90,735 Japanese adults (JPHC cohorts I and II; ages 40-59 and 40-69 respectively), followed 1990-2001. Egg consumption up to almost-daily was not associated with increased coronary heart disease incidence. Serum total cholesterol itself was strongly associated with CHD risk (multivariate HR 2.17, 95% CI 1.22-3.85, comparing ≥2400 vs <1800 mg/L), but this was not mediated by egg intake; frequent egg-eaters actually had somewhat lower hypercholesterolemia prevalence, suggesting dietary self-selection (people with high cholesterol avoiding eggs) rather than eggs driving cholesterol up in this population.
relevance_note: A large, clean null anchor for Japan that explicitly tests and rejects the egg→cholesterol→CHD mechanistic chain as an explanation in this population — useful for the mechanism-vs-population-heterogeneity question.