Could Nightly Drinks Be Good for the Brain?

F. Perry Wilson, MD, MSCE


July 01, 2020

This transcript has been edited for clarity.

Welcome to Impact Factor, your weekly dose of commentary on a new medical study. I'm Dr F. Perry Wilson at the Yale School of Medicine.

This week, in solidarity with many people in this country, I am pretending that coronavirus doesn't exist and talking about a paper that has absolutely nothing to do with that horrible strand of RNA raining destruction around the world.

I apologize for being flip; it's been a frustrating time in the world of epidemiology.

But it's okay to take a break here and there, and this paper appearing in JAMA Network Open caught my eye — not for the purity of the science or the methods, per se, but because the conclusion comported with something that, in the face of lockdowns and shelter-in-place orders, I really wanted to be true: Drinking alcohol might be good for you.

I recognize the bias; cherry-picking studies because they align with what you want is a terrible way to report on science. But fear not, loyal viewers. I have not fully abandoned my principles. In fact, after reading the paper, I'm not any closer to convincing myself that a couple of Dogfish Heads a night are a good thing.

Researchers examined data from the Health and Retirement Study, a nationally representative sample of US adults. More than 19,000 individuals [participated in] biennial surveys of cognitive function looking at domains of mental status, word recall, and vocabulary. You had to have [participated in] at least three such surveys to make it into this analysis, allowing for modeling of cognitive trajectory. Of course, individuals also self-reported on their drinking habits. They were classified as never-drinkers, former drinkers, low to moderate drinkers, and heavy drinkers, based on this rubric:


Now, the analysis I wanted was a nice mixed-effects model, which would examine the rate of drinking over time as the independent variable and cognitive function over time as the dependent variable, with adjustment for a slew of relevant confounders like age, socioeconomic status, comorbidities, smoking history, medication usage, etc.

But this is not what I got.

The researchers used a data-driven approach to model cognitive trajectories. Basically, you feed the cognitive data and time into an algorithm, and it tries to fit all of the data into discrete groups based on some information theoretic principles. You don't know going in how many groups it will find. It might find a group characterized by rapid cognitive decline or slow cognitive improvement. The data drive the grouping. Cool.

But what these data led to were two groups, as you can see here: consistently low cognitive trajectory and consistently high cognitive trajectory.


This is our first sign that we're not going to get the information we want. It seems there is not enough variation in the study cohort to determine how alcohol use may change your cognitive trajectory.

In other words, what the data show is that people who started off with good cognitive scores kept good cognitive scores, and those who started off worse stayed worse.

This does not bode well for my dream analysis. I am not surprised that most people stay in their cognitive lane. Regardless of which lane I am in, what I care about is whether my drinking is going to change my outcome.

The researchers showed, after adjustment for age, sex, race, ethnicity, marital status, smoking status, and BMI, that low to moderate drinking was associated with a significantly lower risk of being in the persistently low cognitive group.


Again, I don't find this terribly interesting. I don't want to know how likely it is that I'm in the persistently low group; I want to know how to get out of it.

To be fair, there was a small analysis looking at rate of change in cognitive score over time that, for whatever reason, was given short shrift in the discussion. It showed that, after adjustment, people who drank more had improved cognitive scores compared with those who never drank.


The authors offered some interesting biologic rationale as to why drinking might be protective, focusing on brain-derived neurotrophic levels and neural plasticity.

But let's be honest. Low to moderate alcohol drinking is likely a sign of socioeconomic status, which can be very difficult to adjust for — and wasn't adjusted for at all in this study, as far as I can tell. Drinking a couple glasses of wine a night takes disposable income and probably a social structure (like marriage, family, and friends) that promotes health and cognition.

An analysis stratifying the results by race confirms this supposition for me. The researchers found that the protection of low to moderate drinking was only seen in white people. No such effect was seen in Black people.


When something like this happens in a study, it's a dead giveaway that there is a sociologic phenomenon driving this rather than a biologic one, because there is no reason to think that there are substantial differences in the effect of alcohol on the brain between Black and white individuals.

So I am left, alas, without confirmation that the second beer of the night is good, healthful behavior. It's probably not. But it's probably not too harmful either. And hey, if we're all destined to stay in our cognitive lane, we might as well enjoy the ride.

F. Perry Wilson, MD, MSCE, is an associate professor of medicine and director of Yale's Program of Applied Translational Research. His science communication work can be found in the Huffington Post, on NPR, and here on Medscape. He tweets @methodsmanmd and hosts a repository of his communication work at

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


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.