The "paleolithic" diet

The idea that a diet containing fewer evolutionary novel foods (esp. grains and dairy) might be more healthy in some regards, always has been a plausible working hypothesis. A hypothesis. And perhaps a decent precautionary principle. So the "paleo hypothesis" always needed to be confirmed by interventional trials and epidemiology.

However, only few of its predictions have turned out true or useful, but attribution is difficult. Is the useful, but sometimes wholly inaccurate, rule of thumb "don't eat too much processed foods" an achievement of the paleolithic hypothesis? The putative dairy-IGF1-mTOR-acne/disease link? Does the latter even matter for primary prevention if we consider the benefits of low fat dairy? (my speculative answer: not at the population level, but preferring yogurt and not overdoing it with dairy might be prudent.)

I want to highlight some issues with the paleo-diet, based on recent articles I have read:

First, we do not know what our ancestors, early h. sapiens and before that, actually ate in much detail. Current evidence suggests that most of our ancestors were indeed omnivores, certainly never pure carnivores. This contradicts your typical Atkins-esque claim of a diet very high in meat, or an ultra-high fat diet.

Second, also evidence from archeology, a recent paper found atherosclerosis in mummies of both hunter-gatherers and farmers (n=137). This has been discussed on

...a study published a week ago online in The Lancet by Prof. Randall C. Thompson of Saint Luke’s Mid America Heart Institute and an international team of investigators entitled Atherosclerosis across 4000 years of human history: the Horus study of four ancient populations...
As is noted in Thompson’s article, ancient Egyptians and Peruvians were agricultural cultures with farms and domesticated animals, Ancestral Puebloans were forager-farmers, and the Unangans were hunter-gatherers without agriculture. Indeed, the Peruvians and Ancestral Puebloans predated the written word and were thus prehistoric cultures.

Third, hard evidence. I have written about vegetarianism before. It's solid. I also reviewed the data on whole grains and found little evidence of harm, certainly some benefit, particularly with non-rice, non-wheat whole-grains. The diet best supported by controlled trials (RCTs) and other data remains the "Mediterranean" diet.

Obviously, this only applies most of the time (primary prevention for the general population). Grains might be less healthy than legumes and oat bran better than whole grains, for instance. If you are gluten intolerant or want to try an experimental diet as an additional treatment for acne, you will deviate from general recommendations. Etc.

Alas, no time to get into the details.