Review Finds No Clear Evidence That GLP-1 Drugs Cause Major Psychiatric Harm

GLP-1 receptor agonists are a group of medicines used for eligible people with type 2 diabetes and obesity. The group includes semaglutide, liraglutide and tirzepatide. As these medicines became more widely used, some patients and clinicians submitted reports of depression, anxiety and suicidal thoughts to medicine-safety databases.

Those early reports raised an important safety question. But a report submitted to a database is not the same as proof that a medicine caused the event. People taking a medicine may have other health conditions, may be taking other medicines, or may have more contact with health services, creating more chances for symptoms to be noticed and reported.

The new paper is an integrative review, meaning its authors brought together different kinds of previously published evidence rather than enrolling a new group of participants. They used targeted searches of peer-reviewed research and regulatory communications through early 2026. The review included studies of voluntary safety reports, cohort and case-control studies based on patient records, systematic reviews, meta-analyses, research on possible biological mechanisms and regulatory assessments.

The authors found that analyses of spontaneous safety reports showed modest signals for depression and suicidality that were mainly linked to semaglutide. Signals for anxiety followed a similar pattern. The abstract says liraglutide and tirzepatide generally did not show the same signals in those reporting databases.

However, the more controlled evidence pointed in a different direction. The review cites large studies that compared similar patients, including propensity-matched cohort studies, a nationwide case-time-control analysis and an administrative-claims study involving more than two million users. It also cites a meta-analysis of 91 placebo-controlled trials enrolling 107,910 participants.

Across this controlled evidence, the review says there was no increased risk of suicidality or major psychiatric harm. Some analyses even found a lower risk of suicidal thoughts among people using GLP-1 medicines. Because these studies were designed to better account for differences between people receiving and not receiving the medicines, they are more useful than voluntary reports for assessing whether a drug may be causing harm.

Still, “no increased risk overall” does not mean every person will have the same experience. The review identified an amplified reporting signal for suicidal thoughts among people who were also prescribed antidepressants or benzodiazepines. Benzodiazepines are medicines sometimes used for anxiety, panic or sleep problems. The authors describe this as possible effect modification: psychiatric vulnerability or related treatment may change how risk appears in a subgroup.

That signal does not establish that GLP-1 medicines caused suicidal thoughts in those patients. Reporting databases cannot fully separate the effects of a medicine from a person’s existing mental health needs, other treatments or the reasons they were prescribed a drug. But it means this group needs more focused research and continued attention in care.

The review also notes emerging research suggesting possible benefits related to substance use, binge eating and sleep-disordered breathing. These findings should not be taken as proof that GLP-1 medicines treat psychiatric conditions. The abstract does not provide enough detail to judge how strong those early findings are.

Overall, this is a synthesis of the current evidence, not a final answer for every person or every GLP-1 medicine. The authors conclude that controlled evidence has not demonstrated a causal relationship between GLP-1 medicines and suicidality or major psychiatric harm. They also say uncertainty remains for people with greater psychiatric vulnerability. More research that specifically includes and follows these groups would help clarify whether any risks differ by medicine, mental health history or other medications.