GLP-1 drugs looked promising for depression. The evidence tells a more complicated story

By Pooja Toshniwal PahariaReviewed by Susha Cheriyedath, M.Sc.Sep 7 2026

From laboratory models to global health records, researchers traced how widely used metabolic drugs interact with mood, cognition, and stress-related biology across strikingly different settings.

Systematic Review: Mind over metabolism: a systematic review of GLP-1 receptor agonists in the treatment of stress-related disorders. Image Credit: Love Employee / Shutterstock

In a recent systematic review published in the journal Translational Psychiatry, researchers evaluated existing evidence on the potential use of glucagon-like peptide-1 receptor agonists (GLP-1RAs) for treating depression and symptoms relevant to post-traumatic stress disorder (PTSD). The data were inconsistent for depression- and anxiety-related outcomes, underscoring the need for targeted clinical and mechanistic studies, particularly because no included study directly evaluated GLP-1RAs as a treatment for PTSD.

Depression and PTSD are debilitating conditions that affect people worldwide. Symptoms may recur even after standard treatment, potentially affecting physical health and quality of life over time. This has prompted interest in developing more effective approaches to reduce the burden of these disorders. As metabolic disturbances may influence mental health, researchers are exploring whether drugs that target metabolic pathways could also help manage psychiatric conditions.

About the review

In the present review, researchers explored GLP-1 RAs as potential treatments for psychiatric disorders, given their effects on metabolism and inflammation, their established use in treating type 2 diabetes (T2D), and the presence of GLP-1 receptors in brain regions involved in stress and mood regulation. Based on peer-reviewed preclinical and human studies identified through searches of Scopus, Web of Science, and PubMed for English-language publications available through December 2025, the authors investigated whether repurposed GLP-1RAs could be used to treat depression or PTSD-related symptoms.

The team used the Hawker tool to assess the quality of the included studies. Studies with overlapping data, unavailable outcome data after author contact, no comparator group, or, for human studies, participants with co-occurring alcohol or drug misuse were excluded. Preclinical studies were required to use a valid model or to examine outcomes relevant to depression or PTSD. In addition, the researchers did not include reviews and studies with multi-receptor agonists or sample populations and outcomes not relevant to depression or PTSD.

Two researchers independently screened the studies and resolved differences through discussion or by consulting another researcher. Measures such as the Beck Depression Inventory (BDI), Clinician-Administered PTSD Scale for DSM-5 (CAPS-5), and Patient Health Questionnaire-9 (PHQ-9) were cited as examples of validated instruments that could establish clinically significant symptoms in eligible human studies.

Results

The initial search yielded 514 records. After excluding duplicates and ineligible records, 43 studies comprising 51 datasets were included in the final review. All the included studies were of at least medium quality. However, these ratings did not remove limitations related to observational designs, indirect outcomes, and nonpsychiatric study populations. The team reviewed 28 rodent studies and 15 human studies, many of which used observational, database, or pharmacovigilance (adverse-event reporting) designs. The data were inconsistent for depression- and anxiety-related outcomes; no studies directly evaluated GLP-1RAs for PTSD, and only one included study reported lixisenatide-specific evidence.

The results were highly heterogeneous across studies. While some reported improvements in anxiety- or depression-related outcomes, others reported minimal or inconsistent effects. Exenatide, in particular, showed different effects on mood-related measures across studies, which appeared to vary with the neurobiological setting, concomitant treatment, and psychosocial stress burden.

In one early animal study, exendin-4 improved cognitive performance and reduced immobility in the forced-swim test after one to two weeks of administration, although this behavioral measure does not capture the complexity of human depression. In a prospective observational study of exenatide in patients with T2D, psychological measures and quality of life also improved after six months of treatment. In a separate randomized trial involving people with Parkinson’s disease, mood-related improvements after 48 weeks were not sustained following treatment withdrawal. Preclinical studies demonstrated conflicting results for anxiety-related outcomes. For exenatide, anxiety-reducing effects were primarily observed in models of metabolic dysfunction and were not found in studies using validated depression models.

Across GLP-1RAs as a class, observational studies produced mixed findings on psychiatric outcomes. In one study with five years of follow-up, GLP-1RA use was associated with a lower incidence of suicide attempts than treatment with dipeptidyl peptidase-4 (DPP-4) inhibitors. This association was also observed in a high-risk subgroup that included people with prior depression, people using antidepressants, and those who had previously attempted suicide. One global real-world analysis reported associations with lower risks of anxiety, depression, and suicide; however, the effects of individual drugs need further clarification.

Data obtained from the United States Medicare claims, the United Kingdom Clinical Practice Research Datalink, and Valencia Health System Integrated Database produced comparator-dependent or null results. Medicare data showed no reduction in incident depression relative to sodium-glucose cotransporter-2 (SGLT2) inhibitors but a modest reduction relative to DPP-4 inhibitors. The UK database found no difference in new depression or self-harm compared with sulfonylureas, while the Valencia database found no association with suicidal ideation or self-injury compared with SGLT2 inhibitors.

In mice exposed to chronic stress, dulaglutide reduced several depression-like behaviors. The treatment reduced signs of anhedonia, social withdrawal, and despair-like behavior in the animals. Neither study included baseline behavioral assessments, limiting confidence that the differences arose entirely from treatment. However, the anxiety-related findings were less consistent. In several stress- or metabolically driven animal models, liraglutide also reduced depression- and anxiety-like behaviors and demonstrated cognitive benefits; however, overall preclinical evidence was mixed, and most clinical studies did not demonstrate clear psychiatric benefits. One positive pharmacovigilance finding was based on reduced reporting of antidepressant treatment failure and cannot establish therapeutic efficacy.

Across two diabetes-based mouse models, semaglutide reduced anxiety-like and diabetes-associated depression-like behaviors, while one study also reported cognitive improvements. However, clinical studies involving people with obesity did not provide consistent evidence of psychiatric benefits. In particular, one small randomized trial involving 35 overweight or obese adults with depression and cognitive dysfunction found no significant improvement in depressive symptoms, cognition, or suicide risk. Pharmacovigilance analyses also produced conflicting signals involving antidepressant treatment failure, depression, and self-injury, but these reporting patterns cannot establish benefit or harm. The authors cautioned that psychiatric use would require careful screening for eating disorders and clinically significant weight loss.

Conclusion

The study findings do not currently support using GLP-1R agonists as treatments for depression or PTSD. The therapeutic benefits, if any, may be particularly relevant among people with metabolic dysfunction, potentially reflecting the effects of GLP-1R agonism on metabolic and inflammatory pathways.

In future studies, researchers should compare individual GLP-1RAs, investigate their mechanisms, and determine whether metabolic comorbidity influences treatment response in depression and PTSD. Including validated models of trauma-related pathology alongside targeted clinical studies could also help determine whether GLP-1RAs could have therapeutic potential for psychiatric disorders.