Exercise for Knee Osteoarthritis

How effective is exercise for knee osteoarthritis?

Compared with control or placebo, exercise improves patient-reported treatment success over 2 years of follow-up.¹ (Strength of Recommendation [SOR]: B, inconsistent or limited-quality patient-oriented evidence.)

Adding exercise to a cointervention (eg, pain coping skills, weight loss program) increases patient-reported treatment success compared with the cointervention alone.¹ (SOR: B, inconsistent or limited-quality patient-oriented evidence.)

Knee osteoarthritis is a degenerative condition that affects the entire synovial joint, including the cartilage, subchondral bone, and menisci. Approximately 23% of people 40 years and older in the United States have knee osteoarthritis.² Osteoarthritis commonly coexists with obesity, cardiovascular disease, diabetes, and mental health conditions. Diagnosis is primarily clinical, based on symptoms such as activity-related joint pain and brief morning stiffness in people 45 years and older.² Exercise may benefit patients through weight loss, increased muscle strength for improved joint function, and reduced inflammation via lower cytokine levels. In addition, exercise may promote cartilage health through inflammatory pathways.¹

This Cochrane review evaluated 139 randomized controlled trials with a total of 12,468 participants (ages 41–81 years; mean body mass index of 22.4–37.9 kg/m²) with knee osteoarthritis diagnosed using clinical or radiographic criteria (presence of osteophytes, joint space narrowing, and subchondral sclerosis using the Kellgren-Lawrence grading system).¹ Interventions were categorized into three groups: (1) exercise vs attention control or placebo; (2) exercise vs no treatment, usual care, or limited education; and (3) exercise with a cointervention (eg, diet, education, physical therapy) vs the cointervention alone. Exercise included supervised and unsupervised land-based exercises such as strengthening (including neuromuscular), aerobics, mind-body interventions (eg, yoga, tai chi), stretching, and balance and proprioception training. Intervention durations ranged from 2 weeks to 2 years.

Outcomes included pain, physical function, quality of life, perceived treatment success, and adverse events.¹ A 0- to 100-point scale was used to measure improvement; minimal clinically important differences were defined as 12 points for pain, 13 for function, and 15 for quality of life. There were statistically significant improvements that did not meet the minimal clinically important differences threshold for reporting.¹

In the 30 trials comparing exercise with attention control or placebo, exercise resulted in a small improvement in pain (mean difference = 8.7 points; control group mean = 44.4 points; low-certainty evidence). Exercise also resulted in a small improvement in function (mean difference = 11.3 points; control mean = 35.8 points; moderate-certainty evidence) and little to no improvement in quality of life (mean difference = 6.1 points; control mean = 48.4 points; moderate-certainty evidence). Exercise likely increased patient-reported treatment success and did not increase study withdrawals. Compared with control or placebo, exercise improved patient-reported treatment success after 4 weeks to 2 years (relative risk = 1.46; 95% CI, 1.11–1.92; number needed to treat = 8; moderate-certainty evidence).

Compared with no treatment or usual care (60 trials), exercise improved pain (mean difference = 13.1 points; low-certainty evidence), improved function (mean difference = 12.5 points; moderate-certainty evidence), and slightly improved quality of life (mean difference = 5.37 points; moderate-certainty evidence). There was no significant difference in patient-reported treatment success and a possible increase in adverse events (low-certainty evidence).