Peak VO2 correction for fat-free mass estimated by anthropometry and DEXA

Anthropometric (ANTHRO) and dual-energy x-ray absorptiometric (DEXA) estimates of total body and leg fat-free masses (FFM) were obtained in 77 randomly selected sedentary men and women, aged 20-80: intermethod limits of agreement and their clinical significance, as inferred from the differences on peak VO2 corrected for FFMANTHRO and FFMDEXA, were determined.

Limits of agreement were calculated as mean bias +/- 95% confidence intervals: peak VO2 at maximum cycle ergometry was related to FFMANTHRO and FFMDEXA by using both standard (y x x(-1)) and power function ratios (allometry).

Data distribution of the ANTHRO-DEXA differences presented significant heteroscedasticity in both sexes, i.e., differences were proportional to the mean (P < 0.05). After logarithmic transformation, the mean bias +/- 95% limits of agreement were expressed as ratios (ANTHRO x DEXA(-1) x// error ratio): these corresponded to 0.95 x// 1.11 or 0.99 x// 1.15 for total body FFM and 0.90 x// 1.10 or 1.02 x// 1.07 for leg FFM in men and women, respectively. In addition, we found different allometric exponents for FFMANTHRO and FFMDEXA: the intermethod differences, therefore, increased after power function expression (P < 0.05).

Discrepancies between ANTHRO and DEXA measurements of FFM depend on the magnitude of the estimate: differences are typically within 10 to 15%. Importantly, FFM-corrected peak VO2 values can vary according to the method chosen for body composition assessment, especially when allometry is used for peak VO2 correction. These results demonstrate that ANTHRO-DEXA differences in FFM estimation do have relevant practical consequences for the analysis of maximum aerobic capacity in nontrained humans.