Lactation and breast carcinoma risk in a South African population

BACKGROUND. A number of epidemiologic studies have reported a reduced risk of breast carcinoma among women who have lactated but others have not. The current study presents data regarding lactation and breast carcinoma risk f rom a hospital-based case-control study of black and colored South African women. METHODS, incident breast carcinoma cases treated between January 1994 and O ctober 1997 (n = 446) at 2 major hospitals in Cape Town and hospital patien ts admitted for conditions unrelated to breast carcinoma (controls, n = 147 1) were queried regarding the duration of breast-feeding each liveborn chil d and breast carcinoma risk factors. Multivariate logistic regression model s were used to calculate odds ratios (ORs) for various categories of lactat ion compared with a reference category of never having breast-fed among wom en who had had at least one full term live birth. RESULTS, Approximately 83% of cases and 85% of controls had ever breast-fed (OR = 0.9; 95% confidence interval 195% Oil, 0.7-1.3). Among all subjects, the ORs for those who lactated for <3 years were near or at unity. Beyond 3 years, ORs extending up to greater than or equal to 7 years were less tha n unity, bur the 95% CIs included 1.0 (OR for duration of greater than or e qual to 7 years = 0.7; 95W CI, 0.4-1.3). ORs did not vary by menopausal sta tus. Breast carcinoma risk was not found to be related to the duration of b reast-feeding the first child, the number of children breast-fed, or the pa tient's age at first lactation. CONCLUSIONS. The results of the current study suggest lactation has little or no protective effect on breast carcinoma risk. (C) 1999 American Cancer Society.