Background Family history of colorectal cancer has been consistently associ
ated with an increased personal risk of this disease. Since evidence sugges
ts that hormones are related to colon cancer risk in women, the effect of f
amily history on large bowel incidence may be modified according to endogen
ous and exogenous hormone levels.
Methods We analysed data from a population-based case-control study of fema
le colorectal cancer to evaluate family history and cancer risk. Cases (n =
702) were female residents of Wisconsin with a new diagnosis of colorectal
cancer, identified through a statewide tumour registry. Controls (n = 2274
) were randomly selected from lists of licensed drivers and from rosters of
Medicare beneficiaries. All relative risks (RR) were adjusted for age, bod
y mass index, smoking and alcohol history, education, and use of hormone re
placement therapy.
Results Compared with women who reported no history of cancer in a first de
gree relative, women with a family history had an RR of 2.07 (95% confidenc
e interval [CI] : 1.60-2.68). Regardless of which parent was affected, risk
s were increased about twofold, while sibling history was associated with a
bout a 50% increase in risk. Risk was greater if more than one family membe
r was affected (RR 3.65, 95% CI: 1.81-7.37). The association between family
history and risk was stronger for colon cancer than for rectal cancer. The
re were no indications that exogenous hormonal factors, notably hormone rep
lacement use, modified these risks. There was a suggestion that high parity
attenuated the risks associated with family history (P = 0.07).
Conclusions These results confirm that family history of colorectal cancer
is associated with a doubling of risk for large bowel cancer in women; some
histories were associated with greater risk. This relation was not substan
tially different among subgroups of women with varying exogenous and endoge
nous hormone exposures.