Large U.S. study finds no significant link between moderate drinking and cancer deaths

A large U.S. study has found no statistically significant link between moderate alcohol consumption and cancer mortality, a result that adds new evidence to a debate that has shaped recent public health guidance on drinking.

Researchers at Weill Cornell Medicine analyzed data from 26,694 Americans and found that people classified as moderate drinkers were no more likely to die from cancer than non-drinkers over an average follow-up of about 13.3 years. Many participants were followed for more than 20 years.

The study, published in May in the Journal of General Internal Medicine, examined data from the REGARDS study, an ongoing National Institutes of Health-funded project that enrolled more than 30,000 Americans between 2003 and 2007 and continues to track them.

The findings stand in contrast to the broad warning issued last year by former U.S. Surgeon General Vivek Murthy, who said alcohol raises cancer risk. According to the Weill Cornell team, some earlier studies used to support such recommendations had important limitations, including incomplete adjustment for factors that can affect cancer outcomes, such as smoking, diet and socioeconomic status.

In the new analysis, moderate drinking was defined as four to seven drinks a week for women and four to 14 drinks a week for men. Lead author Dr. Laura Pinheiro, an associate professor of health services research at Weill Cornell Medicine, said the researchers had expected to find a higher risk of cancer death among moderate drinkers given recent federal messaging.

“It was surprising that we didn’t observe a significant relationship between moderate alcohol consumption and cancer death,” Pinheiro said.

The researchers said they designed the study to address weaknesses in prior work by adjusting for a wide range of potential confounding factors, including smoking status, physical activity, socioeconomic conditions and medical history such as hypertension, high cholesterol, diabetes, heart disease, stroke and obesity.

Their final analysis found no association between moderate drinking and cancer mortality. It also found that light drinkers, defined as people consuming up to three drinks a week, had a lower risk of dying from cancer than abstainers. The authors noted that some non-drinkers may have stopped drinking because of existing health concerns, which can complicate comparisons, but said the apparent protective effect remained after statistical adjustments.

Heavy drinkers showed a different pattern. In the study, heavy drinking meant eight or more drinks a week for women and 15 or more for men. Compared with moderate drinkers, heavy drinkers had about a 57% higher risk of dying from cancer during the follow-up period.

Pinheiro said the association between heavy alcohol use and cancer deaths remained statistically significant even after accounting for multiple confounders.

The authors also stressed that the study does not prove cause and effect. Because it is observational, unmeasured confounding cannot be ruled out. Another limitation is that the mortality records only identified whether participants died from cancer, without details on cancer type, stage, treatment or screening history.

Dr. Erikka Loftfield of QIMR Berghofer in Brisbane, Australia, said that missing detail is especially important for cancers commonly linked to alcohol exposure, including breast and colorectal cancer. She also noted that smoking remains one of the hardest factors to fully separate from alcohol use in this kind of research.

The REGARDS data included participants’ drinking habits only once, at enrollment. That means the analysis could not capture changes in alcohol consumption later in life, including after illness or a cancer scare.

The authors said their results support looking at cancer prevention through broader lifestyle patterns rather than focusing on one behavior alone. Pinheiro said risk is shaped by many factors, including diet, exercise, smoking and social conditions.

The findings may matter beyond medicine because alcohol policy debates can influence consumer behavior across wine, beer and spirits. Research that distinguishes between light, moderate and heavy drinking could affect how regulators, producers and retailers respond to future health warnings and labeling proposals.

Pinheiro also pointed to another question she believes deserves more study: rising cancer rates among younger adults at a time when younger Americans are generally drinking less than earlier generations. She highlighted colorectal cancer in particular, saying its increase among younger people does not fit a simple narrative about alcohol alone.

According to figures cited by the researchers, about one in 26 women and one in 25 men will develop colorectal cancer. The Cancer Research Institute has said cases among people under 50 have been rising by nearly 3% a year and that colorectal cancer is now the leading cause of cancer-related death among young adults.