PMID-39696569 — Danish/Norwegian Cohort: Semaglutide Doubles 5-Year NAION Risk in 424,152 T2D Patients

[DRAFT — authored 2026-04-18. Requires Medical Director review.]

Citation

Grauslund J, et al. Once-weekly semaglutide doubles the five-year risk of nonarteritic anterior ischemic optic neuropathy in a Danish cohort of 424,152 persons with type 2 diabetes. International Journal of Retina and Vitreous. 2024 Dec. doi: 10.1186/s40942-024-00620-x. PMID: 39696569.

Study Design

Design: Longitudinal cohort study, 5-year follow-up
Setting: Denmark, nationwide
Period: 2018–2024
Exposure definition: Once-weekly semaglutide prescription
Comparator: Non-exposure to semaglutide
Primary analysis: Incidence rates and Cox-regression hazard ratios

Population

N: 424,152 persons with type 2 diabetes identified nationally
Exposure group: 106,454 persons exposed to once-weekly semaglutide
Non-exposure group: 317,698 persons not exposed
Total observation: 1,915,120 person-years
Events observed: 218 cases of NAION (67 in exposed)

Intervention

Real-world prescription-based exposure classification; no intervention randomization. Propensity for semaglutide use was not matched but analyzed with multivariable adjustment.

Outcomes

Primary — NAION incidence and HR

Incidence rate in exposed: 0.228 per 1,000 person-years
Incidence rate in non-exposed: 0.093 per 1,000 person-years (p < 0.001)
Hazard ratio: 2.19 (95% CI 1.54–3.12) for NAION in semaglutide-exposed vs. non-exposed
Median time from first semaglutide prescription to NAION event: 22.2 months (IQR 10.2–37.8)

Secondary

Event rates stratified by dose escalation, duration of exposure, and age subgroups
Consistent directional HR > 1 across sensitivity analyses

Key Findings

This large nationwide T2D cohort confirms the NAION signal identified in the Hathaway 2024 single-center cohort (PMID-38958939) at population scale. The magnitude (HR 2.19) is similar to — though slightly lower than — the Mass Eye and Ear single-center estimate (HR 4.28). The 22-month median time-to-event supports an exposure-duration effect consistent with a drug-related rather than acute-risk mechanism.

Cumulative absolute risk remains low — on the order of one additional case per 10,000 person-years of treatment, consistent with EMA PRAC's "very rare" classification. However, in a patient with ophthalmic risk factors (crowded disc, unilateral prior NAION, hypertension, sleep apnea), the relative risk may be clinically meaningful.

Combined with the Lakhani 180-countries pharmacovigilance study (PMID-40383360) and the Hathaway single-center cohort (PMID-38958939), a consistent triangulation of evidence supports the semaglutide-NAION association, with tirzepatide appearing to have a narrower ocular signal profile.

Limitations (Author-acknowledged)

Observational design; cannot establish causation
Exposure to semaglutide and NAION development both correlate with baseline metabolic severity; residual confounding possible despite adjustment
Danish healthcare system characteristics limit generalizability (particularly to US populations)
NAION case ascertainment relies on diagnostic coding; under-reporting likely
No dose-stratification presented in primary analysis
Tirzepatide exposure not analyzed (because of limited Danish tirzepatide exposure during the period)

Evidence Level

Level II (Oxford CEBM) — large, well-conducted cohort study with population-scale capture and 5-year follow-up.

Linked Peptides

Semaglutide

Related Studies

PMID-38958939 – Hathaway Semaglutide NAION Cohort (single-center precursor)
PMID-40383360 – Lakhani 180 Countries GLP-1 Ocular Events (global pharmacovigilance)

Orchestrator Notes

Together with Hathaway 2024 and Lakhani 2025, this paper supports the EMA PRAC June 2025 conclusion classifying NAION as a "very rare" side effect of semaglutide.
The 22-month median time-to-event argues against rapid-onset hemodynamic mechanism; more consistent with cumulative vascular exposure effect.
Implication for tirzepatide: insufficient Danish tirzepatide exposure during study period precluded analysis. Tirzepatide-NAION association remains not-demonstrated in this dataset.

PMID-39696569 – Danish Norwegian NAION Cohort