Scientific deep-dive

GLP-1 Drugs and Suicidal Thoughts: The Warning Came Off in 2026

In February 2026 the Suicidal Behavior and Ideation section was removed from the Wegovy, Zepbound and Saxenda labels, after three large studies looked for a signal and none found one.

By Nora Bissett · Pricing Editor
Editorially reviewed. Not medical advice · How we verify contentLast reviewed
7 min read·7 citations

In February 2026 the Suicidal Behavior and Ideation section was removed from the Wegovy, Zepbound and Saxenda labels.[4][5][6] Each of those labels now contains a single mention of the word, in the table recording that the section is gone. That happened after three large studies looked for a signal and none of them found one; one found the opposite. If you have read that these drugs carry a suicidality warning, that was true, and it is not true now.

If you are having thoughts of harming yourself, this page is not the thing you need right now. In the US you can call or text 988 for the Suicide & Crisis Lifeline, any time. Nothing below is a reason to wait, and nothing below means what you are feeling is not real or not urgent.

Where the concern came from

In 2023 European regulators opened an investigation after reports of suicidal thoughts in people taking semaglutide.[3] That was the right response to a safety signal: reports of that kind are exactly what post-marketing surveillance exists to catch, and the correct next step is to find out whether the drug is causing them or whether they would have occurred anyway in a large population.

There was also a plausible reason to worry. Bariatric surgery, the older standard treatment for obesity, has a small but real association with increased suicidality. A drug producing comparable weight loss raised a reasonable question about whether it might carry a comparable risk.

What the studies found

The three largest investigations of this question. None found an increase.
StudyDesign and sizeResult
JAMA Psychiatry, 2025[1]Meta-analysis of 27 randomized trials; 32,354 on a GLP-1 against 27,042 on placebo0.047 vs 0.042 events per 100 person-years — rate ratio 0.76 (95% CI 0.48–1.21)
JAMA Internal Medicine, 2024[2]Swedish and Danish national registers; 124,517 GLP-1 users against 174,036 on an SGLT2 inhibitor77 vs 71 suicide deaths — HR 1.25 (0.83–1.88), not significant
Nature Medicine, 2024[3]240,618 adults with overweight or obesity, replicated in 1,589,855 with type 2 diabetesLower risk of suicidal ideation on semaglutide — HR 0.27 incident, 0.44 recurrent

Read the register study past its headline. Suicide death alone showed no significant difference. The composite of suicide death and non-fatal self-harm came out lower on GLP-1 drugs, at a hazard ratio of 0.83 (95% CI 0.70–0.97). New diagnoses of depression and anxiety were flat, at 1.01 (0.97–1.06).[2]

Its authors also did what good studies do: they said how much uncertainty is left rather than declaring the matter closed. The upper bound of their confidence interval is compatible with an absolute increase of no more than 0.16 events per 1,000 person-years.[2] That is the ceiling on what the data can still hide.

The randomized evidence, pooled

A 2025 meta-analysis pooled 25 randomized controlled trials and found no significant association between these drugs and suicidal behavior: a risk ratio of 0.84 (95% CI 0.54–1.32, P = 0.46).[7]

What makes it unusually clean is the consistency. No subgroup reached significance, and every one had an I² of 0%, meaning no detectable variation between studies. The point estimates fall on both sides of 1, scattered around no effect: type 2 diabetes (0.74) and obesity (1.07), adolescents (0.91) and adults (0.84), suicidal ideation (1.04), attempts (0.68) and completed suicides (1.06), and each individual drug from dulaglutide (0.46) to lixisenatide (1.25).

Now the limit, which is structural rather than statistical. Randomized trials of these drugs routinely leave out anyone with a current psychiatric illness or past suicidality. So this pools 25 studies that had systematically screened out the population most at risk. The interval, 0.54 to 1.32, is wide because the events are rare. It is reassuring about the trial population and cannot speak to the group the concern was raised about.

What the label change says

Each of the three labels carries a Recent Major Changes table at the top, which records what was added or taken out and when. All three record the same thing in February 2026: the Suicidal Behavior and Ideation section, removed.[4][5][6]

Suicidal Behavior and Ideation (5.10) ……… (Removed) 02/2026
— WEGOVY prescribing information, Recent Major Changes

A regulator removing a warning is a stronger statement than a regulator declining to add one. Warnings are cheap to keep and expensive to take off, because taking one off invites exactly the criticism a regulator least wants. It happened across three products in the same month.

We publish this having just corrected ourselves on it. Our own drug pages still said Zepbound carried “suicidal ideation monitoring recommended” and that Wegovy’s was “under review by FDA”. Both were accurate when written and neither had been re-checked against the current label. They were fixed in the same change that published this page. A safety fact with a date on it needs re-reading, not remembering.

What this does not settle

  • Only 27 of 144 eligible trials systematically recorded these events.[1] The meta-analysis is built from the minority of trials that were looking, which is a real limitation and its authors say so.
  • The register study is mostly people with type 2 diabetes, at a mean age of 60.[2] That is not the same population as a 38-year-old taking Wegovy for weight.
  • Absence of a population signal is not a statement about you. A drug can fail to raise risk across 300,000 people and still be part of what is happening to one person.
  • Continued monitoring is the authors’ own recommendation, not a formality, because extended use in far larger numbers is still accumulating.[1]

If your mood has changed since starting

Tell your prescriber, and do not talk yourself out of it because a study said the risk is not elevated. Population evidence describes populations. Rapid weight loss changes sleep, eating, hormones and how people relate to their bodies, and any of that can move mood without the drug acting on the brain directly.

Stopping abruptly is not obviously the answer either, and it is not a decision to make alone. The same goes for an antidepressant you already take, as covered in antidepressants and a GLP-1. What the evidence supports is telling someone, promptly, and letting a clinician who knows your history weigh it. If you are in crisis, 988 in the US is available now, and is the right first call rather than the last one.

Frequently Asked Questions

The evidence does not support it. A meta-analysis of 27 randomized trials found event rates of 0.047 per 100 person-years on a GLP-1 against 0.042 on placebo, with no significant difference. Swedish and Danish national registers covering nearly 300,000 adults found no increase in suicide death, and a lower rate of the combined suicide-and-self-harm outcome. A 240,000-person cohort found a lower rate of suicidal ideation on semaglutide.
No. The Suicidal Behavior and Ideation section came out of the labels for Wegovy, Zepbound and Saxenda in February 2026, and each label now mentions the word only in the table recording that removal. Many sources still describe the warning as current because it was, until then.
European regulators opened an investigation in 2023 after post-marketing reports of suicidal thoughts, which is what surveillance is for. There was also a reasonable prior concern, since bariatric surgery, the older standard treatment for obesity, carries a small association with increased suicidality.
That is a decision for you and your prescriber, not one to make alone or from an article. Tell them promptly rather than waiting, and do not dismiss it because studies found no population-level signal; evidence about populations does not describe any individual. If you are in crisis, call or text 988 in the US now.
No, and its authors say so. Only 27 of 144 eligible trials systematically recorded these events, the register data are mostly people with type 2 diabetes at a mean age of 60, and the upper bound of that study's confidence interval still allows an absolute increase of up to 0.16 events per 1,000 person-years. The authors call for continued monitoring.

References

  1. 1.Ebrahimi P, Batlle JC, Ayati A, et al. Suicide and Self-Harm Events With GLP-1 Receptor Agonists in Adults With Diabetes or Obesity: A Systematic Review and Meta-Analysis JAMA Psychiatry. 2025. PMID: 40105856.
  2. 2.Ueda P, Söderling J, Wintzell V, et al. GLP-1 Receptor Agonist Use and Risk of Suicide Death JAMA Internal Medicine. 2024. PMID: 39226030.
  3. 3.Wang W, Volkow ND, Berger NA, et al. Association of semaglutide with risk of suicidal ideation in a real-world cohort Nature Medicine. 2024. PMID: 38182782.
  4. 4.Novo Nordisk Inc. WEGOVY (semaglutide) — US Prescribing Information, Recent Major Changes: Suicidal Behavior and Ideation (5.10) removed 02/2026 DailyMed (FDA-approved labeling). 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=ee06186f-2aa3-4990-a760-757579d8f77b
  5. 5.Eli Lilly and Company. ZEPBOUND (tirzepatide) — US Prescribing Information, Recent Major Changes: Suicidal Behavior and Ideation removed 02/2026 DailyMed (FDA-approved labeling). 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=487cd7e7-434c-4925-99fa-aa80b1cc776b
  6. 6.Novo Nordisk Inc. SAXENDA (liraglutide) — US Prescribing Information, Recent Major Changes: Suicidal Behavior and Ideation (5.9) removed 02/2026 (SPL effective 2026-02-25) DailyMed (FDA-approved labeling). 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=3946d389-0926-4f77-a708-0acb8153b143
  7. 7.Chen J, Zhang Q, Wu Q, et al. Impact of GLP-1 Receptor Agonists on Suicide Behavior: A Meta-Analysis Based on Randomized Controlled Trials Journal of Diabetes. 2025. PMID: 40887719.

Exenatide (Byetta, Bydureon): The First GLP-1, Its Weight Loss and Where It Went

Exenatide, sold as Byetta and Bydureon BCise, was the first GLP-1 drug. It is approved only for type 2 diabetes, produces far less weight loss than semaglutide, and both brands are discontinued in the US. Its one generic is now listed as discontinued too.

10 min read

Dizzy on a GLP-1? The Blood-Pressure Question the Labels Skip

Dizziness affects 4% to 8% of people on the weight-loss doses. The labels tell prescribers to lower insulin or sulfonylurea doses, and say nothing about blood-pressure pills, though low blood pressure is more common in people taking them.

7 min read

GLP-1 Allergic Reactions and Anaphylaxis: What the Labels and Cases Show

True allergy to semaglutide or tirzepatide is real but rare: about 3.6 anaphylaxis cases per 10,000 person-years on semaglutide. What each label says, how antibodies and inactive ingredients fit in, whether another GLP-1 is safe after a reaction, and when to call 911.

11 min read

Why a GLP-1 Raises Your Heart Rate, and When Palpitations Matter

Every GLP-1 raises resting heart rate by a few beats a minute, but only Wegovy and Saxenda warn about it. What the labels show, why it happens, whether it matters, and the palpitations that point elsewhere.

9 min read

Ozempic, Wegovy and Zepbound While Breastfeeding: What the Labels Say

Ozempic's label does not forbid breastfeeding; the semaglutide tablets and Foundayo advise against it, and tirzepatide is the only GLP-1 with a human milk study on its label. Each label's lactation wording, LactMed's summaries and the small milk studies.

11 min read

Tirzepatide Week by Week: What the SURMOUNT Trials Recorded

The dose ladder takes months, not weeks. This sets out what SURMOUNT-1, 2 and 4 recorded at each stage on Zepbound and Mounjaro, and the roughly half of the lost weight that came back when people stopped.

7 min read

Where to get tirzepatide online, safely: sellers our editors have checked

These are telehealth sellers our editors have checked. For each one we hold a price, the form the drug comes in, and the states it reaches.

No insurance needed · vetted by our editors

Some of the links on this page earn us money. If you sign up with a provider after following one, that provider may pay GLP Watchdog a commission. Learn more

6.4

Bodybuilding Health+

Seeing what it charges before you answer any medical questions

9.1

Collective

From $69/mo + membership

Get started →
Read full review →

About this analysis

GLP Watchdog publishes independently. Each seller in our dataset is checked against that seller's own website, and the dataset is updated continuously. Read our full methodology →