An allometric large-scale analysis of the observed performance limits and sex differences in elite powerlifting

The objective normalization of athletic performance across diverse body masses remains one of the most persistent methodological challenges in strength and conditioning research. Historically, this normalization has been governed by the theoretical square-cube law (scaling exponent b = 0.67) or, more recently in strength sports, by highly parameterized polynomial algorithms such as the Wilks and International Powerlifting Federation General Load (IPF GL) formulas. This study challenges both geometric scaling and polynomial overfitting by deriving empirical allometric exponents from a tiered stratification design.

A large-scale, highly curated dataset of powerlifters competing under performance-enhancing drug testing protocols was analyzed (N = 457,471). A stratified log-log regression model was implemented, progressively restricted to the absolute Top 20 all-time athletes per weight category (N = 1,833) to identify shifting scaling laws. Comparative statistical equivalence and agreement with the IPF GL formula were evaluated using Williams' t-test and Bland-Altman analysis.

The research identified a systemic phenomenon termed "allometric decay". As athletes approach absolute physiological limits, the scaling exponent meaningfully deviates from 0.67, decaying from general population baselines (males: b = 0.55; females: b = 0.50) down to 0.47 for elite males and 0.41 for elite females, at the Global All-Time Top 20 level. Despite its mathematical simplicity, the empirical allometric model achieved statistical equivalence with the complex IPF GL formula in mass-independence for elite males (Williams' t = -0.72, p = 0.47, Cohen's q = 0.001) and demonstrated extremely high-ranking agreement (mean bias = 0.00; 95% limits of agreement = ± 0.20 Z-scores).

This significant decay indicates that the observed cost of acquiring functional muscle mass exhibits progressive diminishing returns at the elite level. Consequently, current polynomial scoring systems appear to be mathematically overfitting data to compensate for these underlying physiological ceilings. This study proposes a return to sex-specific empirical allometry, advocating for the adoption of these derived exponents as a more parsimonious, transparent, and biologically grounded method for cross-category strength comparison.