Risk factors and 10-year breast cancer survival in northern Alberta.

1,121 women in northern Alberta diagnosed as having breast cancer between 1971-74 were followed for 10 years. Risk factors for breast cancer were studied with respect to their possible influence on survival by comparing survival curves, using both Logrank and Cox's regression model, and controlling for intercurrent death, stage and axillary node status. A complex interaction was found between age and menopausal status and survival rates. Premenopausal women aged 45-55 had a better survival rate than postmenopausal women of the same age. However, one subgroup of premenopausal women aged 35-39 had a significantly worse prognosis than those aged 40-44 as did a group of post menopausal women aged 70-74. No cause was found other than the effect of age. There was a significant trend to worsening survival with heavier weight at time of diagnosis and with breast feeding. Parity of five or more and family history of breast cancer were less consistently associated with worse survival. Oral contraceptive use was only associated with worsened prognosis significantly when stage was controlled for; there was no overall effect. Age at menarche and age at first birth did not influence prognosis. Theories to explain the findings are discussed.