Knee Osteoarthritis Pain Improved With Stable Supportive Footwear

woman tying shoelaces, shoes, exercise
Researchers sought to evaluate the effectiveness of flat flexible shoes for reducing symptoms of knee osteoarthritis compared with stable supportive shoes.

Stable supportive shoes improved knee pain compared with flat flexible shoes among patients with moderate to severe knee osteoarthritis (OA), according to research results published in the Annals of Internal Medicine.

Current knee OA guidelines recommend that patients wear “appropriate” footwear, however, there is a lack of high quality randomized controlled trials and a recent systematic review could not conclude which footwear is appropriate for this patient population. In the current study, researchers sought to evaluate the effectiveness of flat flexible shoes for reducing symptoms of knee OA compared with stable supportive shoes.

Researchers conducted a 2-group, pragmatic, comparative effectiveness, superiority randomized controlled trial of people with knee OA in Melbourne, Australia. Participants were randomly assigned in a 1:1 ratio to either the intervention group, which wore preselected flat flexible shoes, or to the comparator group, which wore preselected stable supportive shoes. Baseline and final assessments were conducted via questionnaires administered 6 months apart.

In total, 164 adults aged 50 years or older with grade 3 or 4 radiographic disease severity were enrolled in the study from August 2018 to November 2019. Between-group participant characteristics were similar, as were treatment expectations.

Mean shoe wear of 8.2 (±3.3) and 8.5 (±2.6) hours per day were reported in the flat flexible and stable supportive shoe groups, respectively. Over the 6-month study period, slightly more participants in the stable supportive shoe group were adherent (70 vs 62 participants).

More participants in the flat flexible shoe group reported adverse events (32% vs 15%), with the relative risk for an adverse event of 0.46 in the stable supportive shoe group compared with the flat flexible shoe group (95% CI, 0.25-0.84). The risk difference was -0.17 (95% CI, -0.30 to -0.05) in favor of stable supportive shoes.

A Poisson regression demonstrated that the adverse event rate in the supportive stable shoe group was 0.36 times that of the flat flexible shoe group (95% CI, 0.19-0.70).

In the flat flexible shoe group, 4 participants discontinued wear due to foot or knee pain, while only 1 discontinuation was noted in the stable supportive shoe group. No between-group difference was seen in terms of the use of cointerventions.

Investigators found no evidence that flat flexible shoes were superior to stable supportive shoes for the primary outcomes (6-month change in overall average knee pain and physical function), but the between-group mean difference in pain change “favored stable supportive shoes” (mean difference, 1.1 units; 95% CI, 0.5-1.8 units). This mean difference was not associated with a change in function (2.3 units; 95% CI, -0.9 to 5.5 units). More participants in the stable supportive shoe group achieved minimal clinically important difference in pain (58% vs 40%).

Study limitations include the lack of a usual shoe control group, a lack of generalizability to populations outside of English-speaking community volunteers, and the evaluation of footwear in isolation rather than in combination with other interventions. Additionally, findings cannot be generalized to people with mild knee OA.

“To our knowledge, our study provides the first [randomized controlled trial] evidence to suggest that stable supportive shoes may be a useful self-management strategy in this subgroup of patients with knee osteoarthritis, supporting clinical practice guideline recommendations that, to date, have been based solely on expert opinion,” the researchers concluded.


Paterson KL, Bennell KL, Campbell PK, et al. The effect of flat flexible versus stable supportive shoes on knee osteoarthritis symptoms: a randomized trial. Published online January 12, 2021. Ann Intern Med. doi:10.7326/M20-6321