Anthropometric indicators of fat-mass excess among women*

The aim of this study was to determine the sensitivity, specificity and the level of association between two indicators of overweight and obesity among women. Sixty-five women (from 50-77 years of age) participated in the study. They showed mean body mass of 70.3 ± 11 kg and mean stature of 158.0 ± 5.5 cm. The body mass index (BMI) and the reciprocal of ponderal index (RPI) were computed. Percent of body fat, estimated by the dual-energy x-ray absorptiometry (%BFDXA), was used as the gold standard measure. Data analyses consisted of descriptive statistics, Pearson’s linear correlation and the Kappa index. The means for BMI, RPI and %BFDXA were, respectively, 28 ± 4.2 kg.m-2, 38 ± 1.9 cm.kg1/3 and 38.1 ± 6.0. A high prevalence of overweight was detected, regardless the method (%BFDXA = 89.2%, > 28%, RPI = 83,1% and BMI = 73,8%). Pearson’s lineal correlation coefficients between %BFDXA and RPI (r = -0.76) and %BFDXA and BMI (r = 0.72) were both significant (p < 0.01). The Kappa index showed an association of 0.31 between the %BFDXA and BMI and of 0.48 between %BFDXA and RPI. The anthropometric indicators presented high sensitivity and specificity values (80% and 55% for BMI 87,7% and 75% for RPI, respectively). The ROC curve analyses showed areas under the curve of 0.80 for BMI and 0.83 for RPI, which did not differ (p > 0.05). The cutoff points of 27.3 kg.m-2 for BMI and 39.3 cm.kg-1/ 3 for RPI presented a better relationship between sensitivity and specificity in determining overweight. Therefore, it is concluded that both BMI and RPI are good indicators for identifying overweight in women aged over 50 years as they presented high sensitivity and specificity.