Tea intake was assessed by questionnaire, once at the beginning of the study and then again five years later. Cancer incidence was assessed by notification from major local hospitals in the study area and data linkage with population-based cancer registries.
