Development and Cross-Validation of a Prediction Equation for Estimating Percentage Body Fat From Body Mass Index in...

The purpose of this study was to develop and cross-validate an equation for estimating percentage body fat (%BF) from body mass index and other potential independent variables among young persons with intellectual disability. Participants were 128 persons with intellectual disability (62 women; age 16-24 years) split between development (n = 98) and cross-validation (n = 30) samples. Dual-energy X-ray absorptiometry served as the reference method for %BF. An equation including 1/body mass index and sex (0 = male; 1 = female) was highly accurate in estimating %BF (p < .001; R2 = .82; standard error of estimate = 5.22%). Mean absolute and root mean square errors were small (3.1% and 3.9%, respectively). A Bland-Altman plot indicated nearly zero mean difference between actual and predicted %BF with modest 95% confidence intervals. The prediction equation was %BF = 56.708 - (729.200 × [1/body mass index]) + (12.134 × sex). Health care professionals may use the prediction equation for monitoring %BF among young people with intellectual disability.