Abstract

This systematic review and meta-analysis investigated the effectiveness of physical activity (PA) and sedentary behaviour (SB) interventions on PA and SB levels in people with persistent musculoskeletal pain. We explored the effectiveness of behaviour change techniques (BCTs), the use of behaviour change theory and non-PA/SB outcomes. Randomised controlled trials of PA or SB interventions for people with persistent musculoskeletal pain were eligible. Twenty-three studies were included. Quality of evidence was assessed using the GRADE approach. Meta-analysis demonstrated a small effect for PA post-intervention (Hedge's g = 0.321, CI 0.136 to 0.507, p = 0.001, very low-quality evidence). There was no effect for longer-term follow-up PA (low quality evidence) or SB outcomes (very low-quality evidence). There was a small effect for studies with low risk-of-bias at longer-term follow-up PA. Self-report PA outcomes, PA and education interventions, non-self-selected PA, a combination of supervised and unsupervised PA and a combination of individual and group-based interventions had larger effects. Heterogeneity was moderate to considerable. Risk-of-bias, assessed using Cochrane risk-of-bias tool (version two), was generally low. Five promising BCTs were identified: ‘adding objects to the environment’, ‘goal setting (outcome)’, ‘action planning’, ‘monitoring outcome(s) of behaviour by others without feedback’ and ‘feedback on outcome(s) of behaviour’. In conclusion, there is evidence for a modest benefit for PA interventions immediately post-intervention, however the quality of evidence is very low. There was no evidence for longer-term follow-up PA or SB. Higher quality studies of PA and SB interventions that use objective measures are needed. PROSPERO registration: CRD42020180260.