A FAERS database study on visual impairment adverse events associated with three long-acting insulin analogues

Purpose To utilize the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS) to systematically evaluate and compare the risk signals, clinical characteristics, and potential influencing factors of visual impairment adverse events (AEs) associated with three major long-acting insulin analogues (insulin degludec, insulin detemir, and insulin glargine). Methods This retrospective pharmacovigilance study analyzed FAERS data from Q1 2004 to Q2 2025. Reports explicitly listing any of the three target insulins as the “primary suspect” were extracted. Data extraction and deduplication were conducted in accordance with FDA guidelines. Disproportionality analysis was performed using the Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and Multi-item Gamma Poisson Shrinker (MGPS). Furthermore, multivariate regression and Weibull distribution analyses were employed to identify potential clinical risk factors and time-to-onset patterns of these ocular events. Results A total of 102,953 primary suspect reports were analyzed. Statistically significant visual impairment signals were detected across all three insulins. Insulin glargine exhibited the strongest overall association (ROR = 3.60, 95% CI : 3.54–3.67), whereas insulin degludec demonstrated the weakest (ROR = 1.59, 95% CI : 1.47–1.72). Regarding specific event types, insulin degludec and insulin detemir were primarily associated with diabetic retinopathy, whereas insulin glargine showed the strongest association with diabetic glaucoma (ROR = 50.36). Among the 5,780 specific “visual impairment” reports, over 60% involved female patients, and 93% originated from the United States. Multivariate regression suggested that higher body weight might be a potential protective factor (P 0.05); however, this finding may be confounded by the database’s inherent limitation in differentiating between Type 1 and Type 2 diabetes characteristics. Weibull analysis confirmed an early-onset pattern (early failure model) for all three insulins, with insulin degludec showing the shortest median time-to-onset (23.6 days). Conclusions Long-acting insulins are associated with varying risks of visual impairment AEs, exhibiting an early-onset temporal pattern. Insulin glargine demonstrated stronger disproportionality signals for glaucoma, while degludec and detemir were more closely associated with retinopathy. These characteristics suggest the necessity of implementing individualized ophthalmic monitoring during the early phase of initiating basal insulin therapy in diabetic patients.