Land-based exercise is the most studied form. A recent Cochrane review summarized the data from 32 RCTs involving almost 3,800 people with symptoms of knee OA of varying duration Subjects were allocated to a group participating in land-based exercise of varying modes, durations, and intensities or to a nonexercise group. A mix of forms of exercise (such as aerobic, strengthening, flexibility) was employed, but aquatic exercise programs were not included.
Subjects in the exercise group demonstrated small but clinically relevant short-term benefits in pain (standardized mean difference [SMD], 0.40; 95% CI, 0.30-0.50) and in physical function (SMD, 0.37; 95% CI, 0.25-0.49). Subgroup analyses indicated that exercise programs for quadriceps and lower-limb strengthening (weight training or the use of resistance bands, for example) provided small but significant self-reported benefits in function and pain. General strengthening and aerobic programs were less effective. Regarding exercise dosage, the authors divided the study data into supervised sessions of 12 or more or less than 12, regardless of program duration, and found that 12 or more supervised sessions yielded greater benefit