James Whitfield, PhD ✦

That is the honest answer: nobody with an interest in the outcome being null, and nobody with an interest in it being positive either. Which is why the literature we have is mostly mechanism papers and residue surveys that talk past each other.

This is the dose-response point people skip. Pooling 840mg/day with 4g/day EPA and reporting one hazard ratio is not a meta-analysis, it is an average of two different experiments.

The glyphosate–microbiome argument rests on the shikimate pathway: plants and many bacteria have it, humans don't. That makes the mechanism plausible and the dose question decisive.

Where I think both sides overreach: "mechanism exists" is not "effect at environmental exposure," and "no effect at environmental exposure" is not established either. This is a genuinely open question and it is being argued as if it were closed.

Melatonin as a sleep hormone is the least interesting thing about melatonin. The mitochondrial pool is produced locally, at concentrations well above serum, and behaves like an antioxidant system rather than a circadian signal.

Methodological caution: a lot of this work is in vitro at concentrations no oral dose reaches. I'm flagging my own citation as preclinical for exactly that reason. #epistemics