> That is condeming oxidised serum fats, not oxidised dietary fats, but it appears to be condemning "omega-6-rich vegetable oils" in the diet because they are more likely to oxidise in the body(?)
We can just bypass that entire speculation, because to get there we've already made a logical leap, namely assuming that oxLDL is even something we should care about. All the evidence suggests that the oxidation of LDL almost entirely occurs within the arterial wall (https://pmc.ncbi.nlm.nih.gov/articles/PMC2924812/). When LDL is retained, oxidation is practically inevitable. The important thing is number of ApoB tagged particles.
So what would we expect to see in the real world if this hypothesis (that oxLDL is not an independent risk factor) was the case? Well, presumably if we adjusted oxLDL for ApoB, we would see the risk association disappear, since all oxLDL would be doing is just proxying for "amount of LDL", and the oxidation itself is not driving the risk. What do we actually see when we do this? When we perform a mutual adjustment of oxLDL and ApoB, and the association of risk from oxLDL disappears: https://pubmed.ncbi.nlm.nih.gov/16949489/ So even if the speculation that PUFAs increase oxLDL count were valid (and I'm not granting it is), it would appear to be irrelevant as far as risk of CVD goes.
> your position boils down to "Science is only for experts, and only the experts I choose, not those they believe the wrong things so they're quacks, trust me" which is .. disagreeable.
That's not my position at all. Please try to employ the principle of charity when interpreting my position, or this is going to be very slow going, and likely unproductive.
My point is that when you look at mechanisms, you're peering at one tiny part of a process of thousands (or more) moving parts. Just stringing together a bunch of those mechanisms into a narrative doesn't provide any kind of sound basis for making an inference about human health outcomes because it ignores every other mechanism at play. For example: mechanistically, exercise induces oxidative stress. I could do as you did above and string it together into a theory about how exercise is bad for your health, but I hope you'd agree that I'd be making a false inference.
My point is not that we then throw our hands up and say "I guess I'll leave science to the experts". I never said that. I said we have to base our inferences off the summation of all the mechanisms at play.
How do we do that? By looking at human outcome data (prospective cohort studies, mendelian randomisation, RCTs). Why don't we think exercise is unhealthy, despite the oxidative stress mechanism? Because we have human outcome data that shows the benefits of exercise. That is, we look at the exposure we're interested in (exercise) and the outcome we actually care about (say, mortality, or cardiovascular disease). Do people who exercise more have lower mortality rates when potential confounders are accounted for? If so, we have good reason to believe exercise is beneficial. This is how we should look at replacing SFA with PUFA. When we do that, what happens to the things we actually care about (heart disease, mortality etc)? Overwhelmingly, we improve outcomes. Why should we prize mechanisms above human health outcomes? It doesn't make sense to me.
> When you say fat quality, do you mean PUFA vs SFA?
Yes.
> Let's be clear that I'm lost in the details of SFA, PUFA, linoleics
The position I take is: - There seems to be a point, around 10% total energy from saturated fats, where risk increases significantly. Either side of that, changes to SFA still make a difference, but not to the same degree. - When it comes to reducing SFA, replacement is important. Replacing SFA with PUFA seems to be the most beneficial approach.
In prospective cohort studies, each 5%E from SFA replaced with PUFA resulted in a 13% lower risk of CVD events, and a 26% reduction in coronary deaths. In RCTs, we see a 21% reduction in CVD events when substituting PUFA in for SFA (https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD...). In general the reduction in risk seems to come out at about 20-30% depending on study characteristics. Considering CVD is one of the top killers in the Western world, I think that's something worth paying attention to.