Dropout from exercise interventions in adults with knee or hip osteoarthritis: a systematic review and meta-analysis

Objective To investigate the prevalence and moderators of dropout rates among adults with knee or hip osteoarthritis participating in exercise randomized controlled trials (RCTs). Data sources Two authors searched Embase, CINAHL, PsycARTICLES and PubMed up to 01/09/2023. Study selection We included RCTs of exercise interventions in people with knee or hip osteoarthritis that reported dropout rates. Data extraction Dropout rates from exercise and control conditions and exerciser/participant, provider, and design/implementation related moderators. Data synthesis In total 209 RCTs involving 277 exercise arms in 13,102 participants were included (mean age at study level=64 years; median prevalence of male participants=26.8%). The trim-and-fill-adjusted prevalence of dropout across all RCTs was 17.5% (95%CI = 16.7%-18.2%), which is comparable to dropout observed in control conditions (trim-and-fill-adjusted odds ratio=0.89; 95%CI=0.71-1.12, P=0.37). Higher prevalence of antidepressant use at study-level predicted higher dropout (R²=0.75, P=0.002, N RCTs=6, n exercisers=412). Supervision by an exercise professional was associated with lower dropout rates, with a trim-and-fill-adjusted rate of 13.2% (95%CI=11.7%-14.9%) compared to 20.8% without supervision (95%CI=18.3%-23.5%) (P<0.001). Conclusions Dropout rates for exercise in RCTs are comparable to control conditions, suggesting that exercise is a generally well-accepted intervention. However, interventions should be supervised by an exercise professional, such as a physiotherapist or exercise physiologist, to further minimize the risk of dropout. Health professionals should consider participants’ use of antidepressants as a risk factor for dropout from exercise.