We determine whether there is a relationship between the number of different lower-limb resistance exercises prescribed in a program and outcomes for people with knee osteoarthritis. We used a systematic review with meta-regression. We searched the Cochrane Central Register of Controlled Trials, MEDLINE, and Embase up to January 4, 2024. We included randomized controlled trials that evaluated land-based resistance exercise for knee osteoarthritis compared with nonexercise interventions. We conducted meta-regressions between number of different exercises prescribed and standardized mean differences (SMDs) for pain and function. Covariates (intervention duration, frequency per week, use of resistance exercise machine[s], and comparator type) were applied to attempt to reduce between-study heterogeneity. Forty-four trials (3,364 participants) were included. The number of resistance exercises ranged from 1 to 12 (mean ± SD 5.0 ± 3.0). Meta-regression showed no relationship between the number of prescribed exercises and change in pain (slope coefficient: -0.04 SMD units [95% confidence interval {95% CI} -0.14 to 0.05]) or self-reported function (SMD -0.04 [95% CI -0.12 to 0.05]). There was substantial heterogeneity and evidence of publication bias. However, even after removing 31 trials that had overall unclear/high risk of bias, there was no change in relationships. There was no relationship between the number of different lower-limb resistance exercises prescribed in a program and change in knee pain or self-reported function. However, given that we were unable to account for all differences in program intensity, progression, and adherence, as well as the heterogeneity and overall low quality of included studies, our results should be interpreted with caution.
Read full abstract