Methods

Overall Study Design

Incident cases aged 20-45 years are being recruited from major cancer hospitals in seven large Latin American cities (Barranquilla, Bogota, Cali, Medellin, Mexico City, San Jose and Santiago del Chile). These centres have been selected based on their previous involvement in breast cancer research and their long-term commitment to this protocol. Controls for the study are selected from the population that have resided in the same city for at least 3 years and matched to cases by age (± 3 years) and health-care institution. Population-based controls are selected within the district of the case using a multilevel sampling frame.