https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1768013/
Correlational studies are inherently flawed when you have something as complex as diet. It's probably quite true that in a random sample of 80,000 health care workers in America, the people with more grease in their diets drop dead faster. It doesn't happen in France, Spain etc etc. Meaning it probably has nothing to do with saturated fatty acids, and probably much to do with whatever goes along with it. Who knows what it actually is? Soybean oil, wonder bread, dunkin doughnuts, HFCS, diet coke, being disgustingly overweight, never walking anywhere, granola, fruit juice (bleh -sugar water), dietary iron, twinkies, nacho "cheese," brined ham, trail mix, marshmallows, potato chips, microwave popcorn, soy protein -any number of things the Frogs and Spaniards don't eat or do, but fat Americans do. Using linear regression, grease/saturated fat probably correlates well with all this other garbage, and with dropping dead sooner, assuming you censor out other world populations with different national diet tendencies. Of course including Spaniards in the data set, you get a different answer.