A comparison of body fat estimates using anthropometry and bioelectrical impedance analysis with distinct prediction equations in elderly persons in the Republic of Guatemala.

Bioelectrical impedance analysis (BIA) has been proposed as a technique to estimate body composition. Its application has extended beyond the clinical setting, however, it remains to be determined if the BIA prediction formulas published to date provide comparable body composition estimates in elderly populations. The objective of this study was to compare the estimates of body fat derived from published prediction equations in a developing country's population such as Guatemala. A total of 234 elderly persons were studied (108 males and 126 females), with ages of 78 +/- 7 and 77 +/- 8 (mean +/- SD) yr, respectively. The height and weight averaged 156.4 +/- 7.9 and 144.0 +/- 7.0 cm, and 54.5 +/- 9.4 and 49.4 +/- 10.6 kg, for males and females, respectively. Measurements included skinfolds (triceps, biceps, subscapular, and suprailiac), and body resistance (BIA). Body fat estimates were derived from four BIA-prediction formulas and three equations based on anthropometry. Mean percent body fat estimates for the overall population obtained from anthropometry and BIA prediction formulas ranged from 22 to 50%. Females had higher estimates compared to males without regard for the method used. Highly significant intermethod correlations were seen, but differences in fat estimation among prediction formulas were noticed and the magnitude of the differences were BIA-formula dependent.