Is medication use associated with the incidence of early age-related maculopathy? Pooled findings from 3 continents

Objective To investigate whether there is an association between the use of medication and the incidence of early age-related maculopathy (ARM). Design Pooled data from 3 prospective, population-based cohort studies. Participants Subjects without early and late ARM at baseline who participated in the follow-up of the Beaver Dam Eye Study (n = 3012), the Rotterdam Study (n = 3434), and the Blue Mountains Eye Study (n = 2203). Methods Stereoscopic color fundus photographs of all participants were graded according to a standardized protocol. At baseline, current use of prescription and over-the-counter medication was assessed by interview, and the drug name was confirmed at the research centers. Procedures and definitions were similar at both baseline and follow-up across the 3 study sites. Main outcome measures Incidence of early ARM, defined as the presence at follow-up of either soft distinct drusen with pigmentary changes or soft indistinct or reticular drusen. Results In the pooled cohort, 53.3% of participants used at least one of the medications selected for this study. Within a mean period of 5.6 years, a total of 683 subjects developed early ARM. Users of antihypertensive medication in general, and β-blockers in particular, had a borderline statistically significant increased risk of early ARM (odds ratio [OR] for β-blockers, 1.3; 95% confidence interval [CI], 1.0-1.6) when adjusted for systolic (or diastolic) blood pressure and other confounders. A protective effect of borderline significance was found among women using hormone replacement therapy (OR, 0.6; 95% CI, 0.4-1.0) and in persons using tricyclic antidepressants (OR, 0.4; 95% CI, 0.2-1.0). In contrast with β-blockers, the direction and magnitude of the association with hormone replacement therapy and tricyclic antidepressants were inconsistent among the 3 study populations. Conclusions Pooled data from 3 population-based studies showed no strong associations between medication use and the incidence of early ARM. Of borderline significance were a slightly increased risk among users of β-blockers and a reduced risk among users of hormone replacement therapy and users of tricyclic antidepressants. Although β-blocker use could be a proxy for systemic hypertension, these findings warrant further investigations, preferably including information on the dosage and duration of drug exposure.