Characteristics of the sub-cohort of the EPIC-InterAct project by categories of tea consumption (n = 15,227).

Values are expressed as Mean (Standard Deviation), Median (p25–p75), or percentage.1Based on n = 13,345, because information about hyperlipidemia was not collected in one center of Sweden.2Based on n = 8,802, because information about family history of diabetes was not collected in Italy, Spain, one center of Germany, and one center of the United Kingdom.3Based on n = 14,262, because information about a history of stroke was not collected in one center of Sweden.4Based on n = 10,168, because information about a history of angina was not collected in Sweden, The Netherlands, and one center of Germany.