The impact of knee positioning on joint space width change: Data from the Osteoarthritis Initiative

Quantify the longitudinal effect of knee positioning on radiographic joint space width (JSW) in knee OA.

This longitudinal cohort study analyzed 1612 knees (806 participants) from the Osteoarthritis Initiative with baseline Kellgren-Lawrence grade = 0 and 12-month follow-up. Images obtained via a standardized fixed-flexion protocol measured minimum JSW (mJSW), fixed-location JSW (fJSW), and a variable related to knee positioning (X_rim). Generalized estimating equation (GEE) models, adjusted for age, gender, race, and BMI, examined associations between changes in ΔfJSW or ΔmJSW and ΔX_rim. A secondary case-control analysis of 69 participants KLG 1-3 over 4 years examined the effect of poor but consistent X_rim values on JSW responsiveness, using standardized response means (SRM).

GEE models showed a significant positive association between ΔXrim and both ΔfJSW (95% CI: 0.07-0.15) and ΔmJSW (95% CI: 0.05-0.13). The association between ΔX_rim and ΔfJSW was significantly stronger in females, with a significant ΔX_rim × gender interaction term (95% CI: 0.03-0.13), with no significant gender difference identified for the association between ΔXrim and ΔmJSW. With ΔfJSW as the outcome, estimated slope for ΔX_rim was 0.19 (95% CI, 0.15-0.22) in females and 0.11 (95% CI, 0.07-0.15) in males. No effect modification by age, race or BMI identified. Secondary analysis revealed fJSW responsiveness was unaffected by poor but consistent positioning (cases: -0.80, controls: -0.76), while mJSW was (cases: -0.81, controls: -0.58).

Inconsistent knee positioning affects JSW measurement. Poor but consistent positioning has a minimal effect on fJSW responsiveness but a larger impact on mJSW.