Humans are not laboratory mice, whose lives can be styled to best eliminate confounding factors from studies aiming to demonstrate a nice, neat correlation between cause A and result B. So when you study humans for decades, living their complicated and individualistic lives, you get all the confounding factors you can handle and then some. Take studies of fat, for example, and the strong correlation between extra body fat carried over the years, age-related illness, and a shorter life : Ame