GLP-1s & Mental Health (2026): What the FDA's January 2026 Suicidality Warning Removal Means for Mood, Anhedonia & Depression

On January 13, 2026, the FDA requested removal of the suicidal-behavior-and-ideation (SI/B) warning from the labels of Wegovy, Saxenda, and Zepbound after a meta-analysis of 91 placebo-controlled trials in 107,910 patients (60,338 on a GLP-1 RA, 47,572 on placebo) found no increased risk. The European Medicines Agency reached the same conclusion in April 2024: “the available evidence does not support a causal association.” Four large, peer-reviewed studies — the Wadden STEP-trials post-hoc (JAMA Intern Med 2024), the Wang TriNetX real-world cohort (Nature Medicine 2024), the Ueda Sweden + Denmark binational registry (JAMA Intern Med 2024), and the McIntyre Bradford-Hill FAERS analysis (Expert Opin Drug Saf 2024) — all align with the regulators. This article covers what changed, what the evidence actually shows, what current FDA labels still say about anhedonia and the “loss of food joy” reports, and what to do if you experience mood changes on a GLP-1.

What changed: January 2026 FDA action

On January 13, 2026, the FDA issued a Drug Safety Communication titled “FDA Requests Removal of Suicidal Behavior and Ideation Warning from Glucagon-like Peptide-1 (GLP-1) Receptor Agonist Labeling.” The DSC announced that the agency had reviewed 91 placebo-controlled clinical trials covering 107,910 patients — 60,338 on a GLP-1 RA (semaglutide, liraglutide, tirzepatide, dulaglutide, exenatide, lixisenatide) and 47,572 on placebo — and concluded that the data did not identify an increased risk of suicidal ideation or behavior, or of related psychiatric adverse events including anxiety, depression, irritability, or psychosis.

The agency formally requested that manufacturers remove the suicidal-behavior-and-ideation §5 subsection from the FDA prescribing information for the three weight-management GLP-1 labels that previously carried it: Wegovy (semaglutide 2.4 mg, Novo Nordisk), Saxenda (liraglutide 3 mg, Novo Nordisk), and Zepbound (tirzepatide, Eli Lilly). The diabetes-only labels — Ozempic, Mounjaro, Trulicity, Victoza, Rybelsus — never carried this specific subsection because the original signal had emerged from weight-management indications, not glycemic.

As of February 2026, the “Recent Major Changes” block at the top of the Wegovy, Saxenda, and Zepbound labels on DailyMed reads “Suicidal Behavior and Ideation … (Removed) 02/2026.” The §5 WARNINGS AND PRECAUTIONS subsection on the topic has been struck. We checked all three labels on 2026-05-09 and confirmed the removal is live.

EMA reached the same conclusion in April 2024

The FDA was second. The European Medicines Agency's Pharmacovigilance Risk Assessment Committee (PRAC) had already concluded its parallel review on April 12, 2024. The PRAC meeting highlights state verbatim:

“the available evidence does not support a causal association between the Glucagon-Like Peptide-1 receptor agonists (GLP-1) – dulaglutide, exenatide, liraglutide, lixisenatide and semaglutide – and suicidal and self-injurious thoughts and actions.”

The PRAC review had been triggered in July 2023 after Iceland flagged a small number of post-marketing reports of suicidal thoughts and self-injury among GLP-1 RA users. The committee spent nine months reviewing all available data — clinical trials, post-marketing surveillance, EudraVigilance reports, and published epidemiologic studies — and concluded that no update to product information was warranted. Tirzepatide (Mounjaro / Zepbound) was added to the review in scope after the original referral and reached the same conclusion.

The published primary-source evidence base

Four large peer-reviewed studies anchor the FDA and EMA decisions. None paraphrase Reddit. All four were published in first-tier indexed journals.

Wadden et al., STEP 1/2/3/5 post-hoc — JAMA Intern Med 2024

This is the largest pooled prospective dataset on semaglutide 2.4 mg and mood. Wadden and colleagues pooled 3,377 STEP 1/2/3 participants + 304 STEP 5 participants — all randomized, all using the validated Patient Health Questionnaire-9 (PHQ-9) for depression and the Columbia-Suicide Severity Rating Scale (C-SSRS) for suicidal ideation. Note that the STEP trials excluded participants with current major depression, serious mental illness, or prior suicide attempt — so this dataset measures GLP-1 mood safety in patients without known active psychopathology, not in the highest-risk populations.

The verbatim findings (PMID 39226070):

“1% or fewer of participants reported suicidal ideation/behavior during treatment, with no differences between semaglutide, 2.4 mg, and placebo.”

On the depression scale, the post-hoc found lower PHQ-9 scores at week 68 in the semaglutide arm than placebo (mean 2.0 [SD 2.9] vs 2.4 [SD 3.3]; treatment difference −0.56, p<0.001) and a lower odds ratio (0.63, p<0.001) of shifting into a more severe depression category over the study period.

Wang et al., TriNetX EHR cohort — Nature Medicine 2024

Wang and colleagues used the TriNetX research network to compare incident suicidal ideation between adults newly prescribed semaglutide and adults newly prescribed a non-GLP-1 anti-obesity or anti-diabetes medication, using propensity-score matching. The overweight/obesity cohort included 240,618 patients; the type 2 diabetes cohort included 1,589,855 patients (PMID 38182782). The hazard ratios:

• Incident suicidal ideation, overweight/obesity cohort: HR 0.27 (95% CI 0.20-0.36)
• Recurrent suicidal ideation, overweight/obesity cohort: HR 0.44 (95% CI 0.32-0.60)
• T2D cohort replicated the direction and significance.

The authors' verbatim conclusion:

“Our findings do not support higher risks of suicidal ideation with semaglutide compared with non-GLP1R agonist anti-obesity or anti-diabetes medications.”

Ueda et al., Sweden + Denmark binational registry — JAMA Intern Med 2024

Ueda and colleagues used nationwide health registries from Sweden and Denmark covering 2013-2021. The exposure cohort was adults with type 2 diabetes initiating a GLP-1 receptor agonist; the active comparator was adults initiating an SGLT2 inhibitor (a non-GLP-1 second-line T2D drug class with no suicidality signal). The primary outcome was suicide death (PMID 39226030):

• Suicide death: HR 1.25 (95% CI 0.83-1.88) — confidence interval crosses 1.0; the comparison is not statistically significant
• Self-harm hospitalization: not significant
• Incident depression and anxiety-related disorders: not significant

The authors' verbatim conclusion:

“use of GLP-1 receptor agonists compared with SGLT2 inhibitors was not associated with an increased risk of suicide death, self-harm, or incident depression and anxiety-related disorders.”

McIntyre et al., FAERS Bradford-Hill analysis — Expert Opin Drug Saf 2024

Roger McIntyre's group analyzed FDA Adverse Event Reporting System (FAERS) reports for semaglutide and liraglutide and applied the Bradford-Hill criteria — the canonical epidemiologic framework for distinguishing association from causation (PMID 38087976). FAERS, on its face, does show disproportionate reporting of suicidal ideation with semaglutide and liraglutide. But disproportionate reporting is not causal evidence — FAERS is a passive spontaneous-reporting system and is highly subject to media attention, regulatory signal cycles, and stimulated reporting. After applying Bradford-Hill (strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, analogy):

“Using the Bradford Hill criteria, however, and taking into consideration confounders, no causal link between GLP-1 RAs and suicidality exists.”

Tobaiqy and Elkout's parallel EudraVigilance analysis (Int J Clin Pharm 2024) found that across all reports for semaglutide, liraglutide, and tirzepatide combined, psychiatric adverse events accounted for 1.2% of total reports — a small fraction of the overall safety signal, dominated by GI events.

SELECT trial — psychiatric AE rates

SELECT (PMID 37952131) is the largest semaglutide trial published to date — 17,604 adults with overweight/obesity and established cardiovascular disease (no diabetes), randomized to semaglutide 2.4 mg vs placebo, followed for a mean of 39.8 months. Across the trial's extensive safety reporting, serious psychiatric adverse events were not more frequent in the semaglutide arm than the placebo arm. SELECT alone contributes nearly 18,000 patients to the FDA's 107,910- patient pooled denominator.

What current FDA labels say after the February 2026 removal

We checked all six labels on DailyMed on 2026-05-09. The current state:

Practical implication for patients: if a 2024-2025 article you find online quotes the prior “Patients should be advised to monitor for the emergence or worsening of depression…” language from the Wegovy or Saxenda label, that text is no longer in the current label. If your prescriber or pharmacist references the older language, ask them to pull the current DailyMed label.

What about “loss of food joy,” anhedonia, and emotional blunting?

Patient reports — particularly on Reddit — frequently describe a phenomenon variously labeled “food noise gone,” “loss of food joy,” “flatness,” or “emotional blunting” on semaglutide and tirzepatide. The American Psychological Association covered this directly in its July 2025 Monitor on Psychology piece “A new era of weight loss: Mental health effects of GLP-1 drugs.”

Here is what the published evidence supports and does not support:

• Reduced appetite and food preoccupation are core intended pharmacologic effects of GLP-1 RAs. The mechanism — slowed gastric emptying plus central GLP-1 receptor signaling that suppresses hedonic eating — is the very effect that produces weight loss. Patients who previously derived strong pleasure from food may experience a real felt-sense reduction in that reward when food preoccupation diminishes.
• GLP-1 receptors are anatomically present in mesolimbic / mesocortical reward circuitry — including the ventral tegmental area, nucleus accumbens, and prefrontal cortex. This provides a biologically plausible mechanism by which a GLP-1 RA could modulate reward signaling beyond food. Plausibility is not causation.
• No primary-source RCT or large prospective cohort has measured anhedonia (Snaith-Hamilton Pleasure Scale or equivalent) as a pre-specified endpoint in adult GLP-1 weight-management patients. The Wadden STEP post-hoc captured PHQ-9 (depression) and C-SSRS (suicidality) — not anhedonia per se.
• Net mood signals from RCTs are neutral-to-favorable (Wadden 2024 PHQ-9 improvement; Wang 2024 lower SI risk). That coexists with individual patients reporting subjective flatness — reasonable explanations include smaller-than-detectable subgroup effects, dose-dependent effects not separated in pooled analyses, regression to the mean masking individual signals, and the difference between item-level scale scoring and lived experience.

Bottom line: “GLP-1s cause anhedonia” is not supported by primary-source evidence as of May 2026. It is also not contradicted as a per-individual lived experience. The honest framing is: some patients report it; published prospective measurement of it does not yet exist; large-cohort net mood signals are neutral or positive; if you experience it, document it for your prescriber and consider whether dose, dose-spacing, or switching agents is warranted. See the maintenance and microdosing decision hub for the four dose-strategy paths and the taper guide if discontinuation is on the table.

Eating disorders, body image, and GLP-1 RAs

Both NEDA (the National Eating Disorders Association) and ANAD (the National Association of Anorexia Nervosa and Associated Disorders) have issued public positions on GLP-1 RAs in 2024-2025. NEDA's position is verbatim:

“There has been very little research done on the impact of taking GLP-1's in people with eating disorders. So the short answer to the question is…we don't know yet.”

NEDA further notes that GLP-1 RAs “can be harmful when not used for their intended purpose, when inadequately monitored or monitored by clinicians without eating disorder expertise, or when used for weight loss motivated by weight stigma or fat phobia in people with eating disorders.” ANAD's parallel statement: “GLP-1s for binge eating disorder or other eating disorders are still not well understood. There is no formal FDA approval for the use of GLP-1s in the treatment of eating disorders.”

The Academy for Eating Disorders flagged the topic at its 2024 International Conference on Eating Disorders plenary, noting verbatim: “There is a lack of evidence to inform whether GLP-1 RAs could be administered in populations with eating disorders.” Bartel et al. (Int J Eat Disord 2024) wrote: “There has been increasing concern about the potential for GLP-1As to impact eating disorder (ED) symptomatology.” For the broader psychosocial picture — disclosure pressure, body-image disruption after rapid weight loss, “ghost fat,” and post-loss grief — see our companion guide on body dysmorphia, disclosure, and post-loss grief on GLP-1.

What to do if you experience mood changes on a GLP-1

1. Track it concretely. Date of onset, what changed (sleep, energy, food motivation, social engagement, interest in things you used to enjoy), severity (1-10 scale), duration. A two-week PHQ-9 self-screen (free at phqscreeners.com) gives a structured number you can hand to a clinician.
2. Tell your prescriber, not Reddit, first. Mood symptoms while on a GLP-1 RA are a clinical conversation. Even with the FDA suicidality warning removed, prescribers should still be willing to assess, adjust dose, or switch agents based on lived experience — especially for patients with prior depression, anxiety, or eating-disorder history (excluded from the published trials).
3. Do not stop abruptly without medical input. Abruptly discontinuing a GLP-1 RA after weeks-to-months of appetite suppression can produce a fast return of food preoccupation, hunger surges, and weight regain — all of which can themselves destabilize mood. The taper decision belongs with your prescriber.
4. If you are in crisis, call or text 988. The 988 Suicide & Crisis Lifeline is free, confidential, and 24/7 in the United States. You do not need a clinician referral to reach 988.
5. For eating-disorder-specific support, call the ANAD helpline at (888) 375-7767 (Mon-Fri 9 am-9 pm CST). The NEDA Helpline (1-800-931-2237) was permanently disbanded in June 2023; do not rely on it. NEDA's web screening tool at nationaleatingdisorders.org remains live.

Who needs extra caution

• Patients with current major depressive disorder, bipolar disorder, prior suicide attempt, or active eating disorder. The STEP and SURMOUNT programs excluded these populations. Net safety signals from large observational cohorts (Wang, Ueda) do not separate them out. The honest answer is: prospective high-quality data in these populations does not yet exist; clinical judgment rules.
• Patients on antidepressants. See our dedicated review of antidepressants, weight, and GLP-1 RAs for the drug-interaction landscape — the SSRI/SNRI weight- gain literature, bupropion considerations, and the tricky case of MAOIs (Qsymia is contraindicated within 14 days of an MAOI; the GLP-1 labels are silent).
• Patients reporting subjective “loss of food joy” or anhedonia. Document, do not self-diagnose. Most patients return to a more typical reward baseline after acclimation; some don't. Dose reduction, dose-spacing, or switch-agent strategies are reasonable discussion points with your prescriber.

Bottom line

Across 91 placebo-controlled trials in 107,910 patients, two regulators (FDA Jan 2026, EMA April 2024), and four large peer-reviewed studies (Wadden, Wang, Ueda, McIntyre), the evidence does not support a causal association between GLP-1 RAs and suicidal ideation, suicidal behavior, depression, or anxiety. The Wegovy / Saxenda / Zepbound labels no longer carry the suicidality warning as of February 2026. Patient reports of anhedonia and emotional blunting are real lived experiences that have not yet been measured prospectively at population scale, and the biological plausibility (GLP-1 receptors in reward circuitry) is documented even though the population-level signal is neutral. Patients with prior or active mental health conditions were excluded from the published trials and deserve more cautious individual assessment. If you experience mood changes, document them, talk to your prescriber, and use 988 or ANAD if in crisis.

Related research

• GLP-1 side-effect questions answered (master Q&A hub)
• What the phase 3 trials actually showed for GLP-1 side effects
• Antidepressants, weight, and GLP-1 receptor agonists
• Body dysmorphia, disclosure, and post-loss grief on GLP-1
• Stress, cortisol, and the “food noise” effect
• Life after GLP-1: maintenance, microdosing, tapering
• GLP-1 medications pillar (every approved + investigational agent)
• FDA-approved weight-loss medications hub