Introduction Molecular, epidemiologic and clinical evidences suggest that there are two main breast cancer types: pre-menopausal and post-menopausal. The purpose of this study was to ascertain the differences between distribution and strengths of risk factors in pre-menopausal and post-menopausal breast cancer.

Methods A case control study was conducted. We included 1014 women with histologically confirmed, incident breast cancer and 1014 controls matched by age and province of residency. All information for the exposures was collected during face-to-face interviews. Logistic regression was performed to investigate associations of reproductive and anthropometric factors in the risk of pre-menopausal and post-menopausal breast cancer. We used Receiver Operating Characteristic (ROC) analysis to compare two models.

Results Longer duration of breastfeeding, positive family history, employment vs housekeeping, oral contraceptive (OC) usage and higher parity were shown significant association with pre-menopausal breast cancer. Higher education, employment vs housekeeping, later age at first marriage, OC usage, positive family history and higher BMI were significant association with post-menopausal breast cancer. The comparison of logistic models for pre-menopausal and post-menopausal breast cancer demonstrated that patterns of risk factors were significantly different. Area Under Curve (AUC) =0.7442 for women ≥50 years vs AUC=0.6635 for women <50 years, p<0.001.

Conclusion Established risk factors can predict post-menopausal breast cancer well, and unknown factors have a greater role in the prediction of pre-menopausal breast cancer.