Abstract
Background: The large response variability observed in most training intervention studies is often used as an argument that certain individuals are more responsive to training than others. However, this implicitly relies on the premise that training responses are highly reproducible. Accordingly, the primary objective of the present study was to evaluate the within-individual reproducibility of training responses to repeated endurance exercise training (ET).
Methods: Forty-two initially untrained men and women (age: 53.9 ± 8.7 years; weight: 86.0 ± 19.5 kg; V̇O2max: 32.4 ± 6.9 mL ·min-1·kg-1) underwent two identical 8-week ET interventions separated by an 8-week wash-out period. Each ET-period consisted of 24 supervised and effort-matched interval sessions. Maximal oxygen consumption (V̇O2max), maximal aerobic power output (Wmax), and mean power output during 15-min all out (W15min) was assessed pre and post both training periods.
Results: Intraclass correlations (ICCs) indicated that the training stimuli were consistent across both periods and confirmed that the wash-out period effectively returned participants' values to baseline levels within-individuals. Across both training periods, it was demonstrated “poor” within-individual reliability for changes in V̇O2max and W15min, and “poor to moderate” reliability for Wmax (ICC: 0.19 [-0.12, 0.46], 0.22 [-0.07, 0.48] and 0.35 [0.07, 0.58], respectively. However, the mean group changes were consistent across both periods (V̇O2max, 9.3% vs 9.7%; W15min; 14.4% vs 15.5%; Wmax, 11.2% vs 13.0%).
Conclusion: The present study demonstrates that observed individual responses in endurance performance measures and determinants of endurance performance are inconsistent and poorly reproducible across two identical eight-week ET interventions in untrained middle-aged men and women. However, at the group level, the mean responses remained consistently reproducible across the repeated interventions.
