Is Moderate Drinking Even Moderately Good For Us?

Recent studies have shown that alcohol, coffee, and chocolate are probably good for you, which almost makes up for the fact that bacon, doughnuts, and anything covered with that diabolical orange dust are probably bad.  So it’s quite a blow that one of them might not be good after all.

It’s alcohol.

The benefits of moderate drinking have been widely accepted, and new information that questions those benefits has to be looked at in the context of all the old information that established them. But when two new studies, each coming at the question differently, conclude that the benefits of moderate drinking may be illusory, it’s worth taking a closer look.

The more recent, published just this March, took a fresh look at the population studies that gave rise to the idea that, although heavy drinking is undeniably bad for you, a drink or two each day has health benefits. The problem with many of those studies, according to University of Victoria psychologist Timothy Stockwell, the lead author on the new analysis, is flawed methodology. Correct that, he says, and the effect vanishes.

The trouble is that people who drink tend to be different from people who don’t drink in all kinds of ways. And, people who drink a lot are different from people who drink a little. Trying to isolate that one variable—the amount they drink—is an undertaking fraught with peril.

One of the main problems is that people who don’t drink at all are, according to Stockwell, “a pretty unhealthy group.” Sometimes, they’ve stopped drinking for health reasons, and people who drink moderately look different from that group in a lot of ways: they’re wealthier, they eat more fruit and vegetables, and they exercise more. Stockwell contends that it’s likely that those factors, and not the moderate drinking, that account for any health benefits.

While some of the benefits attributed to moderate drinking are plausible—such as reduced heart disease, because alcohol does affect biomarkers like cholesterol and inflammation—others don’t make much sense.  Are we ready to believe that moderate drinking protects against cirrhosis of the liver?  “Highly unlikely,” say the researchers. That “benefit” is a testament to the perils of trying to tease health effects out of population data.

Stockwell isn’t ready to put the last nail in the coffin of moderate drinking. The paper concludes, soberly (get it?), that “available evidence supports a skeptical position towards the hypothesis that alcohol in moderation confers net mortality benefits.”

Meanwhile, across the Atlantic, Michael Holmes, research fellow at Oxford University’s Nuffield Deparment of Population Health, is looking at drinking from a completely different angle: Mendelian randomization.

The trouble is that people who drink tend to be different from people who don’t drink in all kinds of ways. And, people who drink a lot are different from people who drink a little. Trying to isolate that one variable—the amount they drink—is an undertaking fraught with peril.

The “Mendel” in “Mendelian” is, of course, Gregor. And the technique is all about genetics. People who drink less tend to be very different from people who drink more because they’re a self-selected group. You choose to drink less, or more, for a reason, and that reason may say a lot about you. But what if your reason didn’t say anything at all about you? What if you had a gene that made you feel crappy if you drank too much, and so you didn’t?

There is such a gene. It has a catchy name: the rs1229984 variant in the alcohol dehydrogenase 1B gene. People who have it tend to get flushed and, sometimes, feel ill or uncomfortable when they drink. So it’s not surprising that they drink less—17 percent less—than people who don’t have the gene.

Compare the people who have the gene with a demographically similar group that doesn’t have the gene, and you have a good shot at looking at the health effects of drinking without all those confounding behavioral factors.

Holmes and his co-authors found that the people who drink less have “a more favourable cardiovascular profile and a reduced risk of coronary heart disease.” Their conclusion? Even if you’re a light drinker, you’ll benefit by drinking less.

So Stockwell and Holmes, playing two different angles, arrived at the same conclusions.

Eric Rimm, director of cardiovascular epidemiology at the Harvard School of Public Health (and one of the many, many co-authors on the Holmes paper), points out that we still have a great deal of evidence that alcohol, in moderation, is beneficial.

Some of that evidence doesn’t require us to parse population data—there are controlled trials that pit drinking against non-drinking. “If you take 50 people, and half have one drink and half have water, after 2-3 weeks there are lots of beneficial things in the blood of the alcohol group. Good cholesterol goes up. Inflammation goes down.”

The problem is, while we can assess interim markers like cholesterol, we can’t keep those research subjects prisoner until they die (or don’t).  And if the debate about saturated fat has taught us anything, it’s the difficulty of drawing conclusions about disease and death from markers like cholesterol.

Rimm also points out that the drink-less gene isn’t randomly distributed in the population. It’s overrepresented in, for example, Ashkenazi Jews. The gene may also affect alcohol metabolism in a way that has repercussions for health. We don’t understand it completely.

All in all, It’s too early to give up that glass of wine with dinner (for me, at least), but this could be an early warning that it may not be as good for you as you thought. I think I can live with that but, when they come for the coffee, I’m going back to doughnuts.

 

Tamar Haspel is a food and science journalist. She can be reached on Twitter @TamarHaspel.

 

Comments

Comments (3)

  1. Jason (April 27, 2016)

    Rubbish. People get paid for this? I need to create a study group and snag a check.

  2. Lynne (April 30, 2016)

    the closing sentence made me laugh! i’m with you on that!!!

  3. Jay (May 1, 2016)

    Well thought out critical news approach to methodologically based statistical indicators. Isolation of variables is a complex task which when done differently based on human judgment can have a significant impact.