The idea that research funds have to go to the individual to have a distorting influence is sheer nonsense. One, making sure that research funds go to people likely to give the sorts of answers that you desire--or people likely to do the sort of research that's likely to give the sorts of answers you desire--is of course distorting. Also, claiming that scientists aren't motivated by things they see beneficial to their research might actually be a bit insulting to the scientists. They aren't just motivated by personal financial gain, there's also actually wanting answers to questions they think need answering, the acclaim that comes if you're actually involved in producing these answers, etc.

Collins....

https://www.thelancet.com/journals/...1357-5/fulltext

This has that 1000 fewer events in 10000 statin takers with a history of heart disease that's mentioned here--as well as the claim for 500 fewer events in 10000 people at risk but who hadn't had a previous event. A number I've seen popping around is 1 in 200 to treat to benefit one person. Some digging around to do...

https://www.thelancet.com/journals/...0367-5/fulltext

Okay, here's the actual meta study the claims are made from.

I don't really have an answer I'm happy with about whether I should be worried about my ldl cholesterol. Fortunately, social anxiety keeps me out of the doctor's office, so I have no idea what my ldl cholesterol might be.

But I do think "lowering ldl cholesterol lowers heart disease, and eating butter can raise ldl cholesterol, so not eating butter lowers heart disease" is a questionable logical jump. If somebody "reverses" or puts their diabetes into remission, or loses a bunch of weight, etc. on a low carb diet that happens to include more butter than their previous diet--can you really say none of that matters, if ldl doesn't go down? You can't even predict that eating butter in this way will raise ldl cholesterol. Dr Westman has said that it's roughly 1/3 go up, 1/3 down, 1/3 pretty much stay the same when they switch to his diet.

Collins study--lowering ldl cholesterol by 1 mM, 1000 less events in 10000 people with previous events.

https://academic.oup.com/ajcn/article/102/2/309/4564657

Random study looking at butter versus olive oil that I have no reason to believe is representative of butter's effect on ldl cholesterol generally in studies like this one--other than my past nodding off reading studies like this one suggesting I haven't seen much to get excited about before.

Baseline, 2.88 mM/l, butter 3.04 mM/l, olive oil 2.87 mM/l. Now, comparing to corn oil would give a greater decrease in ldl cholesterol. But suppose that 1 mM decrease gives a 10 percent decrease in events held true. What would we expect if somebody at 3.04 mM went down to 2.87 mM? Suppose a pure, linear, dose response. Between olive oil and butter, 0.17 mM. We're down to 170 less events per 10000, in 5 years. Not insignificant, still. But not significant enough to offset benefits of a low carb diet, if somebody is insulin resistant/diabetic/pre-diabetic.

You could do low carb with olive oil, avocado etc., and I don't think there's anything wrong with that. But it's not perfect--one thing that's good about a ketogenic diet for me is that it becomes my habitual diet, butter and heavy cream is pretty much what I want to eat, in a way that olive oil and avocado oil just is not. I can eat them, I just don't come to prefer them.

It's sort of like the salt thing--high blood pressure is risky, absolutely. Lowering salt intake can lower blood pressure, yes. So the best diet has the least salt. Well, okay, but how much does lowering salt decrease blood pressure? Does it take you from severe hypertension to normal, or optimal? Generally you can hope for a marginal decrease, another situation where it takes tens of thousands of people before you can detect a benefit. I can detect the benefit of Atkins or a ketogenic diet in one person, me. One Shakespeare vs. thousands of monkeys typing on keyboards, to borrow an evolution denialist meme. Us denialists have to stick together. Except for that one type of denialist, I don't want to invoke Godwin's law so I'll be oblique about that. They suck.