Background and objectives: Recent observational studies have suggested a positive association of white rice and protective associations of green tea and coffee with the risk of diabetes. However, none have examined the interaction between these dietary factors on the risk of diabetes. We prospectively investigated the effect modification of green tea and coffee on the association between rice and incident diabetes in elderly Japanese men and women.
Methods and study design: Among subjects who participated in the baseline survey (2004-2007), 11717 (91 %) subjects responded to the follow-up survey (2010-2012). By using multiple logistic regression analysis, ORs of incident diabetes were calculated according to categories of cereal food, green tea, and coffee intakes, examining also the effect modification of green tea and coffee.
Results: 464 new cases of diabetes were identified. Women, but not men, showed a positive association of rice intake (trend p=0.008) and an inverse association of green tea intake (trend p=0.02) with incident diabetes. Coffee showed no association with incident diabetes either in men or women. In the analysis stratified by green tea intake, the association between rice and diabetes disappeared among women with an intake of >=7 cups/d of green tea (interaction p=0.08).
Conclusions: Rice intake was associated with an increased risk of diabetes only in women, and women with a higher intake of green tea had a lower risk of diabetes. A high intake of green tea may be protective against increased risk of diabetes with a higher intake of rice in women.