Abstract

OBJECTIVE: The Women's Health Initiative (WHI) study of postmenopausal hormone therapy (HT) found that estrogen plus progestogen therapy (EPT) decreased colorectal cancer risk. Thus, the decline in EPT use from 2002 to 2003 should have precipitated an increase in the incidence of colorectal cancer. We tested this prediction using the SEER 9 epidemiologic database. METHODS: We analyzed WHI data concerning the effects of EPT and estrogen therapy (ET) on colorectal cancer risks. We also examined HT prescription sales data, as well as SEER 9 colorectal cancer incidences from 2001 to 2004. RESULTS: In the WHI study, the incidence of colorectal cancer was comparable in EPT placebo-users, ET users, and ET placebo-users, but significantly lower in EPT users. Assuming that 30% of eligible women used HT in 2001, the decline in EPT sales from 2002 to 2003 of 63% should have increased the incidence of colorectal cancer by 2.8% in the overall population at risk. However, the SEER 9 colorectal cancer incidence fell by 5.9% in this population, which is comparable to the 6.7% decrease observed for invasive breast cancer from 2002 to 2003. CONCLUSIONS: Declining EPT use from 2002 to 2003 should have precipitated an increase in the incidence of colorectal cancer, but the opposite trend was seen in the SEER 9 database during this time. The incidences of invasive breast cancer and colorectal cancer both declined by a similar amount from 2002 to 2003, despite the results of the WHI study predicting opposing trends for the two different types of cancer. Thus, the SEER 9 findings are fundamentally incompatible with expectations from the WHI findings. This implies that reductions in HT use from 2002 to 2003 cannot account for the contemporaneous changes in invasive breast cancer and colorectal cancer incidences. Alternative explanations must be found.