GLP-1 receptor activation linked to lower depression and bipolar disorder odds

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Jul 15 2026

Lifelong genetic differences associated with GLP-1 receptor activity point to possible mental health benefits beyond weight and glucose control, but clinical evidence remains essential.

Study: Glucagon-like peptide-1 receptor activation and mental health: a drug-target mendelian randomization study. Image Credit: Bacsica / Shutterstock

In a recent 'article in press' published in the journal Translational Psychiatry, researchers evaluated the association between genetically predicted activation of the glucagon-like peptide-1 receptor (GLP-1R) and mental health outcomes.

GLP-1R agonists (GLP-1RAs) can effectively reduce blood glucose and body weight, and have beneficial cardiovascular and mortality outcomes in randomized controlled trials. These agents act on vagal afferents, the hypothalamus, and the hindbrain, suppressing food intake and reducing body weight. However, obesity drugs acting on the central nervous system (CNS) could produce psychiatric effects.

For instance, the phentermine-topiramate combination increases anxiety and depression. Nevertheless, GLP-1RAs were found to moderately decrease depressive symptoms in individuals with type 2 diabetes (T2D), obesity, or overweight. In addition, animal, observational, and genetic studies suggest benefits of GLP-1RAs in people with alcohol use disorder, although randomized trials have produced mixed results, with ongoing trials assessing their effectiveness in those with substance use disorders (SUDs) and mental illnesses.

About the Study

In the present study, researchers examined the associations between genetically predicted GLP-1R activation and mental health outcomes. The study used data from publicly available genome-wide association studies (GWASs). Genetic instruments for GLP-1R activation were selected based on associations with traits linked to drug target effects, such as body mass index (BMI) and glycated hemoglobin (HbA1c).

Genetic associations with HbA1c and BMI were obtained from a GWAS of the United Kingdom Biobank (UKB) and a meta-analysis of the Genetic Investigation of Anthropometric Traits (GIANT) study and UKB, respectively. Genetic variants within or near the GLP1R gene associated with BMI or HbA1c at genome-wide significance were used for primary analyses. A missense GLP1R variant (rs10305492) associated with T2D and fasting glucose was used in secondary analyses.

Genetic associations with mental health outcomes were derived from the largest GWASs. Primary outcomes were SUDs, well-being spectrum, and mental health disorders, such as autism spectrum disorder (ASD), attention deficit/hyperactivity disorder [ADHD], posttraumatic stress disorder [PTSD], schizophrenia, bipolar disorder [BD], major depressive disorder [MDD], Tourette’s syndrome, and anorexia nervosa.

Secondary outcomes were SUD subtypes (alcohol dependence and cannabis use disorder), mental health disorder subtypes (BD I, BD II, and postpartum depression), and the four mental well-being dimensions (positive affect, depressive symptoms, life satisfaction, and neuroticism). Alcohol Use Disorders Identification Test total, alcohol consumption, and alcohol problems scores were also assessed. In addition, associations between GLP-1R activation and coronary artery disease and parental mortality, a proxy for all-cause mortality, were investigated, as GLP-1RAs have been reported to reduce cardiovascular events and deaths.

For replication analysis, genetic associations from the FinnGen dataset were used. The researchers performed a drug-target Mendelian randomization (MR) analysis. The F-statistic was used to evaluate instrument strength, and MR estimates were obtained using inverse-variance weighting. A Bonferroni-corrected significance threshold of p < 0.005 was applied to the primary outcomes. Pairwise colocalization analyses were performed to assess whether associations were driven by a shared causal variant.

Findings

Genetic variants used in the primary analysis for BMI were rs4714290 and rs17757975, and the variant used for HbA1c was rs10305518. Genetically predicted lower HbA1c and lower BMI via GLP-1R activation yielded point estimates consistent with lower coronary artery disease risk and lower parental mortality, a proxy for all-cause mortality, although the confidence intervals were wide. Genetically predicted lower BMI via GLP-1R activation was associated with better well-being, including better life satisfaction, lower neuroticism, fewer depressive symptoms, and greater positive affect. The overall well-being score was 0.06 standard deviations higher per genetically predicted 1 kg/m² reduction in BMI.

Further, genetically predicted lower BMI was associated with reduced risks of MDD, BD, and BD I after correction for multiple comparisons, while associations with SUDs, postpartum depression, ADHD, and BD II were suggestive. In the genetic analysis, each genetically predicted 1 kg/m² lower BMI through GLP-1R activation corresponded to approximately 18% lower odds of major depressive disorder and 39% lower odds of bipolar disorder. The association with MDD appeared stronger in females than in males, although the formal test for a sex difference was not statistically significant. Genetically predicted lower HbA1c via GLP-1R activation showed limited associations with mental well-being and SUDs. It showed suggestive associations with a reduced risk of Tourette’s syndrome but a higher risk of anorexia nervosa.

The missense variant was unavailable for the well-being spectrum analysis and showed little evidence of association with the assessed mental health outcomes. In FinnGen, the BMI-related GLP-1R proxy was associated with reduced risks of PTSD, ADHD, BD, alcohol dependence, substance abuse, and MDD. Genetically predicted lower HbA1c showed limited association with mental health outcomes, except for MDD. Colocalization analyses for BMI and significant outcomes revealed posterior probabilities of 76.9% for MDD, 59.3% for the well-being spectrum, and 45.9% for BD. These estimates were below the prespecified 80% threshold for colocalization. Conditional analyses exceeded 80% for the well-being spectrum and BD, but not MDD, suggesting that limited statistical power may have influenced the results.

Conclusions

In sum, the findings provide genetic evidence of associations between BMI-related GLP-1R activation and reduced risks of BD and MDD, as well as a better well-being spectrum. Moreover, there was suggestive evidence of an association between GLP-1R activation and a lower risk of SUDs.

The analysis relied on Mendelian randomization assumptions, used only a small number of GLP1R-region instruments, and could not completely exclude effects from neighboring genes or other biological pathways. It also modeled lifelong genetic differences rather than the shorter-term effects of taking GLP-1RAs. These results may inform research into psychiatric safety and potential repurposing opportunities for GLP-1RAs, especially in people with mental health disorders, but randomized clinical trials are needed to determine whether the genetic associations translate into treatment benefits.