Weight Loss Program Eases OA Knee Pain ... But Not Much

Overweight and obese patients with knee osteoarthritis assigned to a diet and exercise program in a randomized trial had greater reductions in pain than a control group, but the difference was small, researchers said.

After 18 months, pain scores on the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) declined an average 2.5 points in the intervention group from a baseline mean of 7.5, whereas the control group had a 2.0-point decline from 7.5 at baseline, according to Stephen Messier, PhD, of Wake Forest University in Winston-Salem, North Carolina, and colleagues.

The difference was 0.6 points after adjustment for covariates (95% CI 0.1-1.0), Messier's group reported in JAMA. But while the difference was statistically significant, its clinical significance was "uncertain," the investigators acknowledged. They suggested that social activities and information provided to the control group could have had "a positive effect on clinical outcomes."

But it's also uncertain whether the absolute changes from baseline in either group were clinically meaningful either. Previous studies to establish a "minimal clinically important difference" (MCID) for WOMAC pain scores provided set the threshold much higher than 2.5 points.

One study in 2021 put the MCID at 4.2 points for osteoarthritis knee pain. Another, from 2018, yielded an estimate of 11 points. A systematic review in 2019 found MCIDs ranging from 13.3 to 36.0. Participants in those studies were slated for knee replacement surgery, whereas those in the new trial had generally less severe disease and less pain. In this population, a small absolute reduction in pain might be more noticeable.

Messier and colleagues explained that the trial was intended to examine whether diet and exercise programs that had tested well in academic settings can also work in routine clinical practices. Dubbed WE-CAN, the trial enrolled 823 community-dwelling osteoarthritis patients age 50 and older (mean age 65) in North Carolina who had obesity or overweight (BMI ≥27; mean 36.8). They were randomized in equal numbers to the intervention or control program. Change in WOMAC pain score was the primary outcome.

The intervention consisted of dietary instruction in seven sessions (three in groups and four individual), plus 1-hour group exercise sessions provided at community facilities, 3 days per week over the 18-month trial. Intervention-group participants could also opt to receive Lean Shake meal-replacement powders to substitute for 1-2 regular meals per day.

Control participants met five times during this period for 1-hour discussions about diet and health, with written materials handed out; the control group also had telephone follow-ups every other month. In both groups, everything was provided gratis to participants.

Weight loss was clearly greater with the intervention: 7.7 kg (16.9 lb) versus 1.7 kg (3.7 lb), for a mean difference of 6.0 kg (95% CI 4.7-7.3). Most secondary outcomes such as 6-minute walk distance and WOMAC functional score also favored the diet-and-exercise program. One exception was use of pain medications, which didn't decline noticeably in either group.

Criticism from Accompanying Editorial

An accompanying editorial by Rena Wing, PhD, of Brown University in Providence, Rhode Island, criticized the intervention as a nonstarter from the beginning, if the goal was to identify a workable diet-and-exercise regimen for the real world.

"It is unlikely that this type of intensive intervention could be used in a pragmatic trial because the costs of the interventionist contact, meal replacements, and exercise sessions would be prohibitive," wrote Wing, who is also director of the Weight Control & Diabetes Research Center at the Miriam Hospital in Providence. She was also unhappy that data were insufficient to determine which component(s) of the intervention were most effective for inducing weight loss.

Wing's assessment of the pain reduction reported in the trial was harsh, too, declaring it "not clinically significant" despite the seemingly important degree of weight loss.

"That osteoarthritis pain was reduced in both groups raises concerns about potential efficacy of the attention and nutrition education provided to the control group and the use of a self-report measure as the primary outcome," Wing cautioned. She pointed out that earlier studies had included objective radiographic outcomes, whereas the current trial didn't even require x-rays to confirm the osteoarthritis diagnosis.

Messier and colleagues acknowledged the lack of radiographic measures as a limitation, along with the fact that participants didn't have to pay for anything in the trial.

Comment