Best evidence is that high cholesterol is indeed associated with an increase in coronary disease, inc. death. The confounding factor in unselected population studies is that several cancers are known to LOWER cholesterol levels, so those with low cholesterol at initial testing who do NOT have cancer do indeed tend to live longer with less heart issues.
Low Serum Cholesterol and Mortality
FWIW- Here's link to the horse's mouth on Framingham-
Framingham Heart Study
And here's their calculator developed using Framingham data-
Heart Failure Framingham Heart Study
Plug in normal systolic BP (120) and normal BMI (23), then play with different cholesterol & HDL numbers. Framingham predicts a 65yr old with bad cholesterol profile (250 total/40 HDL) has almost double the 30 year risk of cardiac events (60 %) vs a good profile (170 total/ 65 HDL @ 33%).