Breast cancer risk in healthy and symptomatic women: results of a multivariate analysis. A case-control study

Several risk factors for breast cancer (BC) have been investigated in different reports, but none has been really useful in preventing BC development. The aim of this study was to evaluate the risk of BC in self-selected symptomatic women in comparison with the healthy population residing in an urban area of Italy. A group of 404 women (median age 59 years, range 26-89 years) with confirmed BC (cases) were age-matched with 389 healthy women (Group A), and 391 (Group B) symptomatic non-screened patients without BC, who were referred to our Breast unit. The results of univariate analysis showed a significant (P < 0.01) difference between cases and controls in (1) age at menarche, (2) number of birth and age at first births, (3) lactation and months of lactation, and (4) estrogen replacement therapy (ERT) and duration of ERT. Multivariate analysis using a logistic regression model adjusted for age showed that five independent parameters (no pregnancy, age at first birth > 30 years, no lactation, use of ERT, ERT > 40 months) significantly (P < 0.01) correlated with BC onset. The relative odds ratios (ORs) at 95% confidence interval (95% CI) were 5.25, 2.47, 2.82, 2.80, and 5.56, respectively. The cumulative OR (95% CI) calculated from the observed vs. predicted values was 7.15. No differences (P = NS) were found between groups A and B. In conclusion, in our study population, the prolonged use (> 40 months) of ETR in menopausal women resulted in an increased risk of BC, and represented the only risk factor that could be removed.