Background/Purpose: Knee osteoarthritis (OA) significantly affects physical function, making function a core outcome in OA research. The Osteoarthritis Research Society International recommends the six-minute walk test (6MWT) as a performance-based measure of physical function in knee OA. Originally developed for cardiopulmonary conditions, the 6MWT is now used in OA populations, though its validity in this context remains unclear. Limited and conflicting evidence regarding its correlation with other function measures in OA raises questions about its utility and interpretation. This study aimed to evaluate the construct validity of the 6MWT in individuals with symptomatic knee OA and identify patient factors influencing six-minute walk distance (6MWD).

Methods: In this cross-sectional study within the BEST-Knee cohort, participants with knee OA completed standardized questionnaires and the 6MWT one month before total knee arthroplasty. Construct validity was assessed per COSMIN guidelines. Convergent validity was tested through hypothesized moderate-to-strong Spearman correlations (r > 0.4) between 6MWD and patient-reported physical function (KOOS-PS), pain (WOMAC pain subscale), and perceived walking ability (in blocks). Discriminant validity was tested examining correlation between 6MWD and the Lubben Social Network Scale, a measure unrelated to physical function. Known-groups validity compared 6MWD in participants who used gait aids versus not. A multivariable linear regression model assessed the association between demographic (age, sex), OA-related (WOMAC pain, KOOS-PS), biomedical (symptomatic joints, comorbidities, obesity, back pain), and psychosocial (PHQ-8, coping, social network, education) factors with 6MWD.

Results: Among 278 participants (mean age 67 years, 65% female; Table 1), greater perceived walking difficulty, higher WOMAC pain, worse KOOS-PS, and gait aid use were associated with lower 6MWD (Figure 1). Correlations were moderate with perceived walking ability (ρ = -0.451, p < 0.001) and fair with WOMAC pain (ρ = -0.323, p < 0.001) and KOOS-PS (ρ = -0.328, p < 0.001). No correlation was seen with the Lubben scale (ρ = 0.067, p = 0.281) as hypothesized. In multivariable analysis, WOMAC pain, KOOS-PS, age, sex, obesity, comorbidities, education level, and arthritis coping were associated with 6MWD (Table 2).

Conclusion: While 6MWD declines with increasing OA symptom severity, its correlations with patient-reported function are only fair-to-moderate so we cannot confirm construct validity. Performance on the 6MWT is influenced by multiple biomedical and psychosocial factors which should be accounted for in knee OA research.

WOMAC Western Ontario and McMaster Universities Osteoarthritis Index, KOOS Knee injury and osteoarthritis outcome score, PHQ-8 Patient health questionnaire – 8.

Bars represent mean six-minute walk distance (6MWD) for each category. Error bars indicate standard deviations. Mean (standard deviation) and sample size (N) indicated for each bar.

Positive beta-estimates indicate greater walking distance associated with higher values or presence of the respective independent variable; negative estimates indicate reduced distance. Confidence intervals that do not include zero suggest statistically significant associations. KOOS-PS = Knee Injury and Osteoarthritis Outcome Score–Physical Function Shortform; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; PHQ8 = Patient Health Questionnaire–8. Obesity was defined as a BMI ≥30 kg/m2

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/construct-validity-of-the-six-minute-walk-test-in-knee-osteoarthritis/