Abstract / Summary
Background/Objectives: Delirium is a common and serious neuropsychiatric complication in hospitalized and critically ill patients and is associated with prolonged hospitalization and increased mortality. Previous meta-analyses found no clear association between statin use and delirium; however, newly published studies and emerging mortality data warrant re-evaluation. Methods: This updated systematic review and meta-analysis extended a previous meta-analysis and re-evaluated the association between statin use and delirium risk. PubMed, Embase, and the Cochrane Library were searched for eligible studies published from January 2022 through October 2025. Study-level estimates were synthesized using random-effects models. Because ORs, RRs, and HRs are not directly interchangeable, the combined synthesis was considered exploratory. Sensitivity analyses included an OR-only dataset, exclusion of the cross-sectional study, REML models with Hartung–Knapp adjustment, and restriction to observational studies reporting multivariable-adjusted ORs. Results: Nineteen studies (5 randomized controlled trials and 14 observational studies) involving 440,737 participants were included. The exploratory combined analysis yielded an OR of 0.79 (95% confidence interval [CI] 0.66–0.95; p = 0.01; I2 = 90%). Results differed by study design: randomized trials yielded a lower pooled point estimate (OR 0.41, 95% CI 0.20–0.82; I2 = 58%), whereas the pooled association among observational studies was not statistically significant (OR 0.86, 95% CI 0.71–1.03; I2 = 91%; p for subgroup difference = 0.04). However, in the three randomized trials contributing to incident delirium, REML with Hartung–Knapp adjustment substantially widened the confidence interval (OR 0.41, 95% CI 0.09–1.86; p = 0.126). In the overall analysis, REML–Hartung–Knapp inference also increased uncertainty (OR 0.79, 95% CI 0.62–1.00; p = 0.047), as did restriction to multivariable-adjusted observational ORs (OR 0.76, 95% CI 0.57–1.01; p = 0.062). In four studies comprising 21,687 participants, statin use was associated with lower 30-day all-cause mortality under the conventional random-effects model (OR 0.52, 95% CI 0.35–0.79; p = 0.002; I2 = 88%); this estimate remained similar under REML with Hartung–Knapp adjustment (OR 0.52, 95% CI 0.32–0.84; p = 0.022). No significant association was observed for delirium-free days. Conclusions: The available evidence suggests a possible association between statin exposure and lower delirium occurrence, but substantial heterogeneity, differences between randomized and observational evidence, and sensitivity to analytic assumptions limit causal interpretation. The findings do not establish statins as a preventive treatment for delirium or prove a mortality benefit. Further adequately powered randomized trials with clearly defined exposure timing and standardized delirium assessment are warranted.