The MOST interesting finding is how linoleic acid (omega 6) is surprisingly protective against BOTH dementia and diabetes. Linoleic acid seems to have an opposite sign to broader omega-6 though?
[also MUFA has the wrong “sign” and mgiht be “slightly bad”?]
Cholesterol is dissected into SO many subtypes - only a tiny fraction of the types really contribute to disease. OVERALL cholesterol ON SUM seems very mildly “good” which is consistent with DNAm age data (and everything else) I’ve seen (eg mortality in late life people).
The most surprising finding is that ketone bodies/BHB are “slightly associated with bad outcomes” (i can see how they’re generated in diabetic ketosis)
Linoleic acid might correlate negatively with AA because higher levels MIGHT mean less is being converted into AA.
Wkipedia says AA has its benefits and is not clearly bad either.
FWIW, mainstream science (eg Harvard nutrition) is more on the side of PUFAs being "good’ than “bad”, but the ““really bad”” tail risk of PUFA is way worse than MUFA which is why I would still choose MUFAs [also we know highly MUFA diets are fine, we don’t have the same ancestral data for PUFAs]
Thanks
I have often thought that this seed oil thing is overblown. Typically I would be spraying less than 1 gram into the bottom of a frying pan just to keep the food from sticking. Say it was 1 gram, how much of that would be transferred to my steak or whatever?
Why don’t people just use olive oil? Because olive oil is not a neutral-tasting oil and other seed oils such as canola are. I use olive oil when it is appropriate to the recipe.
plug these diagrams/heatmaps all into NotebookLM or Claude and get it to interpret all the weird TG/phospholipid data for you on top of your iollo data
Correlations are problematic when not age-stratified (eg lower cholesterol still seems better for younger/CR’d individuals, the associations are the opposite for older people)
