Separate meta-analyses based on case-control and cohort studies have reported different results on the relationship between
coffee consumption and
colorectal cancer risk. To clarify the effect of
coffee intake on
colorectal cancer risk, we performed a meta-analysis based on both case-control and cohort studies.
DESIGN: Review study.
SETTING: We identified case-control and cohort studies related to
coffee consumption and
colorectal cancer risk listed on MEDLINE, the Cochrane Controlled Trials Register, EMBASE, Science Citation Index and PubMed (until May 2011).
SUBJECTS: Twenty-five case-control (15 522 cases) and sixteen cohort studies (10 443 cases) were included in the meta-analysis. Comparing the highest v. the lowest/non category of
coffee consumption, the combined results from case-control studies showed a significant relationship with
colorectal cancer (OR = 0·85, 95 % CI 0·75, 0·97) and
colon cancer (OR = 0·79, 95 % CI 0·67, 0·95), but not
rectal cancer (OR = 0·95, 95 % CI 0·79, 1·15). For cohort studies, there was a slight suggestion of an inverse association with
colorectal cancer (relative ratio = 0·94; 95 % CI 0·88, 1·01) and
colon cancer (OR = 0·93, 95 % CI 0·86, 1·01), rather than
rectal cancer (OR = 0·98, 95 % CI 0·88, 1·09). In subgroup analyses using case-control studies, significant inverse associations were found in females for
colorectal cancer and in Europe for colorectal and
colon cancer, while the subgroup analyses of cohort studies found that
coffee drinks substantially decreased risk of
colon cancer only in Asian women.
CONCLUSIONS: