Metformin linked to lower odds of intermediate macular degeneration in people with diabetes

By Hugo Francisco de SouzaReviewed by Susha Cheriyedath, M.Sc.Feb 4 2026

A large retinal imaging study suggests a common diabetes drug may help slow progression toward sight-threatening macular degeneration, though clinical trials are still needed to confirm any protective effect.

Study: Metformin and incidence of age-related macular degeneration in people with diabetes: a population-based 5-year case-control study. Image Credit: reddees / Shutterstock

In a recent study published in the journal BMJ Open Ophthalmology, researchers conducted a retrospective analysis of prospectively collected screening data, described by the authors as a population-based case–control study, to investigate whether metformin prescription was associated with the incidence of age-related macular degeneration (AMD). The study analysed retinal fundus photographs from more than 2,000 patients over a five-year period.

The findings indicated that metformin use was associated with an approximately 37% lower odds of developing intermediate AMD (p = 0.01). However, causality cannot be inferred, and the results are considered hypothesis-generating rather than confirmatory.

Background: Age-Related Macular Degeneration and Therapeutic Gaps

Age-related macular degeneration (AMD) is a chronic, progressive eye disease that primarily affects individuals aged 60 years and older. Characterised by damage to the macula, AMD is a leading cause of irreversible vision loss. The disease is broadly classified into dry (more common, slower progressing) and wet (less common, more aggressive) forms, with risk influenced by ageing, smoking, and genetic susceptibility.

AMD represents a growing health and economic burden in high-income countries. Current effective treatments, such as intravitreal anti-vascular endothelial growth factor injections, are largely limited to advanced wet-stage disease. As a result, there is a substantial therapeutic gap in preventing progression in the early stages.

Rationale for Investigating Metformin

Metformin is an inexpensive, widely prescribed medication with a well-established safety profile, primarily used in the management of type 2 diabetes (T2D). Beyond glycaemic control, metformin has been associated with potential anti-ageing properties, including anti-inflammatory effects, reduction of oxidative stress, and activation of pathways that mimic caloric restriction.

Because AMD is fundamentally an age-related and inflammatory condition, prior studies have hypothesized that metformin may exert protective effects on the retina. However, much of the existing evidence relies on insurance claims or billing data, which may inadequately capture disease severity and progression.

Study Design and Data Source

This study sought to address these limitations by analysing prospectively collected retinal imaging data from the Individualised Screening for Diabetic Retinopathy (ISDR) study conducted in Liverpool, United Kingdom.

The initial cohort comprised 2,600 randomly selected participants aged 50 years or older with clinically validated type 2 diabetes (T2D). After excluding individuals with ungradable images or type 1 diabetes, the final analytic sample consisted of 2,089 participants.

Retinal Imaging and AMD Classification

Participants underwent standard retinal fundus photography at baseline (2011) and at five-year follow-up (2016). Images were graded using a modified Age-Related Eye Disease Study (AREDS) classification system, categorising disease severity as no AMD, early, intermediate, or late AMD. Grading reliability was supported through masked re-grading of a subset of images.

Statistical Analysis and Confounder Adjustment

Multivariable logistic regression models were used to evaluate associations between metformin prescription and AMD incidence. Models were adjusted for potential confounders, including age, sex, glycated haemoglobin (HbA1c), diabetes duration, presence of diabetic retinopathy, and other prespecified variables. Data on smoking, diet, and supplement use were limited and not fully captured.

Key Results: Association With Intermediate AMD

The analysis identified a statistically significant association specifically with intermediate AMD. Metformin use was associated with reduced odds of developing intermediate-stage disease over the five-year follow-up period.

Adjusted models yielded odds ratios ranging from 0.63 to 0.66 (p = 0.01–0.02), corresponding to approximately 37% lower odds of intermediate AMD among metformin users compared with non-users. These estimates reflect relative odds rather than absolute risk reduction.

Findings for Early and Late AMD

In unadjusted analyses, metformin use appeared to be associated with a reduced incidence of late-stage AMD (OR 0.43, p = 0.02). However, this association lost statistical significance after adjustment for age and sex (p = 0.2), likely due to the small number of late-stage cases and limited statistical power.

No significant association was observed between metformin use and the incidence of early AMD. The authors note that participants not prescribed metformin were, on average, older and had a higher baseline prevalence of AMD. Nonetheless, the association with intermediate AMD persisted after adjustment, although residual confounding cannot be excluded.

Conclusions and Implications for Future Research

This study provides observational, image-based evidence suggesting an association between metformin use and a lower incidence of intermediate AMD, but does not provide definitive proof of a protective effect. By potentially delaying progression to intermediate disease, metformin may represent a candidate for future preventive intervention trials.

The authors hypothesize that metformin’s effects may relate to reduced age-related inflammation and improved mitochondrial function, thereby slowing cellular ageing within the retina. These mechanisms remain speculative and were not directly evaluated.

Limitations include the observational design, incomplete adjustment for lifestyle confounders, the absence of optical coherence tomography imaging, and the lack of data on metformin dose or adherence. Despite these constraints, the findings support the rationale for prospective clinical trials evaluating metformin as a preventive strategy for AMD. If confirmed, this could enable repurposing of a low-cost, widely available medication, particularly among individuals with diabetes, though generalisability beyond this population remains uncertain.