There is strong evidence that higher bone mineral density (BMD) is a risk factor for osteoarthritis (OA), although the association with OA progression is unclear. It is uncertain if observational associations reflect a causal effect of bone on joint deterioration, or shared underlying biology.
I examined the role of BMD in OA progression by determining change in radiographic sub-phenotypes (osteophytes, joint space narrowing) at the hip and knee in individuals with high bone mass (HBM, L1 and/or total hip Z-score>+3.2), comparing progression to their relatives without HBM (Chapters 6,7). HBM individuals had increased progression of radiographic OA sub-phenotypes at both joints, compared to their relatives. Metabolomics analysis did not identify any metabolic traits, and thus metabolic pathways, which could mediate the association between HBM and OA (Chapter 8).
I next performed multivariable Mendelian randomization (MVMR) in UK Biobank to determine if BMD exerts a causal effect on OA, independent of confounding by body mass index (BMI) (Chapter 9). MVMR identified a BMI-independent causal effect of BMD on hip and knee OA (26% and 36% increased odds per SD increase, respectively).
Subsequently, I analysed the underlying biological pathways contributing to both BMD and OA, by calculating the proportion of variance in hospital-diagnosed hip and knee OA explained by genetic variation in the Wnt signalling, TGFβ superfamily and osteoclast differentiation pathways (Chapter 10). This analysis provided evidence for a BMD-independent contribution of these pathways to OA. Further MVMR analyses identified a BMI-independent causal effect of circulating IGF-1 on risk of hospital-diagnosed hip and knee OA (49% and 22% increased odds per SD increase, respectively, Chapter 11).
This work suggests a role of BMD in structural OA progression at the hip and knee, reflecting a direct role of bone in joint deterioration, as well as shared underlying pathways, such as the IGF-1 axis.