Yes, largely junk. As I mentioned the literature is full of studies that are nothing but some regressions on top of self reported dietary data. It’s almost all very low quality[1].

[1] https://www.science.org/content/article/people-are-bad-repor...