Study design

This observational study examined data collected in the Diagnostic Innovations in Glaucoma Study (DIGS) from 2008 to 2016. The cohort comprised 32 patients (55 eyes) with medicated diabetes and 111 age-matched controls (142 eyes) without diabetes. All diabetic patients self-reported their diabetic history and used antidiabetic medication. Patients with diabetic retinopathy or macular edema were excluded.

All study participants had follow-ups every 6 months. Investigators measured structural changes by retinal nerve fiber layer (RNFL) thinning and functional changes using visual field tests.

Outcomes

After a median follow-up time of 5.7 years, the diabetes group had a significantly slower rate of RNFL thinning than the control group (-0.40 µm/year vs -0.83 µm/year; P=0.01). However, the difference in rates of visual field loss did not reach significance (P=0.07).

Limitations

Patients self-reported type 2 diabetes history, possibly introducing some selection bias. The study lacked patients with diabetic retinopathy and had a small sample size. Based on the study design, the authors were unable to separate the treatment effect from that of the disease.

Clinical significance

These findings present 2 possible theories about the observed “protective” effect. It is possible metformin, as part of antidiabetic regimen, provides protection against RNFL thinning. Alternatively, protection may come from the greater amount of VEGF in the vitreous of patients with diabetes. The challenge is extrapolating this to a broader demographic of patients with diabetes. It is also difficult to interpret these results given the conflicting evidence from other longitudinal studies suggesting diabetes either provides protection for patients or is an associated risk factor. These findings, however, should encourage larger, longitudinal future studies.