Coffee Consumption and the Incidence of Colorectal Cancer in Women

Table 2

Proportional Hazards analysis of the relationship between regular coffee consumption and incident colorectal cancer.

Cancer site	Level	Events	Ann rate	Multivariate adjusted^1
				HR (95% CI)	Trend; value^2
All colorectal	None	513	0.108	1.0 (ref)	0.04
	>0–<4 cups/day	634	0.127	1.15 (1.02, 1.29)
	4+ cups/day	135	0.129	1.14 (0.93, 1.38)
Colon	None	438	0.092	1.0 (ref)	0.07
	>0–<4 cups/day	527	0.106	1.11 (0.98, 1.27)
	4+ cups/day	118	0.113	1.17 (0.94, 1.44)
Rectum	None	60	0.013	1.0 (ref)	0.53
	>0–<4 cups/day	86	0.017	1.30 (0.93, 1.82)
	4+ cups/day	14	0.013	0.95 (0.52, 1.72)
Rectosigmoid	None	27	0.006	1.00 (ref)	0.40
	>0–<4 cups/day	33	0.007	1.15 (0.69, 1.94)
	4+ cups/day	9	0.009	1.39 (0.63, 3.07)

Adjusted for age, ethnicity, education, alcohol, smoking/pack-years, BMI, physical activity, energy intake, red meat intake, fruit/vegetable intake, percent calories from fat, fiber intake, calcium intake, hormone use, NSAID use, history of treated diabetes, and family history of colorectal cancer.
Trend value calculated from a separate model with the outcome of interest as a function of linear coffee level (None = 1, >0–4 = 2, ≥4 = 3).