Studies with high credence value for making definitive health advice must involve large numbers of people, span decades, and examine hard endpoints like mortality.

“for a study to have a high credence value um or to say something more definitively to give people advice that's going to they can hold dear their heart they know it's true we have to have some long-term studies that corroborate short-term studies and investigate all the data to see where the overwhelming amount of evidence lies and i could say with clarity that we look at studies with high credence value which means three things one when we're talking about diseases and causes of death we have to study large numbers of people so we have to have many thousands of people in the study not fifty or a hundred but ten thousand twenty thousand hundred thousand people number two the study has to go on for decades not years and number three we have to look at hard end points which means mortality cardiovascular mortality or cancer mortality will all cause mortality and age of death”