Search for a command to run...
Two doctors, both credentialed, both careful with their words, saying closer to opposite things about the same question. This is one of the cleanest disagreements in our dataset, and it is worth sitting with, because the resolution is not "one of them is wrong." It is "they are answering different questions, and the research lets you see exactly how."
On one side, Peter Attia, MD, on The Drive:
Any standard-deviation increase in alcohol consumption leads to a greater increase in the risk — from hypertension to dementia to cardiovascular disease, cancer, and all-cause mortality.
On the other, Dr. Tommy Wood, on Dr. Rangan Chatterjee's Feel Better, Live More:
You might see that a small amount of alcohol intake, or regular alcohol intake, is associated with lower risk of cardiovascular disease, but that's balanced by a slightly higher risk of certain cancers.
Listen to how precisely each is worded. Attia frames it as a dose-response: any increase in consumption raises risk across a basket of outcomes. Wood frames it as an association at low intake for one outcome, and immediately attaches the cancer qualifier that balances it. Neither man is being careless. The disagreement is real, but it lives in the framing, not just the conclusion.
We hold both of these in the database with full provenance — the video, the timestamp, the exact window. Attia's passage runs from 22:48 to 23:05 in the source. Wood's runs from 1:04:40 to 1:04:54. They are not paraphrased. They are the speakers' words, at the second.
Here is the part that matters. These claims sound contradictory, and in a thumbnail they are. But read closely and they are pointing at different slices of the same messy literature.
The "small amount of alcohol is associated with lower cardiovascular risk" finding is a real and well-known pattern in observational data. It is the famous J-shaped curve: moderate drinkers appear to have lower heart-disease risk than both heavy drinkers and non-drinkers. Wood is describing that.
The problem, and the reason Attia lands somewhere harsher, is that the J-shaped curve is one of the most confounded findings in nutrition research. Non-drinkers include former heavy drinkers who quit for health reasons, and people who abstain because they are already unwell. The "protective" effect of moderate drinking may be an artifact of who ends up in the comparison groups. Newer methods that try to handle this confounding — using genetic variants as natural experiments, for example — tend to erase the benefit and push the safest level of consumption toward zero.
So both statements can be honestly made from the literature. Wood is reporting what observational studies associate. Attia is reading the methodological critique and the all-cause picture — where cancer risk rises with alcohol and erases any narrow cardiovascular advantage — and drawing a stricter line. They are not equally right about what you should do, but they are both faithfully representing a strand of the evidence.
This is why "who is right?" is the wrong question. The right question is: which version of the evidence should I weight, and why? That is a question about study design, and it has an answer you can inspect.
Even after you understand the confounding argument, the research has limits it is honest to name.
Most of the data is observational. It shows association, not causation, and observational nutrition data is notoriously noisy. The genetic-instrument studies that cast doubt on the J-curve have their own assumptions and their own uncertainties. The cancer signal at low intake is small in absolute terms. And almost none of it speaks to your specific genetics, metabolism, or baseline risk.
So the defensible summary is not "alcohol is good" or "alcohol is bad." It is closer to: the apparent cardiovascular benefit of moderate drinking is probably overstated by simple observational comparisons, the all-cause picture is less favorable than the heart-only picture, and the safest intake is lower than the old "a glass of wine a day" consensus suggested. How much that matters depends on your other risks.
A claim that admits its own limits is more useful than a confident one that doesn't. We try to write toward the first.
When we render these claims — in the database, on the site, in the short video we cut of this exact disagreement — we keep the hedged language. "Associated with." "Balanced by." These words are not throat-clearing. They are load-bearing. Strip them and the claims become false.
Wood's claim is only defensible because of the cancer qualifier. Attia's is only defensible because of "all-cause mortality" and the dose-response framing. The precision is the content. A version of this disagreement that rounded off the edges into "alcohol is good for your heart" versus "alcohol is bad for you" would be simpler, more shareable, and wrong.
We would rather be slightly harder to read than easy to refute.
This disagreement is also a short. We took the two clips at their exact timestamps and cut them into a single vertical video so you can hear them back to back — same supplement topic, opposite framing, both from named sources with the show and speaker attributed. The captions preserve every qualifier.
The point of a format like that is not to stage a fight. It is to make the disagreement legible in fifteen seconds, so you have a reason to go read the studies. Controversy is the hook. The research is the product.
Explore the alcohol and cardiovascular-risk consensus page →
