing patterns and lifestyles. The Population Panel of the National Choles- terol Education Program offers a set of recom- mendations designed to help healthy Americans lower their blood cholesterol levels through Population Panel Report changes in eating patterns and thus reduce their likelihood of developing CHD. Recognizing that Americans’ food consumption is influenced by many factors, the panel directs its recommenda- tions to individuals, to special population groups, to health professionals, to the many components of the food industry, to relevant government agen- cies, and to public and private education systems. In addition, the panel’s recommendations address cholesterol screening, and the need for continued research and evaluation as eating patterns change, blood cholesterol levels decrease, and CHD con- tinues to decline. The panel recommends the following nutrient intakes for healthy Americans: * Less than 10 percent of total calories from saturated fatty acids. * An average of 30 percent of total calories or less from all fat. * Dietary energy levels needed to reach or maintain a desirable body weight. * Less than 300 mg of cholesterol per day. Each is intended to be achieved by an individ- ual as an average of nutrient intake over sev- eral days. These recommendations concerning nutrient intakes are appropriate for the general popu- lation, including healthy women and individu- als 65 years of age and older. As healthy children join in the eating patterns of others in the family, usually at about 2 years of age or older, they should follow the recommended nutrient intake and eating pat- terns. To achieve more healthful eating patterns, the panel recommends that healthy Americans select, prepare, and consume foods that contain lower amounts of saturated fatty acids, total fat, and cholesterol; choose a variety of foods to ensure recommended intakes of carbohydrates, protein, and other nutrients; and consume only enough calories to maintain desirable weight. The panel also makes recommendations for other groups: Health professionals should both practice and advocate the recommended eating patterns; ensure that education of future health profes- sionals includes appropriate nutrition educa- tion; and work with industry, government, voluntary groups, and health care agencies to facilitate adoption of the recommended eat- ing patterns. The food industry, food and animal scientists, and food technologists should increase efforts to design, modify, prepare, promote, label, and distribute good-tasting, safe foods that are lower in saturated fatty acids, total fat, and cholesterol. Government agencies should provide consistent, coordinated nutrition statements and policies emphasizing low saturated fatty acid, low-fat, and low-cholesterol eating patterns; should expand and standardize food labeling require- ments to identify clearly the content of satu- rated fatty acids, total fat, cholesterol, and to- tal calories; and should take other steps to im- prove the consumer comprehension necessary to achieve the recommended eating patterns. Educational programs at all levels should incor- porate curricula that emphasize the back- ground, benefits, and methods of achieving eating patterns that are lower in saturated fatty acids, total fat, and cholesterol. This rec- ommendation includes elementary through high schools, vocational programs (especially in culinary arts), colleges, universities, and health professional schools. Measurement of blood cholesterol, followed by appropriate education and counseling, is best initiated in the health care setting; but in spe- cific circumstances and especially for selected segments of hard-to-reach population groups, public screening for blood cholesterol, when carried out with high quality standards, is ap- propriate. Research and surveillance must be ongoing to develop new information concerning diet, Population Panel Report blood lipids, and CHD; the development of better data bases concerning food composi- tion, food consumption patterns, illness rates, food product development, and nutrition education and communication is critical. Implementation of these recommendations will promote adoption of eating patterns that will help most Americans lower their levels of blood cholesterol. The result, an approximate reduction of 10 percent or more in the average blood choles- terol level of the U.S. population, will lead to an approximate reduction of 20 percent or more in coronary heart disease and, in consequence, to significant improvement in the health and quality of life of Americans. ST TN eee aR, pee rome wEpTeess ns = = = - Population Panel Report II. Background and Introduction Cholesterol was identified many years ago as a prominent chemical component of the athero- sclerotic lesion in human arteries. Following those early observations, a large base of scientific evi- dence has accumulated linking dietary saturated fatty acid (SFA) and cholesterol intake, blood cho- lesterol, atherosclerosis, and coronary heart dis- ease (CHD). Animal, biochemical, genetic, meta- bolic, clinical, and epidemiologic research have all made important contributions to understand- ing these relationships. Clinical intervention trials have shown reduced CHD rates when cholesterol levels are reduced. While it is apparent that indi- viduals with high blood cholesterol levels are at high risk of CHD, a larger number of individuals with more moderately elevated cholesterol also have increased risk of coronary events; such indi- viduals comprise a substantial proportion of the adult U.S. population. From these observations, it follows that a two-pronged strategy is appro- priate to address the high prevalence of CHD in Americans. The patient-based approach seeks to identify those at highest risk and to treat them to lower their blood cholesterol level. The popula- tion approach seeks to influence the population at large to alter current eating patterns, and thereby to lower individual and average population lev- els of blood cholesterol. These two approaches are complementary means of achieving lower rates of CHD and better health of the public. This report deals with the population approach. In 1984, the National Institutes of Health (NIH) convened a Consensus Development Conference on Lowering Blood Cholesterol to Prevent Heart Disease. Among the recommendations from this conference were identification and treatment of individuals with high-risk cholesterol levels, changes in eating patterns for members of the general public, and the creation and implementa- tion of a national cholesterol education program, which the National Heart, Lung, and Blood Insti- tute (NHLBI) had begun to plan in early 1984. The National Cholesterol Education Program P (NCEP) was launched in November 1985. The goal of the NCEP is to reduce the prevalence of elevated blood cholesterol in the United States, thereby contributing to the reduction of CHD morbidity and mortality. The NCEP is designed to reach and influence both health professionals and the public. This program provides an oppor- tunity for individuals to assume an active role in reducing their risk of CHD. Under the direction of a multidisciplinary coordinating committee rep- resenting a variety of agencies and organizations, the program has accelerated the process of in- creasing public and professional awareness of the importance of reducing elevated blood choles- terol levels. The NCEP has issued reports devel- oped by its Expert Panel on Detection, Evalu- ation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel or ATP) and its Laboratory Standardization Panel (LSP). The pres- ent report has been developed by a third panel, the Population Panel, which has conducted an in- tensive review of the scientific basis for making blood cholesterol-lowering recommendations, especially eating pattern recommendations, to the general public. A fourth panel will report later on blood cholesterol reduction in children and ado- lescents. A. Coronary Heart Disease Atherosclerosis is a complex disorder, the development of which involves the interplay of many factors. Blood cells such as monocytes and platelets, the blood coagulation system, endothe- lial cell damage, and deranged function of arterial Population Panel Report wall macrophages and smooth muscle cells all contribute to the atherosclerotic process. Many other risk factors play important roles. The proc- ess involves the deposition of cholesterol in arte- rial walls and progresses with cellular prolifera- tion, fibrous tissue deposition, and further choles- terol deposition, creating progressively larger atherosclerotic plaques. As atherosclerosis pro- gresses, the opening in the artery through which blood flows is gradually reduced in size. The degree of arterial narrowing may be abruptly in- creased by thrombosis, by plaque disruption, or by smooth muscle spasm. Closure of an artery may cause injury and death of tissue beyond the site of blockage. When a coronary artery closes, myocardial infarction or sudden death is a com- mon outcome. Approximately 1,250,000 Americans each year suffer a myocardial infarction. More than 500,000 die annually from CHD. More than 6,000,000 Americans have symptomatic CHD, while it is estimated that nearly 70 percent of adult Ameri- cans have some degree of atherosclerotic narrow- ing of their coronary arteries. From a public health perspective, CHD is a major health problem. The economic costs have been estimated to be over $50 billion annually for illness care and lost earn- ings and productivity related to CHD. Although the precise causes of atherosclerosis have not been identified fully, clinical, epidemi- ological, animal, and biochemical evidence dem- onstrate that many major factors contribute to the process. Three of these factors—gender, advanc- ing age, and genetic composition—cannot be changed, but their effects can be mitigated. Several other important risk factors for CHD are modifiable and provide opportunities for in- dividuals and populations to reduce their likeli- hood of developing CHD. The most important are high blood cholesterol, high blood pressure, and cigarette smoking. Numerous epidemiologic studies show that the smoking habit increases the chances of developing CHD, as well as the proba- bility of lung cancer or emphysema. The risk of both CHD and lung disease decreases with smok- ing cessation. High blood pressure accelerates the atherosclerotic process; control of high blood pressure reduces the likelihood of stroke and probably decreases the incidence of CHD. Other risk factors have also been identified. Three are of particular note: overweight, diabetes mellitus, and a sedentary lifestyle. Maintaining or achieving a desirable weight is important both by itself (because excess weight is probably an in- dependent risk factor for CHD) and because people who have a desirable body weight are less likely to have high blood cholesterol, a low high density lipoprotein (HDL) cholesterol level, or high blood pressure and are also less likely to de- velop diabetes mellitus than obese people. It may be possible to prevent many cases of non-insulin- dependent diabetes mellitus by avoiding obesity; furthermore, there is hope that control of the three major modifiable CHD risk factors will help dia- betics avoid many vascular complications. Mod- erate, regular aerobic exercise throughout life is also important, in part because it enhances car- diovascular fitness and facilitates control of high blood pressure and excess body weight, and in part because of its generally favorable effect on HDL-cholesterol. As has been reviewed extensively in section V of this report, entitled Scientific Evidence for Rec- ommendations Affecting the General Public, con- sistent evidence from multiple avenues of research implicates high blood cholesterol in the causation of atherosclerosis and CHD. Both genetic and dietary factors influence blood cholesterol levels. Saturated fatty acids (SFA), cholesterol, and ex- cessive calories in the diet tend to elevate the blood cholesterol level. There is now conclusive evidence that lowering elevated blood cholesterol decreases CHD risk; this has special importance for dietary recommendations. The NCEP represents a major effort to reduce CHD rates in the United States by helping to lower elevated blood cholesterol levels, in part through favorable alterations in eating patterns. Together, the National High Blood Pressure Edu- cation Program, federally sponsored smoking education programs, and the NCEP are designed to reduce the three major modifiable risk factors for CHD and improve the health of the American public. Population Panel Report B. Rationale for Both a Population Approach and a High-Risk Approach to Blood Cholesterol Epidemiological, clinical, and experimental evidence clearly demonstrate that the likelihood that a person will develop or die from CHD is directly related to the level of blood cholesterol. This relationship is well illustrated by the 10-year followup experience of men, initially ages 35 to 57 years, screened for the Multiple Risk Factor Inter- vention Trial (MRFIT).! As shown in figure 1, using data from MRFIT, the relationship is con- tinuous and the risk increases with rising choles- terol levels. The risk of dying from CHD increases slowly between 150 mg/dL (3.88 mmol/L) and 200 mg/dL (5.17 mmol/L) and more rapidly when the cholesterol level exceeds about 200 mg/dL (5.17 mmol/L). For those individuals with cho- lesterol levels between 240 mg/dL (6.21 mmol/L) and 300 mg/dL (7.76 mmol/L), the likelihood of CHD is increased up to fourfold above that for those with levels below 200 mg/dL (5.17 mmol/ L). Similar relationships between blood choles- terol and CHD have been found in many other studies. In the United States, as in many other nations of the world, a majority of people have a blood cholesterol level high enough to increase signifi- cantly the likelihood of CHD. Thus, data from the National Health and Nutrition Examination Sur- vey, known as NHANES II (1976-80) (figure 2, page 10), showed the mean cholesterol level for the adult U.S. population to be about 210-215 mg/dL (5.43-5.56 mmol/L).> Approximately 30 percent of adult Americans had cholesterol levels of 200-239 mg/dL (5.17-6.18 mmol/L), and more than 25 percent had levels of 240 mg/dL (6.21 mmol/L) and above. Figure 1 shows the high CHD death rate of those in this upper 25 percent of the cholesterol distribution. Despite the lower relative risk of the 75 percent of the population with blood cholesterol levels below 240 mg/dL (6.21 mmol/L), this group experiences about 60 percent of all CHD deaths, reflecting both the larger number of people at risk and the effects of mildly to moderately elevated cholesterol levels as well as the influence of other risk factors. The fact that risk of CHD is by no means restricted to those with high blood cholesterol levels supports the need for a combined population and high-risk approach to CHD prevention. Based on the scientific evidence of increased CHD risk with elevated blood cholesterol, along with the evidence that risk can be reduced by lowering high blood cholesterol levels, the ATP of the NCEP issued a series of recommendations for a strategy of detecting and treating individu- als with a high blood cholesterol level.” This panel, building on the work of the Consensus Develop- ment Conference on Lowering Blood Cholesterol to Prevent Heart Disease, classified blood choles- terol levels for adults. Levels under 200 mg/dL (5.17 mmol/L) are termed "desirable." Levels be- tween 200 mg/dL and 239 mg/dL (5.17 and 6.18 mmol/L) have been classified "borderline-high,” while levels of 240 mg/dL (6.21 mmol/L) and Figure 1 Relationship Between Serum Cholesterol Level and CHD Death Rate (MRFIT Screenees) 50 40+ w o T 10 Year CHD Death Rate (Deaths/1,000) nN o 10 , A 1 Lo 1 | 1 Lo 11 3 ed 1 | 1 0 150 200 250 300 Serum Cholesterol (mg/dL) From 361,662 Men Screened for MRFIT Program Population Panel Report above have been designated as "high." The ATP has presented a comprehensive set of recommen- dations for intervention in individuals with cho- lesterol levels above the desirable range. These recommendations emphasize that after initial screening for total cholesterol level, therapeutic decisions should be based on the level of low density lipoprotein (LDL) cholesterol (see table 1). Adult Americans with LDL-cholesterol levels that are persistently high-risk (or borderline-high- risk if definite CHD or two or more of the other major CHD risk factors listed in table 2 are pres- ent) should follow the recommendations of the ATP of the NCEP. The ATP set forth detailed recommendations indicating that dietary therapy is the mainstay of treatment. Drug therapy should generally be considered only after 6 months of intensive diet- ary therapy reinforced by the help of a health professional with special nutrition expertise (e.g., a registered dietitian), and then only if the LDL- cholesterol level remains significantly elevated. This conservative approach to drug therapy is wise because drugs can not only bring benefits but can also have deleterious effects. The potential impact of the recommendations of the ATP on the blood (serum) cholesterol dis- tribution of the adult U.S. population is indicated conceptually by the broken line in figure 3, page 10. These recommendations of the ATP should, over the years to come, reduce the proportion of individuals in the United States with high choles- terol levels (240 mg/dL [6.21 mmol/L] and above), and should also reduce the proportion of indi- viduals with borderline-high cholesterol levels (200-239 mg/dL [5.17-6.18 mmol /L]). Even in the event that the high-risk approach recommended by the ATP is largely effective, however, many people will remain at risk of CHD. The recommendations from the Population Panel of the NCEP follow. They are intended to extend the benefits of cholesterol lowering to the population as a whole by promoting adoption of eating patterns that will help lower the blood cholesterol levels of most Americans. In turn, these eating patterns should help to shift the curve representing the population distribution of cho- lesterol to the left, as shown in figure 4, page 11. As the recommended eating patterns become widely adopted, the average blood cholesterol level should decrease by approximately 10 per- cent or greater. Table 1 Recommendations of the Adult Treatment Panel of the National Cholesterol Education Program for Classification of Patients Classification Based on Total Cholesterol Classification Based on LDL-Cholesterol <200 mg/dL (<5.17 mmol/L) Desirable Blood Cholesterol 200-239 mg/dL (5.17-6.18 mmol/L) Borderline-High Blood Cholesterol 2240 mg/dL (26.21 mmol/L) High Blood Cholesterol Source: NCEP ATP, January 1988.° <130 mg/dL (<3.36 mmol/L) Desirable LDL-Cholesterol 130-159 mg/dL (3.36-4.11 mmol/L) Borderline-High-Risk LDL-Cholesterol 2160 mg/dL (=4.13 mmol/L) High-Risk LDL-Cholesterol Population Panel Report Table 2 CHD Risk Factors Other Than LDL-Cholesterol - Male sex Cigarette smoking + Hypertension measurement) « Diabetes mellitus disease + Severe obesity (>30% overweight) Source: NCEP ATP, January 1988.° - Family history of premature CHD (definite myocardial infarction or sudden death before age 55 in a parent or sibling) « Low HDL-cholesterol concentration (below 35 mg/dL confirmed by repeat - History of definite cerebrovascular or occlusive peripheral vascular The anticipated combined and compatible ef- fects of the recommendations from the Adult Treat- ment Panel and from the Population Panel of the NCEP are estimated in figure 5, page 11. Most Americans would have a desirable blood choles- terol level. Far fewer Americans would require drug therapy for persistently high blood choles- terol. The anticipated effect would be to reduce the morbidity from CHD and to produce a contin- ued or accelerated decline in the mortality rate from CHD in the United States. C. Guidelines From Other Authoritative Sources 1. International The work of many scientists has shown that atherosclerosis and CHD constitute a major health problem afflicting numerous nations of the world. The linkage of CHD to national eating patterns has been established. There is widespread inter- national consensus, based upon repeated reviews of the scientific evidence, that the average popu- lation intakes of total fat, SFA, dietary cholesterol, and food energy are excessive in many countries and should be reduced. As a result, since 1973 more than 40 groups of scientists and health pol- icy makers from other nations have recommended changes in population eating patterns with the purpose of reducing blood cholesterol. Most have focused on the prevention of CHD. These recom- mendations have been summarized by Truswell (1983),* and in the report of the National Research Council (NRC), Diet and Health: Implications for Re- ducing Chronic Disease Risk.” A World Health Or- ganization (WHO) study group in 1989 proposed "population nutrient goals" to control noncom- municable diseases and suggested that these be introduced as part of nutrition policy in all coun- tries®”. These nutrient goals are remarkably simil- iar to those of the NRC 1989 and NCEP 1990. Population Panel Report Cholesterol Distribution in the U.S. Population (Males/Females Combined, Ages 20-74, NHANES Il 1976-80),2 and Potential Changes in the Distribution Figure 2 Population distribution of serum cholesterol values. The borderline- high and high cutoff levels are shown as dotted lines to indicate the proportions of the population above or below 200 or 240 mg/dL (5.17 or 6.21 mmol/L). Figure 3 Expected shift in population distri- bution of serum cholesterol val- ues with widespread application of ATP guidelines. The dotted- dashed line represents an esti- mate of the effect of treating many people with elevated cholesterol levels. 10 10 c sr 02 dt 8 3 Q 6 o a [— o — Ar jo [«}] o 1 [+ }] a 2 0 1 1 100 200 300 400 Serum Cholesterol (mg/dL) 10 ah \ 5 \ £ \ FA > 9) \ a \ © — 4 \ c \ [V] : \ Qa 2+ MN / N 0 1 1 | i 1 A 1 1 1 1 1 1 1 ~~ 100 200 300 400 Serum Cholesterol Level (mg/dL) Population Panel Report Cholesterol Distribution in the U.S. Population (Males/Females Combined, Ages 20-74, NHANES Il 1976-80),2 and Potential Changes in the Distribution 0 Figure 4 oy / Expected shift in population dis- £ 8h / \ tribution of serum cholesterol val- = / AN ues if the recommendations of the = / : \ Population Panel result in a 10% 2 6 / : \W\ decrease in blood cholesterol of a / Pou Americans. The dashed line = : \ \ shows the effect of the recom- = ar / : \iN mendations. Q / : * \ 5 / i, \ ar / : :o\, / FNS 0 Lona i I : Lo A Loy et" 100 200 300 400 Serum Cholesterol Level (mg/dL) 10 Figure 5 7 "4n Anticipated combined effects of c sf iN the recommendations of the ATP S I 0% (dotted-dashed line) and the = y, Population Panel (dashed line). go y % ), 9 E\ o : 5 © PA o ~ a ob \ : : "tn, 0 Aaa et dh — Le eid dT] 100 200 300 400 Serum Cholesterol Level (mg/dL) 11 Population Panel Report 2. United States Over the years, many reports linking CHD and diet have appeared in the United States. These have included several diet statements from the American Heart Association (AHA) issued be- tween 1961 and 1988, the 1970 and 1984 reports of the Inter-Society Commission for Heart Disease Resources concerning primary prevention of ath- erosclerotic diseases,’ the Bethesda Conference on Prevention of Coronary Heart Disease," and the recent NRC report on diet and health.” 12 The dietary recommendations issued since 1983 by major health organizations in the United States are summarized in table 3. There are com- mon themes—recommendations for a lower in- take of SFA and total fat, a reduction in dietary cholesterol, an increased intake of complex carbo- hydrates, and the control of obesity. The advice given is similar from all these organizations, al- though some groups are more specific than oth- ers. Several recommend reductions of total fat intake to 30 percent or less of calories, of SFA to less than 10 percent of calories, and of dietary cholesterol to less than 300 mg per day. Population Panel Report Table 3 Dietary Recommendations Related to Blood Cholesterol Levels by United States Organizations Since 1983 Agency/Report Year Audience Saturated Total Cholesterol Carbohydrate Body Weight Fat Fat USDA/DHHS 1985 Public Avoidtoo Avoidtoo Avoid too Avoid too Maintain Dietary Guidelines much much much much sugar; desirable for Americans Increase weight starch & fiber NCEP-ATP 1987 Report of the Expert Panel on. . .High Blood Cholesterol in Adults Step 1 Adult <10% <30% <300mg/day 50-60% Achieve & maintain patients desirable weight Step 2 with high <7% <30% <200mg/day 50-60% blood cholesterol AHA 1988 Dietary Treatment of Hypercholesterolemia Steps 1 & 2 for patients identical to NCEP Steps 1 & 2 AHA 1988 Public <10% <30% <300 mg/day 50-55% with Maintain best body Dietary Guidelines for emphasis on weight Healthy American Adults complex carbs. AHA 1983 Children <10% Approx. <300 mg/day Approx. Caloric intake Diet in the Healthy 30% 55% based on growth Child rate & activity level and to maintain de- sirable body weight AAP 1986 Children Decrease 30-40% Decrease with Maintain ideal body Prudent Life-style for with moderation weight Children: Dietary Fat moderation and Cholesterol DHHS 1988 Public Reduce Reduce Reduce Increase Achieve & maintain The Surgeon General's for most complex desirable weight Report on Nutrition carbs.; and Health” Limit sugars NRC 1988 Public <10% <30% <300 mg/day Achieve & maintain Designing Foods: Animal desirable weight Product Options in the Marketplace™* NRC 1989 Public <10% <30% <300 mg/day >55% with >6 Balance food intake Diet and Health: Implications for Reducing Chronic Disease Risk * A review supporting USDA/DHHS (1985) recommendations. servings grains & legumes/day, with physical activity to maintain >5 servings fruits body weight & vegetables per day ** The report accepted these target levels from other authorities; it also made recommendations on marketing and policy issues and research imperatives. Cm eye ye apr See Pp i . Population Panel Report III. Conclusions of the Population Panel After carefully considering the extensive sci- entific evidence linking blood cholesterol, athero- sclerosis, CHD, and diet, this panel concludes that excessive intakes of saturated fatty acids, to- tal fat, and dietary cholesterol, together with ex- cessive body weight, all contribute importantly to biologically unnecessary and undesirable eleva- tions of blood cholesterol. The panel also reaf- firms the conclusion of previous panels (AHA, ATP, NRC) that elevated levels of blood choles- terol produce a high prevalence of severe athero- sclerosis in coronary arteries, resulting in a high incidence of CHD and premature death. Accord- ingly, the panel concludes that it is important for Americans to change their eating patterns to re- duce the average intakes of saturated fatty acids, total fat, and dietary cholesterol, and to eliminate excess body weight. The panel concludes that changing eating pat- terns will influence blood cholesterol levels and that eating patterns can be changed, while pre- serving the nutritional adequacy, variety, and af- fordability of good-tasting food. The panel re- gards reduction of dietary saturated fatty acid intake from both animal and vegetable sources as being of greatest importance. The work of the National Research Council" concerning the po- tential for reduction of fat in foods of animal ori- gin indicates the feasibility of change in saturated fatty acid content of foods. Furthermore, the panel P agrees that observance of the Dietary Guidelines for Americans issued by the U.S. Department of Agriculture (USDA) and the Department of Health and Human Services (DHHS) can help promote health for individuals beyond the age of 2 years." Although there are challenges to implementing some of the recommended changes, the panel be- lieves that the potential benefits outweigh the ex- penditure of resources needed to achieve these changes. The panel believes that eating patterns of most Americans can change through a process of combining education for the public and actions by the many sectors of society that influence the availability, purchase, preparation, and consump- tion of foods. The panel agrees with estimates that the health benefits to be expected for the average American, based upon available evidence, approximate a 2 percent reduction in CHD risk for each 1 percent reduction in blood cholesterol level. It is from these perspectives that the Popula- tion Panel of the NCEP offers recommendations to help individuals in the United States lower their blood cholesterol levels. 15 Population Panel Report IV. These recommendations are designed to en- courage and assist each individual to adopt eat- ing patterns to lower his or her blood cholesterol level and to reduce the number of people in the United States who develop or die from CHD. A blood cholesterol level under 200 mg/dL (5.17 mmol/L) is termed desirable for adults. To help Americans learn whether they have an elevated blood cholesterol level, this panel concurs with the ATP recommendation that all adult Ameri- cans should have their blood cholesterol level measured at least every 5 years. The average level of blood cholesterol in the U.S. adult popu- lation is approximately 210-215 mg/dL (5.43-5.56 mmol/L). It is anticipated that the recommended changes in eating patterns will result in a reduc- tion of 10 percent or greater in the average person's blood cholesterol level and shift the population distribution curve for blood (serum) cholesterol values as shown in figure 4, page 11. The nature of the food consumed by Ameri- cans is influenced by many factors. These recom- mendations are intended to provide guidance to many segments of the food production-consump- tion chain. Success will require adoption of these recommendations by all healthy Americans, health professionals, the food industry, educators at all levels, mass media, and many government agen- cies. Accordingly, the panel makes the following recommendations for the entire population of healthy Americans. This population approach aims to promote healthful food choices and to make good- tasting food, lower in saturated fatty acids and cholesterol, more widely available to all Ameri- cans. Those whose blood cholesterol levels are Recommendations PS higher than desirable can be expected to pay the most attention and to benefit the most from these recommendations. In addition, most individuals with desirable cholesterol levels can participate in the recommended eating patterns, and expect to lower CHD risk. The recommendations in this report are compatible with the Dietary Guide- lines for Americans as set forth by the USDA and the DHHS. A. Nutrient Intake The panel concludes that health and nutrition professionals, the food industry, and the general public need specific guidance concerning intake of certain nutrients. RECOMMENDATION A. 1: The panel recommends the following pattern of nutrient intake for all healthy Americans*: ¢ Less than 10 percent of total calories from saturated fatty acids. * An average of 30 percent of total calories or less from all fat.’ * Dietary energy (calorie) levels needed to reach or maintain a desirable body weight. * Less than 300 mg of cholesterol per day. * Concerning the very young, as healthy children join in the eating patterns of others in the family, usually at about 2 years of age or older, they should follow the recommended nutrient and eating patterns. t Each gram of fat contains 9 Kilocalories (often called calories). A person consuming 1,500 calories daily should consume no more than 450 calories (1,500 calories x 0.30 = 450 calories) derived from all fat. At 9 calories per gram, less than 50 grams of fat (450 calories + 9 calories/gram = 50 grams) should be consumed on the average day, and less than one-third or 17 grams of the 50 gram total should be from saturated fatty acids (1,500 calories x 0.10 = 150 calories; 150 calories + 9 calories/gram = <17 grams). 17 Population Panel Report Because food intake varies from day to day, these recommendations are meant to be achieved by each individual as an average of nutrient in- take over several days. The nutrient intakes recommended are criti- cal to health-promoting eating patterns. They are essentially the same as those recommended by the American Heart Association, the National Research Council, and in the Step-One therapeu- tic diet of the ATP of the NCEP. The eating patterns that will be required in order to achieve these nutrient intakes by the public are to be at- tained gradually over time. Ultimately, changes in the food system will facilitate these changes in population-wide eating patterns. By contrast, the Step-One diet is intended to treat individuals who are at high risk of CHD because of substantially elevated blood cholesterol levels, through inten- sive nutritional counseling in a clinical setting, and with the intent of achieving a rapid and cer- tain response. What makes the Step-One diet therapeutic for those at high risk is prescription in a medical setting along with the monitoring and followup offered by health professionals. The same eating pattern is suitable for adoption by the population at large in order to achieve a popu- lation-wide reduction of blood cholesterol. It is the intent of the recommendations of the Popula- tion Panel to foster healthful eating for Ameri- cans, young and old. 1. Rationale for Nutrient Intake Recommendations This recommended pattern of nutrient intake is intended for healthy individuals. Strong scien- tific evidence indicates that attainment of each of the four recommendations will help most indi- viduals lower their blood cholesterol and, in turn, reduce their probability of developing CHD. The specific nutrient intake recommendations are based on the scientific evidence, but the exact numerical values reflect both science and prag- matism. Thus, there is virtually unanimous agree- ment that SFA intake should be reduced. Most agree that the science base makes less than 10 18 percent of calories from SFA an appropriate level, although there is no precise guidance as to which level under 10 percent is best. For the sake of practicality, consistency, and comprehensibility, the phrase “less than 10 percent” has been cho- sen. A similar process led to recommendations for total fat and cholesterol intakes. The available knowledge concerning the mechanisms through which these nutrients affect blood cholesterol is summarized in section V, Scientific Evidence for Recommendations Affecting the General Public. The panel's rationale for each of the recom- mended nutrient intakes follows. a. Fat Intake Most of the U.S. population should reduce total dietary fat intake—by an amount that varies according to how much fat is currently con- sumed—to help achieve and maintain desirable body weight and to reduce the risk of certain diseases, particularly CHD. (See table 3, page 13, for a summary of recommendations from other authoritative groups.) Different kinds of fatty acids have different effects on blood cholesterol levels. In general, SFA raise blood cholesterol levels. Polyunsaturated fatty acids (PUFA) of the omega-6 series lower blood cholesterol levels, as do monounsaturated fatty acids (MUFA) and omega-3 PUFA, when substituted for SFA. Total Fat. The panel recommends a target for individuals of an average of 30 percent of total calories or less from total fat. The percentage of calories from total fat intake, independent of the relative content of the different types of fatty ac- ids, does not determine the level of blood choles- terol or CHD risk. As stated in the NRC report on Diet and Health: Implications for Reducing Chronic Disease Risk, “Fat intake should be reduced by curtailing the major sources of dietary fats rather than by eliminating whole categories of foods.”” It is not necessary to resort to very low fat diets (e.g., 10-20 percent of total calories) to achieve a maximal reduction of blood cholesterol levels through dietary means, provided that intakes of saturated fatty acids are kept low. Therefore, the Population Panel Report panel accepts a target for individuals of an aver- age of 30 percent or less of calories from fat. The panel does not recommend severe reduction of total fat intake; however, a limit to total fat intake is a rational part of a program aimed at reducing the risk for CHD because such a limit facilitates reduction of saturated fatty acid intake, and fa- cilitates maintenance of desirable body weight. There is also evidence that suggests that a diet lower in fat may prevent certain types of can- ger Survey data indicate that for children ages 1 to 5 years, the average proportion of calories de- rived from fat was 34 percent; for women ages 19 to 50 years, 37 percent; and for men ages 19 to 50, 36 percent.” Thus, to achieve an average intake of 30 percent of calories or less from fat, reduc- tions in calories from fat of at least 4-7 percent of total calories will be required, on average, for most Americans. For many individuals with fat intakes that are higher than average, however, an even greater reduction in fat intake will be neces- sary. Saturated Fatty Acids. The major decrease in total fat should be in calories from SFA. Data from the USDA" indicate that SFA, as a propor- tion of calories, average about 13.2 percent in the diets of adults ages 19 to 50 years and 13.9 percent in those of children ages 1 to 5 years. Thus, the average reduction in the percentage of calories from SFA that is required to achieve the recom- mended nutrient intake is at least 3-4 percent. In- dividuals currently deriving higher percentages of calories from SFA will need to make greater re- ductions. The greater the decrease in SFA intake, the greater will be the decline in blood cholesterol levels. Therefore, the reduction in saturated fatty acids in the diet should be as great as possible, consistent with palatability and practicality. This panel recommends a target of less than 10 percent of calories from SFA as both practical and palat- able for Americans. Other fatty acids. The remaining fatty acids, up to 20 percent of total calories, should come from unsaturated fatty acids. PUFA can provide up to, but no more than, 10 percent of total calories. The average intake of omega-6 PUFA in the American diet is about 7 percent of total calories,>*!” which is an acceptable intake. Fish oil supplements gen- erally do not reduce blood cholesterol levels. Omega-3 PUFA, however, reduce blood choles- terol when substituted for SFA. The other major type of unsaturated fatty acids, MUFA, should provide the remaining dietary fat. The current average intake of MUFA for the population is ap- proximately 14-16 percent of calories. A slight re- duction in average MUFA intake can be expected because foods of animal origin containing SFA often also have substantial amounts of MUFA. Other Nutrients. Reducing fat intake will re- duce energy intake unless other nutrients replace fat. Two other major macronutrients, protein and carbohydrates, as well as many micronutrients, are of importance. The panel recommends that the intake of carbohydrates be increased to 50-60 percent of total calories, primarily by increasing the intake of complex carbohydrates. This can be achieved by increasing the intake of vegetables and fruits, and of breads, legumes, and whole grain cereals. These plant products generally contain low levels of fat and are good sources of several vitamins, minerals, and dietary fiber. Protein intake should not be increased to com- pensate for the decrease in calories derived from fat. The recommended intake of protein is be- tween 10 percent and 20 percent of calories. In general, average protein intake by adults in the United States considerably exceeds the recom- mended dietary allowance (RDA), which is 0.8 g/ kg of desirable body weight for adults. Overall, the diet consumed should meet the RDA's for all other nutrients. 19 Population Panel Report b. Body Weight Body weight is influenced by many factors; energy balance between caloric intake and caloric expenditure largely dictates whether excess body fat accumulates. Many health problems are re- lated to obesity. Obese people are more likely to have elevated blood cholesterol and triglyceride levels, a reduced level of HDL-cholesterol, high blood pressure, and adult-onset diabetes melli- tus, and are more likely to develop CHD indepen- dently of these other factors. Loss of excess weight and maintenance of a desirable body weight is an important way of reducing blood cholesterol lev- els. The exact dietary energy (calorie) level needed to attain or maintain desirable body weight must be individually determined from observed changes in body weight. c. Cholesterol Intake Many studies have shown that dietary choles- terol raises the blood cholesterol level. It has been estimated that with a 2,500 calorie diet, for every 100 mg/day decrease in cholesterol intake, the blood cholesterol will decrease by about 4 mg/ dL." This response holds even at low intakes, and thus, the lower the cholesterol intake, the lower will be the blood cholesterol on the aver- age. There appears to be considerable inter-indi- vidual variability in response of blood cholesterol to dietary cholesterol intake. Based on the ob- served effects of dietary cholesterol intake on blood cholesterol levels, the panel concludes, as have many investigators and expert panels, that high dietary cholesterol intake contributes to the de- velopment of atherosclerosis and increased CHD risk in the American population and should be reduced. Data from the USDA" indicated that the average daily intake of dietary cholesterol is 304 mg and 435 mg for women and men, respec- tively. The Population Panel recommends an in- take of less than 300 mg of cholesterol per day. 20 B. Eating Patterns The nutrient intakes recommended in this report are consistent with the Dietary Guidelines for Americans as set forth by the USDA and the DHHS." These guidelines are suggested for most Americans. They do not apply to people who need special diets because of disease or condi- tions that interfere with normal nutrition. Such people may require special instruction from a health professional with expertise in nutritional counseling, such as a registered dietitian, in con- sultation with their own physicians. These USDA / DHHS guidelines'” are: e Eat a variety of foods; * Maintain desirable weight; * Avoid too much fat, saturated fatty acids, and cholesterol; e Eat foods with adequate starch and fiber; e Avoid too much sugar; e Avoid too much sodium; and e If youdrink alcoholic beverages, do so in mod- eration. The first two guidelines form the framework of a good diet—eating a variety of foods to ensure getting the essential nutrients, and eating only enough calories to maintain desirable weight. The next five guidelines describe special characteris- tics of a good diet—avoiding too much fat, satu- rated fatty acids, and cholesterol; getting adequate starch and fiber; and avoiding too much sugar, sodium, and alcohol. This report focuses on the two guidelines that are particularly related to blood cholesterol and the prevention of CHD: Avoid too much fat, saturated fatty acids, and cholesterol; and main- tain desirable weight. Population Panel Report The guidelines also set forth the principle that reductions in energy derived from fat should be offset by increased consumption of foods contain- ing complex carbohydrates such as starch and fiber. The Population Panel recommends that between 50 percent and 60 percent of calories come from carbohydrates. Several valuable documents have been devel- oped to assist people to follow the Dietary Guide- lines for Americans. These documents can also be used to follow the recommendations in this re- port. These include four new booklets to help consumers put the Dietary Guidelines into prac- tice," and Dietary Guidelines and Your Diet,* which includes “A Pattern for Daily Food Choices.” The pattern suggests daily servings from the major food groups.” Valuable information is also con- tained in booklets and brochures available from the National Institutes of Health and the Ameri- can Heart Association. RECOMMENDATION B. 1: The panel recommends that healthy Americans, both adults and children, select, prepare, and con- sume foods that contain lower amounts of saturated fatty acids, total fat, and cholesterol. The following specific practices can help Americans, both adults and children, lower blood cholesterol levels. These food choices and prepa- ration methods form the framework of a good diet. Eating a variety of foods to get essential nutrients and only enough calories to maintain desirable weight is also important. These eating patterns are fully compatible with cultural and ethnic considerations and with personal prefer- ences for good food. * Fruits, vegetables, whole grain products, legumes such as beans and peas. These prod- ucts contain complex carbohydrates, fiber, and minimal amounts of saturated fatty acids. Accordingly, it is recommended that healthy Americans: Eat a greater quantity and variety of fruits, vegetables, breads, cereals, and legumes. These choices will help to meet nutritional needs for minerals, vitamins, dietary fiber (including soluble fiber), and complex carbohydrates, and to replace calories from fat. Low-fat dairy products. Dairy products con- stitute a major source of fat and saturated fatty acids in the average American diet. For this reason, it is recommended that healthy Americans: Eat more low-fat dairy products such as skim or low-fat milk and skim or low-fat milk products. Choose them, in place of regular whole milk, cheese, cream, ice cream, butter, or other butterfat-rich products, to help meet nutritional needs for calcium, protein, and vitamin D (skim milk is an excellent source of calcium with minimal accompanying fat). Low-fat meat, poultry, fish. Red meats, both in surface fat and fat within the meat tissue, provide a substantial proportion of fat and saturated fatty acids for the average Ameri- can. Choosing lean cuts of beef, pork, veal, or lamb, removing surface fat, and substituting fish or skinless poultry can significantly re- duce saturated fatty acid intake. Thus, the panel recommends that healthy Americans: Eat moderate amounts (e.g., about 6 ounces per day, cooked) of trimmed, lean red meat, poultry without skin, or fish in place of choices high in saturated fatty acids. These choices will help meet nutritional needs for protein, iron, and zinc. Cholesterol is found only in products from animals. - Organ meats—Iliver, sweetbreads, kidneys, and brain—are very rich in cholesterol, and consumption should be limited. Eggs. Egg yolk serves as an important source of dietary cholesterol for most Americans. An average large egg yolk contains 213 mg of cholesterol. These facts lead the panel to rec- ommend that healthy Americans: 21 Population Panel Report 22 Eat egg yolks only in moderation. Egg whites do not contain cholesterol and they can be eaten often. Fats and oils. All vegetable products are free of cholesterol. Many oils come from vege- table products. Most contain large propor- tions of monounsaturated and polyunsatu- rated fatty acids and small proportions of satu- rated fatty acids. Tropical oils such as palm kernel oil, coconut oil, and palm oil contain relatively large proportions of saturated fatty acids. Accordingly, the panel recommends that healthy Americans: Use oils, margarines, and shortenings with vegetable oils containing primarily unsatu- rated fatty acids instead of saturated fatty acids. Consumption of tropical oils, especially the most highly saturated palm kernel and coconut oils, should be minimized. Baked goods. Most commercially prepared baked goods contain ingredients composed of saturated oils or fats; many contain egg yolks. For these reasons, the panel recommends that healthy Americans: Choose prepared baked goods that have been made with unsaturated vegetable oils and, at most, small amounts of egg. Breads and most rolls are low in saturated fatty acids and cho- lesterol while other commercial products such as croissants, cakes, biscuits, doughnuts, and muffins are often high in saturated fatty acids. Convenience Foods. Increasingly, Americans are consuming meals, entrees, mixes, and other packaged foods that are conveniently recon- stituted and / or reheated. While many of these products have constituents containing large proportions of saturated fatty acid, many do not. Accordingly, it is recommended that healthy Americans: Choose “convenience foods” guided by low saturated fatty acid, total fat, and choles- terol content as well as by cost considera- tions. Food labels. Increasingly, valuable informa- tion is presented on food labels concerning amounts and sources of saturated fatty acids, fat, cholesterol, and other nutrients. For this reason, the panel recommends that all healthy Americans: Know your food by reading food labels, in- cluding both ingredient and nutrition infor- mation. Choose more often foods with lower amounts or proportions of saturated fatty ac- ids and/or fat and of cholesterol. Saturated fatty acids and cholesterol are often found together in foods. Saturated fatty acids are provided primarily by animal products such as the fat in meat, butter, whole milk, cream, cheese, and ice cream. These foods are also major sources of cholesterol. Food preparation. Preparation of food can significantly increase the proportion of calo- ries from all fat and from saturated fatty ac- ids. For this reason, it is recommended that healthy Americans: Keep use of fats to a minimum when prepar- ing foods. Use the smallest amount possible of fat and fatty foods as ingredients. Cook foods in ways that minimize or eliminate fat content, such as broiling, boiling, steaming, microwaving, or baking, instead of frying in fats; Use smaller amounts of ingredients high in saturated fatty acids. Where fats are neces- sary, substitute fats high in unsaturated fatty acids for fats high in saturated fatty acids; substitute equal or smaller amounts of un- saturated margarine or oil for butter; Use low-fat alternatives in food preparation. For example, substitute 1%, skim, or reconsti- tuted nonfat dry milk for whole milk or higher fat choices. Use low-fat yogurt, buttermilk, or evaporated skim milk in place of sour cream, cream, or cream cheese. Eating out. Americans, on average, derive a significant portion of their nutrition outside of the home. Thus, the panel recommends that healthy Americans: Population Panel Report In restaurants, select menu items that are low in saturated fatty acids, total fat, and choles- terol as well as cooked foods that are baked, boiled, or broiled without fat. Ask for lean cuts of meat trimmed of excess fat. Request that sauces and salad dressings be served in separate containers if high in fat and use them sparingly. Ask for information on ingredients and preparation of restaurant food. When eating out and on other social occasions, choose foods with less fat—e.g., larger amounts of vegetables, fruits, and grain products, and smaller amounts of foods that are likely to be high in saturated fatty acids. - Fast foods. While increasing numbers of “fast food” outlets have choices in which low propor- tions of calories are derived from saturated fatty acids, many others contain large amounts of saturated fatty acids and cho- lesterol. For these reasons, the panel rec- ommends that healthy Americans: Ask for information on the fat and satu- rated fatty acid content of “fast food” selections and choose foods that are nutri- tious and low in saturated fatty acids, total fat, and cholesterol. Overly simple solutions. An habitual pat- tern of eating that is consistently low in satu- rated fatty acids, total fat, and cholesterol is recommended. Focusing solely on the elimi- nation of a single food such as egg yolks, however, is not necessary and will not, by itself, achieve sufficient blood cholesterol lowering. As indicated earlier, skim or low- fat dairy products and small portions of trimmed, lean red meat are desirable and nu- tritionally valuable. These foods and eggs in moderation can all be part of a blood choles- terol-lowering eating pattern. The panel there- fore recommends that all healthy Americans: Recognize that no single food or supplement is the answer to achieving a desirable blood cholesterol level. - Fish oil supplementation will not lower blood cholesterol levels in most people. However, consumption of fish, a food low in saturated fatty acids, instead of foods that are higher in saturated fatty acids, is desirable. - Dietary fiber supplements are not a pana- cea for blood cholesterol problems. Foods rich in soluble dietary fiber are, however, a useful addition to a low saturated fatty acid, low-fat, and low-cholesterol eating pattern. - Alcohol is not recommended as a means of preventing CHD because it has not been shown to be effective and because of the potentially harmful effects of excessive al- cohol intake. RECOMMENDATION B. 2: The panel urges the public to recognize that an elevated blood cholesterol level is one of the important modi- fiable CHD risk factors together with smok- ing, high blood pressure, excess body weight, and physical inactivity. The recommended eating patterns should be considered part of a personal health program which also emphasizes the following risk factors for coronary heart disease. Smoking, high blood pressure. Being a non- smoker, and preventing and controlling high blood pressure are both important factors in prevention of CHD. Excess body weight. It is important for all healthy Americans to achieve and maintain desirable body weight by limiting energy (calorie) intake and by engaging in regular moderate exercise. Maintenance of desirable body weight is a useful means of controlling blood cholesterol. Physical inactivity. Apart from beneficial effects on body weight and on blood pressure, regular exercise throughout life is associated with a lower risk of CHD. 23 Population Panel Report C. Healthy Children and Adolescents RECOMMENDATION C. 1: The panel rec- ommends that healthy children follow the recommended eating patterns that are lower in saturated fatty acids, total fat, and choles- terol as they begin to eat with the family, usually at 2 years of age or older. The Population Panel has reviewed evidence concerning benefits and safety for children and adolescents of an eating pattern that is lower in saturated fatty acids, total fat, and cholesterol than the usual American diet. The panel also recog- nizes that the caloric and nutrient needs of chil- dren are critical for supporting growth and de- velopment. Since eating habits developed during childhood can influence lifetime practices, the panel urges prudent movement to the recom- mended eating pattern. Healthy children should be allowed to share in family food choices, reci- pes, and menus. Adolescents are particularly en- couraged to maintain a low saturated fatty acid, low total fat, and low-cholesterol eating pattern. Infants and children under the age of 2 years have dietary requirements different from those of older people. Infants whose diet is primarily mother’s milk or formula often appropriately consume 40 percent or more of calories from fat. The Population Panel recommends that this well- established pattern of infant nutrition be encour- aged. Care must be taken to ensure sufficient energy and nutrient intake to meet the needs of the growing child. A forthcoming report from the Expert Panel on Blood Cholesterol Levels in Chil- dren and Adolescents will deal with strategies for detection, evaluation, and treatment of children and adolescents at high risk of later CHD because of high blood cholesterol levels and will provide additional details about implementation of strate- gies for encouraging desirable eating patterns for healthy children. 24 D. Special Groups RECOMMENDATION D. 1: The panel recommends that particular attention be paid to nutritional concerns of special segments of the population. The eating patterns of specific groups of Americans should be considered. Population groups with special dietary concerns include women, older Americans, cultural and ethnic groups, low-income groups and individuals with special nutritional needs. e Women. Women, like men, can and should adopt the recommended eating patterns that are lower in saturated fatty acids, total fat, and cholesterol as a means of reducing their likelihood of disease. The special needs of women for nutrients such as calcium, iron, and zinc can be met within these eating pat- terns. e The elderly. For most individuals 65 years of age and older, the recommended eating pat- terns that are lower in saturated fatty acids, total fat, and cholesterol, that help maintain desirable body weight, and that are nutrition- ally adequate, are appropriate. Special atten- tion, possibly including nutrition guidance from a qualified health professional with spe- cial nutrition expertise, may be needed for some elderly people who have special nutri- tional needs or who are limited in their ability to procure, prepare, or consume food. e Cultural groups. Foods and menus should accommodate cultural, regional, and ethnic preferences, while incorporating the recom- mended intakes of saturated fatty acid, total fat, and cholesterol. e Low-income groups. Low-cost foods meet- ing the recommended eating patterns are avail- able; individuals of lesser economic means can and should select these to improve their diets and share in better health. Population Panel Report * People with special nutritional needs. Health professionals should be attentive to individu- als who have special nutritional needs, such as pregnant or lactating women, patients with food allergies, and many patients with diabe- tes mellitus, or chronic kidney, heart, or liver disease, who often need counseling from a health professional with special nutritional ex- pertise. E. Health Professionals RECOMMENDATION E. 1: The panel rec- ommends that health professionals advise patients and the public to attain the recom- mended eating patterns. Physicians, nurses, registered dietitians, nu- trition and health educators, pharmacists, and other health professionals have the opportunity to be important influences on patients and the public concerning the attainment of the panel's recommendations. Health professionals should both practice and advocate the recommended eating patterns. Initial and continuing nutrition education for health professionals should empha- size the panel’s recommendations. Health pro- fessionals, as needed, should obtain assistance from others with special nutrition counseling ex- pertise to facilitate achievement by patients and the public of the recommended eating patterns. In particular, health professionals should: * Ensure that adult patients have their blood cholesterol measured accurately and that the meaning of the results is clearly explained in keeping with ATP guidelines; * Encourage and counsel their patients, as members of the public, to follow eating pat- terns that are low in saturated fatty acids, total fat, and cholesterol but that are nutri- tionally adequate; * Support the use of more effective communica- tion approaches for populations such as low- income, low-education, and minority groups, encouraging them to adopt these eating pat- terns; Provide shopping and food preparation guide- lines that include economical food alterna- tives that are lower in saturated fatty acids, total fat, and cholesterol and that incorporate cultural, ethnic, and personal food preferences; Develop simple dietary self-assessment and self-monitoring techniques to help individu- als evaluate and monitor their own eating patterns; Work with industry, government, and volun- tary groups to improve, expand, and simplify consumer-oriented nutrition information, advertising, and labeling; Encourage health departments, hospitals, and clinics to provide nutrition education and serv- ices consistent with the nutrient intakes and eating patterns recommended by this panel through efforts such as printed materials, videotapes, and nutrition information serv- ices; Encourage all medical and other health pro- fessional schools to incorporate educational components consonant with the eating pat- terns and principles recommended by the panel; Work with health care agencies such as the medical centers of the Department of Veter- ans Affairs, the military health care system, private and public hospitals, local neighbor- hood ambulatory health centers, and other health care delivery organizations, to help patients and staff adopt and follow the eating patterns recommended by the Population Panel; Identify individuals with special nutritional needs or eating problems who may benefit from assessment and intervention by a quali- fied health professional with expertise in nu- trition counseling such as a registered dieti- tian; and Serve as role models for the public by adher- ing to the recommended eating patterns. 25 Population Panel Report FE. The Food Industry What people eat is influenced by many fac- tors. These include both individual factors and the nutritional environment of the Nation. Indi- vidual factors include the variety of cultural heri- tages, economic factors, the availability of food, the taste of food, and knowledge about food. The environment includes a broad array of influences on the food chain. Recognizing the influence of each of these factors on the food that Americans eat, the panel makes the following recommenda- tions concerning the diverse elements in the U.S. food chain. * Encourage food industry and agricultural commodity groups to adopt products, poli- cies, and positions consistent with these rec- ommendations while assuring both continued responsiveness to consumer desires and nec- essary profitability; and * Provide professional education materials, and consumer brochures, cookbooks, and recipes, as well as other print and nonprint educa- tional materials indicating ways of practicing the recommended eating patterns. RECOMMENDATION F. 1: The panel rec- ommends that food producers, manufacturers, and distributors increase the availability of good-tasting foods that are lower in saturated fatty acids, total fat, and cholesterol. RECOMMENDATION F. 2: The panel rec- ommends that the food industry participate actively in helping the public attain desir- able eating patterns through labeling and advertising activities. This can be accomplished through the follow- ing activities. * Expand efforts to produce, manufacture, and market animal products lower in saturated fatty acids, total fat, and cholesterol with par- ticular emphasis on leaner cuts of meat and lower fat dairy products; * Continue efforts by the agriculture industry to produce, manufacture, and market proc- essed plant products containing smaller amounts of saturated fatty acids, and larger amounts of complex carbohydrate and die- tary fiber, including soluble fiber; * Continue development of safe food preserva- tion techniques suitable for products contain- ing unsaturated fatty acids; ¢ Continue development of methods for mass preparation of food containing lower levels of saturated fatty acids, total fat, and cholesterol; * Continue development of foods and menus that will help people attain the recommended intakes of saturated fatty acids, total fat, and cholesterol and that will ensure adequate nu- trient intakes; 26 This can be accomplished through the follow- ing activities. * Promote foods that are low in saturated fatty acids, total fat, and cholesterol, working with the advertising industry and the media to ensure that accurate messages reach large segments of the public; * Provide comprehensible quantitative informa- tion on most food labels concerning saturated fatty acids, total fat, cholesterol, and calories to enable consumers to make informed choices; * Avoid misleading the public through adver- tising or product labeling (e.g., refrain from promoting the “no cholesterol” claim for plant origin products when these products contain relatively large amounts of fat, especially satu- rated fatty acids, and avoid making health claims that are not supported by scientific evi- dence); * Publicize the good taste and economy as well as the health-promoting characteristics of an eating pattern low in saturated fatty acids, total fat, and cholesterol; and * Encourage hard-to-reach segments of the population such as low-income, low-educa- tion, low-literacy, and minority groups to choose food products that are low in satu- rated fatty acids, total fat, and cholesterol. Population Panel Report RECOMMENDATIONF. 3: The panel rec- ommends that food vendors and other food distribution sites participate actively in the national effort. This can be accomplished through the fol- lowing activities. Develop and disseminate point-of-purchase information in individual grocery stores, su- permarket chains, restaurants, and cafeterias to identify foods low in saturated fatty acids, total fat, and cholesterol; Label, promote, and expand the selection of food choices that are lower in saturated fatty acids, total fat, and cholesterol in places pro- viding food to large numbers of people such as institutional food services, restaurants, convenience or “fast food” outlets, schools, worksites, government agencies, hospitals, transportation services, catering services, and other feeding programs; Expand the availability and encourage the selection of foods that are lower in saturated fatty acids, total fat, and cholesterol in gov- ernment-subsidized or -supervised food pro- grams (e.g., school lunch programs); and Continue to develop ethnically and culturally appropriate food selections that are lower in saturated fatty acids, total fat, and cholesterol. RECOMMENDATION F. 4: The panel rec- ommends that the food industry, including food and animal scientists, food technolo- gists, and nutritionists, continue to develop and modify foods to help the public meet the recommended eating patterns. This can be accomplished through the follow- ing activities. Intensify research and development of food products that are lower in saturated fatty ac- ids, total fat, and cholesterol; Continue development of economically, fea- sible uses of food processing by-products— such as meat trimmings—containing saturated fatty acids, fats, and cholesterol, other than as human food; and Assure that as substitutes for food compo- nents such as fat are developed, there is clear evidence of safety for human consumption. G. Mass Media RECOMMENDATION G. 1: The panel rec- ommends that the mass media provide infor- mation on a lower saturated fatty acid, lower total fat, and lower cholesterol eating pattern. Through their news, information, entertain- ment, and advertising policies and programming, the mass media have the ability to influence large numbers of people. As critical channels of com- munication to the public, the mass media are ex- tremely important allies in the effort to reduce blood cholesterol levels throughout the popula- tion. H. Government RECOMMENDATION H. 1: The panel rec- ommends that government facilitate attain- ment of healthful eating patterns by modify- ing policies and approaches. Federal, state, and local governments are re- sponsible for many programs and policies related to nutrition and to health. The following govern- mental activities will help promote cholesterol- lowering eating patterns. Increase collaboration among government agencies—e.g., the USDA, the Federal Trade Commission, the Federal Communications Commission, the U.S. Department of Veter- ans Affairs, and the DHHS including the Food and Drug Administration (FDA), the National Institutes of Health (NIH), and the Centers for 27 Population Panel Report Disease Control—to provide consistent, coor- dinated nutrition statements and policies em- phasizing low saturated fatty acid, low-fat, and low-cholesterol eating patterns; * Help consumers make informed choices through expansion and standardization of food labeling requirements, especially through collaborative efforts of the FDA and USDA with the food industry, to include on most packaged foods easily comprehensible labels that clearly identify, among other nutritional characteristics, the content of saturated fatty acids, total fat, and cholesterol, as well as total calories; ¢ Develop, implement, and educate consumers concerning quantitative definitions for such terms as “lean,” “low-fat,” “low in saturated fatty acids,” and “lite”; * Improve labeling of meat to help consumers select lean cuts; * Increase state and federal government efforts to provide consistent nutrition education and food choice guidance for the public empha- sizing the desirability of foods with low-satu- rated fatty acid, low total-fat and low-choles- terol content; and * Revise food-related policies to stimulate pro- duction and distribution of foods low in satu- rated fatty acids, total fat, and cholesterol (e.g., the recent encouraging movement toward reduction of saturated fatty acid content of school lunch programs). I. Educational Systems RECOMMENDATION I. 1: The panel recom- mends that all public and private educational systems become active partners by dissemi- nating information about the role of eating patterns in CHD prevention. 28 Attainment of lower blood cholesterol levels and reduced risk of CHD will be enhanced if educational programs at all levels—elementary through high school grades, vocational schools, colleges, universities, and medical schools—in- crease nutrition education. Accordingly, the panel recommends a focus on low saturated fatty acid, low-fat, and low-cholesterol eating patterns in the following nutrition education materials and ac- tivities: e¢ Curricula in health education, science, and home economics classes as appropriate for kindergarten through 12th grades; * Curricula of culinary arts education programs emphasizing purchase, preparation, and pres- entation of foods and menus low in saturated fatty acids, total fat, and cholesterol; e Curricula of university food science depart- ments, including consideration of how ad- vances in molecular biology, genetic, and pro- duction techniques can facilitate development of new food products with reduced saturated fatty acid, total fat, and cholesterol content; * Curricula of college and university under- graduate programs, especially biological sci- ence, health, nutrition, and physical educa- tion; and * Curricula of all pre- and postdoctoral physi- cian and other health professional education programs (e.g., for nurses, dietitians, and pharmacists). RECOMMENDATION I. 2: The panel recom- mends that existing nutrition education programs expand and be compatible with the recommended eating patterns, and that mes- sages from major organizations concerning health-promoting eating patterns be consistent. Many organizations already offer nutrition education programs to help Americans reduce cholesterol levels or meet other dietary goals. Examples are programs offered or supported by voluntary organizations (e.g., American Heart Population Panel Report Association, American Cancer Society) or by fed- eral, state, or local government agencies. Educa- tion programs offered through the network of USDA's Extension Service can effectively com- municate nutrition information to targeted audi- ences. Divergent messages can confuse the pub- lic; consistent, coherent information is needed. J. Measurement of Blood Cholesterol RECOMMENDATION J. 1: The panel rec- ommends that public screening for blood cholesterol be undertaken only under condi- tions that ensure adherence to high quality standards. Measurement of blood cholesterol serves to augment awareness of elevated blood cholesterol as a personal health problem. Recognizing the vital importance of cholesterol awareness, the NCEP has promoted a broadly based public edu- cation program—"Know Your Cholesterol Num- ber”—to encourage all adult Americans to know their blood cholesterol level. The ATP recommended that all American adults have their blood cholesterol measured at least once every 5 years. Cholesterol measure- ments should meet the accuracy and precision standards of the Laboratory Standardization Panel. The usual approach to detecting high blood cho- lesterol is through the physician's office. In addi- tion, public screening provides the possibility of detecting large numbers of individuals with high blood cholesterol. However, the reliability of cho- lesterol measurements, the education of partici- pants, and proper referral and followup in public screening programs all need special attention. Only under specific conditions should gen- eral public screening for blood cholesterol be used to supplement screening in the health care set- ting. The NHLBI Workshop Regarding Public Screening for Measuring Blood Cholesterol* rec- ommended the following criteria for public screening. e Use recruitment approaches that attract all adult segments of the community and develop special approaches for harder-to-reach target groups. These include men, younger adults, low-income or low-education groups, and minorities; e Evaluate the performance of each cholesterol analyzer before conducting a public choles- terol screening, and document ongoing inter- nal and external quality control procedures, to ensure that cholesterol measurements are precise and accurate, and that they meet the standards of the Laboratory Standardization Panel of the NCEP; e Ensure that public screening programs are managed by staff trained and supervised by appropriate health professionals and are sup- ported by a qualified health institution and that the quality of all aspects of screening programs is evaluated; e Provide cholesterol screening at a reasonable cost to the participants, with moderation of the cost for certain low-income group partici- pants as necessary; e Provide reliable verbal and printed educa- tional information about cholesterol levels from knowledgeable staff, so that individuals understand the meaning and limitations of a single cholesterol measurement, receive in- formation on dietary practices to lower cho- lesterol, and are given clear instructions with respect to their own medical followup; e Recommend referrals on the basis of the NCEP guidelines shown in table 4, page 30; and * Emphasize that screening is not a substitute for health care and cannot replace medical monitoring of blood cholesterol levels of indi- viduals already under treatment. Recommendations for screening children and adolescents and for managing younger people with elevated cholesterol levels will emerge later from the NCEP’s Expert Panel on Blood Choles- terol Levels in Children and Adolescents. 29 Population Panel Report Table 4 Guidelines for Referral of People Based on Their Blood Cholesterol Level Blood Total Cholesterol Level Recommended Action Greater than or equal to 240 Refer to physician for followup.” mg/dL (High) Between 200 and 239 mg/dL (Borderline-High) Refer to physician for followup” if history of coronary heart disease (CHD) or if two or more other CHD risk factors (excluding HDL-cholesterol) detected on interview (see table 2, page 9). If no reported history of CHD or less than two other risk factors, refer to physician within 1 year for repeat cholesterol measurement. Recommend a repeat blood cholesterol followup in 5 years. Less than 200 mg/dL (Desirable) Individuals should be seen by their physicians within 2 months. Source: NCEP ATP, January 1988° K. Research and ¢ Improve analytical methods for food compo- . sition and continue development and applica- Surveillance tion of the National Nutrient Databank of commodity and commercial products, includ- ing brand name information as appropriate to provide reliable information about nutrient composition of foods consumed by Ameri- cans; RECOMMENDATION K. 1: The panel recommends that research and evaluation on the relationships between food and health continue and that the panel’s recommenda- tions be reconsidered in a timely manner as new scientific information becomes available. ¢ Collect and evaluate data, individual and so- cietal, pertaining to dietary change along with evaluations of effects on CHD and other health status indicators. These data should include information on food production, processing, and consumption patterns and costs; popula- The following specific activities should be undertaken: * Continue efforts to clarify even further the 30 role and the interactions of dietary compo- nents and nutrients, and blood lipids, athero- sclerosis, and CHD; tion levels of blood cholesterol; and trends in cause-specific morbidity and mortality. In- formation concerning the efficacy of reaching all segments of the population should be Population Panel Report collected at the national, state, regional, and local levels. Many of these databases are on- going through the work of government agen- cies—e.g., NHANES surveys; National Cen- ter for Health Statistics (NCHS) mortality sta- tistics; FDA /NHLBI consumer and NHLBI health professional surveys; FDA total diet surveys; USDA nationwide food consump- tion surveys; state-based behavioral risk fac- tor surveys—and should be continued or im- proved, while other data needs will require development of new monitoring systems; Expand research and development related to the reduction or alteration of fat content in foods of animal origin as recommended in the National Research Council report, Designing Foods; and Continue to identify, develop, and evaluate effective nutrition education and communica- tion methods. 31 Population Panel Report V. Scientific Evidence for Recommendations Affecting the General Public A. Diet Patterns, Blood Cholesterol, and Health 1. Introduction This review examines the scientific literature concerning the etiologic factors for increased blood cholesterol and CHD. It presents the diversity of observations and interpretations that link diet, blood cholesterol, and CHD. The preponderance of the data in the literature indicate that habitual diet directly influences the level of blood choles- terol, and that in turn, elevated levels of blood cho- lesterol result in high rates of CHD. 2. Clinical, Laboratory, and Metabolic Research a. Lipids and Lipoproteins Lipids circulate in blood plasma in association with certain specific proteins (apolipoproteins) as large, macromolecular complexes called plasma lipoproteins. The lipoproteins are classified as chylomicrons, very low density lipoprotein (VLDL), low density lipoprotein (LDL), and high density lipoprotein (HDL). Lipoproteins are lighter than the other plasma proteins because of their high lipid content. In turn, their weight or density varies from high to very low largely as a function of their relative contents of protein and various lipids. The primary function of plasma lipoproteins is lipid transport. The major lipid transported in lipoproteins—triglyceride—is insoluble, yet up to several hundred grams must be transported through the blood daily. Cholesterol is the other major lipid transported in lipoproteins. It is not used for energy production, as are triglycerides, but is the precursor of steroid hormones and bile acids and serves as an important structural com- ponent of cellular membranes. In higher animals, it is transported in plasma predominantly as cholesteryl esters (CE). The transport of choles- terol in lipoproteins permits specific targeting of cholesterol to tissues that require it for structural purposes or to make metabolic products. Two of the lipoprotein classes, chylomicrons and VLDL, contain large quantities of triglyceride and represent, respectively, the transport form of exogenous (dietary) and endogenous triglyceride. Chylomicrons normally are not present in post- absorptive plasma after an overnight fast. The VLDL normally contains 10-15 percent of the total plasma cholesterol. LDL contains cholesterol as its major component and normally contains most (60-70 percent) of the plasma cholesterol. HDL is comprised of approximately half protein and half lipid, and usually contains 20-30 percent of the total plasma cholesterol. Each lipoprotein class is heterogeneous with respect to its protein constituents. At least nine distinct apolipoproteins have been separated and described. Most investigators group the apolipo- proteins into five families (designated apoA, apoB, apoC, apoD, and apoE) on the basis of their 33 Population Panel Report chemical, immunologic, and metabolic character- istics”? In addition, a distinct LDL-related par- ticle called lipoprotein (a) consists of an unusual apolipoprotein, designated apo (a), attached to an LDL particle. Elevated levels of lipoprotein (a) have been associated with an increased risk of ath- erosclerotic disease.” b. Atherogenesis The strong relationship between high levels of plasma LDL-cholesterol and atherosclerosis in man has stimulated in vitro studies dealing with the cellular and molecular mechanisms of atherogene- sis. These studies have sought to determine how LDL-cholesterol and other “atherogenic” lipopro- teins could promote the formation of foam cells and atherosclerotic lesions. The results of such in vitro studies have, in turn, supported the human and animal data that suggest a fundamental etio- logic role for LDL and related lipoproteins in the cause of atherosclerosis. A prominent and early feature of atheroscle- rotic lesions is the CE-filled macrophage, or foam cell. To understand the mechanism(s) of foam cell formation, investigators have studied the inter- actions of plasma lipoproteins with various types of cultured macrophages. Native LDL does not lead to CE accumulation in most types of macro- phages.” This is due mainly to the fact that pro- duction of the LDL receptor on these cells is down- regulated by small amounts of excess cellular cho- lesterol, thus preventing a large influx of LDL- cholesterol.” However, various modified forms of LDL suchasacetyl-LDL, oxidized LDL, and malon- dialdehyde-LDL do lead to marked accumulation of CE in cultured macrophages. These modified forms of LDL enter the cell by a receptor (called the “scavenger” receptor) that is distinct from the LDL receptor and that is not subject to down-regula- tion”? Thus, cellular influx of these modified forms of LDL continues over time at a high rate, leading to significant CE accumulation. These in vitro observations have spawned a widespread search for evidence of LDL modifica- tion in vivo. Several investigators have reported 34 the occurrence of various kinds of modified LDL in atherosclerotic arteries. A potential role for native LDL in foam cell formation has also been suggested by the finding that a particular type of cultured macrophage has a poorly down-regu- lated LDL receptor and accumulates large amounts of CE in the presence of native LDL.» Foam cell formation has also been demon- strated in vitro by lipoproteins other than modi- fied and native forms of LDL. One such lipopro- tein, called B-VLDL, is a cholesterol-enriched lipo- protein found in the genetic disease, familial dysbetalipoproteinemia. B-VLDL causes massive deposition of cholesteryl ester in macrophages in vitro.” The receptor by which 3-VLDL enters the macrophage demonstrates poor down-regulation.” Chylomicron remnants, which may occur in humans, appear to interact with macrophages ina manner similar to that of B-VLDL and result in some degree of CE accumulation, as well as a large amount of triglyceride accumulation.’ These and other data have led some investigators to suggest that some postprandial lipoproteins, particularly cholesterol-enriched lipoprotein particles and remnants, may be atherogenic in man. Research onlipid-rich monocytes, or foam cells, represents an active area of in vitro investigation exploring the relationship between LDL-choles- terol and atherogenesis. However, other ap- proaches to the exploration of this relationship have also been attempted.” For instance, LDL, either native or oxidized, has been demonstrated to cause endothelial cell injury in vitro. Endo- thelial injury, in turn, has been theorized to be a potentially important initiating event in athero- genesis, leading to platelet adherence and release of growth factors. In addition, LDL-cholesterol has been reported to directly augment platelet ag- gregation and stimulate the growth of smooth muscle cells, two important features of the athero- sclerotic lesion. It is clear, moreover, that a variety of cellular and vessel wall factors, not yet well- defined, are also important in atherogenesis. Ross* has reviewed the potential roles of growth factors, of endothelial injury, and of the different arterial wall cells in the pathogenesis of atherosclerosis. Population Panel Report In summary, there have been numerous in vitro studies demonstrating the ability of LDL and related lipoproteins to cause the formation of foam cells and other elements of the atherosclerotic le- sion. Although other metabolic, cellular, and vessel wall factors important in atherogenesis exist and remain to be better defined, the in vitro data cur- rently available strongly suggest an important causative role for LDL-cholesterol and related lipo- proteins in atherogenesis. c. Experimental Studies in Animals Studies in laboratory animals have clearly shown that atherosclerosis can be induced by diet. The first studies of the relationship of diet to ath- erosclerosis were initiated in the early 1900's. It was not until the 1950's that investigators began to study in animals the effects of fat and different fatty acids.” * Atherosclerosis was first produced experimentally in rabbits by feeding pure choles- terol in 1913."! Certain species—i.e., rabbits, a number of avian varieties, pigs, and nonhuman primates—are especially susceptible to severe hypercholesterolemia induced by dietary cholesterol.*>*” When hypercholesterolemia is maintained for long periods, these animals de- velop advanced atherosclerotic lesions, similar to those found in humans, and eventually myocar- dial infarction occurs.®** The close link between dietary cholesterol and atherosclerosis in various animal species,” including nonhuman primates, underlies one argument for the importance of dietary cholesterol as an atherogenic factor in hu- mans. Variability exists among different animals in their susceptibility to atherosclerosis induced by dietary cholesterol. Some species such as rats and dogs do not develop significant hypercholesterol- emia or atherosclerosis when fed large amounts of dietary cholesterol.”’*? Although these animals absorb a high proportion of excess dietary choles- terol, they can rapidly degrade this excess into bile acids, thereby protecting themselves from developing hypercholesterolemia and athero- sclerosis.” These species can develop hyper- cholesterolemia when their normal ability to dis- pose of excess dietary cholesterol is blocked by induction of hypothyroidism.* Jokinen et al.*> have recently reviewed the ad- vantages and disadvantages of various primates as models for dietary-cholesterol-induced athero- sclerosis. The amount of cholesterol required to induce atherosclerosis is variable between differ- ent species. The administration of dietary choles- terol to susceptible species produces changes in plasma lipoproteins that are similar, or at least analogous, to those associated with enhanced atherogenesis in humans. In nonhuman primates, high-cholesterol intakes cause LDL-cholesterol to rise to levels that rapidly induce atherosclerosis; furthermore, in the same animals, low levels of HDL are inversely related to atherosclerosis.” In fact, approximately 50 percent of the variance in atherosclerosis in primates can be explained by differences in levels of LDL and HDL.*** More- over, dietary cholesterol modifies the types of LDL particles present in plasma of some types; for ex- ample, Rudel et al.” reported that feeding excess cholesterol to primates induces formation of cho- lesterol-enriched LDL particles that are highly atherogenic. Dietary cholesterol theoretically could increase LDL-cholesterol levels in laboratory animals by multiple mechanisms, e.g., increased production of cholesterol-enriched lipoproteins and inhibi- tion of clearance of LDL from the circulation. Studies in tissue culture indicate that increasing the cellular content of cholesterol suppresses LDL receptor synthesis and activity.” It has been shown in many species that raising dietary cholesterol will cause an increase in hepatic cholesterol con- tent. Spady and Dietschy® demonstrated that high cholesterol intakes suppress LDL-receptor- mediated uptake of LDL by hamster liver. Thus, suppression of LDL-receptor activity is probably a major mechanism whereby dietary cholesterol raises LDL-cholesterol levels in plasma. Different animal species develop different pat- terns of hyperlipidemia upon feeding of high- cholesterol diets. Forexample, rabbits show greater increases in VLDL-cholesterol than in LDL-choles- terol levels. Some primates, in contrast, develop abnormally large, cholesterol-rich LDL in prefer- ence to increased VLDL-cholesterol.” In contrast to both of these species, humans normally show increases of blood cholesterol almost exclusively 35 Population Panel Report in normal-sized LDL particles.” The underlying mechanism of hyperlipidemia, i.e., reduced LDL- receptor activity, may nonetheless be similar in all these species; only the pattern of response in plasma lipoproteins may be species specific. Still another mechanism for hypercholesterolemia in different species could be replacement of triglycerides by cholesteryl ester in newly secreted lipoproteins. This effect has been noted in lipoproteins secreted by perfused livers of cholesterol-fed animals,” and it could contribute to the cholesterol-rich VLDL (B-VLDL) noted in rabbits and hypothyroid dogs. Thus excess dietary cholesterol may raise the plasma cholesterol level by more than one mechanism. One theory holds that postprandial lipopro- teins play an important role in atherosclerosis.” According to this view, dietary cholesterol trans- ported in chylomicrons and chylomicron remnants is directly atherogenic independent of its effects on fasting levels of LDL-cholesterol. Research in labo- ratory animals provides a unique opportunity to evaluate this hypothesis. Only a few studies have actually examined postprandial lipoproteins in laboratory animals.***> Studies in primates have shown that cholesterol-rich, postprandial lipopro- teins, resulting from high intakes of dietary choles- terol, may produce atherosclerosis. Future re- search in various animal species may provide new insights into the role of cholesterol-rich, postpran- dial lipoproteins in atherogenesis. Not only will the feeding of high levels of die- tary cholesterol increase the plasma cholesterol level and induce atherosclerosis in many animals, but removal of cholesterol from the diet will lower plasma cholesterol levels and cause regression of atherosclerotic lesions.” This finding adds fur- ther support for a role of dietary cholesterol in the genesis of atherosclerosis. Nonhuman primates, as well as other animal species, are affected by the types of fatty acids in the diet, especially during feeding of high-choles- terol diets.” In primates, when the diet is rich in cholesterol, saturated fatty acids raise blood cho- lesterol levels, whereas unsaturated fatty acids lower them.”"”* In baboons, selective breeding can influence the response to saturated fatty acids.” In 36 some studies polyunsaturated fatty acidsappeared to lower cholesterol levels more than monoun- saturated fatty acids, but the studies by Spady and Dietschy” in hamsters suggested that in this spe- cies the two types of unsaturates have about the same effect on LDL-cholesterol levels. Laboratory animals should be good models for investigations of mechanisms whereby SFA raise the cholesterol level. Studies by Spady and Dietschy®”® showed in hamsters fed with choles- terol that SFA suppress hepatic LDL-receptor ac- tivity. Fox et al.”* demonstrated further that SFA reduce levels of messenger RNA for synthesis of LDL receptors. These studies point to a suppres- sion of LDL receptors as a primary mechanism whereby SFA raise LDL-cholesterol levels in labo- ratory animals. d. Human Studies Metabolic ward studies and other research with carefully controlled diets provide important insights into the effects of individual dietary com- ponents on the plasma lipids and lipoproteins in humans. These investigations have lent support to other work indicating that at least three factors raise the blood cholesterol level: an increased intake of SFA, an increased intake of cholesterol, and the presence of obesity. Among primate spe- cies, humans appear to be less susceptible to the hypercholesterolemic effects of dietary cholesterol than most other primates (see section c¢, Experi- mental Studies in Animals). In contrast, humans seem to be more sensitive to the influence of differ- ent dietary fatty acids than are most species of nonhuman primates. The following summarizes the effects of different nutrients on lipoprotein levels as revealed by studies on the metabolic ward or under tightly controlled experimental conditions. The major action of SFA is to raise the LDL- cholesterol concentration.””* Several different types of SFA are present in the diet, and these produce different responses in LDL-cholesterol levels. Dietary SFA generally range from 4 to 18 carbon atoms (C4, C6, C8, C10, C12, C14, C16, and C18). Several investigators®* have reported that Population Panel Report the C6-C10 SFA do not raise the blood total choles- terol level. The major cholesterol-raising SFA are the C14 and C16 acids.”*”* Hegsted et al.” reported that lauric acid (C12) is less hypercholesterolemic than myristic acid (C14) and palmiticacid (C16). In contrast, Keys et al.* reported that C12-C16 SFA are similar in their cholesterol-raising properties. Finally, it was postulated more than two decades ago that stearic acid (C18) does not raise blood cho- lesterol levels; 75% recent studies support this claim and have further shown that stearic acid, at least when occupying the 1 and 3 positions of the triglyc- eride molecule, does not increase LDL-cholesterol concentrations.* Moreover, since hydrogenation of vegetable oils that are rich in oleic acid and linoleic acid produces stearic acid as one of its end products, many oils subjected to this process may not yield a product that will raise blood cholesterol levels. Although further research is needed to identify the particular effects of individual fatty acid constituents, as well as their effects when combined in actual dietary settings, the fact that most SFA in the American diet raise blood choles- terol level is firmly established. How then do SFA raise LDL-cholesterol lev- els? The best available answer, based on animal studies, is that high intakes of SFA suppress the synthesis of LDL receptors®” leading to increased LDL levels. The final word on this question, however, has yet to be written; SFA may have more than one action on cholesterol and lipo- protein metabolism. The major PUFA in the diet is linoleic acid, which is an omega-6 fatty acid. For many years linoleic acid held a special place among other nutrients for its putative “cholesterol-lowering” action.”778891 A favorite question was, “How do polyunsaturated fatty acids lower the cholesterol level?” A variety of mechanisms have been pro- posed, e.g., increasing the excretion of cholesterol and bile acids from the body,” redistributing cholesterol from the plasma into tissues,” alter- ing the structure of lipoproteins so that they can hold less cholesterol,” increasing cholesteryl ester polyunsaturates, and other mechanisms.” More recent work suggests that when PUFA replaces SFA in the diet, the major action of linoleic acid is to increase clearance of LDL from plasma through an increased activity of LDL receptors. Although the latter mechanism is probably a major action of linoleic acid, this fatty acid may have multiple effects on lipoprotein metabolism, as suggested by previous studies. When linoleic acid replaces SFA in the diet, the major response is a lowering of LDL-cholesterol levels. Keys et al.” and Hegsted et al.”® pro- posed that linoleic acid has a unique action to reduce total cholesterol more than other nutrients (i.e., MUFA and carbohydrates); it has since been widely assumed that this effect extends to LDL- cholesterol levels as well. A recent study® showed that replacement of SFA by either linoleic acid or oleic acid caused significant reductions in total cholesterol and LDL-cholesterol levels; no signifi- cant differences were seen between linoleic and oleic acids. Similar results have been found by Mensink and Katan.'”? High intakes of linoleic acid in some people can reduce levels of VLDL-triglyceride, VLDL- cholesterol, and HDL-cholesterol.?>19110%104 These reductions are greater than those seen for mono- unsaturates® and thus may partly explain enhanced total-cholesterol lowering by linoleic acid com- pared with oleic acid, as reported by Keys et al.” and Hegsted et al.”® On the other hand, the lower- ing of VLDL and HDL is pronounced only at high intakes of linoleic acid; at the more moderate in- takes of typical diets, a significant reduction in VLDL and HDL likewise has been difficult to detect. 105100 37 Population Panel Report The omega-3 PUFA, eicosapentaenoic acid (EPA) (20:5, w —3) and docosahexaenoicacid (DHA) (22:6, ® — 3), are found mainly in fish oils. These fatty acids have unique effects on lipid metabo- lism. Their most striking action is to lower plasma triglyceride.'””'" This action is a result of their inhibition of the production of VLDL-triglyceride. Seemingly, omega-3 polyunsaturates act in the liver to prevent the incorporation of other fatty acids into triglycerides. As a result, total secretion of VLDL-triglyceride is reduced.’ The omega-3 PUFA probably do not have an independent effect by which they lower LDL lev- els. Like other unsaturates, when omega-3 PUFA replace SFA in the diet, the LDL-cholesterol level will fall."'" This response probably is, at least in part, the result of removing SFA from the diet. Supplementation of the diet with omega-3 poly- unsaturates, without altering the intake of satu- rated fatty acids, does not cause a lowering of the LDL-cholesterol level."*' In hypertriglyceridemic patients who are treated with supplements of omega-3 fatty acids, the decline in plasma triglyc- erides frequently is accompanied by an increase in LDL-cholesterol concentration.'"’ The actions of omega-3 PUFA on HDL-choles- terol levels are variable. The decrease in triglyc- eride levels in hypertriglyceridemic patients can be associated with an increase in HDL-cholesterol levels. In the absence of hypertriglyceridemia, however, replacement of SFA by omega-3 PUFA can cause a fall in HDL-cholesterol levels''” as also occurs when large amounts of omega-6 fatty acids are substituted for saturated fatty acids. Finally, supplementation of the diet with fish oils without other changes in the diet usually has little or no effect on HDL concentrations. The major fatty acid in most diets is oleic acid—a MUFA. In metabolic ward studies by Keys et al.* and Hegsted et al.,”* monounsaturates were found to be “neutral” in their action on plasma total cholesterol levels, i.e., monounsatu- rates neither raised nor lowered the cholesterol levels. This conclusion was based on responses relative to carbohydrates, which likewise were reported to be “neutral” with respect to total cho- lesterol levels. These general conclusions have 38 been confirmed by Mattson and Grundy.® Al- though the latter workers questioned whether linoleic acid has more LDL-cholesterol lowering action than oleic acid, they obtained results similar to those of previous investigators for oleic acid. In contrast to linoleic acid, high intakes of oleic acid do not reduce concentrations of HDL-cholesterol when it is substituted for SFA.5-82114 Qverall, therefore, oleic acid appears to have favorable effects on lipoprotein levels, especially when sub- stituted for SFA in the diet. When carbohydrates replace SFA in the diet, they produce a lowering in total cholesterol’ and LDL-cholesterol.®'#*!!* These reductions in choles- terol levels are essentially equivalent to those of monounsaturates. In short-term studies, triglyc- erides rise more in response to carbohydrate in- take than to intakes of SFA and MUFA. This reflects the “hypertriglyceridemic” response to high-carbohydrate diets reported by Ahrens et al.'® and Knittle and Ahrens." Limited evidence suggests that the elevation in triglycerides in re- sponse to high-carbohydrate diets does not persist for long periods,’ although this long-term mitiga- tion of increased triglyceride levels has by no means been confirmed. High-carbohydrate diets will reduce HDL-cholesterol concentrations in short-term studies,” **'"* and the evidence avail- able suggests that lower HDL levels are a sus- tained effect in populations consuming high car- bohydrate diets."® The HDL-lowering effect of high carbohydrate diets extends to apoprotein A-I levels, which also are reduced by high carbo- hydrate intakes." In spite of the differences among the various nutrients—MUFA, PUFA, and carbohydrates— on the metabolism of lipoproteins, the differences are not particularly great. Both carbohydrates and polyunsaturates lower LDL-cholesterol levels when substituted for saturates, and at high intakes they lower HDL-cholesterol levels, although the latter response usually is relatively small and may not be of clinical significance. Carbohydrates raise triglyceride concentrations at least in the short term. MUFA may have a favorable effect on all serum lipoproteins, but the most favorable mix of the nutrients (linoleic acid, oleic acid, and Population Panel Report carbohydrates) that should be substituted for SFA in the most desirable diet remains to be clarified. It is likely that a considerable range of choices will produce similar effects on plasma lipid and lipo- protein levels. On the basis of metabolic ward studies alone, the desirable percentage of total fat in the diet cannot be ascertained. In the metabolic ward, a diet moderately high in fat but low in SFA will induce as much lowering of cholesterol levels as one that is low in total fat and SFA and high in car- bohydrates. In general, however, lower intakes of total fat will help to reduce the intake of SFA and also facilitate maintenance of desirable body weight. The key to reducing the LDL-cholesterol level is to decrease intakes of SFA; little is gained by reducing unsaturated fatty acid intake to bring total fat intake to very low levels? #25411 A second major cholesterol-raising constituent of the diet is cholesterol itself. Some studies done on outpatients have failed to show that increasing dietary cholesterol raised the level of plasma cho- lesterol;'212* but when investigations are done on the metabolic ward or in tightly supervised set- tings, where all nutrients can be carefully con- trolled, increasing dietary cholesterol raises blood total cholesterol and LDL-cholesterol concentra- tions.*!7879125130 Responses of individuals to high cholesterol intakes appear to be variable.””!!'* However, this variability is not always reproduc- ible when multiple studies are carried out in indi- viduals. Therefore the response may be more uniform than revealed by a single study in a group of individuals. Certainly on the average, dietary cholesterol intake raises the blood choles- terol level as concluded in the NIH workshop summarized by Grundy et al." and in the NRC report Diet and Health: Implications for Reducing Chronic Disease Risk.” The rise occurs mainly in LDL-cholesterol, but VLDL-cholesterol and HDL- cholesterol levels also can increase." The third cholesterol-raising factor is excess caloric intake resulting in obesity. The effects of obesity on blood cholesterol and lipoprotein lev- els, however, are complex.” Several epidemio- logic studies support a positive relationship be- tween body weight and total cholesterol levels; still, this is not a universal finding and confound- ing variables cannot always be excluded. In a supervised setting, Anderson et al.’ noted that weight gain generally causes a rise in total choles- terol concentration. Conversely, weight loss in obese individuals has been reported to lower total cholesterol levels." Obesity has been associ- ated with increased production of low density lipoprotein.'* More recently, Davis et al." showed that weight reduction in hypercholesterolemic obese patients caused a significant reduction in LDL-cholesterol levels. Wolf and Grundy'** were unable to demonstrate that weight loss in obese, normocholesterolemic individuals produced a de- crease in LDL-cholesterol concentrations, but they did show a decrease in VLDL-triglyceride and VLDL-cholesterol levels and an increase in HDL- cholesterol concentrations. In this latter study," the diet composition was unchanged before and after weight reduction, which may have mini- mized any change in LDL-cholesterol concentra- tions. For example, Caggiula etal.,'* in the MRFIT, demonstrated a close correlation between changes in body weight and responses in blood cholesterol levels to changes in diet composition. Those indi- viduals who lost weight in addition to reducing intakes of SFA and dietary cholesterol had greater reductions in LDL-cholesterol concentrations than those who merely changed diet composition with- out weight loss. The possible relationships between dietary fiber and certain chronic diseases has stimulated much interest in the last decade. Dietary fiber is a generic term to describe substances that are not metabo- lized by human digestive secretions. Dietary fiber can be categorized as soluble or insoluble. The soluble fibers are pectin and gums. Soluble fibers often retard gastric emptying, possibly because of their gelling properties. Insoluble fibers—e.g., wheat fiber—have a high water-holding capacity and exert laxative effects. In general, insoluble fibers have no effect on blood cholesterol or lipoprotein levels. Oat bran exhibits hypocholesterolemic properties due to its appreciable content of oat gum." Soluble fibers such as pectin, guar gum, locust bean gum, or psyllium in large quantity supplementation have 39 Population Panel Report been shown to lower total and LDL-cholesterol levels.'* 148 The absolute effect on LDL-cholesterol concentrations is modest even when the amount of soluble fiber such as oat bran is consumed in ap- preciable amounts (60 gm)."* This effect, how- ever, represents a useful adjunct to an eating pat- tern low in SFA and cholesterol. In summary, research in humans has made valuable contributions to the understanding of the effects of diet on the metabolism of cholesterol and lipoproteins, and lends support to many of the public healthrecommendations for dietary change. In turn, the lipid responses to many recommended dietary changes have been characterized; how- ever, many important diet-lipid questions have not yet been addressed, much less answered. 3. Epidemiologic Research a. Blood Lipids and CHD Blood Cholesterol and CHD. Substantial epi- demiologic evidence has accumulated over the past 40 years indicating that blood levels of total cholesterol and LDL-cholesterol are highly corre- lated with severity of coronary atherosclerosis and rates of CHD. Three types of epidemiologic investigations support these relationships and provide strong support for a causal connection. These are (a) within-population studies, (b) cross- cultural studies, and (c) migration studies. Each type is reviewed in the following section. Within-Population Studies. The most extensive and long-term study of this type has been the Framingham Heart Study. This project began in 1948 and continues to the present time, over 40 years. At the outset, 5,209 men and women living in Framingham, Massachusetts, entered the study. Individuals have been examined every 2 years and have been evaluated fora host of factors that might be related to subsequent development of disease. Over the years, data emerging from this project have demonstrated that age, male sex, elevated blood cholesterol level, high blood pressure, ciga- rette smoking, obesity, and diabetes mellitus are characteristics that increase the probability for 40 heart attack and stroke.” These characteristics, among others, generally have become known as risk factors.! The Framingham study has firmly established high blood cholesterol as a major risk factor for all clinical manifestations of CHD. High blood cho- lesterol also was found to be a significant predictor of intermittent claudication, although not of stroke or transient cerebral ischemic attacks. When the predictive power of total blood cholesterol is ex- pressed as a risk ratio (the CHD rate at high choles- terol levels divided by the CHD rate at lower cho- lesterol levels), a high blood cholesterol was a better predictor of CHD in younger men than in older men." However, the number of CHD cases that are attributable to elevated blood cholesterol (attributable risk = the CHD rate at higher choles- terol levels minus the CHD rate at lower choles- terol levels) does not decrease with age.” In other words, high blood cholesterol continues to be an important risk factor throughout life. Increased levels of blood cholesterol are very common in the elderly.” The Framingham Heart Study has shown that elevated blood cholesterol is also an inde- pendent risk factor in women. In later years and in numerous publications, the Framingham study has extended its analysis of the role not only of total cholesterol but also of cho- lesterol lipoprotein subfractions and triglyc- erides."™" LDL-cholesterol levels have been shown to be positively and strongly related to CHD rate, even more so than total cholesterol con- centrations. Furthermore, low HDL-cholesterol concentrations are powerful predictors of CHD.” More recently, total cholesterol/ HDL-cholesterol (and LDL-cholesterol/HDL-cholesterol) ratios were also shown to be useful predictors, indicating important interactions between the lipid risk fac- tors. The predictive value of HDL-cholesterol has been affirmed in many other large studies, includ- ing the Lipid Research Clinic's Coronary Primary Prevention Trial (LRC-CPPT) and the MRFIT experience. Although blood triglycerides were positively correlated with CHD rates, earlier data did not reveal an “independent” relationship be- tween triglycerides and CHD." Recently, how- ever, the triglyceride level has emerged as an Population Panel Report independent predictor, especially in women, at Framingham.'”” The relationship between blood triglyceride level and CHD risk will be considered in more detail later in this section. One of the more important findings of the Framingham Heart Study is the fact that blood cholesterol level remained a powerful predictor of CHD over many years of followup." Thirty-year followup data from this study have recently shown not only that incremental cholesterol levels are increasingly predictive of CHD, but they also are correlated with total mortality." For example, participants who had total cholesterol levels be- low 180 mg/dL (4.65 mmol/L) at entry had the lowest 30-year total mortality. This trend was maintained with progressively higher cholesterol levels. This finding has important implications. Most importantly, a lifelong low level of choles- terol is associated not only with a reduced rate of CHD but also with an increased life expectancy. The issue of possible risks of very low blood cho- lesterol levels is considered subsequently in this report. Another major finding of the Framingham study is the multiplicative interaction of risk fac- tors. In other words, the occurrence of several risk factors together produces a greater risk than the sum of individual risks.” This means that any given level of blood cholesterol assumes increased significance in the presence of other risk factors. Thus, in the presence of smoking, hypertension, or low HDL-cholesterol, even a near average blood total cholesterol concentration is associated with relatively high risk of CHD. It must be noted that CHD was still the number one killer in the Fram- ingham study in that portion of the population with relatively low cholesterol levels, a fact reflect- ing, in no small part, the coexistence of other risk factors. Although the Framingham Heart Study is the best known within-country study carried out in the United States, similar results have been ob- served with many other studies from this country. The 9-year CHD rates in the Honolulu Heart Pro- gram have also shown the same trend related to blood cholesterol level.'"!*" The CHD risk model and data derived from Framingham have been compared with those of the First National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. Using blood cholesterol, sys- tolic blood pressure, and cigarette smoking, the Framingham predictive model was found to be generalizable to this randomized sample of U.S. citizens.” The data from several of these surveys were combined in the pooling project.”' The pool- ing project data also revealed a strong relationship between total blood cholesterol and CHD; again the correlation was magnified by the presence of other risk factors. However, these studies con- ducted over a relatively few years did not fully re- solve the issue of whether the relationship between cholesterol levels and CHD risk is con- tinuous over a broad range of cholesterol levels from low to high, or whether there is a threshold level, in the range of 200-220 mg/dL (5.17-5.69 mmol/L), only above which does risk for CHD rise appreciably. This latter issue was partially resolved by the 30-year followup of the Framingham study, which showed that individuals with cholesterol levels below 180 mg/dL (4.65 mmol/L) had lowest rates of CHD and total mortality.” Additional evi- dence on this point has been provided from those men who were screened but not enrolled in the Multiple Risk Factor Intervention Trial (MRFIT). In the followup of this population sample of more than one-third of a million men, ages 35 to 57 years, for whom blood cholesterol measurements were obtained in the early 1970's, there was a strong, positive relationship between total blood choles- terol level and CHD death."*'®> The increase in CHD mortality was continuous from low choles- terol levels to high, demonstrating that no thresh- old relation exists between cholesterol levels and CHD risk. This finding has been confirmed in a 10-year followup of the same population’ (figure 1, 41 Population Panel Report page?7). Therelation between cholesterol level and CHD mortality is curvilinear; risk is progressively greater at higher cholesterol levels. More direct evidence that high blood choles- terol is atherogenic comes from angiographic and autopsy studies in which coronary arteries have been examined. Several investigations have shown that total cholesterol and LDL-cholesterol levels are correlated directly with severity of coronary atherosclerosis as revealed by coronary angio- graphy.'®>® The same relationship holds for aor- ticand coronary atherosclerosis in prospective au- topsy studies.” These reports provide strong evidence that high total and LDL-cholesterol lev- els are directly atherogenic in humans. The link between blood cholesterol and CHD observed in the United States has been confirmed by epidemiologic studies carried out in other coun- tries."”!"'”> The connection is a general relationship, whether occurring in so-called “high-risk” or “low- risk” populations. There appear to be no subpopu- lations that are immune to the CHD risk-enhanc- ing effect of high cholesterol levels. In the Israel Ischemic Heart Disease Study, CHD death rates were directly related to the cho- lesterol level." In Japan, where lower cholesterol levels are common, a clear relationship has also been found with CHD morbidity rates nearly quin- tupling from the lowest (<180 mg/dL; 4.65 mmol/ L) to highest (>220 mg/dL; 5.69 mmol/L) tertiles of cholesterol level."”” Similar results with steadily increasing CHD mortality rates by cholesterol level quintile have been found in Great Britain.'” These studies provide further support for the concept that the lower the cholesterol level the greater will be the reduction in CHD risk. Indeed, these stud- ies go even further. They suggest that total blood cholesterol, or more specifically LDL cholesterol, is a fundamental risk factor for development of coronary atherosclerosis. In certain countries such as Japan and China where the average blood cho- lesterol level has been low, CHD is relatively rare even in the presence of other risk factors. In such countries, a high prevalence of hypertension commonly produces stroke, but does not produce high CHD rates, while a high prevalence of smok- ing commonly causes lung cancer, but again not a 42 high CHD rate. Even diabetes mellitus or low HDL-cholesterol levels do not greatly increase risk for CHD in populations having very low concen- trations of LDL cholesterol. These important epi- demiologic observations have led many inter- national groups to recommend national dietary changes as part of population strategies to reduce the prevalence of CHD.* The importance of blood cholesterol as a risk factor for CHD in various populations has been further emphasized by Peto.” He has pointed out that typical correlations obtained in epidemiologic studies tend to underestimate the strength of the connection. For example, the Framingham study indicated that for every 1 percent increase in cho- lesterol level, the risk for CHD increased by 2 percent on the average over the whole range of cholesterol levels. However, according to the cal- culation proposed by Peto, the duration of the observation plays a role. He has pointed out that analyses of long-term observational studies sug- gest that a 1 percent increase in cholesterol level may actually be associated with a 3 percent in- crease in risk. Nonetheless, it must be emphasized that these are average relationships, and the impact is less when low cholesterol levels are present and greater with high cholesterol levels. Cross-Population Studies. The most extensive and long-term epidemiologic investigation com- paring populations has been the Seven Countries Study." This project was initiated to examine the effects of culture, differences in diet, and other health habits on risk factors (such as cholesterol levels), morbidity, and mortality from CHD. The 16 population groups studied, about 500 to 1,000 persons per cohort, were chosen to represent con- trasting experiences with CHD. A total of about 11,000 men, ages 40 to 59 years at entry, were se- lected from Finland, Greece, Italy, Japan, The Neth- erlands, the United States, and Yugoslavia. The study was carefully designed and conducted, util- izing standardized procedures for examinations, laboratory testing, quality control, 7-day food rec- ords, and chemical analysis of duplicate meals. Blood lipids were determined at the outset and analyzed centrally at the University of Minnesota. The mean total cholesterol levels for the national Population Panel Report population samples ranged from 156 to 264 mg/ dL (4.03-6.83 mmol/L). The 10-year incidence and mortality from CHD showed that the population age-adjusted risk of death from CHD rose pro- gressively with increasing total cholesterol level of 180 mg/dL (4.65 mmol/L) and above (figure 6). The cholesterol-disease relationship was not sta- tistically significant within the five cohorts that ex- hibited the lowest level of blood cholesterol (three Italian and two Yugoslav), although longer fol- lowup of these cohorts revealed clear, positive re- lationships between cholesterol and CHD. The Seven Countries Study provides addi- tional evidence against the concept of a threshold relationship between cholesterol levels and CHD rates, a possibility that was raised by early Fram- ingham and pooling project data. As shown in figure 6, those with cholesterol levels below 200 mg/dL (5.17 mmol/L) had lower CHD rates than those with levels between 220 and 239 mg/dL (5.69-6.18 mmol/L), and the reduction in CHD was even greater when levels were below 180 mg/dL (4.65 mmol/L). These results thus are concordant with those of the 30-year followup of the Framing- ham study” and the 10-year followup of the MRFIT screenees! (figure 1, page 7). They all confirm the general dictum: the lower the blood cholesterol level, the lower the CHD risk. Migration Studies. The classical migration study is the Ni-Hon-San Study." In this study, compari- sons were made in populations of Japanese living in Japan, with people originally from the same regions of Japan who were living in Hawaii or in California. The study established rigorous quality Figure 6 The Seven Countries Study: Total Blood Cholesterol and Percentage of CHD Deaths in Men Originally Free of Cardiovascular Disease 10.0 9.0 8.0 7.0 6.0 5.0 4.0 10-Year CHD Deaths in % 3.0 2.0 1.0 0.0 Entry Serum Cholesterol <160 (mg/dL) CHD Events 10 9 25 32 Men Exposed 1007 1157 1525 1682 Source: Keys, A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press, 1980.8 160-179 180-199 200-219 220-239 240-259 260-279 280-299 >300 31 29 29 26 60 1480 1181 806 551 662 43 Population Panel Report control of measurement and used standardized CHD prevalence and incidence data. There was a marked difference between all three cohorts in the prevalence of CHD observed at entry—the lowest being in Japan and the highest in California. In a similar pattern the age-adjusted mean total cho- lesterol was 176 mg/dL (4.55 mmol/L) for the Japanese living in Japan, 219 mg/dL (5.67 mmol/ L) for the Hawaiian Japanese, and 226 mg/dL (5.85 mmol/L) for the Japanese in California. These blood cholesterol differences were associated with stepwise increases in CHD. The CHD incidence in California Japanese immigrants was two to three times higher than that of the CHD incidence for the Japanese in Japan. Similar changes in cholesterol level related to shifts in dietary composition have been noted among groups of people migrating into Israel.’ Conversely, groups of Irish migrating from Ire- land to Boston did not undergo major nutrient intake changes and did not display major differ- ences in blood cholesterol level or in CHD rates." The migration studies not only support the concept that populations with the lowest choles- terol levels have the lowest rates of CHD, but they also demonstrate that human populations are sus- ceptible to coronary atherosclerosis when blood cholesterol levels are raised by environmental factors. Thus, they add strength to the blood cholesterol-CHD link as revealed by within- and between-population studies. Blood Triglycerides and CHD. The relation be- tween blood triglycerides and CHD risk is not as well delineated as that between total blood choles- terol (and LDL-cholesterol) and CHD. Even so, most epidemiologic studies show a positive corre- lation between triglyceride levels and CHD risk, 44 but often only on univariate analysis.'®'* This re- lationship may be weakened somewhat because of the lability of triglyceride levels with recent food intake, compared to more constant cholesterol con- centrations,’ and by the likelihood that very high triglyceride levels do not raise risk more than mod- erately raised concentrations." Moreover, triglyc- eride levels generally are correlated inversely with HDL-cholesterol levels and directly, but weakly, with total cholesterol concentrations. When the latter two are given priority in multivariate analy- sis, triglycerides lose much of their ability to pre- dict CHD. Such analyses, however, do not take into consideration the possibility that defects in triglyceride metabolism may be the underlying causes of many cases of elevated cholesterol levels or of reduced HDL-cholesterol. Even if the blood cholesterol serves as a better predictor of CHD, this does not eliminate the possible importance of disorders of triglyceride metabolism in the causa- tion of CHD." Furthermore, there is growing evi- dence that certain triglyceride-rich lipoproteins, like VLDL remnants, are atherogenic; these lipo- proteins are increased in many hypertriglyc- eridemic states. Although several epidemiologic studies do not reveal an “independent” correlation between triglyceride concentrations and CHD, as reviewed by Hulley et al.,'"™ data from several studies’! as well as recent Framingham data” have reported an independent effect of high blood triglyceride levels on CHD risk. This effect may be more pronounced in women. Recently, Avins et al.'* reviewed over 30 studies that have been published in the 1980's that contain information relevant to this issue. These studies were broadly classified into: cross-sectional studies, prospective cohort studies, and intervention trials. These authors concluded as follows: The debate surrounding the independent role of TG [triglycerides] in the pathogenesis of CHD cannot be resolved based on available data. Current evidence indicates that, overall, the association often noted between TG and CHD is frequently due to confounding caused by the correlation between TG and cholesterol subfractions. Unlike the data implicating Population Panel Report increased LDL cholesteroland diminished HDL cholesterol in the causation of coronary artery disease, a strong and consistent thread of epi- demiologic evidence does not exist for isolated HTG [hypertriglyceridemia]. However, be- cause of several reports of positive associa- tions even after adjustment for LDL choles- terol and HDL cholesterol, and because infer- ences based on statistical adjustment are sus- ceptible to statistical modeling and other er- rors, a definite conclusion cannot be reached at this time. Nonetheless, even if high blood triglycerides per se are not an independent predictor of CHD risk, available evidence indicates that hypertriglyc- eridemic patients in general are at increased risk. From a population viewpoint, the most common cause of elevated blood triglyceride levels is obe- sity. Thus, the tendency of obesity to raise triglyc- eride and cholesterol levels, and to reduce HDL- cholesterol levels, provides additional reasons to recommend maintenance of a desirable body weight. Another frequent cause of high blood triglycerides is an excessive intake of alcohol, pro- viding an important rationale for people to limit their alcohol consumption. HDL-Cholesterol and CHD. Most within-coun- try studies show an inverse correlation between HDL-cholesterol concentrations and CHD risk.!s71761911%2 This inverse relation holds even within countries in which total cholesterol levels and CHD rates are relatively low. In the Framing- ham study,” the relationship of low HDL-choles- terol to CHD for both sexes ages 50 to 80 was found to be inverse and highly significant. The relation- ship held even into older age groups. Gordon et al." reviewing four major epidemiologic studies, found that each 1 mg/dL decrement in HDL- cholesterol was associated with an increment in CHD risk of between 2.0 to 3.7 percent in men and 3.0 to 4.7 percent in women. In spite of these within-country findings, cross- country studies do not reveal as strong a relation- ship between HDL and CHD as is found between total blood cholesterol (or LDL-cholesterol) and CHD risk.” This lesser HDL-CHD connection may have two explanations. First, in populations that have relatively low HDL levels, LDL concen- trations likewise tend to be low, and, as indicated before, CHD risk is relatively low when LDL levels are low, even in the presence of other risk factors (including reduced HDL levels). And second, there is not as wide a range of HDL-cholesterol levels between countries, regardless of CHD risk, as there is for LDL-cholesterol levels. Furthermore, epidemiologic studies have un- covered strong associations between other risk factors for CHD and HDL concentrations. Ciga- rette smoking, obesity, and lack of exercise are each associated with lower HDL-cholesterol lev- els. Itis difficult to disentangle these confounding variables when assessing the role of HDL in CHD risk. The issue is further complicated by the complex interaction of HDL with other lipopro- teins, notably potentially atherogenic, triglyceride- rich lipoproteins (e.g., VLDL remnants) and small, dense LDL." Thus, available data on HDL, taken as a whole, do not provide as strong a rationale for attempting toraise HDL levels in the general popu- lation as for lowering LDL levels." They strongly support general recommendations to stop smok- ing, to eliminate obesity, and to increase exercise— all of these being actions likely to raise HDL- cholesterol concentrations. b. Diet and CHD Epidemiologic data provide strong evidence that dietary factors are a major cause of CHD. As reviewed in preceding sections, animal studies first suggested a relationship between diet and atherosclerosis, a relationship mediated through diet-induced hypercholesterolemia. Clinical and metabolic-ward studies have demonstrated that certain dietary factors can raise blood cholesterol levels. Such short-term dietary studies do not, of course, demonstrate the critical link between diet and CHD. Linkage between dietary patterns and coronary atherosclerosis at autopsy has been re- ported by Moore and colleagues.” Ideally, a definitive diet-heart primary prevention trial would prove the diet-CHD connection; but after careful analysis of feasibility, such a trial was judged to be unrealistic for many reasons." While 45 Population Panel Report drug trials provide strong evidence that choles- terol lowering will reduce CHD risk, they do not fully answer the diet-heart question. For many disorders, epidemiologic studies yield clues to possible causal connections between various environmental and ethnic factors and disease endpoints; these possible connections must be substantiated by investigations of other kinds. For the diet-heart issue, the available information from epidemiologic studies has been reinforced by many other kinds of research. The issue no longer belongs in the category of “association,” but the totality of data indicate a causal link. The epidemi- ologic data on the diet-heart connection derives from population samples representing literally millions of people. Thus, by the sheer number of subjects under investigation, it has been possible to minimize the effects of confounding variables. Epidemiologic studies have three valuable fea- tures for assessing the diet-CHD relationship. These are sample size (number), duration, and ad- herence. Enormous numbers of people can be sampled after a lifetime of adherence to a particu- lar eating pattern. Estimates of the impact of diet on CHD risk have been made from the three types of epidemi- ologicstudies described previously—within-popu- lation, cross-population, and migration studies. During the early part of the 20th century, there were reports that atherosclerotic disease was rare in nonindustrialized countries in contrast with in- dustrialized countries and that the differences varied with the habitual diet of the country. Rosen- thal,'”” reviewing this literature up to 1934, de- duced a relationship between intake of fat and cho- lesterol and atherosclerosis. The role of dietary factors in the manifestation of CHD was observed in Norway during World War Il when deaths from CHD fell dramatically as the Nazi occupiers di- verted dietary fat to munitions production.” Although certain within-country studies have provided valuable information, they have a disad- vantage in that they contain a relatively narrow range of differences in eating patterns within a single population. On the other hand, wide differ- ences in eating patterns exist between countries and for migrating populations, and thus more 46 meaningful comparisons can be drawn from stud- ies of the latter two types. As indicated before, three major dietary factors that raise the blood cholesterol level and thus potentially increase risk for CHD have been iden- tified. These are saturated fatty acids, dietary cholesterol, and obesity. Thus, it is appropriate to review epidemiologic data as they pertain to each of these factors and their relation to CHD. In addition, epidemiologic studies provide sugges- tive evidence about the role of other factors— unsaturated fatty acids, carbohydrates, and fiber— although the available data are not as firm as for the major dietary factors affecting blood lipid levels. c. Epidemiologic Studies of Diet and CHD Saturated Fatty Acids. The strongest epidemio- logic support that diets high in SFA contribute sig- nificantly to increased CHD risk comes from the Seven Countries Study." This study found a high correlation between intakes of SFA and rates of CHD in several different population groups of the seven countries under investigation. The chemical analysis of duplicate diets, collected over four sea- sons of the year, showed substantial variation in the amount and type of fat among the populations sampled. The saturated fat was low (9 percent of calories or less) in Greece, Italy, Japan, and Yugo- slavia, and high (18-20 percent of the calories) in the populations from Finland, The Netherlands, and the United States that were studied. The average cholesterol levels and CHD rates were highly correlated with the percent of calories de- rived from SFA. In Crete, where the average total fat intake was 40 percent of calories but the SFA intake only 8 percent of calories, there were no CHD deaths and few incident cases after 10 years. This relationship appeared to be mediated through the action of SFA to raise the blood cholesterol level. In the 15-year followup of the Seven Coun- tries Study,” intakes of SFA were found to be cor- related with total mortality rates as well as with ‘CHD rates. Other epidemiologicstudies havealso observed high correlations between dietary SFA and CHD. Population Panel Report One of these was the Ni-Hon-San Study'®**** de- scribed previously. In this rigorous and detailed study, prevalence rates of CHD were found to be lowest in Japan and highest in California. The intake of SFA and cholesterol increased with mi- gration and westernization of the population, and were associated with higher CHD mortality rates.?>?* Further, CHD rates in the three popula- tions were correlated with blood cholesterol lev- els. There was a particularly high correlation between percent of calories from dietary SFA and CHD rates in these populations, a finding in ac- cord with the Seven Countries Study." Similar data have been found when other population groups are compared.”” Still other epidemiologic studies showing a correlation between CHD and intake of SFA have been the Western Electric Study,” the Ireland- Boston Study, the Honolulu Heart Program," and the Zutphen Study.?’’ Similar observations have been made in migrants to Israel’ and South Sea Islanders who have moved to industrialized cities in the South Pacific.>"? In each case, changes in dietary habits toward higher SFA and choles- terol consumption were associated with increas- ing incidence of CHD. Finally, a survey by the Food and Agriculture Organization of the United Nations of CHD rates in 20 countries indicated that intakes of SFA were highly correlated with CHD." Many other smaller epidemiologic studies pro- vide data also supporting a connection between intakes of SFA and rates of CHD. Populations with specific dietary practices pro- vide additional information. Within the United States, studies of the Seventh Day Adventists show differences in plasma lipid levels and CHD inci- dence of men according to dietary habit, e.g., nonvegetarian, lacto-ovo-vegetarian, and pure vegetarian. In contrast with a non-Adventist com- parison population, nonvegetarian men age 35 years and above had a CHD rate that was 44 percent lower, the lacto-ovo-vegetarian group 57 percent lower, and the pure vegetarian group 77 percent lower.” Other studies associate low blood cholesterol levels and low CHD risk among vege- tarians.?’>?" Although dietary differences such as low SFA probably account for part of the differences in CHD risk in vegetarians, abstinence from tobacco and alcohol use and other health characteristics are probably also involved. Dietary Cholesterol. For many years it has been known that feeding cholesterol in the diet to vari- ous laboratory animals such as rabbits, chickens, and monkeys produces hypercholesterolemia and atherosclerosis. However, since humans are less susceptible to the blood cholesterol-raising effects of dietary cholesterol than some other animal species, there has been an ongoing debate whether high intakes of cholesterol are an important cause of human atherosclerosis. Indeed, humans can vary in their response to dietary cholesterol.’ Although most epidemiologic studies havestressed the importance of SFA as a dietary atherogenic factor, a growing body of evidence implicates dietary cholesterol as well. This evidence includes results of the Western Electric Study?" the Honolulu Heart Program,” the Ireland-Boston Study,'®* and the Zutphen Study .”"” In all of these studies, intakes of dietary cholesterol were corre- lated positively with rates of CHD, sometimes in- dependently of blood cholesterol-CHD correla- tion. In addition, in studies based on the Food and Agriculture Organization and World Health Or- ganization data, mean per capita intakes of dietary cholesterol correlate with CHD mortality rates. It must be noted, however, that intakes of SFA and cholesterol, in general, are highly correlated in populations making it difficult to distinguish clearly the independent effect of each. The epi- demiologic data relating dietary cholesterol to human CHD have been reviewed in detail by Stamler and Shekelle.””! These investigators con- clude that “both cross-population and within- population epidemiologic data indicate that die- tary cholesterol is atherogenic in humans, over and above—i.e., independent of—its effect on se- rum cholesterol.” This implies that dietary choles- terol may promote atherosclerosis both by raising total blood cholesterol and LDL-cholesterol levels and by mechanisms independent of fasting choles- terol levels, possibly through the formation of cho- lesterol-rich postprandial lipoproteins (e.g., chylomicron remnants). 47 Population Panel Report Obesity. There is a common belief that obesity contributes to cardiovascular disease. In fact, several reports” ** indicate a connection between being overweight and developing CHD. The mechanisms for this connection are not entirely clear, however, because obesity has so many dif- ferent metabolic effects. It raises blood pressure, increases blood cholesterol, reduces HDL-choles- terol, and induces resistance to the peripheral ac- tion of insulin. Any of these factors may heighten CHD risk. Thus, the influence of obesity on CHD risk might be explained entirely by its action to augment known risk factors. On the other hand, data from the Framingham Heart Study*” indicate that obesity also is an independent risk factor for CHD. This report examined the relationship be- tween the degree of obesity and incidence of cardio- vascular disease in more than 5,000 men and women of the original Framingham cohort. These individuals were followed for a period of 26 years. Analyses have shown that the percentage of desir- able weight on initial examination, according to Metropolitan relative weight tables, predicted 26- year incidence of coronary disease (both angina pectoris and CHD other than angina) and CHD deathsindependent of age, blood cholesterol, blood pressure, cigarette smoking, left ventricular hy- pertrophy, and glucose intolerance. Moreover, weight gain after the initial measurement also predicted an increased risk for CHD. This impor- tant study shows that excess caloric intake leading to obesity is a major dietary factor predisposing to CHD. Unsaturated Fatty Acids. There are three major categories of unsaturated fatty acids—omega-6 polyunsaturates (mainly linoleic acid), omega-9 monounsaturates (mainly oleic acid), and omega- 3 polyunsaturates (linolenic acid, EPA, DHA). In 48 the 1950's and 1960's, there was intense interest in the putative cholesterol-lowering action of linoleic acid, and some investigators of that era advocated high intakes (i.e., 10-20 percent of total calories). However, no large populations have consumed such high intakes of linoleic acid for prolonged periods, and, thus, the long-term consequences of such intakes are not known. Accordingly, there are no epidemiologic results to support very high intakes of linoleic acid. Consequently, most expert panels have recommended that individuals con- sume no more than 10 percent of calories as poly- unsaturated fatty acids. That this is prudent is bol- stered by clinical trial data suggesting that high intakes of linoleic acid may increase risk for gall- stones’ and animal data that high PUFA intake may promote development of tumors*’** and suppress the immune system.” On the other hand, very low intakes of linoleic acid (less than 5 percent of calories) have been related to increased CHD rates in high-risk popu- lations.”*?* The Western Electric Study** also noted that modest increases in linoleic-acid intake were associated with decreased CHD rates. All of these data provide some justification for the rec- ommendation’ that omega-6 PUFA intake be main- tained at about the current level (approximately 7 percent of calories). There are no epidemiologic data to suggest that intakes in the range of 5-10 percent of calories are accompanied by an in- creased risk of diseases such as cancer or gallstones. Substitution of dietary saturated fatty acids with oleic acid results in a fall in LDL-cholesterol levels. However, in contrast to linoleic acid, there are epidemiologic data that demonstrate long- term safety of high intakes of oleic acid. In the Mediterranean region of the world, where large quantities of olive oil are consumed, rates of CHD are relatively low, and there is no evidence for harmful effects (e.g., increased rates of cancer). In fact, in the 15-year followup of the Seven Countries Study,”! intakes of monounsaturated fatty acids (oleic acid) were inversely correlated with both CHD rate and total mortality. This finding thus justifies encouraging higher intakes of oleic acid in the diet, i.e., 14-16 percent of calories, as a substi- tute for saturated fatty acids. Population Panel Report In recent years, there has been a growing inter- est in the possible beneficial effects of omega-3 polyunsaturated fatty acids. This interest stems, in part, from the limited observations suggesting that Greenland Eskimos, who consume large quantities of omega-3 PUFA, have low rates of CHD.> These fatty acids are known to have an antithrombotic effect; many Greenland Eskimos have a bleeding tendency. The low risk of CHD in this population thus might be explained by the absence of coronary thrombosis or perhaps some other protective effect of omega-3 PUFA. On the other hand, there could be a variety of confound- ing factors in the Eskimo population that could contribute to low rates of CHD. For example, they generally have low levels of other major risk fac- tors. Nonetheless, some other population stud- ies? suggest that consumption of fish may exert a protective action against CHD. Although these population studies alone do not show a clear pro- tective effect of omega-3 PUFA, they provide another reason for a more liberal intake of fish, especially because fish contain protein with low proportions of saturated fatty acids. Carbohydrates. There are no large population studies which suggest that high intakes of carbo- hydrates are accompanied by increased risk for CHD. If anything, the opposite is true. Popula- tions that consume high-carbohydrate, low-fatdiets generally have low rates of CHD." This finding is in accord with metabolic studies which show that replacement of saturated fatty acids by carbohy- drates causes a lowering of total cholesterol and LDL-cholesterol levels. As reviewed in a previous section, metabolic studies indicate that high-car- bohydrate diets can raise triglyceride levels and lower HDL-cholesterol levels,” but these changes apparently do not translate into increased risk for CHD.? Thus, cross-country, epidemiologic stud- ies in general support the conclusion that intakes of carbohydrates in the range of 55 percent of total calories are acceptable and that low-fat, high-car- bohydrate diets are associated with low rates of CHD and of certain forms of cancer.”*!! In summary, a broad base of epidemiologic data is available to support the conclusion that there are strong and causal relationships between diet, blood cholesterol level, and CHD. These data provide overwhelming evidence that elevated levels of blood cholesterol and LDL-cholesterol are major risk factors for CHD. In addition, they indicate that dietary composition is highly corre- lated with rates of CHD. In particular they support the conclusion that three major dietary influences on blood cholesterol level—saturated fatty acids, cholesterol, and maintenance of desirable body weight—directly affect CHD rates. 4. Observations in Special Populations a. Children Children representan extremely important spe- cial population. A separate Expert Panel on Blood Cholesterol Levels in Children and Adolescents will provide more detail about how to implement eating pattern changes in healthy children and will recommend approaches for detection and treat- ment of children with high-risk cholesterol levels. In its report, that panel will discuss more fully the issues of dietary safety and adequacy for children and adolescents. The present report reviews the evidence in summary fashion. Surveys of childhood levels of cholesterol and lipoprotein fractions show that North American children have higher total and LDL-cholesterol levels than children in countries where athero- sclerosis is less frequent.”*> A number of surveys of childhood lipid and lipoprotein levels have been published in the United States. One survey that provides nearly a complete age span and values for the lipoprotein fractions is that of the Lipid Research Clinics (LRC) Prevalence Study.” The LRC study sampled populations from multiple sites in the United States. The blood cholesterol levels in children and adolescents are significantly lower than those in adults. In pubertal years, there is a slight transient fall in total cholesterol and LDL-cholesterol levels presumably due to hormonal changes. Blood cholesterol levels in children have a predictive value for levels of cholesterol in young adult life. 49 Population Panel Report In general, children with high cholesterol levels are most likely to have high cholesterol levels as young adults.?**?* However, there is considerable variability. Of children whose blood cholesterol levels are greater than the 90th percentile during childhood, most of them will remain above the 50th percentile.” A correlation between children’s habitual dietary intake of fat and blood cholesterol levels has been shown.** Since no groups of children have been fol- lowed for decades after having had their choles- terol or lipoprotein fractions measured, the rela- tionship of high total cholesterol, LDL-cholesterol, and low HDL-cholesterol during childhood and the atherosclerotic process must be inferred from other data. Obesity plays an important role in the devel- opment of hyperlipidemia, hypertension, diabetes mellitus, and other chronic illnesses in adult- hood.?*#2¥ Obesity acquired in childhood most often persists into adult life.”**" Obese children generally have higher blood cholesterol levels.” In studies following children from school age to age 25 to 30 years, those who gained excess weight for their heights were found to have the greatest increases in their cholesterol levels.?* Several lines of evidence point to a genetic role in the hypercholesterolemia of some children. Children with homozygous familial hyper- cholesterolemia with extreme elevation of total cholesterol and LDL-cholesterol levels have been shown to develop CHD at a very early age.” In addition, many children whose cholesterol levels were greater than the population-based 95th per- centile had a familial prevalence of hyperlipidemia as well as premature CHD.>*** When the progeny of young ischemic heart disease victims were examined, many had hypercholesterolemia.*” Morrison et al.» found that children whose par- ents had experienced a myocardial infarction earlier than age 50 had significantly higher blood cholesterol levels, on average about 195 versus 175 mg/dL (5.04 versus 4.53 mmol/L), than children with parents without early myocardial infarction. Fatty streaks are found in the vascular system of most children, regardless of age, race, sex, or 50 environment, by the age of 10 years.” These fatty streaks cause little or no obstruction and therefore no clinical symptoms. Since these lesions are ubiquitous, some have questioned whether these are the earliest manifestations of the atheroscle- rotic process. Newman and colleagues'’ con- ducted a postmortem study of children whose lipids and lipoproteins had been previously meas- ured and who died accidentally. They showed a relationship between the amount of fatty deposits in arteries and LDL-cholesterol levels. The chil- dren with larger proportions of aortic wall in- volved with fatty streaks were those with higher total cholesterol and LDL-cholesterol levels. In- tensive postmortem examination of large num- bers of other children has shown that a progres- sion of lesions could be traced from the fatty streaks in coronary arteries to the fibrous plaques.” These observations suggest that many coronary artery fatty streaks and some aortic fatty streaks in chil- dren are converted to fibrous plaques in adoles- cents and young adults. Fibrous plaques are the lesions most characteristic of progressing athero- sclerosis.” With age, the fibrous plaques may become vascularized, calcified, and more obstruc- tive. Autopsy studies of U.S. war casualties in Ko- rea and Vietnam confirmed the frequent presence of grossly detectable coronary artery lesions in late teenage and early adulthood”! This was also confirmed in similar-age young people dying of accidental death in the United States.”” Since the early stages of the atherosclerotic process often have origins in childhood *? dietary changes must be viewed in the context of the im- portant nutrient needs for the growth and devel- opment of a child. The dietary requirements for a growing child must provide more calories per kilogram of body weight for heat production to offset the relatively high heat losses over a larger surface area per kilogram for children compared to adults.**?** In addition, bodily growth in chil- dren requires not only further augmentation of caloricintake, butalso intake of a diversity of foods that provide the nutrients needed for growth. Particular attention has been drawn to calcium, iron, and zinc.” Shifts away from fats, which provide a high caloric density, require special care to assure adequate intake of recommended daily Population Panel Report nutrients. Reliance on foods that are low in fat and calories requires that children, in order to eat a nutritionally adequate diet, spend the necessary time to consume more food. An important question concerns the age at which fat consumption can safely be reduced. In infants whose total nutrition is dependent on milk, there is evidence that very low-fat or skim milk formulas do not provide sufficient energy for normal growth.” However, infants fed 30 percent or more of their total calories as fat grow nor- mally.” In countries of the Mediterranean and Far East, children eat with apparent safety a diet con- taining an average of 10 percent or less of calories from SFA and characterized by consumption of grains, legumes, vegetables, fruits, fish, lean meats, and poultry. The human is capable of synthesizing most of the lipid compounds necessary for health. The fatty acids essential for normal growth and de- velopment, which cannot be synthesized de novo by humans, are the omega-6 PUFA (e.g., linoleic) and the omega-3 PUFA. Furthermore, lipid- soluble vitamins require fat intake for their absorption. Surveys of childhood diets were carried out by the National Center for Health Statistics from 1971 to 1974 and again from 1976 to 1980.*%*" These surveys showed that mean fat intake was about 32- 33 percent of total calories consumed per day in the age group 6 to 12 months, and was about 35-37 percent of calories from 1 to 18 years, with approxi- mately one-third of children from 1 to 18 years eating more than 40 percent of calories from fat. Within this same age span, 79 percent received >10 percent of their calories from saturated fats. The data from the NHANES II”? survey showed that 30 percent of children 1.1 to 1.9 years of age con- sumed more than 300 mg dietary cholesterol per day. For those 2 to 18 years old the mean choles- terol intake was about 280 mg/day, with 34 per- cent having an intake exceeding 300 mg/day.” Thus, many U.S. children, along with many adults, presently consume diets that are rich in SFA and cholesterol 264% The data related to childhood have been inter- preted by expert panels with different emphases. The first expert group to recommend that infants and children modify their diets to prevent CHD was the Inter-Society Commission for Heart Dis- ease Resources.” The American Academy of Pedi- atrics (AAP) has recommended an individual ap- proach. The AAP committee on nutrition state- ment, “Prudent Life-style for Children: Dietary Fat and Cholesterol,”?? has concentrated on efforts to identify children with a family history of heart disease or high blood lipid levels in a family member younger than 60 years of age. In addition, the AAP suggests that children with blood choles- terol >176 mg/dL (4.55 mmol/L) be given dietary intervention.”? This strategy requires individual nutritional, therapeutic, and genetic counseling programs. In 1983 the American Heart Association adopted the public health approach to preventing CHD in children. The AHA statement, “Diet in the Healthy Child,”?* recommended that all children over 2 years reduce fat intake to approximately 30 percent of calories with 10 percent or less from SFA, less than 10 percent from PUFA, about 10 percent as MUFA, and cholesterol intake to 100 mg/1,000 calories. This recommendation by the AHA was endorsed by a consensus develop- ment panel on lowering blood cholesterol to pre- vent heart disease that was convened by the Na- tional Institutes of Health.””> These reports view the high incidence of CHD in the U.S. population and the high cholesterol distribution in American children as justification for this dietary interven- tion. Inaddition, they suggest that dietary changes in early life and in the family setting have the potential for healthier lifestyles and decreased CHD risk in adulthood. 51 Population Panel Report The positions of the committees on nutrition of the AHA and the AAP have many more similari- ties than differences. Some critics of the AHA statement have suggested that there is no evidence that the recommended diet is adequate to support growth and development. Some have noted that the diet recommended is bulky and low in energy and essential nutrients because it restricts fat and SFA and promotes an increase in consumption of cereal grains and plant products. Concern has been expressed about the adequacy of intake of high-quality protein, iron, calcium, and other minerals in such diets.”* However, there is no evidence in the scientific literature to support a conclusion that the decreased intake of fat, SFA, and cholesterol to the levels suggested by the AHA would compromise growth or neurological devel- opment. In fact, the diet recommended by the AHA maintains protein and essential mineral and vitamin intake with increased carbohydrate con- sumption substituting for fat and SFA intake. Recently the committees on nutrition of the AHA and the AAP?” met and agreed on several points summarized in the following: 1. No changes in the current feeding pattern for infants and children less than 2 years of age are needed. 2. The present trend toward lower SFA intake in diets for children 2 to 18 years of age should be encouraged. The AHA and AAP committees stated that, although it is not unanimous, there is consensus that 30 percent of total calories from fat (with 10 percent from SFA) is a reason- able target for children. 3. Educational programs for physicians and the public should be undertaken so as to discour- age excessive limitation of fat and animal pro- tein intake that might lead to mineral deficien- cies and growth failure. 4. Research is under way and should be contin- ued on the specific effects on growth and development of diets containing approximately 30 percent of calories from fat. Many American families have already changed their eating habits toward lower-fat eating 52 patterns without evident untoward effects. Only in children with severe fat-malabsorption syn- dromes have linoleic acid or fat-soluble vitamin deficiencies occurred.”® There have been reports of a few zealous families that over-restricted fat and protein intake and whose very young children under 2 years of age consequently failed to thrive.” Thus, as dietary changes are recommended to the general population, some caution must be ob- served with respect to adequacy of energy intake for children. However, a modest reduction of total fat in the diet to 30 percent of calories will allow sufficient energy for growth and development. The Diet and Health report of the NRC,” based on its extensive review of the relevant scientific evidence, has recommended that children over the age of 2 years follow the same dietary guidelines as adults. Specifically, the NRC recommends that both children over the age of 2 years and adults consume 30 percent or less of calories as fat, with less than 10 percent of calories from SFA. The NRC has also recommended that children limit their dietary cholesterol intake to under 100 mg per 1,000 calories consumed, with the total not to exceed 300 mg per day. In summary, data derived from children and adolescents indicate that diet influences blood cholesterol level and that elevated blood choles- terol is often associated with early stages of the atherosclerotic process. Accordingly, it appears reasonable for children, ages 2 years and older, to have SFA, total fat, and cholesterol intakes at the same levels recommended for the adult popula- tion. Available data support the safety of follow- ing these recommendations. A number of profes- sional and scientific organizations have already made these recommendations. b. Older Adults Elderly people form an increasing proportion of the U.S. population. In this age group, CHD is a leading cause of death and disability. Increased levels of blood cholesterol are also common in the elderly, and are frequently higher than the levels at which treatment is recommended by the Adult Population Panel Report Treatment Panel of the NCEP. Thus, there is a need for information and recommendations concerning blood cholesterol in the population of older adults. Little is also known about older adults” aware- ness of blood cholesterol levels and CHD, or of the cholesterol levels above which medical treatment is commonly initiated in this population. Popula- tion surveys in urban Minnesota have shown that 12.4 percent of men and 13.9 percent of women ages 60 to 74 years were aware of their own diag- nosis of hypercholesterolemia; **' relatively few were under any medical treatment. The absolute rates of CHD are also high among the elderly population. CHD mortality rates in the 65 and above age categories are many times those found among middle-age adults.” In addition, CHD rates, which are much higher in men than women in younger age categories, tend to equalize in older age groups, with women exhibiting con- siderable rate increases as they age. A substantial burden of CHD morbidity and mortality in the elderly isamong women. Because of the relatively high event rate among older people, small per- centage shifts in disease rates may produce large differences in number of cases. For example, even a 1 percent fall in CHD mortality rates in the 65 and over age category could result in 4,300 fewer CHD deaths a year in the United States. Mean blood cholesterol levels tend to rise with age among young and middle-age adults in the United States. For men, they peak in the sixth decade and then decline. For women, the average blood cholesterol level continues to rise into the seventh decade and then decreases slightly .?***% Increased blood cholesterol, by the standards of the ATP, is very common among the older population.” Based on the two criteria of the ATP of >200 mg/dL (>5.17 mmol/L) and >240 mg/dL (>6.21 mmol/L), substantial numbers of individu- als will be classified as having borderline-high or high blood cholesterol levels. For menand women, this would include the majority of the population age 60 to 89 years. Epidemiologic data for persons over the age of 65 years are sparse. Data from early studies indicated little association between total blood cholesterol and CHD mortality in groups over 60 years of age. Long-term followup (12 years) of a population of white males in Los Angeles studied from 1950 to 1962 suggested that blood cholesterol level was unrelated to myocardial infarction rates in older men. Compared with men with a blood cholesterol less than 270 mg/dL (6.98 mmol/L), the relative risk in men above 270 mg/dL (6.98 mmol/L) from ages 21 to 59 years ranged from 1.6 to 3.9. However, in the age category of 60 to 70 years, there was no increased risk (relative risk = 1.0) associated with a cholesterol level over 270 mg/dL (6.98 mmol /L).** Gofman et al.” reported that blood total cho- lesterol lost its predictive power as men aged and that it disappeared as a risk factor in men above age 65. A Danish cohort studied in the same era demonstrated some predictive power for choles- terol, even in the 70 to 80 year age group.” The effect, however, was much weaker than that ob- served for younger men. Studies conducted more recently, however, in- cluding further followup data on previously ex- amined populations, indicate that both total cho- lesterol and the lipoprotein cholesterol subfrac- tions are predictive of disease in the elderly. In contrast to earlier reports from the Framingham study, which showed little predictive power for cholesterol in the elderly, the 30-year followup data from this study indicate that blood choles- terol levels continue to predict CHD incidence and mortality in older men and women but that the strength of the association lessens with age. Both elevated LDL-cholesterol and reduced HDL-cho- lesterol were predictive of CHD in men and women through age 82.2 In the Puerto Rico heart study, elevated cholesterol >200mg/dL (>5.17 mmol/L) atbaseline was found to be predictive of CHD mortality in the male 60- to 79-year age group. The augmentation in risk was only slightly lower than that found in the age group less than 60 years.” The Lipid Research Clinics (LRC) prevalence study surveyed healthy populations from 1972 to 1974. Followup data from one center showed that cholesterol pre- dicted ischemic heart disease mortality in the 50 to 53 Population Panel Report 64 age group and in the older 65 to 79 group.” More recent data from the LRC program for men 65 years or older followed for an average of 8.5 years found that both total cholesterol and LDL- cholesterol were predictive of CHD outcomes af- ter adjustment for other risk characteristics.” Similarly, Benfante and Reed”! have analyzed blood cholesterol and CHD in males over 65 years of age who were followed for an average of 12 years. The incidence of CHD increased steadily from the lowest to highest quartiles of cholesterol level. Gordon and Rifkind*? examined the issue of age and cholesterol as CHD risk factors in the data from the 356,222 males, ages 35 to 57 years old, screened for the Multiple Risk Factor Intervention Trial. This analysis shows that CHD mortality rates increased progressively with serum choles- terol within each age group. Italso documents that the relationship is greatest in the youngest age group and declines with increasing age. Gordon and Rifkind state that although a smaller propor- tion of CHD cases may be prevented by cholesterol reduction in older than in younger patients, the potential benefit of treatment, when expressed as the number of events prevented per year per 1,000 treated, is greater in older than in younger pa- tients. Between ages 65 and 74 years, 16 percent fewer men (38 versus 54 percent) and 9 percent fewer women (21 versus 30 percent) would be ex- pected to have a new manifestation of CHD with a cholesterol level of 200 mg/dL (5.17 mmol/L) than with a cholesterol of 285 mg/dL (7.37 mmol/L). A report by Garber et al.” suggests that the magni- tude of benefit of blood cholesterol lowering in the elderly remains to be fully clarified. 54 The question of modifying cholesterol levels in older adults is a difficult one with fewer data available than for younger adults. Among the few dietary intervention studies aimed at lowering blood cholesterol levels and CHD, some have used study populations of elderly subjects. Albanese et al. studied nursing home populations age 70 and older to test the effect of a polyunsaturated to saturated fatty acid ratio (P/S ratio) of 0.7, com- pared with a P/S ratio of 0.1 to 0.3, which was es- timated to be the norm for institutional diets at that time. The dietary changes resulted in a fall in blood cholesterol that averaged 20 percent. In the same era, an 8-year controlled dietary trial in which an institutionalized population of men substituted vegetable for animal fats was begun in a Los Angeles Veterans Administration hospital. The study suffered from poor adherence rates and small sample size. However, blood cholesterol was reduced, and there was a reduction in disease rates in the under-65 population. In the older group (>65 years) there were only small differ- ences between the experimental and control groups.” The issue of adequate nutrition in the elderly is important. Both poor dietary intake and mal- absorption of important nutrients have been found in many elderly people.” Accordingly, there is a concern that some older people may over-adhere to an eating plan that is lower in total fat, SFA, and cholesterol and hence suffer from lack of essential nutrients. However, in a recent series of papers concerning the nutritional status of non-institu- tionalized elderly people (ages 60 to 93 years), this did not occur (e.g., Goodwin et al.).*” In this group, individuals who were concerned about lowering blood cholesterol levels and who fol- lowed dietary principles to achieve this were not deficient in intakes of protein, calcium, and fat- soluble vitamins. Thus it appears that for this group, the diet both lowered consumption of fats and preserved the intake of essential nutrients. In summary, population data show that both total blood cholesterol and LDL-cholesterol levels are predictive of CHD events among adults older than 60 years, that CHD is highly prevalent in the elderly, that blood cholesterol levels are elevated Population Panel Report in the majority of this group, and that blood choles- terol levels can be reduced by dietary change. c. Women Women's dietary intakes and overall health with respect to CHD have only recently received the attention they deserve. CHD accounts for about 250,000 deaths each year among women in the United States, and is responsible for about one- third of all deaths of women in the United States. In age-adjusted data, CHD is the leading cause of death among women, and it is the second most common cause of death among women 35 to 74 years of age. Rates of death due to CHD are 1.5 times higher among black women than among white women. The trend in the CHD death rate in the 1940's and 1950's was upward among men and older women, flatamong middle-age women, and down- ward among younger women. The decline be- tween 1962 and 1978 in CHD death rates began earlier among women than among men, and the percentage decline in CHD mortality has been steeper for younger than for older women.*" However, since 1979, declines in CHD mortality have been steeper among men than women, low- ering the male/female ratio of CHD deaths for the first time. First coronary events are more often fatalamong women (39 percent) than among men (31 per- cent). Other initial manifestations of CHD are dissimilar in men and women. Fifty percent of men with CHD are first seen with myocardial infarction, compared with 34 percent of women.” In the Framingham study, angina was the initial manifestation of CHD in women more frequently than in men. Although women experience an age- adjusted CHD mortality rate less than one-half the rate among men, the total CHD rate in women is similar to that in men. Women in 1980 were re- sponsible for 58 percent of total expenditures for heart disease. The risk factors for CHD in women are the same as in men, namely, age, hypertension, ciga- rette smoking, high total or LDL-cholesterol level, low HDL-cholesterol level, diabetes mellitus, and obesity. However, diabetes appears to be rela- tively more important for women. Cigarette smoking is a major risk factor, but somewhat less so than in men.” Additional characteristics unique to women that may exacerbate the effect of se- lected risk factors include oral contraceptive use, menopause, hysterectomy, and exogenous post- menopausal hormone use. According to the sec- ond National Health and Nutrition Examination Survey (NHANES II), more than 58 percent of U.S. adult women have blood cholesterol levels that are above 200 mg/dL (5.17 mmol /L).> Other cardiac risk factors are also common, including hyperten- sion and smoking.*" Considerable data on risk factors and CHD as- sociations in women are available from the Fram- ingham study. For women ages 40 to 49 years, smoking and relative weight were predictors of CHD. Systolic blood pressure and blood choles- terol concentration both were positively related but did not reach statistical significance, probably because of the small number of CHD cases. Among women ages 50 to 59 years at baseline, systolic blood pressure, relative weight, and blood choles- terol levels were significant predictors of total CHD.* For women age 60 to 69 years, glucose in- tolerance and high blood pressure (systolic blood pressure) were independently related to the devel- opment of CHD.*” It has been argued that at least a portion of the sex difference in CHD susceptibility is related to sex differences in blood lipoprotein levels. In men, fasting levels of VLDL-cholesterol and triglycerides rise to a peak at about 45 to 50 years of age and then decline. In women, VLDL-cholesterol and triglyc- erides rise steadily with age and are lower than in men at all ages up to about 60 years. LDL-choles- terol rises with age in men, but the increase peaks in the sixth decade. In young women, LDL-choles- terol is lower than in men, but rises rapidly with age, overtaking levels in men in the sixth decade. In adults, HDL-cholesterol shows little variation with age, being consistently lower in men. The sex difference in HDL remains even after differences in body weight, smoking, and physical activity have been taken into account.’” 55 Population Panel Report The data are inconclusive regarding the effect of exercise on plasma cholesterol and lipoprotein levels in women. A few studies report a significant increase in HDL-cholesterol levels in women undergoing strenuous endurance training pro- grams.” Other studies have reported no ef- fect. It is hypothesized that HDL-cholesterol levels may not increase in women who exercise because of the decreased circulating estrogen concentrations that are associated with endurance training of women. Although CHD accounts for one-third of all deaths among American women, evaluations of medical therapies among women with CHD have been limited. Clinical trial data to establish the benefit of reducing blood cholesterol levels in women are not available. Expectations about the effects of intervention to prevent CHD among women will need to be based for the most part on the result of available data from trials conducted among men, as well as on inferences from the ex- tensive observational epidemiologic research concerning women. The data on the prevalence of high blood cho- lesterol and CHD in women, and data on the nutrient needs of women, particularly adolescents and women of childbearing age, emphasize the importance of dietary recommendations to lower high blood cholesterol levels in women. The data also indicate that needed dietary changes can be accomplished while ensuring full nutritional adequacy. d. Socioeconomic Status Groups Socioeconomic status (SES) is usually defined by the level of one or more of the following vari- ables: income, education, and occupation. The levels of these three variables are highly correlated and interrelated. A substantial minority of the U.S. population are members of low SES groups. For example, approximately 15 percent live in poverty defined by family income levels.” Poverty rates vary significantly by race, however: 11 percent of whites live in poverty, compared to 34 percent of blacks and 30 percent of Hispanics. 56 Numerous studies have examined the rela- tionship between SES and risk of developing CHD." Some studies have found higher CHD rates among higher SES groups while others have found little relationship with SES, and still others have found an inverse relationship. Some studies have shown different relationships between SES and risk of developing CHD for men and women. Others have found different relationships among different racial groups. Some studies have shown a reversal over time in terms of which SES groups are at higher risk of CHD. In general, lower SES groups currently appear to have higher CHD mortality rates than do higher SES groups. This increase in relative risk is generally in the range of 1.25 to 1.50. Mortality rates for CHD have fallen among all SES groups in the United States in the past 20 years, but the declines have been greater in the higher SES groups. Part of the differences in CHD mortality rates among different SES groups may be explained by differences in the prevalence of some risk factors and by changes in that prevalence over time.***"” Higher prevalence rates of uncontrolled high blood pressure, smoking, physical inactivity, and obe- sity are currently found among lower socioeco- nomic status groups.’®*"” In contrast to other risk factors, mean blood cholesterol levels are quite similar in both higher and lower SES groups in the United States.” The 1976-80 NHANES II’ found that mean blood cholesterol levels were about 6 percent lower in both white and black men who live in poverty compared to the levels in white and black men with family incomes above $20,000. Be- cause of the apparent increase in other risk factors and CHD in lower SES groups, reduction in blood cholesterol may be particularly important in this population. 5. Intervention Studies a. Clinical Trials The broad spectrum of data reviewed in this report suggests that elevated levels of cholesterol, whatever their cause, lead to accelerated athero- sclerosis and CHD, and that a high intake of Population Panel Report saturated fatty acids and cholesterol leads to high blood cholesterol levels. The data strongly suggest that lowering blood cholesterol would reduce the incidence of CHD. Until recently though, direct evidence from well-conducted clinical trials that lowering cholesterol levels is safe and beneficial was not available. Given the role that diet is held to have in pre- venting CHD, the ideal study of cholesterol lower- ing would be a clinical trial of diet. This possibility was considered by the Arteriosclerosis Task Force of the National Heart and Lung Institute in 1971. The task force had available to them the results and conclusions of the Diet/Heart Feasibility Study. The Arteriosclerosis Task Force recommended against a dietary clinical trial based on several con- siderations. For example, the required sample size was variously estimated to range from 24,000 to 115,000 individuals, and 7 to 10 years of followup would be necessary. The managerial problems of carrying out a well-controlled study under such circumstances would make the study difficult to complete. The projections of costs then ranged up to $1 billion. Of special importance was the possi- bility that subjects adhering to a dietary regime would most likely modify other risk factors, thus rendering interpretation of a positive outcome extremely difficult. It should be noted that these considerations equally apply now, just as they did in 1971, and that it is most unlikely that such a trial will ever be performed. Accordingly, testing the hypothesis that cholesterol lowering will reduce therisk of CHD has largely been attempted through drug studies in relatively high-risk segments of the population. There are several ways by which blood choles- terol levels can be reduced pharmacologically.™! A number of effective cholesterol-lowering agents, many acting by different mechanisms, are now available for clinical use. The use and mechanisms of action of these agents have been discussed in the NCEP Adult Treatment Panel Report.’ Prior to the 1980's there were many clinical trials of cholesterol lowering by diet or drugs in patients with or without preexisting clinical CHD. Although the results of some of these studies were encouraging, none of them was held to be conclusive. The limitations of these studies in- cluded small sample size, a short period of fol- lowup, inadequate randomization procedures, absence of a double-blind design, high dropout rates, or problems in statistical analysis. However, a number of well-designed major studies have been reported more recently. One of these, the Lipid Research Clinics Coro- nary Primary Prevention Trial, was directed by the National Heart, Lung, and Blood Institute. The LRC-CPPT tested the efficacy of cholesterol lower- ing in reducing the risk of CHD in 3,806 asympto- matic middle-age men with hypercholesterol- emia.” The treatment group received the bile acid sequestrant cholestyramine resin, and the control group received a placebo for an average of 7.4 years. Both groups followed a moderate cho- lesterol-lowering diet. The cholestyramine group had 8.5 percent and 12.6 percent greater reduc- tions, respectively, in total and LDL-cholesterol levels compared with the placebo group. The cholestyramine group experienced a 19 percent re- duction in risk (p<0.05) of the primary end point— definite CHD death and/or definite nonfatal myo- cardial infarction. Further analysis within the cholestyramine-treated group in the LRC-CPPT showed that men who took the full dose of the medication had a fall of 25 percent in total choles- terol (with an average 35 percent decrease in LDL- cholesterol) and had a CHD incidence that was half that of men who remained at pretreatment levels. In addition, the incidence rates for new positive exercise tests, angina, and coronary by- pass surgery were reduced by 25, 20, and 21 per- cent, respectively, in the cholestyramine group. Taking all endpoints into account, 577 of the 1,740 subjects taking placebo developed an endpoint, compared with 477 of 1,741 men taking choles- tyramine." It was concluded from these findings that reducing total cholesterol by lowering LDL- cholesterol levels can diminish the incidence of CHD morbidity and mortality in men at high risk for CHD because of elevated total and LDL-choles- terol levels. Another primary prevention trial using a lipid- altering drug has recently been reported.’ The Helsinki Heart Study was also a randomized, double-blind, placebo-controlled study involving 57 Population Panel Report 4,081 middle-aged men in a 5-year followup. To qualify, participants had to have concentrations of LDL-cholesterol plus VLDL-cholesterol greater than 200 mg/dL (5.17 mmol/L). The study as- sessed the efficacy of treatment with 600 mg of gemfibrozil twice daily in reducing CHD inci- dence. Treatment resulted in, on average, 10 per- cent reductions in total and LDL-cholesterol, a 35 percent decrease in triglycerides, and a 10 percent increase in HDL-cholesterol. The cumulative com- bined incidence of fatal and nonfatal myocardial infarction and cardiac death, the principal end- point, was reduced by 34 percent in the gemfi- brozil-treated group (p<0.02). Further analysis within the drug-treated group showed that while decreases in LDL-cholesterol levels were associated with a decreased risk of CHD, the subset in whom LDL-cholesterol did not decrease may also have benefited from the lower CHD rate, possibly be- cause of the associated HDL-cholesterol increase. The studies described above show the benefits of cholesterol lowering on clinical events. Their findings have been complemented by several re- cent studies that have focused attention on the impact of cholesterol-lowering on the coronary atherosclerotic lesion itself as seen through angio- graphic studies. Uncontrolled or partially-con- trolled angiographic studies using either diet or drugs to modify lipid concentrations had reported decreased progression of coronary lesions. These findings have been confirmed in two controlled trials, the NHLBI Type II Coronary Intervention Study and the Cholesterol-Lowering Atherosclero- sis Study.” 7% In the latter, marked lipid changes over 2 years produced by a combination of a resin (colestipol hydrochloride) with niacin therapy and a strict diet resulted in less progression, greater stabilization, and apparent regression of some coro- nary lesions. These benefits were observed both in aortocoronary bypass grafts and in native coro- nary vessels, and were irrespective of pretreat- ment cholesterol levels. More recent work by Brown et al.” lends additional support that ath- erosclerosis progression can be slowed and regres- sion induced by aggressive cholesterol-lowering therapy in men with coronary artery disease. Two major studies have reported on the effect of the simultaneous intervention on several CHD 58 risk factors including diet-induced cholesterol lowering. The Oslo study of diet and smoking intervention was conducted in 1,232 healthy normotensive men ages 40 to 49 years at risk for CHD with cholesterol levels of 290-380 mg/dL (7.50-9.83 mmol/L).*** Mean blood cholesterol levels were approximately 13 percent lower in the intervention group than in the control group dur- ing the trial. The mean tobacco consumption per man decreased by 45 percent more in the interven- tion group. The 5-year incidence of acute myocar- dial infarction was reduced by almost 50 percent in the intervention group. Statistical analysis sug- gested that the reduction in the incidence of myo- cardial infarction was correlated with the reduc- tion in total cholesterol levels and, to a lesser extent, with smoking reduction. The Multiple Risk Factor Intervention Trial was a randomized primary prevention trial of 12,866 high-risk men ages 35 to 37.%"' The special intervention (SI) group received treatment for hypertension, counseling for cigarette smoking, and dietary advice for lowering blood cholesterol. Despite an overall improvement in their risk factor status, CHD mortality of the men in the SI groups was reduced by only 7.1 percent compared with that in the usual care (UC) group. Analysis of blood cholesterol data from the SI group showed a moderate reduction in their average level over the course of the study. However, a reduction also occurred in the UC group so that the SI-UC dif- ference amounted to only 4-6 mg/dL (0.10-0.16 mmol/L), or 2 percent, about half of the original goal. Thus the ability of MRFIT to evaluate the benefit of cholesterol lowering was compromised. Analyses of a subgroup of the MRFIT participants that resembled the Oslo study participants sug- gested that diet and smoking interventions may have been beneficial: within the SI group there was a statistically significant reduction in the risk of CHD death in those who both had a reduced cholesterol level and reduced smoking.” A 10 mg/dL (0.26 mmol/L) reduction in cholesterol was estimated to be associated with a 6.6 percent reduction in the CHD death rate. Peto and colleagues,” using a meta-analysis technique, evaluated 18 published and 2 unpub- lished randomized trials of cholesterol lowering Population Panel Report by diet and/or drugs, as described previously. Analysis of all data from these trials indicates that a 10 percent reduction in cholesterol was associ- ated with an average reduction of 16 percent in CHD risk in trials lasting 4 years. Peto also noted that the reduction in CHD risk varied according to the duration of treatment: an average of 11 percent in 13 shorter trials and 21 percent in 7 longer trials, for a 10 percent reduction in blood total choles- terol. The findings in clinical trials are consistent with those of observational epidemiologic stud- ies.’ There have been no trials lasting a decade or longer. Peto concludes that results from the obser- vational studies suggest that a reduction of 10 percent in total cholesterol over decades would be associated with a 33 percent reduction in rate of fatal CHD. In a similar overview of 22 randomized trials of cholesterol lowering, Yusuf et al.** found significant reductions in coronary mortality and morbidity. For a standardized reduction in choles- terol level of a fixed duration, there was no appar- ent heterogeneity in the reduction of CHD ob- served in the drug or dietary intervention trials. Although cholesterol-lowering clinical trials have provided conclusive evidence that both nonfatal and fatal coronary events are significantly reduced, overall mortality has tended to remain unchanged. It has been difficult to draw sound conclusions regarding this issue, especially given the small numbers of events that are usually in- volved, and the lack of statistical power in the experimental design of many trials to detect changes in non-CHD mortality. In those studies in which noncoronary mortality has tended to in- crease, little consistency has been observed in the various causes of noncoronary death, which have included diverse accidents and violent deaths (homicides, suicides, road accidents), cancer, and miscellaneous other medical causes of death. In the World Health Organization study of clofibrate therapy, overall mortality was increased.” The question of whether an excessively low blood cholesterol carries adverse health risk is further considered later in this report. Some reassurance has been gained from the re- sults of the niacin-treated subgroup of the Coro- nary Drug Project. The Coronary Drug Project, a long-term study of lipid-influencing drugs in male survivors of myocardial infarction, was concluded in early 1975.3 At that time niacin treatment showed benefit in decreasing definite nonfatal recurrent myocardial infarction but did not de- crease total mortality. Later analyses, with a mean followup of 15 years, nearly 9 years after termina- tion of the trial, showed that mortality from all causes in the niacin group was 11 percent lower than in the placebo group (52.0 versus 58.2 percent; p=0.0004). This late benefit of niacin, occurring after discontinuation of the drug, is likely to be the result of a translation into mortality benefit over subsequent years of the early favorable effect of niacin in altering lipoproteins and decreasing the rate of recurrence of nonfatal myocardial infarc- tion. In another trial of pharmacologic reduction of blood cholesterol in patients after myocardial infarction, Carlson and Rosenhamer®® reported significant reductions of both cholesterol and triglycerides in the group receiving clofibrate and nicotinic acid. In this Stockholm Ischemic Heart Disease Secondary Prevention Study, total mortal- ity was significantly reduced, and CHD mortality was reduced to an even greater degree. In addi- tion, the Oslo Study Diet and Antismoking Trial has reported results after 102 months, describing significant reductions in sudden death and total coronary events, with marginally significant re- ductions in total mortality. Any single clinical trial has design features specific to it that limit its extrapolation. However, the broad consistency of the results of the many major cholesterol-lowering studies appears to have general implications. The cholesterol-lowering trials were designed to test the hypothesis that reducing elevated cholesterol levels would reduce the risk of CHD. Most studies support this hy- pothesis, indicating that reduction of blood 59 Population Panel Report cholesterol has a beneficial effect on CHD irrespec- tive of whether it is produced by cholesterol-low- ering drugs or by dietary means. These benefits are found in individuals with or without prior manifestation of CHD. The findings also suggest that benefits are seen whether pretreatment cho- lesterol levels are especially high or only moder- ately elevated. Finally, the greater the cholesterol reduction, the greater the reduction in risk. It was pointed out earlier that it is highly un- likely that an entirely satisfactory primary preven- tion study of diet-induced cholesterol lowering to reduce CHD can or will be conducted. In the absence of such a study it is, however, reassuring to note that the diet-induced cholesterol-lowering trials also appear to have reduced the risk of clini- cal CHD in proportion to the amount of cholesterol lowering that they have achieved, and have also been associated with a decreased rate of coronary lesion progression in angiographic studies.” When dietary change has been studied together with other interventions, the results have been more variable, but this may depend on the amount of cholesterol lowering achieved in these studies. On the basis of the experience of the clinical trials, a rule of thumb has been generated, namely that for each 1 percent reduction in blood choles- terol, an approximate 2 percent reduction in CHD incidence can be expected. This relationship has considerable public health and clinical implications. b. Community Trials Epidemiologic studies havelong demonstrated the strong relationship between culture, environ- ment, diet, and cardiovascular disease. Food patterns are strongly rooted in cultural, agricul- tural, and economic factors. Hence community- wide changes in eating patterns may be necessary to produce population changes in disease rates. In fact, Stamler*? suggests that the recent decline in CHD found in many industrialized countries may, to a significant degree, be the result of national shifts in food composition and consumption pat- terns as well as control of other risk factors. These observations have led to a number of experiments 60 to test approaches to educating entire populations on dietary factors that might contribute to lower blood cholesterol levels. While these community studies involve entire populations, most are ex- perimental in that comparison populations are studied to serve as a reference for ascertaining community trends. Multiple educational approaches have been used to attain the goals of these community pro- grams. Prevention of CHD—both through pro- motion of healthy eating patterns and through de- tection and treatment of those with high blood cholesterol levels—has been one important goal. Some studies have used the health care system both to detect high-risk individuals and to educate patient groups. Others have attempted to influ- ence food production and delivery by altering recipes and labeling food products to meet low- cholesterol and low-fat specifications. Mass cho- lesterol screening programs have been started and tested along with community education programs involving both direct and mass media approaches. All have a goal of testing whether CHD can be reduced in entire populations. Noteworthy among the several early studies of community programs is the North Karelia study, which began in the early 1970s. This project was a response to observational studies that had found very high blood cholesterol levels and CHD rates in this part of Finland. Ten-year followup studies demonstrated that the average blood cholesterol level was 3 percent lower in the county in which a cholesterol education program had been con- ducted, compared with a reference area,**** and that this was associated with declining CHD rates. The Stanford Three-Community Study, a project that used a mass media approach, demonstrated lower cholesterol levels in the adult population of the study communities compared with the refer- ence town.” Studies in Switzerland*” and Israel** likewise demonstrated lower cholesterol levels in the study community compared with reference areas. A Polish study was less successful, showing little cholesterol change, although other CHD risk factors were improved.*" There are also several ongoing community studies attempting to lower population Population Panel Report distributions of cholesterol. They all aim at both high-risk individuals and the general population. The Stanford experience has been extended to a larger five-city study.” Preliminary results show that cholesterol has been lowered an average of 6.0 mg/dL (0.16 mmol/L) in the two study communi- ties.®™ In Minnesota, large-scale systematic screen- ing programs supported by multiple community education strategies have shown significant low- ering of cholesterol (6.6 mg/dL [0.17 mmol/L]) in study communities compared with reference popu- lations? In Pawtucket, Rhode Island, a similar screening approach recruiting volunteers has shown significant reduction in cholesterol levels at followup.™ Each of these studies has used different strate- gies to effect dietary change across the population to produce blood cholesterol reduction. Data from these studies also suggest that blood cholesterol reduction can be achieved in people across the entire cholesterol distribution rather than just in those with very high blood cholesterol levels. 6. Other Risk Factors for CHD Other risk factors for CHD fall into two broad categories: “nonmodifiable” or “modifiable.” Increasing age, male sex, and family history of pre- mature CHD are the three major nonmodifiable risk factors.'”*** The three major modifiable risk factors are elevated blood cholesterol, high blood pressure, and cigarette smoking. !3!/163180.200.355357 Other significant modifiable risk factors include physical inactivity and obesity.'>* 53% A low level of HDL-cholesterol also increases CHD risk. The level of HDL-cholesterol, although less modi- fiable than the level of total or LDL-cholesterol, can often be altered by changes in smoking practice, body weight, exercise, and, to a limited extent, by changes in fatty acid intake. Diabetes mellitus is another risk factor, but the effectiveness of therapy in offsetting the increase in CHD risk due to diabe- tes is uncertain. These various risk factors appear to act in concert to increase risk with their effects at least additive and in some cases multiplicative. The three major modifiable risk factors for CHD contribute to the development of a large pro- portion of cases of CHD. As the level of each major risk factor rises, so too does the risk of developing CHD. For example, a blood pressure level of 160/ 95 mm Hg increases risk twofold compared with a level below 140/90 mm Hg, and a level of 200/120 mm Hg increases CHD risk fivefold, at any given cholesterol level. When all three risk factors are present to a moderate degree—for example, blood cholesterol of 240 mg/dL (6.21 mmol/L), blood pressure of 160/95 mm Hg, and one-pack-per-day cigarette smoking—the risk increases manyfold. When the levels of these risk factors are higher, the resultant multiplicative risk increases proportion- ately, rising as high as twentyfold. Almost one-third of all adults in the United States have high blood pressure, defined either as having a systolic blood pressure >140 mm Hg and/or a diastolic blood pressure >90 mm Hg, or as being on antihypertensive therapy.*' By age 65, more than 60 percent of U.S. adults have undesira- bly elevated blood pressure. Nonpharmacologic measures to produce healthful lifestyle habits are now recommended as the first line of therapy for high blood pressure.’ Such measures include reducing dietary consumption of sodium chloride and fat, maintaining desirable body weight, mod- erating alcohol intake, and being physically active. For many individuals, pharmacologic agents are also necessary in order to achieve the goal of maintaining blood pressure <140/90 mm Hg. However, some pharmacologicagents used to treat high blood pressure have adverse effects on serum lipoprotein levels." Such effects include in- creases in the levels of total cholesterol, LDL- cho- lesterol, and triglycerides, and decreases in the levels of HDL-cholesterol. These effects can be reduced or even prevented by modifying the quantity and composition of dietary fat intake, by selecting drugs without these effects, or by limit- ing the use of drugs with undesirable effects on lipids. The risk of developing CHD is also strongly influenced by cigarette smoking. The more ciga- rettes an individual smokes each day, the higher the risk. This increased risk from smoking holds true for both filtered and nonfiltered cigarettes.” Anyone who smokes one pack a day has almost 61 Popuiation Panel Report twice the CHD risk of a nonsmoker, and anyone smoking two packs a day has three times the nonsmoker’s risk. Cigarette smoking is consis- tently associated with lower HDL-cholesterol lev- els.* Smoking cessation is associated with in- creases in HDL-cholesterol levels.” Smoking also increases the risk of developing at least eight types of cancer as well as the risk of a stroke.*”** Recom- mendations to all people who smoke to quit are based on very solid evidence. The most prevalent of the other modifiable risk factors for CHD is physical inactivity. As with the major risk factors, the risk is graded in nature; that is, the more inactive a person is, the greater the risk. Conversely, the more active a person habitu- ally is, the lower the risk.” On the average, physical inactivity increases the risk of developing CHD by one and one-half to two times. Physical inactivity is associated with lower levels of HDL- cholesterol.” Inactivity also is strongly associated with excess weight, which frequently results in increases in the levels of total cholesterol, LDL- cholesterol, blood sugar, and blood pressure.’ Presence of obesity,”?*” defined as a body mass index (weight in kilograms /height in meters squared) of greater than 30, is a major concern. Four studies based upon life insurance data car- ried out between 1903 and 1979 have identified a quadraticrelationship betweenrelative weightand mortality, with the highest mortality ratios being found in those who were extremely obese. For instance, the Build Study** demonstrated a steep, curvilinear relationship between relative weight at entry and subsequent mortality among men age 15 to 39. This relationship is less striking for those in older age groups, which raises the question of whether weight control is more important in terms of overall morbidity and mortality for younger individuals than for those over the age of 50 years. 62 The relationship of obesity to CHD is complex. The Framingham study originally observed that obesity was predictive of the development of CHD and was also closely related to levels of blood total cholesterol, blood pressure, blood sugar, and uric acid. However, when obesity was included in an analytic model that included these other variables, it did not make an independent contribution to risk. After 26 years of followup of the original Framingham cohort, Hubert et al.,*”” however, reported that obesity did make an independent contribution to estimating the risk of cardiovascu- lar disease. They also observed that obesity did not exert its effect exclusively through its associa- tion with serum lipids, blood pressure, and blood sugar. Some other prospective studies of cardiovas- cular diseases have not been able to show such a clear relationship. For instance, the data from eight cohort studies of the U.S. Pooling Project were inconsistent—some showed no relationship, others a U-shaped association, and only one con- firmed the findings of the Framingham Study indicating an independent risk factor status of obesity. However, the pooled data did reveal a positive relationship.”! The most powerful effect of obesity on both cardiovascular morbidity and mortality is proba- bly due to its effect on other cardiovascular risk factors. Not only does obesity increase blood pres- sure, blood sugar, and the risk of noninsulin- dependent diabetes mellitus (NIDDM), but it is as- sociated with an increase in the LDL-cholesterol levels, a decrease in the HDL-cholesterol levels, and an increase in the fasting triglyceride lev- els.?**”> There is a consistent correlation between body mass index and total blood cholesterol level. Alow level of HDL-cholesterol has been linked to an increased risk of developing CHD, particu- larly in persons over age 50." Conversely, a high level of HDL-cholesterol is associated with a re- duced CHD risk. HDL-cholesterol levels <35 mg/ dL (<0.91 mmol/L) are associated with four times the risk compared with levels >75 mg/dL (>1.94 mmol/L). Accordingly, a level of HDL-choles- terol below 35 mg/dL (0.91 mmol/L) has been classified as a major CHD risk factor by the Adult Population Panel Report Treatment Panel of the NCEP.> HDL-cholesterol levels tend to be higher in women than in men and in individuals who are physically active, do not smoke, and are lean. An association of high HDL levels with alcohol intake is also recognized, but it is unclear whether the net effect on health is bene- ficial even at low levels of intake. On the other hand, the undesirable social, economic, and health consequences of high alcohol intakes are well rec- ognized. Accordingly, alcohol has not been rec- ommended as a means of raising HDL- cholesterol.” Diabetes mellitus substantially increases the risk of developing CHD both in men and in women.” This increased risk holds true for both insulin-dependent diabetes and for NIDDM. There are approximately 5 million known people with diabetes mellitus in the United States today. It is estimated that there are an additional 5 million individuals with diabetes mellitus who have not yetbeen diagnosed. More than 85 percent of diabetic persons have NIDDM, the major pre- cipitating cause of which is excess weight. As with nondiabetic individuals, the presence of blood lipid abnormalities in diabetic people in- creases the risk of developing CHD and of stroke.” However, the incidence of these abnormalities is increased in poorly controlled diabetes.” Specifi- cally, blood levels of triglycerides, cholesterol, VLDL-cholesterol, and LDL-cholesterol all tend to be higher and HDL-cholesterol tends to be lower with poor glycemic control. When glycemic con- trol improves, the level of HDL-cholesterol usu- ally rises, and the levels of the other lipids and lipoproteins usually fall. Because of the influence of glycemic control on lipid levels, itis most appro- priate to measure lipids in diabetic people after good glycemic control has been achieved. In the past, it was generally recommended that diabetic persons follow a diet relatively high in proteinand fat. It was thought thata relatively low carbohydrate diet would improve the degree of glycemic control. However, because of the high incidence of CHD in diabetic individuals and because of new evidence concerning the beneficial effects of increased carbohydrate and fiber on glycemic control, dietary recommendations have recently been significantly modified by the Ameri- can Diabetes Association (ADA) nutrition task force. The revised recommendations are similar to those of the ATP report. Specifically, all diabetic persons are urged to follow a diet with about 30 percent of the calories from fat, with less than 10 percent of those calories to come from SFA. For those individuals whose blood cholesterol levels remain high, the ADA recommends that SFA be reduced even further. The American Diabetes Association recom- mends that obese diabetic persons lose excess weightand seek to maintain their desirable weight, thereby often correcting high levels of blood cho- lesterol, blood pressure, and their underlying diabetes. Finally, recommendations have recently been made to have diabetic persons increase their fiber consumption substantially, in an effort both to reduce blood cholesterol levels and to improve glycemic control. In summary, there are numerous risk factors in addition to elevated blood cholesterol that play an important role in the development of CHD. 7. Low Blood Cholesterol and Risk An important question is whether there are any inherent undesirable health effects that may come from an excessive lowering of blood choles- terol levels. Concern over a possible relationship between low blood cholesterol levels and risk of cancer was raised by the report of Pearce and Day- ton.” Their clinical trial of the effects of a diet high in polyunsaturated fat on blood cholesterol showed more carcinoma deaths in the diet groups. Shortly after publication of those results, an analysis of five clinical trials relating to cholesterol lowering was published.” The conclusion reached was that cholesterol-lowering diets do not influence cancer risk, but the need for more data was expressed. Since 1971, there have been a number of publi- cations concerning low blood cholesterol and cancer risk.’ McMichael et al.* reviewed the 63 Population Panel Report data available through 1983. An inverse relation- ship between blood cholesterol levels and cancer risk was found in 12 studies, but notin 8 others. All the reports cited had originally been studies of cardiovascular disease. The general characteristics of these studies were that they all related to men. The cancer most often observed in these studies was colon cancer. Several of the studies showed an inverse relation- ship that did not attain statistical significance. One study™ found a significant inverse relationship with total cancer mortality. In about half the studies, the inverse relation between blood choles- terol and cancer was observed during the early followup period, suggesting that the low choles- terol levels may have been due to preclinical ef- fects of cancer. However, in a few studies the as- sociation with low cholesterol persisted for 10 years. This evidence has prompted the suggestion that males with naturally occurring low blood cholesterol may be at increased risk of colon cancer. Since 1984, several studies concerning the rela- tionship between blood cholesterol and non- cardiovascular disease risk have been reported. Martin et al." presented a graph in which age- adjusted 6-year total mortality and CHD mortality were plotted against serum cholesterol for the 361,662 men screened for the MRFIT program. The CHD mortality curve rose continuously from 140 mg/dL (3.62 mmol/L) of serum cholesterol. The curve representing total mortality rates showed the lowest rates for an average cholesterol level be- tween 165 mg/dL (4.27 mmol/L) and 185 mg/dL (4.78 mmol/L). A smallincrease in total mortality rate occurred in the men with cholesterol levels below 165 mg/dL (4.27 mmol/L) although the CHD death rate was lower. Ten-year followup data on this group of men show similar results.’ Schatzkin et al.®**" have analyzed blood choles- terol in the NHANES 1 followup, using data de- rived from a cohort of 5,125 men and 7,363 women. Among men, cancer (all site) morbidity and mor- tality was somewhat higher in the bottom two quintiles of cholesterol, namely 183-205 mg/dL (4.73-5.30 mmol/L) and below 183 mg/dL (4.73 mmol/L). In women, cancer mortality and 64 morbidity was higher only in those groups whose cholesterol levels were below 179 mg/dL (4.63 mmol/L). Wingard et al.” found no relation between cholesterol levels and cancer morbidity or mortality in 4,035 Californians. Hiatt and Fire- man™? examined data on more than 160,000 men and women in a California health plan. Their con- clusion was that low blood cholesterol did not increase the risk of cancer but that preclinical cancer reduced blood cholesterol levels. Mannes et al.” have actually suggested a small positive association between blood cholesterol and colo- rectal cancer. Tornberg et al.” have found a similar association. On the other hand, Salmond et al.” in reporting on 630 New Zealand Maori followed over 17 years, found significant inverse relationships between cholesterol and either can- cer or all-cause mortality in men and women and suggest that the association may not be due to undetected illness. Neither the LRC-CPPT** nor the Helsinki Heart Study* showed an increase in incident cancer causes in treatment compared to control groups. Overall, the evidence available in- dicates that dietary change to lower blood choles- terol is highly unlikely to increase the risk of colon cancer; indeed, there is some epidemiologic evi- dence suggesting that a habitual diet high in fat and SFA (which would be expected to raise the blood cholesterol level) increases the risk of colon cancer.’ Several studies suggest that hemorrhagic stroke occurs more often in population groups with low levels of blood cholesterol. Hemorrhagic stroke is less common in Americans than in some other groups.’®*” The relationship between hemor- rhagic stroke and low blood cholesterol has been noted in people of Japanese heritage both in Ja- pan™® and in Hawaii.” A similar relationship has been noted among women in the Framingham Study,” and in the men screened for the MRFIT study. MRFIT results are particularly notewor- thy. There was a statistically significant increased likelihood of hemorrhagic stroke, especially intracranial, in men in the lowest quintile (<160 mg/dL or <4.14 mmol/L) of total blood choles- terol. Overall, men with nonhemorrhagic strokes had a significantly higher average blood choles- terol level than did those who experienced no Population Panel Report stroke during the 6 years of followup. The average blood cholesterol levels of those with subarach- noid or intracranial hemorrhage were slightly, but not significantly, lower than those who had no stroke. Men with any of the types of cerebral lesions were more likely to be smokers and to have elevated blood pressure. Given that the age-ad- justed death rates for CHD in the MRFIT screenees were 60.5 per 10,000 men, compared with 5.48 for nonhemorrhagic stroke and only 2.36 for intracra- nial hemorrhage, and given the clear curvilinear relationship between CHD and serum cholesterol levels (figure 1, page 7), Iso and colleagues*” con- clude that the small risk of a low blood cholesterol level in American men is far outweighed by the risks from CHD and nonhemorrhagic stroke asso- ciated with even mild elevations of blood cholesterol. The benefit-risk relationship is important in any public health initiative. While the relation- ships among diet, blood cholesterol, and disease require further research, the data suggest that any additional risk of blood cholesterol lowering is either nonexistent or very small. The issue of potential competing risks and benefits associated with dietary changes has been discussed in detail in the report Diet and Health: Implications for Reduc- ing Chronic Disease Risk.” That report concludes that the public health benefits that will result from dietary changes to lower blood cholesterol greatly exceed any small adverse effects that might also occur. B. Eating Patterns in the United States 1. Current Dietary Patterns in the United States In September 1989, the ad hoc Expert Panel on Nutrition Monitoring (EPONM), a committee es- tablished at the request of the U.S. Department of Agriculture and the U.S. Department of Health and Human Services, issued a report containing current data on the nutritional status of the United States population.*! The EPONM concluded that the principal nutrition-related health problems in the United States are associated with excess con- sumption of food energy (calories), fat, saturated fatty acids, cholesterol, sodium, and alcohol. The report also stated that one-fourth of the adults in the United States are overweight"! Planning for achievement of the nutrition objectives for the nation are consistent with these nutrition-related problems. *” The EPONM report noted that certain sub- groups within the U.S. population had low intakes of iron and vitamin C, including young children and females of childbearing age, especially those who were black, poor, or both. Concern was also expressed that calcium intakes of young women may be inadequate for the development of maxi- mal bone mass. In addition, intakes of zinc, vita- min B_, and folacin were lower than recommended levels in some groups, especially women. 2. Fat in the American Diet The Continuing Survey of Food Intakes by Individuals" showed that the fat intake by men and women 19 to 50 years of age averaged 36-37 percent of total energy intake. Using a “food group” approach which classifies foods by their major ingredients, about 50-55 percent of the fat and 55-60 percent of the SFA was estimated to come from meat, poultry, fish, egg and dairy prod- ucts, and from mixed dishes containing these ani- mal products as major ingredients.” These esti- mates include fats of other origins such as frying fats and salad dressings used in preparation of these foods. The estimates exclude the fatand SFA from the animal products in a mixed dish included in another food group, for example the meat and cheese included in spaghetti or pizza. An “ingre- dient approach” that separates foods into their in- gredients has also been used to estimate sources of fat, although the analysis is limited to women’s diets. These preliminary data show that about 50 percent of the fat and about 59 percent of SFA came from meat, poultry, fish, egg, and dairy products. The ingredient approach also showed that fats and oils added at the table or in food processing and preparation accounted for at least 33 percent of the fat in women’s diets." 65 Population Panel Report The palm, palm kernel, and coconut oils (so- called “tropical oils”) contain as high or higher proportions of SFA’s than animal fats." In 1985 these tropical oils provided about 3 percent of the total amount of fat in the U.S. food supply.*” Current intakes of the three tropical oils combined are estimated to be about 2 to 4 grams per person per day. These estimates represent less than 4 percent of total fat intake, less than 2 percent of the daily energy intake, and 8 percent or less of the SFA intake by the U.S. population.*” In the recommendations for dietary change to lower blood cholesterol level, much of the positive contribution of animal products to vitamin and mineral nutritional status can be maintained. This can be achieved by placing emphasis on lean and lower-fat animal product choices and by limiting portion sizes, rather than avoidance or elimination of foods of animal origin.*” 3. Change Required in American Diet To Achieve Various Targets Current dietary intakes of Americans have been compared to those recommended by the American Heart Association’s Step-One diet (which is identical to the Step-One diet of the NCEP Adult Treatment Panel).>** If in fact the entire popula- tion changed its eating patterns to meet the AHA / NCEP Step-One recommendations, intakes of the target nutrients, total fat, and saturated fatty acids would have to be altered significantly downward from current consumption levels." It is likely that such changes are feasible in the future, given tech- nological improvements in food processing and preparation that facilitate lowering dietary fat and SFA while maintaining acceptable palatability and nutritional adequacy. 4. Dietary Safety and Adequacy Hansen and Windham*” have expressed con- cern that if fat and cholesterol were decreased to levels recommended in the Step-One diet, defi- ciencies might occur in certain “problem” 66 nutrients. These nutrients are regarded as prob- lematic because intakes are already low in Ameri- can diets in comparison with nutrient standards such as the recommended dietary allowances (RDA) of the National Academy of Sciences.” These problem nutrients include zinc, iron, cal- cium, and folacin (folic acid). The proportion of the population consuming less than the RDA for iron and folacin will be reduced substantially with the reduction in several of the relevant RDA's in the 1989 recommendations. Choosing low-fat ani- mal foods permits utilization of the highly bioavailable sources of those nutrients in animal foods. Adequate attention to nutrition education for the public will help ensure achievement of RDA levels including those for zing, iron, calcium, and folacin. Overall, there is extensive epidemio- logic evidence that low-fat diets are not harmful to health; this has been carefully reviewed recently.” Menus for adults that meet the RDA's for all nutrients while meeting low-fat, low-SFA, and low-cholesterol targets can be created by helping individuals make informed food choices. For example, Dougherty et al.*” found, with diets planned and served in a metabolic ward, that it was possible to improve intakes of vitamins and minerals while decreasing SFA, total fat, and cho- lesterol intake to target levels. Usual intakes of meats, milk products, fish, and eggs were sus- tained. Intakes of vitamins B, B,, B,, B ,, and C in- creased, as did potassium, calcium, magnesium, phosphorus, zinc, and iron. Eating patterns de- signed to lower blood cholesterol can be as high or higher in nutritional quality as other diets." This also applies to diets of children and ado- lescents. The feasibility of such menu planning for children and adolescents, with their increased needs for nutrients per calorie, has also been demonstrated previously using the 30 percent fat, 10 percent SFA, and low dietary cholesterol speci- fications.” Moreover, observational studies indi- cate that even children who are vegetarian and consume diets very low in fat and cholesterol can achieve and maintain good nutritional status, whereas those who are fed nutritionally unplanned diets are the individuals who may fall into difficulties." Population Panel Report 5. Special Concerns Involving Women Although the nutrient densities of men’s and women’s diets are similar, the considerably smaller energy intakes of women, coupled with increased requirements for some essential nutrients such as iron and calcium, highlight the need for guidance for women to ensure nutritional adequacy. This becomes even more important with the reduced energy intakes that occur as women become older. Information on the reported dietary intake of women, available from the USDA Continuing Sur- vey of Food Intakes by Individuals (CSFII) isshown in table 5, page 68."'> For women ages 19-50 years, the percentage of calories from fat was 37 percent. The corresponding percentage distribution of calo- ries from fat was as follows: SFA, 13; MUFA, 14; and PUFA, 7. Only 10 percent of women ages 19- 50 years were below the level of 10 percent of calories from SFA. About 65 percent of women at- tained a level of 10 percent or more of calories from MUFA, while 86 percent of women consumed less than 10 percent of calories from PUFA. About 62 percent of women ages 19-50 years had diets that provided 300 mg or less of cholesterol per day. As previously described, the EPONM*" con- cluded that women tend to have a lower calcium intake than recommended and that intakes of iron and vitamin C do not meet RDA's for some women of childbearing age. According to the 1985 USDA CSFII data—for calcium-78 percent of women did not meet 100 percent of the RDA; and 54 percent did not achieve 70 percent of the RDA; several other nutrients—including B, and folacin—were below recommended levels in diets of many women. The new RDA standards for folacin and iron will mean that more women meet these standards.’ According to NHANES II,> more than 28 per- cent of U.S. adult women have blood cholesterol levels at or over 240 mg/dL (6.21 mmol/L). More than 58 percent of women have blood cholesterol levels that are above 200 mg/dL (5.17 mmol/L). These data on the prevalence of high blood choles- terol and data on the nutrient intakes of women, particularly adolescents and women of childbearing age"? emphasize the importance of developing detailed dietary recommendations to lower blood cholesterol levels in women. Cleve- land*"” used a computer-generated model to esti- mate changes in women's reported intakes to meet a target diet containing 1,600 calories, <30 percent of calories from fat, <10 percent of calories from saturated fatty acids, <300 mg/day cholesterol, over 50 percent of calories from carbohydrates, 3,300 mg/day sodium, 20 g/day fiber, and 100 percent RDA for vitamins and minerals. Within major food groups, the model relied on lower-fat, leaner choices. A pattern meeting the specified nutritional criteria contained at least twice as much low-fat milk, lean red meat, legumes, whole-grain bread, cereal, and vegetables as the estimated usual consumption pattern. It contained almost no bacon, sausage or lunchmeat, cheese, grain- based snacks, desserts, fats added at the table, or sugars and candies. C. What Influences Eating Patterns? 1. Overview Various factors influence eating patterns and ultimately many aspects of health.*'* These factors fall into four major groups: food composition (e.g., genetics, manufacturing); food supply (production, processing, distribution, cost, and many other factors affecting accessibility); food demand (in- come, relative price structures of food and other needs, education, cultural influences, food habits, food taste, taxes, advertising, public relations and consumption stimulation, programs sponsored by government or special interest groups); and nutri- tional needs (factors influencing either the individ- ual's state of health or biological utilization of food). Figure 7, page 70, lists many of the factors that influence food consumption, and the pages that follow describe much of the existing research on these factors. Some, but not all, of these factors have been carefully studied, and some have been subjects of intervention research. Other factors, such as new methods of distributing, preparing, or cooking food, have not been carefully analyzed, 67 Population Panel Report Table 5 Mean Daily Intake? of Food Energy, Nutrients, and Food Components for Men, Women, and Young Children from the Continuing Survey of Food Intakes by Individuals (CSFII), 1985* Men Women Children Total Food Energy (% REI)¢ (94) (82) (100) Fat [% total energy] Total fat [36] [37] [34] Saturated fatty acids [13] [13] [14] Monounsaturated fatty acids [14] [14] [12] Polyunsaturated fatty acids [7] [7] [6] Cholesterol mg 435 304 254 Protein [% total energy] [16] [16] [16] (% RDA) (175) (144) (222) Carbohydrates [% total energy] [45] [46] [52] Dietary Fiber g 18 12 10 Vitamins (% RDA) Vitamin A (122) (127) (215) Vitamin E (98) (97) (108) Vitamin C (182) (133) (186) Thiamin (124) (110) (153) Riboflavin (129) (115) (197) Niacin (146) (130) (151) Vitamin B, (85) (61) (127) Vitamin B,, (245) (156) (192) Folacin (76) (51) (157) Minerals (% RDA) Calcium (115) (78) (105) Phosphorus (192) (126) (132) Iron (159) 61) (88) Zinc (94) (60) (84) Magnesium (94) (72) (121) Minerals (ESADDI)® Sodium (exceeds) (within) (exceeds) Potassium (within) (within) (1-3 years exceeds) (4-5 years within) Copper (below) (below) (below) * The Surgeon General's Report on Nutrition and Health. U.S. Department of Health and Human Services. Public Health Service USDHHS (PHS)Publication No. 88-50210, 1988.4'° a Estimated mean daily intake is expressed in several ways: amount of intake, percent of total energy intake, percent of 1980 Recom- mended Dietary Allowance, or comparison with Estimated Safe and Adequate Daily Dietary Intake. b Data based on 1-day dietary recalls obtained by personal interview for 658 men 19 to 50 years of age, for 1,459 women 19 to 50 years of age, and for 489 of their children 1 to 5 years of age in 1985 (unweighted numbers). Nutrient intakes do not include vitamin and mineral supplements or sodium from salt added at the table. ¢ Recommended Energy Intake;*'' Source of percentages: NFCS, CSFII Report Nos. 85-1 and 85-3." 9 Recommended Dietary Allowance;*"' Source of percentages: NFCS, CSFII Report Nos. 85-1 and 85-3." © Estimated Safe and Adequate Daily Dietary Intake.*'! i Figures given are based upon 1980 RDA's; the 1989 RDA's for these nutrients are substantially lower. 68 Population Panel Report even though some of them, such as microwave cooking, have potentially major effects. 2. Factors That Influence Eating Patterns and Approaches to Changing Them a. Food Supply Food producers and processors influence the food supply in a variety of ways, which are briefly considered in the following section. Many of these food supply factors are amenable to change by societal or industry-wide efforts but are less influ- enced by individual efforts. e Food Production, Processing, Distribution, and Marketing The production, processing, distribution, and marketing of food influences eating patterns in ways that may either promote or detract from cholesterol-lowering efforts.*> Designing Foods" describes some of the many barriers that hin- der the animal agriculture industry from achieving a food supply that is lower in fat and cholesterol and suggests ways of overcoming these barriers. e (Consumer Information on Foods The American food supply is one of the most abundant, varied, and economical in the world. With the average supermarket selling more than 24,000 items, the consumer is faced with a myriad of food choices. For individuals who wish to alter the type and amount of fat and the amount of cholesterol in their diets, some of these food choices are more appropriate than others.*® The problem is not that suitable foods are unavailable, but that it is difficult while food shopping to know which choices are best.*!” Increased access to information to guide food choices can help.*"® The knowledge, skills, and behavior of those who prepare foods is also important. * Ecosystem On a worldwide basis, the ecosystem—as de- termined by geography, climate, labor, capital, and other factors—is critical in shaping the ag- ricultural base and the food supply. It is a potent and changing influence on food availa- bility and thus on eating patterns. b. Societal Factors Influences on eating behavior are of two differ- ent kinds: those that operate at the societal level, over which individuals have relatively little influ- ence, and those at the personal level, over which individuals exert more control. This review em- phasizes factors subject to individual control in addition to societal policies and practices that affect eating and lifestyle behaviors. This empha- sis is not because changes at the societal level are unimportant, butbecause, in the short run, changes are more likely to occur first at the individual level. Collectively, such changes at the individual level foster society-wide change since they influence demand and strengthen the political will that is necessary to change policies and regulations in the public, private, and voluntary sectors. Eating pat- terns are also affected by policies related to the food chain (food production, agriculture, food science, economics, food manufacturing or distri- bution) and to health care, as well as to broader economic and educational factors. e (Cultural Factors Culture is a major influence on food selec- tion.**#! Culture influences food-related be- havior—the ways in which people select, consume, and use the available food supply in response to social and cultural pressures. #042 A great deal of intra- and interindividual vari- ability exists in eating habits even in this coun- try, which has a relatively consistent national food supply.*” In a given society, there are subcultures that have very different norms and values, which may influence the effective- ness of health-related interventions (see sec- tion VI, Ethnic, Cultural, and Minority Charac- teristics That Influence Diet and Health). 69 Population Panel Report Figure 7 Eating Pattern Determinants Food Food Food Food Science Manufacture Distribution Purchasing Base Nutrition Agriculture Wholesale Cost Biochemistry Synthesis Retail Culture Preservation Processing Product- Advertising Genetics Additives Specific Knowledge Modifiers Route- Health Hybridizers Specific Mass Preparation The boxes depict the major elements in the food chain from the food science base to the outcomes of food consumption. Major influences on the food chain, each with an impact on several of the food chain steps, are listed. Food Food Food Preparation Consumption Consumption Outcomes In-House Socialization Pleasure Restaurants Education Health Institutions Nutritive Deficiencies Value Surpluses Health Culture Taste Cost Mood Major influences on the food chain: Agribusiness Conglomerates Grocery Chains Advertising Profitability Special Interest Groups Media Government Agencies Health Professionals General Education Nutrition Education Food Science Education Culinary Education 70 Other Societal Influences A wide variety of other societal factors influ- ence food demand. Economic factors include stimulation of consumption, rationing, and dif- ferential taxes (such as taxes on alcohol and cigarettes). These and similar measures have attempted to alter demand for food, some- times in directions counter to an objective of lowering blood cholesterol levels. Some pro- motional efforts to stimulate consumption are viewed as successful enough to be supported on a continuing basis by producers. Examples include programs of the National Dairy Re- search and Promotion Board, the Egg Board, the Beef Check Off Program, the National Dairy Council's Bridge Project, and some programs administered by the Government with contri- butions by producers. Several of these have emphasized lower fat and lower cholesterol consumption options. Population Panel Report e Political Factors Historically, many attempts have been made to influence what people eat by changing poli- cies, laws, and regulations. More recently, public health research and practice have turned toward efforts to modify eating behavior by modifying the environment within institutions and communities. If the political will is present for making changes, such measures can be very effective, as recent smoking-related inter- ventions have shown. Political factors can both hinder and foster healthful changes in food consumption. In making nutrition policy, political, social, and economic considerations must be weighed along with the science base. Therefore, it is important to distinguish clearly between ad- vice based on scientifically established evi- dence, deductions from indirect evidence, and conjectures based on assumptions, no matter how logical they seem to be.*** ¢. Consumer Characteristics Influencing Eating Patterns Historically, most dietary interventions have focused on modifying the knowledge, attitudes, and behavior of individuals. Recently, strategies to achieve changes in eating patterns have ex- panded to include a greater consideration of the social and cultural characteristics of the individ- ual's immediate or microenvironment, with re- spect to influences on eating behavior and health status.” Alteration of the microenvironment (in contrast to the larger environment) is more ame- nable to individual control. Many theoretical models that seek to identify predictors of food behavior in individuals have not been validated. They lack the specificity re- quired to quantify or qualify the various influ- ences and how they relate to a specific individ- ual. In general, it is thought that two influences on food-related behavior are sociodemographic and psychosocial factors. These are reviewed briefly in the following discussion. Social and Demographic Influences on Eating Behavior Income. When personal income increases, the relative amount of money spent on food pur- chases decreases in comparison with other ex- penses.*” This phenomenon, known as income- related elasticity of demand, is uncommon in the United States, presumably because the food supply is plentiful and relatively inexpensive.* Some foods have higher income elasticities than do others. Foods with high income elas- ticities include meats, fresh fruits, and vege- tables.** Foods with low income elasticities include eggs, cereal products, fresh milk, and bread.*”” Purchase of dried beans, peas, rice, and eggs is inversely related to income. Two otheritems, low-fat milk and whole-grain bread usage, are positively related to income, per- haps reflecting the growing concern about health in the higher socioeconomic groups.*” Nevertheless, the effects of income on food consumption patterns are still apparent. For example, data from the Nationwide Food Con- sumption Survey (NFCS) demonstrate that high-income households spend five times as much as low-income households on food eaten away from home.*” Household Size. Larger households spend more money on food than smaller households, even at the same income level. However, the per- person food cost decreases as the number of family members increases. *** The net effect on food purchasing behavior of larger house- hold size is the same as a decrease in income elasticity. Thus, income spent on food eaten away from home is lower in larger households. In the 1977-78 NFCS, the foods most respon- sive to household size were fresh milk, dairy products, cereal products, bread, bakery prod- ucts, sugar products, potatoes, fats and oils, and eggs, with some of these increasing and some decreasing as size of household increased. Surprisingly, meat usage remained remarka- bly high despite its high income elasticity, whereas the use of fresh vegetables, fresh fruits, and fruit juices decreased as household size increased." 71 Population Panel Report 72 In regard to nutrient density, Windham et al." reported that households with five or more members had a slightly but significantly lower fat intake than smaller households (difference of 2 grams per 1,000 calories). Households with three or more members had significantly higher carbohydrate intakes per 1,000 calories. In contrast, intakes of vitamins C and B, per 1,000 calories were significantly and inversely related to household size, as might be expected from the lower income elasticities for fresh fruits and vegetables in larger households. Education. A number of studies have evaluated food behavior as a function of educational level for both males and females.*"** The most commonly reported association is a positive relationship between educational level of a female head-of-household and adequacy of dietary intake. Education-related elasticities of demand for fruits and milk are even higher than their income elasticities. Educational level has been positively correlated with nutrition knowledge, #74" but educational level and dietary behavior are not always positively re- lated”? In more educated households, use of "convenience" foods is lower,*’#! more meals are eaten together,*? and more meals are eaten away from home. "04? Gender. Gender differences in food and nutri- ent consumption in the United States are not striking. Women, because of their smaller size and lower energy needs, eat less than men, but the variety of foods they select is generally similar.*’#? When nutrient densities are com- pared, women report consuming diets higher in vitamins A and C per 1,000 calories than men. Gender differences also have been noted in the Coronary Artery Risk Development in Young Adults Study (CARDIA), which is ex- amining longitudinally the development of risk of coronary artery disease in 18- to 30-year- old black and white adults. Preliminary data from this study reveal that white women re- port dietary intakes that are more consistent with dietary recommendations for lowering high blood cholesterol than do males or black women. ** In comparing male and female food usage using 1977-78 NFCS data, Cronin and colleagues found no differences in 54 out of 65 food groups examined. In the 11 remaining groups, women reportedly ate more citrus fruit, yogurt, coffee, tea, and low-calorie beverages over the 3-day recording period, whereas men ate more whole milk, luncheon meats, meat, fish, and poultry sandwiches, desserts, sugar, and sweet spreads.** Women reportedly pre- fer fruits and vegetables more than men do, but within a specific culture such as that of the United States, the relationship between stated food preferences and food consumption is only low to moderate. *>#¢ Age. There are few longitudinal studies report- ing the effect of aging on dietary intake. Most studies are cross-sectional, and thus they re- flect not only the effect of age but also the effects of the dietary habits of different age cohorts. In the data reported, changes in nutri- ent intake patterns are generally consistent with the decreases in energy intake that occur with aging.*7*“ This is probably related both to a tendency of the elderly to be more seden- tary and to a slowing of metabolism with ag- ing. Data from NHANES land Il are currently being evaluated by age, gender, and ethnic background.” Preliminary results reveal that within ethnic groups, intakes of dairy prod- ucts, meats, and low-calorie beverages are consistent throughout an age range within the population. The greatest differences appear in intakes of sugar-rich products, with those under 25 years of age consuming more of them. Cultural, Ethnic, and Minority Considerations. Many important factors relating to health and nutrition in population subgroups have been described. Many of these differences are proba- bly related to income rather than ethnicity, because income and ethnicity often vary si- multaneously. Regional factors may also be operating.”® Many of these differences are discussed in section VI, Ethnic, Cultural, and Minority Characteristics That Influence Diet and Health; others are considered here. As time in residence in a new country increases, the food habits of cultural subgroups Population Panel Report gradually assume the characteristics of the dominant culture. Migrant studies support this observation, 2% including studies of first generation Chinese living in California, who reported reduced consumption of crab, oys- ters, prawns, and other foods popularin China, even though the foods were widely available in California as well.*! Recent population- based surveys have reported differences in food group usage by age and race based on the NHANES II data.” The greatest black-white differences appear to be in the usage of dairy products. Blacks reported lower intakes of dairy products and increased intakes of sugar-rich products from age 12 onward.”” This phe- nomenon may reflect, in part, dietary altera- tions resulting from lactose intolerance, which is common in blacks. Increased intake of other sugars is not easily explained on a physiologic basis, yet the NHANES II data suggest that blacks eat more sugar-rich products, jelly, jam, candy, and low-calorie beverages (diet soft drinks, coffee, tea, artificially sweetened drinks) than whites throughout life.” Data from the Consumer Expenditure Survey of 1972-74 indicated that, compared with whites, blacks purchased more beef, pork, poultry, fish, and seafood but fewer cereal and bakery products, sugar-rich products, dairy products, and nonalcoholic beverages.*> While some other studies confirm that blacks pur- chase more of these products than whites, evidence from NHANES II?” is not as strong. For other ethnic groups, there are few data on dietary intake of either a cross-sectional or lon- gitudinal nature. Analyses of Hispanic NHANES data are now in progress. One study of Mexican-Americans reported that they con- sumed greater amounts of eggs, white bread, and tortillas but less dry cereal, pastry, wine, and beer than did a reference group of non- Mexican-Americans.** Both groups reportedly consumed the same amount of "convenience" foods, soft drinks, coffee, and tea.*>® These au- thors concluded that Mexican-Americans adopted eating patterns that were more like the stereotypical American patterns of several years ago, prior to the widespread interest in nutrition and health, than eating patterns re- flecting either their culture of origin or their culture of residence. Employment Status of Female Spouse. The Ameri- can family is changing; as more women work, the associations between female employment status and food-related behavior are of inter- est.*’ Most studies have reported that as the number of hours at work increases, and the number of hours spent doing housework de- creases, people spend less time in meal prepa- ration, +45 eat more meals away from home, and use more convenience foods. *0:442:457.458 However, not all studies have found this to be the case.*!! Women's employment does not appear to be related to other factors such as food prepara- tion style or the number of meals eaten to- gether as a household. These factors are more strongly associated with the age of the children of the household.*” The few studies that evalu- ate the relationship between quality of dietary intake from the standpoint of nutrients and employment status of the female head of the household #%4 reveal no link between the two. There are few data on the effects on family nu- tritional status when households are headed by female single parents." Psychosocial Determinants of Nutritional Status Nutrition Knowledge. Nutrition knowledge in- fluences food-related behavior, but the effect does not appear to be large. *31433437-439461-465 Methodological inconsistencies may account 73 Population Panel Report 74 for the weakness of the positive correlations observed in most studies between nutrition knowledge and dietary intake.*’** Studies in which the food-related behavior of interest is specifically defined show the strongest asso- ciations with nutrition knowledge. Little infor- mation is available on the extent to which nutrition knowledge affects eating behavior among those who are deliberately attempting to alter their diet for health reasons. Attitudes. Attitudes toward food and eating derive from a variety of different beliefs, feel- ings, and intentions, which are poorly docu- mented but are thought to be important deter- minants of food habits. Food attitudes are also influenced by general health, health concerns, aesthetics, economics, convenience, creativity, adventuresomeness, sociability, prestige, fa- miliarity, meal planning, and preparation styles.*” Cultural and ethnic considerations also play an important role.*” Correlations between food-related attitudes and eating behaviors are highest when the attitude and outcome measures are specifically defined. Childhood Influences on Eating Behavior. Child- hood influences on diet and eating behavior may carry over into adulthood. Biological, cul- tural, and psychological factors play a role in the food behavior of children.*” During the first year of life, children learn what is edible and what is not. Once the edible foods have been identified, issues of taste preference arise. Some workers have suggested that there is an innate, but variable, preference for sweet tastes thatis exhibited atbirth and continues through- out life for most people.*®4 It was long con- tended that infants left to their own choices would continue to have normal growth #04! More recently, it has been recognized that this is not necessarily the case and that "hidden hungers" or the "biological wisdom of the body" do not ensure good nutritional status.*’ In general, culture has the greatest impact on children's food choices and eating behavior. ** Simply being exposed to a food increases chil- dren's liking for it.**4’! Parents generally serve as the guides and gatekeepers to their children's food choices until children are able to obtain food on their own. Thus parents have a potent influence on children's food choices. However, siblings are more likely to have similar taste preferences with each other than even those between children and their parents. Preferences of mothers and fathers are also more highly correlated with each other than are those of parents with their children. Spousal correlations increase with years married.*® Eating pattern correlations between parents and their children may be higher initially but are low by the time children are of college age. 72473 Much of the unexplained variability in food preferences within cultures is probably due to social encounters outside the family, especially between peers and respected adults.**4* Mass media and food advertisements in particular have capitalized on this method by selecting actors popular with children's audiences to endorse their products. The nutrition education of children involves anticipatory guidelines for parents to help them avoid the "new faces of malnutrition in Ameri- can children today,” which consist of over- nutrition, imbalances, and excesses of food in- takes, sometimes coupled with poverty-related undernutrition such as vitamin-mineral deficiencies.*”® Health System. Many parts of the health system influence eating patterns. The role of the Fed- eral Government is exemplified by this report, along with efforts by the Surgeon General, the USDA, and many other groups. Many state and local departments of health supply nutri- tion services to groups such as children and pregnant women and to the general public (e.g., through publications and nutrition infor- mation sources). Physicians, registered dieti- tians, registered nurses, and other health pro- fessionals often provide nutritional guidance during illness as well as counseling designed to promote health maintenance. Population Panel Report D. Can Intervention Modify Dietary Behavior? Historically, dietary interventions focused ei- ther on modifications of individual knowledge, at- titudes, and behavior, or on changing governmen- tal policies, regulations, or laws relating to food. More recently, interventions seeking to change in- dividual behavior through environmental modifi- cation have been increasing." 1. The Multiple Levels at Which Change is Possible: Individual, Network, Organizational, and Community This section reviews selected interventions de- signed to change eating patterns. Such interven- tions have been implemented in many settings, such as schools, worksites, families, and fast food restaurants—and for many audiences, such as youth, adults, workers, restaurant patrons, and those with elevated blood cholesterol. The results of these programs can be categorized according to the level of the target: individuals, social networks, organizations, and communities, states, or socie- ties. A framework for reviewing blood cholesterol lowering interventions at these target levels is presented in figure 8, page 76. Primary prevention programs are often planned to encourage individual behavior change in order to reduce risk for a particular disease or health problem. For example, a blood cholesterol- lowering program to reduce CHD risk might en- courage people to eat more fruits, grains, low-fat dairy products, and vegetables, and to consume only moderate portions of lean meat, poultry (without skin), or fish in place of high-fat meats, poultry fried or with skin, or fried fish. However, some have argued that to be truly supportive of in- dividual changes, changes at the social network, organizational, and community levels must also occur.”74% This approach views the individual as part of a larger system made up of levels of social organization that are increasingly complex. Each level, or part of the system, affects and is affected by the individual's health behavior. Brody and Sobel’ argue that "disease most often involves multiple levels disrupting the person and social group, and therefore requires multiple interven- tions directed at different levels." This supports the idea that family, friends, and the community at large can play important roles in modeling and re- inforcing changes in the individual's health behavior. Targets of change at the individual level in- clude: youth, the elderly, women, men, ethnic and racial minorities, the poor, and those at high risk. Interventions at this target level include interper- sonal strategies such as classes or individual die- tary counseling including telephone-mediated ap- proaches; *" minimal contact strategies such as self-help printed materials, home videotapes, or special television programming; and technologies such as computer-based assessment and feedback.*!*? Social-network-level interventions target families and larger social networks such as circles of friends as a means to influence dietary be- havior. This level of intervention uses similar strate- gies to those of individual-level interventions (e.g., interpersonal, media, and technologies), empha- sizing that information be disseminated by key persons such as opinion leaders." Organization- level interventions seek to make changes in or- ganizations such as worksites, restaurants or res- taurant associations, food boards, and physicians’ offices. Some examples of strategies for change at this level are to increase the amount of health-pro- moting information in the organizational environ- ment; to provide programs to initiate and support dietary change (such as offering classes and stock- ing vending machines with low-SFA and -choles- terol foods); and to promote organizational poli- cies on types of foods advertised or sold.**>*% Com- munity-level interventions are intended to affect the entire community (e.g., entire populations of individuals, organizations, or networks). Strate- gies include using the mass media to persuade and educate, implementing health-promoting policies and regulations, and changing the activities and programs of health agencies (e.g., the addition of blood cholesterol screening programs and changes in the curriculum requirements of institutions of 75 Population Panel Report Figure 8 A Multilayered View of Cholesterol-Lowering Strategies: Levels of Intervention and Corresponding Targets of Change INDIVIDUAL-LEVEL TARGETS Men/Women Youth Elderly Minorities Low income NETWORK-LEVEL TARGETS Extended social network Peer groups Families ORGANIZATION-LEVEL TARGETS Worksites Restaurants Grocery stores Schools Cafeterias COMMUNITY-LEVEL TARGETS Media Social norms Public opinion Local legislation Food production Institutions of higher education 76 Population Panel Report higher education). Many examples of such com- munity programs exist in the United States and abroad 34748749 This section reviews selected research results from cholesterol-lowering and dietary change pro- grams at each of the four levels. These studies have been selected primarily because they: (1) investi- gate a diet change/cholesterol-lowering interven- tion that is congruent with the NCEP goals for cholesterol lowering,*! (2) are related to primary prevention of cardiovascular disease, (3) exem- plify one of the four levels of intervention, and/or (4) include a rigorous evaluation. Also included are relevant interventions in other health areas that illustrate important and successful models for behavior change and that have not been conducted for blood cholesterol reduction. Not included are weightloss and exercise programs, which, although they may have a substantial influence on blood cholesterol and CHD risk, are beyond the scope of this review. 2. Individual-Level Interventions Various strategies may encourage individuals to change their diets through individual-level inter- ventions. Typical are interpersonal strategies such as lectures, counseling, or groups and classes, supplemented with written materials. Intensive individual dietary counseling, group programs, or adjunctive counseling in community studies also can have favorable effects on both eating patterns and blood cholesterol level. #3 34 #24 Combina- tions of mass media and individual dietitian coun- seling have also proven to be effective. ** ** In addition, minimal contact programs have been used increasingly to inform and to teach dietary change skills. a. Individual and Group Face-to-Face Interventions Efforts aimed at increasing adherence to die- tary modifications for health purposes, especially efforts to lower blood cholesterol, require motiva- tion and encouragement from all of those involved. Successful methods for one-to-one health care encounters and in group situations have been described. Meichenbaum and Turk,*” for example, suggest several steps that ensure adherence to modified eating behaviors: "anticipate nonadher- ence; consider the prescribed self-care regimen from the patient's perspective; foster collaborative relationships based on negotiation; be patient- oriented; customize treatment; enlist family sup- port; provide a system of continuity and accessibil- ity; make use of other health care providers and personnel as well as community resources; repeat everything and don't give up.” In the past two decades, nutrition education and dietary counseling for blood cholesterol low- ering have become popular. Several examples of such counseling strategies are available.**5% In several lipid-lowering investigations such as the North Karelia study, MRFIT, and the LRC-CPPT, individual counseling has been used effectively (in North Karelia, individual counseling supple- mented group and mass media efforts). ?343344340 Snetselaar®” has reviewed methods that have proven effective in blood cholesterol lowering with individuals. These methods can be used for groups as well. Johnson and Johnson! identified 673 studies of nutrition education programs conducted since 1910, and analyzed 303 programs that had been quantitatively evaluated. The analysis was re- stricted to programs that had measurements of the participants’ attitudes, knowledge, and behavior on diet, and either a pre- or post-change score or a score comparing treatment with control groups. Some of the major programs included in the analy- sis are the Special Supplemental Food Program for Women, Infants, and Children (WIC), Expanded Food and Nutrition Education Program (EFNEP), and the National Dairy Council's Food...Your Choice.* All of the programs used individual-level change strategies, with lectures, written materials, and directed small group activities ranking at the top. * It should be noted that the dietary objectives of these studies varied. They are included here because they were dietary behavior change programs, not because of the actual changes they advocated. 77 Population Panel Report The findings of Johnson and Johnson™' sug- gest thatindividual-level dietary change programs are successful about half of the time. While very few programs produced negative results (i.e., par- ticipants did not regress in their knowledge or be- haviors), many programs had no measurable ef- fect after the intervention. Knowledge and atti- tude were significantly changed in about half of the programs. Somewhat more than half of the studies showed significant differences in behavior change. Of the total of 4,108 findings from the 303 studies, 67 percent focused on consumption of nu- tritious foods. The changes observed were reported for all age levels, for both males and females, and for all socioeconomic levels regardless of the type of individual-level instructional procedures used. The results were especially strong for positive changes in the consumption of fluid milk, of dairy products as a whole, of meat and fruits, and of protein, fat, calcium, thiamin, riboflavin, and nia- cin.”! This analysis did not review the effects of the dietary changes on blood cholesterol or other physiological indices related to diet. In another review of nutrition interventions for reducing chronic disease risk factors, Glanz>” categorized the individual-level interventions into two types: nutrition education and nutrition coun- seling. Nutrition education is defined as the proc- ess of imparting knowledge that is aimed at the general improvement of nutritional status through elimination of unsatisfactory dietary practices, promotion of adequate food habits and greater food hygiene, and more efficient use of food re- sources.’ Nutrition counseling is described as an individualized process by which a person is helped to acquire the ability to manage nutritional care, to gain nutritional knowledge, and to im- prove decisionmaking and behavior change skills”? In practice, the distinction is primarily that of the size of the audience; in counseling the audience is one individual, whereas in nutrition education the audience is most likely to be a group or population. Moreover, counseling tends to be more focused with respect to goals (e.g., reduce SFA intake to lower blood cholesterol) and usually takes place when an individual is motivated to change. Nutrition education, on the other hand, does not necessarily assume that motivation exists or that goals for change have been identified. 78 Intervention strategies have been classified as instructional, motivational, or behavioral. Instruc- tional approaches are concerned primarily with imparting information about disease processes or eating patterns. Motivational strategies focus on verbal persuasion to change attitudes and increase motivation to change; they do not usually include use of incentives. Behavioral approaches to change identify the antecedents and consequences of behavior, determine barriers to change, and ar- range behavioral contingencies to facilitate change. Reports of nutrition education programs are rarely detailed enough to determine the ways in which these three strategies differ. A few com- parative studies seem to indicate that behavioral strategies may be more effective than other ap- proaches in achieving dietary behavior change.” However, this has not been rigorously evaluated in terms of physiological endpoints, suchas changes in blood cholesterol level. The theoretical basis that governs the choice of each of these strategies is rarely reported, especially when instructional and motivational strategies are used. Studies of behavioral strategies more commonly report the theoretical base, usually social learning theory,” which postulates that interactions among individ- ual and environmental factors augment the likeli- hood of behavior. Specific details on the behav- ioral strategies studied are beyond the scope of this review. Glanz>” reviewed 10 studies on lowering die- tary fats and found that 7 produced changes in blood cholesterol via nutrition counseling or edu- cation. The context of these interventions ranged from clinical trials and community studies to worksite investigations. Approximately one-third of the study results that related to low-fat, low- SFA diets were from community studies, each of which carried out nutrition education and coun- seling as part of a comprehensive dietary change strategy. The other two-thirds involved individual strategies. No single strategy was consistently the most successful or cost-effective. Many of the stud- ies aimed at changing multiple risk factors. Thus, the possible independent contributions of smok- ing cessation or changes in physical activity or nu- trition education are unknown. Most studies failed to demonstrate differences in educational Population Panel Report strategies, but the intervention groups showed im- provements compared with controls. Glanz*” has also reviewed dietary interven- tions for a number of outcomes in addition to lowered blood cholesterol, such as treatment or control of hypertension, reducing the risk of can- cer, and weight loss. This study recommends that nutrition education programs be tailored to indi- viduals’ needs, include social support, supply skills and information in nutrition education materials, provide adequate patient-provider communica- tion, and pay attention to patient followup, moni- toring, and feedback. A recent review by Dis- brow, focusing on the cost-benefit relationship of nutrition services for the prevention of CHD, has summarized many individual and group coun- seling programs and found that such approaches are both promising and cost-effective. b. Individual Interventions Involving Minimal Contact With Providers or Facilitators Several promising individual-based strategies to assist dietary change have emerged in recent years. One is the use of materials that require minimal contact with providers. Such programs are behavior oriented, require minimal therapist time, are disseminated through a variety of chan- nels of communication, and rely to a great extent on self-administration by the individual.” Fea- tures include self-help printed kits and booklets; electronic media programs (radio, television, or home video); and new technologies, such as com- puter software, electronic mail, and other com- puter-based programs. For example, researchers working with the Stanford Five City Project (FCP) developed a skills- oriented dietary change "kit," consisting of four pages that outlined steps for changing eating behavior. In a pilot study of the effectiveness of the kit, FCP workers compared a traditional nutrition information booklet with the skills-oriented, self- help kit. Participants in the study were recruited through the mail and were randomly assigned to receive the booklet, the kit, or no materials for 6 months. Results indicated that use of the kit significantly reduced reported dietary intake of cholesterol and increased nutrition knowledge at 6 weeks and 29 weeks after completion.” A second minimal-contact program developed in the context of the Stanford FCP was a multiple- risk reduction newsletter-based project titled "Healthy Living." More than 6,000 self-selected in- dividuals signed up for and received the newslet- ter over a period of one and a half years. Most par- ticipants were female, over 35 years of age, and interested in health. The newsletter contained recent health information, tips for changing be- havior, stories of people in the comunity who were making health-related changes, and news about health promotion programs, materials, and pre- sentations available locally. The program was evaluated by surveying a random sample of par- ticipants before and after the program. Results showed no consistent behavior changes but in- creases in perceived ability to choose "healthful foods." It is not known whether this perceived ability may influence future behavior. As a third example, computer-based dietary assessment and feedback was a key element in a collaborative effort between Stanford researchers and private business aimed at reducing cardio- vascular disease risk in the workplace.” At the be- ginning of the program, participants viewed a series of health videotapes and were counseled by a specially trained nurse. The information pro- vided after the initial and monthly computer- based assessments and feedback was enhanced by attractive materials with tips for changing eating behavior. This computer-aided risk reduction ef- fort led to statistically significant changes in blood 79 Population Panel Report cholesterol (and smoking) in these workplaces when compared with a randomly assigned control workplace within the same corporation.”’ This same group of researchers carried out a second cholesterol-lowering intervention in a worksite using a brief questionnaire of the frequency of consumption of selected foods, computer-gener- ated individualized feedback, brief counseling at the time of assessment, printed materials on die- tary goals, and dietitian-initiated phone contacts on progress. This minimal-contact, individualized treatment resulted in a significant reduction in total cholesterol in the treatment group (n=114) from 238 + 25 mg/dL (6.15 + 0.65 mmol/L) to 223 + 33 mg/dL (5.77 + 0.85 mmol/L)(p<0.002). One- third of participants in the treatment group had a 20 mg/dL or more decrease.* Finally, incentive-based programs, such as lot- teries or contests, have been shown to be a cost- effective means of helping individuals quit smok- ing and lose weight.”'! Incentives have a potential role in reinforcing and maintaining behavior changes. ¢. Nutrition Education and Dietary Change Programs for Groups of Children and Adolescents Individual-level interventions have also been designed for children and adolescents. These pro- grams often are found in school settings and use educational and skill-based strategies to change the dietary habits of young people. School-based programs for children. In a review of school-based nutrition education programs, Say- lor and colleagues concluded that most inter- ventions are informational rather than skill-based, are not theoretically derived, use self-report change measures, and often do not have control groups and long-term followup. There are, however, some exceptions. Coates and colleagues®” reported that a 12-class program for fourth- and fifth-grade students, which also targeted parents, produced changes in the chil- dren's choice of school lunches and in the parents’ self-reported eating behavior. While exercise was 80 also targeted for change, the results were less substantial. The curriculum, delivered by univer- sity undergraduates, included techniques such as modeling, behavioral rehearsal, goal specification, feedback, and reinforcement for behavior change. Perhaps most important, the study included ob- servation of the students’ behavior to assess the outcome of the program. Researchers at the University of California at Los Angeles produced "Know Your Body," a chronic disease risk reduction program for ele- mentary students. Youth from 18 schools were assigned to one of four comparison groups: cur- riculum and health screening, curriculum alone, screening alone, and control. One of the nine cur- riculum modules focused on nutrition. Students in all three of the intervention groups scored more points on all six measures of knowledge of caloric value, nutrients, and basic concepts of blood cho- lesterol than did controls.”* Eating behavior was notobserved, butself-reported behavior was meas- ured. The treatment groups and controls did not differ in consumption of dairy products and foods high in SFA and cholesterol. However, "high-risk" students (49 percent of the screened population) in the screening-and-curriculum group did report such dietary changes. The first application of the "Know Your Body" program was with New York elementary school children in 22 schools from either a lower-income district in the Bronx or an upper-income suburb in Westchester. Schools were randomly assigned to intervention or control groups. Beginning in the fourth grade and continuing to the eighth or ninth grade, students were taught the "Know Your Body" curriculum. The principal findings from this longi- tudinal study were that the program was associ- ated with favorable changes in blood cholesterol level and the rate of initiation of cigarette smoking. No effect was seen on body mass, physical fitness, or blood pressure. The researchers report that the findings may be generalized to similar popula- tions of school children elsewhere." Application of the "Know Your Body" pro- gram with black elementary school students in Washington, D.C., indicated that after 1 to 4 years of "Know Your Body," changes in the desired di- rection occurred for blood pressure and blood Population Panel Report cholesterol level. The differences between the in- tervention and control groups were consistently statistically significant only for diastolic blood pressure.’ The Minnesota Healthy Heart curriculum is a 5-week, 15-session school curriculum taught by third-grade classroom teachers. The program tar- gets changes in environmental, personality, and behavioral factors that are likely to influence chil- dren's eating patterns. The study results indicated changes in the desired direction for 5 of 12 targeted foods.”™® These three examples, particularly the two re- sulting in behavior change,”*"® highlight the po- tential for changing dietary behavior of young children. These successful programs (1) integrate behavior change theory into the development and implementation of curricula, (2) use behavioral rehearsal and role playing as part of the interven- tion, (3) focus on specific foods rather than general information, (4) have a component that includes parents and children working together to make changes, and (5) concentrate on skills for behavior change combined with factual information. School-based programs for adolescents. Working with seventh graders in Finland,’ the North Karelia Youth Project compared a community- based intervention with a combined school- and community-based intervention. The goal of the ten 45-minute sessions was to reduce the five risk factors for cardiovascular disease. The researchers noted a significant decrease in fat intake (meas- ured by amount of fat in milk) among the intensive school and community intervention groups com- pared with school and community groups that re- ceived no special instruction. Adietary intervention school program for older adolescents based on an earlier program delivered to elementary students and parents was designed by Coates and colleagues.”*** The new program with black high school students focused on food behaviors particularly salient to adolescents (snack- ing), used behavior principles to develop the inter- vention, conducted an in-school media campaign, and concentrated on reducing intakes of salty, high-fat, high-SFA, and high-sugar snacks. The study compared classroom instruction with and without a parental involvement component. Ini- tial changes in the desired direction for snack foods were not maintained at followup except for the group with the parent involvement compo- nent; these students reported eating fewer target snack foods. Study investigators also reported that participation on a sports team predicted desirable snack consumption, while beer drinking and smok- ing were associated with undesirable snacks. In a study by King and colleagues,’ a 15-ses- sion curriculum, taught by specially trained in- structors, focused on snack foods, utilized social learning theory, incorporated behavioral meas- ures, and conducted a long-term followup. At the end of the program there were significant changes in knowledge, reported availability of healthful food at home, and self-reported behavior change. At 1-year followup, knowledge was still greater among students who had participated in the program. In Texas, two of four elementary schools were assigned to a diet and exercise intervention group and two to a measurement-only group. The inter- vention group included third and fourth graders whose teachers received two in-service programs twice each year. The school curriculum was modi- fied to include two 4-week healthful eating mod- ules and a 6-week physical activity module. The students were taught to recognize foods by their sodium and fat content and practiced making healthful food choices at fast food restaurants and at parties. The school lunch menu was modified to be lower in fat and sodium. The results indicated that, in the intervention groups, self-reported salt use declined in comparison to controls; however, there was no statistical difference in reported lower fat food selections. The authors note that changes in one intervention group were offset by improve- ment in control groups due to secular trend, con- tamination, or social desirability.” Another study of dietary modification of ado- lescents was conducted by Perry and colleagues™ in Minnesota. This pilot study for the Minnesota Heart Health Program was unique in its curricu- lum design and findings. Approximately half of the 10-session curriculum for the participating 81 Population Panel Report tenth graders was presented by classroom peers. Knowledge, food choice, and self-reported diet and salt intake were measured. While girls in general performed better on all measures, boys and girls made significant changes in knowledge and salt intake scores. These studies illustrate the utility of theory, the need for rigorous research design and evaluation, the necessity forbehavioraland physical outcomes, and the benefits of long-term followup. The Min- nesota study points out the need to tailor messages to audience subgroups for maximal effectiveness.” Except for the Coates study,’ minorities were not analyzed separately. 3. Network-Level Interventions Network-level interventions attempt to mod- ify the quantity of food-related communications within a network, defined as "the web of ties that surrounds an individual." Despite its potential for health education, this type of intervention is the least studied of the four levels. Such programs present difficulties such as achieving and sustain- ing adequate access to participants. Networks encompass family members, friends, members of a church or club, or any other formal or informal contacts one has on a regular basis. In a research context, networks are studied according to the quality of interactions, and their frequency, intensity, durability over time, and strength.” a. Family Networks While there are numerous studies of the effects of social networks on health status, only a handful of programs have intervened at this level. Families are the most common target for such interventions (see the successful results of youth programs that have incorporated family members, above). Inter- vention strategies used with families have included interpersonal communication such as group meet- ings and discussions and print media such as handouts and direct mail. 82 Minnesota researchers compared a home-based eating patterns program for third-grade children to a school-based approach. The "Home Team" program, a 5-week correspondence course with the children and their parents, mailed weekly packets containing behavior tip sheets and work- sheets to the children's homes. The packet de- scribed 2 to 3 hours of activities concerning eating pattern changes. Finally, a grand prize was offered as an incentive for participation, as evidenced by participation points. The results showed that the self-reported food selections, the 24-hour dietary recall data for fat, SFA, and complex carbohydrate consumption, and the food-shelf inventory for encouraged foods were all more positive for chil- dren and homes involved in the "Home Team." At a l-year post-test, most of the changes were no longer significant. In general, associations were statistically stronger when the student and not the school was the unit considered in the analysis.” Another study in California evaluated the ef- fectiveness of a family-based cardiovascular dis- ease risk reduction program in two ethnic groups.” The program reached 206 low- to middle-income Mexican-American and non-Hispanic white (An- glo-American) families with fifth or sixth graders. Half of the families were randomized to a 1-year intervention to decrease intake of foods high in fat, SFA, and sodium and to increase regular physical activity. The intervention sessions, taught by two graduate students, included 12 intensive, weekly sessions over a 3-month period and 6 maintenance sessions taught monthly or bimonthly for the subsequent 9 months. Followup continued for 24 months. Theresultsindicated anincrease in knowl- edge of the skills needed to change diet and exer- cise habits in both Mexican-Americans and Anglo- Americans, compared with the control groups. The intervention groups reported improved diets in the food questionnaire. There were no signifi- cant differences in physical activity. A significant decrease was shown in LDL-cholesterol levels for Anglo versus control participants. There were significant differences in blood pressure levels, which were 2.2 to 3.4 mm Hg lower in both Anglo- and Mexican-American intervention groups than in the control groups. Population Panel Report b. Peer Networks Researchers have suggested several models of communication flow for understanding peer net- work influences. For example, an intervention could influence the target audience directly or by reaching the target audience indirectly through opinion leaders and other people influential in the network. Using the indirect approach to promote diffu- sion of health information, the North Karelia proj- ect trained lay health opinion leaders to promote the dietary and other messages central to the pre- vention of cardiovascular disease.” These inter- personal messages reinforced similar messages broadcast in the media, presented at classes, and provided through events in the community. Thus, several levels of intervention provided stimuli for change to the networks. A second strategy is to actually develop social networks. The Stanford Three-Community Study (TCS) showed that opinion leadership emerged from high-risk participants who were exposed to both media-based and interpersonal interven- tion.” As the magnitude of input increased (from mass media only, to mass media plus screening plus change agent contact), so did the number of contacts and the number of health conversations. This study illustrated how the media alone and the media supplemented with interpersonal contact affected the diffusion of health information. Simi- larly, in a recent study, groups of women at risk for cancer used self-help networks to achieve reduced dietary fat intake over a period of 2 years.” Promoting changes in communication behav- ior within social networks may be another impor- tant avenue for dietary change. While the effect of the content of network relationship on health out- comes has not been directly studied, spouse and peer support have been shown to be important elements in health behavior change and mainte- nance.>2*%0531 Both the number of supportive inter- actions and the kinds of supportive interactions may be important. Changes in network interac- tions about diet and health can potentially influ- ence the short- and long-term outcomes of pro- grams in organizations and in whole communities. 4. Organization-Level Interventions As intervention targets, institutions and or- ganizations such as worksites, schools, primary health care settings, supermarkets, and other retail food outlets offer many potential benefits. The two most studied settings for organization-level inter- ventions are the worksites and point-of-purchase settings such as cafeterias, restaurants, and gro- cery stores. There arealso several studies of schools as settings for organization-level change. Using organizations as channels for delivering health programs, health educators can reach target audi- ences; tailor interventions to specific contexts; multiply efforts via social support, instructional expertise, and facilities; influence organizational structure in ways that will support individual- level behavior change; and persuade organiza- tions to become healthful models for other com- munity institutions. McGandy et al. changed the foods available in a school food service and found lowered blood cholesterol levels in students. Ellison et al.”* have also attempted cardiovascular disease modifica- tion by a change in the school environment, espe- cially by reducing the fat and sodium content of food made available through the food service in two boarding high schools. By implementing the program in each of the schools in alternate school years, researchers were able to compare students ataschoolin which there were food service changes and students in a school without such changes. These students were not counseled to change their eating practices since the intent was to study envi- ronmental changes, but they were asked to com- plete periodic 24-hour food diaries. The study demonstrated that, at the end of the academic year, the change in food service was associated with an estimated 15-20 percent decrease in sodium in- take. In addition, preliminary results indicate that total fat decreased about 9 percent from 35.4 per- cent of calories at baseline to 32.1 percent and that SFA decreased 21.8 percent from 12.4 to 9.7 per- cent. The latter study showed the effects of an en- vironmental program alone and cannot predict the influence of an additional education component.” 83 Population Panel Report Typically, interventions delivered through or- ganizations targetand measure individual change. Recently, however, some research has targeted and measured both organization- and individual- level change.” In one weight-loss program, for example, organizational goal setting, public dis- plays of group achievements, and use of the or- ganization's newsletter, as well as intergroup com- petition were used to enhance generalization and maintenance of lifestyle change. The individuals demonstrated weight loss at a 6-month followup, and the workplace found the program feasible to maintain.” Organizations not only can serve as channels for reaching individuals, but can also be targets of change. In terms of CHD risk and food, these organization-level interventions attempt to change conditions or policies within an organization in order to promote diets likely to lower blood cho- lesterol levels. They are characterized by strategies to increase the availability of appropriate food choices, improve access to these foods, increase the appeal of low-fat, low-SFA, and low-cholesterol foods, increase the information about these foods, and create incentives to eat healthful foods." a. Worksites as Targets of Change In a review of nutrition programs at the worksite, Glanz and Seewald-Klein®* found that most worksite health promotion programs offer- ing nutrition education have been developed since 1980. The most commonly used educational method is printed brochures or newsletters. These authors report that the Johnson and Johnson Company's Live for Life program, based in part on an eight-session nutrition education course, is re- inforced with policy changes that provide scales in restrooms and food choices that are low in calories and fat in vending machines and the cafeteria. An evaluation of this program concluded that there was a significant decrease in the number of over- weight employees at program sites. In addition, Southern New England Bell Telephone Company's 6-week nutrition course, plus short courses and point-of-purchase information at company cafete- rias, resulted in measurable reductions in weight, blood triglycerides, and cholesterol.” 84 Evaluation of two other worksite nutrition programs provide some insight into components that are linked to successful behavior change. In a worksite screening and education program for employees of the Social Security Administration, the treatment group received individual counsel- ing, and spouses were invited to an informational meeting. Although the program resulted in a re- duction of blood cholesterol levels by 12-17 per- cent compared with 4 percent for the control group, these results disappeared after the completion of the program. The Heart Disease Prevention Project of the World Health Organization, conducted in places of work, found that cholesterol reduction was greatest when participants received interper- sonal communication about diet rather than print communication.” The New York Telephone Company's short nutrition course was designed for high-risk indi- viduals with total cholesterol levels over 240 mg/ dL (6.21 mmol/L). The treatment consisted of an eight-session program with monthly behavior maintenance meetings for the following 6 months. Cholesterol declined a little more in the treatment group (8.8 percent) than in the control group (6.4 percent). These studies suggest that worksite-based classes and individual counseling can be effective, and that individual dietary change can be sup- ported and enhanced by changes in the environ- ment such as providing low-cholesterol entrees in the cafeteria. Recommendations for enhancing the effectiveness of worksite programs have been pub- lished,” and the American Heart Association has developed a detailed worksite program entitled Heart at Work.>" b. Point-of-Purchase Interventions Point-of-purchase interventions are those con- ducted at sites where people purchase food either to prepare or to eat directly.” These programs aim to influence the food purchases people make; however, learning, deciding, choosing, and con- suming may also be targeted outcomes at point-of- purchase sites.” Restaurants, grocery stores, cafe- terias, "mini" markets, and even vending machines Population Panel Report are all examples of potential sites for point-of- purchase interventions.” Generally, desired insti- tution-level outcomes are increased availability, accessibility, and purchase of healthful foods, along with the improvement of the institutional environ- ment through increased informationand reinforce- ment of healthful actions. One would typically work toward these general goals by increasing the number of advertisements for healthful food within the store, labeling the healthful foods, organizing food tastings for targeted foods, and distributing informational brochures.>*!** Little has been done to test the effectiveness of changing the in-store en- vironment (e.g., optimal positioning of healthful food on shelves or packaging of foods to encour- age consumers to eat smaller amounts, such as providing 4-ounce packages of meat to yield 3-ounce cooked portions), but some observations have been made.” Researchers at the Minnesota Heart Health Program have been successful in col- laborating with local beef and pork producers to provide consumers with information about lean cuts of red meat.** Training meat department per- sonnel and taste-testing recipes are all part of the educational effort. While not yetexamining all the potential means of influencing people in food outlets, studies of point-of-purchase interventions have already yielded some important results.” Although the impact of these interventions on individual eating patterns and health outcomes is unclear, they have demonstrated that organization-level change is possible. Grocery Store Programs. Programs in grocery stores use nutritional information displayed at various points in the store as a means of increasing the knowledge of consumers and promoting a change in food purchasing behavior.” In a na- tional survey of 83 grocery store chains, Penning- ton and colleagues™® found that only 30 (36 per- cent) had some type of in-store food health label- ing program. Moreover, there was a lack of consis- tency across programs in the use of terminology, labeling format, and color coding of calorie and nutrient labels (such as "low calorie"), indicating a potential for consumer confusion. A number of studies of grocery store interven- tion programs have placed particular emphasis on purchase behavior as an evaluation criterion. The studies have measured changes in the actual sales of entire food categories or specific brands tar- geted for health-related nutrient characteristics. A cooperative program between NHLBI and Giant Food Inc. in the Washington, D.C., area—"Food for Health"—used special brochures and sectional shelf labeling over a 1-year period. Evaluation results indicated a significantincrease in knowledge at the intervention sites compared with a control group of stores in Baltimore. The campaign also demon- strated the feasibility of point-of-purchase infor- mation programs and generated a high demand for materials. Sales data, however, did not indicate a shift in food purchases for intervention versus control stores.>*¥ In another early attempt to influence food be- havior through grocery store interventions, Jeffery and colleagues™' carried out an experimental study in eight Minnesota stores over 6 months. They monitored weekly sales of 25 items and admini- stered pre- and post-test questionnaires to shop- pers on dietary knowledge. Although shoppers’ knowledge increased in both experimental and control stores, there were no differences in food purchasing behavior. Building on the earlier NHLBI in-store pro- gram, Giant Food Inc. and the FDA jointly con- ducted a 2-year point-of-purchase labeling study, "Special Diet Alert,” in the same intervention and control markets. This program included a longer test period and the highlighting of specific food brands by both special shelf labels and an alpha- betized booklet containing the names of all partici- pating brands. The average market share of flagged 85 Population Panel Report products increased by a statistically significant 8 percent in intervention versus control stores for 8 of the 16 product categories.” The results of this study were interpreted in the context of a review of seven in-store purchase behavior studies reported in the literature (two with positive and five with negative results). This review led to the conclusion that positive outcome was related to some combi- nation of a longer program duration, simpler nu- trition messages, brand-specific labeling, and ex- perimental or statistical control of environmental variables in the supermarket. The FDA-Giant study was enlarged to include more brands and food categories and was placed for an additional 2 years in stores in Baltimore, the control market for the original study. Original test market stores in Washington also carried the ex- panded program. A nominal price was charged for the in-store booklet. Statistically significant mar- ket share gains occurred for labeled foods in Balti- more stores in half of the product categories that had been part of the initial program. Furthermore, in 25 newly added product categories, flagged products had significant market share gains in 10 categories as opposed to share declines in only 4.7 In-store booklet sales were not strongly related to usual shopper demographics, but were concen- trated among shoppers with a household member on a special diet. The booklet purchase rate aver- aged 9 percent of households patronizing the test stores; this equaled the average U.S. household re- demption rate for direct mail food coupons during the same period, according to national clearing- house statistics. The use of food coupons is consid- ered an effective means of persuading consumers to try products, and the nutrition booklet pur- chases compared quite favorably to coupon re- demptions. Across stores, average booklet sales were moderately and positively correlated with the magnitude of gains in the market share of labeled products.” The FDA recently completed a third in-store labeling program of similar design to the Giant Food study in cooperation with the Stop and Shop food chain in the New England area. Preliminary results in 60 participating food categories essen- tially replicated the findings of the earlier studies, 86 with statistically significant gains in market share registered for labeled products in half of the cate- gories. The shifts in purchase behavior occurred in markets that were smaller, as well as socio- demographically and ethnically more diverse, than the Washington and Baltimore markets.” The Minnesota Heart Health Program initi- ated a community "Eating Pattern” task force that developed the "Shop Smart for Your Heart" pro- gram. This consisted of an integrated system of shelf labeling, posters, shopping bag stuffers, a preprinted grocery list, and a brochure all empha- sizing low-cholesterol, low-fat, low-SFA foods. Telephone surveys of randomly selected partici- pants made 2 to 4 months after the start of the program showed that in the two intervention cit- ies, 25 and 47 percent of the respondents, respec- tively, reported that labels influenced their food purchasing decisions.” The Pawtucket Heart Health Program intro- duced, in 1983, the "Four Heart" Grocery Program, which used shelf labels, product-specific cards, posters, and brochures to inform consumers about foods low in fat, cholesterol, or sodium. The pro- gram was evaluated using consumer exit inter- views and food sales. The result showed that in- store programs increased consumer awareness and influenced their food choices.” Knowledge of the outcomes that can be ex- pected from point-of-purchase programs is in- complete; there is ample evidence that consumers are aware of the programs, are satisfied with their content, and report that the programs influence their food purchases. The growing body of sales evidence confirms that shifts in purchase behavior do in fact occur in many food categories, but there is no evidence to date on whether the new pur- chase habits are sustained over time. Restaurants and Cafeterias. Point-of-purchase programs also have been implemented in busi- nesses that sell prepared foods. These include res- taurants, cafeterias, and convenience food outlets. These programs also rely on print media to edu- cate and persuade consumers to buy particular foods. One study used a single poster advertising the health benefits of the low-fat entree, which Population Panel Report resulted in a 15 percent increase in purchases of that entree.” Wagner and Winett™ replicated these findings by successfully using signs and posters to cue healthful food choices in a fast food restaurant. They found that the cues for salads increased the sales of salads; sales of higher fat foods such as hamburgers and french fries did not decrease. By contrast, a study of the impact of nutrition information placed in vending machines concluded that availability, rather than attractive nutrition information, was the best predictor of whether a consumer would purchase a food item that was lower in calories, sodium, and fat than other choices.” In a study of a restaurant menu labeling pro- gram in four family-style restaurants in four com- munities (two in the Stanford Five City education communities), Albright and colleagues™' found that sales of targeted entrees increased signifi- cantly in two communities and stayed the same in two other communities. Neither food sales data nor exit surveys of patrons explained the differ- ences in impact. Cafeterias serve millions of students and em- ployees in the private and public sector. A game developed to promote healthful food choices in a large institutional cafeteria was associated with reduced purchases of desserts and total calories per customer and increased sales of skim milk compared with controls.** Another cafeteria-based program, using labels to emphasize the calorie content of foods, found reduced purchases of high- calorie side dishes but not entrees so that the over- all calorie content of meals purchased was mini- mally affected.” Ata university dormitory cafete- ria, a food labeling program was developed that consisted of food labels with information about the amount of fat, cholesterol, and calories for each item. Compared with the baseline data, selections of meals lower in fats, cholesterol, and calories increased from 31 to 47 percent depending on the meal and intervention phase; selections of such meals remained higher than baseline levels 5 weeks after the intervention stopped.*” In another cafete- ria program, Cincirpini,** over the period of an academic year, assessed the effects on food pur- chases of various methods, i.e., calorie listing, labeling of target items, and token rebates (incen- tives) for labeled items. The tokens produced the most uniform changes toward low-fat and low- calorie food choices. Point-of-purchase programs in grocery stores, restaurants, and cafeterias are feasible and in most cases welcomed by store managers.>**** Consumer satisfaction and awareness are typically increased after intervention.”*>* Consumers report that la- bels and other informational cues influence their choices, yet sales data do not consistently confirm these reports. This may be due, in part, to other factors that also influence shifts in sales, such as price and seasonal variation. In addition, many studies lack controls, thereby rendering interpre- tation difficult. There has been little comparison of different types of approaches to promote healthful foods. The usual messages are informational and empha- size health values. In fact, other factors such as price, agricultural season, culture, and geography may influence food purchasing behavior. Colby and colleagues™ compared different influences for a low-fat entree in a family restaurant: health, price, taste, and no special influence. Sales data showed that consumers were more likely to pur- chase the low-fat entree promoted as a low-priced daily special. Overall, labeling programs in grocery stores and restaurants have yielded mixed results. In the literature reviewed, consumer knowledge seems to have increased in all of the studies that meas- ured knowledge change. Evidence of behavior change is less consistent, although each of the more recentsupermarket studies has demonstrated shifts toward more healthful food purchases in about half of the food categories studied. Success- ful grocery store programs have combined the following elements: a focus on specific foods; an emphasis on a few diet/health issues; and a pro- gram duration of more than 1 year. In addition, they have used a rigorous experimental design for program evaluation. Considering the food distri- bution channels as a whole, however, results con- cerning the amount of labeled products are incon- sistent. Some studies document changes while others do not. Increases in purchases of identified 87 Population Panel Report foods have been reported in some instances with- out a corresponding decrease in the amount of purchase of foods that are high in fat, SFA, choles- terol, and calories. c. Schools as Targets for Change Many workers have considered schools as set- tings for influencing individual behavior; schools themselves can also be targets of change.” Rede- fining the cholesterol, fat, and sodium levels in school lunches is a policy change with far-reaching implications.” A study of school lunches in Texas revealed that local school lunches had 30 percent more fat and 100 percent more sodium than rec- ommended by the study standard.” Thus in the research study, school cafeteria lunches were changed to low-fat and low-sodium meals. The success of this change reflected the change in pol- icy, rather than a primary change in consumer behavior. The USDA has recently published reci- pes recommended for school lunch programs that are lower in total fat, SFA, cholesterol, and sodium than previous recipes. In some parts of the country such as New York State and Massachusetts, pro- grams that include low-fat choices have been developed. Several studies of changes in school nutritional environments, food service, and health education to favor "heart-healthy" eating patterns have been reported.” In the school environment, the behavior of teachers can also be influenced. In a worksite inter- vention®” in a Dallas school district, pre- and post- screening and a 10-week behavior-oriented class were offered to teachers. Significant reductions occurred in 11 factors related to cardiovascular disease. Job satisfaction, freedom from worry, energy level, life satisfaction, cheerfulness, relaxa- tion, emotional control, and general well-being all improved. d. Health Care Facilities Several national reports’! recommend that pa- tient contacts with health professionals include 88 dietary education and counseling. National sur- veys, however, indicate that fewer than 30 percent of men and women report that food habits were discussed when they visited a doctor or health pro- fessional .*>*"°70 In a study of medical residents in a university-based training program concerning compliance with eight recommended preventive screening practices, the physicians performed only three of the practices (Pap test, breast examination, and influenza vaccination) as much as 20 percent of the time. *”' A recent report by the Food and Nu- trition Board of the National Academy of Sciences indicated that many nutrition education programs in U.S. medical schools are inadequate.” On the other hand, there is some evidence that the treat- ment practices of family practice residents can be improved with education, feedback, and monitor- ing.” Physicians’ Offices. Physicians are often viewed as expert sources on most health topics and as ac- curate sources of health information.” Studies of physicians and their role in cholesterol reduc- tion fall generally into three categories: (1) studies of physicians’ attitudes and practices regarding cholesterol management, (2) studies of interven- tions in physicians’ offices, and (3) investigations of methods to increase the likelihood and effec- tiveness of physician management of cholesterol. Ellison” has reviewed practice patterns, pointing out problems that require attention. Studies of the attitudes and activities of family practice physicians show that most physicians report giving some nutrition advice to patients visiting them.” However, almost half of these physicians give advice about dietary fat, sodium, and fiber to fewer than 20 percent of their patients. Only 10 percent of physicians give advice to more than 80 percent of their patients. Half of physicians reported not giving dietary advice because the pa- tient did not have an elevated blood cholesterol level. Other reasons included expectations of pa- tient noncompliance and perceived lack of need for dietary fat reduction because there was no weight problem. About one-third reported not being convinced of the value of saturated fat re- duction and an equal proportion indicated that a lack of patient interest prevented them from Population Panel Report making recommendations. Physician perspectives concerning cholesterol have undergone signifi- cant change.” Public expectations may influence physicians behavior. For example, physicians in the Pawtucket Heart Health Program identified requests from the publicas animportant factor significantly more than physicians in a comparison city.” These and other studies indicate that physicians’ treatment practices are the outcome of multiple influences. At the First National Conference on Choles- terol in November 1988, many of the presentations focused on physicians and physician-based inter- ventions.”*50%2 In a survey of Vermont physi- cians’ attitudes toward screening and dietary advice, Secker-Walker et al.”® reported that physi- cians decide which patients will receive dietary advice according to their perception of patient interest. Many believe that nutrition education is the responsibility of the physician and does not require referral to a registered dietitian or nutri- tionist. These Vermont physicians reported that 35 percent of patients are regularly screened and only 23 percent receive dietary advice. These results are similar to those reported 4 years earlier by Kottke and colleagues.” Neither the effectiveness nor the cost of physician nutritional advice has been thor- oughly evaluated. A study of physician-based intervention was conducted in two multiple-specialty clinics (one intervention clinic and one control) in Minnesota to determine the effect of systematic screening and educational intervention for hypercholesterol- emia. Patients, whose cholesterol level was in the high- or moderate-risk group according to NIH Consensus Development Conference” guidelines, were invited by the intervention clinic to attend a two-session nutrition class. Of those invited, 27.5 percent attended at least one class. In the control clinic, participants were invited to come only for a repeat dietary and cholesterol assessment; a similar recruitment rate was achieved. After 6 months, there was a significant fall in cholesterol (from a mean of 273.4 mg/dL to 248.9 mg/dL; 7.07 mmol/L to 6.44 mmol/L) and a significant change in diet in the intervention group.” In a second Minnesota study of cholesterol management in five physicians’ office practices, researchers reported that physicians thought that the program was satisfactory, workable, and helpful to patients, and that it improved the clinic's image and re- duced cholesterol. All reported that they would continue the program.®> 5% In a study of physician office intervention in California, patients within a family practice clinic were recruited, had their blood cholesterol as- sessed, and were randomly assigned to a treat- ment group or delayed treatment control group. All patients received standardized instructions from their physicians. Patients in the special treat- ment group subsequently received computer- based feedback on a dietary assessment, and ap- proximately 70 percent of the patients returned for a 6-month assessment. There were marginally sig- nificant changes in blood cholesterol between in- tervention and control patients, but there were sig- nificant improvements on 6 of the 11 targeted food items.” Some insight into the components of a success- ful office-based cholesterol management program was gained from a study of dietary reduction of hypercholesterolemia by Crouch and colleagues.™ They reported that multiple face-to-face visits and telephone and mail contacts over 4 months pro- duced comparable and significant reductions in cholesterol, whereas one-time counseling or no contact produced no change or changes in the wrong direction. Other recent studies have investigated strate- gies to increase physicians’ treatment of high blood cholesterol. Rhode Island researchers, forexample, studied three family practice residency programs. In one, residents were trained in blood cholesterol measurement using a fingerstick and desktop 89 Population Panel Report analyzer, a dietary assessment technique, the NCEP guidelines, and followup protocols in dietary and drug treatment. In addition, these residents were given feedback on their performance from a re- view of charts.” The other two programs served as controls. Early results from the intervention group indicate a 100 percent increase in blood cho- lesterol measurement over 1 year and a threefold increase in referrals for nutrition counseling. In another innovative investigation of methods to improve physicians’ cholesterol management practices, hospital house staff were randomized into one of four treatment groups.” Treatment groups received either a cholesterol reminder checklist in patient charts, performance feedback of management for each high-risk patient, or both feedback and checklist. Those in the feedback group showed a significant increase in noting elevated cholesterol levels, giving orders for diets to lower blood cholesterol, and ordering lipid profile tests. The combined (checklist and feedback) group showed smaller improvements. In another hospi- tal-based education program, awareness of cho- lesterol was assessed by notation in charts; the awareness level (20 percent) did not change over 2 years in spite of changes in laboratory criteria for cholesterol. After a 1-hour lecture on cholesterol management, awareness shifted to 69 percent 2 weeks after the program and to 52 percent 6 weeks later.” In San Francisco, researchers achieved a 35 percent recruitment rate of community physicians to participate in a study of office-based cholesterol management practices. Reasons for participation included personal interest, patient demand, availability of continuing medical education credit, and willingness to contribute to research. Those who refused (65 percent) gave lack of time, inter- est, and financial incentive as reasons for nonparticipation.’” These studies suggest that physician interven- tions can be effective in helping people change their eating patterns and, at times, lower their blood cholesterol level. Concern about the cost- effectiveness of such efforts has been expressed by Weinstein and Stason™' in a review of five studies, and subsequently by Disbrow.”" 90 Other Health Professionals and Other Health Care Settings. In recent years, group practices, health maintenance organizations, and other forms of practice have brought large groups of patients into prevention-oriented health care settings where they are seen by teams of providers. These changes, along with changes in reimbursement methods, have led to consideration of more cost-effective methods for delivering health care. The role of other health professionals such as nurses, registered dietitians, pharmacists, and so- cial workers has been considered in providing lower-cost nutrition and health education. Three recent papers examined cost-effectiveness of blood cholesterol lowering using various mixes of health professionals.”**?>** The analyses of various fac- tors, such as salary, fringe benefits, overhead costs, and numbers of clients, clearly show that non- physician health professionals can provide indi- vidual and group nutrition counseling at rela- tively low cost. Many of the studies reported here were pilot projects and lack long-term followup. However, several tentative inferences can be drawn. First, blood cholesterol management in physicians’ of- fices and other health care settings is accepted and approved by patients. Second, physicians and other health professionals will participate in and main- tain cholesterol management programs, even though they may need additional stimuli to initi- ate such programs on their own. Third, the num- bers of physicians and other health professionals practicing vigorous cholesterol management can be increased. Health professionals are likely to respond positively to patient interest and demand for advice about high blood cholesterol manage- ment and diet. Finally, the entire health commu- nity should join in the effort to find cost-effective ways to help people adopt new eating patterns and decrease blood cholesterol levels. Health Facility Food Services. Many hospitals and other health facilities provide food to large numbers of people. Despite the opportunity to model food choice behavior to achieve low blood cholesterol levels, little research has been reported in this area other than the National Institutes of Health study previously described. Population Panel Report 5. Community-Level Interventions This section reviews three kinds of interven- tions: media-, policy-, and community-based. a. Media-Based Interventions The media influence community practices re- garding politics, economics, and public health. Media attention to disease prevention and health promotion can influence the status of health as a topic of community importance.” The media are particularly powerful at influencing individual behavior directly through mass media campaigns and indirectly by influencing public opinion, pri- oritizing issues, and helping to create and rein- force social standards. Mass Media Campaigns. Mass media campaigns are one important means of influencing the infor- mation environment, and they in turn supplement the effects of other social, organizational, network, or individual change efforts.””*>* One among many examples of media campaigns designed to change a particular dietary habit is the National Cancer Institute's (NCI) Cancer Prevention Awareness Program. This nationwide program to prevent cancer seeks to change knowledge, attitudes, and behavior through information and education for the general public and health professionals. A 2-year cereal advertising campaign, which in- cluded NCI's diet message and telephone number in television ads and on the backs of boxes of abran cereal, produced more than 20,000 telephone and 30,000 written inquiries to the NCI. A study of cereal sales in the greater Washington, D.C. area revealed that the sales of all high-fiber cereals in- creased, not just the brand advertised .”*>* b. Policy-Based Interventions The laws, policies, and allocation of public re- sources at all levels of government can have im- portant impacts on the health of the public. Alco- hol and cigarette taxes and seat belt laws are but a few examples of the potent effect of government regulation. Local or national policy changes can also influence eating patterns.” The effects of wartime food rationing serve as one example. Another is the recent change in the term for grad- ing lean beef from "good" to "select," a policy change that may influence consumers to choose leaner meat." ® In terms of food choices, the food industry, wholesale and retail food vendors and food service establishments, and public health agencies are influences that are increasing the availability of foods that are low in calories, fat, SFA, cholesterol, and sodium. These groups can further support activities such as education of food servers, grading, and labeling of foods.” The efforts of the National Livestock and Meat Board to stress availability and consumption of smaller, leaner meat portions represent a helpful organiza- tional policy change. Regulation affects the actions of organizations. Using regulations as an incentive for health pro- motion, a community in Pennsylvania allowed food service institutions to receive reduced fees if they adopted healthful policies in their food out- lets. These policies included designated nonsmok- ing areas, certification of a percentage of employ- ees in cardiopulmonary resuscitation, and attain- ing certification in food sanitation principles and practices." While not all of these policies were food related, this study exemplifies the ways that communities can stimulate and support organiza- tional adoption of healthful policies. c¢. Community-Based Interventions A growing number of community-based stud- ies have been reported or are underway. The Stan- ford Three Community Study (TCS) and the Finn- ish North Karelia Project*3#34527.602605 achieved similar cholesterol, blood pressure, and smoking results after 3 years of education in the TCS and after the first 5 years of education in the North Karelia Project. In the Finnish study, less cardio- vascular disease morbidity and mortality were re- ported after 10 years in comparison not only with the adjoining county but with the remaining part of Finland. Preliminary results from the Stanford Five City Study suggest favorable trends.” Re- sults from two other large Heart Health Studies in 01 Population Panel Report Minnesota and Rhode Island, discussed previ- ously, will be available in the early 1990's. Four other community-based studies using methods similar to the TCS and the North Karelia Project have reported their main findings. Only one of them, in three small rural South African towns, reported significant decreases in blood cholesterol and blood pressures as well as in smoking.” Two others, a four-town study in Switzerland and a three-town study in Australia, reported signifi- cant drops in smoking rates.®”**® The favorable effects of these studies present an encouraging picture for community-based intervention. 6. Conclusions The foregoing section has reviewed investiga- tions of interventions targeted at individuals, net- works, organizations, and communities / societies. Studied strategies for change have included face- to-face persuasion, skills training, information dissemination, minimal contact programs, use of incentives, policy and regulatory change, and use of influential individuals. The channels for com- municating dietary information have included use of the media, new technologies, printed materials, health professionals, and other credible sources. , Some intervention strategies have been more rigorously evaluated than others, some have more successfully integrated behavior theory than oth- ers, and some have used more formative research in the development and implementation of inter- ventions than others. Not all of the potential points of influence have yetbeen investigated in the effort to improve the diets of Americans. Nonetheless, many feasible and well-evaluated interventions do provide scientific support for recommenda- tions designed to influence national eating patterns. 7. Summary The data reviewed here and elsewhere>'! im- plicate elevated blood cholesterol in the patho- genesis of atherosclerosis, especially CHD. Fur- thermore, while there is evidence that blood 92 cholesterol levels are affected by many factors, food habits are among the most important. The strongest dietary relationship with blood choles- terol level is between saturated fatty acid intake (especially those SFA containing 12 to 16 carbon atoms) and elevation of both total and LDL-choles- terol levels. The average current U.S. consumption of saturated fatty acids (14 percent of total calories) is high and warrants a significant reduction. A rec- ommendation that less than 10 percent of calories be derived from SFA is based on epidemiologic evidence that healthy populations with SFA levels of 10 percent or less have low CHD rates and on a large body of evidence from many kinds of studies that reducing SFA will lower the level of blood cholesterol and thus lower coronary risk. PUFA and MUFA also influence circulating lipids. Their major effect derives from their substi- tution in the diet for SFA. The data on which to base quantitative dietary recommendations con- cerning PUFA or MUFA are less clear than are those for SFA. Given current knowledge, the best approach appears to be to emphasize a reduction of SFA, as above, and of total fat. Several beneficial effects can be expected. Reducing total fat facilitates at- tainment of the SFA goal. Limiting all fat intake may also help to prevent obesity, an important factor in control of high blood cholesterol, and some forms of cancer.” Finally, presenting quanti- tative recommendations for SFA and total fat is likely to enhance public comprehension and compliance. High levels of dietary cholesterol have, in most people, some impact on blood cholesterol level. The evidence for limiting intake is strong. A reduc- tion to an average cholesterol intake below 300 mg per day will produce a significant reduction in the average blood cholesterol and, correspondingly, an important anticipated decrease in CHD rates. While achieving even lower cholesterol intake would likely lower CHD even more, a target of 300 mg per day appears to be realistic in light of the current eating pattern of many people. Population Panel Report VI. Ethnic, Cultural, and Minority Characteristics That Influence Diet and Health A. Introduction Itis a truism that people eat food, not nutrients. The food we consume determines our energy and nutrient intakes. The adequacy of our diets is a consequence of the food we select and its nutrient content. Culture, society’s design for living, is a major determinant of food habits in any society. Cultural values help determine what edible items are se- lected from the environment, and how they are transformed into what the society considers food. Customs govern the preparation, flavoring, pres- entation, combination, order, and eating rituals relating to food. Violation of these customs by well-meaning health professionals or others seek- ing to reform food habits may threaten community identity and solidarity, and may result in rejection of the advice. Therefore, to be effective in the long run, public health efforts need to encourage changes in those customs and values that promote excess consump- tion of fat, saturated fat, and cholesterol.®”” And in the short term, specific guidance, to be acceptable, must respect cultural customs and work within them. Public health approaches must also con- sider risk of malnutrition versus risk of CHD in groups where poverty and social isolation are factors. The following section deals with the many fac- tors that influence diet and health in all popula- tions, especially in ethnically diverse or minority groups. Caution is indicated when interpreting health data because ethnic and cultural factors are highly confounded with economicand educational status and with access to both preventive and curative health care. Y 1. Dynamics of Culture Change In order to understand the characteristics of ethnic, cultural, and minority groups that may in- fluence intervention efforts, one needs to examine the dynamics of culture change. American eating habits and traditions reflect the many customs brought to the country by immigrant groups. During the immediate post World War II pe- riod, as a result of a new national consciousness, technological developments in mass communica- tion, and a rapidly expanding food industry, a “na- tional diet” emerged. It synthesized many ele- ments into the lowest common denominator, or what has been otherwise described as the “blan- derizing of America.” Reaction to this “leveling” process began in the late 1960's. The food industry reaction has been one of market segmentation, and the growth of conveniently packaged ethnic spe- cialties and of ethnic restaurants. These develop- ments have created an environment of greater flexibility and have weakened some of the cultural rules governing food selection and preparation. This flexibility and greater respect for ethnic foods by the majority culture provide an opportunity to enhance intervention for dietary change. 2. Health Belief Systems An important component of culture is how the society organizes and processes information. Understanding this process is crucial to communi- cation. The belief system of the majority American culture is strongly rooted in the scientific method. This is reflected in the Western system for 93 Population Panel Report treatment and prevention of illness. It is also evi- dent in delineation of food groups based on their nutrient content. A common and persistent alter- native health belief system, among such diverse populations as some in Latin America and Asia, is the hot-cold theory or yin-yang. According to this theory, health is maintained by the proper balance of hot and cold factors. Foods are classified as hot, cold, or neutral. A dietary intervention strategy developed within this framework might be far more effective than nutrient-based food groups in sub-populations that adhere to the hot-cold the- ory. Itis not necessary to adopt Western thinking in all respects to preserve or promote good health. In fact, many cultural rules for food use in tradi- tional societies are directed at perceived health benefits. Other cultural characteristics may also affect intervention strategy. Weaver® has proposed a model for comparing two basic culture-and-per- sonality systems. It represents two ways of organ- izing society, its values, harmony, and percep- tions. In Weaver's model, cultures are placed along a continuum ranging from abstractive to associative worlds. Mainstream American culture falls on the side of the abstractive world. Ethnic minorities may fall anywhere along the contin- uum depending on the degree of acculturation. Since many of the factors affect communication, it is important to recognize differences that may exist among ethnic minorities. For example, ac- cording to Weaver, the abstractive world is loosely integrated, urban, and heterogeneous; social struc- ture is based on achieved status, and the nuclear family is important; there is a high degree of social and physical mobility. Basic values include change/action and individualism. Interaction is written, electronic, and impersonal, and time is future or outcome oriented. By contrast, the asso- ciative world is based on a highly integrated homogeneous community whose roots are rural. The social structure is based on ascribed status, kinship, and the extended family. There is little social mobility. Basic values include stability, harmony, and the sense of belonging. Interaction is face to face, personal, verbal, and nonverbal. Time orientation is past-present. The model pro- vides a theoretical framework for understanding 94 cultures, which is crucial to designing effective strategies for ethnic minority groups in the United States. Intervention strategies in groups with more as- sociative characteristics should use kinship and community resources as much as possible. Plan- ners must take into account the implications of past-orientation for an intervention whose out- come is future oriented. At an extreme, looking into the future may be seen as prophecy and thus to violate religious principles. However, among the tenets of most major religious and ethical sys- tems is a moral responsibility to maintain a healthy body. 3. Acculturation Acculturation is the process whereby the char- acteristics (design for living) of the dominant cul- ture are adopted and assimilated. The process is complex, varies among individuals, and depends onmany factors, including the extent and duration of exposure to the dominant culture. Low levels of educational attainment, lack of proficiency with the language of the dominant culture, social isola- tion, subtle discrimination, low income, and high unemployment retard acculturation. 4. Poverty Many ethnic minorities have a substantial pro- portion of educationally and economically disad- vantaged persons (table 6). For some groups, this may be a temporary situation, whereas for others it may become a long-term reality. Much has been written about the so-called culture of poverty. Adherents to this theory contend that the culture of poverty has certain characteristic traits includ- ing strong feelings of fatalism, helplessness, de- pendence, and inferiority. The poor, it is claimed, have a strong present-time orientation and thus do not defer gratification or plan for the future. They often believe in male superiority and are sensitive to external status symbols. Once the culture of poverty comes into existence, some suggest that it often persists in succeeding generations.'' In any Population Panel Report Table 6 Sociodemographic Characteristics of Population Subgroups in the United States Characteristics Ethnic/Racial Background White Black Hispanic Asian/Pacific Native Islander American Total Percent 79.6 11.5 6.4 1.6 0.9 Distribution (1980) Median Age (1980) 31.6 24.9 23.2 28.7 22.4 Percent High School 88 79 58 75 31 Graduates (1980) Median Household $23,270 1 $13,270 $16,228 $22,713" $15,900" Income Percent Below 111 341 30% 13.1* 29* Poverty Level Percent Unemployed 861% 18.9% 13.8% 4.7" 13% Average Fertility Rate 1.7 23 2.3 -- 4.0 (1980) Percent of Households 10.9 37.7 23 11 24 Headed by Women (1980) *1979 T1981 $1982 Source: Adapted from the Report of the Secretary's Task Force on Black and Minority Health, Volume 1.357 95 Population Panel Report event, sensitivity to the psychosocial needs of the poor and the dignity of the individual is important in intervention programs. In this country, more women than men are poor. B. Characteristics of Important Population Subgroups Undoubtedly the summaries of the important population subgroups that follow reflect the influ- ences of a mix of the original cultures, the culture change process, health belief systems, and socio- economic and educational status. 1. Blacks a. Social and Demographic Characteristics Blacks, the largest single minority, represent nearly 12 percent of the U.S. population (table 6, page 95). The relatively low median age of 24.9 years reflects both birth rates and slightly shorter average life expectancy. Blacks have the highest rate among minorities of high school completion, but lag behind Anglos in receipt of high school di- plomas. The median income is lower for blacks than for any other group, with 34 percent living beneath the poverty level. Further, the average unemployment rate is substantially higher than that of other major minority groups. Household size is larger among blacks than among Anglos, and nearly 40 percent of black households are headed by a woman. b. Health Indicators According to the data summarized from 1979- 81,7 59,000 excess deaths occur each year among blacks compared with mortality rates for whites. Homicides alone account for 38 percent of the ex- cess deaths among black males younger than 45 years of age. Cardiovascular disease and cirrhosis account for another 18 percent, and infant 96 mortality for 23 percent. The infant mortality rate among blacks is twice that of whites. For black females younger than 45 years of age, infant mor- tality is responsible for 35 percent of the excess deaths, cardiovascular disease accounts for 17 percent, and homicide for 14 percent. Among black males younger than 70 years of age, the three leading causes for excess deaths are cardiovascu- lar disease (24 percent), homicide (19 percent), and cancer (16 percent). For black women younger than 70 years of age, cardiovascular disease ac- counts for 41 percent of excess deaths and cancer for 10 percent. In 1986, age-adjusted mortality rates from CHD were about 8 percent lower for black males than for white males. Age-adjusted CHD mortality rates for black women were almost 20 percent higher than for white women.*"? c. Risk Factors for Coronary Heart Disease in Blacks A National Center for Health Statistics report revealed that the mean total blood cholesterol levels in black and white adult men and women are similar," whereas the HDL-cholesterol levels are higher in black men than in white men.” Cigarette smoking prevalence is greater among blacks than whites, but the prevalence of heavy smoking is greater among white than black adults."**"> Obesity is also common in black women and may provide a partial explanation for their excess coronary disease risk.’ Blacks have a higher prevalence than whites of high blood pres- sure and of diabetes. Each of these risk factors causes roughly similar increases in risk for blacks and whites. d. EatingPatterns Among Blacks According to the 1985 Continuing Survey of Food Intakes by Individuals (CSFII), dietary in- takes of fat, saturated fatty acids, monounsatu- rated fatty acids, and polyunsaturated fatty acids, expressed as average percentages of total energy, are similar for black and white women. Data from Population Panel Report several surveys show that black adults on average consume fewer calories and more cholesterol than whites.*1°°!8 Mean intakes of some essential nutri- ents for black women, expressed as a percent of the RDA, are lower than for white women. As shown in the CSFII, mean intakes of vitamin B,, folacin, calcium, magnesium, iron, and zinc are lower for black than for white women, but only for calcium and magnesium are the differences such that black women are likely to be at much greater risk of de- ficiency compared with white women.'*"” Intakes of these nutrients by black men are lower than by white men, but black men achieve greater than 70 percent RDA average levels for all nutrients except folacin."” Although the CSFII sample of blacks contained a larger proportion of individuals be- low the poverty level, living in central cities, and residing in the Northeast and South, the differ- ences in intakes of essential nutrients do not ap- pear to be attributable to the sample selection. Mainstays of traditional black food patterns have tended to reflect a southern rural heritage, i.e, fried, barbecued, roasted, or smoked pork and chicken; rice and grits; dried legumes; sweet pota- toes; hot breads (biscuits, corn bread, and rolls); buttermilk; custards; and vegetables seasoned with smoked bacon, salt pork, ham hocks, or neck bones. In one study, use of some foods apparently was re- lated to perceptions of social status. Recipients of public welfare considered some “modern” items, such as unseasoned vegetables and variety in meals, beyond their reach.®"” Traditional black recipes are sometimes modified to suit modern tastes®’ as part of a process of acculturation, together with avoidance of many fatty, fried, and overcooked foods. It is likely that the food patterns of blacks vary according to the economic, regional, and social influences of communities. Thus generali- zations are inappropriate. 2. Hispanics a. Social and Demographic Characteristics Hispanics are the second largest minority in the United States, with approximately 18.9 million residents in 1987 or 7.5 percent of the United States population.’*”**" The Hispanic growth rate is now the fastest of any population group in the United States. The Hispanic minority has four major subgroups: Mexican-Americans at 10.9 million (58 percent); Puerto Ricans at 2.7 million (14 percent); Cubans at more than 1 million (6 percent); and other Latinos at 4 million (22 percent) including Dominicans, Haitians, and others from South and Central America. Of this Hispanic minority, 60 percent live in California, Texas, Colorado, New Mexico, Arizona, and Florida, and 50 percent live in cities of more than 1 million in population. The Hispanic population has a median age of 23 years. As of 1980, only 58 percent of Hispanics had earned high school diplomas, and 10 percent had graduated from college. In 1981, the median in- come for Hispanic families was $16,228. Thirty percent of Hispanics lived below the poverty level in 1982, and nearly 14 percent were unemployed. Hispanics have relatively large families, aver- aging 2.3 children per family. Fifty percent of all Hispanic families have five or more children. Twenty-three percent of all Hispanic households are headed by women; among Puerto Rican house- holds the proportion is even higher, at 40 percent. b. Health Indicators National mortality rates among Hispanics are unknown in the absence of a national data collec- tion system. Many states with substantial num- bers of Hispanic residents (e.g., Florida and New Jersey) have no Hispanic identifier on death cer- tificates. However, Spanish surname data from Texas and national death certificate data listing Mexico or Cuba as the country of birth provide the following information. Among Mexican-born Hispanic males, homicides and unintentional in- juries account for 98 percent of the total excess deaths. Cirrhosis accounts for the other 2 percent. For Mexican-born women, homicide and diabetes accounted for 100 percent of the total excess deaths.”” Hispanic men and women appear to have CHD mortality rates that are 5-15 percent lower than those of non-Hispanics. 97 Population Panel Report c. Risk Factors for Coronary Heart Disease in Hispanics Diabetes mellitus is a major health problem for Hispanics, especially for Mexican-Americans and Puerto Ricans.®””** The prevalence of obesity, noninsulin-dependent diabetes, high LDL-choles- terol and low HDL-cholesterol levels in Hispanics might be expected to increase their CHD risk. The prevalence of high blood pressure and physical inactivity is higher in Hispanics than in non-His- panics. However, the scarce data that exist indi- cate that Hispanic males have rates similar to those of other white males. Because the Hispanic catego- rization includes many different populations, avail- able data on the prevalence of elevated blood cho- lesterol are difficult to interpret. Cholesterol levels have been reported as being higher in Hispanics than in non-Hispanic whites in Texas, equal to those of non-Hispanic whites in California, and lower than those of non-Hispanic whites in Puerto Rico. Mexican-Americans with low socioeconomic status (SES) and low levels of acculturation have significantly worse CHD risk factor profiles than those in higher SES groups.®*#> d. Eating Patterns Among Hispanics Hispanics are not a homogeneous group, ei- ther in heritage or food patterns. Dominicans, Mexican-Americans, Cuban-Americans, and Puerto Ricans have distinctive native cuisines. The impact of these differences on nutrient intake is not well documented. Many traditional dishes are complicated and time consuming to prepare and hence tend to be reserved for special occasions. Traditional core dietary constituents such as rice, beans, cornmeal, squash, and other starchy vege- tables and chicken are compatible with choles- terol-lowering dietary recommendations. 98 3. Asian/Pacific Islanders a. Social and Demographic Characteristics Asian/Pacific Islanders accounted for an esti- mated 1.6 percent (3.7 million) of the total popula- tion in the United States in 1980," although the actual population has increased substantially be- cause of the immigration of Indochinese under the Refugee Resettlement Program. In 1980, 58 per- cent of Asian/Pacific Islanders were foreign born, the largest proportion of all minority groups. Asian/Pacific Islanders are also the most diverse minority group, originating from more than 20 different countries. The three largest subgroups are Chinese (812,000), Filipinos (781,000), and Japanese (716,000). The data on Asian/Pacific Islanders are heavily influenced by these three largest subgroups, which would mask important socioeconomic characteristics of the more recent Indochinese immigrants. According to the 1980 census, 56 percent of Asian/Pacific Islanderslived in the western United States. Thirty-three percentlived in California and 10 percent lived in Texas. The median age of Asian/Pacific Islanders is higher than that of other minorities, at 28.7 years. The median education level of Asian/Pacific Islanders is similar to that of nonminorities. Of Asian/PacificIslanders, 75 percent have completed high school, and approximately 33 percent hold a college degree. In 1980, the median household income for Asian/Pacific Islander families was higher than that of any other U.S. minority, at $22,713 (see table 6, page 95). However, this figure masks the true extent of poverty as it often represents the total of several incomes when families share houses. Median income levels vary substantially between groups, with the lowest, $12,840, being earned by the Vietnamese. The unemployment rate among Asian/Pacific Islanders in 1979 was 4.7 percent, and the poverty rate was 13.1 percent. Only 11 percent of Asian/Pacific Islander families are headed by women. Population Panel Report b. Health Indicators Compared with all other population groups in the United States, Asian/Pacific Islanders are on average reported to be the healthiest and have the lowest risk of early death.” For the entire popu- lation, there are no excess deaths reported for any cause. However, substantial variation does exist regarding the relative health status among Asian/ Pacific Islander subgroups. For example, native Hawaiians have excess deaths owing to cardiovas- cular disease, cancer, diabetes, infant mortality, and accidents. Other Asian/Pacific Islander groups such as recent immigrants from Southeast Asia have higher rates of tuberculosis, hepatitis, and anemia. c. Risk Factors for Coronary Heart Disease in Asian/Pacific Islanders In their native countries, Asians tend to have low blood cholesterol levels, a low prevalence of obesity, and a low prevalence of CHD. As immi- grants move from areas of low to high CHD preva- lence, they exhibit risk factor levels that approach those of the host-country population. For ex- ample, a comparison of risk factors in Japanese men with risk factors in Japanese-American men living in Hawaii and California showed clear gra- dients in blood cholesterol—181 mg/dL (4.68 mmol/L) for the Japanese, 218 mg/dL (5.64 mmol/L) for their counterparts in Hawaii, and 228 mg/dL (5.90 mmol/L) for the Japanese-Ameri- cans in California.*** Moreover, the Japanese men had lower intakes of SFA (16 g/day) and dietary cholesterol (464 mg/day), compared with their counterparts in Hawaii (SFA, 59 g/day; choles- terol, 545 mg/day) and California (SFA, 66 g/day; cholesterol, 533 mg/day). Similar changes could be expected for other Asian immigrants to the United States. d. Eating Patterns Among Asian/Pacific Islanders Asian Americans are a heterogeneous group; however, the nutritional habits of Asian/Pacific Islanders can be generalized to some extent. Rice is a primary source of calories, although the varie- ties used are often culturally specific.*” On the average, Asian/Pacific Islanders, compared with other groups, consume more vegetables, fruits, fish, and shellfish and consume less meat and dairy products. Different flavoring ingredients, cooking methods, and cultural rules governing the manner of service make each group's diet distinc- tive. As with many other ethnic groups, many traditional foods are receiving less emphasis and are being replaced by popular American foods. Hence, secondary and tertiary generations of Asian/Pacific Islander immigrants are becoming increasingly acculturated to nonminority dietary practices and, as a result, the consumption of ani- mal proteins, fats, and refined sugar is on the increase. 4. Native Americans a. Social and Demographic Characteristics The native American minority is heterogene- ous. Although it is composed of American Indi- ans, Aleuts, Alaskan Eskimos, and native Hawai- ians, the data reported from most national surveys refer primarily to American Indians and Alaskan natives.’”” According to the 1980 census, the native American minority is the smallest of all four groups—its total population of approximately 1.5 million represents less than 1 percent of the total U.S. population.” Fifty percent of native Ameri- cans live in the West or Southwest. Twenty-four percent live on reservations, and 8 percent live on historic trust sites. Most reservations have fewer than 1,000 inhabitants, and only one reservation has a population exceeding 100,000. Tribal cus- toms vary considerably. The median education level among native Americans is lower than that of all other minority groups. Only 31 percent of native Americans have graduated from high school, and only 7 percent hold a college degree. The median family income among native Americans in 1979 was $15,900. Twenty-nine percent of native Americans lived below the poverty level in 1979, and, in 1982, 13 percent were unemployed. 99 Population Panel Report The average native American family has 4.0 members, the largest of any population subgroup in the United States. In addition, native Americans have the highest total birth rate—approximately twice that of the nonminority rate. Twenty-four percent of native American households are headed by women. b. Health Indicators The native American life expectancy is 6 years less than that of nonminorities, and the median age is the lowest of all groups, at 22.4 years. Native American males under the age of 70 years experienced a 28 percent excess in deaths in 1979-81.%7 The three primary causes of the excess deaths were accidents (44 percent), cirrhosis (13 percent), and homicide (8 percent). Native Ameri- can women younger than 70 years of age also expe- rienced a 28 percent excess death rate. The three major causes of excess mortality were accidents (29 percent), cirrhosis (22 percent), and infant mor- tality (9 percent). Heart disease is a significant con- tributor to all-cause mortality of native Americans but is proportionately less a contributor in this subgroup than in the general population. The health of native Americans is a serious problem that must be addressed with a great deal of skill since these individuals have a special relationship with the Federal Government. c. Risk Factors for Coronary Heart Disease in Native Americans Both obesity and diabetes are major public health concerns for native Americans,*” especially among Southwestern desert tribes. Prevalence es- timates of obesity and diabetes as high as 75 per- cent have been reported. Alcohol use is widely prevalent. The data indicate a lower prevalence of elevated blood cholesterol, high blood pressure, and physical inactivity in native Americans than in the majority population. 100 d. Eating Patterns Among Native Americans Native Americans are a culturally diverse group. Data on the nutrient intakes of native Americans are scarce. A common impression is that the diet is high in fat. However, in one study of Navajo women, fat contributed only 30 percent of calories, mostly saturated (animal) fat.* An- other study in women indicated a fat level of about 37 percent of calories.” Dietary patterns are highly variable, depend- ing on the tribe, geographic location, and eco- nomic resources. Although most traditional foods are preferred and continue to be a source of pride, they are reserved mostly for ceremonial occa- sions.'#2062 Many elements of native American diets such as corn, beans, squash, wild game, fish, and wild plants are health promoting. There are Federal programs on reservations (e.g., EFNEP) that are targeted toward helping native Americans follow the dietary guidelines using traditional foods and commodities supplied by the Federal Government. 5. Summary More systematic investigation and documen- tation of the health, nutritional status, food prac- tices, and nutrient content of native foods of mi- nority groups are needed. In addition to food and nutrient intakes of individuals, information on the importance of particular foods, preparation tech- niques, format of meals, health beliefs, and related social and ecological factors should be collected, to help in designing strategies relevant to cultural practices.®® Population Panel Report VII. Screening the Population for Blood Cholesterol Level The development of consensus and recom- mendations by the scientific community for the definition and management of high blood choles- terol plus the advent of new technologies for blood cholesterol measurement have resulted in numer- ous blood cholesterol screening programs both within health care facilities and in the community. Many in the population have been screened out- side health care settings, and many more will be in the coming years. Public cholesterol screening that meets NCEP standards? has value as a supplement to the measurement of blood cholesterol that occurs as part of routine health care. Adherence to these standards will minimize the risk that poor screen- ing programs will be conducted that fail to meas- ure cholesterol accurately or educate screenees adequately about cholesterol, or that lead to inap- propriate referral and/or false reassurance. The rationale for public screening for blood cholesterol is found in two major areas. The first of these relates to the educational value of these pro- grams. Public screening programs must provide screened individuals with health information about cholesterol beyond the simple knowledge of their “cholesterol number.” This information is avail- able in the form of printed documents. The pres- ence of knowledgeable staff to answer reasonable questions from participants is also important. Documents should include a description of the meaning of a single blood cholesterol measure- ment, the importance of a prudent diet in lowering blood cholesterol, specifics about such a healthy diet, and the appropriate followup steps that indi- viduals should take depending on their blood cho- lesterol level and other CHD risk factors. The second major role that public screening programs can play is the traditional case-finding y function. Public screening programs can be par- ticularly helpful in detecting individuals with ele- vated blood cholesterol levels who might not be detected by the health care system. This includes young adults, middle-age males, individuals with low income or education, and minorities. Public screening programs should target their efforts to populations that have high proportions of these underserved individuals. One of the dangers of public screening pro- grams is the tendency of some individuals to use them as an alternative source of medical care. This is inappropriate. Screening programs should not be a substitute for a health care visit, nor should they replace routine medical monitoring of blood cholesterol levels for anyone already under treat- ment. While self responsibility for control of cho- lesterol is important, issues such as drug side effects, changes in therapy, and complications require the judgment and experience of a health care professional. Public screening programs, ex- cept in rare instances, are not equipped to provide this medical advice. Public screening programs should provide blood cholesterol readings at reasonable cost. It is recognized thatany recommendation that targeted screening should include low-income and low- education groups means that subsidization by pub- lic and private resources may be necessary to en- courage these individuals to attend a screening program. In addition, efforts must be made to develop sources of referral and followup that pro- vide care for these target groups. 101 Population Panel Report Only the measurement of total blood choles- terol is currently recommended for public screen- ing programs. New technologies for the direct measurement of LDL- and HDL-cholesterol and the potential benefit of such measurements have been considered by the panel. At present, reliable measures for field settings, while under development, are not yet available. Because of concerns about cost, the meaning of a single HDL- cholesterol measurement, and quality control, the panel does not recommend HDL-cholesterol measurement as presently suitable for public screening efforts.” When better technology be- comes available and has validated accuracy and precision, it may become appropriate to add this measurement. Estimation of the effects of screening and treat- ment for cholesterol on the costs of health care is important. Reasonable projections may rely on methods similar to those used for hypertension™ and could be based on the results of epidemiologic research and on the various trials that have exam- ined the efficacy of cholesterol-lowering agents. Cholesterol screening programs, even if they lead to net increases in the costs of health care, are reasonably cost-effective if they are efficiently run, targeted at groups likely to contain high-risk indi- viduals, and successful in achieving long-term reductions in blood cholesterol. Of particular importance is the need to main- tain high quality programs while maximizing the cost-effectiveness of screening programs. Cost- effective cholesterol screening programs will have the following characteristics: * Program marketing strategies that reach the hard-to-reach; e Staffing matched to level of program activity; e Well-organized screening and educational processes; eo Well-trained staff; e Staff incentives for productivity; * Low program overhead; 102 * Equipment whose cost is suitable to needs; e Screening targeted at high-risk, underserved populations; * Rescreening limited to pre-established inter- vals, with avoidance of rescreening “worried well” on demand; eo Accurate cholesterol measurement, to mini- mize false-positive and false-negative values; * Clearbehavior change messages tailored to the educational level and cultural patterns of the population; * Referral to medical care limited to high-risk individuals; and * Followup to ensure successful completion of referral. Cholesterol screening programs, as well as phy- sicians who treat identified high-risk individuals, are urged to give high priority to opportunities to reduce costs, as well as enhance the effectiveness of their activities. In October 1988, the NHLBI convened a Work- shop on Public Screening for High Blood Choles- terol, to review the available scientific knowledge concerning screening and to develop guidelines for screening. The report from that workshop contains further details.” Population Panel Report VIII. AHA ATP CHD DHHS FDA HDL LDL LSP MRFIT MUFA NCEP NCHS NHANES NHLBI NIH NRC PUFA RDA SFA USDA Glossary American Heart Association Adult Treatment Panel coronary heart disease United States Department of Health and Human Services Food and Drug Administration high density lipoprotein low density lipoprotein Laboratory Standardization Panel Multiple Risk Factor Intervention Trial monounsaturated fatty acids National Cholesterol Education Program National Center for Health Statistics National Health and Nutrition Examination Survey National Heart, Lung, and Blood Institute National Institutes of Health National Research Council polyunsaturated fatty acids recommended dietary allowance saturated fatty acids United States Department of Agriculture 103 Ser were vv rs ee BE } Re htm ante om - ) ) a - EY i Population Panel Report IX. Bibliography 1. MRFIT Research Group, personal communica- tion, 1989. 2. National Center for Health Statistics— National Heart, Lung, and Blood Institute Collaborative Lipid Group. Trends in serum cholesterol level among U.S. adults aged 20-74 years. Data from the National Health and Nutrition Examination Survey, 1960 to 1980. Journal of the American Medical Association 257:937-42, 1987. 3. National Cholesterol Education Program. Re- port of the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Bethesda, MD: U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, Na- tional Heart, Lung, and Blood Institute, NIH Pub. No. 88-2925, January 1988. 4. Truswell AS. The development of dietary guidelines. Food Technology in Australia 35:498-502, 1983. 5. National Research Council. Diet and Health: Implications for Reducing Chronic Disease Risk. Washington, DC: National Academy Press, 1989. 6. Gyarfas I, Stanley K. Diet and chronic dis- eases: A global perspective. Acta Cardiologica XLIV:487-89, 1989. 7. World Health Organization Study Group on Diet, Nutrition, and Prevention of Noncom- municable Diseases. Diet and chronic dis- eases: A global perspective (in press). 8. Inter-Society Commission for Heart Disease Resources. Primary prevention of the athero- sclerotic diseases. Circulation 42:A55, 1970. 9. Inter-Society Commission for Heart Disease Resources. Optimal resources for primary prevention of atherosclerotic diseases. Circulation 70:155A-96A, 1984. 10. 11. 12. 13. 14. 15. American College of Cardiology. Elev- enth Bethesda Conference: Prevention of coronary heart disease, September 27- 28, 1980, Bethesda, MD. American Journal of Cardiology 47:713-76, 1981. National Research Council. Committee on Technological Options to Improve the Nutri- tional Attributes of Animal Products, Board on Agriculture. Designing Foods: Animal Products Options in the Marketplace. Washing- ton, DC: National Academy Press, 1988. U.S. Department of Agriculture/Department of Health and Human Services. Nutrition and Your Health: Dietary Guidelines for Americans, 2nd ed. Washington, DC: U.S. Government Printing Office, HG-232, 1985. National Cancer Institute. Cancer Control Objectives for the Nation—1985-2000. NCI Monographs, No. 2. Bethesda, MD: U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, 1986. Human Nutrition Information Service. Nationwide Food Consumption Survey: Continu- ing Survey of Food Intakes by Individuals, Women 19-50 Years and Their Children 1-5 Years, 1 Day. Hyattsville, MD: U.S. Depart- ment of Agriculture, CSFII Report No. 85-1, November 1985. Human Nutrition Information Service. Nationwide Food Consumption Survey: Con- tinuing Survey of Food Intakes by Individuals, Men 19-50 Years, 1 Day. Hyattsville, MD: U.S. Department of Agriculture, CSFII Report No. 85-3, 1985. 105 Population Panel Report 16. Human Nutrition Information Service. Nationwide Food Consumption Survey: Continu- ing Survey of Food Intakes by Individuals, Women 19-50 Years and Their Children 1-5 Years, 4 Days, 1985. Hyattsville, MD: U.S. Department of Agriculture, CSFII Report No. 85-4, 1985. 17. Human Nutrition Information Service. Nationwide Food Consumption Survey: Continu- ing Survey of Food Intakes by Individuals, Women 19-50 Years and Their Children 1-5 Years, 4 Days, 1986. Hyattsville, MD: U.S. Department of Agriculture, CSFII Report No. 86-3, 1986. 18. Grundy SM, Barrett-Connor E, Rudel LL. Workshop on the impact of dietary choles- terol on plasma lipoproteins and atherogenesis. Arteriosclerosis 8:95-101, 1988. 19. Human Nutrition Information Service. Diet- ary Guidelines and Your Diet: Preparing Foods and Planning Menus; Making Bag Lunches, Snacks and Desserts; Shopping for Food and Making Meals in Minutes; Eating Better When Eating Out. Hyattsville, MD: U.S. Depart- ment of Agriculture, HG-232 #8-11, 1989. 20. Human Nutrition Information Service. Dietary Guidelines and Your Diet. Hyattsville, MD: U.S. Department of Agri- culture, HG-232 #1-7, 1986. 21. Cronin FJ, Shaw SM, Krebs-Smith SM, et al. Developing a food guidance system to implement the dietary guidelines. Journal of Nutrition Education 19:282-302, 1987. 22. National Heart, Lung, and Blood Institute. Recommendations regarding public screen- ing for measuring blood cholesterol: Sum- mary of a National Heart, Lung, and Blood Institute workshop, October 1988. Archives of Internal Medicine 149:2650-54, 1989. 23. Stanbury JB, Wyngaarden JB, Fredrickson DS, et al., eds. The Metabolic Basis of Inherited Disease Part 4: Disorders of Lipoprotein and Lipid Metabolism, 5th ed. New York: McGraw-Hill Book Co., 589-750, 1983. 106 24. 25. 26. 27 28. 29. 30. 31. Havel RJ. Origin, metabolic rate, and meta- bolic function of plasma lipoproteins. In: Steinberg D, Olefsky JM, eds. Hyper- cholesterolemia and Atherosclerosis: Pathogene- sis and Prevention. New York: Churchill Livingstone, Inc., 117-41, 1986. Dahlen GH. Lipoprotein (a) in relation to atherosclerotic disease. In: Widhalm K, Naito HK, eds. Recent Aspects of Diagnosis and Treatment of Lipoprotein Disorders: Impact on Prevention of Atherosclerotic Disease. New York: Alan R. Liss, Inc., 27-36, 1988. Hoefler G, Harnoncourt F, Paschke E, et al. Lipoprotein Lp(a): A risk factor for myocar- dial infarction. Arteriosclerosis 8:398-401, 1988. Brown MS, Goldstein JL. Lipoprotein me- tabolism in the macrophage: Implications for cholesterol deposition in atherosclerosis. An- nual Review of Biochemistry 52:233-61, 1983. Goldstein JL, Brown MS. Regulation of low- density lipoprotein receptors: Implications for pathogenesis and therapy of hyper- cholesterolemia and atherosclerosis. Circula- tion 76:504-7, 1987. Brown MS, Goldstein JL. A receptor- mediated pathway for cholesterol homeostasis. Science 232:34-47, 1986. Pitas RE, Innerarity TL, Mahley RW. Foam cell in explants of atherosclerotic rabbit aortas have receptors for 3-very low density lipoprotein and modified low density lipoproteins. Arteriosclerosis 3:2-12, 1983. Kita T, Nagano Y, Yokode M, et al. Probucol prevents the progression of atherosclerosis in Watanabe heritable hyperlipidemic rabbits, and animal model for familial hyper- cholesterolemia. Proceedings of the National Academy of Sciences USA 84:5928-31, 1987. Population Panel Report 32. 33. 34. 35. 36. 37. 38. 39. Carew TE, Schwenke DC, Steinberg D. Anti- atherogenic effect of probucol unrelated to its hypocholesterolemic effect: Evidence that antioxidants in vivo can selectively inhibit low density lipoprotein degradation in macrophage-rich fatty streaks and slow the progression of atherosclerosis in the Watanabe heritable hyperlipidemic rabbit. Proceedings of the National Academy of Sciences USA 84:7725-29, 1987. Haberland ME, Fong D, Cheng L. Malondial- dehyde-altered apo B protein occurs in arterial lesions of Watanabe heritable hyper- lipidemic rabbits. Arteriosclerosis 7:526-27a, 1987. Tabas I, Weiland DA, Tall AR. Stimulation of cholesteryl ester storage in macrophage- like ]J774 cells by unmodified human low density lipoprotein. Proceedings of the Na- tional Academy of Sciences USA 82:416-20, 1985. Koo C, Wernette-Hammond ME, Innerarity TL. Uptake of canine B-very low density lipoproteins by mouse peritonealmacro- phages is mediated by a low density lipopro- tein receptor. Journal of Biological Chemistry 261:11194-201, 1986. Van Lentern BJ, Fogelman AM, Jackson RL, et al. Receptor-mediated uptake of remnant lipoproteins by cholesterol-loaded human monocyte-macrophages. Journal of Biological Chemistry 260:8783-88, 1985. Steinberg D. Lipoproteins and atherosclero- sis: A look back and a look ahead. The George Duff Memorial Lecture. Arteriosclero- sis 4:283-301, 1983. Ross R. The pathogenesis of atherosclero- sis—an update. New England Journal of Medicine 314:488-500, 1986. Kritchevsky D, Moyer AW, Tesar WC, et al. Effect of cholesterol vehicle in experimental atherosclerosis. American Journal of Physiology 178:30-32, 1954. 40. 41. 42. 43. 44. 45. 46. 47. 48. Olson RE. Atherosclerosis—a primary hepatic or vascular disease. Perspectives in Biology and Medicine 2:84-90, 1958. Anitschkow N, Chalatow S. Ueber experi- mentelle cholesterinsteatose und ihre be- deutung fur die entstehung einiger patholo- gischer prozesse. Zentralblatt fur Allegmeine Pathologie und Pathologische Anatomie 24:1-9, 1913. McGill HC Jr, Strong JP, Holman RL, et al. Arterial lesions in the Kenya baboon. Circu- lation Research 8:670-79, 1960. Adams CWM. Atherosclerosis in man, other mammals, and birds. Biological Reviews of the Cambridge Philosophical Society 39:372-423, 1964. Strong JP, Eggen DA, Jirge SK. Atheroscle- rotic lesions produced in baboons by feeding an atherogenic diet for four years. Experi- mental and Molecular Pathology 24:320-32, 1976. Kritchevsky D, Davidson LM, Kim HK, et al. Influence of semiurified diets on athero- sclerosis in African green monkeys. Experi- mental and Molecular Pathology 26:28-51, 1977. Kritchevsky D, Davidson LM, Kim HK, et al. Influence of type of carbohydrate on athero- sclerosis in baboons fed semi-purified diets plus 0.1% cholesterol. American Journal of Clinical Nutrition 33:1869-87, 1980. Babiak J, Nichols AV, Gong EL, et al. Effects of dietary polyunsaturated and saturated fats on lipoproteins in the baboon. Athero- sclerosis 57:1-17, 1985. Taylor CB, Patton DE, Cox GE. Athero- sclerosis in Rhesus monkeys. VI. Fatal myocardial infarction in a monkey fed fat and cholesterol. Archives of Pathology 76:404- 23, 1963. 107 Population Panel Report 49. 50. 51. 52. 53. 54. 55. 56. 57. 108 Bullock BC, Lehner ND, Clarkson TB, et al. Comparative primate atherosclerosis. Experi- mental and Molecular Pathology 22:151-75, 1975. Wissler RW, Vesselinovitch D. Comparative pathogenic patterns in atherosclerosis. Advances in Lipid Research 6:181-206, 1974. Pertsemlidis D, Kirchman EH, Ahrens EH |r. Regulation of cholesterol metabolism in the dog. I. Effects of complete bile diversion and of cholesterol feeding on absorption, synthe- sis accumulation, and excretion rates meas- ured during life. Journal of Clinical Investiga- tion 52:2353-65, 1973. Pertsemlidis D, Kirchman EH, Ahrens EH Jr. Regulation of cholesterol metabolism in the dog. II. Effects of complete bile diversion and of cholesterol feeding on pool sizes of tissue cholesterol measured at autopsy. Journal of Clinical Investigation 52:2368-78, 1973. Wilson JD. The qualification of cholesterol excretion and degradation in the isotopic steady state in the rat: The influence of dietary cholesterol. Journal of Lipid Research 5:409-17, 1964. Mahley RW, Weisgraber KH, Innerarity T. Canine lipoproteins and atherosclerosis II. Characterization of the plasma lipoproteins associated with atherogenic and nonathero- genic hyperlipidemia. Circulation Research 35:722-33, 1974. Jokinen MP, Clarkson TB, Prichard RW. Animal models in atherosclerosis research. Experimental and Molecular Pathology 42:1-28, 1985. Rudel LL. Plasma lipoproteins in athero- genesis in nonhuman primates. In: Kalter SS, ed. The Use of Nonhuman Primates in Cardiovascular Disease. Austin, TX: Univer- sity of Texas Press, 37-57, 1980. McGill HC Jr, McMahan CA, Kruski AW, et al. Relationship of lipoprotein cholesterol 58. 59, 60. 61. 62. 63. 64. 65, 66. concentrations to experimental atherosclero- sis in baboons. Arteriosclerosis 1:3-12, 1981. McGill HC Jr, Carey KD, McMahan CA, et al. Effects of two forms of hypertension on atherosclerosis in the hyperlipidemic baboon. Arteriosclerosis 5:481-93, 1985. Rudel LL, Parks JS, Bond MG. LDL hetero- geneity and atherosclerosis in nonhuman primates. Annals of the New York Academy of Sciences 454:248-53, 1985. Spady DK, Dietschy J. Dietary saturated triglycerides suppress hepatic low density lipoprotein receptors in the hamster. Pro- ceedings of the National Academy of Sciences USA 82:4526-30,1985. Schonfeld G, Patsch W, Rudel LL, et al. Effects of dietary cholesterol and fatty acids on plasma lipoproteins. Journal of Clinical In- vestigation 69:1072-80, 1981. Nicolosi RJ, Hayes KC. Composition of plasma and nascent very low density lipo- protein from perfused livers of hyper- cholesterolemic squirrel monkeys. Lipids 15:549-54, 1980. Zilversmit DB. A proposal linking athero- sclerosis to the interaction of endothelial lipoprotein lipase with triglyceride-rich lipoproteins. Circulation Research 33:633-38, 1973. Mahley RW, Weisgraber KH, Innerarity T. Atherogenic hyperlipoproteinemia induced by cholesterol feeding in the patas monkey. Biochemistry 15:2979-85, 1976. Melchior GW, Mahley RW, Buckhold DK. Chylomicron metabolism during dietary- induced hypercholesterolemia in dogs. Journal of Lipid Research 22:598-609, 1981. Armstrong ML. Regression of atherosclerosis. Atherosclerosis Review 1:137- 88, 1975. Population Panel Report 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. Stary HC. Regression of atherosclerosis in primates. Virchows Arch, A: Pathological Anatomy and History 383:117-34, 1979. Malinow MR. Experimental models of atherosclerosis regression. Atherosclerosis 48:105-18, 1983. Fincham JE, Woodroof CW, van Wyk M], et al. Formation and regression of athero- sclerosis in Vervet monkeys by diets realistic for westernized people. Atherosclerosis 66:205-13, 1987. McGill HC Jr, McMahan CA, Kruski AW, et al. Responses of serum lipoproteins to dietary cholesterol and type of fat in the baboon. Arteriosclerosis 5:337-44, 1981. Nicolosi RJ, Hayes KC, el Lozy M, et al. Hypercholesterolemia and triglyceride secretion rates in monkeys fed different dietary fats. Lipids 12:936-40, 1977. Nicolosi RJ, Hojnacki JL, Llansa N, et al. Diet and lipoprotein influence on primate atherosclerosis. Proceedings of the Society of Biology and Medicine 156:1-7, 1977. Stephan ZF, Hayes KC. Mechanism underly- ing the dietary fat impact on lipoproteins, atherogenesis, and gallstones. Pathology and Immunopathology Research 5:468-90, 1986. McGill HC Jr, McMahan CA, Mott GE, et al. Effects of selective breeding on the cholesterolemic responses to dietary satu- rated fat and cholesterol in baboons. Arterio- sclerosis 8:33-39, 1988. Spady DK, Dietschy J. Interaction of dietary cholesterol and triglycerides in the regula- tion of hepatic low density lipoprotein transport in the hamster. Journal of Clinical Investigation 81:300-09, 1988. Fox JC, McGill HC Jr, Carey KD, et al. In vivo regulation of hepatic LDL receptor mRNA in the baboon: Differential effects of saturated and unsaturated fat. Journal of Biological Chemistry 262:7014-20, 1987. 77. 78. 79. 80. 81. 82. 84. 85. 86. Ahrens EH, Hirsch J, Insull W Jr, et al. The influence of dietary fats on serum-lipid levels in man. Lancet 1:943-53, 1957. Hegsted DM, McGandy RB, Myers ML, et al. Quantitative effects of dietary fat on serum cholesterol in man. American Journal of Clinical Nutrition 17:281-95, 1965. Keys A, Anderson JT, Grande F. Serum cholesterol response to changes in the diet: II. The effect of cholesterol in the diet. Me- tabolism 14:759-65, 1965. Mattson FH, Grundy SM. Comparison of effects of dietary saturated, monounsatu- rated, and polyunsaturated fatty acids on plasma lipids and lipoproteins in man. Journal of Lipid Research 26:194-202, 1985. Grundy SM. Comparison of monounsatu- rated fatty acids and carbohydrates for lowering plasma cholesterol. New England Journal of Medicine 314:745-48, 1986. Grundy SM, Florentin L, Nix D, et al. Com- parison of monounsaturated fatty acids and carbohydrates for reducing raised levels of plasma cholesterol in man. American Journal of Clinical Nutrition 47:965-69, 1988. . Grundy SM, Vega GL. Plasma cholesterol responsiveness to saturated fatty acids. American Journal of Clinical Nutrition 47:822- 24, 1988. Bonanome A, Grundy SM. Effect of dietary stearic acid on plasma cholesterol and lipo- protein levels. New England Journal of Medi- cine 318:1244-48, 1988. Beveridge JMR, Connell WF, Haust HL, et al. Dietary cholesterol and plasma cholesterol levels in man. Canadian Journal of Biochemical Physiology 37:575-82, 1959. Hashim SA, Arteaga A, van Itallie TB. Effect of a saturated medium-chain triglyceride on serum lipids in man. Lancet 1:1105-08, 1960. 109 Popr tlation Panel Report 87. 88. 29. 90. 91. 92. 93. 94. 95. 96. 110 Grande F. Dog serum lipid responses to dietary fats differing in the chain length of the saturated fatty acids. Journal of Nutrition 76:255-64, 1962. Keys A, Anderson JT, Grande F. Serum cholesterol response to changes in the diet: IV. Particular saturated fatty acids in the diet. Metabolism 14:776-87, 1965. Kinsell LW, Michael GD. Letter to the editor. American Journal of Clinical Nutrition 3:247-53, 1955. Keys A, Anderson JT, Grande F. Prediction of serum cholesterol responses of man to change in fats in the diet. Lancet 2:959-66, 1957. Spritz N, Grundy SM, Ahrens EH Jr. Studies of the mechanism of diet-induced alterations of plasma cholesterol. Journal of Clinical In- vestigation 42:981, 1963. Hellman L, Rosenfeld RS, Insull W Jr, et al. Intestinal excretion of cholesterol: A mecha- nism for regulation of plasma levels. Journal of Clinical Investigation 36:898-99, 1957. Moore RB, Anderson JT, Taylor HL, et al. Effect of dietary fat on the fecal excretion of cholesterol and its degradation products in man. Journal of Clinical Investigation 47:1517-34, 1968. Connor WE, Witiak DT, Stone DB, et al. Cholesterol balance and fecal neutral steroid and bile acid excretion in normal men fed dietary fats of different fatty acid composi- tion. Journal of Clinical Investigation 48:1363- 75, 1969. Grundy SM. Effects of polyunsaturated fats on lipid metabolism in patients with hypertriglyceridemia. Journal of Clinical Investigation 55:269-82, 1975. Spritz NE, Ahrens EH Jr, Grundy SM. Sterol balance in man as plasma choles- terol concentrations are altered by ex- change of dietary fats. Journal of Clinical Investigation 44:1482-93, 1968. 97. 98. 99. 100. 101. 102. 103. 104. Grundy SM, Ahrens EH Jr. The effects of unsaturated dietary fats on absorption, ex- cretion, synthesis, and distribution of cholesterol in man. Journal of Clinical Inves- tigation 49:1135-52, 1970. Spritz N, Mishkel MA. Effects of dietary fats on plasma lipids and lipoproteins: A hypothesis for the lipid-lowering effect of unsaturated fatty acids. Journal of Clinical Investigation 48:78-86, 1969. Nichaman MZ, Olson RE, Sweeley CC. Metabolism of linoleic acid-1-14-C in normolipemic and hyperlipemic humans fed linoleate diets. American Journal of Clinical Nutrition 20:1070-83, October 20, 1967. Shepherd J, Packard CJ, Grundy SM, et al. Effects of saturated and polyunsaturated fat diets on the chemical composition and metabolism of low density lipoproteins in man. Journal of Lipid Research 21:91-99, 1980. Vega GL, Groszek E, Wolf R, et al. Influ- ence of polyunsaturated fats on composi- tion of plasma lipoproteins and apolipo proteins. Journal of Lipid Research 32:811-22, 1982. Mensink RP, Katan MB. Effect of a diet en- riched with monounsaturated or polyun- saturated fatty acids on levels of low- density and high-density lipoprotein cholesterol in healthy women and men. New England Journal of Medicine 321:436- 41, 1989. Nestel PJ, Barter P. Metabolism of palmitic and linoleic acids in man: Differences in turnover and conversion to glycerides. Clinical Science 40:345-50, 1971. Shepherd J, Packard CJ, Patsch JR, et al. Effects of dietary polyunsaturated and saturated fat on the properties of high density lipoproteins and the metabolism of apolipoprotein A-I. Journal of Clinical Investigation 61:1582-92, 1978. Population Panel Report 105. 106. 107. 108. 109. 110. 111. 112. 113. Schwandt P, Janetschek PP, Weisweiler P. High density lipoproteins unaffected by dietary fat modification. Atherosclerosis 44:9- 17, 1982. Weisweiler P, Janetschek PP, Schwandt P. Influence of polyunsaturated fats and fat restriction on serum lipoproteins in humans. Metabolism 34:83-87, 1985. Harris WS, Connor WE, McMurry MP. The comparative reduction of the plasma lipids and lipoproteins by dietary poly- unsaturated fats: Salmon oil versus vege- table oils. Metabolism 32:179, 1983. Phillipson BE, Rothrock DW, Connor WE, et al. The reduction of plasma lipids, lipoproteins, and apoproteins in hyper- triglyceridemic patients by dietary fish oils. New England Journal of Medicine 312:1210, 1985. Nestel PJ, Connor WE, Reardon MF, et al. Suppression by diets rich in fish oils of very low density lipoprotein production in man. Journal of Clinical Investigation 74:82-89, 1984. Connor WE. Hypolipidemic effects of dietary omega-3 fatty acids in normal and hyperlipidemic humans: Effects and mechanisms. In: Simopoulos A, Kifer RR, Martin RE, eds. Health Effects of Polyun- saturated Fatty Acids in Seafoods. New York: Academic Press, 173-210, 1986. Illingworth DR, Harris WS, Connor WE. Inhibition of low density lipoprotein syn- thesis by dietary omega-3 fatty acids in humans. Arteriosclerosis 4:270-75, 1984. Simons LA, Hickie JB, Balasubramaniam S. On the effects of dietary n-3 fatty acids (MaxEpa) on plasma lipids and lipoproteins in patients with hyperlipidaemia. Athero- sclerosis 54:75-88, 1985. Rogers S, James KS, Butland BK, et al. Effects of a fish oil supplement on serum lipids, blood pressure, bleeding time, 114. 115. 116. 117. 118. 119. 120. 121. haemostatic and rheological variables: A double blind randomised controlled trial in healthy volunteers. Atherosclerosis 63:137- 43,1987. Mensink RP, Katan MB. Effects of monoun- saturated fatty acids versus complex carbo- hydrates on high-density lipoproteins in healthy men and women. Lancet 1:122-25, 1987. Ahrens EH Jr, Hirsch J, Oette K, et al. Car- bohydrate-induced fat-induced lipemia. Transactions of the Association of American Physicians 74:134-46, 1961. Knittle JL, Ahrens EH Jr. Carbohydrate metabolism in two forms of hyper- glyceridemia. Journal of Clinical Investigation 43:485-95, 1964. Antonis A, Bersohn I. The influence of diet on serum lipids in South African White and Bantu prisoners. American Journal of Clinical Nutrition 10:484-99, 1962. Knuiman JT, West CE, Katan MB, et al. Total cholesterol and high density lipopro- tein cholesterol levels in populations differ- ing in fat and carbohydrate intake. Arterio- sclerosis 7:612-19, 1987. Grundy SM, Nix D, Whelan M, et al. Com- parison of three cholesterol-lowering diets in normolipidemic men. Journal of the American Medical Association 256:2351-55, 1986. Slater G, Mead J, Dhopeshwarkar G, et al. Plasma cholesterol and triglycerides in men with added eggs in the diet. Nutrition Reports International 14:249-60, 1976. Porter MW, Yamanaka W, Carlson SD, et al. Effects of dietary egg on serum cholesterol and triglycerides in human males. Ameri- can Journal of Clinical Nutrition 30:490-95, 1977. 111 Population Panel Report 122. 123. 124. 125. 126. 127. 128. 129, 130. 131. 112 Flynn MA, Nolph GB, Flynn TC, et al. Effect of dietary egg on human serum cholesterol and triglyceride. American Journal of Clinical Nutrition 32:1050-57, 1979. Dawber TR, Nickerson RJ, Brand FN, et al. Eggs, serum cholesterol, and coronary heart disease. American Journal of Clinical Nutrition 36:617-25, 1982. Edington JD, Geekie M, Carter R, et al. Serum lipid response to dietary cholesterol in subjects fed a low-fat, high-fiber diet. American Journal of Clinical Nutrition 50:58- 62, 1989. Mattson FH, Erickson BA, Kligman AM. Effect of dietary cholesterol on serum cho- lesterol in man. American Journal of Clinical Nutrition 25:589-94, 1972. Nestel P, Tada N, Billington T, et al. Changes in very low density lipoproteins with cholesterol loading in man. Metabo- lism 4:398-405, 1982. Beynen AC, Katan MB. Effect of egg yolk feeding on the concentration and composi- tion of serum lipoproteins in man. Athero- sclerosis 54:157-66, 1985. Hegsted DM. Serum-cholesterol response to dietary cholesterol: A re-evaluation. American Journal of Clinical Nutrition 44:299- 305, 1986. Katan MB, Beynen AC. Characteristics of human hypo- and hyperresponders to die- tary cholesterol. American Journal of Epi- demiology 125:387-99, 1987. Katan MB, Berns MAM, Glatz JFC, et al. Congruence of individual responsiveness to dietary cholesterol and to saturated fat in man. Journal of Lipid Research 29:883-92, 1988. McNamara JD, Kolb R, Parker S, et al. Heterogeneity of cholesterol homeostasis in man: Response in changes to dietary fat quality and cholesterol quantity. Journal of Clinical Investigation 79:1729-39, 1987. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. Frantz ID Jr, Dawson EA, Ashman PL. Test of effect of lipid lowering by diet on cardio- vascular risk: The Minnesota Coronary Survey. Arteriosclerosis 9:129-35, 1989. Beynen AC, Katan MB. Reproducibility of the variations between humans in the re- sponse of serum cholesterol to cessation of egg consumption. Atherosclerosis 57:19-31, 1985. Mahley RW, Bersot TP, Innerarity TL, et al. Alterations in human high-density lipopro- teins, with or without increased plasma- cholesterol, induced by diets high in cholesterol. Lancet 2:807-09, 1978. Egusa G, Beltz WF, Grundy SM, et al. The influence of obesity on the metabolism of apolipoprotein B in man. Journal of Clinical Investigation 76:596-603, 1985. Anderson JT, Lawler A, Keys A. Weight gain from simple overeating. II. Serum lipids and blood volume. Journal of Clinical Investigation 36:81-88, 1957. Galbraith WB, Connor WE, Stone DB. Weight loss and serum lipid changes in obese subjects given low calorie diets of varied cholesterol intake. Annals of Internal Medicine 64:268-75, 1966. Contaldo F, Strazzullo P, Postiglione A, et al. Plasma high density lipoprotein-choles- terol levels in severe obesity after stable weight loss. Atherosclerosis 37:163-67, 1980. Brownell KD, Stunkard AJ. Differential changes in plasma high-density lipopro- tein-cholesterol levels in obese men and women during weight reduction. Archives of Internal Medicine 141:1142-46, 1981. Kesaniemi YA, Grundy SM. Increased low density lipoprotein production associated with obesity. Arteriosclerosis 3:170-77, 1983. Davis TA, Anderson EC, Ginsberg AV, et al. Weight loss improves lipoprotein lipid profiles in patients with hyper- cholesterolemia. Journal of Laboratory and Clinical Medicine 104:447-54, 1985. Population Panel Report 142. 143. 144. 145. 146. 147. 148. 149. 150. Wolf RN, Grundy SM. Influence of weight reduction on plasma lipoproteins in obese patients. Arteriosclerosis 3:160-69, 1983. Caggiula AW, Christakis G, Farrand M, et al. The Multiple Risk Factor Intervention Trial (MRFIT): IV. Intervention on blood lipids. Preventive Medicine 10:443-75, 1981. Anderson JW, Story L, Sieling B, et al. Hy- pocholesterolemic effects on high fiber diets rich in water soluble plant fiber. Journal of the Canadian Dietetic Association 45:140-48, 1984. Jenkins DJA, Reynolds D, Slavin B, et al. Dietary fiber and blood lipids: Treatment of hypercholesterolemia with guar crispbread. American Journal of Clinical Nutrition 33:575-81, 1980. Zavoral JH, Hannan P, Fields DJ, et al. The hypolipidemic effect of locust bean gum food products in familial hypercholesterol- emic adults and children. American Journal of Clinical Nutrition 38:285-94, 1983. Superko HR, Haskell WL, Sawrey-Kubicek L, et al. Effects of solid and liquid guar gum on plasma cholesterol and triglyceride concentrations in moderate hyper- cholesterolemia. American Journal of Cardiol- ogy 62:51-55, 1988. Bell LP, Hectorne K, Helenbeth R, et al. Cholesterol-lowering effects of psyllium hydrophilic mucilloid. Journal of the Ameri- can Medical Association 261:3419-23, 1989. Van Horn L, Liu K, Parker D, et al. Serum lipid response to oat product intake with a fat-modified diet. Journal of the American Dietetic Association 86:759-64, 1986. Wilson WS, Hulley SB, Borrows MI, et al. Serial lipid and lipoprotein responses to the AHA fat controlled diet. American Journal of Medicine 51:491-504, 1971. 131. 152. 153. 154. 155. 156. 157. 158. Pooling Project Research Group. Relation- ships of blood pressure, serum cholesterol, smoking habit, relative weight and ECG ab- normalities to incidence of major coronary events: Final report of the Pooling Project. Journal of Chronic Diseases 31:201-306, 1978. Kannel WB, Dawber TR, Kagan A, et al. Factors of risk in the development of coro- nary heart disease—six year follow-up ex- perience: The Framingham Study. Annals of Internal Medicine 55:33-50, 1961. Stokes J 111, Kannel WB, Wolf PA, et al. The relative importance of selected risk factors for various manifestations of cardiovascular disease among men and women from 35 to 64 years old: 30 years of follow-up in the Framingham Study. Circulation 75(suppl V):65-73, 1987. Sempos C, Fulwood R, Haines C, et al. The prevalence of high blood cholesterol levels among adults in the United States. Journal of the American Medical Association 262:45-52, 1989. Gordon T, Castelli W, Hjortlan M, et al. High density lipoprotein as a protective factor against coronary heart disease: The Framingham Study. American Journal of Medicine 62:707-14, 1977. Gordon T, Kannel WB, Castelli WP, et al. Lipoproteins, cardiovascular disease, and death. The Framingham Study. Archives of Internal Medicine 141:1128-31, 1981. Castelli WP, Garrison RJ, Wilson PW, et al. Incidence of coronary heart disease and lipoprotein cholesterol levels. The Fra- mingham Study. Journal of the American Medical Association 256:2835-38, 1986. Gordon T, Garcia-Palmieri MR, Kagan A, et al. Differences in coronary heart disease mortality in Framingham, Honolulu and Puerto Rico. Journal of Chronic Diseases 27:329-44, 1974. 113 Population Panel Report 159. 160. 161. 162. 163. 164. 165. 166. 167. 114 Anderson KM, Castelli WP, Levy D. Cho- lesterol and mortality: 30 years of follow- up from the Framingham Study. Journal of the American Medical Association 257:2176-80, 1987. Yano K, Blackwelder WC, Kagan A, et al. Childhood cultural experience and the inci- dence of CHD in Hawaii Japanese men. American Journal of Epidemiology 109:440-50, 1979. Kagan A, McGee DL, Yano K, et al. Serum cholesterol and mortality in a Japanese- American population. American Journal of Epidemiology 114:11-20, 1981. Leaverton PE, Sorlie PD, Kleinman JC, et al. Representativeness of the Framingham risk model for coronary heart disease mortality: A comparison with a national cohort study. Journal of Chronic Diseases 40:775-84, 1987. Kannel WB, Neaton JD, Wentworth D, et al. Overall and coronary heart disease mortal- ity rates in relation to major risk factors in 325,348 men screened for the MRFIT. Multiple Risk Factor Intervention Trial. American Heart Journal 112:825-36, 1986. Stamler J, Wentworth D, Neaton JD. Is the relationship between serum cholesterol and risk of premature death from coronary heart disease continuous and graded? Findings in 356,222 primary screenees of the Multiple Risk Factor Intervention Trial (MRFIT). Journal of the American Medical Association 256:2823-28, 1986. Martin MJ, Hulley SB, Browner WS, et al. Serum cholesterol, blood pressure, and mortality: Implications from a cohort of 361,662 men. Lancet 2:933-36, 1986. Holmes DR, Elveback LR, Frye RL, et al. Association of risk factor variables and coronary artery disease documented with angiography. Circulation 63:293-99, 1981. Schmidt SB, Wasserman AG, Muesing RA, et al. Lipoprotein and apolipoprotein levels in angiographically defined coronary 168. 169. 170. 171. 172. 173. 174. 175. 176. atherosclerosis. American Journal of Cardiol- ogy 55:1459-62, 1985. Hamsten A, Walldius G, Szamosi A, et al. Relationship of angiographically defined coronary artery disease to serum lipopro- teins and apolipoproteins in young survi- vors of myocardial infarction. Circulation 73:1097-1110, 1986. Solberg LA, Strong JP. Risk factors and ath- erosclerotic lesions. A review of autopsy studies. Arteriosclerosis 3:187-98, 1983. Newman WP III, Freedman DS, Voors AW, et al. Relation of serum lipoprotein levels and systolic blood pressure to early athero- sclerosis: The Bogalusa Heart Study. New England Journal of Medicine 314:138-44, 1986. Kahn HA, Medalie JH, Neufeld HN, et al. Serum cholesterol: Its distribution and as- sociation with dietary and other variables in a survey of 10,000 men. Israel Journal of Medical Sciences 5:1117-27, 1969. Keys A. Coronary heart disease—the global picture. Atherosclerosis 22:149-92, 1975. Lewis B, Chait A, Sigurdsson G. Serum lipoproteins in four European communities: A quantitative comparison. European Journal of Clinical Investigation 8:165-73, 1978. Halfon ST, Ritkind BM, Harlap S, et al. Plasma lipids and lipoproteins in adult Jews of different origin: The Jerusalem Lipid Research Clinic prevalence study. Israel Journal of Medical Sciences 18:1113-20, 1982. Simons LA. Interrelations of lipids and lipoproteins with coronary artery disease mortality in 19 countries. American Journal of Cardiology 57:5G-10G, 1986. Goldbourt U, Holtzman E, Neufeld HN. Total and high density lipoprotein choles- terol in the serum and risk of mortality: Evidence of a threshold effect. British Medical Journal 290:1239-43, 1985. Population Panel Report 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. Johnson KG, Yano K, Kato H. Coronary heart disease in Hiroshima, Japan: A report of a six-year period of surveillance, 1958- 1964. American Journal of Public Health 58:1355-67, 1968. Rose G, Shipley MJ]. Plasma lipids and mortality: A source of error. Lancet 1:523- 26, 1980. Peto R, Oxford University, personal com- munication, 1989. Keys A. Coronary heart disease in seven countries. Circulation 41(suppl 1):1-211, 1970. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press, 1980. Kagan A, Harris BR, Winkelstein W, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii, and California: Demo- graphic, physical, dietary and biochemical characteristics. Journal of Chronic Diseases 27:345-64, 1974. Toor M, Katchalsky A, Agmon J, et al. Ath- erosclerosis and related factors in immi- grants to Israel. Circulation 22:265-79, 1960. Kushi LH, Lew RA, Stare FJ, et al. Diet and 20-year mortality from coronary heart disease: The Ireland-Boston Diet-Heart Study. New England Journal of Medicine 312:811-18, 1985. Hulley SB, Rosenman RD, Brand R]. The association between triglycerides and coro- nary heart disease. New England Journal of Medicine 302:1383-89, 1980. Avins AL, Haber RJ, Hulley SB. The status of hypertriglyceridemia as a risk factor for coronary heart disease. In: Rifkind BM, Lippel K, eds. Clinics in Laboratory Medicine 9:153-68, 1989. 187. 188. 189. 190. 191. 192. 193. 194. 195, Gordon T, Fisher M, Ernst N, et al. Relation of diet to LDL cholesterol, VLDL choles- terol, and plasma total cholesterol and triglycerides in white adults: The Lipid Research Clinics Prevalence Study. Arterio- sclerosis 2:502-12, 1982. National Institutes of Health. Treatment of hypertriglyceridemia: Consensus confer- ence. Journal of the American Medical Asso- ciation 251:1196-1200, 1984. Carlson LA, Bottiger LE. Risk factors for ischaemic heart disease in men and women. Results of the 19-year follow-up of the Stockholm Prospective Study. Acta Medica Scandinavica 218:207-11, 1985. Aberg H, Lithell H, Selinus I, et al. Serum triglycerides are a risk factor for myocardial infarction but not for angina pectoris. Results from a 10-year followup of Uppsala Primary Preventive Study. Atherosclerosis 54:89-97, 1985. Knuiman JT, West CE. HDL-cholesterol in men from thirteen countries. Lancet 2:367- 68, 1981. Watkins LO, Neaton JD, Kuller LH. Racial differences in high-density lipoprotein cho- lesterol and coronary heart disease inci- dence in the usual-care group of the Mul- tiple Risk Factor Intervention Trial. Ameri- can Journal of Cardiology 57:538-45, 1986. Castelli WP, Abbott RD, McNamara PM. Summary estimates of cholesterol used to predict coronary heart diseases. Circulation 67:730-34, 1983. Gordon DJ, Probstfield JL, Garrison RJ, et al. High-density lipoprotein cholesterol and cardiovascular disease: Four prospec- tive American studies. Circulation 79:8- 15, 1989. Wallace RB, Anderson RA. Blood lipids, lipid-related measures, and the risk of ath- erosclerotic cardiovascular disease. Epi- demiology Review 9:95-119, 1987. 115 Population Panel Report 196. 197. 198. 199. 200. 201. 202. 203. 204. 116 Grundy SM, Goodman DS, Rifkind BM, et al. The place of HDL in cholesterol management. Archives of Internal Medicine 149:505-10, 1989. Moore MC, Guzman MA, Schilling PE, et al. Dietary-atherosclerosis study on de- ceased persons. Relation of selected dietary components to raised coronary lesions. Journal of the American Dietetic Association 68:216-23, 1976. National Heart and Lung Institute. Arterio- sclerosis: A Report by the National Heart and Lung Institute Task Force on Arteriosclerosis, Volume I. Bethesda, MD: U.S. Department of Health, Education, and Welfare, Public Health Service, National Institutes of Health, Pub. No. (NIH) 71-137, 1971. Rosenthal SR. Studies in atherosclerosis: Chemical, experimental and morphologic. Archives of Pathology 18:473-698, 1934. Malmros H. The relation of nutrition to health. Acta Medica Scandinavica 246(suppl):137-53, 1950. Keys A, Aravanis C, Blackburn H, et al. Se- rum cholesterol and cancer mortality in the Seven Countries Study. American Journal of Epidemiology 121:870-83, 1985. Kato H, Tillotson J, Nichaman MZ, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii, and California: Serum lipids and diet. American Journal of Epidemi- ology 97:372-85, 1973. Marmot MG, Syme SL, Kagan A, et al. Epi- demiologic studies of CHD and stroke in Japanese men living in Japan, Hawaii, and California: Prevalence of coronary and hy- pertensive heart disease and associated risk factors. American Journal of Epidemiology 102:514-25, 1975. Worth RM, Kato H, Rhoads GG, et al. Epi- demiologic studies of coronary heart dis- ease and stroke in Japanese men living in Japan, Hawaii, and California: Mortality. 203. 206. 207. 208. 209. 210. 211. 212. 213. American Journal of Epidemiology 102:481-90, 1975. Gordon T, Kagan A, Garcia-Palmieri M, et al. Diet and its relation to coronary heart disease and death in three populations. Circulation 63:500-15, 1981. Shekelle RB, Shryock AM, Paul O, et al. Diet, serum cholesterol, and death from coronary heart disease: The Western Elec- tric Study. New England Journal of Medicine 304:65-70, 1981. Shekelle RB, Stamler J. Dietary cholesterol and ischaemic heart disease. Lancet 1:1177- 79, 1989. McGee DL, Reed DM, Yano K, et al. Ten- year incidence of coronary heart disease in the Honolulu Heart Program: Relationship to nutrient intake. American Journal of Epi- demiology 119:667-76, 1984. McGee DL, Reed DM, Stemmerman G, et al. The relationship of dietary fat and choles- terol to mortality in 10 years: The Honolulu Heart Program. International Journal of Epi- demiology 14:97-105, 1985. Kromhout D, de Lezenne Coulander C. Diet, prevalence and 10-year mortality from coronary heart disease in 871 middle-aged men. The Zutphen Study. American Journal of Epidemiology 119:733-41, 1984. Toor M, Katchalsky A, Agmon J, et al. Serum-lipid and atherosclerosis among Yemenite immigrants in Israel. Lancet 1:1270-73, 1957. Prior I, Tasman-Jones C. New Zealand Maori and Pacific Polynesians. In: Trowell HC, Burkett DP, eds. Western Diseases: Their Emergence and Prevention. Cambridge, MA: Harvard University Press, 227, 1981. Jolliffe N, Archer M. Statistical associations between international coronary heart disease death rates and certain environ- mental factors. Journal of Chronic Diseases 9:636-52, 1959. Population Panel Report 214. 215. 216. 217. 218. 219. 220. 221. 222. 223, Beaglehole R, LaRosa JC, Heiss G, et al. Se- rum cholesterol, diet, and the decline in coronary heart disease mortality. Preventive Medicine 8:538-47, 1979. Walden RT, Schaefer LE, Lemon FR, et al. Effect of environment on the serum choles- terol-triglyceride distribution among Sev- enth-day Adventists. American Journal of Medicine 36:269-76, 1964. Phillips RL, Lemon FR, Beeson WL, et al. Coronary heart disease mortality among Seventh-Day Adventists with differing dietary habits: A preliminary report. American Journal of Clinical Nutrition 31(suppl 10):5191-98, 1978. Thorogood M, Carter R, Benfield L, et al. Plasma lipids and lipoprotein cholesterol concentrations in people with different diets in Britain. British Medical Journal 295:351-53, 1987. Fraser GE. Determinants of ischemic heart disease in Seventh-day Adventists: A review. American Journal of Clinical Nutrition 48:833-36, 1988. Dwyer JT. Health aspects of vegetarian diets. American Journal of Clinical Nutrition 48:712-38, 1988. Katan MB, Beynan AC, de Vries JH, et al. Existence of consistent hypo- and hyperre- sponders to dietary cholesterol in man. American Journal of Epidemiology 123:221-34, 1986. Stamler J, Shekelle RB. Dietary cholesterol and human coronary heart disease: The epidemiologic evidence. Archives of Pathol- ogy and Laboratory Medicine 112:1032-40, 1988. Paul O, Lepper MH, Phelan WH, et al. A longitudinal study of coronary heart disease. Circulation 28:20-31, 1963. Chapman JM, Coulson AH, Clark VA, et al. The differential effect of serum cholesterol, blood pressure and weight on the incidence 224. 225. 226. 227. 228. 229. 230. of myocardial infarction and angina pectoris. Journal of Chronic Diseases 23:631- 45, 1971. Rabkin SW, Mathewson FA, Hsu PH. Rela- tion of body weight to development of is- chemic heart disease in a cohort of young North American men after a 26 year obser- vation period: The Manitoba Study. Ameri- can Journal of Cardiology 39:452-58, 1977. Robertson TL, Kato H, Gordon T, et al. Epi- demiologic studies of coronary heart dis- ease and stroke in Japanese men living in Japan, Hawaii and California: Coronary heart disease risk factors in Japan and Hawaii. American Journal of Cardiology 39:244-49, 1977. Stokes J III, Garrison RJ, Kannel WB. The independent contributions of various indices of obesity to the 22-year incidence of coronary heart disease: The Framingham Heart Study. In: Vague J, Bjorntorp P, Guy-Grand B, et al., eds. Metabolic Compli- cations of Human Obesities. New York: Elsevier, 49-57, 1985. Hubert HB, Feinleib M, McNamara PM, et al. Obesity as an independent risk factor for cardiovascular disease: A 26-year follow-up of participants in the Framing- ham Heart Study. Circulation 67:968-77, 1983. Sturdevant RA, Pearce ML, Dayton S. In- creased prevalence of cholelithiasis in men ingesting a serum-cholesterol-lowering diet. New England Journal of Medicine 288:24-27, 1973. Howe GR. The use of polytomous dual re- sponse data to increase power in case- control studies: An application to the association between dietary fat and breast cancer. Journal of Chronic Diseases 38:663-70, 1985. Ip C. Fat and essential fatty acid in mam- mary carcinogenesis. American Journal of Clinical Nutrition 45:218-24, 1987. 117 Population Panel Report 231. 232. 233. 234. 235. 236. 237. 238. 239, 240. 118 Erickson KL. Mechanisms of dietary fat modulation of tumorigenesis: Changes in immune response. Progress in Clinical and Biological Research 222:555-86, 1986. Nikkari T, Salo M, Maatela J, et al. Serum fatty acids in Finnish men. Atherosclerosis 49:139-48, 1983. Riemersma RA, Wood DA, Butler S, et al. Linoleic acid content in adipose tissue and coronary heart disease. British Medical Jour- nal 292:1423-27, 1986. Wood DA, Riemersma RA, Butler S, et al. Linoleic and eicosapentaenoic acids in adi- pose tissue and platelets and risk of coro- nary heart disease. Lancet 1:177-83, 1987. Abraham R, Riemersma RA, Wood D, et al. Adipose fatty acid composition and the risk of serious ventricular arrhythmias in acute myocardial infarction. American Journal of Cardiology 63:269-72, 1989. Bang HO, Dyerberg J. Lipid metabolism and ischemic heart disease in Greenland Eskimos. In: Draper HH, ed. Advances in Nutritional Research, Volume 3. New York: Plenum Press, 1-22, 1980. Kromhout D, Bosschieter EB, de Lezenne Coulander C. The inverse relation between fish consumption and 20-year mortality from coronary heart disease. New England Journal of Medicine 312:1205-09, 1985. Shekelle RB, Missell L, Paul O, et al. Fish consumption and mortality from coronary heart disease. New England Journal of Medi- cine 313:820, 1985. Katan MB. Diet and HDL. In: Miller NE, Miller GJ, eds. Metabolic Aspects of Cardio- vascular Disease, Volume 3. Clinical and Metabolic Aspect of HDL. Oxford: Elsevier, 103-31, 1984. Garcia-Palmieri MR, Sorlie P, Tillotson J, et al. Relationship of dietary intake to subse- quent coronary heart disease incidence: 241. 242. 243. 244. 245. 246. 247. 248. The Puerto Rico Heart Health Program. American Journal of Clinical Nutrition 33:1818-27, 1980. Carroll KK. Experimental studies on dietary fat and cancer in relation to epi- demiological data. In: Ip C, Birt DF, Rogers AE, et al., eds. Progress in Clinical and Bio- logical Research, Volume 222: Dietary Fat and Cancer. New York: Alan R. Liss, Inc., 231- 48, 1986. Knuiman JT, Hermus RJ], et al. Serum total and high-density lipoprotein cholesterol concentrations in rural and urban boys from 16 countries. Atherosclerosis 36:529-37, 1980. National Heart, Lung, and Blood Institute. The Lipid Research Clinics Population Studies Data Book. Volume 1: The Prevalence Study. Bethesda, MD: U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, NIH Pub. No. 80-1527, July 1980. Laskarzewski P, Morrison JA, deGroot I, et al. Lipid and lipoprotein tracking in 108 children over a four-year period. Pediatrics 64:584-91, 1979. Orchard TJ, Donahue RP, Kuller LH, et al. Cholesterol screening in childhood: Does it predict adult hypercholesterolemia? The Beaver County experience. Journal of Pediat- rics 103:687-91, 1983. Lauer RM, Lee J, Clarke WR. Factors affecting the relationship between child- hood and adult cholesterol levels: The Muscatine Study. Pediatrics 82:309-18, 1988. Berenson GS, Epstein FH. Conference on blood lipids in children: Optimal levels for early prevention of coronary heart disease. Preventive Medicine 12:741-97, 1983. Clarke WR, Woolson RF, Lauer RM. Changes in ponderosity and blood pressure in childhood: The Muscatine Study. Ameri- can Journal of Epidemiology 124:195-206, 1986. Population Panel Report 249. 250. 251. 252. 253. 254. 253. 256. 257. Smoak CG, Burke GL, Webber LS, et al. Re- lation of obesity to clustering of cardiovas- cular disease risk factors in children and young adults: The Bogalusa Heart Study. American Journal of Epidemiology 125:364-72, 1987. Johnson AL, Cornoni JC, Cassels JC, et al. Influence of race, sex and weight on blood pressure behavior in young adults. Ameri- can Journal of Cardiology 35:523-30, 1975. Frerichs RR, Webber LS, Srinivasan SR, et al. Relationship of serum lipids and lipo- proteins to obesity and sexual maturation in white and black children. American Journal of Epidemiology 108:486-96, 1978. Goldstein JL, Brown MS. Familial hyper- cholesterolemia. In: Stanbury JB, Wyngaarden |B, Fredrickson DS, et al., eds. The Metabolic Basis of Inherited Disease. New York: McGraw-Hill Book Co., 672-712, 1983. Schrott HG, Clarke WR, Wiebe DA, et al. Increased coronary mortality in relatives of hypercholesterolemic school children: The Muscatine Study. Circulation 59:320-26, 1979. Moll PP, et al. Total cholesterol and lipo- proteins in school children: Prediction of coronary heart disease in adult relatives. Circulation 67:127-34, 1983. Lee J, Lauer RM, Clarke WR. Lipoproteins in the progeny of young men with coronary artery disease: Children with increased risk. Pediatrics 78:330-37, 1986. Morrison JA, Kelly KA, Mellies MJ, et al. Parent-child associations at upper and lower ranges of plasma cholesterol and triglyceride levels. Pediatrics 62:468-77, 1978. Strong JP, McGill HC Jr. Pediatric aspects of atherosclerosis. Journal of Atherosclerosis Research 9:251-65, 1969. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. McGill HC Jr. The George Lyman Duff Memorial Lecture: Persistent problems in the pathogenesis of atherosclerosis. Arterio- sclerosis 4:443-51, 1984. Ross R, Glomset JA: The pathogenesis of atherosclerosis. New England Journal of Medicine 295:369-77, 1976. Enos WF, Holmes RH, Beyer J. Coronary artery disease among United States soldiers killed in action in Korea. Journal of the American Medical Association 152:1090-93, 1953. McNamara JJ, Molst MA, Stremple JF, et al. Coronary artery disease on combat casual- ties in Vietnam. Journal of the American Medical Association 216:1185-87, 1971. Boulton TJC. Patterns of food intake in childhood and adolescence and risk of later disease. Australian and New Zealand Journal of Medicine 15:478-88, 1985. Darrow DC. The significance of body size. AMA Journal of Diseases of Children 98:416- 23,1939. Finberg L, Kravath RE, Fleischman AR. Water and Electrolytes in Pediatrics Physiol- ogy, Pathophysiology, and Treatment. Phila- delphia: W.B. Saunders and Co., 17-22, 1982. National Research Council, Commission on Life Sciences, Food and Nutrition Board. Recommended Dietary Allowances, 10th ed. Washington, DC: National Academy Press, 1989. Fomon SJ, Filer L] Jr, Ziegler EE, et al. Skim milk in infant feeding. Acta Paediatrica Scandinavica 66:17-30, 1987. Fomon SJ, Thomas LN, Filer L] Jr, et al. In- fluences of fat and carbohydrate content of diet on food intake and growth of male infants. Acta Paediatrica Scandinavica 35:136- 44, 1976. 119 Population Panel Report 268. 269, 270. 271. 272. 273. 274. 120 Abraham S, Carroll MD, Johnson CL, et al. Caloric and selected nutrient values for per- sons 1-74 years of age: First Health and Nutrition Examination Survey, United States, 1971-1974. Vital and Health Statistics, Series 11, No. 209. Hyattsville, MD: U.S. Department of Health, Education, and Wel- fare, Public Health Service, National Center for Health Statistics, Pub. No. (PHS) 79- 1657, 1979. Carroll MD, Abraham S, Dresser CM. Die- tary intake source data: United States 1976- 1980. Vital and Health Statistics, Series 11, No. 231. Hyattsville, MD: U.S. Department of Health and Human Services, Public Health Service, National Center for Health Statistics, Pub. No. (PHS) 83-1681, 1983. Dresser C, Carroll M. Major contributing sources of macro- and micronutrients for persons ages 1-74 years by age and race: United States, 1971-1980. U.S. Department of Health and Human Services, Public Health Service, National Center for Health Statistics (in preparation). Martinez GA, Ryan AS. What are children in the United States eating? In: Prevention of Adult Atherosclerosis During Childhood, Re- port of the 95th Ross Conference on Pediatric Research. Columbus, OH: Ross Laborato- ries, 53-61, 1988. American Academy of Pediatrics Commit- tee on Nutrition. Prudent lifestyle for children: Dietary fat and cholesterol. Pedi- atrics 78:521-25, 1986. American Academy of Pediatrics Commit- tee on Nutrition. Indications for cholesterol testing in children. Pediatrics 83:141-42, 1989. Weidman W, Kwiterovich P, Jesse MJ, et al. Diet in the healthy child: Report of the Task Force Committee on Nutrition and the Cardiovascular Disease in the Young Council of the American Heart Association. Circulation 67:1411A-14A, 1983. 275. 276. 277. 278. 279. 280. 281. 282. 283, 284. Consensus Conference. Lowering blood cholesterol to prevent heart disease. Journal of the American Medical Association 253:2080- 86, 1985. Mauer AM. Prudent diet and lifestyle rec- ommendations of the American Academy of Pediatrics. In: Prevention of Adult Athero- sclerosis During Childhood: Report of the 95th Ross Conference on Pediatric Research. Co- lumbus, OH: Ross Laboratories, 83-88, 1988. LaRosa J, Finberg L. Preliminary report from a conference entitled “Prevention of adult atherosclerosis during childhood.” Journal of Pediatrics 112:317-8, 1988. Srikantia SG. Use of fats and oils in child feeding. Food and Nutrition (Roma) 5:18-22, 1979. Pugliese MT, Weyman-Daum M, Moses N, et al. Parental health beliefs as a cause of non-organic failure to thrive. Pediatrics 80:175-82, 1987. Kritchevsky D. How aging affects choles- terol metabolism. Postgraduate Medicine 63:133-38, 1978. Luepker RV. Coronary heart disease risk factor reduction in the elderly. Quality of Life and Cardiovascular Care 2:159-75, 1986. National Center for Health Statistics. Vital Statistics of the United States. Hyattsville, MD: U.S. Government Printing Office, 1985. Rastam L, Luepker RV, Mittelmark MB, et al. Population screening and referral for hypercholesterolemia. American Journal of Preventive Medicine 4:249-54, 1988. Chapman JM, Massey FJ]. The interrelation- ship of serum cholesterol, hypertension, body weight, and risk of coronary disease: Results of the first ten years’ follow-up in the Los Angeles Heart Study. Journal of Chronic Diseases 17:933-49, 1964. Population Panel Report 285. 286. 287. 288. 289. 290. 291. 292. 290. 294. Gofman JW, Young W, Tandy R. Ischemic heart disease, atherosclerosis, and longevity. Circulation 34:679-97, 1966. Agner E. Epidemiology of coronary heart disease in the elderly patient. In: Coodley EL, ed. Geriatric Heart Disease. Littleton, MA: PSG Publications, 114-126, 1985. Kannel WB, Doyle JT, Shephard RJ, et al. Prevention of cardiovascular disease in the elderly. Journal of the American College of Cardiology 10:25A-28A, 1987. Kittner SJ, Garcia-Palmieri MR, Costas R Jr, et al. Alcohol and coronary heart disease in Puerto Rico. American Journal of Epidemiol- ogy 117:538-50, 1983. Barrett-Connor E, Suarez L, Khaw KT, et al. Ischemic heart disease risk factors after age 50. Journal of Chronic Diseases 37:903-08, 1984. Bush T, Knoke J, Criqui M, et al. Total and LDL cholesterol as predictors of CHD death in elderly men: The Lipid Research Clinics’ Program Follow-up Study. CVD Epidemiol- ogy Newsletter 43:22, 1988. Benfante R, Reed D. Is elevated serum cho- lesterol level a risk factor for coronary heart disease in the elderly? Journal of the Ameri- can Medical Association 263:393-96, 1990. Gordon DJ, Rifkind BM. Treating high blood cholesterol in the older patient. American Journal of Cardiology 63:48H-52H, 1989. Garber AM, Littenberg B, Sox HC Jr.; and Gluck ME, Wagner JL, Duffy BM. Costs and Effectiveness of Cholesterol Screening in the Elderly. Washington, DC: Health Program, Office of Technology Assessment, Congress of the United States, GPO No. 052-003- 01151-4, April 1989. Albanese A, Woodhull ML, Lorenze EJ, et al. Effect of diet on blood cholesterol of the elderly. Geriatrics 22:133-44, 1967. 295. 296. 297. 298. 299. 300. 301. 302. Dayton S, Pearce ML, Hashimoto S, et al. A controlled clinical trial of a diet high in un- saturated fat in preventing complications of atherosclerosis. Circulation XL(1)(suppl ID):111-1163, 1969. Schneider EL, Vining EM, Hadley EC, et al. Recommended dietary allowances and the health of the elderly. New England Journal of Medicine 314:157-60, 1986. Goodwin JS, Hunt WC, Hooper P, et al. Re- lationship between zinc intake, physical ac- tivity, and blood levels of high-density lipo- protein: Cholesterol in a healthy elderly population. Metabolism 34:519-23, 1985. Levy RI. Declining mortality in coronary heart disease. Arteriosclerosis 1:312-25, 1981. Kannel WB, Abbott RD. Incidence and prognosis of myocardial infarction in women: The Framingham study. In: Eaker ED, Packard B, Wenger NK, et al., eds. Coronary Heart Disease in Women: Proceed- ings of an NIH Workshop. New York: Hay- market Doyma, 208-14, 1987. Eaker ED, Packard B, Wenger NK, et al., eds. Coronary Heart Disease in Women: Pro- ceedings of an NIH Workshop. New York: Haymarket Doyma, Inc., 1987. Dustan HP. Nutrition and cardiovascular diseases of women. Public Health Reports 102(suppl):22-25, 1987. Eaker ED, Castelli WP. Coronary heart dis- ease and its risk factors among women in the Framingham study. In: Eaker ED, Packard B, Wenger NK, et al., eds. Coronary Heart Disease in Women: Proceedings of an NIH Workshop. New York: Haymarket Doyma, Inc., 122-30, 1987. . Gosland BA, Wynn V, Crook D, et al. Sex, lipoproteins, and atherosclerosis: Prevail- ing assumptions and outstanding questions. American Heart Journal 114:1467- 1503, 1987. 121 Population Panel Report 304. 303. 306. 307. 308. 309. 310. 311. 312. 313. 122 Brownell KD, Bacherik PS, Ayerle RS. Changes in plasma lipid and lipoprotein levels in men and women after ten weeks of exercise. Metabolism 31:1142-46, 1982. Lennon DL, Stratman FW, Shrago E, et al. Total cholesterol and HDL-cholesterol changes during acute, moderate intensity exercise in men and women. Metabolism 32:244-49, 1983. Frey MA, Doerr BM, Laubach LL, et al. Ex- ercise does not change high-density lipo- protein cholesterol in women after ten weeks of training. Metabolism 31:1142-46, 1982. Department of Health and Human Services. Report of the Secretary's Task Force on Black and Minority Health. Volume I: Executive Summary. Washington, DC: U.S. Govern- ment Printing Office, 1985. Antonovsky A. Social class and the major cardiovascular diseases. Journal of Chronic Diseases 21:65-106, 1968. Shekelle RB, Ostfeld A, Paul O. Social status and incidence of coronary disease. Journal of Chronic Diseases 22:381-94, 1969. Marmot M, Adelstein A, Robinson N, et al. Changing social class distribution of heart disease. British Medical Journal 2:1109-12, 1978. Kraus J, Borhani N, Franti C. Socio- economic status, ethnicity, and risk of coronary heart disease. American Journal of Epidemiology 111:407-14, 1980. Rose G, Marmot M. Social class and coro- nary heart disease. British Heart Journal 45:13-19, 1981. Salonen J. Socioeconomic status and risk of cancer, cerebral stroke, and death due to coronary heart disease: A longitudinal study in Eastern Finland. Journal of Epi- demiology and Community Health 36:294-97, 1982. 314. 315. 316. 317. 319. 320. 321. 322. Marmot M, Kogevinas M, Elston M. So- cial /economic status and disease. Annual Review of Public Health 8:111-35, 1987. Wing S, Dargent-Molina P, Casper M, et al. Changing association between community occupational structure and ischemic heart disease mortality in the United States. Lancet 1:1067-70, 1987. Holme I, Hjermann I, Helgeland A, et al. Coronary risk factors and socioeconomic status. Lancet 2:1396-98, 1976. Pocock SJ, Shaper AG, Cook DG, et al. Social class differences in ischaemic heart disease in British men. Lancet 2:197-201, 1987. . Holme I, Helgeland A, Hjermann I, et al. Physical activity at work and leisure in rela- tion to coronary risk factors and social class. Acta Medica Scandinavica 209:277-83, 1981. Khoury P, Morrison J, Laskarzewski P, et al. Relationships of education and occupation to coronary heart disease risk factors in schoolchildren and adults: The Princeton School District Study. American Journal of Epidemiology 113:378-95, 1981. Fulwood R, Kalsbeek W, Rifkind B, et al. Total serum cholesterol levels of adults 20- 74 years of age in the United States, 1976- 80. Vital Health Statistics, Series 11, No. 236. Hyattsville, MD: U.S. Department of Health and Human Services, Public Health Service, National Center for Health Statis- tics, Pub. No. (PHS) 86-1686, May 1986. Fears R. Pharmacological Control of Cho- lesterol 7-Hydroxylase Activity. In: Fears R, Sabine JR, Eds. Cholesterol 7-Hydroxylase (7-monooxygenase). Boca Raton, FL: CRC Press, 115-32, 1986. Lipid Research Clinics Program. The Lipid Research Clinics Coronary Primary Preven- tion Trial results: I. Reduction in incidence of coronary heart disease. Journal of the American Medical Association 251:351-64, 1984. Population Panel Report 323. 324. 325. 326. 327. 328. 329. Lipid Research Clinics Program. The Lipid Research Clinics Coronary Primary Preven- tion Trial results: II. The relationship of re- duction in incidence of coronary heart dis- ease to cholesterol-lowering. Journal of the American Medical Association 251:365-74, 1984. Gordon DJ, Knoke J, Probstfield JL, et al. High-density lipoprotein cholesterol and coronary heart disease in hypercholesterol- emic men: The Lipid Research Clinics Coronary Primary Prevention Trial. Circu- lation 74:1217-25, 1986. Frick MH, Elo O, Haapa K, et al. Helsinki Heart Study: Primary-prevention trial with gemfibrozil in middle-aged men with dyslipidemia. Safety of treatment, changes in risk factors, and incidence of coronary heart disease. New England Journal of Medi- cine 317:1237-45, 1987. Manninen V, Elo MO, Frick MH, et al. Lipid alterations and decline in the inci- dence of coronary heart disease in the Helsinki Heart Study. Journal of the Ameri- can Medical Association 260:641-51, 1988. Brensike JF, Levy RI, Kelsey SF, et al. Effects of therapy with cholestyramine on progression of coronary arteriosclerosis: Results of the NHLBI Type II Coronary Intervention Study. Circulation 69:313-24, 1984. Blankenhorn DH, Nessim SA, Johnson RL, et al. Beneficial effects of combined coles- tipol-niacin therapy on coronary athero- sclerosis and coronary venous bypass grafts. Journal of the American Medical Association 257:3233-40, 1987. Brown BG, Lin JT, Schaefer M, et al. Niacin and lovastatin combined with colestipol re- gress coronary atherosclerosis and prevent clinical events in men with elevated apo- lipoprotein B. Circulation 80(suppl 11):266, 1989. 330. 331. 332. 333. 334. 335. 336. 337. Hjermann I, Velve Byre K, Holme I, Leren P. Effect of diet and smoking intervention on the incidence of coronary heart disease. Lancet 2:1303-10, 1981. MRFIT Research Group. Multiple Risk Fac- tor Intervention Trial. Risk factor changes and mortality results. Journal of the Ameri- can Medical Association 248:1465-77, 1982. Cutler JA, Neaton JD, Hulley SB, et al. Coronary heart disease and all-causes mortality in the Multiple Risk Factor Inter- vention Trial: Subgroup findings and com- parisons with other trials. Preventive Medicine 14:293-311, 1985. Tyroler HA. Review of lipid-lowering clini- cal trials in relation to observational and epidemiologic studies. Circulation 76:515- 22,1987. Yusuf S, Wittes J, Friedman L. Overview of results of randomized clinical trials in heart failure. II. Unstable angina, heart failure, primary prevention with aspirin and risk factor modification. Journal of the American Medical Association 260:2259-63, 1988. Oliver MF, Heady JA, Morris JN, et al. WHO cooperative trial on primary preven- tion of ischaemic heart disease using clofi- brate to lower serum cholesterol: Mortality follow-up. Report of the Committee of Principal Investigators. Lancet 2:379-85, 1980. Coronary Drug Project Research Group. The Coronary Drug Project: Clofibrate and niacin in coronary heart disease. Journal of the American Medical Association 231:360-81, 1975. Canner PL, Berge KG, Wenger NK, et al. Fifteen year mortality in coronary drug project patients: Long-term benefit with niacin. Journal of the American College of Car- diology 8:1245-55, 1986. 123 Population Panel Report 338. 339. 340. 341. 342. 343. 344. 345. 346. 347. 124 Carlson LA, Rosenhamer G. Reduction of mortality in the Stockholm Ischaemic Heart Disease Secondary Prevention Study by combined treatment with clofibrate and nicotinic acid. Acta Medica Scandinavica 223:405-18, 1988. Hjermann I, Holme I, Leren P. Oslo Study Diet and Antismoking Trial: Results after 102 months. American Journal of Medicine 80:7-11, 1986. Arntzenius AC, Kromhout D, Barth JD, et al. Diet, lipoproteins, and the progression of coronary atherosclerosis. The Leiden Intervention Trial. New England Journal of Medicine 312:805-11, 1985. Keys A. Coronary heart disease, serum cholesterol, and the diet. Acta Medica Scandinavica 207:153-60, 1980. Stamler J. Population studies. In: Levy RI, Rifkind BM, Dennis BH, Ernst N, eds. Nu- trition, Lipids and Coronary Heart Disease. New York: Raven Press, 25-88, 1979. Puska P, Tuomilehto J, Nissinen A, et al. The 10 year results of the North Karelia Project. CVD Epidemiology Newsletter 37:88, 1985. Puska P, Nissinen A, Tuomilehto J, et al. Communication based strategy to prevent coronary heart disease: Conclusions from 10 years of North Karelia. Annual Review of Public Health 6:147-93, 1985. Pietinen P, Nissinen A, Vartianinen E, et al. Dietary changes in the North Karelia Project (1972-1982). Preventive Medicine 17:183-93, 1988. Farquhar JW, Maccoby N, Wood PD, et al. Community education for cardiovascular health. Lancet 1:1192-95, 1977. Gutzwiller F, Junod B, Epstein FH, et al. First results of a cardiovascular community control programme in Switzerland: The National Research Programme IA “Primary 348. 349. 350. 351. 352. 353. 354. 355. prevention of cardiovascular diseases.” Heart Beat, Journal of the International Society and Federation of Cardiology, Spring 1988. Abramson JH, Gofin R, Hopp C, et al. Evaluation of a community program for the control of cardiovascular risk factors: The CHAD program in Jerusalem. Israel Journal of Medical Sciences 17:201-12, 1981. Rywik S, Sznajd |, Kiec E, et al. The changes in risk factors levels after two years of implementation of the coronary diseases prevention in Poland. I. The dynamics of changes in total population. Przeglad Lekarski 38:769, 1981. Farquhar JW, Fortmann SP, Maccoby N, et al. The Stanford Five City Project: Design and methods. American Journal of Epidemiol- ogy 63:171-82, 1985. Farquhar JW, Fortmann SP, Flora JA, et al. The Stanford Five City Project: Results after 51/3 years of education. Paper presented at the 28th Conference on Car- diovascular Disease Epidemiology, Santa Fe, NM, March 1987. Murray DM, Luepker RV, Pirie PL, et al. Systematic risk factor screening and educa- tion: A community-wide approach to pre- vention of coronary heart disease. Preven- tive Medicine 15:661-72, 1986. Lefebvre RC, Peterson GS, McGraw SA, et al. Community intervention to lower blood cholesterol: The “Know Your Cholesterol” campaign in Pawtucket, Rhode Island. Health Education Quarterly 13:117-29, 1986. Gotto A, Farmer JA. Risk factors for coro- nary artery disease. In: Braunwald E, ed. Heart Disease: A Text Book of Cardiovascular Medicine. Philadelphia, PA: W.B. Saunders Co., 1153-90, 1988. Keys A, Taylor HL, Blackburn H, et al. Mortality and coronary heart disease among men studied for 23 years. Archives of Internal Medicine 28:201-14, 1971. Population Panel Report 356. 357. 358. 359. 360. 361. 362. 363. 364. 365. Blackburn H. The public health view of diet and mass hyperlipidemia. Cardiovascu- lar Diseases 1:361-69, 1980. Dawber T. The Framingham Study: The Epi- demiology of Atherosclerotic Disease. Cam- bridge, MA: Harvard University Press, 1980. Paffenbarger R, Laughlin M, Hyde R. Physical activity as an index of heart attack risk in college alumni. American Journal of Epidemiology 108:161-75, 1978. Kannel WB, McGee D. Diabetes and car- diovascular disease: The Framingham Study. Journal of the American Medical Association 241:2035-38, 1979. Morris J, Pollard R, Everitt M, et al. Vigor- ous exercise in leisure time: Protection against coronary heart disease. Lancet 2:1207-10, 1980. Joint National Committee. The 1988 report of the Joint National Committee on detec- tion, evaluation, and treatment of high blood pressure. Archives of Internal Medicine 148:1023-38, 1988. Weidman P, Uehlinger D, Gerber A. Edito- rial review: Antihypertensive treatment and serum lipoproteins. Journal of Hyperten- sion 3:297-306, 1985. Weinberger M. Antihypertensive therapy and lipids: Paradoxical influence on cardio- vascular disease risk. American Journal of Medicine 80(suppl 2A):64-70, 1986. Castelli WP, Garrison RJ, Dawber TR, et al. The filter cigarette and coronary heart dis- ease: The Framingham Study. Lancet 2:109- 13, 1981. Criqui M, Wallace R, Heiss G, et al. Ciga- rette smoking and plasma high-density lipoprotein cholesterol: The Lipid Research Clinics Program Prevalence Study. Circula- tion 62(suppl 1V):70-76, 1980. 366. 367. 368. 369. 370. 371. 372. 373. 374. Fortmann SP, Haskell WL, Williams PT. Changes in plasma high density lipoprotein cholesterol after changes in cigarette use. American Journal of Epidemiology 124:706-10, 1986. Gordon T, Kannel WB, McGee D, et al. Death and coronary attacks in men after giving up cigarette smoking: A report from the Framingham Study. Lancet 1:1345-48, 1974. Abbott R, Yin Y, Reed D, et al. Risk of stroke in male cigarette smokers. New England Journal of Medicine 315:717-20, 1986. Ekelund L-G, Haskell WL, Johnson JL, et al. Physical fitness as a predictor of cardiovas- cular mortality in asymptomatic North American men. The Lipid Research Clinics Mortality Follow-up Study. New England Journal of Medicine 319:1379-84, 1988. Haskell W, Taylor H, Wood P, et al. Strenu ous physical activity, treadmill exercise test performance and plasma high-density lipo- protein cholesterol: The Lipid Research Clinics Program Prevalence Study. Circula- tion 62(suppl IV):53-61, 1980. Hickey N, Mulcahy R, Bourke G, et al. Study of coronary risk factors related to physical activity in 15,171 men. British Medical Journal 3:507-09, 1975. Borkin G, Sparrow D, Wisniewski C, et al. Body weight and coronary disease risk: Patterns of risk factor change associated with long-term weight change. American Journal of Epidemiology 124:410-19, 1986. Millar WJ], Stephens T. The prevalence of overweight and obesity in Britain, Canada, and the United States. American Journal of Public Health 77:38-41, 1987. Society of Actuaries and Association of Life Insurance Medical Directors of America. Build Study 1979. Chicago: Society of Actuaries and Association of Life Insurance Medical Directors of America, 1980. 125 Population Panel Report 375. 376. 377. 378. 379. 381. 382. 383. 384. 385. 126 Baumgartner RN, Roche AF, Chumlea C, et al. Fatness and fat patterns: Associations with plasma lipids and blood pressures in adults, 18 to 57 years of age. American Jour- nal of Epidemiology 126:614-28, 1987. Higgins M, Kannel WB, Garrison RJ, et al. Hazards of obesity: The Framingham experience. Acta Medica Scandinavica 723(suppl):23-36, 1987. American Heart Association. Dietary Guide- lines for Healthy American Adults. Dallas: American Heart Association, 1986. Abbott R, Donahue R, MacMahon S, et al. Diabetes and the risk of stroke: The Hon- olulu Heart Program. Journal of the Ameri- can Medical Association 257:949-52, 1987. Rosenstock J, Strowig S, Cercone S, et al. Reduction in cardiovascular disease risk factors with intensive diabetes treatment in insulin-dependent diabetes mellitus. Diabe- tes Care 10:729-34, 1987. . Wingard DL, Barrett-Connor E, Criqui MH, et al. Clustering of heart disease risk factors in diabetic compared to nondiabetic adults. American Journal of Epidemiology 117:19-26, 1983. Kissebah A, Schectman G. Polyunsatu- rated fat, cholesterol, and fatty acid supplementation. Diabetes Care 11:129-42, 1988. Vinik A, Jenkins D. Dietary fiber in man- agement of diabetes. Diabetes Care 11:160- 73, 1933, Pearce ML, Dayton S. Incidence of cancer in men on a diet high in polyunsaturated fat. Lancet 1:464-67, 1971. Ederer F, Leren P, Turpeinen O, et al. Cancer among men on cholesterol lowering diets. Lancet 2:203-06, 1971. Dales LG, Friedman GD, Ury HK, et al. A case-control study of relationships of diet and other traits to colorectal cancer in 386. 387. 388. 389. 390. 391, 392. 393. 394. American blacks. American Journal of Epi- demiology 109:132-44, 1979. Miller AB. Nutrition and cancer. Preventive Medicine 9:189-96, 1980. McMichael AJ, Jensen OM, Parkin DM, et al. Dietary and endogenous cholesterol and human cancer. Epidemiology Review 6:192- 216, 1984. Yaari S, Goldbourt U, Evan-Zohar S, et al. Associations of serum high density lipopro- tein and total cholesterol with total, cardio- vascular, and cancer mortality in a 7-year prospective study of 10,000 men. Lancet 1:1011-14, 1981. Schatzkin A, Hoover RN, Taylor PR, et al. Serum cholesterol and cancer in the NHANES I epidemiologic followup study: National Health and Nutrition Examination Survey. Lancet 2:298-301, 1987. Schatzkin A, Hoover RN, Taylor PR, et al. Site-specific analysis of total serum choles- terol and incident cancer in the National Health and Nutrition Examination Survey I. Epidemiologic follow-up study. Cancer Re- search 48:452-58, 1988. Wingard DL, Criqui MH, Holdbook M]J, et al. Plasma cholesterol and cancer morbid- ity and mortality in an adult community. Journal of Chronic Diseases 37:401-06, 1984. Hiatt RA, Fireman BH. Serum cholesterol and the incidence of cancer in a large cohort. Journal of Chronic Diseases 29:861-70, 1986. Mannes GA, Maier A, Thieme C, et al. Re- lation between the frequency of colorectal adenoma and the serum cholesterol level. New England Journal of Medicine 315:1634-38, 1986. Tornberg SH, Holm LE, Carstensen JM, et al. Risks of cancer of the colon and rectum in relation to serum cholesterol and beta lipoprotein. New England Journal of Medicine 315:1629-33, 1986. Population Panel Report 393. 398. 399. 400. 401. 402. 403. 404. Salmond CE, Beaglehole R, Prior IA. Are low cholesterol values associated with excess mortality? British Medical Journal 290:422-24, 1985. . Ostfeld AM. A review of stroke epidemiol- ogy. Epidemiology Review 2:136-52, 1980. . Takeya Y, Popper JS, Shimizu Y, et al. Epi- demiologic studies of coronary heart dis- ease and stroke in Japanese men living in Japan, Hawaii and California: Incidence of stroke in Japan and Hawaii. Stroke 15:15-23, 1984. Tanaka H, Ueda Y, Hayashi M, et al. Risk factors for cerebral hemorrhage and cere- bral infarction in a Japanese rural community. Stroke 13:62-73, 1982. Kagan A, Popper JS, Rhoads GG, et al. Die- tary and other risk factors for stroke in Ha- waiian Japanese men. Stroke 16:390-96, 1986. Iso H, Jacobs D, Wentworth D, et al. Serum cholesterol levels and six-year mortality from stroke in 350,977 men screened for the Multiple Risk Factor Intervention Trial. New England Journal of Medicine 320:904-10, 1989. Department of Health and Human Serv- ices/U.S. Department of Agriculture. Nutrition Monitoring in the United States— An Update Report on Nutrition Monitoring. DHHS Pub. No. (PHS) 89-1255, 1989. Stephenson MG, Levy AS, Sass NL, et al. NHIS findings: Nutrition knowledge and baseline data for weight-loss objectives. Public Health Reports 102:61-67, 1987. Park YK, Yetley EA. Trend changes in use and current intakes of tropical oils in the United States. American Journal of Clinical Nutrition 51:738-48, 1990. Harris S, Welsh S. How well are our food choices meeting our nutrition needs? Nutri- tion Today 24:20-28, 1989. 405. 406. 407. 408. 409. 410. 411. 412. 413. 414. Dougherty RM, Fong AKH, lacono JM. Nutrient content of the diet when the fat is reduced. American Journal of Clinical Nutri- tion 48:970-79, 1988. American Heart Association. Dietary Treat- ment of Hypercholesterolemia. A Handbook for Counselors. Dallas: American Heart Asso- ciation, 1988. Hansen RG, Windham CT. Policies affect- ing food selection. Food and Nutrition News, January-February 1988. Dwyer JT. Diets for children and adoles- cents that meet the dietary goals. American Journal of Diseases of Children 134:1073-80, 1980. Dwyer JT, Dietz WH Jr, Andrews EM, et al. Nutritional status of vegetarian children. American Journal of Clinical Nutrition 35:204- 16, 1982. Department of Health and Human Services. Public Health Service. Surgeon General's Report on Nutrition and Health. Washington, DC: Department of Health and Human Services, Public Health Service, Pub. No. (PHS) 88-50210, 1988. National Research Council, Committee on Dietary Allowances, Food and Nutrition Board. Recommended Dietary Allowances, 9th ed. Washington, DC: National Academy Press, 1980. Block G, Dresser CM, Hartman AM. Nutri- ent sources in the American diet: Quantita- tive data from the NHANES II survey. IL Macronutrients and fats. American Journal of Epidemiology 122:27-40, 1985. Cleveland LE. Presentation at The Ameri- can Dietetic Association Conference for Advanced Practice and Research, Chicago, March 1988. Dwyer JT, Mayer J. Beyond economics and nutrition: The complex basis of food policy. Science 188:566-70, 1975. 127 Population Panel Report 415. 416. 417. 418. 419. 420. 421. 422. 423. 424. 425. 128 Kinsella JE. Trends in new product devel- opment: Modifying the nutrient composi- tion of animal products. Food Technology 41:62, 1987. Heimbach JT. Cardiovascular disease and diet: The public view. Public Health Reports 100:5-12, 1985. Campbell TC, O'Connor TP. Scientific evi- dence and explicit health claims in food ad- vertising. Journal of Nutrition Education 20:87, 1988. Hammond SL. Health advertising: The credibility of organizational sources. Com- munication Yearbook 10. Beverly Hills: Sage Publications, 1987. Hertzler AA, Owen C. Culture, families, and the change process: A systems approach. Journal of the American Dietetic Association 84:535-40, 543, 1984. Axelson ML. The impact of culture on food-related behavior. Annual Review of Nutrition 6:345-63, 1986. Krondl M, Coleman P. Social and biocultu- ral determinants of food selection. Progress in Food Science 10:179-203, 1986. Bass MA, Wakefield LM, Kolasa K. Com- munity Nutrition and Individual Food Behav- ior. Minneapolis, MN: Burgess, 1979. Liu K, Stamler J, Dyer A, et al. Statistical methods to assess and minimize the role of intra-individual variability in obscuring the relationship between dietary lipids and se- rum cholesterol. Journal of Chronic Diseases 31:399-418, 1978. James TN. Presidential address. AHA 53rd Scientific Sessions, Miami Beach, Florida, November 1980. Sure cures, quick fixes and easy answers. A cautionary tale about coronary disease. Circulation 63:1199A- 1202A, 1981. Swagler RM. Caveat Emptor! An Introduc- tory Analysis of Consumer Problems. Lexing- ton, MA: D.C. Heath, 1975. 426. 427. 428. 429. 430. 431. 432. 433. 434. 435. Popkin BM, Haines PS. Factors affecting food selection: The role of economics. Journal of the American Dietetic Association 79:419-25, 1981. Rizek RL, Peterkin BB. Food costs of US households, Spring 1977. Family Economics Review 14-19, Fall, 1979. Schafer R, Reger R, Gillespie A, et al. Diet quality of selected samples of women and sociodemographic and social-psychological correlates. Home Economics Research Journal 8:190-99, 1980. Smallwood D, Blaylock J. Impact of House- hold Size and Income on Food Spending Pat- terns. Washington, DC: U.S. Department of Agriculture, Technical Bulletin 1650, 1981. Windham CT, Wyse BW, Hurst RL, et al. Consistency of nutrient consumption pat- terns in the United States. Journal of the American Dietetic Association 78:587-95, 1981. Eppright ES, Fox HM, Fryer BA, et al. The North Central Regional Study of diets of preschool children: 2. Nutrition knowledge and attitudes of mothers. Journal of Home Economics 62:327-32, 1970. Schorr BC, Sanjur D, Erickson EC. Teen- age food habits. Journal of the American Dietetic Association 61:415-20, 1972. Adrian J, Daniel R. Impact of socioeco- nomic factors on consumption of selected food nutrients in the United States. Ameri- can Journal of Agricultural Economics 58:31- 38, 1976. Abdel-Ghany M, Schrimper RA. Food con- sumption expenditures and education of the homemaker. Home Economics Research Journal 6:283-92, 1978. Caliendo MA, Sanjur D. The dietary status of preschool children: An ecological approach. Journal of Nutrition Education 10:69-72, 1978. Population Panel Report 436. 437. 438. 439. 440. 441. 442. 443. 444. 445. Yperman AM, Vermeersch JA. Factors as- sociated with children’s food habits. Journal of Nutrition Education 11:72-76, 1979. Phillips DE, Bass MA, Yetley E. Use of food and nutrition knowledge by mothers of pre- school children. Journal of Nutrition Educa- tion 10:73-75, 1978. Yetley EA, Roderuck C. Nutritional knowl- edge and health goals of young spouses. Journal of the American Dietetic Association 77:31-41, 1980. Woolcott DM, Kawash GF, Sabry JH. Cor- relates of nutrition knowledge in Canadian businessmen. Journal of Nutrition Education 13:153-86, 1981. Redman BJ. The impact of women’s time allocation on food expenditure for meals away from home and prepared foods. American Journal of Agricultural Economics 62:234-37, 1980. Reilly MD. Working wives and conven- ience consumption. Journal of Consumer Research 8:407-18, 1982. Oritz B, MacDonald M, Ackerman N, et al. The effect of homemakers’ employment on meal preparation time, meals at home, and meals away from home. Home Economics Research Journal 9:200-06, 1981. Cronin FJ, Krebs-Smith SM, Wyse BW, et al. Characterizing food usage by demographic variables. Journal of the American Dietetic Association 81:661-73, 1982. Van Horn L, Ballew C, Liu K, et al. Diet, body size and serum lipids and lipopro- teins in young adults: Differences by sex and race. CARDIA study. American Journal of Epidemiology (in review). Randall E, Sanjur D. Food preferences— Their conceptualization and relationship to consumption. Ecology of Food and Nutrition 11:151-61, 1981. 446. 447. 448. 449. 450. 451. 452. 453. 454. 455. Wyant KW, Meiselman HL. Sex and race differences in food preferences of military personnel. Journal of the American Dietetic Association 84:169-75, 1984. Steinkamp RC, Cohen NL, Walsh HE. Re- survey of an aging population—Fourteen- year follow-up. The San Mateo Nutrition Study. Journal of the American Dietetic Association 46:103-10, 1965. Garcia PA, Battese GE, Brewer WD. Longi- tudinal study of age and cohort influences on dietary practices. Journal of Gerontology 30:349-56, 1975. Ohlson MA, Harper LJ. Longitudinal stud- ies of food intake and weight of women from ages 18 to 56 years. Journal of the American Dietetic Association 69:626-31, 1976. Brittin HC, Zinn DW. Meat buying prac- tices of Caucasians, Mexican-Americans, and Negroes. Journal of the American Dietetic Association 71:623-28, 1977. Grivetti L, Paquette M. Nontraditional eth- nic food choices among first generation Chinese in California. Journal of Nutrition Education 10:109-112, 1978. Salathe LE, Gallo AE, Boehm WT. The Im- pact of Race on Consumer Food Purchases. Washington, DC: U.S. Department of Agri- culture, Report No. ESCS-68, 1979. Wallendorf M, Reilly MD. Ethnic migra- tion, assimilation, and consumption. Jour- nal of Consumer Research 10:292-302, 1983. Nickols SY, Metzen EJ. Housework time of husband and wife. Home Economics Research Journal 7:85-97, 1978. Hafstrom JL, Schram VR. Housework time of wives: Pressure, facilitators, constraints. Home Economics Research Journal 11:245-55, 1983. 129 Population Panel Report 456. 457. 458. 459. 460. 461. 462. 463. 464. 130 Nickols SY, Fox KD. Buying time and saving time: Strategies for managing household production. Journal of Consumer Research 10:197-208, 1983. Gobel KP, Hennon CB. An empirical inves- tigation of the relationship among wife's employment status, stage in the family life cycle, meal preparation time, and expendi- tures for meals away from home. Journal of Consumer Studies and Home Economics 6:63- 78, 1982. Havlicek J Jr, Axelson JM, Capps O Jr, et al. Nutritional and economic aspects of con- venience and nonconvenience foods. In: Proceedings of Outlook 83 Conference, Wash- ington, DC: U.S. Department of Agricul- ture, 539-50, 1983. Windham CT, Wyse BW, Hansen RG, et al. Nutrient density of diets in the USDA Na- tionwide Food Consumption Survey, 1977- 1978: I. Impact of socioeconomic status on dietary density. Journal of the American Die- tetic Association 82:28-34, 1983. Skinner JD, Ezell JM, Salvetti NN, et al. Re- lationships between mothers” employment and nutritional quality of adolescents” diets. Home Economics Research Journal 13:218-25, 1985. Jalso SB, Burns MM, Rivers JM. Nutritional beliefs and practices. Journal of the American Dietetic Association 47:263-68, 1965. Grotkowski ML, Sims LS. Nutritional knowledge, attitudes, and dietary practices of the elderly. Journal of the American Dietetic Association 72:499-506, 1978. Sims LS. Dietary status of lactating women. 2. Relation of nutritional knowledge and at- titudes to nutrient intake. Journal of the American Dietetic Association 73:147-54, 1978. Yudkin S. Objectives in nutrition education. Journal of Human Nutrition 35:203-13, 1931. 465. 466. 467. 468. 469. 470. 471. 472. 473. 474. 475. Beavers I, Kelley M, Flenner J. Nutrition knowledge, attitudes, and food purchasing practices of parents. Home Economics Research Journal 11:134-42, 1982. Axelson ML, Federline T, Brinberg D. A meta-analysis of food and nutrition- related research. Journal of Nutrition Educa- tion 17:51-54, 1985. Pizin P, Vollnecke TA. Food likes and dis- likes. Annual Review of Nutrition 6:433-56, 1986. Weiffenbach JM, ed. Taste and Development: The Genesis of Sweet Preference. Bethesda, MD: Department of Health, Education, and Welfare, Public Health Service, National In- stitutes of Health, National Institute of Dental Research. DHEW Pub. No. (NIH) 77-1068, 1977. Cowart BJ. Development of taste percep- tion in humans: Sensitivity and preference throughout the lifespan. Psychology Bulletin 90:43-73, 1981. Birch L, Martin D. I don’t like it, I never tried it: Effect of exposure to food on two- year-old children’s food preferences. Appe- tite 4:353-60, 1982. Pliner P. The effects of mere exposure on liking for edible substances. Appetite 3:283, 1982. Birch L. The relationship between children’s food preferences and those of their parents. Journal of Nutrition Education 12:14-18, 1980. Pliner P. Family resemblance in food preferences. Journal of Nutrition Education 15:137-40, 1983. Cavalli-Sforza L, Feldman M, Chen K, et al. Theory and observation in cultural transmission. Science 218:19-27, 1982. Dwyer JT. Nutrition education and infor- mation. In: Better Health for Our Children: A National Strategy. The Report of the Select Population Panel Report 476. 477. 478. 479. 480. 481. 482. 483. Panel for the Promotion of Child Health. Washington, DC: U.S. Department of Health and Human Services, Public Health Service, Pub. No. (PHS)79-55071, 1981. Mullis RM, Pirie P. Lean meats make the grades: A collaborative nutrition education program. Journal of the American Dietetic As- sociation 88:191, 1988. Abrams DB, Elder JP, Carleton RA, et al. Social learning principles for organizational health promotion: An integrated approach. In: Cataldo M, Coates T, eds. Health and Industry: A Behavioral Medicine Perspective. New York: John Wiley & Sons, 28-51, 1986. Flora JA, Maibach E, Maccoby N. Mass media across four levels of health promo- tion intervention. In: Breslow J, ed. Annual Review of Public Health. Vol. 10, Palo Alto, CA: Annual Reviews, Inc. 181-201, 1989. Brody H, Sobel DS. A systems view of health and disease. In: Sobel DS, ed. Ways of Health: Holistic Approaches to Ancient and Contemporary Medicine. New York: Har- court, Brace, Jovanovich, 87-104, 1979. Clark M, DeBusk RF, Johannsson M, et al. Telephone-mediated dietary intervention for hypercholesterolemia. Poster presented at The First National Cholesterol Confer- ence, Arlington, VA, November 1988. Mahoney MJ, Caggiula AW. Applying be- havioral methods to nutritional counseling. Journal of the American Dietetic Association 72:372-77, 1978. Gorder DD, Dolecek TA, Coleman GG, et al. Dietary intake in the Multiple Risk Factor Intervention Trial (MRFIT): Nutrient and food group changes over 6 years. Journal of the American Dietetic Association 86:744-51, 1986. Witschi JC, Singer M, Wu-Lee M, et al. Family cooperation and effectiveness in a cholesterol-lowering diet. Journal of the American Dietetic Association 72:384-89, 1978. 484. 485. 486. 487. 488. 489. 490. Basch C, Eveland C, Portnoy B. Diffusion systems for education and learning about health. Family and Community Health 9:1-26, 1986. Zifferblatt SM, Wilbur CS, Pinsky JL. Influ- ence of ecologic events on cafeteria food se- lections: Understanding food habits. Journal of the American Dietetic Association 76:9-14, 1980. Zifferblatt SM, Wilbur CS, Pinsky JL. A new direction for public health care: Changing cafeteria eating habits. Journal of the American Dietetic Association 76:15-20, 1980. Blackburn H, Luepker RV, Kline FG, et al. The Minnesota Heart Health Program: A research and demonstration project in car- diovascular disease prevention. In: Mata- razzo JD, Weiss SM, Herd JA, et al., eds. Behavioral Health: A Handbook of Health En- hancement and Disease Prevention. New York: John Wiley & Sons, 1171-78, 1984. Farquhar JW, Fortmann SP, Maccoby N, et al. The Stanford Five City Project: An overview. In: Matarazzo JD, Weiss SM, Herd JA, et al., eds. Behavioral Health: A Handbook of Health Enhancement and Disease Prevention. New York: John Wiley, 1154-65, 1984. Lasater T, Abrams D, Artz L, et al. Lay volunteer delivery of a community-based cardiovascular risk factor change program: The Pawtucket experiment. In: Matarazzo JD, Miller NE, Weiss SM, eds. Behavioral Health: A Handbook of Health Enhancement and Disease Prevention. New York: John Wiley, 1166-70, 1984. Puska P, Koskela K, McAlister A, et al. Use of lay opinion leaders to promote diffusion of health innovations in a community pro- gramme: Lessons learned from the North Karelia project. Bulletin of the World Health Organization 64:437-46, 1986. 131 Population Panel Report 491. 492. 493. 494. 495. 496. 497. 498. 499. 500. 132 Ernst N, Cleeman J. Reducing high blood cholesterol levels: Recommendations from the National High Blood Cholesterol Edu- cation Program. Journal of Nutrition Educa- tion, 20:23-29, 1988. lacono JM, Puska P, Dougherty RM, et al. Effect of dietary fat on blood pressure in a rural Finnish population. American Journal of Clinical Nutrition 38:860-69, 1981. Enholm C, Huttunen JK, Pietinen P, et al. Effect of diet on serum lipoproteins in a population with a high risk of coronary heart disease. New England Journal of Medicine 307:850-55, 1982. Puska P, Iacono J, Nissinen A, et al. Con- trolled, randomized trial of the effect of die- tary fat on blood pressure. Lancet 1:1-5, 1983. Enholm C, Huttunen JK, Pietinen P, et al. Effect of a diet low in saturated fatty acids on plasma lipids, lipoproteins and HDL subfractions. Arteriosclerosis 4:265, 1984. Christakis G, Rinzler SH, Archer M, et al. The anti-coronary club. A dietary approach to the prevention of coronary heart dis- ease—a seven-year report. American Journal of Public Health—Nation's Health 56:299-314, February 1966. Meichenbaum D, Turk DC. Facilitating Treatment Adherence: A Practitioner's Guide- book. New York: Plenum Press, 1987. Bullet AC. Improving dietary education for patients with hyperlipidemia. Journal of the American Dietetic Association 72:277-81, 1978. Reeves RS, Foreyt JP, Scott LW, et al. Ef- fects of a low cholesterol eating plan on plasma lipids: Results of a 3-year commu- nity study. American Journal of Public Health 73:873, 1983. Snetselaar LG. Nutrition counseling in pre- vention and treatment of coronary heart disease. In: Snetselaar LG, ed. Nutrition Counseling Skills: Assessment, Treatment and 501. 502. 503. 504. 505. 506. 507. 508. 509. 510. Evaluation, 2nd ed. Rockville, MD: Aspen, 231-85, 1988. Johnson DW, Johnson RT. Nutrition Edu- cation: A model for effectiveness, a synthe- sis for research. Journal of Nutrition Educa- tion 17(suppl):S1-44, 1985. Glanz, K. Nutrition education for risk factor reduction and patient education: A review. Preventive Medicine 14:721-52, 1985. Glanz K, Mullis RM. Environmental pro- grams to promote healthy eating: A review of models, programs, and evidence. Health Education Quarterly 15:1-21, 1988. Bruno R, Arnold C, Jacobson L, et al. Ran- domized controlled trial of a nonpharma- cologic cholesterol reduction program at the worksite. Preventive Medicine 12:523-32, 1983. Bandura A. Social Learning Theory. Engle- wood Cliffs, NJ: Prentice-Hall, 1977. Disbrow D. A cost-benefit analysis of nutri- tion services for prevention of coronary heart disease. Topics in Clinical Nutrition 3:31-45, 1988. Glasgow RE, Rosen GM. Behavioral biblio- therapy: A review of self-help behavior therapy manuals. Psychological Bulletin 85:1-23, 1978. Elmore JG, Taylor CB, Flora JA, etal. A controlled trial of direct mail nutrition edu- cation strategies in a community setting. Unpublished report of the Stanford Heart Disease Prevention Program, 1986. Maibach EW, Flora JA, Nass CI. Changes in self-efficacy and health behavior in re- sponse to a minimal contact community health campaign (under revision). Miller N, Wagner E, Rogers PP. Worksite- based multifactorial risk intervention trial. Journal of the American College of Cardiology 11(suppl A):207A, 1988. Population Panel Report 511. 512. 313, 514. 515. 516. 517. 518. 519. King AC, Saylor KE, Foster S, et al. Pro- moting dietary change in adolescents: A school-based approach for modifying and maintaining healthful behavior. American Journal of Preventive Medicine 4:68-74, 1988. Saylor KE, Coates TJ, Killen J, et al. Nutri- tion education research: Feast or famine? In: Coates TJ, Petersen AC, Perry CL, eds. Promoting Adolescent Health. A Dialog on Research and Practice. New York: Academic Press, 355-380, 1982. Coates TJ, Jeffery RW, Slinkard LA. Heart healthy eating and exercise: Introducing and maintaining changes in health behaviors. American Journal of Public Health 71:15-23, 1981. Marcus AC, Wheeler RC, Cullen JW, et al. Quasi-experimental evaluation of the Los Angeles Know Your Body Program: Knowledge, beliefs, and self-reported behaviors. Preventive Medicine 16:803-15, 1987. Walter HJ, Holman A, Vaughan RD, et al. Modification of risk factors for CHD: Five- year results of a school-based intervention trial. New England Journal of Medicine 315:1093-1100, 1988. Walter, HJ. Primary prevention of chronic disease among children: The school-based “Know Your Body” intervention trials. Health Education Quarterly 16:215-28, 1989. Bush PJ, Zuckerman AE, Taggart VS, et al. Cardiovascular risk factor prevention in black school children: The “Know Your Body” evaluation project. Health Education Quarterly 16:215-17, 1989. Perry CL, Mullis RM, Maile MC. Modify- ing the eating behavior of young children. Journal of School Health 55:399-402, 1985. Puska P, Vartianinen E, Pallonen U, et al. The North Karelia Youth Project: Evalu- ation of two years of intervention on health 520. 521. 522. 523. 524. 525. 526. 527. behavior and CVD risk factors among 13-15 year old children. Preventive Medicine 11:550-70, 1982. Coates TJ, Barofsky I, Slinkard KE, et al. Modifying the snack food consumption pat- terns of inner city high school students: The great sensations study. Preventive Medicine 14:234-47, 1985. Parcel GS, Simons-Morton BG, O'Hara NM, et al. School promotion of healthful diet and exercise behavior: An integration of or- ganizational change and social learning theory interventions. Journal of School Health 57:150-56, 1987. Perry CL, Klepp KL, Halper A, et al. Pro- moting healthy eating and physical activity patterns among adolescents: Slice of life. Health Education Research: Theory and Practice 2:93-104, 1987. Berkman LF. Assessing the physical health effects of social networks and social support. Annual Review of Public Health 5:413-32, 1984. Perry CL, Luepker R, Murray D, et al. Parent involvement with children’s health promotion: The Minnesota Home Team. The American Journal of Public Health 78:1156-60, 1988. Perry CL, Luepker RV, Murray DM, et al. Parent involvement with children’s health promotion: A one-year followup of the Minnesota Home Team. Health Education Quarterly 16:171-80, 1989. Nader PR, Sallis JF, Rupp J, et al. San Diego Family Health Project: Reaching families through the schools. Journal of School Health 56:227-31, 1986. Neittaanmaki L, Koskela K, Puska P, et al. The role of lay health workers in commu- nity health education: Experiences of the North Karelia Project. Scandinavian Journal of Social Medicine 8:1-7, 1980. 133 Population Panel Report 528, 529. 530. 531. 532, 533. 534. 535. 536. 537. 134 Meyer AJ, Maccoby N, Farquhar JW. The role of opinion leadership in a cardiovascu- lar health education campaign. In: Ruben B, ed. Communication Yearbook 1. New Brunswick, NJ: Transaction, 579-91, 1977. Gorbach SL, Morrill-LaBrode A, Woods MN, Dwyer J, et al. Changes in food pat- terns during a low-fat dietary intervention in women. Journal of the American Dietetic Association 90:802-09, 1990. Brownell KD, Heckerman CL, Westlake RJ, et al. The effect of couples training and partner co-operatives in the behavioral treatment of obesity. Behavior Research and Therapy 16:323-33, 1978. Abrams DB, Follick MJ. Behavioral weight loss intervention at the worksite: Feasibility and maintenance. Journal of Consulting and Clinical Psychology 47:243-51, 1983. McGandy RB, Hall B, Ford C, et al. Dietary regulation of blood cholesterol in adoles- cent males: A pilot study. American Journal of Clinical Nutrition 25:61-66, 1972. Ellison RC, Capper AL, Goldberg R], et al. The environmental component: Changing school food service to promote cardiovascu- lar health. Health Education Quarterly 16:285-98, 1989. Glanz K, Seewald-Klein T. Nutrition at the worksite: An overview. Journal of Nutrition Education 18(suppl):S1-7, 1986. Carlier J, Joossens JV, Kornitzer M, et al. National experiments in prevention: Bel- gium. Abstract presented at the Seventh Congress of the European Society of Cardi- ology, Amsterdam, June 1976. Lefebvre CR, Flora JA. Social marketing and public health interventions. Health Education Quarterly 15:299-315, 1988. American Heart Association. Heart at Work. Dallas: American Heart Association, 1984. 538. 539. 540. 541. 542. 543. 544. 545. 546. 547. 548. Soriano E, Dozier DM. Selling nutrition and heart health diet behaviors at point of purchase. Journal of Applied Nutrition 30:55- 65, 1978. Bettman JT. An Information Processing Theory of Customer Choice. Reading, MA: Addison Wesley, 1979. St. Jeor ST, Winston M. A national educa- tional event: The food festival designed to help consumers implement American Heart Association Dietary Guidelines for Healthy Americans. Nutrition Today 10, July-August 1987. Muller TE. The use of nutritive composi- tion data at the point of purchase. Journal of Nutrition Education 16:137-41, 1984. Davis-Chervin D, Rogers T, Clark M. Influ- encing food selection with point-of-choice nutrition information. Journal of Nutrition Education 17:18-22, 1985. Labuza TP. A perspective on health claims in food labeling. Cereal Foods World 132:264, 1987. Achabal DD, McIntyre SH, Bell CH, et al. The effect of nutrition P-O-P signs on consumer attitudes and behavior. Journal of Retailing 63:9-24, 1987. Wilkerson JB, Mason B, Paksov CH. As- sessing the impact of short-term supermar- ket strategy variables. Journal of Marketing Research 19:72-86, 1982. Hixson ML. The four heart environmental nutrition programs. Poster presented at The First National Cholesterol Conference, Arlington, VA, November 1988. Olson CM, Bisogni CA, Thorneny PF. Evaluation of a supermarket nutrition education program. Journal of Nutrition Education 14:171-245, 1982. Pennington JAT, Wisniowski LA, Logan GB. In-store nutrition information programs. Journal of Nutrition Education 20:5-10, 1988. Population Panel Report 549. 550. 551. 552. 553. 554. 555. 556. 557. 558. Ernst N, Frommer P, Moskowitz J, et al. Foods for Health: Report of the Pilot Program. U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, NIH Pub. No. 83-2036, August 1983. Ernst N, Wu M, Frommer P, et al. Nutrition education at the point of purchase: The Foods for Health Project evaluated. Preven- tive Medicine 15:60-73, 1986. Jeffery RW, Pirie PL, Rosenthal BS, et al. Nutrition education in supermarkets: An unsuccessful attempt to influence knowl- edge and product sales. Journal of Behavioral Medicine 5:189-200, 1982. Levy AS, Mathews O, Stephenson M, et al. The impact of a nutrition information pro- gram on food purchases. Journal of Public Policy Marketing 4:1-16, 1985. Levy A, Schucker R, Tenney J, et al. Nutri- tion shelf labeling and consumer purchase behavior. Division of Consumer Studies, Center for Food Safety and Applied Nutri- tion, Food and Drug Administration, 1989. Levy A, Schucker R, Tenney J, et al. Con- sumer demand for detailed nutrition: A case study. Journal of Nutrition Education 20:161-66, 1988. Schucker R. Food and Drug Administra- tion, personal communication, 1989. Mullis RM, Hunt MK, Foster M, et al. The Shop Smart for Your Heart grocery program. Journal of Nutrition Education 19:225-28, 1987. Peterson G, Elder JP, Knisley PM, et al. De- veloping strategies for food vendor inter- vention: The first step. Journal of the Ameri- can Dietetic Association 86:659, 1986. Mayer JA, Heins JA, Vogel JM, et al. Pro- moting low-fat entree choices in a public cafeteria. Journal of Applied Behavior Analysis 19:397-402, 1986. 559. 560. 561. 562. 563. 564. 565. 566. 567. 568. 569. Wagner JL, Winett RA. Prompting one low-fat, high fiber selection in a fast food restaurant. Journal of Applied Behavior Analysis 21:179-85, 1988. Wilbur CS, Zifferblatt SM, Pinsky JL, et al. Healthy vending: A cooperative pilot re- search program to stimulate good health in the marketplace. Preventive Medicine 10:85- 93, 1931. Albright CL, Flora JA, Fortmann SP. Res- taurant menu labeling: Impact of nutrition information on entree sales and patron attitudes. Health Education Quarterly 17:157- 67, 1990. Dubbert PM, Johnson WG, Schlundt DG, et al. The influence of caloric information on cafeteria food choices. Journal of Applied Be- havioral Analysis, 17:85-91, 1984. Cincirpini PM. Changing food selections in a public cafeteria. Behavior Modification 8:520-39, 1984. Hixson ML. Personal communication, 1989. Colby JJ, Elder JP, Peterson G, et al. Pro- moting the selection of healthy food through menu item description in a family- style restaurant. American Journal of Preven- tive Medicine 3:171-77, 1987. Montgomery A. School lunch: The case of low-fat, high-fiber meals. Nutrition Action Health Letter 14, September 1987. Frank GC, Vaden A, Martin J. School health promotion: Child nutrition pro- grams. Journal of School Health 57:451-60, 1987. Simons-Morton BG, Parcel GS, O'Hara NM. Implementing organizational changes to promote healthful diet and physical activity at school. Health Education Quarterly 15:115- 30, 1988. Blair SN, Smith M, Collingwood TR, et al. Health promotion for educators: Impact on absenteeism. Preventive Medicine 15:166-75, 1986. 135 Population Panel Report 570. 571. 572. 573. 574. 575. 576. B77. 578. 136 Department of Health and Human Services. Promoting Health/Preventing Disease: Objec- tives for the Nation. Washington, DC: U.S. Government Printing Office, 1980. Morris PD, Morris ER. Family practice resi- dents’ compliance with preventive medi- cine recommendations. American Journal of Preventive Medicine 4:161-65, 1988. National Research Council, Committee on Nutrition in Medical Education, Food and Nutrition Board. Nutrition Education in U.S. Medical Schools. Washington, DC: National Press, 1985. Gans K, Jack B, Lefebvre RC. Cholesterol screening and management: Training pro- gram for family medicine residents. Poster presented at The First National Cholesterol Conference, Arlington, VA, November 1988. Tannon C, Rogers E. Diffusion research methodology: Focus on health care organi- zations. In: Gorden G, Fisher G, eds. Diffusion of Medical Technology: Policy and Research Planning Perspectives. Cambridge, MA: Ballinger Books, 1986. Mittelmark MB, Luepker RV, Grimm R, et al. The role of physicians in a community- wide program for prevention of cardiovas- cular disease: The Minnesota Heart Health Program. Public Health Reports 103:360-65, 1988. Ellison RC. Give diet a chance in lowering cholesterol levels. Archives of Internal Medi- cine 148:1017-19, 1988. Kottke TE, Foels JK, Hill C, et al. Nutrition counseling in private practice: Attitudes and activities of family practice physicians. Preventive Medicine 13:219-25, 1984. Schucker B, Wittes JT, Cutler JA, et al. Change in physician perspective on choles- terol and heart disease: Results from two national surveys. Journal of the American Medical Association 258:3521-27, 1987. 579. 580. 581. 582. 583. 584. 585. 586. Block L, Banspach SW, Gans K, et al. Im- pact of public education and continuing medical education on physician attitudes and behavior concerning cholesterol. American Journal of Preventive Medicine 4:255-60, 1988. Hyman D, Flora JA, Reynolds K, et al. Cho- lesterol screening in primary care practice: A pilot study. Poster presented at The First National Cholesterol Conference, Arling- ton, VA, November 1988. Kimm SY, Payne GH, Lakatos E, et al. Physicians’ attitudes and practices regard- ing cholesterol in children. Poster pre- sented at The First National Cholesterol Conference in Arlington, VA, November 1988. Kofron PM, Mullis R, Bluhm J, et al. Effect of a screening dietary educational interven- tion for hypercholesterolemia in a medical clinic. Poster presented at The First Na- tional Cholesterol Conference, Arlington, VA, November 1988. Secker-Walker RH, Weaver SO, Bach N, et al. Factors affecting the proportion of patients having serum cholesterol meas- ured in family practice, and the threshold for giving dietary advice. Poster presented at The First National Cholesterol Confer- ence, Arlington, VA, November 1988. Admire |B, Bell LP, Harper DE, et al. Physi- cian acceptance of cholesterol management in primary care: Ways of improving it. Poster presented at The First National Cholesterol Conference, Arlington, VA, November 1988. Bache-Wiig ML, Hoffman E, Glanz ML, et al. Patient perceptions of clinic based cholesterol management programs. Poster presented at The First National Cholesterol Conference, Arlington, VA, November 1988. Bache-Wiig ML, McComas KA, Admire |B, et al. Cholesterol management program in Population Panel Report 587. 588. 589. 590. 591. 592. 593. 594. family practice clinics. Poster presented at The First National Cholesterol Conference, Arlington, VA, November 1988. Crouch M, Sallis JF, Farquhar JW, et al. Personal and mediated health counseling for sustained dietary reduction of hyper- cholesterolemia. Preventive Medicine 15:282- 91, 1986. Boekeloo BO, Becker DM, Pearson TA, et al. A randomized trial of strategies to improve cholesterol management practices of physi- cians in training. Poster presented at The First National Cholesterol Conference, Arlington, VA, November 1988. Clearfield MB. Physician awareness of ele- vated cholesterol. Poster presented at The First National Cholesterol Conference, Ar- lington, VA, November 1988. Solkowitz SN, Adler LJ, Baron RB, et al. Recruitment of physician subjects for a ran- domized controlled trial on the efficacy of continuing medical education for the treat- ment and detection of hypercholesterol- emia. Poster presented at The First Na- tional Cholesterol Conference, Arlington, VA, November 1988. Weinstein MC, Stason WB. Cost-effective- ness of interventions to prevent or treat CHD. American Review of Public Health 6:41, 1985. Disbrow D. Cost and Benefits of Nutritional Care, Phase 1. Chicago: American Dietetic Association, 1983. Disbrow D. The costs and benefits of nutri- tion services: A literature review. Journal of the American Dietetic Association 89(suppl): 53-566, 1989. Atkin CK. Effects of television advertising on children. In: Palmer EL, Dorr A, eds. Children and the Faces of Television: Teaching, Violence, Selling. New York: Academic Press, 287-305, 1980. 595. 596. B97. 598. 599. 600. 601. 602. 603. Barcus FE. The nature of television adver- tising to children. In: Palmer EL, Dorr A, eds. Children and the Faces of Television: Teaching, Violence, Selling. New York: Academic Press, 273-85, 1980. Freimuth VS, Hammond SL, Stein JA. Health advertising: Prevention for profit. American Journal of Public Health 78:557-61, 1988. Levy AS, Stokes RC. Effects of a health pro- motion advertising campaign on sales of ready-to-eat cereals. Public Health Reports 102:398, 1987. Heimbach J, Levy A. Changing public be- liefs about diet and health. Paper presented at the Food Safety and Nutrition Update Conference, New York, 1987. Dwyer JT. Dietary recommendations and policy implications: The U.S. experience. In: Weininger J, Briggs GM, eds. Nutrition Update 1:315-56. New York: John Wiley, 1983. Breidenstein BC. Changes in consumer atti- tudes toward red meat and their effects on marketing strategies. Food Technology 112- 16, January 1988. Gurian GL. Promoting the community’s health through a food service establishment incentive program. American Journal of Pre- ventive Medicine, 3:342-44, 1987. Stern MP, Farquhar JW, Maccoby N, et al. Results of a 2-year health education cam- paign on dietary behavior. Circulation 54:826-33, 1976. Maccoby N, Farquhar JW, Wood PD, et al. Reducing the risk of cardiovascular disease: Effects of a community-based campaign on knowledge and behavior. Journal of Com- munity Health 3:100-14, 1977. 137 Population Panel Report 604. 605. 606. 607. 608. 609. 610. 611. 612. 138 Puska P, Tuomilehto J, Salonen J, et al. The North Karelia Project: Evaluation of a Compre- hensive Communication Programme for Control of Cardiovascular Disease in North Karelia, Finland, 1972-1977. Copenhagen: WHO- EURO, 1981. Fortmann S, Williams P, Hulley S, et al. Does dietary health education reach only the privileged? The Stanford Three Com- munity Study. Circulation 66:77-82, 1982. Rossouw JE, Jooste PL, Kotze JP, et al. The control of hypertension in two communi- ties: An interim evaluation. South African Medical Journal 60:208, 1981. Egger G, Fitzgerald W, Frape G, et al. Results of a large scale media antismoking campaign: North Coast “Quit for Life” Programme. British Medical Journal 296:1125, 1983. Gutzwiller F, Nater B, Martin J. Commu- nity-based primary prevention of cardio- vascular disease in Switzerland: Methods and results of the National Research Pro- gram (NRP 1A). Preventive Medicine 14:482- 91, 1985. National Cholesterol Education Program. Strategy Development Workshop for Minorities: Summary Report. Bethesda, MD: U.S. De- partment of Health and Human Services, Public Health Service, National Institutes of Health, National Heart, Lung, and Blood Institute, 1988. Weaver GR. Contrast cultures continuum. In: Weaver GR, ed. Readings in Cross-Cul- tural Communication. Lexington, MA: Ginn Press, 226-28, 1987. Sanjur D. Social and Cultural Perspectives in Nutrition. Englewood Cliffs, NJ: Prentice Hall, 33-83, 1982. National Center for Health Statistics. Monthly Vital Statistics Report, Vol. 37, No. 6, Supplement. Hyattsville, MD: U.S. De- partment of Health and Human Services, 613. 614. 615. 616. 617. 618. 619. 620. Public Health Service, Pub. No. (PHS) 88- 1120, September 30, 1988. National Center for Health Statistics. Total serum cholesterol levels of adults 18-74 years: United States, 1971-74. Vital and Health Statistics, Series 11, No. 205. Hyattsville, MD: U.S. Department of Health, Education, and Welfare, Public Health Service, Pub. No. (PHS) 78-1652, 1978. Schoenborn CA, Danchik KM. Health prac- tices among adults: United States, 1977. Advance Data From Vital and Health Statistics, No. 64. Hyattsville, MD: U.S. Department of Health and Human Services, Public Health Service, National Center for Health Statistics, November 4, 1980. Moss AJ. Changes in cigarette smoking and current smoking practices among adults, United States, 1978. Advance Data From Vital and Health Statistics, No. 52. Hyattsville, MD: U.S. Department of Health, Education, and Welfare, Public Health Service, National Center for Health Statistics, September 19, 1979. Dawson D. Ethnic differences in female overweight: Data from the 1985 National Health Interview Survey. American Journal of Public Health 78:1326-29, 1988. Kumanyika S. Obesity in black women. Epidemiology Reviews 9:31-50, 1987. Block G, Rosenberger WF, Patterson BH. Calories, fat and cholesterol: Intake pat- terns in the US population by race, sex and age. American Journal of Public Health 78:1150-55, 1988. Jerome NW. Diet and acculturation. In: Jerome NW, Kandel RF, Pelto GH, eds. Nu- tritional Anthropology. New York: Redgrave Publishing, 1980. Kaplan AR, Hoover NA, Moore WB. The Minnesota Ethnic Food Book. St. Paul: Min- nesota Historical Society Press, 1986. Population Panel Report 621. 622. 623. 624. 625. Office of Cancer Communications. Background Paper for Hispanic Com- munication Plan. Bethesda, MD: U.S. Department of Health and Human Services, Public Health Service, Na- tional Institutes of Health, National Cancer Institute, 1987. Perez-Stable EJ, McMillen MM, Harris MI, et al. Self-reported diabetes in Mexican Americans: HHANES 1982-84. American Journal of Public Health 79:770- 72,1989. Hanis CL, Ferrell RE, Barton SA, et al. Diabetes among Mexican Americans in Starr County, Texas. American Journal of Epidemiology 118:659-72, 1983. Hazuda HP: Difference in socioeco- nomic status and acculturation among Mexican Americans and risk for cardio- vascular disease. Task Force on Black and Minority Health, U.S. Department of Health and Human Services, unpub- lished manuscript, 1985. Castro FG, et al. Risk factors for coro- nary heart disease in Hispanic popula- tions: A review. Hispanic Journal of Psy- chology (in press). 626. 627. 628. 629. Wolfe WS, Sanjur D. Contemporary diet and body weight of Navajo women receiving food assistance: An ethno- graphic and nutritional investigation. Journal of American Dietetic Association 88:822-27, 1988. Bass MA, Wakefield LM. Nutrient intake and food patterns of Indians of Standing Rock Reservation. Journal of the American Dietetic Association 64:36- 41, 1974. Douglas M. Food in the Social Order. New York: Russell Sage Foundation, 1984. Carrillo JE. A rationale for effective smoking prevention and cessation interventions in minority communities. Smoking Behavior and Policy Discussion Paper Series. Cambridge, MA: Harvard University, John F. Kennedy School of Government, Institute for the Study of Smoking Behavior and Policy, 1987. 139 RETURN PUBLIC HEALTH LIBRARY TO mp. 42 Warren Hall 642-251 NF ~ tery r——s (030913470