[mlyle]

The big issue with public health is that everything is dangerous and people will only comply with so many rules and society will only work towards so many goals.

Public health decided dietary fat was the big problem to focus on, despite imperfect evidence. Three things screwed this all up:

* Dietary fat wasn't nearly as bad as everyone thought, because eating lots of fats was correlated with lots of other negative lifestyle choices. * It looks like dietary sugar was actually worse than the evidence of the time said. * Policymakers didn't understand that an emphasis on reducing fats would cause food chemists and others to add more sugars and salt to food to make it more palatable.

In the end, we need to go where the evidence points. The earlier evidence that was hopelessly confounded was much more clear that fats and salt are universally bad; later evidence painted a much more ambiguous picture. It is so dang hard to measure what actually happens in real populations.

> With sugar we kids were taught all the usual facts at school: rotting teeth, getting fat and the risk of diabetes,

And the last we only have evidence of in the past couple of years, and it's relatively weak evidence.

> but perhaps what made those facts sink in and last a lifetime was that as the same time we were taught the heroic story of how Charles Best and Frederick Banting separated insulin in 1922 and immediately saved lives. This history was even more poignant to us as one our classmates had type-1 diabetes and it was obvious to all that he wouldn't be alive but for them. Both facts sharpened our minds about the dangers of high sugar intake.

Well, sugar intake isn't a risk factor or cause of type 1 diabetes.

[hilbert42]

"And the last we only have evidence of in the past couple of years, and it's relatively weak evidence."

Perhaps so, but it turned out to be good advice—I heeded the advice and have always limited my sugar and salt intake. As I said, at what point or how much evidence is need to heed such advice. Leaving it until factually certain could mean it's too late.

"Well, sugar intake isn't a risk factor or cause of type 1 diabetes."

That was known at the time (and to me also—in fact to all of us). The instance further highlighted the sugar/diabetes issue as we kids actually knew someone who had diabetes and who had to inject insulin (even if we didn't have diabetes, we knew we could still develop it and that excessive sugar would likely be the cause).

The thought of having to inject insulin daily focused our minds. That seems obvious doesn't it?

[mlyle]

> As I said, at what point or how much evidence is need to heed such advice.

This is like a Pascal's wager. There's a bazillion things we could be doing that there's a little evidence for or a plausible explanation. And we can't do them all, or believe in every God, or whatever.

And even things that we had a lot of evidence for (like dietary fat) turned out to be harmful when given as public health advice.

In the end, things are tricky. Best to do whatever seems like it's reasonably careful to you, and for science to try and figure more of this out and decide the best things to give advice on.