Discussion about this post

User's avatar
D.E.Kieras's avatar

Please do not quote an "AI" in the future - these are still, and will almost certainly remain, unreliable and untrustworthy. I've followed your postings because I trust *your* reliability.

I've followed, made use of, and contributed to artificial intelligence research for all of my professional career. I trust you on medical and health issues - kindly trust me on this.

Expand full comment
J Das's avatar

For full impact instead of reading I listened to the AI-generated narration of this post.

More seriously, and to actually engage with the content of this post and the larger conversation about dairy fat, I have been contemplating the following question: for the sake of argument let’s say that at the population level it is true that eating full fat yogurt doesn’t cause obesity compared with eating reduced or fat free yogurt. How then are we to respond to an individual who finds that for them personally it is much, much easier to maintain a healthy body weight eating low fat foods? I’m not talking about foods where the fat has been replaced with sugar, but things like unflavored non-fat Greek yogurt. I think the naive “evidence based” approach would be to just say the individual is wrong and they actually are compensating in some other way that is causing them to gain weight. Okay.

But then if you look at almost any study in which you find even a very strong signal in one direction, you can almost always find individual study participants who had the opposite outcome or no change at all. Are they “wrong?” What about those who had outlier responses in the same direction as the population-level signal?

This is all to say that I think population-level averages are a very good starting place but then you have to look at the individual response and adjust accordingly.

Expand full comment
33 more comments...

No posts