Lead: A multinational observational study published in The Lancet Psychiatry in March 2026 reports that people treated with GLP-1–based weight-loss drugs showed lower rates of mental-health deterioration. Researchers from the University of Eastern Finland, Karolinska Institutet and Griffith University analyzed data from more than 95,000 adults (mean age 50.6) diagnosed with depression or anxiety who used antidiabetic medications between 2009 and 2022. Semaglutide users had substantially lower rates of worsening psychiatric outcomes, with smaller but measurable reductions observed for liraglutide. The authors stress that the study is observational and does not prove a causal link between the medications, weight loss and improved mental health.
Key takeaways
- Study cohort: more than 95,000 people with a diagnosis of depression or anxiety, mean age 50.6, who used antidiabetic medication from 2009–2022.
- Any worsening mental illness: semaglutide associated with a 42% lower risk; liraglutide associated with an 18% lower risk.
- Depression-specific outcome: a 44% lower risk observed (reported for semaglutide users in the analysis).
- Anxiety disorders: a 38% lower risk reported (reported for semaglutide users in the analysis).
- Substance use outcomes: hospital care and medically certified work leave related to substance use were 47% lower for semaglutide users.
- Self-harm: the analysis found lower recorded risk among people taking semaglutide, though authors gave no single summary percentage for this outcome.
- Design caveat: the study is observational and cannot establish direct causality; authors call for randomized clinical trials to test mechanisms.
Background
GLP-1 receptor agonists such as semaglutide and liraglutide mimic the gut hormone glucagon-like peptide-1, which helps regulate appetite and blood glucose. These agents are widely used in diabetes care and, more recently, at higher doses for treating obesity. Their rapid expansion into weight management has prompted intensive research into wider physiological and psychological effects.
Mental illness and metabolic disorders are closely linked. The World Health Organization estimates roughly one in six people in Europe—about 140 million—live with a mental health condition. International Diabetes Federation data indicate depression is nearly twice as common among adults with diabetes, and people with severe mental illness have two to three times higher rates of diabetes. Biological, behavioural and social factors contribute to a reciprocal relationship that complicates treatment and raises the stakes for interventions that might break the cycle.
Main event
The analysis, reported in The Lancet Psychiatry, pooled health-record data covering 2009–2022 and focused on adults who had a recorded diagnosis of depression or anxiety and who received antidiabetic medications during the study window. Investigators compared periods when individuals were taking GLP-1 receptor agonists to periods on other antidiabetic treatments to assess changes in psychiatric hospital care, medically certified leave for psychiatric reasons, substance-use outcomes and self-harm.
Semaglutide use was linked to a 42% lower risk of deteriorating mental illness overall; liraglutide showed an 18% lower risk for the same composite outcome. For specific diagnostic categories, semaglutide users experienced a 44% lower recorded risk of depression-related worsening and a 38% lower risk for anxiety disorders. Hospital care and work-leave related to substance use fell by 47% for semaglutide users in the reported models.
The study team repeatedly cautioned that these are associations observed in routine-care data, subject to confounding by indication, differences in baseline health, concurrent treatments and socioeconomic factors. They recommend randomized trials to test whether GLP-1 agents directly affect psychiatric symptoms, whether benefits are mediated by weight loss, glycaemic control or other biological pathways, or whether healthier patients are simply more likely to receive these treatments.
Analysis & implications
If confirmed in randomized trials, the findings could reshape clinical thinking about the broader benefits of GLP-1 treatments beyond metabolic endpoints. Potential biological mechanisms include central nervous system effects of GLP-1 signalling, reductions in systemic inflammation related to improved metabolic control, and psychosocial improvements tied to weight loss and better physical health.
However, the observational design leaves several competing explanations plausible. Patients prescribed semaglutide may differ from others in unmeasured ways—access to specialty care, concurrent psychotherapy or medication adherence—that also reduce psychiatric risk. Residual confounding and indication bias remain important caveats that the authors acknowledge.
Policy and clinical practice implications depend on replication. Clinicians should not assume a direct psychiatric benefit when prescribing GLP-1 drugs for obesity or diabetes, but the findings justify prospective trials that include psychiatric endpoints. Health systems should also consider equity: if GLP-1 drugs prove beneficial for mental health, access and cost barriers could widen disparities unless addressed.
| Outcome | Semaglutide (relative reduction) | Liraglutide (relative reduction) |
|---|---|---|
| Any worsening mental illness | 42% | 18% |
| Depression (worsening) | 44%* | — |
| Anxiety disorders (worsening) | 38%* | — |
| Substance use: hospital care & leave | 47% | — |
| Self-harm (recorded events) | Lower risk reported | — |
The table summarizes percentage reductions reported in the study. Absolute risks, confidence intervals and model adjustments are necessary to interpret clinical impact but were not reproduced here; the original paper includes those details.
Reactions & quotes
“Our observational results do not show causation and should be tested in randomized trials,”
The Lancet Psychiatry study authors
The study team emphasised careful interpretation, noting multiple possible confounders and urging clinical trials with psychiatric outcomes built in.
“About one in six people in Europe lives with a mental health condition,”
World Health Organization
Public-health bodies say the potential to reduce psychiatric burden by improving metabolic health would be welcome, but policymakers will require more rigorous evidence before changing guidance.
Unconfirmed
- Whether weight loss itself, improved glycaemic control, direct central nervous system effects, or patient selection explains the observed psychiatric improvements remains unconfirmed.
- Generalizability to people without diabetes or to younger or older age groups beyond the study mean (50.6 years) is uncertain.
- Long-term psychiatric outcomes and harms related to extended GLP-1 use were not established by this study.
Bottom line
The Lancet Psychiatry analysis finds substantial associations between semaglutide use and lower recorded risks of worsening depression, anxiety, substance-use outcomes and self-harm among more than 95,000 adults treated with antidiabetic drugs from 2009–2022. Liraglutide showed smaller but detectable associations on the composite psychiatric outcome.
These findings are promising but preliminary. They provide a strong rationale for randomized clinical trials that include psychiatric endpoints and careful measurement of weight change, metabolic markers, and psychosocial factors. Until such trials are completed, clinicians and patients should view any mental-health benefit as possible but not proven, and continue evidence-based psychiatric care alongside metabolic treatments.
Sources
- Euronews — news report summarizing the study (press)
- The Lancet Psychiatry — academic journal where the study was published (academic)
- World Health Organization — mental health prevalence and public-health context (official)
- International Diabetes Federation — diabetes and mental-health burden (industry/advocacy)