We Actually Have No Idea How Much We Drink

A few months ago, after reading Sarah Probst's thoughtful story about her month without booze, I took the so-called "alcohol abuse self-test" from the National Council on Alcoholism and Drug Dependence. I didn't really think I had an alcohol problem — I certainly wasn't pouring whiskey into my Starbucks in the morning — but I was curious about where I fell on the seemingly arbitrary spectrum of "abuse." Much to my surprise, after answering the questions as honestly as possible, I scored pretty darn high.

Of course, previous research has shown that surveys like this one are notoriously inaccurate, as most of us can't remember — or don't want to disclose — how much we really drink. A new study published in the journal Addiction tried to figure out if we're kidding ourselves about our drinking habits.

Researchers at the University of Victoria in Canada conducted a phone-based survey of 43,371 Canadian subjects between 2008 and 2010. In addition to the standard alcohol-questionnaire items ("How many times do you consume alcohol per month?"), the survey also asked participants about how many drinks they had yesterday. The idea was that subjects' answers about their regular drinking habits could be fact-checked against how much they drank the day before the study. For example, someone who reportedly had 10 drinks per month but six drinks last night is probably underestimating his or her total.

The researchers calculated the chances that each subject was low-balling his or her habits (based on those "yesterday" responses). Then, they compared the reported habits with data on alcohol sales in Canada to figure out how misrepresentation differed by drink type. The verdict: Survey participants underreported their hard alcohol consumption by almost 66%, while beer and wine drinking were underestimated by 38% and 49%, respectively. Unsurprisingly, young people underreported their drinking to a greater degree than their older counterparts (83% vs. 70%). And, low-risk drinkers (for example, women who consume no more than three drinks in one day and no more than seven drinks in a week) tended to underestimate their consumption more significantly than medium- and high-risk drinkers (76% versus 62% and 49%, respectively).

Of course, this study doesn't provide an explanation for such a large discrepancy between reported activity and actual drinking patterns — which, I'd argue, is probably the most important part of the story. Lying about your boozing is one thing; it implies that you are at least tangentially aware that your drinking could be deemed "excessive." But, if the problem is that most of us simply can't remember how much we drank on a given night, we're talking about a much more dangerous situation. A good rule of thumb? If you've completely lost track of how much you've knocked back, the answer is probably "too much."