Mechanistic evidence and exercise interventions: Causal claims, extrapolation, and implementation.

Abstract

Exercise interventions and policies are widely prescribed in both sport and healthcare. Research investigating exercise interventions and policies is generally conducted using an Evidence-Based framework, placing an emphasis on evidence gathered from randomised controlled trials (RCTs). To explore the idea that, in addition to the assessment of evidence from RCTs when investigating exercise interventions, mechanistic studies ought to also be assessed and considered. This article assesses the rationale supporting the use of RCTs as evidence for exercise interventions, and the use of evidence of mechanisms in establishing efficacy, determining external validity, and tailoring interventions. The article argues that evidence from mechanistic studies ought to be considered alongside evidence from RCTs because: as RCTs investigating exercise interventions tend to be of low quality, mechanistic studies ought to be used to reinforce the evidence base; further, evidence from mechanistic studies is highly useful for both questions of extrapolation and implementation. This article argues for this on theoretical grounds, and also draws on a number of case studies.