Association Between Sodium Glucose Co-transporter 2 Inhibitors and Incident Glaucoma in Patients with Type 2 Diabetes: A Multi-institutional Cohort Study in Taiwan

Background: Type 2 diabetes (T2D) is an important risk factor for glaucoma, and sodium-glucose co-transporter 2 (SGLT2) inhibitors have been shown to protect the optic nerves. However, no current studies have investigated the risk of glaucoma following SGLT2 inhibitor treatment.

Objectives: To evaluate the association between SGLT2 inhibitors and incident glaucoma.

Methods: This retrospective cohort study analyzed the largest multi-institutional electronic medical records database in Taiwan, containing 6% of the Taiwanese population, i.e. 1.3 million patients. We included T2D patients newly prescribed SGLT2 inhibitors or glucagon-like peptide-1 receptor agonists (GLP-1 RAs) from 2016 to 2018. Our primary outcome was incident glaucoma diagnosis between initiation of SGLT2 inhibitors or GLP-1 RAs, and 31st March 2021. After applying inverse probability of treatment weighting (IPTW) to increase homogeneity of co-morbidities, co-medications, glycemic controls and renal functions between the two treatment groups at baseline, we estimated hazard ratios (HR) with 95% confidence intervals (CI) for the risk of glaucoma, based on Cox proportional hazards regression models. To demonstrate the robustness of our study results, we conducted several sensitivity analyses by using other propensity score methods (1:1 matching and standardized mortality ratio weighting methods) and on-treatment analysis. We also used a falsification endpoint (pseudophakia or aphakia) to determine the likelihood of false positive results of the original study outcomes.

Results: We included 9,927 and 1,065 T2D patients who had been newly prescribed SGLT2 inhibitors or GLP-1 RAs, respectively. The mean age of SGLT2 inhibitor new users was 59.5 (standard deviation, SD 12.1) years old, with a median HbA1c of 8.6 (interquartile range, IQR: 7.7-9.7) % and eGFR of 89.2 (IQR: 71.4-108.8) ml/min/1.73m2, which was similar to GLP-1 RA new users. We observed a total of 623 incident glaucoma events during the follow-up period. Patients newly receiving SGLT2 inhibitors were associated with a lower risk of glaucoma (7.9 / 1,000 person-years), compared to those newly receiving GLP-1 RAs (10.0 / 1,000 person-years) with an HR of 0.81 (95% CI: 0.69-0.95). Multiple sensitivity analyses and a negative control outcome analysis confirmed the robustness of our main findings.

Conclusions: This study suggests that T2D patients newly prescribed SGLT2 inhibitors have a reduced risk of incident glaucoma, compared to those prescribed GLP-1 RAs, in clinical practice. Future prospective studies are suggested to confirm this association.