Background: A meta-analysis published in 2005 reported that the risk of developing diabetes mellitus was approximately two-thirds lower among persons who consume high levels of coffee compared with those who consume less. Since then, additional studies have suggested that tea and decaffeinated coffee may also provide a similar benefit. This suggests that ingredients other than caffeine may have a protective effect (e.g., lignans or chlorogenic acid, which have antioxidant properties). Huxley and colleagues updated the previous meta-analysis, and conducted an overview of the evidence for decaffeinated coffee and tea consumption on subsequent risk of developing diabetes.

The Study: The authors systematically reviewed prospective studies from 1966 to July 2009, and examined the association between new-onset type 2 diabetes and the intake of coffee, decaffeinated coffee, or tea (i.e., green, black, and oolong). Cross-sectional studies and studies that did not specify the number of cups of beverages consumed per day were excluded.

Results: Twenty studies including 517,325 participants were reviewed, with a median follow-up of two to 20 years. Most participants were white; 21 percent were Asian. After controlling for multiple variables, each cup of coffee consumed daily was associated with an approximately 7 percent reduction in the risk of developing diabetes. Overall, consuming three to four cups of coffee per day was associated with a 25 percent risk reduction compared with persons drinking two or fewer cups per day; however, a greater effect was reported in the smaller studies. Estimates using the six largest studies found a 15 percent reduction in diabetes risk at this level of intake.

Persons who drank more than three to four cups of decaffeinated coffee per day were less likely to develop diabetes than those who did not consume decaffeinated coffee (relative risk = 0.64). Three to four cups of tea per day also showed some protective effect (relative risk = 0.82).

Conclusion: The authors conclude that high intake of coffee, decaffeinated coffee, or tea is associated with substantial reductions in the risk of new-onset diabetes. The authors believe that a causal effect is supported by the presence of an apparent dose-response relationship. However, they caution that it is unknown whether their findings are applicable to nonwhite populations because of the predominantly white study population.