Early Return to Drinking After Alcohol-Associated Hepatitis Remains Common Despite Recovery

New research from Cleveland Clinic has identified risk factors for recurrent alcohol use among patients recovering from alcohol-associated hepatitis (AH). The findings also indicate that patients with alcohol-associated hepatitis, regardless of severity, may benefit from earlier alcohol use disorder (AUD) interventions.

The research, which was published in Alcohol, Clinical and Experimental Research, centers around two concurrent studies: a randomized controlled trial (RCT) and an observational (OBS) cohort in patients with AH. Through both studies, the research group looked at alcohol use among AH survivors and risk factors for recurrent alcohol use.

“Recurrent alcohol use is a major determinant of liver-related outcomes in patients recovering from AH,” explains Srinivasan Dasarathy, MD, senior author on the paper and a transplant hepatologist at Cleveland Clinic. “While return to drinking (RTD) following liver transplantation has been studied quite a bit, there hasn’t been as much research on the clinical characteristics of patients who remain abstinent after hospitalization for an episode of AH. Patients with alcohol-associated hepatitis face two major challenges: “During the first few weeks, survival is driven primarily by the severity of liver disease. After recovery from the acute illness, however, long-term outcomes are strongly influenced by whether patients return to drinking. That is why identifying patients at high risk for early return to drinking is so important.”

Study design

The research used data from two prospective clinical studies, a randomized Phase 2b double-blind controlled trial (RCT) (ClinicalTrials.gov NCT04072822) and a prospective observational cohort study (OBS) (ClinicalTrials.gov NCT03850899), both conducted by the Alcoholic Hepatitis Network (AlcHepNet). The trials included patients who were 21 and older and who were clinically diagnosed with AH.

Enrolled RCT participants were assessed during follow-up visits at three, seven, 14, 28, 60, 90 and 180 days, and enrolled OBS participants at four, 12 and 24 weeks. Clinical evaluation included in-person or telehealth interviews using Timeline Follow-back (TLFB) to assess alcohol consumption since the last study visit.

Severe AH (sAH) was defined by a MELD score ≥20, while moderate AH was defined by a MELD score of 11–19. Only patients with sAH and MELD scores 20–35 were enrolled in the RCT, while the OBS cohort included participants with MELD scores of ≥11.

Study findings

The study included 518 patients who were prospectively followed at visit Day 28 or after.

Compared with patients who did not return to drinking, patients who did relapse had lower MELD scores (23.4 vs. 25.2; P = .01) and were less likely to have ascites (61.4% vs. 75.2%; P < .01). They also consumed more drinks per month (233.8 vs. 130.3; P < .01), drank on more days per month (20.9 vs. 13.9; P < .01), had higher baseline AUDIT scores (25.3 vs. 22.7; P <.01) and were more likely to report a natural parent with AUD (69.5% vs. 54.0%; P =.01).

The typical number of standard daily drinks was 12.1 ± 10.3 among patients who returned to drinking versus 9.3 ± 8.1 among those who did not (P < .01).

“We also found that patients who returned to drinking had lower BMI and waist circumference compared with those who did not, but these findings did not reach statistical significance,” adds Dr. Dasarathy.

In univariate regression, higher educational status, higher MELD scores, and the presence of ascites were associated with significantly reduced RTD hazard. In contrast, higher AUDIT scores, a higher number of daily drinks and more drinking days were associated with significantly increased hazards of RTD (all P < .05).

From patients' responses in the TLFB assessment about drinking behaviors, the researchers categorized the weekly average number of drinks as 0 drinks per week, 1–6 drinks per week, and seven or more drinks per week.

“We found that although drinking amounts decreased significantly from baseline, they gradually increased over time,” explains Dr. Dasarathy.

By 180 days, only 76% of the cohort remained completely abstinent, while 12% started moderate drinking (1–6 drinks per week), and 12% returned to heavier drinking (7 or more drinks per week).

Patients with moderate AH (mAH, MELD 11–19, n= 103) had a higher RTD incidence at 180 days than those with severe AH (sAH, MELD ≥20, n= 415) (44.3% vs. 27.5%; P = .01). This may be because patients with moderate disease recover more quickly and may feel less urgency to remain abstinent.

Takeaways

“Our findings reinforce the idea that we need to do more to provide early AUD interventions for patients with AH,” says Dr. Dasarathy. “In our study, we were able to identify risk factors for RTD, which may help clinicians focus additional support on patients who need it most.”

He continues, “For example, the number of drinking days prior to hospitalization is often overlooked as a risk factor for RTD, but in our cohort, it appeared to be stronger than other alcohol parameters for predicting RTD. The strongest predictor of early return to drinking was frequent alcohol use before hospitalization. Patients who had been drinking on more than 20 of the previous 30 days had more than a threefold higher risk of returning to alcohol.”

Education level also emerged as an important factor, notes Dr. Dasarathy. Patients with a college education or higher had about half the risk of returning to drinking compared with those with less education. However, education may reflect broader differences in income, health literacy, and access to alcohol use disorder treatment rather than education itself.

“This finding suggests that we may need to tailor interventions to the characteristics of our patients,” he explains. “But there’s an important caveat here — education level is also a surrogate for income and potentially for access to AUD treatments, which we did not measure in this study. So, while every patient with AH merits AUD evaluation and treatment, these findings may help identify patient subgroups who could benefit from improved access to AUD care and/or more intensive treatment.”

Similarly, the authors note that although medications for AUD (MAUD) were associated with higher RTD rates in this study, the finding may reflect greater AUD severity among patients receiving MAUD or treatment initiated in response to RTD, rather than a causal effect on relapse risk. Additionally, prior studies have indicated that MAUD may be effective in reducing RTD events. These findings support integrating addiction treatment into routine hepatology care. Patients recovering from alcohol-associated hepatitis should receive early evaluation and treatment for alcohol use disorder, including counseling, consideration of medications when appropriate, and close follow-up during the first three months after discharge.

“If MAUD were consistently implemented, the proportion of RTD events that might be preventable could be substantial,” says Dr. Dasarathy. “A 2025 study on treating AUD in alcohol-associated liver disease estimated that RTD could be significantly reduced with effective MAUD treatment among eligible patients.”

Identifying which patients with AH are most likely to benefit from intervention — and which are at greatest risk for RTD — is critical to delivering targeted, effective care. The authors believe this study is an important step toward that goal.

“Taken together with prior research, our findings suggest that post-AH relapse risk is highly modifiable,” says Dr. Dasarathy. “The fact that RTD occurred in more than one in five patients with AH is concerning, and it underscores the need for more effective relapse-prevention strategies. A better understanding of which patients are at highest risk can help clinicians tailor treatment and provide the support patients need to succeed.”