Gaping hole in scientific case for moderate drinking

Past studies missed reason non-drinkers didn't live as long as people who imbibed a bit

It's an irresistible headline: People who drink alcohol in moderation actually live longer than those who abstain entirely. Counterintuitive studies that show the purported benefits of a drink or two a day prompt flurries of bright news reports. You can hear the glasses clinking.

Buzzkill: It's probably not true, according to a new analysis of existing research published in the peer-reviewed Journal of Studies on Alcohol and Drugs.

It turns out many studies showing that moderate drinkers live longer than non-drinkers suffer from a big flaw: The "abstainers" category includes people who used to drink but have stopped, sometimes for health reasons. They may be inherently less healthy, as a group, than people who drink in moderation. That doesn't mean that drinking in moderation causes people to live longer.

Researchers at the University of Victoria in British Columbia examined 87 studies on the relationship between alcohol and mortality that together involved nearly four million people. Only 13 of them strictly separated life-long non-drinkers from people who used to drink or those who imbibe only occasionally. Analyzing the studies that were free from "abstainer bias" showed no significant benefits for moderate drinkers compared to lifetime abstainers. And when they compared moderate drinkers with those "occasional drinkers" who consumed alcohol less than once a week, they found no benefit to drinking more often, said Tim Stockwell, lead author of the paper and director of the Center for Addictions Research of BC.

"It's just an example of why this observational research is very fraught," he said.

Some scholars have been questioning the purported health benefits of alcohol for years. The debate is complicated because most of the evidence on the long-term effects of alcohol-or diet, exercise, and other lifestyle factors, for that matter-is observational. Scientists examine people's behaviour to try to find associations with different health outcomes.

That contrasts with experimental research, which is the method used to test new drugs: People are randomly assigned to get either the drug or a placebo, and neither the patients nor the scientists know which is which. Such double-blind, randomised control trials are designed to reduce biases that could inadvertently skew the results.

They are also nearly impossible to do for long-term research on drinking or diet.

Observational research can show associations between two things, such as drinking and mortality. But proving that one affects the other is much harder. Despite the inherent limits, these studies have sometimes been taken as evidence that drinking in moderate amounts causes people to live longer than not drinking at all.

"As an intoxicating, addictive, toxic, carcinogenic drug, alcohol is not a good choice as a therapeutic agent," wrote Jennie Connor, chair in preventive and social medicine at the University of Otago in New Zealand, in a commentary accompanying the new study.

Emanuel Rubin, a pathologist and professor at Thomas Jefferson University, suggested last year that the evidence was strong enough that doctors should consider recommending that patients start drinking. While medical advice should be tailored to patients' specific circumstances, "the overwhelming evidence suggests that physicians should counsel lifelong nondrinkers at about 40 to 50 years of age to relax and take a drink a day, preferably with dinner," Rubin wrote in the journal Alcoholism: Clinical & Experimental Research.

Rubin concedes that there may be "possible confounders" that affect the observational studies, but he said that "the evidence is pretty good that there's a protective effect to moderate alcohol consumption" for several health conditions, like coronary artery disease and stroke. (Rubin served in the 1990s on an advisory council for a non-profit group funded by the brewing industry. He said he had no current ties to the alcohol industry and hasn't for years.)

The new study didn't address the specific effects of red wine, for example, which has been linked in some research to heart benefits. The researchers said their findings apply only to alcohol's overall effect on people's mortality, and that "different risk relationships likely apply" to specific diseases.

Others argue that there's no place for doctors or public health officials to recommend drinking, and that talk of potential health benefits is a distraction from the real harms of excessive consumption.

The research on potential health benefits of alcohol needs a more skeptical evaluation, by scientists, journalists, and the public alike, said Stockwell, the study's lead author.

"In fairness to the journals, this is a contested area, and it's an illustration of how subjective the peer review process is," he said. "We hope our contribution is to put the skepticism back in there."