Causal associations of obesity related anthropometric indicators and body compositions with knee and hip arthritis: A large-scale genetic correlation study.

Backgrounds: Epidemiological studies have repeatedly investigated the association between obesity related anthropometric indicators and body compositions and osteoarthritis (OA). However, the results have remained inconsistent. This work aimed to investigate the genetic correlation and causal associations of obesity related anthropometric indicators and body compositions with knee and hip OA. Methods: Single-nucleotide polymorphisms associated with the exposures were searched from the recent genome-wide association studies (GWAS) to obtain full statistics. Summary-level results of knee and hip OA were from the UK Biobank and arcOGEN. First, linkage disequilibrium score regression (LD score regression) was applied to detect the genetic correlation (rg). We further performed a series of sensitivity analyses as validation of primary mendelian randomization (MR) results and the specific evidence of potential causal effects was defined. Results: We found that genetic components in OA had significant correlation with obesity related traits, except waist-to-hip ratio. In the univariable MR analysis, with the exception of waist-to-hip ratio, obesity related anthropometric indicators were causally associated with increased risks of knee and hip OA. For obesity related body compositions, higher fat-free mass in arm, leg, and whole body increased the risk of knee OA but only fat-free mass in leg showed a significant association with hip OA. Meanwhile trunk fat mass and trunk fat percentage, were associated with knee but not with hip OA. Higher fat mass, and fat percentage in arm, leg, and whole body increased the risk of both knee and hip OA. After adjusting for BMI, the multivariable MR showed maintained results in knee OA. However, in hip OA, only fat mass and fat-free mass in arm, leg, trunk and whole body were significantly associated with the risk of hip OA. Conclusion: The present study suggests genetic evidence for certain causal associations of obesity related anthropometric indicators and body compositions with knee and hip OA, which may provide important insights for the prevention and treatment on OA.