A Comparison between Composition and Density of the Fat-Free Mass of Young Adult Singaporean Chinese and Dutch Caucasians

Normative values in healthy adults for the 6-minute and 2-minute walk tests in Belgium and Vietnam: implications for clinical practice

OBJECTIVE

To establish reference values for the 6-minute walk test (6MWT) and 2-minute walk test (2MWT) distances, to investigate the correlation between these 2 tests, and to establish prediction equations for these distances in healthy populations of Belgium and Vietnam.

DESIGN

Cross-sectional study.

SUBJECTS AND METHODS

The 6MWT and 2MWT were administered to a convenience sample of 239 Belgian and 303 Vietnamese participants between the ages of 18 and 80 years.

RESULTS

The mean (standard deviation; SD) 2MWT distances were 215 (32.8) m for Belgian participants and 156 (25.5) m for Vietnamese participants. The mean (SD) 6MWT distances were 625 (90.7) m for Belgian participants and 449 (70.4) m for Vietnamese participants. The Pearson correlation coefficient between the 2 tests was 0.901 (p < 0.001) for Belgian participants and 0.871 (p < 0.001) for Vietnamese participants. Age and sex were the 2 most important predictors of walking distance, followed by body mass index for Belgium and height for Vietnam. The adjusted R² ranged from 0.31 to 0.49 across 4 predictive equations.

CONCLUSION

These results can be used to determine the presence of walking performance deficits and to guide future studies. The 2MWT is suggested as a useful and convenient alternative to the 6MWT for assessing walking performance in clinical practice.

Predictive equations for fat mass in older Hispanic adults with excess adiposity using the 4-compartment model as a reference method

BACKGROUND

Predictive equations are the best option for assessing fat mass in clinical practice due to their low cost and practicality. However, several factors, such as age, excess adiposity, and ethnicity can compromise the accuracy of the equations reported to date in the literature.

OBJECTIVE

To develop and validate two predictive equations for estimating fat mass: one based exclusively on anthropometric variables, the other combining anthropometric and bioelectrical impedance variables using the 4C model as the reference method.

SUBJECTS/METHODS

This is a cross-sectional study that included 386 Hispanic subjects aged ≥60 with excess adiposity. Fat mass and fat-free mass were measured by the 4C model as predictive variables. Age, sex, and certain anthropometric and bioelectrical impedance data were considered as potential predictor variables. To develop and to validate the equations, the multiple linear regression analysis, and cross-validation protocol were applied.

RESULTS

Equation 1 included weight, sex, and BMI as predictor variables, while equation 2 considered sex, weight, height squared/resistance, and resistance as predictor variables. R2 and RMSE values were ≥0.79 and ≤3.45, respectively, in both equations. The differences in estimates of fat mass by equations 1 and 2 were 0.34 kg and -0.25 kg, respectively, compared to the 4C model. This bias was not significant (p < 0.05).

CONCLUSIONS

The new predictive equations are reliable for estimating body composition and are interchangeable with the 4C model. Thus, they can be used in epidemiological and clinical studies, as well as in clinical practice, to estimate body composition in older Hispanic adults with excess adiposity.

Determination of Fat-Free Mass Density and its Components in Older Hispanic Adults by In Vivo Methods

The densitometry method estimates body composition based on cadaver reference values, mainly the fat-free mass density value of 1.100 g/cm³. However, several changes in fat-free mass components by aging, ethnicity, and excess adiposity could influence their density and affect body composition estimations. The present study aimed to compare the mean fat-free mass component values in older Hispanic adults to cadaver reference values. This cross-sectional study included a sample of 420 subjects aged ≥60 yr from northern Mexico. Fat-free mass was determined by the four-compartment model using air displacement plethysmography, the deuterium dilution technique, and dual-energy X-ray absorptiometry for body density, aqueous and mineral fractions of body weight, respectively. A 1-sample t test was used to compare the fat-free mass density and aqueous, mineral, and residue fractions of fat-free mass from subjects in the study to the assumed cadaver reference values. The mean fat-free mass density value for the total sample of older Hispanic adults (1.096 ± 0.011 g/cm³) was significantly (p < 0.001) lower than the assumed value of 1.100 g/cm³, except in obese older men. The mean aqueous fraction of fat-free mass (74.8 ± 3.3%) was higher than the assumed value of 73.8%, and the mean residue fraction of fat-free mass value was lower (18.3 ± 3.4%) than the reference value of 19.4%. Indeed, only the mean mineral fraction of fat-free mass value (6.8 ± 0.8%) was similar to the reference value. In the total sample, all characteristic mean fat-free mass values in these older Hispanic adults differed from cadaver reference values, except the mineral fraction of fat-free mass value.

Examining Race-Related Error in Two-Compartment Models of Body Composition Assessment: A Systematic Review and Meta-Analysis

BACKGROUND

The Brozek and Siri formulas estimate relative adiposity (%Fat) from total body density (D_b) using a 2-compartment (2C) model. Racial/ethnic differences in D_b have been reported, along with subsequent errors in estimated %Fat.

OBJECTIVE

The primary aim of this systematic review and meta-analysis was to examine potential race/ethnic differences in the accuracy of the Brozek and Siri 2C formulas using aggregate-level data.

METHODS

Peer-reviewed studies available in English that provided 2C and 4C estimates of %Fat were located using searches of the PubMed (n = 150), Scopus (n = 170), and Web of Science (n = 138) online electronic databases. Random-effects models were used to determine potential differences between racial groups using a mean ES and 95% confidence intervals.

RESULTS

The cumulative results from 78 effects indicate that the relative accuracy of the Brozek equation did not vary between racial groups (between group p = 0.053). In contrast, the Siri equation slightly underestimated %Fat for Asian adults (ES_WMD = -1.40%, 95%CI -2.33% to -0.46%; p = 0.004) and Black adults (ES_WMD = -1.10%, 95%CI -2.11% to -0.08%; p = 0.034), with no significant differences observed in Hispanic adults (ES_WMD = 0.64%, 95%CI -1.02% to 2.31%; p = 0.448) and White adults (ES_WMD = 0.08%, 95%CI -0.42% to 0.57%; p = 0.766) (between group p = 0.019).

CONCLUSION

Small, but statistically significant, error was found between racial groups when estimating %Fat using the 2C Siri equation when compared to 4C models. However, the observed error due to race/ethnicity appears to be of little clinical or practical significance when using either equation.

Differences in Fat Mass Estimation Formulas in Physically Active Adult Population and Relationship with Sums of Skinfolds

Changes in body composition and specifically fat mass, has traditionally been used as a way to monitor the changes produced by nutrition and training. The objective of the present study was to analyse the differences between the formulas used to estimate fat mass and to establish the existing relationship with the body mass index and sums of skinfolds measurement in kinanthropometry. A total of 2458 active adults participated in the study. Body mass index (BMI) and skinfolds were measured, and the Kerr, Durnin-Womersley, Faulkner and Carter equations were used to assess fat mass. Significant differences were found between all the formulas for the percentage of fat mass, ranging from 10.70 ± 2.48 to 28.43 ± 5.99% (p < 0.001) and fat mass from 7.56 ± 2.13 to 19.89 ± 4.24 kg (p < 0.001). The correlations among sums of skinfolds and the different equations were positive, high and significant in all the cases (r from 0.705 to 0.926 p < 0.001), unlike in the case of BMI, were the correlation was lower and both positive or negative (r from -0.271 to 0.719; p < 0.001). In conclusion, there were differences between all the formulas used to estimate fat mass; thus, for the evaluation of fat mass with kinanthropometry of an active adult, the use of the same formula is recommended on all occasions when the results are going to be compared or when an athlete is compared with a reference.

Validity of Deuterium Oxide Dilution for the Measurement of Body Fat among Singaporeans

Body fat percent (BF%) was measured in 108 adult Chinese, 76 Malays, and 107 Indians in Singapore by densitometry, deuterium oxide dilution (hydrometry), dual energy x-ray absorptiometry (DXA) and a chemical four-compartment model (BF%4c). The hydration of the fat-free mass (FFM) was calculated. Subjects ranged in age from 18 to 69 years and their body mass index ranged from 16 to 40 kg/m2. BF%4c for the various subgroups were: Chinese females (33.5 ± 7.5%), Chinese males (24.4 ± 6.1%), Malay females (37.8 ± 6.3%), Malay males (26.0 ± 7.6%), Indian females (38.2 ± 7.0%) and Indian males (28.1 ± 5.5%). Biases were found between BF%4c and BF% measured by 2-compartment models (hydrometry, densitometry, DXA), with systematic underestimation by DXA and densitometry. On a group level hydrometry had the lowest bias while DXA gave the highest bias. When validated against BF%4c, 2-compartment models were found to be unsuitable for accurate measures of body fat due to high biases at the individual level and the violation of assumptions of constant hydration of FFM and density FFM among the ethnic groups. On a group level the best 2-compartment model for measuring body fat was found to be hydrometry.

Relationships Among C-Reactive Protein, Alanine Aminotransferase, and Metabolic Syndrome in Apparently Healthy Chinese Subjects

AIM

Elevated levels of C-reactive protein (CRP) and alanine aminotransferase (ALT) are tightly associated with metabolic syndrome (MetS). However, it is unclear whether the combination of normal CRP and ALT increases probability of the presence of MetS.

MATERIALS AND METHODS

We enrolled 433 Chinese in health screening with both CRP and ALT in normal range. They were divided into four groups: both low (BL, both low baseline CRP and ALT), only high CRP (HCRP), only high ALT (HALT), and both high (BH, both high baseline CRP and ALT) in different genders. The receiver operating characteristic curve was used for comparing the sensitivity and assessing the cutoff point.

RESULTS

The BH group had more numbers of MetS components than the BL group. Addition of CRP and ALT did not further increase the accuracy of predicting MetS with a sensitivity of 65.1% and a specificity of 59.1% in men, and a sensitivity of 77.8% and a specificity of 60.2% in women. The area under curve for ALT was greater than that for CRP (0.62 vs. 0.524 in men; 0.663 vs. 0.598 in women). After adjusting for age, ALT could significantly predict the incidental MetS in both genders; nevertheless, CRP failed to demonstrate prediction. The formula of probability of occurrence of MetS was established on CRP and ALT.

CONCLUSIONS

Healthy subjects with high baseline ALT may be at risk of developing MetS, and early recognition and prevention are important for health providers. We concluded that local liver inflammation is significantly associated with the presence of MetS compared with systemic inflammation in health screening subjects.

Estimating Body Fat from Anthropometry and Isotopic Dilution: A Four-Country Comparison

OBJECTIVE

The goal was to assess the ability of BMI to predict body fat (BF) among youths in four countries and identify the degree to which additional anthropometric measures improve this prediction. BMI is widely recommended as an indicator of overweight. However, whether BMI adequately estimates BF and has the same meaning in different ethnic groups and youths has been questioned.

RESEARCH METHODS AND PROCEDURES

Data come from 456 Filipino, Chinese, Russian, and black South African youths, 6 to 16 years old. Percentage BF and fat mass index (FMI) were estimated by the deuterium dilution method. Skinfold thicknesses (triceps, subscapular, and suprailiac) and weight and height measures were collected. Percentage BF was regressed first on BMI and age and then with the addition of the skinfold measures. Linear models were run separately by country and sex. The models were repeated with FMI as the outcome.

RESULTS

The R2 values from the percentage BF models ranged from 0.13 to 0.69 in the first models to 0.38 to 0.81 in the full models. The values were lowest among Russian males > or = 13 years and Russian females > or = 13 years of age in the reduced and full models, respectively, and were highest among Chinese females. Using FMI as the outcome did not meaningfully change the results.

DISCUSSION

The ability of BMI to adequately predict BF and the additional predictivity of anthropometric measures varied widely across the samples, making its uniform use as a proxy for BF in youths from different countries questionable.

Comparisons of body-composition prediction accuracy: a study of 2 bioelectric impedance consumer devices in healthy Chinese persons using DXA and MRI as criteria methods

We compared the accuracy of body-composition estimation for 2 commercial single-frequency bioelectric impedance analysis (BIA) devices in 200 healthy Chinese adults using magnetic resonance imaging (MRI) and dual-energy X-ray absorptiometry (DXA) as criterion methods. We evaluated the fat mass percentage (%FM), skeletal muscle mass percentage (%SM), or total-body bone-free lean mass percentage (%TBBLM), and level of visceral fat mass (VF(level)) using the Omron HBF-359 (SF-BIA8) and Tanita BC-532 (SF-BIA4) BIA devices, MRI, and DXA. Both devices showed a similarly high correlation with DXA for %FM prediction (r=0.89 for SF-BIA8 and 0.90 for SF-BIA4) and with MRI and DXA for %SM and %TBBLM prediction (r=0.85 for SF-BIA8 and 0.89 for SF-BIA4). There were small but significant biases in all body-composition parameter evaluations except for %SM assessed by the SF-BIA8. Both the SF-BIA8 and SF-BIA4 provided small, insignificant mean biases but wide limits of agreement with MRI for VF(level) assessments. Both BIA devices can relatively accurately predict %FM and %SM in healthy Chinese adults. The SF-BIA8 is suitable for individual prediction of %SM, whereas the SF-BIA8 is required to eliminate systematic errors in this population by improving population-specific prediction equations from height, weight, and age to increase estimation accuracy.

Predicting fat percent by skinfolds in racial groups: Durnin and Womersley revisited

PURPOSE

Despite their widespread use in research and fitness settings, Durnin and Womersley's (DW) 1974 prediction equations using skinfold thickness to estimate body fat percent by hydrodensitometry have not been systematically evaluated in racial or ethnic groups using body fat percent measured by dual-energy x-ray absorptiometry (%BF(DXA)) as the standard.

METHODS

This cross-sectional, population-based study examined whether the DW skinfold equations predict %BF(DXA) in a large, multiracial sample. Four skinfold measures (biceps, triceps, subscapular, and suprailiac), other clinical anthropometrics, and %BF(DXA) were obtained from 1675 healthy adults, age 18-110 yr, who were classified into four racial or ethnic categories: Caucasian, African American, Hispanic, or Asian. Predicted body fat percent using DW equations was compared with %BF(DXA) and evaluated within race/ethnicity- and sex-specific groups.

RESULTS

Mean body fat percent predicted by DW equations was significantly different from %BF(DXA) in four of eight race/ethnicity- and sex-specific groups, particularly in Asian women and African American men (3.3 and 2.4 percentage point overestimates, respectively, P < 0.0001). New linear regression equations were developed estimating %BF(DXA) specific to each race/ethnicity and sex group, using the original DW skinfold sites. Body weight, height, and waist circumference independently predicted fat percent and were also included in the new equations.

CONCLUSIONS

The 1974 DW equations did not predict %BF(DXA) uniformly in all races or ethnicities. Using %BF(DXA) as the criterion measure, the original DW skinfold equations have been updated specific to sex and race/ethnicity while maintaining the DW options for a minimalistic model using fewer predictors.

Body composition in childhood: effects of normal growth and disease

Body composition in children is of increasing interest within the contexts of childhood obesity, clinical management of patients and nutritional programming as a pathway to adult disease. Energy imbalance appears to be common in many disease states; however, body composition is not routinely measured in patients. Traditionally, clinical interest has focused on growth or nutritional status, whereas more recent studies have quantified fat mass and lean mass. The human body changes in proportions and chemical composition during childhood and adolescence. Most of the weight gain comprises lean mass rather than fat. In general, interest has focused on percentage fat, and less attention has been paid to the way in which lean mass varies within and between individuals. In the general population secular trends in BMI have been widely reported, indicating increasing levels of childhood obesity, which have been linked to reduced physical activity. However, lower activity levels may potentially lead not only to increased fatness, but also to reduced lean mass. This issue merits further investigation. Diseases have multiple effects on body composition and may influence fat-free mass and/or fat mass. In some diseases both components change in the same direction, whereas in other diseases, the changes are contradictory and may be concealed by relatively normal weight. Improved techniques are required for clinical evaluations. Both higher fatness and reduced lean mass may represent pathways to an increased risk of adult disease.

Differences in the relationship between BMI and percentage body fat between Japanese and Australian-Caucasian young men

This cross-sectional study aimed to determine ethnic and environmental influences on the relationship between BMI and percentage body fat, using a sample of 144 Japanese and 140 Australian-Caucasian men living in Australia, and eighty-eight Japanese men living in Japan. Body composition was assessed by anthropometry using standard international methods (International Society for the Advancement of Kinanthropometry protocol). Body density was predicted using Durnin and Womersley's (1974) equation, and percentage body fat was calculated from Siri's (1961) equation.Significant (P<0·05) ethnic differences in stature, body mass and BMI were observed between Japanese and Australian men, but no ethnic differences were observed in their percentage body fat and height-corrected sum of skinfold thicknesses. No differences were found in the BMI–percentage body fat relationship between the Japanese subjects living in Australia and in Japan. Significant (P<0·05) ethnic differences in the BMI–percentage body fat relationship observed from a comparison between pooled Japanese men(aged 18–40 years, BMI range 16·6–32·8kg/m2) andAustralians (aged 18–39 years, BMI range 16·1–31·4kg/m2) suggest that Japanese men are likely to havea greater percentage body fat than Australian men at any given BMI value.From the analyses, the Japanese men were estimated to have an equivalent amount of body fat to the Australian men at BMI values that were about 1·5 units lower than those of the Australians (23·5kg/m2 and 28·2kg/m2, respectively).It was concluded that Japanese men have greater body fat deposition than Australian-Caucasians at the same BMI value. Japanese men may therefore require lower BMI cut-off points to identify obese individuals compared with Australian-Caucasian men.

Body composition predicted from skinfolds in African women: a cross-validation study using air-displacement plethysmography and a black-specific equation

Skinfold thickness (SF) measurements are commonly used for the indirect assessment of body composition. It is necessary to know how large the bias is when using Caucasian SF-based prediction equations Africans, as no specific equations exist. Our first aim was to test the validity of the equation of Durnin & Womersley for predicting body density from SF in Africans. The second aim was to determine the effect of calculating percentage body fat (%BF) from body density using a black-specific formula rather than the Siri equation, thus taking into account the higher fat-free mass (FFM) density in blacks than in whites. A total of 196 African women volunteered. Mean age was 29·5 (sd 8·7) years and mean BMI was 22·5 (sd 4·6) kg/m2. We compared body density values predicted from SF with those measured by air-displacement plethysmography, and %BF values obtained from body density using the Siri equation or the black-specific calculation. The bias (reference minus prediction) was 0·0100 kg/cm3 in body density (P<10−4) and 6·5 % BF (P<10−4), and the error (sd of the bias) 0·0097 kg/l and 4·5 % BF. With the black-specific equation, the bias was reduced by 1·9 % BF, while error remained similar. As the %BF prediction required an SF-based equation followed by a body density-based calculation, the lack of validity we observed in Africans may be due to known differences between blacks and whites in the distribution of subcutaneous adipose tissue and, as demonstrated, in the FFM density. Equations thus need to be established using SF values specific to Africans.

Laboratory and field measurements of body composition

Objective

This background paper was prepared in response to a request to review the concepts related to measurement of body composition, to discuss laboratory and field methods of assessing body composition and to discuss the practical applications of the methods – how they might be used singly or in combination to provide data for a selected population.

Design

The common laboratory and field methods are described and discussed, with particular attention to the assumptions involved and the applicability of the methods to the different population groups. Most measurements of body composition are made in the field, at the bedside or clinic as opposed to in the laboratory. The laboratory methods have a role to play in their own right, in research into new concepts, models and methods. However, they are particularly important in establishing the accuracy of the field methods.

Setting

Field, bedside and laboratory studies.

Subjects

Children, adults, the elderly, ethnic groups.

Results

Laboratory estimates of body compositions are best performed by multi-component methods or by 2-component methods adjusted for to the populations under investigation. There is a scarcity of data for most of the populations in the world.

Conclusions

Energy requirements based on body weight are an approximation since they do not take into account differences in body composition, which will better determine the true requirements. The measurement of body composition occurs in many branches of biology and medicine. It is used in the assessment of nutritional and growth status and in disease states and their treatment. Energy stores, skeletal muscle and protein content can be determined and changes monitored. In human energetics, body composition is widely used for the standardisation of other variables, such as basal metabolic rate (BMR), in the assessments of ethnic and environmental differences and of variability and adaptation to different levels of nutrition. Choosing a method is very problematic. Users want simple, inexpensive, rapid, safe accurate methods to measure body composition but speed and simplicity come at the expense of accuracy. Recommendations are made for age, sex, and in some cases, fatness and ethnic specific methods.

Development of methods for body composition studies

This review is focused on experimental methods for determination of the composition of the human body, its organs and tissues. It summarizes the development and current status of fat determinations from body density, total body water determinations through the dilution technique, whole and partial body potassium measurements for body cell mass estimates, in vivo neutron activation analysis for body protein measurements, dual-energy absorptiometry (DEXA), computed tomography (CT) and magnetic resonance imaging (MRI, fMRI) and spectroscopy (MRS) for body composition studies on tissue and organ levels, as well as single- and multiple-frequency bioimpedance (BIA) and anthropometry as simple easily available methods. Methods for trace element analysis in vivo are also described. Using this wide range of measurement methods, together with gradually improved body composition models, it is now possible to quantify a number of body components and follow their changes in health and disease.

Six-minute walk distance in healthy Singaporean adults cannot be predicted using reference equations derived from Caucasian populations

OBJECTIVES

The 6-min walk test (6MWT) is commonly used to assess the functional exercise capacity of individuals with cardiopulmonary disease. Recent studies have established regression equations to predict the 6-min walk distance (6MWD) in healthy Caucasian populations; however, regression equations have yet to be established for the Singaporean population. The aim of this study was to determine 6MWD in healthy Singaporeans and identify contributors to 6MWD in this population. We also compared measured 6MWD with predicted 6MWD from two regression equations derived in Caucasian subjects.

METHODOLOGY

Thirty-five healthy subjects (32 Chinese, 16 men) aged between 45 and 85 years performed three walking tests using a standardized protocol. 6MWD was defined as the greatest distance achieved from the three tests. Heart rate (HR) was recorded each minute during the 6MWT. Other measurements included age, height, leg length, smoking history and self-reported physical activity.

RESULTS

6MWD was 560 +/- 105 m and was not significantly different between men and women (P = 0.19). 6MWD was related to age (r = -0.36, P = 0.03), height (r = 0.35, P = 0.04), leg length (r = 0.38, P = 0.02) and the maximum HR achieved on the 6MWT when expressed as a percentage of the predicted maximum HR (%predHRmax, r = 0.73, P < 0.001). Stepwise multiple regression analysis showed that age, height, weight and %predHRmax were independent contributors (P < 0.01) to 6MWD, explaining 78% of the variance. Predicted 6MWD using regression equations derived from Caucasian subjects exceeded measured 6MWD by more than 75 m (P < 0.001).

CONCLUSIONS

This is the first study to report 6MWD for healthy Singaporeans aged 45-85 years. The regression equation developed in this study explained 78% of the variance in 6MWD. Published equations derived from Caucasian subjects overestimate 6MWD in Singaporean Chinese.

QDR 4500A dual-energy X-ray absorptiometer underestimates fat mass in comparison with criterion methods in adults

BACKGROUND

Dual-energy X-ray absorptiometry (DXA) has become one of the most frequently used methods for estimating human body composition. Although the DXA technique has been validated for the measurement of fat-free mass and fat mass, differences in calibration between instruments produced by different manufacturers, as well as between different models produced by the same manufacturer, have been reported.

OBJECTIVE

The objective was to compare the calibration of the QDR 4500A against criterion methods in a large heterogeneous population.

DESIGN

DXA-derived body-composition data were obtained from 7 studies: 6 data sets were provided by the investigators, one of which was published. The data included fat mass and fat-free mass measured with a QDR 4500A and criteria measurements of body composition from total body water by dilution at 4 centers, densitometry from 1 center, and four-compartment analysis at 2 centers.

RESULTS

In the cohort of 1195 subjects, 602 men and 593 women aged 19-82 y with a body mass index (in kg/m2) of 16-44, the fan-beam DXA overestimated fat-free mass (P < 0.05). A significant difference was observed in all 7 data sets, and the mean (+/-SE) was 5 +/- 1%.

CONCLUSIONS

It is recommended that the lean soft tissue mass estimate with the fan-beam QDR 4500A be reduced by 5% and that for fat mass be increased by that same mass. This finding is particularly important because the National Health and Nutrition Examination Survey is using the QDR 4500A to assess body composition in a nationally representative sample of persons in the United States.

Asians are different from Caucasians and from each other in their body mass index/body fat per cent relationship

The objective was to study the relationship between body mass index (BMI) and body fat per cent (BF%) in different population groups of Asians. The study design was a literature overview with special attention to recent Asian data. Specific information is provided on Indonesians (Malays and Chinese ancestry), Singaporean Chinese, Malays and Indians, and Hong Kong Chinese. The BMI was calculated from weight and height and the BF% was determined by deuterium oxide dilution, a chemical-for-compartment model, or dual-energy X-ray absorptiometry. All Asian populations studied had a higher BF% at a lower BMI compared to Caucasians. Generally, for the same BMI their BF% was 3-5% points higher compared to Caucasians. For the same BF% their BMI was 3-4 units lower compared to Caucasians. The high BF% at low BMI can be partly explained by differences in body build, i.e. differences in trunk-to-leg-length ratio and differences in slenderness. Differences in muscularity may also contribute to the different BF%/BMI relationship. Hence, the relationship between BF% and BMI is ethnic-specific. For comparisons of obesity prevalence between ethnic groups, universal BMI cut-off points are not appropriate.

The validity of predicted body fat percentage from body mass index and from impedance in samples of five European populations

OBJECTIVES

To test and compare the validity of a body mass index (BMI)-based prediction equation and an impedance-based prediction equation for body fat percentage among various European population groups.

DESIGN

Cross-sectional observational study.

SETTINGS

The study was performed in five different European centres: Maastricht and Wageningen (The Netherlands), Milan and Rome (Italy) and Tampere (Finland), where body composition studies are routinely performed.

SUBJECTS

A total of 234 females and 182 males, aged 18-70 y, BMI 17.0-41.9 kg/m(2).

METHODS

The reference method for body fat percentage (BF%(REF)) was either dual-energy X-ray absorptiometry (DXA) or densitometry (underwater weighing). Body fat percentage (BF%) was also predicted from BMI, age and sex (BF%(BMI)) or with a hand-held impedance analyser that uses in addition to arm impedance weight, height, age and sex as predictors (BF%(IMP)).

RESULTS

The overall mean (+/-s.e.) bias (measured minus predicted) for BF%(BMI) was 0.2+/-0.3 (NS) and-0.7+/-0.3 (NS) in females and males, respectively. The bias of BF%(IMP) was 0.2+/-0.2 (NS) and 1.0+/-0.4 (P<0.01) for females and males, respectively. There were significant differences in biases among the centres. The biases were correlated with level of BF% and with age. After correction for differences in age and BF% between the centres the bias of BF%(BMI) was not significantly different from zero in each centre and was not different among the centres anymore. The bias of BF%(IMP) decreased after correction and was significant from zero and significant from the other centres only in males from Tampere. Generally, individual biases can be high, leading to a considerable misclassification of obesity. The individual misclassification was generally higher with the BMI-based prediction.

CONCLUSIONS

The prediction formulas give generally good estimates of BF% on a group level in the five population samples, except for the males from Tampere. More comparative studies should be conducted to get better insight in the generalisation of prediction methods and formulas. Individual results and classifications have to be interpreted with caution.

Differences in body-composition assumptions across ethnic groups: practical consequences

Body-composition methodologies rely on assumptions that are not constant across ethnic groups. The consequence is that the methods used need to be population specific to guarantee unbiased conclusions. Alternatively, assumption-free methods such as chemical multiple-compartment models or scanning techniques should be used. Adequate and unbiased scientific data on body fat among ethnic population groups are necessary for a better understanding of the aetiology of obesity and its co-morbidities. The World Health Organization cut-off points for underweight, overweight, obesity and fat distribution, now universally defined, may need revision because the relationship between body mass index and body composition, and between indices of fat distribution and the actual amount of visceral fat, differ across ethnic groups. The need for ethnically specific cut-off points for obesity and fat distribution is substantiated, for example, by elevated disease risks at low levels of body mass index in several populations. Properly designed and conducted studies are needed to obtain unbiased information on these topics.