The Lancet published the SEMALCO trial on May 2 and the NIH summarized it on May 12. [1][2] The study randomized 108 adults at the Mental Health Center Copenhagen to once-weekly subcutaneous semaglutide at 2.4 mg or placebo for 26 weeks; both arms received up to ten 45-minute cognitive-behavioral-therapy sessions with a trained nurse. The primary endpoint was the change in heavy drinking days from baseline to week 26. [3]

The eligibility line is the most under-reported piece. The trial enrolled only adults aged 18 to 70 with a diagnostic-interview-confirmed alcohol use disorder, a body mass index of 30 kg/m² or higher, a minimum of six heavy drinking days in the past 30, and an AUDIT score above 15. [3] The findings apply, by protocol, only to people with AUD and comorbid obesity. The paper's May 20 brief on GLP-1 and alcohol staying inside the Lancet trial named the eligibility frame as the gap between the headline and the prescription pad. Wednesday's edition adds the dose-response specificity and the NIH's institutional read.

The numbers, in the trial's units. At baseline, participants reported a mean of approximately 17 heavy drinking days in the previous 30. At week 26, the semaglutide arm reported a change of −41.1 percentage points versus −26.4 in placebo — a between-group difference of −13.7 points (p=0.0015). [3][4] Total alcohol consumption fell by −1,550.2 g per 30 days versus −1,025.9 g in placebo. Body weight fell 11.2 kg versus 2.2 kg. The Lancet's secondary biomarker — plasma phosphatidyl ethanol, the gold-standard alcohol biomarker — confirmed the self-reported reductions. [5]

The NIH summary, posted under the National Institute on Alcohol Abuse and Alcoholism's "Research Matters" series May 12, is the institutional read. [2] NIAAA Director George Koob, MD, a study co-author: "Very few medications are currently approved for alcohol use disorder. A new option that is more accessible and more effective could be a gamechanger for closing the treatment gap." [4] NIDA Director Nora Volkow, MD, added the design caveat: "We're beginning to see some of that potential for GLP-1s to treat drug addiction turn into reality. This is nonetheless very encouraging." [4]

Three FDA-approved drugs treat alcohol use disorder: naltrexone, acamprosate, and disulfiram. About 2% of U.S. adults with AUD receive any of them. [6] The Number Needed to Treat for the SEMALCO endpoint was 4.3 — lower (better) than any approved AUD pharmacotherapy. [4] SEMALCO was a treatment-seeking cohort with obesity, and the field-deployed comparator drugs are not; the comparison is therefore directional, not strict. The trial result is one of the strongest single-trial AUD signals on record and remains inside its enrollment criteria.

Mechanistically, semaglutide is a GLP-1 receptor agonist approved for chronic weight management and type 2 diabetes. The trial's authors point to GLP-1 receptors expressed in mesolimbic dopamine pathways — ventral tegmental area, nucleus accumbens — that overlap with alcohol's reinforcing circuits. [7] The trial cannot distinguish that mechanism from the weight-loss-mediated effects on alcohol pharmacokinetics. The authors flag this directly. [5]

Adverse events tracked the GLP-1 class. Gastrointestinal symptoms — nausea (57% versus 7%), loss of appetite, vomiting, abdominal pain — were common in the semaglutide arm, mostly transient. Five participants discontinued for adverse effects; one serious adverse event in the semaglutide arm. [3] No suicidality signal appeared. [7]

The trial collected no alcohol data after week 26, so durability is unknown. The population was predominantly white. The trial cannot answer whether semaglutide works in AUD patients without obesity. The U.S. VA has registered the CRAVE trial (NCT07218354), a Phase 3 study in veterans with moderate-to-severe AUD; CRAVE will not enroll on obesity criteria. [8] That study is the closest registered confirmation.

The Wednesday read is that the trial is real and the eligibility line is real. Novo Nordisk does not yet have an AUD indication in front of the FDA; the SEMALCO authors do not endorse off-label use. [5] The next clinical step is replication in non-obese AUD patients. The next reporting step is to keep the headline inside the protocol.

-- KENJI NAKAMURA, Tokyo