Causal association between smoking behavior and the decreased risk of osteoarthritis: a Mendelian randomization

Young Ho Lee

This study aimed to examine whether smoking behavior is causally associated with osteoarthritis. A two-sample Mendelian randomization (MR) analysis using the inverse-variance weighted (IVW), weighted median, and MR-Egger regression methods was performed. We used the publicly available summary statistics datasets of smoking behavior genome-wide association studies (GWASs; n = 85,997) as an exposure, and a GWAS in 7410 patients with osteoarthritis in the arcOGEN study and 11,009 controls of European ancestry as an outcome. We selected four single nucleotide polymorphisms (SNPs) from GWASs of smoking behavior as instrumental variables (IVs) to improve inference. These SNPs were located at CHRNA3 (rs1051730), SLC25A5P5A9 (rs215614), CHRNB3 (rs6474412), and CYP2B6 (rs7260329). The IVW method showed evidence to support an inverse causal association between smoking behavior and osteoarthritis in the knee and hip (beta = -0.056, standard error [SE] = 0.027, p = 0.035). MR-Egger regression revealed that directional pleiotropy was unlikely to be biasing the result (intercept = -0.005; p = 0.848), but showed no causal association between smoking behavior and osteoarthritis (beta = -0.048, SE = 0.048, p = 0.427). However, the weighted median approach yielded evidence of a negative causal association between smoking behavior and osteoarthritis (beta = -0.056, SE = 0.028, p = 0.046). Cochran's Q test and the funnel plot indicated no evidence of heterogeneity between IV estimates based on the individual variants. The results of MR analysis support that smoking behavior was causally associated with a reduced risk of osteoarthritis.