| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research Communications |
1 From the Obesity Research Center, St Luke's–Roosevelt Hospital, Columbia University College of Physicians and Surgeons, New York; MRC Human Nutrition Research, Cambridge, United Kingdom; and Kashiwa Hospital, Jikei University, Chiba, Japan.
2 Presented in part at the 1999 European Congress on Obesity, Milan, Italy.
3 Address reprint requests to SB Heymsfield, Weight Control Unit, 1090 Amsterdam Avenue, 14th Floor, New York, NY 10025. E-mail: sbh2{at}columbia.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: The aim of this study was to examine a potential new approach for developing percentage body fat ranges. The approach taken was to link healthy body mass index (BMI; in kg/m2) guidelines established by the National Institutes of Health and the World Health Organization with predicted percentage body fat.
Design: Body fat was measured in subjects from 3 ethnic groups (white, African American, and Asian) who were screened and evaluated at 3 universities [Cambridge (United Kingdom), Columbia (United States), and Jikei (Japan)] with use of reference body-composition methods [4-compartment model (4C) at 2 laboratories and dual-energy X-ray absorptiometry (DXA) at all 3 laboratories]. Percentage body fat prediction equations were developed based on BMI and other independent variables.
Results: A convenient sample of 1626 adults with BMIs 
35 was evaluated. Independent percentage body fat predictor variables in multiple regression models included 1/BMI, sex, age, and ethnic group (R values from 0.74 to 0.92 and SEEs from 2.8 to 5.4% fat). The prediction formulas were then used to prepare provisional healthy percentage body fat ranges based on published BMI limits for underweight (<18.5), overweight (
25), and obesity (30).
Conclusion: This proposed approach and initial findings provide the groundwork and stimulus for establishing international healthy body fat ranges.
Key Words:
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The main assumption of BMI guidelines is that body mass, adjusted for stature squared, is closely associated with body fatness and consequent morbidity and mortality (3, 4). However, some individuals who are overweight are not overfat (eg, bodybuilders). Others have BMIs within the normal range and yet have a high percentage of their body weight as fat. Although these misclassified persons are uncommon relative to the population as a whole (1), the question arises as to how they might be evaluated correctly according to body fatness. Moreover, screening and retention of military recruits (5, 6), policemen, firemen (7), and other workers in whom high fitness levels are required are often based on BMI standards and in some cases on a second-tier body fat evaluation (5).
Unfortunately, there is no consensus on how body fat is linked with morbidity and mortality because of the absence of appropriate prospective studies. Specifically, no accepted published body fat ranges exist; those reported based on empirically set limits, population percentiles, and z scores have serious limitations. Additionally, methods of limited accuracy such as anthropometry are typically used to estimate fatness in population surveys (8).
The aim of the present study was to examine an approach for developing percentage body fat ranges that correspond to published BMI guidelines. Sex-specific formulas were first developed for estimating relative body fatness from BMI and other potential independent variables such as age and ethnicity. These formulas, based on reference methods for evaluating total body fat, were then used to derive percentage body fat levels corresponding to the BMI thresholds for underweight (<18.5), overweight (25), and obesity (30).
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adult subjects with BMIs 35 were evaluated. Subjects were excluded if they had a history of recent acute illness (eg, pneumonia or myocardial infarction), had a chronic condition (eg, cancer, uncontrolled high blood pressure, or collagen vascular disease), or were actively engaged in a vigorous (>6 h/wk) physical activity training program.
Subjects were screened through a medical history questionnaire, physical examination, and measurement of routine blood chemistry indexes. Healthy subjects were enrolled in the study and completed up to 5 evaluations: weight, height, DXA for body fat and bone mineral mass, tritium or deuterium dilution for total body water, and hydrostatic weighing for body density and volume. The measured bone mineral mass, total body water, and body volume values were then used to calculate total body fat by using a 4-compartment model (4C) (9). The study was performed in accord with the Helsinki Declaration of 1975 as revised in 1983.
Experimental design
Three groups of subjects (white, African American, and Asian) were evaluated at MRC Nutrition Research (United Kingdom), Columbia University (United States), and Jikei University (Japan). White subjects were evaluated at both the UK and US sites whereas African American subjects were evaluated at the US site only and all Asians were Japanese evaluated at the Japan site. Body fat was measured by DXA at all 3 centers and, additionally, tritium or deuterium dilution volume, bone mineral mass, and body density were measured at the UK and US sites. Two measures of percentage body fat were used at the UK and US sites, DXA and 4C, whereas DXA alone was used at the Japan site.
Body composition
DXA scanners were used to measure body composition [Japan: DPX-L with software version 1.3z (Lunar Radiation Corp, Madison, WI); United Kingdom: QDR 1000 with enhanced software (Hologic Inc, Bedford, MA); and United States: DPX with software version 3.6 (Lunar Radiation Corp)]. These had CVs of 
1.5% for bone mineral (10) and 3–4% for body fat (11).
Tritium space (3H2O, in L) was measured at the US site with a CV of 1.5% (12). Deuterium space (D2O) was evaluated at the UK site by using infrared spectroscopy with a CV of <1% (13). The tritium and deuterium spaces were then converted into total body water (in kg) with a correction factor for nonaqueous hydrogen exchange and water density at 36°C (total body water = 3H2O or D2O x 0.96 x 0.994) (14).
Body density and volume were measured by underwater weighing in water tanks according to standard methods with a technical error of 0.0020 g/cm3 (15). Residual lung volume was estimated after immersion of subjects in a sitting position by means of the closed-circuit oxygen dilution method in the United States (16) and at the time of immersion by helium dilution in the United Kingdom (9, 17).
Statistical analysis
Body fat estimates made with the 4C method were available from the UK and US sites. The 4C method is generally accepted as a reference method for measuring body fat (17, 18). Body fat estimates by DXA, a second reference method (17, 18), were also available from all 3 sites. The same DXA systems and software were used at the US and Japan sites, facilitating conversion of percentage body fat by DXA in Japanese subjects to a corresponding 4C percentage body fat value. Specifically, the DXA-4C conversion was carried out by using simple regression analysis with 4C percentage body fat as the dependent variable and DXA percentage body fat as the independent variable with use of data collected at the US site. The DXA-4C conversion was intended solely to make Japanese DXA percentage body fat estimates consistent in magnitude with 4C percentage body fat estimates from the UK and US sites. We were therefore able to create, for exploratory purposes, 2 complete sets of operational percentage body fat formulas based on BMI and other potential independent variables: one with DXA percentage body fat as the dependent variable and the other with 4C percentage body fat as the dependent variable.
Models for predicting percentage body fat were developed by using multiple regression analysis with 1/BMI, age, sex, and ethnicity evaluated as potential independent variables. The 1/BMI term was used to linearize the data and to avoid the need for logarithmic conversion or inclusion of power terms (19, 20). Potential interaction terms were explored in model development and a forward-backward stepwise selection procedure was applied for the derivation of prediction equation models. Group data are presented as means ± SDs. All analyses were carried out with the statistical software program SPSS (version 8.0, 1998; SPSS Inc, Chicago).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The total subject pool consisted of 1626 subjects, 1013 women and 613 men. The subject pool included 254 African Americans, 955 Asians, and 417 whites.
|
. Mean age varied from 39.3 y in Asian women to 56.2 y in African American women. African American women had the highest mean weight (71.5 kg) among the women and both male and female Asians had the lowest mean weights within their sex groups. Similarly, Asians were shorter than their white and African American counterparts. BMI was lowest in Asian men and women and highest in African American women (Figure 1).
|
|
for DXA percentage body fat versus BMI in all women. This relation became linear when BMI was replaced by 1/BMI, as shown in the right-hand portion of Figure 2. Although logarithmic transformations and various power terms improved the linearity of and correlation coefficients for the relation between percentage body fat and BMI, we chose for simplicity to use 1/BMI instead as an independent variable in all percentage body fat prediction models developed. Additionally, regression models with 1/BMI provided higher R2 and SEE values than did those with BMI or transformations of BMI as independent variables.
|
. The slope and intercept of this relation differed significantly from 1.0 and 0 (both P < 0.001), respectively. The bias between the 2 methods was relatively small. For example, when 4C body fat was 10%, 30%, and 50%, DXA body fat was 8.9%, 30.1%, and 51.3%, respectively.
|
. The univariate correlations for DXA percentage body fat versus 1/BMI ranged from R = 0.68 to R = 0.89 (all P < 0.001).
|
 |
 |
 |
 |
 |
 |
where multiple R = 0.90 and SEE = 4.31% body fat, sex = 1 for male and 0 for female, and Asian = 1 for Asians and 0 for the other races. The model predicted higher percentage body fat in men and in older subjects. The model did not show a significant difference between African Americans and whites for either sex. Asians, however, according to the model, had a significantly higher percentage body fat for any given BMI than did the other 2 ethnic groups.
Separate analyses for DXA percentage body fat with use of 1/BMI and age by sex and ethnicity were also carried out, yielding multiple correlations (ie, R values) ranging from 0.74 to 0.88 and SEEs from 3.8% to 5.4% body fat. These models provided results similar to those for Equation 1
and are not reported herein. Solutions for DXA equations at the 3 BMI levels are not presented because these can be calculated directly from Equation 1. The patterns of estimated percentage body fat ranges are similar to those presented below for the 4C method.
Four-compartment models
We calculated a 4C percentage body fat value for each Asian subject by using the following equation to convert DXA to 4C percentage body fat, which was derived by simple regression of 4C percentage body fat on DXA percentage body fat from measurements taken at the US site:
 |
where multiple R = 0.92 and SEE = 3.08% fat, and
 |
where multiple R = 0.92 and SEE = 2.79% fat.
The forward-backward stepwise selection produced the following equation for the 4C percentage body fat estimates:
 |
 |
 |
 |
 |
where multiple R = 0.89 and SEE = 3.97% fat, sex = 1 for male and 0 for female, Asian = 1 for Asians and 0 for the other races, and African American = 1 for African Americans and 0 for the other races.
The corresponding 4C percentage body fat ranges are presented in Table 3. According to this model, Asians had higher percentages of body fat at lower BMIs, particularly at younger ages, than did the other 2 ethnic groups. These developed ranges also indicate a small difference between African Americans and whites for males and females, with African American subjects having lower percentages of body fat, particularly at older ages. The difference between African Americans and whites remained significant even when the analysis was repeated for the US data alone.
|
), the magnitude of this effect was rather small (1–2% fat). For practical purposes, we therefore developed another set of simplified equations by combining the data from white and African American subjects for 4C percentage body fat, as follows:
 |
 |
 |
where multiple R = 0.86 and SEE = 4.98% fat and sex = 1 for male and 0 for female. The resulting estimated 4C percentage body fat ranges by sex and age are presented in Table 4 and the presented values consolidate the 4C studies in African American and white subjects into a single chart.
|
 |
where n = 322, multiple R = 0.88, and SEE = 2.91% fat.
 |
where n = 633, multiple R = 0.77, and SEE = 3.49% fat. The resulting estimated 4C percentage body fat ranges, converted from those derived by DXA, are presented by sex and age in Table 5; these values differ slightly from the Asian values summarized in Table 3.
|
. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Model development
BMI transformation
As shown in this study and others (19, 20), the function percentage fat versus BMI is curvilinear and the best fit is often accomplished by using power functions or logarithmic conversions, adding complexity to developed formulas. In creating our prediction models for percentage body fat, we applied the inverse of BMI, 1/BMI, as the main predictor variable. This approach improved the linearity of the association between percentage body fat and BMI and simplified model development. Our observations combined with that of earlier investigations (20) suggest a utility in prediction model development for 1/BMI when subject populations include a wide range of body fatness values.
Population specificity
The prediction models developed for percentage body fat had significant age and ethnicity terms as independent variables. This highlights a critical concern regarding population specificity with these and similar empirical prediction models. Our population included adults aged 97 y. All of our developed models indicate that, after 1/BMI is first controlled for, greater age is associated with a higher percentage body fat. Many cross-sectional and longitudinal studies now indicate that relative fatness in adults increases with age (21–31). Although the mechanisms leading to increasing fatness with age are not fully understood, our findings and those of others suggest that sex, ethnic, and individual differences exist in the rate at which percentage body fat changes with senescence (25, 30, 31). Some of our cross-sectional prediction models for percentage body fat had significant age-by-sex interaction terms (eg, Equation 1), usually in the direction of a greater relative increase in percentage fat with older age in men than in women. An important and as yet unanswered question is whether the greater fatness with older age, even after BMI is first controlled for, poses additional health risks. This important question, highlighted by the present investigation, needs to be examined in future mechanistic and clinical studies.
In addition to an age effect on BMI-predicted percentage body fat, we observed an independent effect of ethnicity or study site. Others made similar observations when studying various ethnic groups at the same or different evaluation sites (32–36). For example, Deurenberg et al (35) found that American blacks had a 1.3-unit lower and Polynesians a 4.5-unit higher BMI than whites with the same body fatness (35). Even within the white cohort, the investigators observed small differences between Americans and Europeans.
The underlying causes of ethnic variation in relations between BMI and percentage body fat are likely due to small between-center body fat measurement differences and biological between-group differences (35). The evaluation of Asians was confined to Japan and that of African Americans to the United States. Therefore, the underlying causes of observed ethnic differences in terms of measurement, environmental, and genetic factors are difficult to ascertain. Nevertheless, it appears evident that a single set of universal percentage body fat ranges cannot be easily developed without considerable additional analysis of this problem. Our equations and associated tables provide several ethnic-specific ranges as working guidelines. Because African Americans and whites differed only slightly in percentage body fat (by 1–2%) after BMI was first controlled for, we presented a combined equation (Equation 5) and table (Table 4) based on 4C percentage body fat for these 2 groups.
Methodologic differences
An important issue highlighted by the present study is between-method and instrument differences in body fat measurement. Although the various DXA and 4C models we developed provided similar body fat ranges, some notable differences were found (eg, significant African American term in 4C Equation 4 and no comparable term in DXA Equation 1). Moreover, as might be expected between any 2 measurement methods, there was a small but statistically significant bias for DXA versus 4C percentage body fat. Ideally, the identical calibrated measurement system would have been used at all 3 sites for body fat measurement. This optimum situation was not possible in the present study and it is likely that any large-scale international study would face similar methodologic issues.
Study limitations
Our subjects, by necessity, were a convenient sample and may not be representative of the populations from which they were recruited. We excluded subjects with recent weight change, those who had acute or chronic illnesses, and those engaged in physical training programs. Moreover, the aim of this study was not to provide population ranges for body fatness as might be a goal of epidemiologic studies. Ideally, however, future prospective studies should consider optimum sampling strategies when prediction formulas for percentage body fat are developed based on BMI.
A second concern involves subjects at or below the lower BMI limit of 18.5. Males with BMIs <18.5 who are otherwise healthy are relatively uncommon in industrialized nations. Only 2.6% of our male subjects and 6.0% of our female subjects had a BMI <18.5 (Figure 1). Hence, our lower BMI percentage body fat ranges, by necessity, have large CIs and should be applied cautiously. Moreover, the limited number of subjects with low BMI values suggests that very large subject samples are needed in future studies for model development or that subject evaluations include much leaner populations than those evaluated in the present 3-country study.
Last, our assumption is that percentage body fat is an improved phenotypic characteristic over BMI when functionality and mortality risk are considered. This hypothesis has some support (37), although additional studies with appropriate methods are needed to fully explore the range of issues surrounding body composition compared with body weight as "ponderal" measures. Moreover, visceral adiposity, separate from body composition or weight, is an independent predictor of morbidity and mortality (1) and development of improved clinical quantification methods remains a high priority.
Conclusion
This study presents a working approach to developing body fat ranges by linking current BMI guidelines with predicted percentage body fat. The developed provisional equations and tables in this report can fill an information gap because no comparable percentage body fat ranges exist for review for evaluation of potentially misclassified subjects referred for body-composition analysis. Our effort highlights important issues for future consideration, such as the appropriateness of increasing fatness with aging even when BMI remains constant, the causes of country or ethnicity differences in BMI–percentage body fat relations, whether misclassified subjects are more or less healthy than their counterparts with similar BMIs, and how to develop appropriate sampling strategies to prospectively develop percentage body fat ranges. These important issues require discussion, debate, and future investigation.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. Later, A. Bosy-Westphal, B. Hitze, E. Kossel, C-C Gluer, M. Heller, and M. J Muller No evidence of mass dependency of specific organ metabolic rate in healthy humans Am. J. Clinical Nutrition, October 1, 2008; 88(4): 1004 - 1009. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-E. Yim, S. Heshka, J. B. Albu, S. Heymsfield, and D. Gallagher Femoral-gluteal subcutaneous and intermuscular adipose tissues have independent and opposing relationships with CVD risk J Appl Physiol, March 1, 2008; 104(3): 700 - 707. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Nooyens, L. L. Koppes, T. L. Visscher, J. W. Twisk, H. C. Kemper, A J. Schuit, W. van Mechelen, and J. C Seidell Adolescent skinfold thickness is a better predictor of high body fatness in adults than is body mass index: the Amsterdam Growth and Health Longitudinal Study Am. J. Clinical Nutrition, June 1, 2007; 85(6): 1533 - 1539. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Fontana and S. Klein Aging, Adiposity, and Calorie Restriction JAMA, March 7, 2007; 297(9): 986 - 994. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Stevenson and W. A. Woods Jr Condition indices for conservation: new uses for evolving tools Integr. Comp. Biol., December 1, 2006; 46(6): 1169 - 1190. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-f. Guo, H. Shen, Y.-j. Liu, W. Wang, D.-h. Xiong, P. Xiao, Y.-z. Liu, L.-j. Zhao, R. R. Recker, and H.-w. Deng Assessment of Genetic Linkage and Parent-of-Origin Effects on Obesity J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 4001 - 4005. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Larsson, B. Henning, A. K. Lindroos, I. Naslund, C. D. Sjostrom, and L. Sjostrom Optimized predictions of absolute and relative amounts of body fat from weight, height, other anthropometric predictors, and age 1 Am. J. Clinical Nutrition, February 1, 2006; 83(2): 252 - 259. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Cesari, S. B Kritchevsky, R. N Baumgartner, H. H Atkinson, B. W. Penninx, L. Lenchik, S. L Palla, W. T Ambrosius, R. P Tracy, and M. Pahor Sarcopenia, obesity, and inflammation--results from the Trial of Angiotensin Converting Enzyme Inhibition and Novel Cardiovascular Risk Factors study Am. J. Clinical Nutrition, August 1, 2005; 82(2): 428 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bosy-Westphal, S. Danielzik, R.-P. Dorhofer, A. Piccoli, and M. J Muller Patterns of bioelectrical impedance vector distribution by body mass index and age: implications for body-composition analysis Am. J. Clinical Nutrition, July 1, 2005; 82(1): 60 - 68. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Jarfelt, B Lannering, I Bosaeus, G Johannsson, and R Bjarnason Body composition in young adult survivors of childhood acute lymphoblastic leukaemia Eur. J. Endocrinol., July 1, 2005; 153(1): 81 - 89. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Chung, M.-Y. Song, H.-D. Shin, D.-Y. Kim, Q. He, S. Heshka, J. Wang, J. Thornton, B. Laferrere, F. X. Pi-Sunyer, et al. Korean and Caucasian overweight premenopausal women have different relationship of body mass index to percent body fat with age J Appl Physiol, July 1, 2005; 99(1): 103 - 107. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Young, L. Morbois-Trabut, B. Couzinet, O. Lascols, E. Dion, V. Bereziat, B. Feve, I. Richard, J. Capeau, P. Chanson, et al. Type A Insulin Resistance Syndrome Revealing a Novel Lamin A Mutation Diabetes, June 1, 2005; 54(6): 1873 - 1878. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Klein, L. E. Burke, G. A. Bray, S. Blair, D. B. Allison, X. Pi-Sunyer, Y. Hong, and R. H. Eckel Clinical Implications of Obesity With Specific Focus on Cardiovascular Disease: A Statement for Professionals From the American Heart Association Council on Nutrition, Physical Activity, and Metabolism: Endorsed by the American College of Cardiology Foundation Circulation, November 2, 2004; 110(18): 2952 - 2967. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P Wildman, D. Gu, K. Reynolds, X. Duan, and J. He Appropriate body mass index and waist circumference cutoffs for categorization of overweight and central adiposity among Chinese adults Am. J. Clinical Nutrition, November 1, 2004; 80(5): 1129 - 1136. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. B. Sopher, J. C. Thornton, J. Wang, R. N. Pierson Jr, S. B. Heymsfield, and M. Horlick Measurement of Percentage of Body Fat in 411 Children and Adolescents: A Comparison of Dual-Energy X-Ray Absorptiometry With a Four-Compartment Model Pediatrics, May 1, 2004; 113(5): 1285 - 1290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. L Blackburn National Health and Nutrition Examination Survey: where nutrition meets medicine for the benefit of health Am. J. Clinical Nutrition, August 1, 2003; 78(2): 197 - 198. [Full Text] [PDF]
|
|
|
|
|
|
S. Zhu, Z. Wang, W. Shen, S. B Heymsfield, and S. Heshka Percentage body fat ranges associated with metabolic syndrome risk: results based on the third National Health and Nutrition Examination Survey (1988-1994) Am. J. Clinical Nutrition, August 1, 2003; 78(2): 228 - 235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Savage, G. D. Tan, C. L. Acerini, S. A. Jebb, M. Agostini, M. Gurnell, R. L. Williams, A. M. Umpleby, E. L. Thomas, J. D. Bell, et al. Human Metabolic Syndrome Resulting From Dominant-Negative Mutations in the Nuclear Receptor Peroxisome Proliferator-Activated Receptor-{gamma} Diabetes, April 1, 2003; 52(4): 910 - 917. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B Racette, S. S Deusinger, and R. H Deusinger Obesity: Overview of Prevalence, Etiology, and Treatment Physical Therapy, March 1, 2003; 83(3): 276 - 288. [Full Text] [PDF]
|
|
|
|
|
|
G. E. van der Ploeg, R. T. Withers, and J. Laforgia Percent body fat via DEXA: comparison with a four-compartment model J Appl Physiol, February 1, 2003; 94(2): 499 - 506. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. J. Moran, M. Noakes, P. M. Clifton, L. Tomlinson, and R. J. Norman Dietary Composition in Restoring Reproductive and Metabolic Physiology in Overweight Women with Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., February 1, 2003; 88(2): 812 - 819. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R Fernandez, M. Heo, S. B Heymsfield, R. N Pierson Jr, F X. Pi-Sunyer, Z. M Wang, J. Wang, M. Hayes, D. B Allison, and D. Gallagher Is percentage body fat differentially related to body mass index in Hispanic Americans, African Americans, and European Americans? Am. J. Clinical Nutrition, January 1, 2003; 77(1): 71 - 75. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W Taylor, I. E Jones, S. M Williams, and A. Goulding Body fat percentages measured by dual-energy X-ray absorptiometry corresponding to recently recommended body mass index cutoffs for overweight and obesity in children and adolescents aged 3-18 y Am. J. Clinical Nutrition, December 1, 2002; 76(6): 1416 - 1421. [Abstract] [Full Text] [PDF]
|
|
|
, African American; 
, Asian.
