Our combined model accounted for just shy of 85% of the risk of incident MI, but factors like diet and exercise underperformed.
We wonder if healthy participant bias, or self report were to blame here?
In a cohort with more lifestyle phenotyping would this improve?
Criticism welcome
