key: cord-0715518-7grnn2bz
authors: Moonen, Hanneke PFX.; Hermans, Anoek JH.; Jans, Inez; van Zanten, Arthur RH.
title: Protein requirements and provision in hospitalised COVID-19 ward and ICU patients: agreement between calculations based on body weight and height, and measured bioimpedance lean body mass
date: 2022-03-04
journal: Clin Nutr ESPEN
DOI: 10.1016/j.clnesp.2022.03.001
sha: ab0191411462d72609da32ec00223c65ad9d455b
doc_id: 715518
cord_uid: 7grnn2bz

Background A large proportion of hospitalised COVID-19 patients are overweight. There is no consensus in the literature on how lean body mass (LBM) can best be estimated to adequately guide nutritional protein recommendations in hospitalised patients who are not at an ideal weight. We aim to explore which method best agrees with lean body mass as measured by bioelectric impedance (LBMBIA) in this population. Methods LBM was calculated by five commonly used methods for 150 hospitalised COVID-19 patients previously included in the BIAC-19 study; total body weight, regression to a BMI of 22.5, regression to BMI 27.5 when BMI>30, and the equations described by Gallagher and the ESPEN ICU guideline. Error-standard plots were used to assess agreement and bias compared to LBMBIA. The actual protein provided to ICU patients during their stay was compared to targets set using LBMBIA and LBM calculated by other methods. Results All methods to calculate LBM suffered from overestimation, underestimation, fixed- and proportional bias and wide limits of agreement compared to LBMBIA. Bias was inconsistent across sex and BMI subgroups. Twenty-eight ICU patients received a mean of 51.19 (95%-BCa CI 37.1;64.1) grams of protein daily, accumulating to a mean of 61.6% (95%-BCa CI 43.2;80.8) of TargetBIA during their ICU stay. The percentage received of the target as calculated by the LBMGallagher method for males was the only one to not differ significantly from the percentage received of TargetBIA (mean difference 1.4% (95%-BCa CI -1.3;4.6) p=1.0). Conclusions We could not identify a mathematical method for calculating LBM that had an acceptable agreement with LBM as derived from BIA for males and females across all BMI subgroups in our hospitalised COVID-19 population. Consequently, discrepancies when assessing the adequacy of protein provision in ICU patients were found. We strongly advise using baseline LBMBIA to guide protein dosing if possible. In the absence of BIA, using a method that overestimates LBM in all categories may be the only way to minimise underdosing of nutritional protein. Trial registration The protocol of the BIAC-19 study, of which this is a post-hoc sub-analysis, is registered in the Netherlands Trial Register (number NL8562).

Obesity is a significant independent risk factor for hospitalisation in Coronavirus Disease 2019 (COVID- 19) patients (1, 2) . A large proportion of hospitalised COVID-19 patients, and by extend, ICU patients, are thus overweight. The prevalence of sarcopenic obesity has increased infection rates and morbidity related to COVID-19 (3) . A positive correlation between high nutritional risk and adverse clinical outcomes of COVID-19 has been observed (4) .

It is suggested that a high protein diet is beneficial during COVID-19 (3), as protein provision may prevent further breakdown of muscle protein for gluconeogenesis and thereby the patient from going into a further catabolic state.

Nutrition guidelines advise increasing the given protein quantity as the illness becomes more severe but vary between prescribing 1.2 to 2.5 g/kg of protein a day in the intensive care unit (ICU) (5) (6) (7) . One study showed that although targets of >1.2 g/kg/day of protein were hard to achieve in COVID-19 ICU patients, a supply of at least 0.8 g/ideal body weight (IBW)/day was already related to lower mortality (8) .

However, setting protein targets is challenging when patients are not at IBW. As in general, the overweight (Body mass index (BMI) ≥ 25 kg/m 2 ), or obese body (BMI≥ 30 kg/m 2 ) contains less protein per kilogram of body weight, use of total body weight (TBW) likely result in an overestimation of protein needs in overweight and obese persons. Currently, numerous mathematical formulas try to account for variations in body composition (such as between biological sexes) by estimating fat-free or lean body mass (LBM), which is assumed to be the true determinant of protein requirement (9) . It is still unclear which method is superior, which is reflected by discrepancies, or vagueness in recommendations between, and sometimes within, nutritional guidelines, which suggest multiple methods, or fail to state whether TBM, LBM, or IBW should be used (7, (9) (10) (11) . Slight variations in the definitions of fat-free mass (FFM) and LBM between sources further confuse the discussion.

Bioelectric impedance analysis (BIA) is a technique that calculates the volume of body water compartments through the use of measured electric reactance and resistance. The incorporated software then derives LBM through validated regression analyses based on a healthy reference population. BIA derived LBM (LBMBIA) for calculating protein needs has substantial theoretical advantages over mathematical methods regarding body composition (12) . In addition, BIA measurements can be performed at the bedside, in contrast to other direct methods such as dual-energy X-ray absorptiometry. However, BIA is not ubiquitously available and can pose J o u r n a l P r e -p r o o f challenges related to disinfection when used on a high volume of patients with a transmittable disease such as COVID-19. Therefore, it is worth exploring the agreement between BIA and commonly used mathematical formulas.

We previously conducted a prospective observational study in which all hospitalised patients for COVID-19 underwent BIA measurements within 24 hours of hospital admission (13) . The current post-hoc study compares the agreement between LBMBIA and five mathematical methods in estimating LBM in this COVID-19 population.

In addition, we retrospectively compare protein provision adequacy in our COVID-19 ICU population based on LBMBIA to that based on LBM predicted by other methods.

For this post-hoc sub-study, baseline data previously collected for the prospective BIAC-19 study were used. The

Bioelectric impedance body composition and phase angle concerning 90-day adverse outcome in hospitalised COVID-19 ward and ICU patients: the prospective BIAC-19 study aimed to associate baseline (<24 hours of hospital admission) BIA body composition parameters with 90-day adverse outcome of COVID-19 (13) . The BIAC-19 study protocol has been registered in the Netherlands Trial Register (number NL8562).

The study was performed between April 10 th and 17 th , 2020, and again between October 10 th 2020 and February 11 th 2021, at Gelderse Vallei Hospital, a teaching hospital in Ede, The Netherlands. The hospital has two ICU units, with a combined capacity of 18 beds. Thirty-eight general ward COVID-19 beds were available.

Protein targets in the wards are set according to actual (BMI 20-30 kg/m 2 ) or corrected body weight (BMI <20 kg/m 2 adjusted to 20 kg/m 2 ; BMI >30 kg/m 2 adjusted to 27 kg/m 2 ). In addition, the Gallagher method is described in the local protocol. Gallagher et al. developed an equation to calculate percentage body fat through sex, age, BMI, ethnicity and regression models based on the measured (by 4-compartment model (4C) or dual-energy Xray absorptiometry (DXA)) body fat of 1626 healthy adults with a BMI ≤ 35 kg/m 2 (14) . The Dutch dietary guidelines use a transformation of the original Gallagher formula, to approximate LBM at which protein provision J o u r n a l P r e -p r o o f is targeted (15) . This method is currently not routinely used in our hospital but is regarded as a potentially superior method (9, 16) .

Protein targets in the ICU are calculated by our computerized nutrition protocol, and are set according to actual (BMI < 27 kg/m 2 ), corrected (BMI 27-30 kg/m 2 ; regression to BMI of 27 kg/m 2 ), or ideal body weight (BMI > 30 kg/m 2 ; regression to BMI 21 kg/m 2 in women and BMI 22.5 kg/m 2 in men), and amount to 1.5g/kg/day in BMI <30 kg/m 2 , 2.0g/kg/day in BMI 30-40 kg/m 2 or 2.5g/kg/day in BMI ≥40 kg/m 2 . A progressive feeding strategy towards 100% of targets at admission day four is used (10) . Actual enteral/parenteral nutritional and nonnutritional energy and protein provision are automatically calculated hourly. Oral nutrition is currently not incorporated, as it cannot be done automatically and oral nutrition is not usually a substantial contribution to the total nutritional intake in ICU patients.

The BIAC-19 study included patients aged 18 years or above admitted to the hospital with COVID-19 symptoms and proved SARS-CoV-2 positive through polymerase chain reaction-test and had BIA measurements within 24 hours after hospital admission. Exclusion criteria were pregnancy, electrical implants, wounds or skin damage at the designated electrode sites, or inability to maintain posture for 5 minutes.

Patient subgroups for the current study were defined by biological sex (female/male) and BMI category. Normal weight was defined as a BMI <25 kg/m 2 , overweight as BMI 25-30 kg/m 2 and obese as BMI>30 kg/m 2 .

For the secondary research question addressing protein provision adequacy in the ICU, patients admitted to the ICU after transfer to another hospital were excluded, as no nutrition records were available. In addition, patients who only received oral nutrition were excluded, as protein contents of oral nutrition are not registered.

Trained researchers conducted BIA measurements with the InBody S10 ® (InBody Co., Ltd., Seoul, Korea). This multi-frequency, segmental impedance analyser requires height, weight, and sex as input parameters. Height and weight as measured upon admission were used. When circumstances did not allow measurements, height as J o u r n a l P r e -p r o o f provided by the patient or their representative was entered. BIA measurements were performed in a supine position with reusable electrodes attached to the left and right thumb and middle finger, and both ankles.

Inbody regards FFM and LBM as synonyms, defined as TBW minus non-essential storage fat mass (FM), corrected for hydration status through extracellular/total body water ratio (12) . In this definition, FFM/LBM includes essential fats, such as those stored in organs, the central nervous system and bone marrow. In other sources TBW minus FM is usually regarded as the LBM, whereas FFM is defined as LBM minus essential body fat. To avoid confusion, we choose to use only the term LBM for TBW minus FM.

Demographic and clinical data previously collected for the BIAC-19 study from local electronic medical record systems MetaVision ® (iMDsoft, Tel Aviv, Israel) and NeoZIS ® (MI Consultancy, Katwijk, The Netherlands) and

NeoZIS® (MI Consultancy, Katwijk, The Netherlands) were reused for the current study, i.e., age, sex, ethnicity, height, weight, and protein provision, specifics of the length of stay (LOS) and ventilation in ICU patients.

In addition to measured TBW (kg), four equations for LBM were chosen for comparisons with LBMBIA (kg). The methods aim to approximate IBW (1), adjusted body weight (2/3) or LBM (4) , which in all methods is regarded as a proxy for the true determinant of protein requirement: LBM (9). To improve readability, 'LBM' is used in all equations from hereon 

Descriptive statistics were calculated for demographics and protein provision in ICU patients. The quantilequantile plots were visually assessed for the normality of the distribution of continuous data. Continuous values are reported as mean (95% bias-corrected accelerated bootstrap confidence intervals (95%-BCa CI) based on 1000 samples) or median (interquartile range), discrete data are presented as numbers (%). Biological males were compared to female patients. Differences were assessed using independent samples t-tests for continuous data or chi-squared tests for categorical data. When test assumptions were not met, Mann-Whitney U tests or Fisher's exact tests were used.

We visually checked that the scatter plots showed a monotonic relation between LBMBIA and each method for all subgroups. Subsequently, a correlation analysis was conducted using Spearman's rank correlation coefficient, as the distribution of the variables was not normal. For this and all subsequent agreement analyses, the normal weight and overweight groups were disregarded when considering the LBM27.5 method, as it uses TBW in BMI <30 kg/m 2 . As Spearman's correlation only reveals the strength and mean direction of the association but does not reveal information on the presence of a systematic bias, we continued to construct error-standard plots. In this method, the difference or error between two measurements is plotted against the reference or standard method, in this case, BIA-LBM. This method was chosen over the Bland-Altman plot, where the difference is plotted against the mean of the two methods, as this can lead to underestimation of proportional bias, and in this case, the BIA-LBM method was considered the reference/standard method (Concept illustrated in Additional File 1). The 95-% Limits of agreements (average difference ± 1.96 standard deviations) with their 95% confidence intervals were calculated and plotted for each comparison. A significant result on a one-sample t-test comparing the mean of the differences to 0 was used to confirm fixed bias whenever visual inspection of the plots was suggestive of one (males and females separately). Where relevant, a sensitivity analysis of the t-test excluding visual outliers was conducted. The presence of proportional bias (i.e. a relationship between the size of the error and size of the reference value) was assessed visually and formally by regressing the difference on the reference value (i.e. J o u r n a l P r e -p r o o f LBMBIA) (males and females separately). The assumption for homogeneity of variance for linear regression was confirmed through non-significant Levene's test. Proportional bias was considered proven when a relationship was identified (i.e., a significant slope of the regression line).

Protein targets were calculated as 1.3g/day/LBM and incorporated progressive feeding during the first three days of ICU admission (i.e., 

One-hundred-and-fifty patients were included in the BIAC-19 prospective study and subsequent post-hoc analyses. All the included patients were of white of Western-European descent. Table 1 summarises baseline characteristics and measurements and compares those of biological males and females.

All mathematical methods for calculating LBM correlated significantly with LBMBIA at the level of p-value <.001 (Additional File 2, Table 1 ). LBMGallagher showed the highest correlation coefficient for all subgroups except overweight females, where LBM27.5 reached the same coefficient as LBMGallagher. Table 2 ). The visual outlier on all plots except LBMTBW discerned herself from the cohort with an LBM% of 80% compared to a mean of 62% (95%-BCa CI 59.5 -64.9) for females. A sensitivity analysis excluding this outlier did not change the significance of these findings.

Proportional bias was suspected from visual inspection of all plots except for TBW and confirmed by regressing the difference between the methods and LBMBIA, separately for males and females. A relationship between the error size and the reference value size was confirmed in all methods except TBW (males p=.8; females p=.087) (Additional File 2, Table 3 ).

Forty-one (27%) patients eventually had to be admitted to the ICU. Two ICU patients (5%) were admitted to the ICU after transfer from another hospital, and eleven (27%) only received oral nutrition, so no nutrition records were available. Consequently, 28 (68%) of the ICU patients could be included in the protein provision ICU subanalyses (Table 2) 

We aimed to assess which method approximates lean body mass best compared with bioelectric impedance in the hospitalised COVID-19 population. Total body weight and four other common methods were used; regression to a BMI of 22.5 kg/m 2 , regression to BMI 27.5 kg/m 2 when BMI>30 kg/m 2 , and the equations described by

Gallagher and the ESPEN ICU guideline (14, 10) . Although all methods were correlated with the reference method LBMBIA, we could not identify a mathematical method for calculating LBM that had an acceptable agreement with LBMBIA for males and females across BMI subgroups.

Although the LBMGallagher had the smallest overall 95%-CI, it could still lead to an over-or underestimation of the LBM of 16.4kg. Furthermore, all methods were subject to fixed bias (mean difference deviates from 0) when assessing males and females separately, except the LBMESPEN for males. All methods except TBW also had proportional bias (association between the difference between measurements and the size of the value measured).

The confidence intervals were wide for all methods studied, and visual inspection of the plots suggested that the regression slopes for proportional bias were different per sex/BMI subgroup. We are confident that there is no easy workaround to correct both fixed and proportional bias and make one of the methods agree on an acceptable level with LBMBIA across the whole cohort.

The overestimation of LBM based on TBW (figure 1-2 panels D) could be expected, as the fat% is never zero, especially in the current population. Our results show that the size of the overestimation varied widely, although it understandably increased with BMI. The same can be said for LBM27.5, as this method essentially presumes a weight equivalent of BMI 27.5 kg/m 2 to be the LBM. For example, a person of 170 cm in height with a BMI of 31 kg/m 2 , is presumed to have a LBM of (27.5 * 1.7 2 =) 79.5 kg on a weight of 89,6 kg, giving him a LBM% of (78.5/89.6 * 100=) 89%. In reality, excluding the very athletic, most of our patients with a BMI of 31 kg/m 2 will not have a fat% of (100-89=) 11%. Thus, the LBM27.5 method becomes more realistic as actual BMI increases (up to a certain point), explaining the proportional bias that can be seen in Figure 1 panel C. Indeed a previous study compared protein targets considering LBMBIA , TBW and ABW (BMI < 20 kg/m 2 adjusted to BMI = 20 kg/m 2 and BMI> 27.5 kg/m 2 adjusted to BMI = 27.5) in 115 hemodialysis patients and concluded that mean protein needs to be estimated by (adjusted) TBW were higher compared to those based on LBMBIA, across all BMI categories (P < 0.01) and most explicitly in obese patients (18) . This overestimation occurred even though a correction factor in grams/kg was advised (LBMBIA * 1.5, whereas (adjusted) TBW * 1.2). A Dutch study comparing protein targets (1.2g * LBM) set by LBMBIA, ABW (BMI < 20 kg/m 2 adjusted to BMI = 20 kg/m 2 and BMI> 30 kg/m 2 adjusted to BMI = 27.5) or TBW in 661 outpatients, showed that adjusted BW estimated LBMBIA correctly (<5% over-or underestimation) in only 33% of their obese patients, whilst LBMTBW estimated between J o u r n a l P r e -p r o o f 1% (obese persons) and 33% (underweight persons) correctly (16) . These reports are in line with our findings that TBW and regression to a BMI of 27.5 severely overestimated LBM and thereby protein requirements.

The same explanation can be offered for the proportional bias seen in LBM22.5. Similar to the LBM27.5 method, this method led to more overestimation in females than males ( figure 1 Panel B) . Underestimation occurred in more males than females, which is likely the result of the difference in the relationship between TBW and LBM in males and females. Forbes described a semilogarithmic relation between LBM and TBW, with slightly different coefficients for men and women (19) . Indeed when we plot TBW and LBM in our cohort (excluding outliers of the mean±2SD), quadratic regression lines for men and women are different, and a common one for both does neither justice ( Figure 3) . Thus, the same is likely the case for LBM equations.

The Gallagher formula and the ESPEN method were the only two LBM equations to acknowledge the difference in body composition between males and females. Although ESPEN offers no reference for their method, the Gallagher formula uses regression models based on DXA studies (14) . As BIA is also validated against DXA, a strong agreement was expected and found (Additional file 2, Table 1 ). In addition, LBMGallagher had the smallest overall 95%-CI. Nevertheless, LBM was often underestimated in women. The previously mentioned Dutch study by Velzeboer et al. (16) found that although LBMGallagher was an improvement over LBMTBW and LBM27.5, protein targets set by LBMGallagher * 1.5g agreed (<5% over-or underestimation) with LBMBIA * 1.2g in only 9% (underweight persons) to 54% (obese persons) of the cases. A possible explanation could be differences in body composition between Gallagher's cohort of (white) British and Northern American volunteers and the Dutch cohorts. Indeed white women in the Gallagher cohort had a BMI of 24.5 ± 4.5 kg/m 2 , compared to 30 (95%-BCa CI 28-32) kg/m 2 in ours. The LBMESPEN method was not subject to fixed bias in males, although gross over-and underestimation were still common and only appeared to cancel each other out around a mean of 0 (Figure 1-2) .

Notably, for the female outlier with an LBM% of 80%, underestimation of LBM occurred in all methods except LBMTBW, alluding to the fact that the studied equations may be even less appropriate for non-sarcopenic obese persons.

J o u r n a l P r e -p r o o f As a real-world exploration of the subject, a secondary aim of this study was to retrospectively compare actual protein provision adequacy in our COVID-19 ICU population based on LBMBIA to that based on LBM predicted by other methods. As such, we found that ICU patients received a mean of 38.7% protein of the local target, or 61.6% (95%-BCa CI 43.2; 80.8) of TargetBIA during ICU admission. This discrepancy shows that our local targets overestimated protein requirements by a third. However, proteins were generally underdelivered by either target.

Our findings align with findings from other studies proving that adequate protein provision is difficult to achieve in the ICU population, including COVID-19 patients (8, 20, 21) . When comparing the percentage of target delivered as calculated by the other methods to TargetBIA, all methods except TargetGallagher for males differed significantly. Therefore, using targets set to LBM based on mathematical methods or TBW is likely to lead to significant over-or underdosing of protein in all other groups. This is in line with findings in other patient categories (16, 18) .

In practice, it has proven difficult to achieve even low-end protein targets in hospitalised COVID-19 patients (8) .

Additionally, there is reason to assume that a high protein diet is beneficial during COVID-19 (3, 4) . Therefore, to guide protein provision we strongly recommend measuring LBMBIA upon hospital admission (as quickly as possible, to prevent bias through hydration shifts).

However, as protein overdoses based on any target have proven less likely to happen than underdosing, we argue that if admission LBMBIA measurements are not feasible, it is probably safer to accept a certain degree of overestimation rather than the underestimation of LBM by formulas. Consequently, our results may argue a preference towards the use of Target22.5, as it had the lowest overestimation with its entire confidence interval above 0 for both sexes in the ICU cohort (Table 3) . Nevertheless, regarding the entire cohort (Figure 1 ), the use of LBM22.5 still led to underestimating LBM in quite a few cases, mostly overweight and obese males. On the other hand, target27.5 and TargetTBW have a confidence interval above 0 for both sexes on the LBM plots of the entire cohort ( Figure 1 ) and regarding targets in the ICU (Table 3) . However, this would mean excepting a mean overestimation of LBM of 23.4 kg or 29.9 kg (Figure 1 ), respectively. It is up to the dietician and clinical to decide whether this is acceptable for their patient.

Although a practical exploration of the subject goes beyond the scope of the current paper, future research could explore the possibility of stratifying methods for estimating LBM according to which works best for which sex/BMI group, if not devising a new universal method based on LBMBIA. Alternatively, the difference between LBM and TBW is sometimes acknowledged through a correction of the amount of protein per kilogram of either (i.e. 1.9g/kg LBM or 1.5g/kg TBW) (15, 16) . However, this correction is based on the assumption of a fixed LBM/TBW ratio, which is an oversimplification that leads to a large error in many individuals (Figure 1 ). Based on our findings we think it is highly unlikely that a static correction such as the one in the example will improve accuracy of protein targets, and we do not recommend its use without further scientific exploration of the subject.

This study was done solely in white Dutch COVID-19 patients and results should be interpreted with caution before the value has been confirmed in other populations.

This research is subject to several limitations. No sample size calculation was performed as the data were dependent on the sample size of the mother study, and not all ICU patients could be included in the protein adequacy analyses. The subsequent relatively small cohort size prevented subdividing into BMI categories for these analyses. Segmenting data could be a point of attention for future studies focusing more specifically on protein provision in the ICU.

The formulas used by the Inbody S10 software to calculate the derived BIA parameters (such as LBM) are not publicly available and therefore cannot be provided here. However, Inbody S10 (LBM) calculations are based on regression formulas derived from reference groups, and have independently been validated against other methods such as Dual-Energy X-ray Absorptiometry in peer-reviewed studies in various populations (22) (23) (24) . Nevertheless, caution is warranted when comparing the results of this study to other populations or BIA devices.

We did not regard underweight persons as a separate category for this study. When regarding BMI 18.5 kg/m 2 as the lower limit of normal weight, the current cohort included three underweight persons (two males with BMI 16 kg/m 2 and 17.3 kg/m 2 , one female with BMI 18 kg/m 2 ), who were grouped in with 30 others in the normal weight category. None of these patients was in the ICU cohort. We do not expect this to have impacted the main findings of this study.

Although we incorporated progressive feeding during the first three days of ICU admission into our targets, accounting for a possibly incomplete first day of admission, we did not account for a possibly incomplete last day. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f 

Obesity and diabetes as high-risk factors for severe coronavirus disease 2019 (Covid-19)

Development and external validation of a prognostic tool for COVID-19 critical disease

Sarcopenia: An underlying treatment target during the COVID-19 pandemic

Evaluation of Nutrition Risk and Its Association With Mortality Risk in Severely and Critically Ill COVID-19 Patients

Nutrition guidelines for critically ill adults admitted with COVID-19: Is there consensus?

Nutrition risk prevalence and nutrition care recommendations for hospitalised and critically-ill patients with COVID-19

Compher C; Society of Critical Care Medicine; American Society for Parenteral and Enteral Nutrition. Guidelines for the Provision and Assessment of Nutrition Support Therapy in the Adult Critically Ill Patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (ASPEN)

Protein provision and lower mortality in critically ill patients with COVID-19

Protein recommendations in the ICU: g protein/kg body weightwhich body weight for underweight and obese patients? Clin Nutr

ESPEN guideline on clinical nutrition in the intensive care unit

Protein for the critically ill patient--what and when?

Bioelectric impedance analysis for body composition measurement and other potential clinical applications in critical illness

Bioelectric impedance body composition and phase angle in relation to 90-day adverse outcome in hospitalized COVID-19 ward and ICU patients: The prospective BIAC-19 study

Healthy percentage body fat ranges: an approach for developing guidelines based on body mass index

Zakboek Dietetiek: compleet herziene uitgave

Hoe berekenen we de eiwitbehoefte bij ondergewicht en overgewicht

Dutch Institute for Healthcare Improvement CBO

Are we overfeeding hemodialysis patients with protein? Exploring an alternative method to estimate protein needs

Forbes Dagger GB Some adventures in body composition, with special reference to nutrition

Protein Provision in Critically Ill Adults Requiring Enteral Nutrition: Are Guidelines Being Met?

The prevalence of iatrogenic underfeeding in the nutritionally 'at-risk' critically ill patient: Results of an international, multicenter, prospective study

The Comparison of Segmental Multifrequency Bioelectrical Impedance Analysis and Dual-Energy X-ray Absorptiometry for Estimating Fat Free Mass and Percentage Body Fat in an Ambulatory Population

Assessment of body composition in peritoneal dialysis patients using bioelectrical impedance and dual-energy x-ray absorptiometry

Evaluation of Body Composition in Hemodialysis Thai Patients: Comparison between Two Models of Bioelectrical Impedance Analyzer and Dual-Energy X-Ray Absorptiometry