It's a public health mantra that regular exercise can benefit rheumatologic conditions, but a close look at the latest evidence shows considerable gaps in that knowledge. Overall, the studies back exercise programs as benefitcial. But there is still little good evidence about which kind of exercise (aerobic versus weight training, for instance) is best for which conditions, according to a review of systematic reviews created or updated since 2007.
For two common conditions, knee osteoarthritis and low back pain, solid evidence shows that structured exercise programs provide a "small to modest" benefit in pain relief and physical function, write the rheumatologists from Norway who have published their review in BMC Medicine. For other inflammatory joint conditions such as rheumatoid arthritis or ankylosing spondylitis, not enough good studies have been conducted to settle the question.
Only one previous overview of systematic reviews on exercise in musculoskeletal conditions has been published, in 2007, and it did not include inflammatory joint diseases. Thus these authors from the National Resource Centre for Rehabilitation in Rheumatology in Oslo chose to analyze more recent systematic reviews and to include rheumatologic conditions.
They assessed only two outcomes, pain and physical function, ignoring evidence about general heatlh, cardiovascular prevention, and related issues such as work-related disability or quality of life.
Among 224 trials with 24,059 subjects, they found 32 that focused on knee osteoarthritis, involving 3,600 patients. This evidence was robust enough to support the conclusion that the treatment effect increases with the number of exercise sessions (though not robust enough to prove a clinical benefit). Pooled data for conditions such as low back pain and fibromyalgia were also sufficient to document a positive effect.
For other conditions such as rheumatoid arthritis and ankylosing spondylitis, only one or two trials have compared exercise therapy with moderate exercise or non-exercise interventions, among only 50 to 150 patients each. The evidence of a benefit is also not strong for neck pain.
Except for osteoporosis, it was not possible to detect an effect on pathogenesis for any of the musculoskeletal conditions. Therefore we still have no idea why exercise relieves pain when it does, the authors write. Nor do we have a good answer to the most interesting clinical question: Which exercise regimen for which patient?