Semaglutide tied to lower seizure risk

A population-based target trial emulation published in Neurology on 17 June 2026 reports that adults with type 2 diabetes who started semaglutide had a significantly lower incidence of adult-onset seizures compared with those who started other glucose-lowering drugs or SGLT2 inhibitors^[1]. The study used data from the All of Us Research Program, a US-based longitudinal cohort, covering January 2018 to October 2023.

Semaglutide is a GLP-1 receptor agonist prescribed for type 2 diabetes (as Ozempic) and weight management (as Wegovy). Most research on the drug has focused on metabolic and cardiovascular outcomes. The emergence of seizure-related findings reflects a broader pattern: observational databases are surfacing neurological associations that were never part of the original clinical development programme.

What the study found

The researchers emulated two parallel trials. In the first, 2,586 semaglutide initiators were compared with 7,627 adults who started other glucose-lowering drugs. The weighted hazard ratio for adult-onset seizure was 0.44 (95% CI 0.25 to 0.79), meaning the semaglutide group experienced roughly 56% fewer seizure events over up to four years of follow-up^[1]. The absolute four-year risk difference was -1.78 percentage points, translating to a number needed to treat of 70.

In the second comparison, 2,814 semaglutide initiators were matched against 5,791 adults who started an SGLT2 inhibitor. Here the hazard ratio was 0.48 (95% CI 0.27 to 0.85), with a four-year risk difference of -1.46 percentage points and a number needed to treat of 131^[1]. Both comparisons point in the same direction: semaglutide users experienced fewer seizures than either comparator group.

The mechanism is not explained by weight or blood sugar

A mediation analysis tested whether the lower seizure risk could be accounted for by semaglutide's known effects on HbA1c and BMI. It could not. HbA1c mediated only 2.4% of the effect versus other glucose-lowering drugs and 6.5% versus SGLT2 inhibitors. BMI mediated 0% and 0.7% respectively^[1]. In other words, the association appears largely independent of glycaemic control and weight loss. The authors suggest that GLP-1 receptor signalling in the brain may play a direct neuroprotective role, though the study was not designed to test that hypothesis.

How this fits the wider evidence

This is not the first study to link GLP-1 receptor agonists with lower seizure or epilepsy risk. A large Taiwanese cohort study published in the same journal earlier in 2026 found that GLP-1 RA users had a 16% lower epilepsy risk overall (HR 0.84), with semaglutide specifically showing a 32% reduction (HR 0.68) compared with DPP-4 inhibitor users^[2]. A 2024 meta-analysis of 27 randomised controlled trials covering nearly 200,000 patients found that newer glucose-lowering drugs, GLP-1 agonists in particular, reduced the combined risk of seizures and epilepsy by 24% (RR 0.76, 95% CI 0.62 to 0.95)^[3].

Taken together, these three studies from different databases, populations, and methodologies all point in the same direction. That convergence is notable, though none of the studies is a randomised trial designed with seizure as the primary endpoint. GLP-1 receptors are expressed throughout the central nervous system, and preclinical work has shown that GLP-1 agonists can reduce neuroinflammation and oxidative stress in animal models of neurological injury. Whether those laboratory findings translate to a clinically meaningful seizure benefit in humans remains to be proven. Semaglutide is already the subject of several studies exploring its effects beyond metabolism, including trials examining cognitive decline and neurodegeneration.

Limitations to keep in mind

The authors themselves note several caveats. Event counts were low, which limits statistical precision. Follow-up was relatively short for a seizure-risk study, capped at about four years. The All of Us cohort, while diverse, does not capture the full US population. And as with any observational study, residual confounding cannot be ruled out, even after inverse probability weighting and multiple sensitivity analyses^[1].

The study also examined only semaglutide, not the broader GLP-1 agonist class. Whether liraglutide, dulaglutide, or the dual agonist tirzepatide carry the same association remains an open question. The classification of evidence is Class II, one step below the gold standard of a randomised controlled trial.

What we do not yet know

No randomised trial has tested a GLP-1 agonist with seizure reduction as a primary outcome. The biological mechanism behind the observed association is not established. Whether the effect is specific to semaglutide or shared across the GLP-1 agonist class is unclear. And the clinical magnitude of the benefit, while statistically significant, is modest in absolute terms: roughly 1 to 2 fewer seizures per 100 treated patients over four years.

Patients with type 2 diabetes who have concerns about seizure risk should discuss this evidence with a healthcare provider in the context of their full medical history. This study does not support using semaglutide specifically to prevent seizures.