In a Tokyo liver clinic, the doctor treating your cirrhosis also ran the conversation about your drinking. For most patients, no psychiatrist, no separate referral. Over six months, the ones in the middle risk band cut their intake by almost half.
That's the setup I found interesting here. Alcohol is the leading cause of cirrhosis and liver failure in Japan, and psychiatric slots for alcohol problems are scarce. So this team asked a blunt question: can the hepatologist just handle it?
What they did
Single-center, prospective study, running from August 2022 to September 2025. Researchers screened 932 patients with chronic liver disease using AUDIT (Alcohol Use Disorder Identification Test), a 10-question screen scored from 0 to 40. Higher score, riskier drinking.
They sorted patients into three lanes by score:
- Under 10. Education only. 823 people landed here.
- 10 to 19. The hepatologist-led group. 78 people. This is the one the study is really about.
- 20 and up. Referred to a psychiatrist. 31 people.
The middle group got the full package from their liver doctor: drinking guidance, psychosocial support, and nalmefene (a drug that dampens the urge to drink) when it was indicated. Over 24 weeks the team tracked total alcohol, heavy-drinking days, drinking risk level, and liver function.
Worth saying up front: this is observational, with no control group. So there's no clean way to separate the intervention from what these patients might have done anyway once they were sitting in a liver clinic.
What they found
Of the 78 in the middle group, 74 finished the follow-up.
- Median intake dropped from 63.1 to 35.1 grams a day (p=0.001). That's roughly a 44% cut.
- Heavy-drinking days fell from 17.0 to 11.3 per month (p=0.002).
- 52 patients, 70.2% of the group, moved down into a lower drinking-risk category.
The people who downstaged weren't just drinking less on paper. Their AST and their γ-GTP, both liver enzymes that climb with alcohol damage, came down (p=0.027 and p=0.014). They also had fewer cirrhosis-related complications than the patients who didn't downstage.
Then the odd one. When researchers ran a regression to find what predicted downstaging, the strongest signal was having liver cancer at baseline (odds ratio 5.23). The patients with the most frightening diagnosis were the most likely to cut back. The confidence interval is wide (1.10 to 15.3) and the group is small, so I wouldn't lean on the exact number. But the direction is worth sitting with.
What it means
The headline for me isn't the 44%. It's who delivered it. A hepatologist, in a normal clinic visit, running the drinking conversation that usually gets punted to a specialist who may not exist in your area. AUDIT did the triage: one short questionnaire splits a waiting room into education, hands-on help, and genuine psychiatric referral, so the scarce psychiatric slots go to the 31 people who most need them.
The cancer finding points somewhere uncomfortable. A cirrhosis diagnosis is abstract. Liver cancer is not. It reads like the moment the risk stopped being a number on a chart and became personal, which is exactly when people move. That fits a broader pattern in the liver literature, where cutting back can pull disease into a lower-risk category even after years of heavy use. It shouldn't take a tumor to get there, but for some of these patients it did.
Keep the limits in view. Drinking was self-reported, and people under-report, especially to the doctor treating their liver. There was no control group and only 24 weeks of data, so we don't know if the cut holds. The role of nalmefene is muddy too, since it went to some patients and not others with no clean breakdown of who or how much it mattered.
Still, the practical shape is clean. In a country where alcohol wrecks more livers than anything else, and where East Asian genetics push liver risk up at lower thresholds than Western guidelines assume, moving the drinking conversation into the room where the liver is already being treated looks like it does real work.
Source: Internal Medicine (Tokyo), DOI
