40% Fewer Opioid Cravings: Inside the Penn State Liraglutide Trial

In February 2024, neuroscientist Patricia "Sue" Grigson walked onto a stage at the American Association for the Advancement of Science meeting in Denver and presented data that would have sounded absurd five years earlier: an obesity drug had significantly reduced opioid cravings in people with opioid use disorder.

The study was small — just 20 participants. But it was the first randomized controlled trial to test any GLP-1 receptor agonist against opioid addiction in humans. And the results, while preliminary, were striking enough to shift the conversation from "interesting hypothesis" to "this deserves serious investigation."

Study Design

The trial (NCT04199728) was conducted at the Caron Treatment Center in Wernersville, Pennsylvania — a residential rehabilitation facility. All 20 participants had recently completed opioid withdrawal and were in early recovery. Some were also taking buprenorphine, an approved opioid use disorder medication.

Participants were randomized to receive either daily liraglutide (brand names: Saxenda for obesity, Victoza for diabetes) or placebo via subcutaneous injection. The dose escalated over the 3-week trial period. Cravings were measured using ecological momentary assessment — smartphone notifications four times daily asking participants to rate their opioid craving in real time.

This design captured craving as it actually happens, not just what patients remember reporting at weekly clinic visits. It's a more granular and ecologically valid measure than traditional retrospective questionnaires.

What the Data Showed

Relative to placebo, patients receiving liraglutide showed a 40% reduction in craving that began in the first week of treatment. The effect was consistent across the study period.

The most dramatic finding was in the subgroup taking both liraglutide and buprenorphine. By day 10, these patients were significantly more likely to report zero opioid cravings than patients on buprenorphine plus placebo. For people in early recovery — when craving is most intense and relapse risk is highest — any period free from craving is clinically meaningful.

The Dropout Problem

The trial's biggest limitation was attrition. Of the 20 enrolled participants, 11 dropped out before the three-week trial ended. The primary reason: gastrointestinal side effects from liraglutide, which were twice as common in the active drug group as in placebo. Nausea, vomiting, and abdominal discomfort drove most early exits.

However, a crucial detail mitigated this concern. Patients who received both liraglutide and buprenorphine had lower GI distress and lower dropout rates than those on liraglutide alone. This suggests the combination approach may be more tolerable — and it's the combination that showed the most impressive craving reduction.

Furthermore, the craving reduction was significant even at the lowest liraglutide dose. This suggests that a low-dose strategy — staying below the threshold that triggers severe GI symptoms — might capture most of the craving benefit without the tolerability cost. This is a testable hypothesis for future trials.

From Rats to Humans

Grigson's work represents one of the cleanest translational pipelines in the GLP-1/addiction field. The progression was methodical: first, a 2022 preclinical study in rats demonstrating that liraglutide reduced heroin-seeking behavior and altered brain activity in reward regions. Then, a human pilot RCT using the same molecule and measuring the analogous outcome (craving as a proxy for drug-seeking).

The rat data showed that liraglutide not only reduced heroin self-administration but also decreased relapse-like behavior after a period of abstinence — the preclinical equivalent of early recovery. The human trial essentially asked: does this translate? And the answer, with appropriate caveats about sample size, was yes.

Context: The Opioid Crisis in Numbers

Opioid overdose kills approximately 80,000 Americans per year. While medications like buprenorphine and methadone are effective, they are stigmatized and dramatically underutilized — fewer than 20% of people with opioid use disorder receive any medication-assisted treatment. Adding a well-tolerated GLP-1 medication to the toolkit wouldn't replace existing treatments, but it could reach patients who refuse or can't access traditional options.

The Wang/Volkow JAMA Network Open study (covered in our opioid deep dive) provided the population-level signal: 78% lower overdose risk in semaglutide patients. The Grigson trial provides the mechanistic evidence: the craving pathway appears to be real. Together, they make a compelling case for larger trials.

What's Next

Grigson and her team have stated that if liraglutide shows a trend toward efficacy and safety in this population, the next step would be a Phase 2 multi-center clinical trial with larger enrollment, longer duration, and formal dose-finding. The combination of liraglutide with buprenorphine — which showed both better tolerability and better craving outcomes — is the most promising configuration to test.

Meanwhile, the NIAAA (National Institute on Alcohol Abuse and Alcoholism) is running separate semaglutide trials for alcohol use disorder, and if those show positive results, the case for expanding to opioid-specific trials with more potent GLP-1 agonists will strengthen considerably.