The idea made so much sense that it was rarely questioned: exercise to strengthen muscles around the knee helps patients with osteoarthritis, making it easier and less painful to move the inflamed joint.
Nearly 40 percent of Americans over age 65 have knee osteoarthritis, and tens of millions of patients have been instructed to do these exercises. Indeed, the American College of Rheumatology and the Arthritis Foundation routinely advise strength training to improve symptoms.
Stephen Messier, a professor of biomechanics at Wake Forest University, believed in the guidance. But he decided to put the prescription to the test in a rigorous 18-month clinical trial involving 377 participants. The verdict appeared in a study published this week in JAMA: Strength training did not seem to help knee pain.
One group lifted heavy weights three times a week, while another group tried moderate strength training. A third group was counseled on “healthy living” and given instructions on foot care, nutrition, managing medications and better sleep practices.
Dr. Messier had expected that the group performing the heavy lifting would fare best and that the participants who received only counseling would see no improvement in knee pain. But the outcomes were the same in all three groups. Everyone reported slightly less pain, including those who had received only counseling.
You might expect some easing in pain in the patients who exercised. But why would those who didn’t exercise also report an improvement? “It’s an interesting dilemma we’ve been put in,” Dr. Messier said.
A simple placebo effect might explain why they felt better, he said. Or it could be something scientists call a regression to the mean: Arthritis symptoms tend to surge and subside, and people tend to seek out treatments when the pain is at its peak. When it declines, as it would have anyway, they ascribe the improvement to the treatment.
“The natural history of osteoarthritis of the knee includes waxing and waning of symptoms,” said Dr. Adolph Yates, vice chair of orthopedic surgery at the University of Pittsburgh School of Medicine, who was not associated with the study. “It is what makes studying osteoarthritis of the knee interventions difficult.”
Dr. David Felson, a professor of medicine at Boston University, argued that the study did not find strength training was useless. But instead, the trial showed that very aggressive strength training was not helpful and might actually be harmful, he said, especially if the arthritic knees are bowed inward or outward, as is common.
Strong muscles can act like a vise, putting pressure on tiny areas of the knee that bear most of the load when we’re walking. When Dr. Felson scrutinized the study’s data, he saw signs that the high-intensity group experienced slightly more pain and worse functioning.
Patients tend to resist advice to exercise at all, said Dr. Robert Marx, a professor of orthopedic surgery at Weill Cornell Medical College in New York City: “They want a reason not to exercise, asking: ‘Will it make my arthritis better? Will it make my X-rays better?’”
He tells them that the answer to their questions is no, but that exercise stabilizes the joints. Although it is not as effective against pain as anti-inflammatory drugs, “it is one piece of arthritis treatment.”
For Dr. Messier, who has been researching arthritis and exercise for over 30 years, the new findings are a bit of a departure. His first study, published in 1997 in JAMA, found that exercise groups had less pain in the end than the control group, but that was not really because the participants improved. It was because those in the control group got worse.
He also noticed that half of the participants in his study were overweight or obese. “What if we added weight loss to exercise?” he asked.
He tried that in another study, which was published in JAMA in 2013, that showed that a combination of weight loss and exercise decreased pain to a greater extent than either alone.
But he had long wondered whether the intensity of strength training mattered. In previous studies, participants had used weights that fell far short of what they could actually lift; the trials lasted just six to 24 weeks, and patients showed only modest improvements in pain and functioning.
Despite the new, unexpected results, Dr. Messier still urges patients to exercise, saying it can stave off an inevitable decline in muscle strength and mobility. But now it seems clear there is no particular advantage to strength training with heavy weights instead of a moderate-intensity routine with more repetitions and lighter weights.
Arthritis, he noted, is a chronic degenerative disease of the entire joint. “There are a lot of things going on,” Dr. Messier said. “It’s not just degradation of the cartilage.”
But, he added, he believes the best non-pharmaceutical intervention for knee arthritis pain is a 10 percent weight loss and moderate exercise.
Dr. Messier is now planning that his next study will combine weight loss with exercise in people at risk for knee osteoarthritis, in the hopes of preventing the onset of this disabling disease.