Hacker News new | past | comments | ask | show | jobs | submit login

They use statistical methods to eliminate the confounding variables. But given that they do not understand the fallibility of doing that on extremely noisy data, the conclusions are telling us nothing.

They cannot control anything and it's all just questionnaires.

It's fine too look at these things but they seem to get blown out of proportion every single time, even when industry funded studies come out saying eggs or bacon is fine.

And yes, 7% increased risk is practically nothing. That's baseline risk multiplied by 1.07. It's noise all the way.




> They use statistical methods to eliminate the confounding variables.

Some of those anyway. You can only account for confounders that you know about a priori.

And given they used Food Frequency Questionnaires, it's mostly wasted efforts imho.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: