Brian Lee, MD, is a hepatologist and liver transplant specialist with Keck School of Medicine of USC.

Brian Lee, MD

Alcohol use among US adults declined between 2022 and 2024, marking the first confirmed reduction since the onset of the COVID-19 pandemic, according to a national cross-sectional study published in Annals of Internal Medicine.¹˒² Any alcohol use fell by nearly 2%, and heavy drinking fell by more than 8%, though both remain above 2018 levels.¹˒²

"It's encouraging to see alcohol use decline for the first time since the start of the pandemic," said Brian P. Lee, MD, MAS, 

 and liver transplant specialist with Keck Medicine of USC and principal investigator of the study in a statement. "However, use remains above pre-pandemic levels and certain groups continue to have concerning trends."²

Lee and colleagues analyzed National Health Interview Survey data from 114,352 US adults between 2018 and 2024, classifying heavy drinking by National Institute on Alcohol Abuse and Alcoholism criteria.² These criteria define heavy drinking as ≥5 drinks per day or ≥15 drinks per week for men and ≥4 drinks per day or ≥8 drinks per week for women.²

The declines were uneven across age groups. Gen Z adults showed the largest drop in any alcohol use, at more than 5%, while heavy drinking among millennials fell by 17% between 2022 and 2024.² Among adults aged 50 to 64, any alcohol use rose by nearly 4% and heavy drinking rose by more than 36% from 2018 to 2024.²

Men saw a larger decline in any alcohol use than women (2.4% vs 1.1%), while women saw a larger decline in heavy drinking (9% vs 7.7%).²

In the following interview, Lee discusses what clinicians can do to sustain the decline, why adults aged 50 to 64 remain a concern, and how much pandemic-era drinking explains the rise in alcohol-associated liver disease.

Q&A: Declining Alcohol Use and Heavy Drinking in Adults 50 to 64

: What can clinicians and patients do to help the recent decline in alcohol use continue?

Brian P. Lee, MD, MAS: The decline occurred largely independent of clear clinical intervention as far as we know. What clinicians can do is to stop relying on patients to bring alcohol use up in conversation. Alcohol use should be asked about at every visit, like smoking, with a validated screen like the AUDIT-C, and for patients where the history is uncertain there's now a biomarker, phosphatidylethanol, that gives an objective measure of intake over the prior 4 weeks. 

The thresholds for heavy drinking are lower than most patients and a fair number of physicians assume, so being knowledgeable about those thresholds (≥15 drinks per week or ≥5 drinks on any given day for men and ≥8 drinks per week or ≥4 drinks on any given day for women) is important. Physician messaging is part of this too. Many patients, and some physicians, still believe a glass of wine a day is good for the heart. That idea came from observational studies that compared drinkers with abstainers who had often quit because they were already sick, and when that bias is accounted for, the protective effect largely disappears. 

The current evidence is that cancer and liver risk can begin even at low alcohol levels and rise with dose, and the message coming from physicians should reflect that. A patient who hears "moderate drinking is fine" from their doctor has no reason to cut back. And when a patient does meet criteria for alcohol use disorder, we have effective medications that are prescribed far less often than they should be.

HCPLive: Heavy drinking among adults 50 to 64 rose by more than a third since 2018. What are the implications for liver disease in this group?

Brian P. Lee, MD, MAS: This is the group in which the combination of exposure and timing is most concerning. Fibrosis accumulates over decades, so a 55-year-old who increased their drinking in 2020 is often adding to damage that's already there rather than starting from a healthy liver. 

This is also the decade when cirrhosis and hepatocellular carcinoma most often declare themselves, and there's less physiologic reserve to absorb an insult. In practical terms, patients with heavy drinking, particularly those above the age of 50, should be considered for liver fibrosis screening.

HCPLive: Women reduced heavy drinking more than men, yet their alcohol-associated liver disease has been rising faster. What might explain that discrepancy?

Brian P. Lee, MD, MAS: The reduction from 2022 to 2024 is real, but it's a recent step down from a large step up. Heavy drinking among women is still about 13% above 2018, and I should mention that in this survey the prevalence of heavy drinking among women is now slightly higher than among men, because the thresholds are sex-specific and women's drinking has been converging with men's for years. 

Women develop fibrosis at lower cumulative exposure, they progress faster once it's established, and the increase in women's drinking predated the pandemic by at least a decade. So the liver disease we're seeing in women now reflects 15 or 20 years of rising intake. Two years of modest decline won't change that trajectory quickly.

We've also published on co-factors for the rise in alcohol-associated liver disease; if a co-factor rises in prevalence, alcohol-associated liver disease can increase even if alcohol consumption is stable. In a separate study that we published in Annals of Internal Medicine a couple years ago, we showed that the rise of metabolic syndrome among women may be contributing to the rise in alcohol-associated liver disease, independent of changes in alcohol consumption.

HCPLive: With drinking now falling, to what extent do you attribute the rise in alcohol-associated liver disease among young adults to pandemic-era drinking?

Brian P. Lee, MD, MAS: Only partly. Alcohol-associated hepatitis in young adults was rising well before 2020, and the pandemic accelerated a trend rather than creating one. If younger adults' drinking stays below 2018 levels, I'd expect those acute presentations to ease sooner than the cirrhosis burden in older patients. 

That's a hypothesis rather than something this study can show, and it doesn't account for co-factors of alcohol-associated liver disease that are changing in prevalence independent of alcohol consumption.

Editor's Note: This transcript has been edited for grammar and clarity using artificial intelligence tools. Lee reports consulting for Gilead Sciences, Novo Nordisk, GlaxoSmithKline, Altimmune, Bausch Health, and others.

Ayyala-Somayajula D, Dodge JL, Lee BP. Alcohol use between 2018 and 2024: a national cross-sectional study. 

. Published online September 22, 2026. doi:10.7326/ANNALS-26-01824

Keck Medicine of USC. U.S. alcohol use falls for first time since the COVID-19 pandemic, but remains above pre-pandemic levels. Published September 21, 2026. Accessed September 30, 2026. 

https://news.keckmedicine.org/us-alcohol-use-falls-for-first-time-since-the-covid-19-pandemic-but-remains-above-pre-pandemic-levels/

Articles

Alcohol use among US adults declined from 2022 to 2024, with any drinking dropping nearly 2% and heavy drinking falling more than 8%, although both remain above 2018 baselines. Trends diverged by cohort: Gen Z had the largest reduction in any use, while adults aged 50–64 increased both any use and heavy drinking substantially since 2018. Sustaining declines requires routine screening (eg, AUDIT-C), objective testing when history is unclear (phosphatidylethanol), and updated counseling on cancer/liver risk even at low intake.

Heavy drinking rose >36% since 2018 among middle-aged adults. A hepatologist on screening, sex-specific thresholds, and pandemic effects.

US Alcohol Use Falls, but Not for Adults Aged 50 to 64, With Brian Lee, MD

Improving Fibrosis Risk Stratification in MetALD, with Brian Lee, MD

In 2023, A modified Delphi process led by the American Association for the Study of Liver Disease, the European Association for the Study of the Liver, and the Latin American Association for the Study of the Liver yielded new liver disease nomenclature to address concerns regarding the terminology “nonalcoholic fatty liver disease” and “metabolic-associated fatty liver disease.”

The revised nomenclature introduced the term 

“metabolic dysfunction-associated steatotic liver disease” (MASLD)

 to describe patients who have hepatic steatosis and ≥ 1 of 5 cardiometabolic risk factors (CRMF). Additionally, the updated nomenclature added a new classification, MetALD, to represent a separate group of patients with MASLD who consume alcohol at a rate of 140–350 g/wk for females and 210–420 g/wk for males.

“There have been several questions that have come up because of the nomenclature, but one of the controversies was whether MetALD should be defined by just 1 metabolic risk factor, Brian Lee, MD, a hepatologist and liver transplant specialist with Keck Medicine of USC, explained to 

. “We know that alcohol can actually cause metabolic risk factors. So if you have just 1 and increased alcohol intake, is it truly a distinct disease?”

Lee noted this was one of several questions he and a group of colleagues sought to answer in their recent analysis of NHANES data assessing FIB-4 scores in adults ≥ 20 years of age with complete alcohol and CMRF status.

Results showed the prevalence of FIB-4 >2.67 among individuals with and without increased alcohol use was higher at each quantity of CMRFs and with each incremental CMRF:

0 CRMFs: (2.3%; 95% CI, 1.0-5.0% vs 0.7%; 95% CI, 0.5-0.9%)

1 CRMF: (3.0%; 95% CI, 1.6-5.6% vs 1.7%; 95% CI, 1.4-2.1%)

2 CRMFs: (3.3%; 95% CI, 2.1-5.1% vs 2.1%; 95% CI, 1.8-2.4%)

3 CRMFs: (5.9%; 95% CI, 3.5-9.6% vs 2.5%; 95% CI, 2.1-2.9%)

4 or 5 CRMFs (6.1%; 95% CI, 3.3-9.7% vs 4.0%; 95% CI, 3.5-4.5%)

Among individuals with increased alcohol, in multivariable logistic regression analysis, 3 (adjusted odds ratio [aOR], 2.57; 95% CI, 0.93-7.08) and 4 or 5 (aOR, 2.64; 95% CI, 1.05-6.67) CMRFs were associated with 2-fold greater odds of high FIB-4 versus 0 CRMFs, but not 1 (aOR, 1.24; 95% CI, 0.41-3.69) or 2 (aOR, 1.39; 95% CI, 0.56-3.50) CMRFs.

“The cutoff where you have a different risk of liver fibrosis is not with 1 cardiometabolic risk factor, but it's really with 3 or more,” Lee said. “So, the take-home message here is that we should be classifying MetALD with not just 1, but really 3 or more cardiometabolic risk factors, and maybe having this higher threshold for cut-off would better differentiate the patient populations in steatotic liver disease.”

Editors’ note: Lee has no relevant disclosures.

Brooks A. From NAFLD to MASLD: 2023 Brings New Liver Disease Nomenclature. 

. December 13, 2023. Accessed February 13, 2025. https://www.hcplive.com/view/from-nafld-to-masld-2023-new-liver-disease-nomenclature

Lee BP, Molina J, Kim S, et al. Association of Alcohol and Incremental Cardiometabolic Risk Factors with Liver Disease: A National Cross-Sectional Study. 

Videos

Lee explains study findings suggesting MetALD classification should be based on ≥ 3 cardiometabolic risk factors rather than just alcohol intake.