ARS | Publication request: Obesity classification in military personnel: A comparison of body fat, waist circumference, and body mass index measurements

Submitted to: Military Medicine
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: October 1, 2007
Publication Date: January 1, 2008
Citation: Heinrich, K.M., Jitnarin, N., Suminski, R.R., Berkel, L., Hunter, C.M., Alvarez, L., Brundige, A.R., Peterson, A.L., Foreyt, J.P., Haddock, C.K., Poston, W.S.C. 2008. Obesity classification in military personnel: A comparison of body fat, waist circumference, and body mass index measurements. Military Medicine. 173(1):67-73.

Interpretive Summary: The purpose of this study was to evaluate how common methods of assessing body composition affected obesity classification in a sample of overweight military personnel. The assessment methods that were analyzed included bioelectrical impedance-determined body fat% (BF%), waist circumference (WC), and body mass index (BMI). The results indicated that both WC and BMI underestimated obesity compared with BF%. Additionally, more non-obese men were misclassified as obese using the BMI method than the WC method. Although both WC and BMI tended to underestimate obesity, both were determined to be statistically acceptable for accurately predicting obesity in women, and WC was determined acceptable for predicting obesity in men. These results should inform research and policy in how WC- and BMI-derived data are used. The latter finding suggests that both WC and BMI are adequate for use in large clinical and population studies. However, the authors believe that the ideal field situation may be to combine WC and BMI data to improve prediction.

Technical Abstract: The purpose of this study was to evaluate obesity classifications from body fat percentage (BF%), body mass index (BMI), and waist circumference (WC). A total of 451 overweight/obese active duty military personnel completed all three assessments. Most were obese (men, 81%; women, 98%) using National Institutes of Health (NIH) BF% standards (men, >25%; women, >30%). Using the higher World Health Organization (WHO) BF >35% standard, 86% of women were obese. BMI (55.5% and 51.4%) and WC (21.4% and 31.9%) obesity rates were substantially lower for men and women, respectively (p < 0.05). BMI/WC were accurate discriminators for BF% obesity (theta for all comparisons >0.75, p < 0.001). Optimal cutoff points were lower than NIH/WHO standards; WC = 100 cm and BMI = 29 maximized sensitivity and specificity for men, and WC = 79 cm and BMI = 25.5 (NIH) or WC = 83 cm and BMI = 26 (WHO) maximized sensitivity and specificity for women. Both WC and BMI measures had high rates of false negatives compared to BF%. However, at a population level, WC/BMI are useful obesity measures, demonstrating fair-to-high discriminatory power.