Résumé
ABSTRACT BACKGROUND: The largest radiological accident to occur in any urban area happened in Goiânia, Brazil, in 1987. OBJECTIVE: To evaluate the association between breast cancer incidence and ionizing radiation levels. DESIGN AND SETTING: Ecological study among residents of the city of Goiânia, Brazil. METHODS: The central region of Goiânia, with seven major sources of contamination from cesium-137, was defined as the study area. The addresses of women diagnosed with breast cancer were identified between 2001 and 2010. The data were geographically referenced and, using census data, the annual averages of crude incidence rates were estimated. The existence of clusters of new cases was ascertained by means of the Moran index. Correlations of radiometric measurements with the incidence were assessed using unconditional linear regression. RESULTS: A total of 4,105 new cases were identified, of which 2,233 were in the study area, and of these, 1,286 (57.59%) were georeferenced. The gross rates of total and referenced cases were 102.91 and 71.86/100,000 women, respectively. These were close to the average for Brazilian state capitals, which is 79.37/100,000 women. The cluster analysis showed slight correlations in three small sets of census tracts, but these were far from the sources of contamination. The scatter plot of points and the R2 value close to zero indicated that there was no association between the variables. CONCLUSION: This study reinforces the hypothesis that the ionizing radiation levels to which women living in Goiânia are now exposed to are not associated with the onset of new cases of breast cancer.