Relationship between body fat and anthropometric variables in a large sample of young males

To establish the relationship between fat mass (FM) and other anthropometric variables, 2152 males of mid-European extract aged between 18 and 24 years were examined. Based on these data, three models for either fat percent (FP) or FM prediction were calculated from body mass (BM), body height (BH) and body mass index (BMI = BM/BH2). The best models for fat percent prediction both explained only 51% of the variance (FP = 0.02711.BMI2.042 (0.043 SE); and FP = 720310.BM2.05 (0.043).BH-3.76 (0.155), R2 = 0.51 for both models). There was a remarkable increase in R2 when FP was replaced by FM. The best model explained 70% of the variance for FM prediction (FM = 7207.BM3.05 (0.043).BH-3.76 (0.155), R2 = 0.70). We have incorporated these data into a nomogram applicable to an active, young male mid-European population. From this study we conclude that for the estimation of body fat, models based upon FM (absolute value) are preferred over models for prediction of FP (relative value).