Target trial emulation finds semaglutide or tirzepatide linked to higher optic nerve disorder risk

A new cohort study published in JAMA Network Open reports that adults with type 2 diabetes who were prescribed semaglutide or tirzepatide had a higher two-year risk of being newly diagnosed with nonarteritic anterior ischemic optic neuropathy (NAION) and other optic nerve disorders than similar patients taking different diabetes drugs — though the absolute number of cases in both groups was very small ^[1].

NAION is a sudden loss of blood flow to the optic nerve that can cause permanent vision loss in one eye. Researchers have been examining a possible link to semaglutide since 2024, but as the study authors note, "reported associations were inconsistent," and it was not known whether these drugs were tied to other optic nerve and visual pathway problems ^[1].

What the study did

The researchers used a method called target trial emulation, which applies the design rules of a randomized trial to real-world records. They drew on a nationwide, population-based database of US electronic health records covering December 2017 through January 2023 ^[1].

Eligible patients had type 2 diabetes, no prior diagnosis of eye disorders, and a prescription for semaglutide, tirzepatide, or another diabetes medication. From 1,511,637 eligible patients, propensity score matching produced a study population of 159,398 people: 79,699 prescribed semaglutide or tirzepatide and 79,699 in the comparison group. Mean age was 56.5 years (standard deviation 13.3), and 83,123 patients (52.15%) were women ^[1].

The comparison group included people taking insulins, metformin, DPP-4 inhibitors, SGLT2 inhibitors, sulfonylureas, thiazolidinediones, and other GLP-1 receptor agonists — a group that included dulaglutide (83,501 patients), liraglutide (53,189), exenatide (17,046), lixisenatide (2,465), and albiglutide (1,108) ^[1].

The numbers

Over two years of follow-up, 35 patients (0.04%) in the semaglutide-or-tirzepatide group were diagnosed with NAION, compared with 19 patients (0.02%) in the matched comparison group. That worked out to a hazard ratio of 1.76, with a 95% confidence interval of 1.01 to 3.07 ^[1].

For other optic nerve disorders, there were 93 cases (0.12%) in the semaglutide-or-tirzepatide group versus 54 cases (0.07%) among comparators, a hazard ratio of 1.65 (95% CI, 1.18–2.31) ^[1].

The authors reported no association with other disorders of the optic nerve or visual pathways ^[1]. They concluded that patients prescribed these drugs "had an increased risk of NAION and other optic nerve disorders, although the overall risk was low," and said the findings "highlight the need for close monitoring of these conditions" ^[1].

An important limitation for readers weighing specific products: the exposure group combined semaglutide (sold as Ozempic, Wegovy and Rybelsus) and tirzepatide (Mounjaro and Zepbound) into a single category. Because the two medicines were analyzed together, the study cannot say whether the signal comes from one drug, the other, or both ^[1]. The study also included only people with type 2 diabetes who had no prior eye disorders, so it does not directly address people taking these drugs for obesity alone ^[1]. The authors reported no conflicts of interest ^[1].

Why it matters for patients

The practical takeaway is about scale. Even in the higher-risk group, NAION was diagnosed in about 4 of every 10,000 patients over two years, versus about 2 of every 10,000 in the comparison group ^[1]. The lower edge of the confidence interval for NAION was 1.01 — just barely above the level at which a result would be considered statistically inconclusive — meaning the NAION estimate is less precise than the finding for other optic nerve disorders, where the interval ran from 1.18 to 2.31 ^[1].

This is also an observational study built from prescription and diagnosis records, not a randomized trial. Propensity score matching balances measured characteristics between groups, but it cannot rule out differences the records did not capture ^[1]. The study does not establish that the medications cause optic nerve injury.

What happens next

JAMA Network Open published an accompanying commentary on the paper ^[1]. The study itself points to continued monitoring of optic nerve conditions in people taking these drugs ^[1]. Whether regulators will act on these findings, and whether further research can separate semaglutide from tirzepatide, is not addressed in the study.