Abstract
Ankylosing spondylitis (AS) and thyroid dysfunction are prevalent conditions, with observational studies suggesting a possible association. However, the causal relationship remains unclear. This study employed two-sample Mendelian randomization (MR) to evaluate the bidirectional causal association between AS and thyroid dysfunction using genetic summary data. The inverse variance weighted (IVW) method was the primary analytical approach, with multivariate MR (MVMR) adjusting for body mass index (BMI) and daily smoking quantity. A series of sensitivity analyses and visualization were performed to test the stability and reliability of causal estimates. Genetically predicted AS increased hyperthyroidism risk (OR = 1.0006, 95% CI = 1.0001-1.0011; P = .018), but not hypothyroidism (P = .266). No causal effect was found of hyperthyroidism or hypothyroidism on AS. In MVMR, the association between AS and hyperthyroidism remained significant after adjusting for covariates (P < .01). Sensitivity analyses confirmed the robustness of these findings. This MR study provides evidence for a causal effect of AS on hyperthyroidism, but not hypothyroidism or reverse causation. Further research should explore underlying mechanisms.