Drinking Habits Barely Budge Among U.S. Cancer Survivors, National Survey Reveals

Alcohol is now firmly established as a modifiable cancer risk factor, linked by decades of epidemiological evidence to cancers of the head and neck, esophagus, liver, breast, and colorectum, with suggestive associations extending to melanoma, pancreatic, prostate, and stomach cancers. Yet a sweeping new analysis of nearly four million American adults suggests that the people most directly affected by a cancer diagnosis are not dramatically changing their drinking habits, and that the broader population’s consumption is declining only at a glacial pace. The study, published in Cancer Causes & Control, offers one of the most detailed national portraits to date of how recent alcohol consumption differs between people with and without a history of cancer, and how those patterns have shifted across nearly a decade.

Researchers at Washington University in St. Louis, led by Kimberly J. Johnson, drew on the Behavioral Risk Factor Surveillance System, or BRFSS, a giant telephone survey run by the U.S. Centers for Disease Control and Prevention that collects health information from more than 400,000 residents each year across all fifty states, the District of Columbia, and U.S. territories. For their trend analysis, the team pooled data from 2016 through 2024, encompassing 3,968,799 respondents after imputation of missing values. For a finer-grained comparison of drinking prevalence by cancer type, they used the most recent 2022 to 2024 data from the 38 states and territories that included an optional cancer survivorship module, yielding 465,870 respondents. Because BRFSS uses complex, stratified sampling, the analysts applied survey weights and multiple imputation, then pooled estimates using established statistical rules to produce nationally representative figures.

The survey questions distinguished three patterns of consumption. Current drinking was defined as having had at least one alcoholic drink in the past 30 days. Binge drinking meant four or more drinks on a single occasion for women, or five or more for men, while heavy drinking was classified as more than seven drinks per week for women and more than fourteen for men. Cancer history came from a question asking whether a respondent had ever been told by a doctor that they had cancer, and a follow-up item identified the specific cancer type. The researchers classified head and neck, esophageal, liver, breast, and colorectal cancers as alcohol-related, following the National Cancer Institute’s categorization.

The headline finding is a story of divergence, albeit a subtle one. Among adults without a cancer history, the prevalence of current drinking declined significantly, from 53.9 percent in 2016 to 51.5 percent in 2024, an average drop of 0.28 percentage points per year. Binge drinking in this group also fell, by about 0.20 percentage points annually. Among people with a cancer history, however, current drinking was essentially flat, hovering at 45.8 percent in both 2016 and 2024, with a statistically non-significant upward drift of 0.06 percentage points per year. Binge and heavy drinking trends in survivors were similarly unremarkable. In other words, whatever forces are nudging the general population toward slightly less drinking, they do not appear to be operating among cancer survivors, who started from a lower baseline and stayed there.

That baseline gap persisted after statistical adjustment. Compared with adults never diagnosed with cancer, those with a cancer history had an adjusted prevalence of current drinking 1.01 percentage points lower and binge drinking 1.16 percentage points lower, while heavy drinking showed no significant difference. People who had survived an alcohol-related cancer, such as breast or colorectal cancer, reported current drinking at a prevalence 2.17 percentage points lower than the cancer-free population, though their binge and heavy drinking rates were not significantly different. The direction of these differences held even when the analysis was restricted to adults aged 55 and older, although the estimates were attenuated and generally lost statistical significance.

The cancer-type breakdown produced some of the study’s most intriguing results. Current drinking prevalence was significantly higher than expected among survivors of testicular cancer, melanoma, and other skin cancers, while most other cancer types showed lower prevalences, with prostate and throat or pharyngeal cancers non-significantly elevated. The most common cancer types in the dataset were breast, non-melanoma skin cancer, melanoma, and prostate cancer. Because small cell sizes limited stable estimation for binge and heavy drinking within individual cancer types, the type-specific models focused on current drinking, with sex-specific restrictions applied for reproductive cancers.

One of the most revealing analyses came from a simple sensitivity check. Because some patients reduce or stop drinking during cancer treatment, the researchers repeated their prevalence difference models excluding respondents who reported currently being in treatment. The result was striking: most of the negative prevalence differences, where survivors drank less than the cancer-free population, shifted toward zero, while some of the positive differences moved further from the null. The authors interpret this as suggesting that many individuals resume alcohol consumption once treatment ends, meaning that the apparent protective gap between survivors and others may partly reflect temporary abstinence during therapy rather than a durable lifestyle change. The cross-sectional nature of the survey data, however, prevents any conclusion that a cancer diagnosis itself causes changes in drinking behavior.

The findings arrive against a backdrop of intensifying public debate about alcohol and cancer. The scientific consensus has hardened over the past decade, with meta-analyses confirming dose-response relationships and population attributable fractions estimated at 4.4 percent for female breast cancer and as high as 31.6 percent for esophageal cancer. In 2025, the U.S. Surgeon General called for cancer warning labels on alcoholic beverages, echoing advocacy from consumer and public health groups dating back to 2019. The American Cancer Society now states flatly that it is best not to drink alcohol, and the U.S. Preventive Services Task Force recommends that adults be screened for unhealthy alcohol use. Google Trends data show a steady rise in searches connecting alcohol and cancer risk since 2004. Yet awareness remains startlingly low: in a 2024 survey, only 39.4 percent of U.S. adults knew that alcohol increases cancer risk, and earlier data suggested more than half were unaware that wine, beer, or liquor carries any such risk.

Against this context, the stability of drinking among survivors is particularly consequential. Alcohol consumption may raise the risk of second primary cancers in survivors, making continued drinking a clinically meaningful concern rather than a matter of taste alone. Prior research offers mixed signals about whether diagnosis changes behavior. A Canadian study of more than 500 survivors found that among 299 patients who drank at diagnosis, 52 percent reported reduced or no consumption one year later. A study of 973 head and neck cancer patients recorded a decline in drinking from 54.3 percent to 41.2 percent within a year of diagnosis. But a German study of 300 survivors found that 70 percent continued to drink six months after diagnosis. The new BRFSS analysis suggests that, at the population level, any such reductions are modest and may fade once treatment concludes.

The study also highlights a persistent gap in clinical practice. Although alcohol screening is recommended in oncology and primary care settings, previous research indicates that heavy drinkers with a cancer history are not consistently counseled to cut back; one study found that only 15 percent of cancer survivors consuming three or more drinks per day were advised by a provider to reduce their intake. The authors note that their results provide a contemporary baseline against which future changes can be measured, as dietary guidance and the alcohol environment evolve. They also acknowledge limitations: cancer history and alcohol use were self-reported and subject to recall and social desirability bias, some states lacked data in certain years, and small samples for some cancer types reduced precision. Survivors aware of alcohol’s risks may also under-report their drinking, potentially widening the apparent gap. Still, with nearly four million respondents, weighted analyses, and careful handling of missing data, the study stands as a sobering snapshot: as national conversation about alcohol and cancer grows louder, the drinking habits of American cancer survivors have barely moved at all.

Subject of Research: Alcohol consumption patterns and trends among U.S. adults with and without a cancer history

Article Title: Prevalence and trends in recent alcohol consumption by cancer history and cancer type in a nationally representative sample of U.S. adults

Article References: Johnson, K. J., Eyler, A. A., & Harris, J. K. (2026). Prevalence and trends in recent alcohol consumption by cancer history and cancer type in a nationally representative sample of U.S. adults. Cancer Causes & Control, 37(10), Article 167. https://doi.org/10.1007/s10552-026-02248-8

Image Credits: AI Generated

DOI: 10.1007/s10552-026-02248-8

Keywords: alcohol consumption, cancer survivors, cancer risk, BRFSS, binge drinking, heavy drinking, cancer epidemiology, public health, surgeon general, health behavior, national survey, alcohol-related cancers