The Effect of GLP-1 Receptor Agonists on Epilepsy: No Appetite for Seizures

Abstract

Association Between GLP-1 Receptor Agonist Use and Epilepsy Risk in Type 2 Diabetes

Cheng CY, Lo SC, Huang CN, Yang YS, Wang YH, Kornelius E. Neurology. 2026;106(1):e214509. doi:10.1212/WNL.0000000000214509. Epub 2025 Dec 10. PMID: 41370744.

Background and Objectives: Individuals with type 2 diabetes mellitus (T2DM) are at an increased risk of developing epilepsy, particularly in later life. While preclinical studies suggest neuroprotective properties of glucagon-like peptide-1 receptor agonists (GLP-1 RAs), real-world comparative effectiveness data remain limited. We aimed to evaluate whether GLP-1 RA use is associated with a lower risk of incident epilepsy compared with dipeptidyl peptidase-4 inhibitor (DPP-4i) use in adults with T2DM. Methods: We conducted a retrospective cohort study using the TriNetX network from 2015 to 2023, including adults aged 18 years or older with T2DM who were new users of either GLP-1 RAs or DPP-4is. Patients with a previous diagnosis of epilepsy or seizure, or those using antiepileptic drugs, were excluded. The primary outcome was incident epilepsy, identified using ICD-10-CM codes. Propensity score matching (1:1) was performed based on demographics, socioeconomic status, body mass index, comorbidities, and baseline medications. Cox proportional hazard models estimated hazard ratios (HRs) with 95% confidence intervals (CIs). We also conducted prespecified subgroup and sensitivity analyses to assess the robustness of the findings. Results: After matching, 452 766 patients were included (226 383 in each group; mean age 60.5 years; 47.1% female). During follow-up, 1670 individuals in the GLP-1 RA group and 1886 in the DPP-4i group developed epilepsy, corresponding to cumulative incidences of 2.35% versus 2.41%. Glucagon-like peptide-1 receptor agonists use was associated with a significantly lower risk of epilepsy (HR 0.84, 95% CI 0.78-0.90), with protective associations evident at 1 year (HR 0.71, 95% CI 0.62-0.80), 3 years (HR 0.81, 95% CI 0.74-0.88), and 5 years (HR 0.82, 95% CI 0.76-0.88). Among individual agents, semaglutide showed the strongest association (HR 0.68, 95% CI 0.60-0.77). The results were consistent across major subgroups, including both age and sex. Sensitivity analyses excluding patients with overlapping or switching exposure yielded similar findings (HR 0.71, 95% CI 0.64-0.78). Discussion: Glucagon-like peptide-1 receptor agonists therapy was associated with a significantly lower epilepsy risk compared with DPP-4i use in adults with T2DM. These results support the hypothesis that GLP-1 RAs may exert neurologic benefits beyond glycemic control. Limitations include the observational design and potential residual confounding.

Seizure Recurrence after GLP-1 Receptor Agonist Initiation in Adults with Epilepsy

AbuAlrob MA, Hussein A, Abdellatif R, Itbaisha A, Zammar K, Mesraoua B. Epilepsia. 2025. doi:10.1111/epi.70022. Epub ahead of print. PMID: 41251033.

Objective: To examine whether initiation of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) is associated with seizure recurrence and related outcomes in adults with epilepsy and type 2 diabetes. Methods: We conducted a retrospective cohort study using de-identified electronic health records from the TriNetX Research Network (January 2003-August 2025), including adults ≥18 years with ≥3 epilepsy or recurrent seizure diagnoses. Patients initiating a GLP-1 RA (exenatide, liraglutide, dulaglutide, lixisenatide, semaglutide, or tirzepatide) without prior comparator therapy were compared with those initiating other glucose-lowering agents (sodium–glucose cotransporter 2 inhibitors, dipeptidyl peptidase 4 inhibitors, sulfonylureas, or insulin) without GLP-1 RA exposure. Propensity score matching (1:1) was performed on 82 covariates, yielding 8688 matched pairs. Outcomes were assessed using Cox proportional hazards models. Results: After matching, the mean age was 52.6 years, and 67.6% were female. Median follow-up was 514 days (interquartile range [IQR] 671) for GLP-1 RA initiators and 415 days (IQR 769) for comparators. Glucagon-like peptide-1 receptor agonists initiation was associated with lower risk of seizure recurrence (HR 0.82, 95% confidence interval [CI] 0.78-0.86; RD −2.1%), hospitalization (HR 0.35, 95% CI 0.29-0.43; RD −2.6%), and all-cause mortality (HR 0.40, 95% CI 0.34-0.47; RD −4.8%). Associations with status epilepticus (HR 0.75, 95% CI 0.66-0.85; RD −0.7%) and ICU admission (HR 0.82, 95% CI 0.69-0.96; RD −0.3%) were smaller; the latter was not statistically significant. Significance: In this large multinational cohort, GLP-1 RA initiation was associated with reduced risks of seizure recurrence, hospitalization, and mortality compared with other glucose-lowering therapies. These hypothesis-generating findings warrant confirmation in prospective studies before translation into clinical practice.

Commentary

Glucagon-like peptide-1 (GLP-1) is a gut hormone secreted from enteroendocrine cells after food intake, facilitating insulin signaling.^1,2 Hence, GLP-1 receptor agonists (GLP-1 RAs) such as exenatide, liraglutide, dulaglutide, lixisenatide, semaglutide, and trizeparide have gained increased popularity in the management of diabetes and obesity over the past 2 decades. Given that GLP-1 receptors are also expressed in the brain and that GLP-1 RAs cross the blood–brain barrier, experimental and clinical data have emerged on the beneficial effects of these agents for several central nervous system disorders, including Alzheimer's disease, Parkinson's disease, stroke, traumatic brain injury, spinal cord injury, and addictive disorders.^2,3 Epilepsy is no exception to that trend, with animal data suggesting an anticonvulsant effect, as well as a positive impact on neuropsychiatric comorbidities.^4–8 Yet, how does that translate into clinical practice?

The current 2 studies^9,10 attempt to resonate this promising bench signal to the bedside. The first study ⁹ compared the incidence of epilepsy in adults with type 2 diabetes mellitus who were treated with GLP-1 RAs versus inhibitors of dipeptidyl peptidase 4 (DPP-4i or gliptins). After extensive propensity score matching of the 2 cohorts, GLP-1 RAs use (specifically semaglutide) was associated with a significant lower risk of epilepsy (hazard ratio [HR] 0.84, 95% CI 0.78-0.90) that sustained over 1, 3, and 5 years across all age groups and in subsequent sensitivity analysis with overlapping or switching exposure. ⁹ The second study ¹⁰ investigated the risk of seizures in patients with established epilepsy on GLP-1 RAs versus other glucose-lowering agents (eg, DPP-4i, sodium-glucose cotransporter 2 inhibitors, sulfonylureas, or insulin) with a median follow-up of 514 days for the target group versus 415 for the comparator group. After adopting a similar propensity score matching approach for the 2 cohorts, GLP-1 RAs initiation was associated with lower risk of seizure recurrence (HR 0.82, 95% CI 0.78-0.86), in addition to lower risk for hospitalization, ICU admission, status epilepticus, and all-cause mortality. ¹⁰

Both studies^9,10 draw from observational data stemming from TriNextX Research Network, a large, multinational database of more than 220 healthcare organizations around the globe. In that sense, they provide robust, real-world, epidemiological data with substantial generalizability. They utilize an active control group and extensive covariate adjustment for potential confounders with a good balance of the baseline characteristics of the populations under scrutiny, as suggested by the absolute standardized mean differences. Moreover, stratified analysis was performed in the former study, ⁹ along with a sensitivity analysis to investigate the effect of the COVID pandemic, yielding similar results. Similarly, sensitivity analysis was performed in the latter study, ¹⁰ restricted only to epilepsy-specific codes to eliminate misclassification, producing consistent results.

On the other hand, retrospective studies based on diagnostic codes are inevitably susceptible to misclassification related to coding procedures, selection bias related to practice habits and patient preference, along with the risk of residual confounding from unmeasured variables, such as genetic susceptibility or lifestyle factors. A more granular analysis on the type, frequency, and severity of incident seizures, as well as their response to antiseizure medication management, was not possible. Despite the inclusion of additional clinical outcomes in the second study, ¹⁰ the long-term effect on epilepsy related comorbidities could not be elucidated.

These limitations notwithstanding, these 2 studies^9,10 highlight the potential antiepileptic effect of GLP-1 RAs. That is in line with a mounting amount of experimental data that show promising potential of GLP1-RAs in numerous animal models of epilepsy (pentylenetetratozol, picrotoxin, and pilocarpine-induced seizures, as well as kainic acid-induced temporal lobe epilepsy, WAG/Rij rat model of absence epileptogenesis, and Dravet syndrome).^4–8 Potential mechanisms of reduced cortical excitability include modulation of cytoplasmic calcium and K_ATP channels, upregulation of the GABAergic and downregulation of the glutaminergic system, regulation of dopaminergic, norepinephrine, and serotonin activity, homeostasis and neuroprotection through neurotrophic factors, as well as antioxidant and anti-inflammatory effects.^1,6,7 These effects appear to act synergistically with coadministered antiseizure medications (ASMs). ³ Moreover, in the laboratory, the benefit appears to extend beyond seizure control to common cognitive and psychiatric comorbidities of epilepsy. ⁶

The current 2 studies^9,10 provide preliminary clinical credibility to this biological plausibility. They attest to the reduction of cortical excitability in humans, not only those with established predisposition to seizures ¹⁰ but also these without documented epilepsy. ⁹ In that sense, they provide support for an antiepileptic and an anticonvulsant effect. That is critical as our armamentarium in epilepsy is based predominantly on symptomatic management and not etiological treatment. Furthermore, the translation of this signal into better clinical outcomes, such as reduced hospitalization and mortality, ¹⁰ delivers a very encouraging message beyond a merely theoretical association. A recent meta-analysis of 27 randomized clinical trials comparing newer glucose-lowering drugs to placebo corroborated this contention, given that patients taking specifically GLP-1 RAs had a 24% lower risk of late-onset seizures and epilepsy combined compared to placebo. ¹¹

Several questions remain to be answered. Are the promising associations seen in these studies a reflection of better metabolic control and improved cerebrovascular and cognitive health, or do they suggest a direct anticonvulsant effect through enhanced neuroprotection and optimized neurotransmitter signaling. In that regard, given that GLP-1 RAs carry minimal intrinsic hypoglycemia risk, can they be extended to nondiabetic patients at risk to develop epilepsy from other causes and established patients with epilepsy without diabetes? Which GLP-1 RA should we use? As the first of these 2 studies suggested with the differential benefit of semaglutide likely to its favorable pharmacokinetics, ⁹ all size may not fit all. How much should we administer and what should we beware of? Can their use be extended to special populations not included in the current studies (eg, children)? What types of seizures and epilepsy syndromes would mostly be amenable to them? From a clinical and health economics viewpoint, when and how should they be combined with traditional treatment approaches, such as ASMs, neurostimulation, and surgery? What is their impact on commonly encountered neuropsychiatric comorbidities and the overall quality of life?

Το sum up, the presented epidemiological data^9,10 remain exploratory and cannot undoubtedly establish causality. Nevertheless and regardless of the underlying mechanism (neurobiologic actions vs broader health benefit), they suggest that GLP-1 RAs reduce the “appetite” for seizures. The risk reduction becomes apparent early and persists in the long run. As such, they merit further investigation with large, prospective, clinical trials to demonstrate their promising neuroprotective effect against cerebrovascular pathology, neurodegeneration and neuroinflammation and, hopefully, enrich our epilepsy management palette.