Swinton · Sports medicine (Auckland, N.Z.) 2024 · Bayesian mixed-effects meta-analysis · n=295 studies (6,710 participants)

Dose-Response Modelling of Resistance Exercise Across Outcome Domains in Strength and Conditioning: A Meta-analysis.

Cited 45 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of 295 interventional training studies (though inclusion was not restricted to controlled trials).

PubMed 38652410 · doi:10.1007/s40279-024-02006-3 · record verified 2026-08-26

What was done

Bayesian mixed-effects meta-analytic modelling was performed on a dataset of studies from 1962 to 2018 evaluating resistance-only or resistance-dominant training in healthy trained and untrained adults. Studies were not required to have a control group, and analyses were restricted to interventions lasting up to 26 weeks. The authors examined linear and non-linear dose-response relationships between training variables (frequency, volume, load intensity as % of 1RM), participant traits (sex, training status), and physical performance domains (maximal strength, power, vertical jump, change of direction, sprint performance).

What was found

Across 295 studies (535 groups, 6,710 participants), the best-performing model included duration from baseline, average number of sets, and non-linear interactions between outcome domain and load intensity (% 1RM). Model predictive accuracy was not improved by including participant sex or training status. Optimal intensity varied by outcome: maximal strength peaked with the heaviest loads, power peaked with 40-70% 1RM, and vertical jump peaked with light loads (~30% 1RM). Specific numerical effect sizes and credible intervals were not reported in the abstract.

Why it matters

This large-scale modelling demonstrates that resistance training load prescription should be specifically tailored to the targeted performance adaptation rather than applying uniform loading parameters across different physical qualities.

Limits

Inclusion was not limited to controlled trials, potentially introducing bias from time or test-retest effects. Exact numerical effect sizes, variance estimates, and uncertainty intervals are omitted from the abstract. Long-term training adaptations beyond 26 weeks were not evaluated.

Cited by