Background: The aim of this paper was to identify the most appropriate allometric scaling model for expressing aerobic fitness, determined by maximal oxygen consumption (VO<inf>2max</inf>), that would allow comparisons across differing body types.
Methods: VO<inf>2max</inf> and body composition data were collected from untrained non-obese and obese participants (N.=126). Allometric models were created using body mass (BM), fat-free mass (FFM), and leg FFM (LFFM) to determine the goodness-of-fit using the Akaike Information Criterion (AIC).
Results: Allometric scaled exponents adjusted for BM, FFM and LFFM were 0.67, 0.68 and 0.55, respectively. VO<inf>2max</inf> scaled to BM was 22% higher in non-obese individuals. Scaled to LFFM, V VO<inf>2max</inf> was only 7.5% higher in non-obese individuals as compared to obese individuals. Data showed a positive correlation (r=0.28; P=0.009) between VO<inf>2max</inf> and BM for non-obese participants and a negative correlation (r=-0.39; P=0.014) for obese participants. AIC values showed the LFFM model as the best fit (AICc = 0 "substantial support) and the AIC differences for FFM and BM were both >10 "no support" for the model (12.1 and 28.2, respectively).
Conclusions: Interpretation of aerobic power and comparisons would be most appropriate when allometrically scaled to the metabolically active tissue (LFFM). Bias is introduced when scaling to BM and comparing individuals of various body compositions.