番茄社区

CHICAGO -- Diabetic patients treated with metformin had almost a 50% lower risk of developing age-related macular degeneration (AMD), a large retrospective study from Taiwan showed.

Overall, metformin users had a 46% reduction in the relative risk of AMD, as compared with nonusers. The magnitude of the protective effect increased with duration of metformin use and with cumulative dose of metformin, reported Yu-Yen Chen, MD, PhD, of Taichung Veterans General Hospital in Taipei.

"In type 2 diabetic patients, those who use metformin will be at a considerably lower risk of AMD," Chen said at the American Academy of Ophthalmology meeting. "The protective effect of metformin is dose responsive, meaning that a longer duration of treatment or a higher dose of metformin will significantly reduce the risk of AMD."

Chen added that "These findings were based on a statistical analysis of patient records. Further basic research will be performed to find more evidence to support our conclusions."

One of the challenges of observational studies is how to assign causality for a statistical association, noted Emily Y. Chew, MD, of the National Eye Institute in Bethesda, Maryland. As a point of clarification, she asked how the investigators defined AMD.

Chen responded that each AMD diagnosis was confirmed by fundoscopy, fluoroangioscopy, optical coherence tomography, or another accepted test. Investigators did not examine the extent or severity of AMD, which could be a component of future studies to confirm and clarify the association between metformin use and risk of AMD, she added.

Metformin is the most widely used oral medication for type 2 diabetes, and the drug has both anti-inflammatory and anti-oxidative effects, Chen noted. Inflammation and oxidative stress have been shown to increase production of vascular endothelial growth factor, a driver component of AMD development, and to play a role in formation of drusen, the defining feature of AMD.

"If metformin can inhibit inflammation and oxidative stress, it may also reduce the risk of AMD," Chen said, explaining the rationale behind the study.

Previous studies failed to demonstrate an association between metformin and AMD, she acknowledged. However, the studies were limited by small patient numbers, limited follow-up, variable diagnostic criteria for AMD, and limited access to prescription records to document metformin use. For their investigation, Chen and colleagues had access to the Taiwan National Health Insurance Database, which provided the quantity and depth of data to address limitations of the previous studies.

Investigators identified 73,118 patients who had new diagnoses of diabetes during 2001-2013. After excluding patients with type 1 diabetes and a diagnosis of AMD at enrollment, they were left with 68,205 patients for the analysis, comprising 45,524 patients with one or more prescriptions for metformin and 22,681 patients with no documented metformin use.

Metformin users were slightly younger (age 55.2 vs 57.8, P<0.0001) and more likely to be men (54.4% vs 49.0%, P<0.0001). Users had significantly higher rates of hypertension (70.3% vs 64.0%, P<0.0001), hyperlipidemia (69.4% vs 59.5%, P<0.0001), and coronary heart disease (31.8% vs 28.0%, P<0.0001).

During the period studied, 3.4% of metformin users and 5.6% of nonusers had new diagnoses of AMD (P<0.0001). The difference translated into a hazard ratio of 0.54 in favor of metformin use as a protective factor in the risk of AMD (95% CI 0.50-0.58, P<0.0001). The only other significant predictive factor was increasing age (HR 2.57 to HR 6.44 for each increasing decade, P<0.0001).

After identifying the association between metformin use and AMD risk, Chen's group sought to determine whether exposure level of metformin affected the risk of AMD. They found that patients who had a treatment history of less than 1.5 years had an HR of 1.0 for AMD, declining to 0.50 among patients treated with metformin for 1.5 to 4.0 years and to 0.15 for patients treated more than 4 years.

A similar relationship existed with cumulative total dose of metformin. As compared with patients who received a total metformin dose <400 g, those who received a total dose of 400-1,400 g had a 50% reduction in the hazard for AMD, increasing to 72% for patients who received total doses exceeding 1,400 g. Increasing daily dose of metformin also was associated with a declining risk of AMD, from 1.0 for an average daily dose of 1.5 g/day to 0.84 among patients who received 1.5 to 2.1 g/day to 0.53 for patients with average doses ≥2.1 g/day.

Comment