The Relations Among Body Mass Index, Percent Body Fat, and Cardiorespiratory Fitness

Abstract

Body mass index (BMI) is used to classify weight status (underweight, normal weight, overweight, obesity) in clinical and research settings. Recent data are inconsistent in regard to the adverse health effects of BMI-defmed overweight and obesity. BMI's relation with adiposity may attenuate its relations with health outcomes and risk factors for chronic disease morbidity and mortality. PURPOSE: 1) To determine relations among BMI, hydro static ally determined percent body fat (PFAT) and cardiorespiratory fitness (CRF); 2) To develop a regression equation using readily measured values of gender, age (yrs), BMI (wt/ht2) to estimate PFAT (EPFAT); and 3) To determine the relation between EPFAT and CRF. METHODS: The study sample included 28,673 participants (Men=23,543; Women=5130) who participated in a clinical evaluation of health and fitness at The Cooper Clinic (Dallas, TX) between 1970–2002. Clinical measures included height and weight from which BMI was computed as kg/m2, PFAT using hydrostatic weighing techniques, and CRF quantified as the duration of a maximal treadmill exercise test using a modified Balke protocol. RESULTS: For the entire sample, the correlations (r) for BMI were .48 (p < .0001) with PFAT and −.28 (p < .0001) with CRF. The partial rs, controlling for gender and age, were .64 (p < .0001) and −.40 (p < .0001). With PFAT as the criterion, a regression equation was developed using gender (Men=0, Women=1), age, and BMI as the predictors. The multiple correlation (R) was .71 (p < .0001, SEE = 5.1%). Gender specific analyses using age and BMI as predictors (Men: R = .68, SEE = 5.0%; Women: R = .71, SEE = 5.6%) produced similar results. A unique cross validation sample (n= 1510) verified the consistency of the regression equation (R = .70, p < .0001, SEE = 5.2%). The rbetween CRF and EPFAT was −.54 (p < .0001), similar to that between CRF and PFAT (r = −.59, p < .0001). CONCLUSIONS: BMI using measured height and weight had moderate to low relations with PFAT and CRF. A cross validated regression equation using gender, age, and BMI can be used to estimate PFAT. The relation between EPFAT and CRF was significantly higher than the relation between BMI and CRF, a known risk factor for all-cause mortality. Supported in part by NIH grant AG06945.