Background: Alcohol-related liver disease is the leading indication for liver transplantation in the USA. After remaining stable for over three decades, the number of deaths due to alcohol-related liver disease has been increasing as a result of increased high-risk drinking. We aimed to project trends in alcohol-related cirrhosis and deaths in the USA up to 2040 and assess the effect of potential changes in alcohol consumption on those trends.

Methods: In this modelling study, we developed a multicohort state-transition (Markov) model of high-risk alcohol drinking patterns and alcohol-related liver disease in high-risk drinking populations born in 1900-2016 in the USA projected up to 2040. We used data from the National Epidemiologic Survey on Alcohol and Related Conditions, National Institute of Alcohol Abuse and Alcoholism, US National Death Index, National Vital Statistics System, and published studies. We modelled trends in alcohol-related liver disease under three projected scenarios: the status quo scenario, in which current trends continued; a moderate intervention scenario, in which trends in high-risk drinking reduced to 2001 levels under some hypothetical moderate intervention; and a strong intervention, in which trends in high-risk drinking decreased by 3·5% per year under some hypothetical strong intervention. The primary outcome was to project deaths associated with alcohol-related liver disease from 2019 to 2040 for each pattern of alcohol consumption under the different scenarios.

Findings: Our model closely reproduced the observed trends in deaths due to alcohol-related liver disease from 2005 to 2018. Under the status quo scenario, age-standardised deaths due to alcohol-related liver disease are expected to increase from 8·23 (95% uncertainty interval [UI] 7·92-9·29) per 100 000 person-years in 2019 to 15·20 (13·93-16·19) per 100 000 person-years in 2040, and from 2019 to 2040, 1 003 400 (95% CI 896 800-1 036 200) people are projected to die from alcohol-related liver disease, resulting in 1 128 400 (1 113 200-1 308 400) DALYs by 2040. Under the moderate intervention scenario, age-standardised deaths due to alcohol-related liver disease would increase to 14·49 (95% UI 12·55-14·57) per 100 000 person-years by 2040, with 968 100 (95% UI 845 600-975 900) individuals projected to die between 2019 and 2040-35 300 fewer deaths than under the status quo scenario (a 3·5% decrease). Whereas, under the strong intervention scenario, age-standardised deaths due to alcohol-related liver disease would peak at 8·65 (95% UI 8·12-9·51) per 100 000 person-years in 2024 and decrease to 7·60 (6·96-8·10) per 100 000 person-years in 2040, with 704 300 (95% CI 632 700-731 500) individuals projected to die from alcohol-related liver disease in the USA between 2019 and 2040-299 100 fewer deaths than under the status quo scenario (a 29·8% decrease).

Interpretation: Without substantial changes in drinking culture or interventions to address high-risk drinking, the disease burden and deaths due to alcohol-related liver disease will worsen in the USA. Additional interventions are urgently needed to reduce mortality and morbidity associated with alcohol-related liver disease.

Funding: American Cancer Society and the Robert Wood Johnson Health Policy Research Fellowship.