Authors :
Presenting Author: Yong Eun, MD – Columbia University
Hyun Yong Koh, MD, PhD – Baylor College of Medicine
Suhwan Bong, BS – Harvard T.H. Chan School of Public Health
Yoonhyuk Jang, MD, PhD – Harvard Medical School
Soon-Tae Lee, MD, PhD – Seoul National University
Rationale:
Adult-onset epilepsy and seizures, occurring after age 18, are increasingly prevalent and often linked to acquired brain insults and neuroinflammatory pathways. Semaglutide, a glucagon-like peptide-1 receptor agonist (GLP-1 RA) approved for type 2 diabetes and obesity management, has shown neuroprotective effects in stroke and dementia. We investigated whether semaglutide reduces the risk of incident adult-onset epilepsy or seizures in patients with diabetes compared to sodium-glucose cotransporter-2 inhibitors (SGLT2i) and other glucose-lowering drugs (GLDs, including sulfonylureas and dipeptidyl peptidase-4 inhibitors).
Methods:
We conducted a retrospective cohort study using the National Institutes of Health All of Us Research Program (Controlled Tier Dataset v8), encompassing 393,596 adults with electronic health records. Adults (≥18 years) with diabetes mellitus newly initiating semaglutide, SGLT2i, or GLDs between December 2017-2021 were included, with follow-up through December 2023. Incident epilepsy/seizures were identified using ICD-9/10 and SNOMED CT codes. We employed target trial emulation methodology with inverse probability of treatment weighting (IPTW) to balance baseline covariates including demographics, comorbidities, medications, and baseline HbA1c and BMI. Cox proportional hazards models were used to estimate hazard ratios (HR). Additionally, we implemented targeted maximum likelihood estimation (TMLE) to calculate risk differences and number needed to treat (NNT). To evaluate whether glycemic control mediated the observed effects, we performed counterfactual mediation analysis.
Results:
From 393,596 participants, 7,326 were eligible for semaglutide vs. GLDs (n=1,964 vs. 5,362) and 5,375 for semaglutide vs. SGLT2i (n=1,650 vs. 3,725) comparisons, with mean age 63.5 years and mean follow-up 2.6 years. IPTW-weighted analyses showed semaglutide significantly reduced epilepsy/seizure risk vs. GLDs (43/3,442 vs. 89/3,884 events; HR 0.51; 95% CI 0.31-0.82; P=0.006) and SGLT2i (23/2,651 vs. 54/2,724 events; HR 0.45; 95% CI 0.23-0.85; P=0.015). TMLE showed similar findings (vs. GLDs: risk difference [RD] -0.010; NNT 96; P=0.004; vs. SGLT2i: RD -0.009; NNT 107; P=0.010).[Figure 1] SGLT2i showed no significant risk reduction vs. GLDs. Mediation analysis revealed semaglutide’s effect was independent of glycemic control (none mediated vs. GLDs, maximum 3.6% mediated vs. SGLT2i).[Figure 2]
Conclusions:
Semaglutide significantly reduced adult-onset epilepsy and seizure risk in patients with diabetes compared to SGLT2i or conventional GLDs, independent of glycemic control. These findings from rigorous target trial emulation suggest a unique neuroprotective mechanism warranting investigation of semaglutide's anti-neuroinflammatory pathways for epilepsy prevention.
Funding: The authors declare that no funding was received for this research.