OBJECTIVE: To investigate if physical activity interventions increase objectively measured physical activity levels of patients with chronic musculoskeletal pain (e.g. osteoarthritis, low back pain) compared to no/minimal intervention. METHODS: We performed a systematic review with meta-analysis (Prospero registration CRD42014015363) searching the MEDLINE, EMBASE, CINAHL, SPORTDiscus and PEDro and the main clinical trial registers. Quasi- or randomized controlled trials investigating the effect of physical activity interventions on objectively measured physical activity levels (e.g. accelerometers, pedometers) of patients with chronic musculoskeletal pain compared with no/minimal intervention were considered eligible. Analyses were conducted separately for short-term ( 3 months and < 12 months); and long-term (>/= 12 months) follow-ups. Pooled effects were calculated using standardized mean difference (SMD), and the GRADE approach was used in summary conclusions. Results Eight published trials and six registered trials were included. For the short-term follow-up, pooling of six trials showed no significant effect (SMD= 0.34, 95% CI: -0.09 to 0.77) between physical activity intervention and no/minimal intervention. Similarly non-significant results were found for the intermediate and long-term follow-ups. The overall evidence according to the GRADE approach was classified as low quality. Conclusion Our findings suggest that physical activity-based interventions may lead to little or no difference in objectively measured physical activity levels of patients with chronic musculoskeletal pain compared with no/minimal interventions. Given the number of registered trials, the pooled effect found in this review is likely to change once the results of these trials become available. This article is protected by copyright. All rights reserved.
a2151-4658 (Electronic)