Cancer Epidemiology, Biomarkers and Prevention 4(1): 21-28, 1995.
Risch HA, Howe GR
It has been suggested that usage of menopausal estrogens may be associated with risk of colorectal cancer. This association was examined in a record linkage cohort study using the Saskatchewan Health Plan Databases. All women ages 43-49 who were residents in Saskatchewan in 1976 were identified from the Saskatchewan Health master registration file. These 33,003 women were linked by registration beneficiary number to the Prescription Drug Plan Database for the period January 1976 through June 1987, and to the Provincial Cancer Registry Database for the period March 1960 through December 1990. Thirty women in the cohort had colon or rectal cancer diagnosed before 1976 and were omitted from the analysis; exposures within 3.5 years of diagnosis or end of follow-up were also omitted. Between 1976 and 1990, 230 first primary colorectal cancer cases occurred. Women who took estrogens had nonsignificantly elevated risk of colon cancer in general (age-adjusted relative risk = 1.29; 95% confidence interval, 0.86-1.93) and of cancer of the distal colon (relative risk = 1.51; 95% confidence interval, 0.90-2.54). Women who took p.o. contraceptives during this follow-up period seemed to be at higher risk of cancer of the proximal colon (relative risk = 2.12; 95% confidence interval, 1.00-4.53), though this apparent association could very well have occurred by chance. No associations were seen between hormone use and risk of rectal cancer. This study provides little evidence that usage of menopausal hormones may be related to risk of colorectal cancer (either positively or negatively), though further follow-up of the cohort appears warranted.
Rheinische Friedrich- Wilhelms- Universität Bonn