This article doesn’t tell us about potential selection bias issues or confounding variables. For instance, do people who don’t drink any coffee tend to be peop

This article doesn’t tell us about potential selection bias issues or confounding variables. For instance, do people who don’t drink any coffee tend to be people who have been advised not to by their doctors due to cardiovascular disease? That might account for their higher mortality than morning coffee drinkers. Similarly, are afternoon/evening coffee drinkers more likely to drink coffee late because they have to stay up late or work the graveyard shift — which has been associated with higher disease and mortality for other reasons?