Ji 1997.

Study characteristics
Methods	PCC in China
Participants	Participants: 2266 (male/female: NR) cases and 1552 (male/female: NR) controls Inclusion criteria: aged 30/74 years, permanent resident of the 10 urban districts of Shanghai, China Recruitment: from October 1990‐June 1993
Interventions	N/A
Outcomes	Number of cases Colon cancer: 885 (male/female: 426/459) out of the 931 eligible cases Rectal cancer: 843 (male/female: 441/402) out of the 884 eligible cases Pancreatic cancer: 428 (male/female: 246/182) out of the 451 eligible cases
Green tea in exposure categories	Exposure assessment A: green tea drinking habit Lowest exposure: nondrinker Highest exposure: regular drinker Exposure assessment B: green tea consumption Men Lowest exposure: nondrinker Intermediate exposure 1: 1‐199 g/month Intermediate exposure 2: 200‐299 g/month Highest exposure: ≥ 300 g/month Women Lowest exposure: nondrinker Intermediate exposure: 1‐199 g/month Highest exposure: ≥ 200 g/month Exposure assessment C: lifetime green tea consumption Men Lowest exposure: nondrinker Intermediate exposure 1: 1‐3499 g/month per years of drinking Intermediate exposure 2: 3500‐8499 g/month per years of drinking Highest exposure: ≥ 8500 g/month per years of drinking Women Lowest exposure: nondrinker Intermediate exposure: 1‐3499 g/month per years of drinking Highest exposure: ≥ 3500 g/month per years of drinking
Notes	Funding: not declared Statistical methods: logistic regression Variables controlled in analysis: age, income, education and cigarette smoking Variables controlled by matching: sex and age (± 5 years)