[fastball]

Can we stop publishing diet studies that use self-reported diet journals/questionnaires as the method for controlling variables? They are completely unreliable and invalidate anything they touch.

In this particular study it is even worse, because there is very likely to be confounding reporting related to what is being tested, i.e. the type of person who specifically seeks out diet soda is probably going to report their other intake in a different way to the person who YOLOs on sugar.