GLP-1 medicines and mental health / mood — UK research summary 2026

How the psychiatric signal emerged

From around 2023, as prescribing volumes for semaglutide and other GLP-1 receptor agonists grew rapidly, the MHRA and FDA began receiving spontaneous adverse-event reports through the Yellow Card scheme and FAERS describing mood changes, depressive symptoms and, in some reports, suicidal ideation in people taking these medicines. A rise in reports within a specific category is a standard trigger for a formal pharmacovigilance review, and both regulators opened reviews on this basis — signal detection followed by formal investigation, which is routine regulatory practice rather than a settled conclusion about risk. See psychiatric adverse-event signal tracking for the underlying signal-tracking detail behind this page.

What the regulatory reviews have found so far

The EMA's Pharmacovigilance Risk Assessment Committee has assessed reports of suicidal and self-harm ideation associated with GLP-1 receptor agonists and, in published review outcomes to date, has not confirmed a causal association based on the evidence available. The FDA has similarly continued to monitor FAERS reporting without issuing a confirmed causal finding. This is meaningfully different from a clearance or "no signal" conclusion: it reflects active, ongoing surveillance of a signal that has been formally investigated and not, at this stage, substantiated as causal. The current published statement from the relevant regulator is the definitive and most up-to-date reference, not this summary.

What pooled trial data show, and their limits

Pooled analyses of randomised controlled trial data across the class have generally not identified a statistically significant increase in suicidality or depressive-episode rates compared with placebo. Randomised evidence is considered stronger than spontaneous reporting because it includes a comparator group and prospective outcome capture. However, pivotal weight-management and diabetes trials were not designed with psychiatric outcomes as a primary pre-specified endpoint, and trial populations have typically excluded people with significant pre-existing psychiatric illness — limiting how confidently the reassuring pooled result generalises to everyone prescribed these medicines in practice.

What FAERS and Yellow Card data can and cannot establish

Spontaneous-reporting systems are valuable for hypothesis generation and early signal detection because they capture real-world use at a scale trials cannot match. They cannot, on their own, establish causation: there is no unexposed comparator group, reporting is voluntary and uneven, and reporting behaviour is itself influenced by media coverage and clinician awareness, independent of any true change in underlying risk. A rise in reports can reflect a genuine safety signal, rising prescribing volume, heightened attention, or a mixture of all three — which is why it triggers a review rather than an automatic labelling change.

Pre-existing mental health conditions and prescriber assessment

Current UK product information for licensed GLP-1 medicines does not carry a categorical contraindication for depression, anxiety or other pre-existing mental health conditions, but prescribers are expected to factor psychiatric history into overall risk-benefit assessment before initiating treatment. Anyone with a current or past significant mental health condition, or a history of self-harm, should raise this directly with the prescriber considering GLP-1 therapy, since individualised monitoring or specialist mental-health input may be appropriate depending on the specific clinical picture.

Weight loss itself as a confounding factor

Any intervention producing substantial, relatively rapid weight loss — pharmacological or otherwise — has been associated in the broader literature with mood changes in some individuals, in both directions: improved mood and self-esteem for many, alongside low mood or adjustment difficulty for others. This makes it genuinely difficult to separate a possible direct pharmacological effect on central appetite and reward circuitry from the psychological and physiological consequences of the weight change itself, and current evidence has not cleanly resolved which mechanism, if either, predominates in a given individual. See GLP-1 side effects A–Z and GLP-1 and alcohol for related central-nervous-system considerations discussed for this medicine class.

When to seek qualified clinical assessment

New or worsening low mood, anxiety, loss of interest in previously enjoyed activities, or any thoughts of self-harm or suicide while taking a GLP-1 medicine warrant prompt contact with a GP, prescriber or mental health service, regardless of whether the medicine is suspected to be a contributing factor. In the UK, NHS 111 and Samaritans (116 123) provide urgent support outside a routine appointment. Decisions about continuing, adjusting or stopping a GLP-1 medicine in response to a mood change are a prescriber decision following proper clinical assessment, not a step to take unilaterally based on general information. See GLP-1 medicines are UK prescription-only for the broader prescribing framework this sits within.

What the evidence does not establish

Current evidence does not establish that GLP-1 receptor agonists cause depression, anxiety or suicidality as a direct pharmacological effect, and it equally does not establish that no such risk exists in any individual or subgroup. Pooled RCT data are reassuring at a population level but were not purpose-built to detect psychiatric outcomes, and spontaneous-reporting signals cannot themselves confirm causation. Regulatory reviews have not substantiated a causal association, but active surveillance continues because the evidence base is still developing, particularly for longer-term use and for populations under-represented in pivotal trials. This page does not provide an individual risk assessment or a substitute for discussing personal or family psychiatric history with the prescriber managing treatment.

Frequently asked questions

Do GLP-1 medicines cause depression or suicidal thoughts?

This is not established. Regulators including the MHRA and FDA identified a potential psychiatric signal through post-marketing surveillance and opened formal reviews, but published reviews to date have not confirmed a causal link between GLP-1 receptor agonists and depression, anxiety or suicidality. The position from regulators is one of continued monitoring — the signal is taken seriously enough to review, but it has not been upgraded to a confirmed causal safety finding. Anyone experiencing mood changes or thoughts of self-harm while on a GLP-1 medicine should seek urgent qualified clinical help rather than relying on this page.

Why did regulators start looking at this in the first place?

From around 2023, the MHRA and FDA received a number of spontaneous reports through the Yellow Card scheme and FAERS respectively describing mood changes, depressive symptoms or suicidal ideation in people taking GLP-1 receptor agonists. Spontaneous-report volume alone does not establish causation, but a rise in reports for a specific adverse-event category is a standard trigger for regulators to open a formal pharmacovigilance review, which is what happened here.

What did the regulatory reviews conclude?

Regulatory reviews conducted so far, including EMA's Pharmacovigilance Risk Assessment Committee assessment, have not established a confirmed causal association between GLP-1 receptor agonists and suicidal ideation or self-harm based on the evidence reviewed to date. Reviews of this kind typically conclude either that a signal is not substantiated, that it requires continued monitoring, or that labelling should be updated. For this psychiatric signal, the outcome so far has generally been continued monitoring rather than either a clearance or a confirmed-risk finding — an evolving position rather than a closed question.

What do the pooled meta-analyses of randomised controlled trials show?

Pooled analyses drawing on randomised controlled trial data across the GLP-1 and dual-agonist class have generally not shown a statistically significant increase in suicidality or depressive-episode rates compared with placebo. Randomised trial data are considered stronger evidence than spontaneous reporting because they include a comparator arm and prospective, standardised outcome capture. However, pivotal weight-management and diabetes trials were not designed with psychiatric outcomes as their primary pre-specified endpoint, and trial populations typically exclude people with significant pre-existing psychiatric illness, which limits how confidently the findings generalise to the full range of people prescribed these medicines in practice.

If trial data look reassuring, why do regulators keep monitoring this?

Randomised trials are well suited to detecting common adverse events but are generally underpowered to detect rare events, including rare psychiatric outcomes, especially when psychiatric assessment was not the trial's primary purpose. Spontaneous-reporting systems like the Yellow Card scheme and FAERS exist precisely to catch signals that trials might miss once a medicine is used by much larger and more varied real-world populations. Continued surveillance reflects appropriate caution given this evidence gap, not a suspicion that the trial data are wrong.

What can FAERS and Yellow Card reports actually tell us, and what can't they?

Spontaneous adverse-event databases can flag a disproportionate reporting pattern worth investigating and can generate hypotheses for further study. They cannot establish causation, because there is no unexposed comparator group, reporting is voluntary and inconsistent, and people may be more likely to report an event they suspect is drug-related (a phenomenon regulators account for but cannot fully eliminate from the data). A rise in reports can reflect genuine increased risk, increased prescribing volume, increased media or clinician attention prompting more reporting, or some combination of all three. This is why regulators treat these databases as a starting point for investigation rather than a source of confirmed conclusions.

Does having a pre-existing mental health condition change the risk-benefit picture?

Product information (SmPC) for licensed GLP-1 medicines generally does not carry a categorical contraindication for pre-existing depression, anxiety or other mental health conditions, but prescribers are expected to take psychiatric history into account as part of routine risk-benefit assessment before and during treatment, as they would for many medicines. Anyone with a current or past significant mental health condition should discuss this directly with the prescriber initiating GLP-1 therapy, since individualised monitoring or a specialist mental-health input may be appropriate depending on the specific history.

Could mood changes on a GLP-1 medicine just be related to rapid weight loss itself, rather than the drug?

This is a genuine confounding factor that complicates interpretation of both trial and observational data. Significant weight change of any kind — whether from a GLP-1 medicine, other pharmacological treatment, bariatric surgery or intensive lifestyle intervention — has been associated in the broader literature with mood changes in some individuals, which can run in either direction: improved mood and self-esteem for some, and low mood, altered body image, or adjustment difficulty for others. Disentangling a possible direct pharmacological effect on central appetite and reward pathways from the psychological and physiological effects of substantial weight change itself is difficult with the data currently available, and both mechanisms may plausibly play some role in individual cases.

What symptoms should prompt someone to seek help while on a GLP-1 medicine?

New or worsening low mood, anxiety, hopelessness, loss of interest in previously enjoyed activities, or any thoughts of self-harm or suicide should prompt contact with a GP, prescriber or mental health service without delay, regardless of whether the person believes the GLP-1 medicine is responsible. In the UK, NHS 111 and Samaritans (116 123) are available for urgent support. This page is an evidence summary, not a substitute for clinical assessment of a specific individual's symptoms.

Should someone stop a GLP-1 medicine if they notice a mood change?

Whether to continue, adjust or stop treatment in response to a mood change is a prescriber decision that should follow a proper clinical assessment, not a step to take unilaterally based on this or any other general information page. A qualified clinician can weigh the specific symptoms, timing, personal and family psychiatric history, and the metabolic benefits of continued treatment for that individual. This page does not provide dosing, titration or discontinuation guidance.