Long-term Metformin Use Associated with Reduced AMD Risk

Metformin has been shown in various studies to decrease AMD risk, but new research highlights the potential influence of the duration of medication usage. Prolonged use, defined as six or more consecutive years, exhibited consistently protective effects, whereas usage lasting one to four years revealed weaker associations. Notably, these protective effects didn’t transfer to wet AMD, which showed a non-significant association with short- and long-term metformin use. Photo: Photo: Jessica Haynes, OD. Click image to enlarge. Numerous studies have indicated a potential protective effect of metformin—a widely used antidiabetic drug—on age-related macular degeneration (AMD) due to the medication’s antioxidative, anti-inflammatory and anti-angiogenic properties, though findings have been largely inconsistent. These heterogeneous results may stem from discrepancies in how the studies account for the duration of therapy, which is likely a crucial mediator of any protective effect of metformin, given the gradual progression of AMD. To address this research gap, a team of investigators recently studied a large cohort from a nationwide database to explore the relationship between varying durations of metformin therapy and AMD incidence.The retrospective study pulled data from patients in the TriNetX network, including those aged 60 or older with diabetes who had no AMD diagnoses at two eyecare visits spaced at least one year apart. Metformin use was assessed over periods ranging from one to 10 years before the first ophthalmology visit, with a threshold of at least five consecutive years of use determining the primary cohort. The researchers used propensity score matching to control for potential confounders, including demographics, comorbidities, medications, laboratory values and healthcare use.The results showed that among the 3,748 patients who met the criteria for prolonged metformin use, only 3.3% were diagnosed with any form of AMD compared to 4.9% in the unexposed cohort. The hazard ratio (HR) for developing AMD among metformin users was 0.68, suggesting a statistically significant reduction in risk. Sensitivity analyses, including comparisons with insulin as a negative control and statins as a positive control, reinforced these findings by demonstrating that insulin usage did not correlate with AMD development, while statins did reflect a protective effect against AMD risk.In subgroup analyses, similar protective effects were shown for dry AMD (HR: 0.69), although for wet AMD, the association was less pronounced and not statistically significant (HR: 0.84). In their paper on the study, the researchers commented, “This observation aligns with prior research suggesting that metformin does not significantly slow the progression of existing geographic atrophy, reinforcing the idea that its protective effects may be most relevant in earlier stages of disease development.”Notably, the study authors observed that longer durations of metformin usage (≥six consecutive years) showed a trend towards greater protective effects, though with smaller sample sizes in the longer cohorts. The authors warned in their paper that these “analyses were done in a series of nested cohorts with varying durations of preceding data,” and, therefore, “patterns observed between the cohorts with longer and shorter durations of metformin are not independent and should be interpreted with caution.” While further research is warranted, these findings support the possibility that prolonged exposure to metformin could reduce the risk of AMD. The authors concluded, “Prospective randomized trials are needed to clarify the underlying mechanisms, assess long-term benefits, and explore broader clinical implications to optimize ocular care strategies in diabetic populations.”Click here for the journal source. Hong AT, Luu IY, Keenan JD, et al. Long-term metformin use and reduced risk of age-related macular degeneration: a large database study. Ophthalmol Ret. August 11, 2025. [Epub ahead of print]. This article was developed by the editorial staff in conjunction with experts in the field. In the process, AI may have been among the editorial tools used to meet the goals of human editors, who approved all content.