Thirteen RCTs were included. Three studies assessed aerobic walking (n=156), nine assessed home-based quadriceps strengthening exercises (n=1,709) and one assessed both aerobic walking and home-based strengthening exercises (n=439).
The Jadad scores ranged from 1 to 3 (out of 5), with none of the trials reporting to be double-blinded.
Pain was reduced significantly more with aerobic exercise than with no exercise (WMD 0.52, 95% CI: 0.34, 0.70; 4 RCTs, n=449). Participants in the aerobic exercise group reported less disability than those in the no exercise group (WMD 0.46, 95% CI: 0.25, 0.67; 2 RCTs, n=385). No statistically significant heterogeneity was detected for either analysis.
Pain was reduced significantly more with home-based quadriceps strengthening exercises than with no exercise (WMD 0.32, 95% CI: 0.23, 0.42; 11 RCTs, n=2,004). One study, with a large effect size, was excluded from this meta-analysis due to concomitant medication and a sham exercise programme delivered in the control group. Participants in the quadriceps strengthening exercise group reported less disability than those in the no exercise group (WMD 0.32, 95% CI: 0.23, 0.41; 11 RCTs, n=2,004). No statistically significant heterogeneity was detected for either analysis.