Alcohol Ups Mortality and Cancer Risk; No Net Benefit

Veronica Hackethal, MD

September 24, 2015

Drinking alcohol does not result in a net health benefit and, in fact, increases the risk for alcohol-related cancers by 51%, according to a study of almost 115,000 people from 12 countries.

It also showed that heavy drinking increases the risk for death by 31% to 54%, and that the highest rates of harmful alcohol use are seen in the lowest-income countries.

The results were published online September 17 in the Lancet.

"The association between alcohol consumption and health is extremely complex. Although alcohol consumption has been associated with some health benefits, it has also been associated with increased risks for other outcomes," said first author Andrew Smyth, MMedSc, a research fellow at the Population Health Research Institute, McMaster University, in Hamilton, Ontario, Canada.

"Our study suggests no overall benefit from alcohol consumption. Importantly, the greatest magnitude of increases in risk were seen in lower-income countries, where harmful alcohol use was highest," he reported. "Our data support global health strategies and national initiatives to reduce harmful alcohol use."

Our study suggests no overall benefit from alcohol consumption.

Alcohol consumption has been linked to more than 60 medical disorders and represents the third most important modifiable risk factor for death and disability.

Low to moderate drinking has been linked to a reduced risk for heart attack, whereas heavy episodic drinking can increase the risk for injury and sudden cardiac death. And the risk for certain cancers increases with the amount of alcohol consumed over time, the researchers report.

The prospective cohort study involved 114,970 adults with no history of heart disease, stroke, or cancer. Participants lived in 12 countries on five different continents and were part of the Prospective Urban Rural Epidemiological (PURE) study.

The participants were categorized into income groups on the basis of country of residence: the 12,904 from Canada and Sweden were categorized as high income (11%); the 24,408 from Argentina, Brazil, Chile, Poland, South Africa, and Turkey were categorized as upper middle income (21%); the 48,845 from China and Colombia were categorized as lower middle income (43%); and the 28,813 from India and Zimbabwe were categorized as lower income (25%).

The researchers assessed potential confounders such as high-density-lipoprotein cholesterol, body mass index, education, diabetes, hypertension, physical activity, diet, and smoking.

Self-reported alcohol use was categorized as abstinence (never drinking), former drinking, (no drinking for at least 1 year), or current drinking (drinking in the previous year). Level of consumption was categorized as low (up to seven drinks a week), moderate (seven to 14 drinks in women and seven to 21 drinks in men), or high (14 or more drinks in women and 21 or more drinks in men). Consumption of five or more drinks in one sitting at least once per month was considered heavy episodic drinking.

Of the 36,030 (31%) current drinkers, 72% had low consumption. Over a follow-up of about 4 years, current drinking was linked to a 24% lower risk for heart attack (hazard ratio [HR], 0.76; 95% confidence interval [CI], 0.63 - 0.93), a 51% increased risk for alcohol-related cancers (mouth, esophagus, stomach, colorectal, liver, breast, ovary, and head and neck) (HR, 1.51; 95% CI, 1.22 - 1.89), and a 29% increased risk for injury (HR, 1.29; 95% CI, 1.04 - 1.61). There was no reduction in the risk for death or stroke among current drinkers.

The risk for cardiovascular disease was lower in wine drinkers than in never drinkers, and the risk for heart attack was significantly lower (HR, 0.55; 95% CI, 0.39 - 0.77).

However, the risk for cancer was 38% higher in wine drinkers than in never drinkers (HR, 1.38; 95% CI, 1.05 - 1.81), 69% higher in spirit drinkers (HR, 1.69; 95% CI, 1.26 - 2.26), and 20% higher in beer drinkers (HR, 1.20; 95% CI, 0.91 - 1.57).

"The reduction in risk of heart attack is consistent with previous literature, both concerning red wine and low alcohol consumption. However, this may be offset by increases in risk for other outcomes," Dr Smyth pointed out.

People with high alcohol intake had a 31% increased risk for death (HR, 1.31; 95% CI, 1.04 - 1.66). Those with heavy episodic drinking had a 54% increased risk for mortality (HR, 1.54; 95% CI, 1.27 - 1.87) and a 71% increased risk for injury (HR, 1.71; 95% CI, 1.14 - 2.56).

More than three-quarters of people in high-income countries consumed alcohol, whereas only one-eighth of those in low-income countries did. However, even though low-income countries had the lowest frequency of current drinking, they also had the highest rates of current drinkers with high intake and heavy episodic drinking patterns.

In higher-income countries, risk on a composite score indicating the net association between alcohol and health outcomes was significantly lower for current drinkers than for never drinkers (HR, 0.84; 95% CI, 0.77 - 0.92). In lower-income countries, there was no reduction in composite score for current drinkers (HR, 1.07; 0.95 - 1.21; P interaction ≤ .0001).

Dr Smyth and his colleagues emphasize that people who do not drink should not be advised to start drinking because of the potential to increase consumption or to start drinking in a heavy episodic pattern.

A detailed assessment of alcohol use during follow-up is lacking in this study, Jason P. Connor, PhD, from the University of Queensland in Herston, Australia, and Wayne Hall, PhD, from King's College London in the United Kingdom, write in an accompanying comment.

Even though outcomes on all health measures assessed were worse in former drinkers, the researchers did not collect data on how much alcohol these people drank before they abstained. In addition, relatively few adverse events occurred during the short follow-up period, which affects the study's statistical power, they note.

Nevertheless, Drs Connor and Hall commend the researchers, noting that the value of the PURE study "will greatly increase as the number of adverse health outcomes accumulates with longer follow-up."

"In the meantime, we should not delay action," they write. "More than sufficient evidence is available for governments to give increased public health priority to reducing alcohol-related disease burden in low-income and middle-income countries."

"This should be done by implementing the most effective population policies to discourage harmful drinking — namely, increasing the price of alcohol and reducing its availability, especially to younger drinkers, and preventing the alcohol industry from promotion of frequent drinking to intoxication," they explain.

The study authors, Dr Connor, and Dr Hall have disclosed no relevant financial relationships.

Lancet. Published online September 17, 2015. Abstract, Comment


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.