Maybe many of those who consumed more eggs or cholesterol simply just ate more.
If the study simply tracked calories, maybe calories would be found to be correlated with CVD risk.
Problem with that, you can't really track what some people eat for 30 years.