Risk of age-related macular degeneration in men receiving 5α-reductase inhibitors: a population-based cohort study.

Background: Recent studies suggest that 5α-reductase inhibitors (5ARIs) for benign prostate hyperplasia (BPH) result in abnormal retinal anatomical alteration.
Objective: To compare age-related macular degeneration (AMD) incidence in BPH patients receiving 5ARIs or tamsulosin.
Design: Retrospective, population-based cohort study using new-user and active-comparator design.
Setting: General population.
Subjects: Males with BPH, newly receiving 5ARIs or tamsulosin from 2010 to 2018.
Methods: Data were extracted from Taiwan's National Health Insurance Research Database. We used Cox proportional hazards model with 1:4 propensity score (PS) matching, based on intention-to-treat analysis to determine the risk of incident AMD. Sensitivity analyses included an as-treated approach and weighting-based PS methods. We also separately reported the risks of incident AMD in patients receiving finasteride and dutasteride to determine risk differences among different 5ARIs.
Results: We included 13 586 5ARIs users (mean age: 69 years) and 54 344 tamsulosin users (mean age: 68.37 years). After a mean follow-up of 3.7 years, no differences were observed in the risk of incident AMD between 5ARIs and tamsulosin users [hazard ratio (HR): 1.06; 95% confidence intervals (95% CI): 0.98-1.15], with similar results from sensitivity analyses. However, increased risk of incident age-related macular degeneration was observed in patients receiving dutasteride [HR: 1.13; 95% CI: 1.02-1.25], but not in those receiving finasteride [HR: 0.99; 95% CI: 0.87-1.12], in the subgroup analyses.
Conclusions: We found no difference between 5ARIs and tamsulosin regarding the incidence of AMD in BPH patients. However, the risk profiles for AMD differed slightly between dutasteride and finasteride, suggesting that the potency of androgen inhibition is a factor related to AMD incidence.
(© The Author(s) 2024. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved. For permissions, please email: journals.permissions@oup.com.)