This page was translated automatically from Turkish. Read the original

Association Between GLP-1 Receptor Agonist Use and Worsening Mental Illness in People with Depression and Anxiety in Sweden: A Nationwide Cohort Study

March 29, 2026Recent Articles

Authors: Heidi Taipale, Mark Taylor, Markku Lähteenvuo, Ellenor Mittendorfer-Rutz, Antti Tanskanen, Jari Tiihonen

Abstract

Background: Individuals with diabetes are at increased risk of depression, anxiety, and suicide. GLP-1 receptor agonists are approved for the treatment of diabetes and obesity; however, evidence is mixed as to whether these drugs alleviate or worsen anxiety, depression, and self-harm. We examined the risk of worsening mental illness in people already diagnosed with depression, anxiety, or both who were prescribed antidiabetic medication, including GLP-1 receptor agonists.

Methods: The study cohort, identified from Swedish national electronic health records, included people diagnosed with depression or an anxiety disorder who used any antidiabetic medication between 2009 and 2022. To reduce confounding, all comparisons used a within-individual design comparing periods of medication use with periods of non-use in the same individual. The primary outcome was worsening mental illness, defined as a composite of psychiatric hospitalization, sickness absence lasting more than 14 days for psychiatric reasons, hospitalization for self-harm, or death by suicide. Secondary outcomes were worsening depression or anxiety, worsening substance use disorder, and self-harm, analyzed separately. Within-individual stratified Cox models were used, with adjusted hazard ratios (aHRs) and 95% CIs. A person with relevant lived experience was involved in the design and writing of the study.

Findings: The cohort included 95,490 people with a mean age of 50.6 years (SD 12.3). During follow-up, 22,480 individuals used a GLP-1 receptor agonist. Compared with periods without GLP-1 receptor agonist use, semaglutide (aHR 0.58 [95% CI 0.51–0.65]) and liraglutide (0.82 [0.76–0.89]) were associated with a lower risk of worsening mental illness, whereas exenatide (1.01 [0.69–1.46]) and dulaglutide (1.01 [0.85–1.20]) were not. Semaglutide was associated with a reduced risk of worsening depression (0.56 [0.44–0.71]), worsening anxiety (0.62 [0.52–0.73]), and worsening substance use disorder (0.53 [0.35–0.80]). Liraglutide was associated only with a lower risk of worsening depression (0.74 [0.64–0.87]). As a group, GLP-1 receptor agonists were associated with a reduced risk of self-harm (0.56 [0.34–0.92]).

Interpretation: Semaglutide and, to a lesser extent, liraglutide may be beneficial, dual-action therapeutic options for anxiety and depression co-occurring with diabetes and obesity. Randomized controlled trials are needed to evaluate these findings.

Funding: Sigrid Jusélius Foundation, Jane and Aatos Erkko Foundation, and the Finnish Ministry of Social Affairs and Health.

Introduction

GLP-1 receptor agonists (also known as incretin mimetics) are approved for the treatment of diabetes and obesity in many countries; they mimic the endogenous hormone GLP-1 by inhibiting glucagon and stimulating insulin secretion. GLP-1 receptor agonists also increase satiety or reduce hunger and cravings. It is estimated that more than 15 million people (6%) in the United States have been formally prescribed a GLP-1 receptor agonist, although the actual number may be higher due to online and overseas supplies.

People with diabetes are more likely to develop depression and anxiety than the general population and have an increased risk of death by suicide. There appears to be a cyclical relationship between metabolic dysregulation, such as diabetes or obesity, and anxiety and depression. In addition, depression and anxiety are currently among the leading causes of health-related sickness absence. Recent research has suggested that GLP-1 receptor agonists may have neuropsychiatric benefits for cognition, substance misuse or addiction, and mood.

Research in Context

Evidence before this study: People with diabetes or obesity are more likely to experience depression and anxiety than the general population. It is unclear whether GLP-1 receptor agonists reduce or increase this risk. A systematic search of PubMed and Google Scholar using the terms "GLP-1", "GLP-1RA", "depression", "anxiety", and "suicide" identified eight systematic reviews and four meta-analyses published in the past 12 months. Adverse psychiatric outcomes associated with GLP-1 receptor agonists in pharmacovigilance and observational studies have included depression, anxiety, and suicidal ideation, raising concerns among regulatory agencies. Individual studies have yielded mixed results; however, the most recent large-scale meta-analytic reviews incorporating placebo data have been more reassuring, noting no increased risk of suicide or greater depression severity across studies of GLP-1 receptor agonists.

Added value of this study: This is the first study to show that people using GLP-1 receptor agonists are less likely to experience worsening mental disorders. Although other recent data have found an increased risk of anxiety, depression, and suicide in a large cohort, these conflicting findings may be due to selection bias. To avoid this bias, we used a within-individual design in which each individual acts as their own control, an approach considered to reduce confounding from individual patient characteristics. Differences in hospitalization, sickness absence, and depression/anxiety outcomes were observed according to the type of GLP-1 receptor agonist, suggesting that this is not a class effect. Semaglutide was the most effective of those examined, followed by liraglutide.

Implications of all the available evidence: GLP-1 receptor agonists may prevent worsening depression and anxiety. Randomized controlled trials are needed to test these dual-action associations.

Methods

Study Design and Population: Data from Swedish national electronic health records were linked using anonymized identification numbers. Sweden has a decentralized, publicly funded healthcare system. The National Patient Register and the MiDAS register, which includes sickness absence data, were used to identify participants. The cohort included people diagnosed with depression (ICD-10 F32-F33) or an anxiety disorder (F40-F43) who used non-insulin antidiabetic medication between 2009 and 2022. People diagnosed with schizophrenia spectrum disorders or bipolar disorder were excluded because these represent more severe clinical conditions. Follow-up ended at death, emigration, onset of schizophrenia/bipolar disorder, or December 31, 2022.

Exposures: The PRE2DUP method was used to construct periods of medication use. The exposures analyzed were the use of semaglutide, liraglutide, exenatide, and dulaglutide. A grouped category including all GLP-1 receptor agonists was also created. Periods without GLP-1 receptor agonist use served as the reference for comparison. To control for diabetes severity, GLP-1 receptor agonists were also compared with other commonly used second-line antidiabetic medications, such as empagliflozin, dapagliflozin, and sitagliptin.

Outcomes: The primary outcome (worsening mental illness) was a composite of psychiatric hospitalizations, sickness absence lasting more than 14 days for psychiatric reasons, hospitalization for self-harm, and death by suicide. Secondary outcomes were defined as worsening anxiety disorders, worsening depression, and worsening substance use disorder.

Statistical Analysis: A within-individual design was used, in which each individual served as their own control. Risks were calculated using Cox regression models. This design automatically eliminates the effects of time-invariant characteristics, such as sex and baseline illness severity. Analyses were adjusted for time-varying variables, such as the use of other antidiabetic medications, antipsychotics, mood stabilizers, and benzodiazepines.

Findings

The total cohort comprised 95,490 individuals, with a mean age of 50.6 years. Of the cohort, 59.7% were women and 40.3% were men. An anxiety disorder was diagnosed in 81.5% and depression in 54.9%. The mean follow-up duration was 5.2 years.

Main Results:

Semaglutide use (aHR 0.58 [95% CI 0.51–0.65]) and liraglutide use (0.82 [0.76–0.89]) were associated with a reduced risk of worsening mental illness compared with periods of non-use.
This association was not observed for exenatide (1.01) or dulaglutide (1.01).
Semaglutide was associated with a lower risk (aHR 0.67) when directly compared with other GLP-1 receptor agonists.
In the analysis by sex, a reduced risk with liraglutide was observed only in women (0.78), not in men (0.95).

Secondary and Other Results:

Semaglutide was found to be associated with both worsening depression (0.56) and worsening anxiety (0.62).
Liraglutide was associated only with worsening depression (0.74).
Semaglutide use was associated with a reduced risk of worsening substance use disorder (0.53).
At the group level, GLP-1 receptor agonists were associated with a reduced risk of self-harm (0.56).
Deaths by suicide (171 deaths in total) contributed only 0.2% to the primary outcome.
Compared with empagliflozin, semaglutide was associated with a lower risk (0.73), whereas dapagliflozin (1.24) and sitagliptin (1.49) were associated with a higher risk.

Discussion

Semaglutide and, to a lesser extent, liraglutide were found to significantly reduce the risk of worsening mental illness. Semaglutide and liraglutide also yielded better outcomes when compared with other second-line antidiabetic medications. Overall, GLP-1 use was associated with a reduced risk of self-harm and substance use disorder.

This association does not appear to be a "class effect", as similar effects were not found for exenatide or dulaglutide. The superiority of semaglutide over the others is consistent with its greater weight loss and better glycemic control. However, our study does not provide evidence that weight loss directly improves mental health; this relationship is likely complex.

Mechanisms: There is a theoretical basis for GLP-1 receptor agonists improving symptoms of anxiety and depression through mesolimbic and frontocortical serotonergic and dopaminergic pathways. The hormone GLP-1 may reduce neuroinflammation by decreasing oxidative stress. Neuroprotective and neurotrophic properties have been demonstrated in animal studies; it may induce the expression of genes involved in neuronal survival and regeneration.

Limitations: The results can only be generalized to systems such as Sweden's, where healthcare is free of charge. Causality cannot be established in an observational study. Individual data such as symptom severity, weight change, or BMI are not available in national registers. Statistical power is low for some medications. Residual confounding is always a possibility.

Conclusion: Using 14 years of nationwide observational data, we showed that semaglutide and liraglutide were associated with a significant reduction in the risk of worsening mental illness. Our findings suggest that a randomized controlled trial of GLP-1 receptor agonist use in people with diabetes and depression/anxiety would be valuable.

Administrative Information

Contributors: Study concept: ML, HT, and JT; Analytical design: HT and AT; Data analysis: HT and AT; Draft writing: MT and HT .

Conflicts of Interest: HT, EM-R, and JT have been involved in research grants provided by Janssen to their institutions. HT has received consulting fees or honoraria from companies such as Gedeon Richter, Lundbeck, and Otsuka. ML has received honoraria from companies such as Johnson & Johnson and Orion Pharma. MT declared no conflicts of interest .

Data Sharing: Swedish data cannot be made publicly available due to privacy regulations; however, access for research purposes may be granted following legal review .

Acknowledgments: This study was funded by the Sigrid Jusélius Foundation, the Jane and Aatos Erkko Foundation, and the Finnish Ministry of Social Affairs and Health.

← Previous

Why does a physical wound heal within weeks, while a wound inflicted by a 'word' continues to bleed for years?

What are the effects of saffron during pregnancy?