Did they control for all of the comorbidities associated with non-drinkers?

A lot of people who don't drink are not abstaining for some health augmentation or lifestyle choice, but because drinking any amount of alcohol would send them into a very early grave.

It seems like these hidden confounding factors are more plausible of an explanation that the "0 drinks per day" group has higher morbidity, compared to "alcohol is healthy in moderate amounts" explanation for the bottom of the J curve.