Association of current Statin use with incident age-related macular degeneration: using time-varying propensity...

Statins exhibit pleiotropic properties that may influence age-related macular degeneration (AMD) progression. However, since their prescription is predominant in older populations at high cardiovascular risk, a lack of robust approach to control for confounding by indication could partly explain controversial results in previous studies. We therefore analyzed the association of current statin use with incident AMD using time-varying propensity score-based inverse probability of treatment weighting (IPTW).

Analysis of data from the ALIENOR study, a 14-year French prospective population-based cohort of 963 adults aged 73 years and older in Bordeaux, followed every 2-3 years (2006-2020), with repeated assessments of statin use through a face-to-face questionnaire. AMD status was assessed using retinal multimodal imaging every 2-3 years (2006-2020). Incident intermediate and advanced AMD were analyzed separately. Combining time-varying IPTW balancing and covariates adjustment, we fitted time-dependent Cox proportional hazards models to estimate adjusted hazard ratios (aHRs [95% confidence intervals]) for incident AMD associated with current statin use (defined as using statin within the last three years prior to the current visit).

Over a median follow-up of 7 years, cumulative incidence was 36% and 10% among the participants analyzed for intermediate (n = 475) and advanced (n = 692) AMD, respectively. After accounting for confounding by indication, current statin use was not associated with increased hazards of intermediate (aHR: 1.02 [0.72-1.45]) and advanced (aHR: 1.08 [0.61-1.91]) AMD. Analyses of statin subtypes (lipophilic and hydrophilic) and all lipid-lowering agents consistently yielded non-significant association with the hazards of AMD. Findings were consistent across sensitivity analyses.

These findings suggest that, after controlling for confounding by indication, statins may not substantially influence the risk of developing intermediate or advanced AMD in older adults, despite the biological plausibility.