Effect of exercise on pain processing and motor output in people with knee osteoarthritis: a systematic review and meta-analysis

Abstract

Guidelines recommend exercise as a core treatment for knee osteoarthritis. However, it is unclear how exercise affects measures of pain processing and motor function. The aim was to evaluate the effect of exercise on measures of pain processing and motor function in people with knee osteoarthritis. We searched five electronic databases (MEDLINE, EMBASE, CINAHL, SCOPUS and Cochrane Central Register of Controlled Trials) for studies on knee osteoarthritis, of any design, evaluating pain processing and motor function before and after exercise. Data were pooled with random-effects meta-analysis. Study quality was assessed using the Downs and Black and quality of evidence was assessed using the GRADE. Eighteen studies were eligible and 16 were included. Following acute exercise, pressure pain threshold increased local to the study limb (standardised mean difference [95% confidence interval (CI)] 0.26, [0.02, 0.51], n=159 from 5 studies), but there was no statistically significant change remote from the study limb (0.09, [-0.11, 0.29], n=90 from 4 studies). Following an exercise program (range 5-12 weeks) there were no statistically significant changes in pressure pain threshold (local 0.23, [-0.01, 0.47], n=218 from 8 studies; remote 0.33 [-0.13, 0.79], n=76 from 4 studies), temporal pain summation (0.38 [-0.08, 0.85], n=122 from 3 studies) or voluntary quadriceps muscle activation (4.23% [-1.84 to 10.30], n=139 from 4 studies). Very-low quality evidence suggests that pressure pain threshold increases following acute exercise. Very-low quality evidence suggests that pressure pain threshold, temporal pain summation or voluntary quadriceps activation do not change statistically significantly following exercise programs.