Certain GLP-1 receptor agonists (RAs) may offer benefits beyond diabetes and weight loss, with emerging evidence suggesting they may have potential positive effects on mental health. 

In a large observational study of adults with pre-existing depression or anxiety, semaglutide use — and, to a lesser extent, liraglutide — was associated with a reduced risk of worsening mental health. Semaglutide was also linked to a lower risk of worsening depression, anxiety, and substance use disorders. 

The study was published online March 18 in Lancet Psychiatry.

Debate Settled?

People with diabetes are at increased risk of depression, anxiety, and suicide. GLP-1 RAs are widely used to treat diabetes and obesity, but data on whether these medications alleviate or exacerbate anxiety, depression, and self-harm are mixed. 

“We would need randomized clinical trials to really settle the dispute for good. However, our study is very large and comprehensive (nationwide and inclusive), so I think we can be pretty certain that there is an association with using GLP-1s and reduced psychiatric symptoms,” Markku Lähteenvuo, MD, PhD, research director, University of Eastern Finland in Kuopio, told Medscape Medical News. 

The researchers analyzed Swedish national health registers spanning from 2009 to 2022. The cohort comprised 95,490 adults (mean age, 50 years; 60% women) with diagnosed depression, anxiety disorders, or both, who were prescribed non-insulin antidiabetic medications during the study period. 

To reduce confounding, a within-individual design was used for all comparisons — meaning each participant acted as their own control. Periods when individuals were using specific GLP-1 RAs were compared with periods when they were not using these drugs.

During a mean follow-up of 5.2 years, GLP-1 RAs were used by 22,480 individuals (23.5% of the total cohort), with roughly 60% prescribed semaglutide and 47% prescribed liraglutide. 

No Causal Link

The primary outcome — worsening mental illness, defined as a composite of psychiatric hospitalization, sick leave of more than 14 days for psychiatric reasons, hospitalization for self-harm, or death by suicide — occurred in 32,638 participants at least once during the study period. 

Compared with no use of a GLP-1 RA, semaglutide use was linked to a 42% decreased risk of worsening mental health (adjusted hazard ratio [aHR], 0.58), whereas liraglutide was associated with an 18% reduction of worsening mental health (aHR, 0.82). 

Semaglutide was also associated with a 44% decreased risk of worsening depression (aHR, 0.56), a 38% decreased risk of worsening anxiety (aHR, 0.62), and a 47% decreased risk of worsening substance use disorder (aHR, 0.53). 

Liraglutide was associated with a 26% lower risk of worsening depression (aHR, 0.74) but no clear reduction in worsening anxiety. 

Other GLP-1 RAs, including exenatide and dulaglutide, showed no significant association with mental health outcomes, although these analyses were underpowered due to smaller sample sizes. As a group GLP-1 RAs were associated with a 44% reduced risk of self-harm (aHR, 0.56).

Lähteenvuo said it’s plausible that semaglutide might have greater mental health benefits than liraglutide. 

“Semaglutide seems to work better for weight loss and glycemic control than liraglutide (or the other studied GLP-1s). Thus, this could indicate that semaglutide is just more potent in general and that the effects we observe for the psychiatric symptoms are related to the overall efficacy of the GLP-1,” Lähteenvuo noted. 

Comparisons with second-line diabetes medications also suggested advantages for semaglutide. Compared with empagliflozin, semaglutide use was linked to a reduced risk of worsening mental illness (aHR, 0.73), while dapagliflozin (aHR, 1.24) and sitagliptin (aHR, 1.49) were associated with increased risk. 

Given that the study was observational, Lähteenvuo said that a causal relationship cannot be established and that it remains uncertain whether GLP-1 use directly leads to these mental health benefits. 

However, he noted that the greater benefits seen with certain GLP-1 agents — broadly consistent with their weight-loss effects — support the possibility that these outcomes are attributable to the drugs themselves rather than other factors. 

Reassuring Results

Reached for comment, Thomas Rutledge, PhD, staff psychologist, VA San Diego Healthcare System, and professor of psychiatry, University of California, San Diego, said the study results “align with the most rigorous prior studies on the topic of GLP-mental health.”

Rutledge cautioned that a major limitation of observational research is that individuals with diabetes or obesity already have higher baseline rates of anxiety and depression, regardless of GLP-1 use. As a result, any association between these medications and mental health outcomes should be interpreted carefully, as such studies cannot establish causality. 

He added that while GLP-1s may improve mental health in some patients, there is considerable variability in the drugs’ effects between individuals. Factors such as reductions in systemic inflammation, meaningful weight loss, improved sleep, and enhanced quality of life may contribute to better psychological outcomes. 

However, in cases where biological improvements are modest, weight loss is limited, or lean tissue loss occurs, mental health may remain unchanged or even worsen. Clinically, he noted, both patterns are observed.

Also commenting on the study, Riccardo De Giorgi, MD, DPhil, MRCPsych, clinical lecturer, honorary consultant in general adult psychiatry, and adjunct professor within the Department of Psychiatry at the University of Oxford, England, told Medscape Medical News that the findings are “reassuring.” 

In this population, GLP-1 RAs do not appear to be associated with worsening mental health and may, in fact, be linked to a lower risk of clinically significant deterioration.

“Still, we should remain vigilant. There continue to be non-sporadic reports of mood worsening and suicidal behavior in GLP-1 RA users, and this study cannot fully address associations outside the metabolic-treatment context, nor in other major psychiatric disorders beyond depression/anxiety,” De Giorgi cautioned. 

Rutledge said the next research step should be “better controlled studies — not giant observational studies — so that individuals can be tracked closely to monitor adherence, biological mechanism and psychosocial changes over time. The latter are necessary if we ever hope to understand who improves, when, and why.”

De Giorgi echoed the need for further study. He noted that only a minority of the patients in this study were taking GLP-1 RAs for weight loss/obesity — “and we know that this population may also be particularly vulnerable to mental health deterioration when taking weight-loss treatments.”

“We still need to understand who are the people that are more likely to benefit from these medications in terms of their mental health, and who are those that might, conversely, experience psychiatric adverse events,” he said. 

Overall, he said, the study is “methodologically strong with valid and important findings, whose key implication is that the evidence-base is now stronger and supports moving to targeted randomized trials measuring mental health outcomes directly in people with pre-existing psychiatric disorders.”

This study was funded by the Sigrid Jusélius Foundation, Jane and Aatos Erkko Foundation, and Finnish Ministry of Social Affairs and Health. Lähteenvuo has received honoraria from Lundbeck, Otsuka Pharma, Janssen, Johnson & Johnson, Orion Pharma, and Recordati. Rutledge and De Giorgi had no relevant disclosures.