@ChristofferBN@nicknorwitz Agreed that MR provides a clean natural experiment! The issue isn't whether MR is randomized, but whether a 60-year genetic exposure model can be used to numerically validate a 7-year drug trial curve. If duration makes them non-comparable Got to run out, but thanks for engaging!
@ChristofferBN@nicknorwitz The results from short-term randomized trials are compatible with the very strong associations between LDL-C and cardiovascular events seen in Mendelian randomization studies."
@ChristofferBN@nicknorwitz If taking a statin at age 50 isn't biologically equivalent to 60 years of genetic expression, then using Mendelian Randomization as the benchmark to validate 7-year drug trials is an apples-to-oranges comparison. If they aren't comparable, why use MR to justify the model?
@ChristofferBN Or, if it really is just serum LDL-C driving the mortality reduction, why wouldn't the narrowing cholesterol gap lead to a narrowing benefit gap over time?
@ChristofferBN If LDL-C levels between groups converge over time while the treatment group's relative benefit keeps growing, doesn't that point to non-cholesterol mechanisms (like statin pleiotropy or anti inflammatory effects) driving long-term protection?