Abstract
Visual impairment is a global health problem that leads to poor quality of life. The aim of the study was to examine the dose-response relationship between alcohol consumption and incident visual impairment (VI). This longitudinal cohort study consisted of 287,352 Korean adults who attended health screenings between March 2011 and December 2017 and were followed for up to 8.8 years (median, 4.9 years). Participants were categorized based on their average alcohol consumption. VI was defined as bilateral visual acuity (VA) worse than 0.3 logMAR. We identified 8320 cases of new-onset bilateral VI (incidence rate, 6.0/1000 person-years). Increased alcohol intake was positively and dose-dependently associated with elevated incidence of VI (p(trend) < 0.001). With lifetime abstinence (reference), the multivariable-adjusted hazard ratios (HRs) (95% confidence intervals (CIs)) for incident VI with alcohol intake of <10, 10 to <20, 20-39.9, and ≥40 g/day were 1.07 (0.96-1.19), 1.15 (1.03-1.30), 1.15 (1.01-1.30), and 1.23 (1.08-1.40), respectively. Frequent binge drinking (≥once/per week) was associated with elevated risk of VI (HRs, 1.22; 95% CIs: 1.13-1.32). Former drinkers, particularly men, were at a higher risk for incident VI than lifetime abstainers. Similar associations were observed on evaluating changes in alcohol consumption and other confounders as time-varying covariates. Alcohol consumption, both in moderation and excess, was associated with increased VI incidence.