Oral contraceptive and breast cancer: a case-control study

• Published in: Revista de saúde pública 2001-02, Vol.35 (1), p.32
• Main Authors: Tessaro, S, Béria, J U, Tomasi, E, Barros, A J
• Format: Article
• Language: Portuguese
• Subjects: Adult; Age of Onset; Brazil - epidemiology; Brazil, Pelotas; Breast Neoplasms - chemically induced; Breast Neoplasms - epidemiology; Case-Control Studies; contraceptives; Contraceptives, Oral - adverse effects; Female; Humans; Middle Aged; Odds Ratio; Time Factors
• ISSN: 0034-8910
• DOI: 10.1590/S0034-89102001000100005

To investigate the association between breast cancer and the duration of use of oral contraceptives (OC), and age it started to be used in a population of Pelotas, Southern Brazil. There were identified 250 incident cases of breast cancer in patients aged 20 to 60 years from records of pathology laboratories and there were enrolled 1,020 controls drawn from hospital and neighbourhood population. For 90 cases identified in Pelotas, 270 hospital controls and 270 neighbourhood controls were selected, for another 78 cases in Pelotas, 234 controls were selected, and for 82 cases from other municipalities, 246 hospital controls were selected. Controls were matched by age. Adjusted analysis was performed using conditional logistic regression. No association between oral contraceptive use and breast cancer was found (OR=1.1;CI95% 0.7 - 1.6 for hospital controls, and OR=0.9;CI95% 0.6 - 1.6 for neighbourhood controls) neither for different duration of use or starting age. To increase the test power, 250 cases and all 1020 controls were analyzed together, and an odds ratio of 1.6 (CI95% 1.0 - 2.4) was found for women older than 45 years of age who had been using oral contraceptives for five years or more. No evidence was found of a general association between oral contraceptive use and breast cancer. When analyzing the whole date set, with all neighbourhood and hospital controls together, for women older than 45 years of age who had been using oral contraceptives for more than 5 years, it was found an increased risk almost statistically significant (p=0.05).