The mediation effects of malalignment on the relation of sex to the risk of incidence tibiofemoral osteoarthritis

Abstract

ObjectiveTo investigate to what extent the higher risk of tibiofemoral radiographic osteoarthritis (TFROA) in females vs. males can be explained by knee malalignment. DesignUsing data from Multicenter Osteoarthritis Study (MOST) and Osteoarthritis Initiative (OAI), we examined the relation of sex to the incident medial and lateral TFROA and performed mediation analyses to assess to what extent varus and valgus malalignments account for sex differences in the incident medial or lateral TFROA. ResultsOf the 3,462 knees without medial and lateral TFROA in MOST, the 7-year risks of medial and lateral TFROA were 16.9% and 10.0% in females, and 15.8% and 4.2% in males, respectively. Females had 2.31-fold (95% CI: 1.73 to 3.08) higher incident lateral TFROA than males, and the relative risk (RR) of the indirect effect of sex on lateral TFROA through valgus malalignment was 1.15 (95% CI:1.09 to 1.20), accounting for 23% of its total effect on lateral TFROA. In OAI (n=3,095 knees), females had 1.54-fold (95% CI:1.15 to 2.04) higher incident lateral TFROA than males, and RR of the indirect effect of sex on lateral TFROA through valgus malalignment was 1.10 (95% CI:1.04 to 1.21), accounting for 26% of its total effect on lateral TFROA. No apparent sex difference in the incident medial TFROA was found in MOST (RR=1.05, 95% CI: 0.89 to 1.25) or OAI (RR=1.02, 95% CI: 0.84 to 1.19). ConclusionFemales had a higher risk of developing lateral TFROA than males; however, valgus malalignment only modestly explained such a difference.