Elsevier

Obesity Research & Clinical Practice

Volume 7, Issue 1, January–February 2013, Pages e55-e66
Obesity Research & Clinical Practice

Original Article
Comparison of anthropometric and body composition measures as predictors of components of the metabolic syndrome in a clinical setting

https://doi.org/10.1016/j.orcp.2012.10.004Get rights and content

Summary

Problem

The use of body mass index (BMI) to assess obesity and health risks has been criticized in scientific and lay publications because of its failure to account for body shape and inability to distinguish fat mass from lean mass. We sought to determine whether other anthropometric measures (waist circumference (WC), waist-to-height ratio (WtH), percent body fat (%BF), fat mass index (FMI), or fat-free mass index (FFMI)) were consistently better predictors of components of the metabolic syndrome than BMI is.

Methods

Cross-sectional measurements of height, weight, waist circumference and percent body fat were obtained from 12,294 adults who took part in annual physical exams provided by EHE International, Inc. Blood pressure was measured during the exam and HDL, LDL, and fasting glucose were measured from blood samples. Pearson correlations, linear regression, and adjusted Receiver Operator Characteristic (ROC) curves were used to relate each anthropometric measure to each metabolic risk factor.

Results

None of the measures was consistently the strongest predictor. BMI was the strongest predictor of blood pressure, measures related to central adiposity (WC and WtH) performed better at predicting fasting glucose, and all measures were roughly comparable at predicting cholesterol levels. In all, differences in areas under ROC curves were 0.03 or less for all measure/outcome pairs that performed better than BMI.

Conclusion

Body mass index is an adequate measure of adiposity for clinical purposes. In the context of lay press critiques of BMI and recommendations for alternative body-size measures, these data support clinicians making recommendations to patients based on BMI measurements.

Introduction

Body Mass Index (BMI) is a widely used measured of body size and standardized BMI thresholds define overweight and obesity [1]. Despite its common usage among clinicians, researchers, and the lay public there is been ongoing criticism of BMI as a measure of obesity and as a predictor of excess body fat related health risk [2], [3], [4]. The general public has become aware of this controversy as the popular press has publicized these criticisms of BMI (e.g. [5], [6], [7], [8]). Purported shortcomings of the BMI measure include its inability to discriminate between lean mass and fat mass and to capture the anatomical distribution of adipose tissue.

Studies have shown a strong association between high percent body fat and an increased risk of chronic diseases such as hypertension, dyslipidemia, diabetes mellitus, and coronary heart disease [9], [10], [11]. While BMI and fat mass are highly correlated at high BMIs, these measures are less well correlated in normal weight ranges [12], [13]. As a result, measures capturing differences in body composition have also been proposed, including percent body fat (%BF fat mass/total mass), Fat Mass Index (FMI – fat in kg/height in m2) [2], [3] and Fat-free Mass Index (FFMI – fat-free mass in kg/height in m2) [14]. The measurement of body fat has traditionally involved imprecise measures such as skinfold thickness or expensive machinery such as dual energy X-ray absorptiometric machines [15]. More recently, inexpensive technologies to clinically measure body-composition such as bio-electric impedance as implemented in the Tanita scale are increasingly becoming available [16]. While calculation of percent body fat from impedance data may be imprecise and unreliable if the scale's body fat calculation equation does not take ethnicity into account [21], the commercial availability of this technology is likely to increase the use of body composition measures in routine clinical practice.

Visceral fat, which is located in the abdominal region, is more strongly associated with adverse health effects than fat that is distributed in other areas, such as the hips [17]. Changes in metabolic intermediates, such as blood pressure, lipid profile, and insulin sensitivity, have also been associated with an excess of abdominal fat [18]. As a result, measures that better reflect body shape and the anatomical distribution of adipose tissue have been proposed as clinically useful measures of obesity, including waist circumference (WC) [19], [20] and the subject's waist circumference divided by the subject's height, or waist-to-height ratio (WtH) [21], [22]. Ultimately, the test of a measure of obesity is its ability to reflect health risk; for an obesity measure to replace BMI in the clinical setting, the measure must consistently improve clinicians’ ability to assess health risks associated with patient adiposity. While many prior studies have assessed the relative value of anthropometric measures in predicting metabolic intermediates, cardiovascular disease, and mortality (e.g. [9], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38]), results have been inconsistent. We are aware of only one study to date that compares the predictive value of impedance-measured body composition to BMI in a multi-ethnic population [9], and that study used raw impedance data rather than the results of a commercially available impedance scale designed for professional use, limiting its conclusions for clinical applications. We assessed alternative anthropometric measures and a commercially available impedance clinical scale's body composition measures, as predictors of components of the metabolic syndrome in an ethnically diverse population.

Section snippets

Sample

Cross-sectional data was obtained from EHE International, Inc. (EHE), which provides annual routine screening physical examinations as part of a corporate wellness plan. The examinations are provided free of charge to the employees and their spouses. The physical examinations take place at six EHE-owned centers and at a network of more than 60 physician offices across the country [39]. Data from visits in 2007 was used in our analyses. Body composition and waist circumference measurements were

Results

Data for all measures were available from 12,294 subjects; Table 1 provides descriptive statistics for the subjects. About 17% of the cohort was obese and prevalence of clinical risk ranged from 16% for low HDL to 4% for IFG. Most anthropometric measures were correlated with each other (Table 2), with between-measure correlations ranging from 0.93 (for WtH:WC) to 0.41 (for WC:%BF). Notably, BMI was strongly correlated (r > 0.8) with all measures except with percent body fat (r = 0.65).

In regression

Discussion

In these cross-sectional analyses of a multi-ethnic patient population, none of the available anthropometric measures was consistently the best predictor of inter-individual variation in the metabolic risk factors studied. Compared with BMI, no alternate anthropometric measure was consistently better under regression or ROC analysis and the magnitude of all differences in predictive value across measures was minor. Overall, while the use of alternate anthropometric measures in place of BMI may

Conflict of interest

AR sits on the Medical Advisory Board of EHE International, Inc. SJM and AB have no personal or financial conflict of interest to disclose.

Authorial roles

AR, SJM and AB designed research; AR provided databases; SJM and AB analyzed data; SJM, AB and AR wrote the paper; SJM had primary responsibility for final content. All authors read and approved the final manuscript.

Acknowledgments

No funds from any grant supported this project.

EHE International, Inc., supplied the de-identified medical record data. Dr. Rundle sits on the Medical Advisory Board of EHE International, Inc. EHE International, Inc. did not have any role in the design of the study, execution of the analyses or in the decision to submit this article for publication. EHE International, Inc. did verify that the information provided about the company and its exam protocol is accurate.

References (56)

  • J.E. Brody

    Weight index doesn’t tell the whole truth

    The New York Times

    (2010)
  • S. Adams

    ‘Forget BMI, just measure your waist and height’ say scientists

    The Telegraph

    (2012)
  • M. Healy

    Move over BMI: There's a new way to measure fat

    Los Angeles Times

    (2012)
  • K. Willett et al.

    Comparison of bioelectrical impedance and BMI in predicting obesity-related medical conditions

    Obesity (Silver Spring)

    (2006)
  • A.M. Sharma et al.

    Obesity, hypertension and insulin resistance

    Acta Diabetol

    (2005)
  • M.Y. Jaffrin

    Body composition determination by bioimpedance: an update

    Curr Opin Clin Nutr Metab Care

    (2009)
  • A. Romero-Corral et al.

    Accuracy of body mass index in diagnosing obesity in the adult general population

    Int J Obes (Lond)

    (2008)
  • M.I. Goran

    Measurement issues related to studies of childhood obesity: assessment of body composition, body fat distribution, physical activity, and food intake

    Pediatrics

    (1998)
  • S.R. Daniels

    The use of BMI in the clinical setting

    Pediatrics

    (2009)
  • D. Canoy

    Coronary heart disease and body fat distribution

    Curr Atheroscler Rep

    (2010)
  • M. Wei et al.

    Waist circumference as the best predictor of noninsulin dependent diabetes mellitus (NIDDM) compared to body mass index, waist/hip ratio and other anthropometric measurements in Mexican Americans – a 7-year prospective study

    Obes Res

    (1997)
  • T.A. Welborn et al.

    Preferred clinical measures of central obesity for predicting mortality

    Eur J Clin Nutr

    (2007)
  • M. Ashwell et al.

    Six reasons why the waist-to-height ratio is a rapid and effective global indicator for health risks of obesity and how its use could simplify the international public health message on obesity

    Int J Food Sci Nutr

    (2005)
  • A.S. Can et al.

    Anthropometric indices and their relationship with cardiometabolic risk factors in a sample of Turkish adults

    Public Health Nutr

    (2009)
  • P.R. Deshmukh et al.

    Relationship of anthropometric indicators with blood pressure levels in rural Wardha

    Indian J Med Res

    (2006)
  • J.R. Ghosh et al.

    Comparative evaluation of obesity measures: relationship with blood pressures and hypertension

    Singapore Med J

    (2007)
  • V. Maher et al.

    Association of central obesity with early carotid intima-media thickening is independent of that from other risk factors

    Int J Obes (Lond)

    (2009)
  • S. Patel et al.

    A comparison of proxy measures of abdominal obesity in Chinese, European and South Asian adults

    Diabet Med

    (1999)
  • Cited by (43)

    • Biomarkers in metabolic syndrome

      2022, Advances in Clinical Chemistry
      Citation Excerpt :

      However, another study that included 1552 Qatari adults > 20 years of age, reported the BMI cut-off at 28 kg/m2 [37]. In a cross-sectional annual physical study of 12,294 adults in the US, BMI was found to be an inexpensive, easy and adequate measure that predicted the presence of metabolic risk factors as accurately as WC, WHR, and percent body fat [38]. In addition, in a cross-sectional study of 981 Spanish adolescents (456 boys and 525 girls), BMI showed an area under the curve (AUC) values for MetS (I) were > 0.75 and > 0.85, respectively [39].

    • Relationship between obesity and structural brain abnormality: Accumulated evidence from observational studies

      2021, Ageing Research Reviews
      Citation Excerpt :

      The results of analyses related to cerebral microbleeds performed in our study were all insignificant. This also confirms the recently proposed idea that central obesity or visceral obesity has a higher predictive power than global obesity for the risk of structural brain abnormalities (S. Chang et al., 2012; Lee et al., 2013; Mooney et al., 2013). Also, a noteworthy aspect is that in recent years, researchers have discovered a U-shaped association between age and the impact of obese status on structural brain abnormalities (Kaur et al., 2015; Kim et al., 2015a, 2015b; Qizilbash et al., 2015; Singh-Manoux et al., 2018).

    • MRI-Derived Radiomics Features of Hepatic Fat Predict Metabolic States in Individuals without Cardiovascular Disease

      2021, Academic Radiology
      Citation Excerpt :

      For benchmarking of RFRAD models, three parameters, where associations with T2DM and MetS had previously been established, were chosen from the KORA-MRI study data. These included the body mass index (BMI) (13,14), as an anthropometric marker and quantifications of hepatic fat content (21,23) from T1-MED and MRS sequences. In analogy to the RFRAD models, benchmark RF models were trained on the SMOTE-rebalanced training set.

    View all citing articles on Scopus
    View full text