Sex differences in bone shape and knee osteoarthritis: the Osteoarthritis Initia

点击翻译按钮获取中文摘要

Incidence and prevalence of either radiographic or symptomatic osteoarthritis (OA) are much higher in women than men; however, the underlying causes for this sex difference in OA remain unknown. Several potential explanations have been proposed for the sex difference in OA, including differences in estrogen level, physical activity, and laxity or alignment, but each has only moderate supporting evidence, and none fully explains the observed sex differences. Recently, several investigators have proposed that aspects of bone shape are associated with an increased risk of incident OA and with severity of OA, based on anthropometric measures, cross-sectional findings, or in one case statistical shape modeling done on 24 knees. Our group demonstrated that several specific proximal femoral shapes, assessed by Active Shape Modeling (ASM), were associated with incident radiographic hip OA in elderly women. Others have also reported that ASM-based knee shape is associated with severity of radiographic knee OA. It is postulated that changes in bone shape occur prior to cartilage damage in the joint. In the proposed study, we will take the advantage of knee radiographs and magnetic resonance images (MRIs) collected annually for 5 years in the Osteoarthritis Initiative (OAI). We will measure femoral and tibial bone shape using an ASM method improved by our group. We will evaluate whether baseline bone shape is associated with incidence or progression of radiographic knee OA, pain and function and whether these associations differ by sex. We will apply group-based finite modeling methods to identify distinctive sub-groups that follow similar trajectories of bone shape change over time in women and men respectively and examine whether associations between various risk factors assessed at baseline (i.e., age, sex, body mass index, alignment, etc.) and membership of trajectories of bone shape differ by sex using a multinomial logistic regression model. Finally, we will evaluate the risk factors that determine sex-specific trajectories of bone shape change over time.