First-ever randomized trial of Ozempic for alcohol use disorder shows significant reductions in drinking
Dr. Christian Hendershot from UNC announced preliminary study results of a 48-person randomized trial that compared semaglutide (Ozempic / Wegovy) to placebo for reduction of drinking in non-treatment seeking participants.
The initial findings, presented today at the annual CPDD Conference in Montreal, indicate greater reductions in drinking for semaglutide vs. placebo groups for several drinking outcomes. For some drinking outcomes, reductions in consumption for the semaglutide group were significantly larger than the placebo group. This is a big result and is the first completed RCT examining effects of semaglutide on alcohol outcomes . Early estimates of the size of the effects suggest substantially stronger effects than are typically seen for currently approved medications for alcohol use disorder.
Only 2% of people with alcohol use disorder (AUD) currently receive medication, because existing AUD medications have moderate efficacy and low patient interest. GLP-1 receptor agonists are a potential revolution in treatment, and may finally allow addiction medication to scale to a significant portion of people in need.
These findings are in line with previous research on GLP-1s and alcohol use that show similarly dramatic reductions. They are also consistent with the flood of reports from patients who say their urge to drink shrinks or disappears when taking GLP-1s.
Key findings of the study
Drinks per day and heavy drinking days were reduced for patients receiving semaglutide versus placebo. Percent reductions in weekly and monthly drinking were greater in magnitude during treatment at a higher dose compared to a lower dose.
In typical use, semaglutide dosing begins at .25 mg and gradually increases to 1mg or 2mg over a series of months. This study was only able to follow patients long enough to complete one month of treatment at the .5mg dose level, which showed increased efficacy vs .25 mg. It seems likely that continuing to increase the dose, as would happen in real world usage, would show increased efficacy, as it does for weight loss. At the same time, there were results seen at the lowest dose which is another good sign for tolerability.
Crucially, the study looked at “non-treatment-seeking participants” – this means people who drink at a medically unhealthy level, but who are not trying to reduce their drinking. This is remarkable because it means that people cut their drinking significantly despite having no intention to do so. It seems very possible that motivated patients would have even higher success.
96% of patients who received semaglutide stayed in the trial, higher than the placebo group. It would have been 100%, but one patient was removed because of protocol deviation. This retention rate is far higher than studies typically find for naltrexone adherence. There was no evidence of serious adverse events. Together, the findings suggest that semaglutide may be effective at reducing heavy drinking and more likely to have patients stick with the treatment. That combination creates the possibility for an exponential performance improvement.
The findings, while preliminary and not yet published, suggest that the heaviest drinkers saw the largest effects. Larger benefits for drinkers who are most at risk would disproportionately reduce alcohol-related health harms at a population level.
Final results, including percent reductions, will be published once the data is peer-reviewed. Dr. Hendershot and his collaborators will be working together on larger followup studies which may bring GLP-1s for alcohol use disorder even closer to broad adoption. While many prescribers are already using GLP-1s for patients with AUD, most are waiting for additional data as well as an FDA-indication that would enable insurance coverage.
At the same session, Dr. Grigson and Dr. Bunce showed results from their study on GLP-1 receptor agonists for reducing opioid cravings, and Dr. Yammine presented her existing research on GLP-1 medications for smoking cessation as well as plans for two larger studies on GLP-1 and smoking.
The Strategic Importance of GLP-1s for Addiction
Together with existing evidence, this study points towards the incredible promise of GLP-1 medications to reduce addiction at a population scale.
Our focus at CASPR is always on scale and strategy: which policies or medications give us the best chance of achieving a society-wide reduction of addiction and its cascade of downstream harms? If we are taking seriously the 767,000 annual deaths from addiction in the US, including the surge in overdoses from fentanyl and methamphetamine, we must be looking for big, game changing opportunities. There have been incremental improvements in existing treatment tools over the past 20 years but we’re still not anywhere close to solving the opioid crisis or reducing alcohol, cocaine, or methamphetamine use.
I believe that GLP-1s are the key breakthrough technology that the field of addiction has been looking for and here’s our long-form argument for the scientific and strategic importance.
GLP-1s don’t just exceed the consumption reduction performance of existing medications, they are far superior across a number of strategic dimensions. This will enable them to reach many times more people than any previous addiction medication.
Unlike existing medications, GLP-1s will be easy to get:
GLP-1s are non-opioid based and they are not primarily medications for addiction. They sidestep all stigma and regulation of medications for SUD.
Most physicians already prescribe GLP-1s. Once recommended for SUD, they will be available to every patient in any type of medical evaluation.
And unlike existing medications, patients will want to take them:
GLP-1 provide holistic health benefits for patients by reducing mortality, reducing anxiety and depression, reducing suicidal ideation, reducing weight, improving metabolic markers, reducing brain inflammation, reducing certain cancers, and more. The decision to take GLP-1s will be a no-brainer for so many patients who would never be willing to take medication specifically for addiction.
They are easy to get on and off– there’s no process for initiation and there are no withdrawal side effects when you stop. If you know you can get off a drug, you are much more likely to try it.
No one knows exactly how many patients will take GLP-1s for substance use disorders once approved. But we have good reason to believe it could be 10X or more vs existing medications for SUD. Since 6% of the population takes GLP-1s for obesity or diabetes, we know that more people with AUD are already incidentally taking a GLP-1 than are taking a medication indicated for AUD (2%).
What’s Next: the Path Towards an FDA Indication
When you’re casually following a field, in this case addiction science, it can be hard to tell which reports of new studies are really meaningful and which are random blips or hyped up. Even more tricky is to get a sense of whether a scientific team is working in a lonely niche or if they are part of a broader scientific project or movement. In my mind, what’s most significant about these Hendershot findings is that they are part of an evidence base being built by multiple labs across multiple countries.
With this study, the ground has been laid for larger studies that could bring GLP-1s for alcohol misuse into the FDA approval process, ultimately leading to an indication for alcohol use disorder (AUD). Currently, some leading-edge prescribers are already using GLP-1s to help their patients reduce drinking and other dangerous substance use– but they are a minority. Once the FDA approves an indication, it will become the new standard of care for clinics and physicians everywhere. As importantly, insurance and other payers will not cover GLP-1s for addiction until there is an FDA seal of approval. When that happens, GLP-1 use for AUD will grow very rapidly and we will be able to see population level impacts on drinking and alcohol related health harms. 178,000 Americans die every year from alcohol-related causes. GLP-1s may be able to reduce that number dramatically over the next 5-10 years.