It’s from his newsletter (which I think I signed up for ages before I subscribed to his site) so, I assume there’s no “paywall” beyond signing up to receive them.
Worth mentioning, but difficulties with assessing the impact of nutrition isn’t confined to cognitive health alone.
Like studying one chemical at a time in search of carcinogens. Very focused looking for specifically what happens. This misses that point. Most DNA chemical reactions that lead to cancer are a mix of hundreds of chemicals altering a lot of DNA.
We study chocolate, coffee, and alcohol more than anything else. A lot of good that does us.
The beginnings of more complex studies are in the offing.
We do not study gasoline because we are sort of assuming we do not consume it internally. That is not true. We do.
As to chemical carcinogens, i think the Ames test is a big improvement. Not perfect but an excellent screening test. Better than painting chemicals on the back of shaved rabbits.
Statistical analysis of nutrition and complex interactions make long term studies difficult. Chemicals often get discovered after someone notices a cluster. Then materials thought to be safe turn out to be toxic under some circumstances. Alcohol is one of those.