Background: The evidence on the relationship between coffee intake and cancer of the oral cavity and oesophagus is conflicting and few follow-up studies have been done.
Methods: A total of 389,624 men and women 40-45 years who participated in a national survey programme were followed with respect to cancer for an average of 14.4 years by linkage to the Cancer Registry of Norway. Coffee consumption at baseline was reported as a categorical variable (0 or <1 cup, 1-4, 5-8, 9+ cups per day).
Results: Altogether 450 squamous oral or oesophageal cancers were registered during follow-up. The adjusted hazard ratios with 1-4 cups per day as reference were 1.01 (95% confidence interval: 0.70, 1.47), 1.16 (0.93, 1.45) and 0.96 (0.71, 1.14) for 0 or <1 cup, 5-8 and 9+ cups per day, respectively. Stratification by sex, type of coffee, smoking status and dividing the end point into oral and oesophageal cancers gave heterogeneous and non-significant estimates.
Conclusion: This study does not support an inverse relationship between coffee intake and incidence of cancer in the mouth or oesophagus, but cannot exclude a weak inverse relationship.