Divergent Responses to Training Load in Professional Australian Football: Are Insights Obtained from Generic and...

Generic neuromuscular assessments are widely used to monitor training responses in team sports. Emerging running-based measures have been proposed to offer greater sensitivity to residual neuromuscular training effects by aligning more closely with the exercise dose, yet their practical utility remains ambiguous. Concurrent evaluation is thus necessary to identify athlete monitoring assessments that best capture meaningful responses.

We examined within-athlete relationships among common training response measures and their associations with prior weekly external training load in professional Australian Football athletes.

Repeated-measures, observational.

Forty-three male athletes were regularly assessed over two pre-seasons following 48-hours of no training. Outcome measures were derived from maximal performances in adductor strength and countermovement jump (CMJ) assessments, the final minute of a continuous-fixed submaximal fitness test (CF-SMFT: 3-minutes, 12 km⋅h^-1), and the plateau-phase of a high-intensity intermittent-fixed run (HI-IR_plateau: 50 m, ~ 25 km⋅h^-1). External loads were derived from Global Navigation Satellite Systems. Repeated-measures correlations assessed relationships between the response measures, while linear mixed-effects models evaluated measurement variability. Generalised Additive Models examined associations between prior weekly external training loads and the response measures.

Adductor strength and CMJ measures showed low within-athlete variability (coefficient of variation, CV = 4.3-6.6%) and limited associations with weekly external loads. Conversely, 2 standard deviation (SD) increases in weekly total distance were associated with reductions in CF-SMFT HR_ex of - 1.3%-points (95% Confidence Intervals, CI: - 0.9, - 1.8), and as large as - 3.2%-points. Similarly, 2 SD increases in very-high speed running (> 25 km.h^-1) were associated with declines in HI-IR_plateau velocity of - 0.4 km·h^-1 (95% CI: - 0.1, - 0.8), and as much as - 1.2 km·h^-1. Both effects exceeded thresholds of practical significance, as indicated by the typical error of measurement. Accelerometer-vector measures presented inconsistent associations.

Weekly external load changes were reflected in CF-SMFT HR_ex and HI-IR_plateau velocity, supporting their practical utility for athlete monitoring; whereas measures derived from generic neuromuscular tests (adductor strength, and CMJ) showed limited responsiveness. Associations between external loads and accelerometer-derived measures presented significant uncertainty, which currently limits confident recommendations on the utility of these emerging techniques.