Development and validation of anthropometric equations for prediction of the body fat in women aged 50 to 75 year

Abstract

This study had as objective the development and validation of specific anthropometric equations for predicting body fat among women between 50 and 75 years of age, using as criterion the dual-energy x-ray absorptiometry (DEXA). Seventy five women participated in the study and were grouped into two sub-samples: one for developing equations (GR, n=60), and the other for cross validation group (GV, n=15). Measurements of 10 skinfolds, 8 circumferences, body mass (MC), stature (EST) and body fat by DEXA (%GDEXA) were performed. Stepwise regression analyses were used for developing the equations. Cross-validation was analyzed through Pearson’s coefficient correlation (r), student-t test for comparison of means, calculation of constant error, standard error of the estimate (EPE) and total error. The sample presented for GR a mean age of 62.9±7.6 years, EST 155±0.05 cm, MC 70±10.8 kg and %GDEXA 38.4±5.6%. GV presented means of 59.3±4.9 years, EST 158±0.06 cm, MC 71.8±12.5 kg and %GDEXA 36.9±8.1%. Seven equations were developed, with a multiple correlation coefficient varying from 0.80 to 0.88 and an EPE from 2.88% to 3.58%. After analyzing adopted criteria, three equations using anthropometric variables and age were validated, presenting a correlation coefficient ranging from 0.86 to 0.90 and an EPE from 3.4% to 3.8%. Therefore, the proposed equations using athropometry and developed from DEXA criterion are valid to predict body fat among 50 to 75 year-old women.