Direct answer
The incretin class has legitimate, early human evidence in alcohol use disorder and a broad preclinical footprint across addiction models — and nothing more than that yet. No incretin is an approved addiction or mood medicine. What exists is a randomized trial, real-world associations, meta-analytic mood data, and a registry of ongoing trials that will decide what the class actually is in psychiatry.19
The story
How a metabolic class reached psychiatry
It did not start in a trial. It started in clinic rooms and on forums: people on GLP-1-based medicines for diabetes or weight kept volunteering that they had stopped wanting alcohol — not white-knuckling it, just not wanting it. A signal like that is easy to dismiss and was not dismissed, because the receptor biology already predicted it might be real: GLP-1 receptors are expressed in the midbrain reward circuitry that addiction dysregulates.2
The sequence since has been textbook evidence-building. Real-world cohort work associated semaglutide exposure with reduced incidence and recurrence of alcohol use disorder.3 A randomized placebo-controlled trial of a weekly GLP-1 agonist in alcohol use disorder followed.4 Reviews from the field's own pharmacology bodies now treat substance use disorders as an emerging incretin indication.25 Each step was smaller than the headlines and more useful than them.
The map
Indication by indication, by evidence tier
Each row names the strongest evidence tier the class has reached in that indication — not the loudest claim made about it. A tier is a statement about what kind of knowing is available.
Alcohol use disorder — strongest tier: randomized human trial
Once-weekly semaglutide against placebo in adults with AUD (JAMA Psychiatry, 2025),4 with real-world cohort support3 and dual-agonist preclinical replication in rodents.6 The class's most developed reward indication.
Mood — strongest tier: meta-analysis plus large safety analyses
A systematic review and meta-analysis reports antidepressant effects across GLP-1 receptor agonist studies.7 Separately, large-scale safety analyses asked whether the class carries suicidality risk and found no elevated risk of suicide death.8 Both directions stand on this page at equal prominence — that is what an honest map looks like.
Nicotine, opioids, other compulsive behaviour — strongest tier: preclinical with registry activity
The preclinical literature across substance models is substantial and consistent in direction.2 Human trials across nicotine relapse, opioid use disorder and related indications are registered and running — the outcomes are not yet in.9
What the evidence does not establish
Long-horizon outcomes in any reward indication. Durability after discontinuation — whether craving returns when the molecule leaves. Head-to-head comparison of dual agonism against GLP-1-only agonism in human reward outcomes. Any approved psychiatric indication. A page that implies otherwise is ahead of the record; this one declines to be.
| Evidence tier | What it can support | Where the class stands |
|---|---|---|
| Randomized controlled trial | A causal claim in the studied population, on the studied agent | Reached in alcohol use disorder (GLP-1-only agent)4 |
| Real-world cohort | Association; hypothesis confirmation at scale | Reached in AUD incidence and recurrence3 |
| Meta-analysis | Pooled direction of effect across studies | Reached for antidepressant signal7 |
| Preclinical | Mechanism plausibility; dose and circuitry hypotheses | Reached across alcohol, nicotine, opioid models, incl. dual agonism6 |
| Registry activity | Evidence that the question is being asked properly | Active across the reward indications9 |
Why the interval matters here
Reward indications are adherence problems wearing a diagnosis
Addiction and mood disorders share an operational property that metabolic ones do not: the condition itself erodes the daily discipline a medicine depends on. A weekly injection asks fifty-two acts of follow-through a year from the population least able to guarantee them. That is not a criticism of the weekly agents — it is the argument for what comes after them.
A once-monthly dual agonist is the class's answer to that argument, and it is the axis on which brenapatide is built: the same dual-receptor mechanism, carried by a backbone and side chain engineered for a month of circulation.10 Whether monthly dual agonism delivers on the reward indications is precisely what the current trial generation exists to answer.9
← Back to the brenapatide monographQuestions
Frequently asked
Do GLP-1 agonists work for addiction?
The evidence is early but real. A randomized trial in alcohol use disorder,4 real-world cohort associations,3 and a substantial preclinical literature across alcohol, nicotine and opioid models2 establish that the question is legitimate. No incretin is an approved addiction medicine; the frontier is active, not settled.
What is the strongest human evidence in alcohol use disorder?
Do incretin therapies carry mood risks?
Does the effect last after stopping?
Not established. Durability after discontinuation is one of the explicit gaps on this page's map: the published record does not yet say whether craving returns when the molecule leaves. The ongoing long-duration trials are designed to answer exactly that.9
Why a monthly molecule for these indications?
Because craving and relapse operate on month timescales while the conditions erode daily adherence. A once-monthly dual agonist asks twelve acts of follow-through a year instead of fifty-two — the design rationale behind brenapatide's half-life engineering.10
Literature
References
Retrieved at the identifiers given, 5 September 2026. Where the record is thin, the map above says so rather than inflating the citation count.
