Genetically predicted circulating linoleic acid levels and risk of osteoarthritis: a two-sample mendelian randomization study

Abstract

Abstract Objectives This study aimed to provide insight into the effect of genetically predicted linoleic acid (LA) levels on osteoarthritis (OA). Methods The LA dataset was obtained from the UK Biobank (UKBB) consortium and contained 114,999 samples. The OA discovery dataset was derived from MRC-IEU consortium and included 38,472 cases and 424,461 controls. The OA validation set was derived from a summary-level genome-wide association study (GWAS) and included 39,427 cases and 378,169 controls. Genetic variants strongly associated with LA (p < 5 × 10− 8) were extracted as instrumental variables (IVs). The inverse variance weighted (IVW) approach was adopted as the primary analysis method in this study. In addition, multiple sensitivity analysis methods were used to assess the reliability of our results. Results The IVW approach showed that circulating LA levels were negatively associated with OA risk in the discovery set (odds ratio (OR) = 0.993, 95% confidence interval (95% CI): 0.988–0.998, p = 0.011). A consistent result was obtained in the validation set (OR = 0.904, 95%CI: 0.845–0.967, p = 0.003). These results were validated by sensitivity analysis. Conclusion This study provides new evidence for the causal relationship between LA and OA, which provides new insights for the treatment of OA.

Keywords

• Osteoarthritis
• Linoleic acid
• Mendelian randomization