Little evidence for meaningful additive intervention effect heterogeneity of resistance training upon strength and hypertrophy: Evidence from causal inference and meta-analytic variance comparisons in a large dataset of randomised controlled trials
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Resistance training studies consistently demonstrate substantial variation in observed outcomes after interventions between participants.This variation is commonly interpreted as evidence that individuals differ meaningfully in their responses to resistance training interventions.However, observed variation in outcomes does not necessarily imply heterogeneity in the underlying causal effects of an intervention.Rather, observed outcomes reflect multiple sources of variation, including between-participant differences, within-participant variability, measurement error, and potentially true intervention effect heterogeneity.Within the potential outcomes framework, heterogeneity of intervention effects is fundamentally a question about the variance of causal effects.Although directly estimating this variance generally requires specialised study designs such as randomised replicated crossover trials, traditional randomised controlled trials (RCTs) comparing an intervention against an inert control can provide indirect evidence regarding intervention effect heterogeneity through comparisons of outcome variances.Here, I outline the causal framework underpinning variance comparisons under additive intervention effects, discuss the assumptions required for causal interpretation, and demonstrate how meta-analytic models can synthesise evidence across studies to compare variances while accounting for the ubiquitous relationship between means and variances.Applying these methods to a large (102 studies, 292 arms, 3784 participants) meta-analytic dataset of RCTs of resistance training interventions reveals little evidence that interventions meaningfully increase outcome variability.Although the conclusions drawn necessarily depend upon assumptions regarding additivity of intervention effects, that controls are inert, and the correlation between potential outcomes, these assumptions appear reasonable for resistance training studies in healthy participants.Taken together, current evidence provides little support for the existence of substantial intervention effect heterogeneity in resistance training adaptations.As such, recommendations, policies, and guidelines for resistance training interventions can be simplified and it can be justifiably assumed that the average intervention effects estimated in adequately powered and precise RCTs are constant and experienced by all undergoing the intervention.
Fields of Study & Subjects
Sports Performance and Training
Meta-analysis and systematic reviews
Advanced Causal Inference Techniques
Publication details
- Publication date
- 27 March 2026
- Language
- EN
- Document type
- Preprint