Defining cutoffs to diagnose obesity using the relative fat mass (RFM): Association with mortality in NHANES 1999–2014

We have recently proposed and validated a simple and accurate method to estimate whole-body fat percentage in adults, the relative fat mass (RFM), derived from the ratio of height to waist circumference. We aimed to identify RFM cutoffs to diagnose obesity based on the association between RFM and all-cause mortality. We used data from adult participants (≥20 years of age, n = 43,793) of the National Health and Nutrition Examination Survey (NHANES) 1999–2014 linked with death certificate records from the National Death Index. Optimal RFM cutoffs were determined using receiver-operating characteristic analysis (the Youden’s index and the Euclidean minimum distance to the coordinate (0,1)). Final dataset for analyses comprised 31,008 adults. During a median follow-up of 8.3 years (IQR, 7.6–13.7), 2,517 deaths occurred. Youden and Euclidean optimal cutoffs of baseline RFM for all-cause mortality were 40.8% and 41.6% for women, and 30.9% and 28.9% for men, respectively. Similar cutoffs were obtained using measured whole-body fat percentage by dual energy X-ray absorptiometry. Adjusting for age, BMI category, ethnicity, education level, and smoking status, the hazard ratio for mortality using Cox proportional hazard regression was 1.41 (95% CI, 1.02–1.95) among women who had an RFM of 40.0–44.9% compared with women who had an RFM <35% (P = 0.035). Among men, the hazard ratio was 1.57 (95% CI, 1.07–2.30) among those with an RFM of 30.0–34.9% compared with men who had an RFM <25% (P = 0.020). Similar adjusted hazard ratios for same RFM categories were obtained in our validation population (NHANES III, n = 12,650, median follow-up: 23.3 years): 1.42 (95% CI, 1.01–2.00) among women (P = 0.043) and 1.50 (95% CI, 1.07–2.10) among men (P = 0.021). We suggest rounded RFM cutoffs of 40% for women and 30% for men to diagnose obesity and identify individuals at higher risk of death.