It's pretty simple. The per-nation health data used to back up the "dietary cholesterol leads to arteriosclerosis” theory was cherry-picked (in the sense of only showing data of nations that fit the hypothesis). If one nation’s data doesn't fit, call it a 'paradox'. Personally, I don't consider it so much a paradox as a sign of scientific dishonesty. Sorry to sound grumpy about this, but IMO Keyes did a lot of damage.

All that said, I'd look more at grass-fed dairy leading to less K2 deficiency. (Clot blocking with warfarin/coumadin leads to a higher incidence of arteriosclerosis — the target is K1, but it also blocks K2 which is needed to direct serum calcium to osteoblasts / osteocalcin). Nutritionists were aware that there was some kind of ‘X factor’ coming from some foods (notably dairy from cows eating bright green grass) that did this, but it was only isolated in the last decade or so. Now we know it’s in a lot of other foods.