There are genetic variants associated with coffee consumption[1] so you're right, in a sense the groups are self selecting. It still could be the coffee and tea

There are genetic variants associated with coffee consumption[1] so you're right, in a sense the groups are self selecting. It still could be the coffee and tea though, this kind of study could be the first step to looking deeper and finding out. [1] https://www.hsph.harvard.edu/news/press-releases/genetic-variants-linked-with-coffee-drinking/ https://www.hsph.harvard.edu/news/press-releases/genetic-var...