Surgeon General's Unwitting Message: Drink More Alcohol, Live Longer
Vivek Murthy may not have read the studies his report is citing
Fifteen years ago at Columbia University my students debated the results of a study on alcohol and cancer, and a fawning editorial that summarized the findings by saying “No level of alcohol consumption can be considered safe.”
But the students, who had learned causation vs association and research methodology, came to a different conclusion: The study’s methods were flawed, the associations were weak, and the editorial was a spin job, cherry-picking and exaggerating to fit an agenda.
Class ended with this cartoon on the screen.
Energized by the students’ smart analysis I fired off an email to the editorial author, a prominent researcher (now director) at the NIH, asking why a scientist would write such things.
To my surprise, he answered. To my shock and admiration, he was honest.
The data were weak, he conceded, and his statements were misleading. But, he reminded me, drunk driving and alcoholism are serious problems that public health officials like him must tackle.
The message was clear: fearmongering dishonestly about alcohol and cancer was justified in the pursuit of public health.
He ended the email with a joke, and an emoticon. “Maybe we can discuss this further with your students over some wine…or beer… :-).”
Affable!
————
On Friday, US Surgeon General Vivek Murthy launched an eerily similar campaign against alcohol, posting a report that claims a “causal link” between cancer and alcohol at any level of consumption.
Sound familiar? These days even the World Health Organization agrees.
The report cites the data I’ve reviewed and described before, including the study my students debated. Even if Murthy somehow mistook the pitifully weak link between alcohol and, for instance, oral cancer as causation rather than association, the finding suggests an annual increased risk for drinkers of 1 in 7,000. Americans are more likely to choke to death on their food. (And yet they keep eating).
For sharper focus, check out this figure from the 2009 study:
In places where the horizontal lines touch the vertical line, it means the study found no link between alcohol and cancer—the finding for most cancers. But the lines don’t touch for a few at the top and bottom. Throat/mouth, liver, esophagus, and breast cancers are right of the line, suggesting a link. But kidney, lymph node, and thyroid cancers are left of the line, suggesting a negative link.
Yes. A negative link. For those cancers, drinking alcohol meant less cancers. Shocker: Murthy doesn’t mention these findings in his report.
Assuming he’s seen the data he cites, however, perhaps the Surgeon General will soon call for new labels on malt liquor: “To prevent kidney cancer, DRINK ME!”
But alas, sadly, moderate alcohol doesn’t actually prevent cancers any more than it causes them. Note the resemblance of the study figure to a ‘normal distribution’, i.e. a Bell Curve.
In a normal distribution of random events like dice rolls, most results fall in the middle. Snake eyes and boxcars are rare, while 5’s, 6’s, 7’s, 8’s and 9’s are common. Here’s the probability of various results when rolling two dice:
Notice a few outliers are high (boxcars!), a few are low (snake eyes!), and most are in the middle. Now, look at the figure from the cancer study. A few to the right, a few to the left, the great majority in the middle. Normal distribution.
But it’s sillier than that, because with alcohol and cancer even the outliers are boring. For comparison, the risk of lung cancer among smokers is 20-40 times higher than non-smokers. That’s not boring. That’s causation. But the risk of oral cancer for drinkers vs non-drinkers is 1.4 times higher. Is that even higher?? Boooooring.
And weak. And silly. And meaningless. And not remotely ‘causal’.
I hesitate to call anyone a liar, and I judge no one. I lied plenty when I was manic, and far worse. Moreover, exaggeration and spin are well-worn tools in public health. Think Reefer Madness (pleeease watch the video). Murthy and the NIH researcher are good people, trying to save lives.
But they lied.
They did it because public health authorities have an agenda. My email buddy plainly admitted he was aiming to change behavior, not simply report facts. And he shares this agenda with the Surgeon General and the WHO, two highly political institutions.
Just days before the Surgeon General’s report, however, the National Academy of Sciences released a “Review of Evidence on Alcohol and Health.” Written by distinguished, prominent, non-political scientists, the report was commissioned by Congress to inform the US dietary guidelines. The results are a stark and embarrassing contrast to the Surgeon General. They find no link between moderate alcohol intake and any cancer, other than a tiny, non-causal association with breast cancer (relative risk 1.1).1
They also find that moderate alcohol intake is associated with living longer. Which, somehow, didn’t make it into the Surgeon General report.
Piling on, Dr. Ashish Jha, Biden’s former COVID advisor, this weekend politely dismissed Murthy’s central claim during an appearance on national television.2
See? I’m not crazy—and neither were my students.
Dishonesty is a losing strategy. COVID failures proved it. Ignore inconvenient facts, spin data, or claim certainty when there is none, and you’ll get burned. Truth rises to the top and when it does people decide who to believe—and who not to.
The Surgeon General, the WHO, and the NIH director want you not to drive drunk or be alcoholic. Maybe next time they could just say that.
Jha talks about the report at the 45-minute mark in the video.
Thank you for this!