Impact of exercise on articular cartilage in people at risk of, or with established, knee osteoarthritis: a systematic review of randomised controlled trials

Bricca A, Juhl CB, et al. · British Journal of Sports Medicine

Nine trials contributing 14 comparisons, two in people at increased risk of knee OA and 12 in people who already had it. Among the at-risk comparisons one found no effect on cartilage defects and one a positive effect on glycosaminoglycans. Among those with OA, six found no effect on cartilage thickness, volume or defects; on glycosaminoglycans one was negative and one null; on collagen two were positive and two null. The authors conclude that 'Knee joint loading exercise seems to not be harmful for articular cartilage', and grade their own evidence low.

Population

9 RCTs (14 comparisons) in people at risk of or with knee OA

Intervention

Knee-loading exercise programs

Comparator

Non-exercise controls

Limitations

The reviewers deem the overall quality of evidence low, for indirectness and imprecision: few studies, few participants, and some interventions they judged to fall outside the therapeutic loading spectrum. Every included trial failed to report dropouts clearly or handled them inadequately. One comparison in people with OA reported a negative effect on glycosaminoglycans, so 'not harmful' summarises a small and mixed set rather than a uniform result, and nothing here addresses horizons beyond the trials' own.