RESUMO
OBJECTIVES: To determine if sleep interventions improve pain and sleep in people with osteoarthritis (OA) and/or spinal pain compared to control/placebo. DESIGN: Medline, Embase, AMED, PsycINFO, CENTRAL, CINAHL and PEDro were searched from their inception date to July 2017. Keywords relating to "sleep", "OA", "spinal pain", and "randomized controlled trial (RCT)" were combined. Included RCTs investigated the use of sleep interventions for people with OA and/or spinal pain, and measured at least one sleep and health related outcome. Meta-analyses were performed to pool mean differences for pain and sleep quality. PROSPERO: CRD42016036315. RESULTS: Of 1445 unique records, 24 studies were included. Sixteen studies included participants with spinal pain, seven with OA, and one included a mixed population. Sleep interventions included established sleep interventions (ESI) [cognitive behavioural therapy (CBT) and pharmacological interventions], and a range of others. Intervention periods ranged from 4 to 10 weeks. Thirteen studies were of moderate to high quality (PEDro ≥ 6/10). Due to high heterogeneity between studies we also performed sub-group and sensitivity analyses. ESI decreased Insomnia Severity Index (ISI) for people with low back pain (LBP) (pooled mean difference: -6.78/28, 95% confidence interval (95% CI): [-9.47, -4.09], I2 = 40%) and OA (-2.41, [-4.19, -0.63], 0%). However ESI decreased pain for people with LBP (pooled mean difference: visual analogue scale (VAS) -12.77/100, 95% CI: [-17.57, -7.97], I2 = 0%), but not OA (-2.32, [-7.18, 2.54], 27%). CONCLUSION: ESI appeared to improve sleep and pain for people with LBP, and sleep for people with OA. However more vigorous studies need to be conducted.