The association between arthritis and insomnia risk: Evidence from NHANES 2005 to 2008 and Mendelian randomization (MR) analyses

Arthritis is a highly prevalent comorbidity linked to insomnia in older U.S. adults, yet the causal relationship remains unclear. This study investigated the association between arthritis and insomnia using data from the National Health and Nutrition Examination Survey (2005-2008) (N = 8173) and Mendelian randomization (MR). Weighted multivariable logistic regression, eXtreme Gradient Boosting, and generalized additive models assessed associations. MR and sensitivity analyses (inverse variance weighted, Steiger test) evaluated causality and robustness. Arthritis was significantly associated with insomnia risk (adjusted odds ratio = 1.479, 95% confidence interval: 1.132-1.932, P = .015), with eXtreme Gradient Boosting identifying arthritis as the strongest predictor and generalized additive models confirming a linear positive correlation. MR supported a causal effect (inverse variance weighted odds ratio = 1.054, 95% confidence interval: 1.007-1.103, P = .024), with no heterogeneity or pleiotropy (P > .05). The Steiger test validated unidirectional causality (arthritis → insomnia, P < .05). Arthritis is an independent risk factor for insomnia, providing critical insights for clinical prevention and management strategies.