Caffeine stimulates insulin secretion of pancreatic beta cells. In this study, Tuomilehto and colleagues studied the relationship between coffee consumption and the onset of diabetes mellitus in a Finnish population, which has the highest percapita coffee consumption in the world.

Using three baseline surveys taken in different regions of Finland, the study identified 6,974 men and 7,655 women between 35 and 64 years of age who did not have diabetes at baseline. Exclusion criteria included a diagnosis of coronary heart disease or stroke and known diabetes at baseline. Participants completed a questionnaire with information on medical history, demographics, and lifestyle factors concerning physical activity, smoking habits, and alcohol, coffee, and tea consumption. Coffee consumption was categorized as: zero to 2 cups, 3 to 4 cups, 5 to 6 cups, 7 to 9 cups, or 10 or more cups daily. Incident cases of diabetes were found through a national registry.

After 12 years, 381 incident cases of diabetes were identified. After adjusting for age and study year, coffee consumption was associated with higher body mass index and cigarette smoking, and lower blood pressure, education level, light occupational physical activity, tea consumption, and alcohol use. The authors found an inverse association between diabetes and coffee consumption. The findings remained significant in women and were observed as a trend in men after adjusting for the above factors, with a significant reduction in diabetes risk in those who drank at least 10 cups of coffee per day.

By combining data for men and women, the sex-adjusted, multivariate hazard ratios were 1.00, 0.76, 0.54, 0.55, and 0.39 for each of the five coffee consumption categories, respectively. Of note, although filtered and pot-boiled coffee were inversely associated with diabetes incidence, filtered coffee was more strongly associated than pot-boiled coffee.

This study found an inverse and graded association between coffee consumption and type 2 diabetes. Although several mechanisms are postulated, a clear understanding of the way coffee limits diabetes risk is unknown.