New study: Do diet and exercise actually help knee pain due to osteoarthritis?

The latest trial data are encouraging, if not overwhelming.

Knee pain affects millions of people. When knee pain is due to osteoarthritis, some people can expect decades of discomfort. Naturally, physicians want to be able to help.

Surgery can help some patients with moderate-to-severe osteoarthritis. But for milder cases, which can be quite debilitating for people, surgery has been found (somewhat infamously) to be no better than a placebo sham procedure. (Yes, they actually did that.) Also, the rate of adverse effects after surgery is not negligible, and some can be serious.

So, researchers and clinicians want to know if they can treat knee pain caused by osteoarthritis conservatively—for example, with diet and exercise. That’s especially true for pain that is mild or moderate; not so bad that surgery is needed, but bad enough to cause genuine reductions in quality of life. A new randomized trial in JAMA found that in older patients with obesity or elevated body mass indexes (i.e., overweight), diet and exercise appear to help. While the effect was quite modest (though statistically significant), in my view a closer look at the data are more ultimately more encouraging.

On a 20-point pain scale (higher is worse), after 18 months, self-reported knee pain scores among patients who were counseled to engage in “diet and exercise” (caloric restriction with meal replacements, and an hour of exercise three times per week) were around 5.0 versus 5.5 among control patients who received “attention control” (comprised of social interactions, nutrition and five 1-hour in-person health education sessions, and five in-person group meetings spaced over 15 months). While both groups improved from their baseline scores of around 7.5, and the statistical difference was notable (if p values mean anything to you), generally a difference of 1.6 points on this scale is considered clinically relevant. So, this result was not thrilling.

That said, if you dig deeper, I see reasons for optimism. First, in the diet and exercise group, patients lost an average of 17 pounds during the 18-month study, compared to 3.7 pounds in the control group. The benefits of this difference might be apparent in other body systems that were not measured here (including other types of pain like back pain, or even metabolic health). Plus, it looks like the diet and exercise group had substantially better mobility at the end of the trial.

Another secondary finding I found interesting was the rate of knee surgery and joint injections during the study in each group. The rate of surgery was around the same in both groups (reflecting the mild-to-moderate osteoarthritis most of the patients in the study had at baseline) but far fewer people in the diet and exercise group received joint injections (8.7%) to help manage their pain than among the control group (13.7%). If that difference was related to the diet and exercise, that’s good because it means that more invasive approaches were less necessary, thereby limiting the risk of adverse effects and keeping costs down. That said, people in the diet and exercise group did indeed have a higher rate of injury during the study. The fact that they fared better overall despite this, is good news.

Of course, interventions are only useful if they are able to be done. Not everyone who enrolled in the study actually adhered to the plan. In the first 6 months, most patients did what they had been randomized to do. But by the end of the study, a lot of fall-off had occurred. For example, only 26% of participants in the diet and exercise group were still consuming the meal replacements.

There are two ways to look at this. You could say, “well, if everyone had done what they were supposed to do, the results might look better.” Or, you could say, “these are the systemic results we can expect if we apply these lessons to all of our patients, because not everyone will stick with the program.” Both are true.

But if you’re a person with osteoarthritis and you think, “Sure, I could do the program they described, and I could stick with it for 18 months,” there would be reason for optimism. When the researchers checked what percent of participants achieved more than a 2 point improvement in pain (which they considered clinically relevant), they found reason for hope. Over 60% of participants in the diet and exercise group reported 2 or more points improvement in pain after 18 months versus under 50% in the control group.

Which leads to a final thought. The control group did not get “nothing.” They actually received fairly intense social interventions, which led to some improvements in pain over time. Assuming those social interventions are better than nothing (which is reasonable), I think diet and exercise are probably somewhat more effective in improving knee pain due to osteoarthritis than this new study suggests.

Do you have questions about this or other studies about knee pain and osteoarthritis? Ask below in the comments and I’ll get an expert to chime in if needed!