This article was published in the Annals of Internal Medicine: The Relationship between Green Tea and Total Caffeine Intake and Risk for Self-Reported Type 2 Diabetes among Japanese Adults.
The conclusion looks good but a retrospective cohort study is not the best study design.