Age at first birth, dietary practices and breast cancer mortality in various Italian regions.

Breast cancer death certification rates in Italy showed a consistent geographic variation, with approximately a 2-fold ratio between the highest rates registered in Northern (and richer) regions, and the lowest ones in the South. This variation in breast cancer mortality was positively and strongly correlated with mean age at first birth (r = + 0.73). This positive correlation could not be totally explained by several other economic or dietary variables considered, though some of them substantially reduced the strength of the association. Breast cancer mortality rates were found to be positively correlated with milk, meat and sugar consumption, and negatively with pasta, thus confirming previous studies on national or international scales. These correlation coefficients, however, were considerably reduced after controlling for mean age at first birth. Further, when allowance was made for economic variables and age at first birth, only the correlation coefficients for milk and cheese remained significantly positive. Thus, the evidence from this study suggests that the Italian geographical correlations between breast cancer mortality and dietary variables may be largely explained in terms of reproductive factors, though there may be some effect of dietary variables on breast cancer risk as well.