Abstract
Purpose :
Metformin, the oral diabetic agent, is a safe and inexpensive medication that may have protective effects against diabetic retinopathy (DR). This study tested the hypothesis that use of metformin was associated with reduced prevalence of DR independent of its blood sugar lowering effects.
Methods :
We analyzed cross-sectional data from patients undergoing diabetic retinopathy screening at Zuckerberg San Francisco General between March 1, 2014 and November 24, 2015. All patients undergoing DR screening were included. A modified Poisson model tested the relative risk of having DR when exposed to metformin, adjusting for age, sex, last available HbA1c value, and kidney function. The University of California, San Francisco IRB approved the use of patient data for this study.
Results :
Of the 1,711 diabetic patients screened, 1,131 (66.1%) did not show signs of DR, 430 (25.1%) had mild non-proliferative DR (NPDR), 88 had moderate NPDR (5.1%), 18 had severe NPDR (1.1%), and 44 had proliferative DR (2.6%). Our population contained 1,330 (77.7%) patients currently on metformin. 49% of patients were male and 51% were female. The mean age of patients at the time of DR screening was 59.0 years old. After adjusting for age, sex, last available HbA1c value, and kidney function, exposure to metformin was associated with an 18% reduction in diabetic retinopathy prevalence (RR = 0.82, 95% CI 0.70 to 0.95, P=0.01). The protective effect of metformin remained statistically significant when adjusting for both last HbA1c value and HbA1c value from one year prior to DR screening in place of the most recent HbA1c value (RR = 0.83, 95% CI 0.72 to 0.97, P=0.02). The results of the analysis also remained essentially unchanged with (RR = 0.82, 95% CI 0.70 to 0.95, P=0.01) and without adjusting for kidney function (RR = 0.83, 95% CI 0.72 to 0.97, P=0.02).
Conclusions :
After adjusting for HbA1c, sex, age and kidney function, exposure to metformin was associated with reduced prevalence of diabetic retinopathy. This supports the hypothesis that there are protective effects of metformin against diabetic retinopathy independent of glycemic control. Further studies analyzing longitudinal and/or more detailed data are needed to confirm whether this inexpensive and safe medication truly has a protective effect against diabetic retinopathy or whether the protective effect is due to confounding differences in exposure groups.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.