Evaluation of the Impact of Gamma-Aminobutyric Acid on Diabetic Retinopathy in a Large US Population-Based Cohort.

Kim, Sonia B; Zhou, Melody Y; Allan, Kevin C; Kaelber, David C; Babiuch, Amy S; Singh, Rishi P; Talcott, Katherine E · Am J Ophthalmol · 2026

prospective_cohort · Level II

Where this comes from

Abstract

To investigate how exposure to GABAergic medications affects diabetic retinopathy (DR) development, progression, and complications. Retrospective clinical cohort study using multi-institutional electronic health record data (TriNetX, US Collaborative Network). Adults aged ≥ 18 years with type 2 diabetes mellitus with ophthalmology follow-up. Study cohorts had GABAergic prescription records for 6-months, 1-year, 3-years, or 5-years; control cohorts had no GABAergic prescriptions ever. Cohorts were propensity-score matched (PSM) on demographics, systemic comorbidities, common indications for GABAergic medications, and ophthalmic confounders. Outcomes included incident DR, progression from mild/moderate nonproliferative DR to severe nonproliferative DR, proliferative DR, or interventions required in advanced DR, and incident DR complications. Hazard ratio (HR) and 95% confidence intervals (CI); significance threshold < 0.9 or > 1.1. After successful PSM, there were 110 495 (6-months), 99 187 (1-year), 63 194 (3-years), and 40 810 (5-years) DR-naive study patients with the respective GABAergic prescription durations of interest. Compared to 109 603 DR-naive control patients, study patients demonstrated a significantly reduced HR for DR development at all time points from 6-months (HR 0.61, 95% CI 0.57-0.65) to 5-years (HR 0.81, 95% CI 0.77-0.85). Study patients (n = 14 644) with baseline mild/moderate nonproliferative DR had a significantly reduced hazard of progressing to severe nonproliferative DR, proliferative DR, or DR interventions with 6-months (HR 0.75, 95% CI 0.68-0.82) and 1-year (HR 0.69, 95% CI 0.73, 0.86) GABAergic exposure. Prescription for 6-months (HR 0.74, 95% CI 0.68-0.80) to 3-years (HR 0.80, 95% CI 0.84-0.86) was associated with a significantly reduced hazard for DR complications. Stratification by 4 specific medication indications consistently showed a reduced hazard for DR development with a 6-month prescription duration. GABAergic medication use, particularly short-term exposure, is associated with a reduced hazard of DR development, progression, and complications. These exploratory findings support a potential role of GABAergic modulation in diabetic retinal disease.