GLP-1 drugs and suicidal thoughts: what FDA’s 2026 label update found

01

What changed in FDA’s January 2026 communication

FDA requested that application holders remove information about suicidal ideation and behavior from the Warnings and Precautions sections of Saxenda, Wegovy, and Zepbound labeling. The agency said the change would make messaging consistent across FDA-approved GLP-1 receptor agonists because diabetes-product labels in the class did not contain that warning.

The action followed a review that began after postmarketing reports prompted further investigation in 2023. In January 2024, FDA had said its preliminary evaluation had not found evidence that use caused suicidal thoughts or actions, while also explaining that small numbers of events limited the ability to rule out a small risk. The 2026 communication reports the result of the expanded analysis.

A request to change labeling is more specific than a general news headline. It does not erase every historical label immediately, determine what a patient experienced, or change unrelated warnings, contraindications, adverse reactions, or monitoring considerations in each product’s current prescribing information.

02

What evidence FDA reviewed

FDA’s meta-analysis included 91 placebo-controlled GLP-1 receptor agonist trials and 107,910 participants: 60,338 assigned to a GLP-1 medicine and 47,572 assigned to placebo. FDA reported that the analysis did not show increased risk for suicidal ideation or behavior or for other examined psychiatric events such as anxiety, depression, irritability, or psychosis.

The agency also used its Sentinel System for a retrospective cohort study of 2,243,138 new users with type 2 diabetes. It compared GLP-1 receptor agonists with SGLT2 inhibitors and reported no increased intentional self-harm risk after adjustment for measured baseline confounders. A subgroup with both diabetes and obesity had a similar result.

FDA additionally reviewed published observational and pooled studies and concluded that the totality did not support a causal relationship. These complementary methods matter because randomized trials and observational data have different strengths and limitations.

03

What the finding does not prove

No increased risk in the reviewed populations is not the same as proof that no person can experience a mood change while using a medicine. Trials may exclude some patients, definitions and reporting can vary, and observational comparisons can retain unmeasured confounding. FDA’s conclusion addresses whether available evidence supports a class-level causal risk, not whether a specific symptom in a specific person has a specific cause.

The update also should not be generalized to unapproved or compounded products with uncertain identity, concentration, ingredients, or instructions. FDA’s review concerns the GLP-1 receptor agonist class and identified approved products; it does not establish the safety or quality of every product marketed with a GLP-1 name.

Finally, removing this warning does not turn a product into a recommendation. Approved indications, contraindications, dosing, adverse reactions, interactions, and patient-specific considerations remain in the current label and belong in a clinician’s assessment.

04

How to check the current label rather than an old claim

Search the exact brand and application in Drugs@FDA, then open the most recent label date. Review the Warnings and Precautions section, Adverse Reactions, Patient Counseling Information, and any Medication Guide. Compare that date with the webpage, handout, or provider FAQ making the claim. An older PDF can remain indexed online after newer labeling is approved.

Record product name and indication because labels in the same pharmacologic class are not necessarily identical. The January communication named three weight-management products. Do not rewrite the action as a removal from every medicine, every country, or every type of patient information unless the current product-specific record supports that statement.

Provider researchers should update language in both directions: remove outdated assertions that FDA found a causal class risk, but do not replace them with claims that the medicines improve mental health or can never be associated with a reported symptom.

• →Exact product and active ingredient
• →Most recent label date
• →Warnings and Precautions
• →Medication Guide or Patient Information
• →FDA safety communication date
• →Provider page update date

05

What patients and caregivers should do with symptoms

FDA advises patients to continue medication as prescribed and discuss concerns with health professionals. It also says to tell a health professional about new or worsening depression, suicidal thoughts, or unusual changes in mood or behavior. A person should not delay seeking help while trying to decide whether a medicine caused a symptom.

In the United States, FDA’s communication directs people in distress to call or text 988 or use the 988 Lifeline website. Immediate danger calls for emergency services. A provider directory, article, portal message, or routine appointment request is not crisis care.

Medication changes can also have consequences, so individual decisions should be made with an appropriately licensed clinician who knows the prescription and health history. This article cannot determine whether to start, stop, or change treatment.

06

Limits and update triggers

This article reports FDA’s conclusion and the evidence summary in its January 2026 communication. It does not independently reanalyze patient-level data, estimate an individual risk, or diagnose a psychiatric condition. The label for each product should be rechecked because approval supplements and safety information can change after publication.

The article should be updated if FDA publishes a materially different safety communication, approves revised product labeling that changes the relevant language, or releases new evidence that changes its conclusion.