How Much Alcohol Is Too Much?
A new analysis explores the relationship between drinking rates and disease.
Posted Apr 26, 2018
Lots of people drink. Most think they “don’t drink too much.” They “only drink socially.” Or “just a couple of drinks a few nights of the week.”
But when it comes to health, how much is "too much"?
A new study looking at national guidelines has come up with surprising results. The short answer is “less is best.”
Perhaps a lot less.
Alcohol and the Heart
About 30 years ago, Richard Peto (now Sir Richard) found that small amounts of drinking led to decreased risks of heart attack. Many patients, and certainly the drinks industry, was heartened by these results. Daily drinking was “healthy"—if you looked at heart attacks.
A new study in The Lancet proved a bit more exhaustive than most. It involved 600,000 people in 19 countries who were current drinkers (teetotalers are epidemiologically different in subtle ways), were followed prospectively, and whose results could be controlled for diabetes, hypertension, and the standard confounders. The main aim was to look at cardiovascular disease, but all-cause mortality—including the threat that alcohol poses for cancer, liver disease, accidents and dementia—was included.
The researchers found that more than 100 grams of alcohol (there are 28 grams in an ounce) a week predicted lower odds of survival.
The finding that drinking small amounts cut back on heart attacks held up. But looking at stroke, coronary disease besides heart attacks, heart failure, fatal hypertension, or aortic aneurysm, the death rates went up. The researchers finally concluded, in language worthy of legal scholars, that regarding cardiovascular disease, “there were no clear risk thresholds below which lower alcohol consumption stopped being associated with lower disease risk.” In English, that means there was no level of drinking that was not associated with some increase in disease risk.
Yet in this study, cardiovascular disease constitutes only about one fifth of alcohol’s overall association with mortality.
What’s In a Drink
Ask Americans “what’s in a drink of alcohol” and you often get very different answers. Some report an “ounce” of alcohol as a “large glass of wine” or “two fingers” in a shot glass.
The actual American "standard drink" is a half ounce of alcohol, 14 grams. It’s what you usually find in one beer, five ounces of “most” wines, and one-and-a-half ounces of vodka or whiskey. When it comes to estimating how much they drink, people are often poor self-reporters. Typically they underestimate. Just as they dramatically underestimate the physical effects of alcohol.
Mortality and Booze
Perhaps the most alarming result of this international study was that as people drank more, the increased death rate became curvilinear—going up faster.
It was particularly bad for those who drank 350 grams of alcohol or more a week. By American standards, that’s 25 drinks, or three and a half drinks a day. The researchers decided to look at this in terms of years of lost life, from age 40 on.
The older one got, the more the effects returned towards a straight line relationship between drinking and death rates. That means in part that people who made it to older ages were more resistant to disease than those who died earlier. But if you looked at steadily drinking 3.5 drinks a day at age 40, the predicted years of life lost was about five. Even for those drinking 200 grams a week, or about two drinks a day, the figure approached two years.
Two years may not sound like much, but that’s close to what all of medical care is supposed to add to one’s lifespan. How many would be willing to foreclose all medical care for their life starting at age 40? More importantly, mortality statistics do not include the many other effects of alcohol on morbidity, including recent research suggesting that “minor” drinking in the middle age increases the risk of dementia.
Heavy drinking, here defined as 3.5 standard American drinks a day, is associated with the kind of mortality statistics you generally see with smoking. And that’s aside from its effects on mood, family, well-being, accidents, work, social and economic and performance, and memory. There are many reasons to drink—socializing, conviviality, culture, amusement, pleasure, tradition. However, physical health, as measured by mortality, may not be one of them.