Dr. Donald Layman

And so that's a meaningful number.

But in a non-growing adult, it's not very meaningful.

And I don't know of any help outcome that relates to it at all.

And I sort of understand the dilemma.

The public would like nice, clean answers.

And so the people in charge would like to give them that.

But the problem is in nutrition, we don't have nice, clean answers.

And I think the difference that Kennedy's looking at between hard science is that we would like to have what we call random control trials, where people are actually

well-controlled, what they're eating is well-established, well-controlled.

A lot of the nutrition right now is using what we call epidemiology, which is where you basically go out and ask someone what they ate yesterday and they give you some answer, probably make it up.

And then 20 years later, you ask how their health turned out.

And I, for one, don't remember what I ate yesterday.

I don't remember what I ate this morning.

And so I think the problem with it, in computer language, you always say garbage in, garbage out.

If you start out with really bad food data, making conclusions from it 20 years later is not very good.

So I think Kennedy is arguing that that's not the way we should do nutrition science.

Back in the old days, epidemiology was good for something like smoking.

You either smoked or you didn't.

And 20 years later, you can say there was lung cancer.

Okay, that's very clean.