Objective This study aimed to compare anthropometry with dual-energy X-ray absorptiometry for estimating the body composition of elderly women with type 2 diabetes. Methods The body composition of 51 volunteers aged 60 to 70 years, with body mass indices ranging from 19 and 43kg/m², was determined by anthropometry (triceps, suprailiac and thigh skinfold thicknesses) and dual-energy X-ray, which is considered the gold standard. The Bland-Altman limits of agreement were the standard deviations of the dual-energy X-ray coefficient of variation and the anthropometric values should be in this range. The means were compared by the Student's t-test followed by the Pearson's correlation. Results Limits of agreement: weight (W, kg) M=0.3, SD=0.1, lean body mass (lear body mass, kg) M=3.0, SD=1.9; fat body mass (fat body mass, kg) M=5.0, SD=3.1, percentage of body fat (% body fat) M=5.0, SD=3.1. The means of the two methods were the same and within the limits of agreement. Anthropometry tended to overestimate lear body mass by 2.31 kg and underestimate fat body mass and % BF by -0.93 kg and -2.7%, respectively. The estimates correlated significantly (W, r=0.99; lear body mass, r=0.79; fat body mass, r=0.93; % body fat, r=0.72; p<0.0001). Conclusion The anthropometric and dual-energy x-ray results were consistent. Although some lear body mass and % body fat results were beyond the limits of agreement, the others were within. Therefore, anthropometry can be used to assess the body composition of this population.
Body composition; Diabetes mellitus, type 2; Aged
