Dr Butt said: "Several previous studies have investigated the association between breast-feeding and breast cancer risk, but, to our knowledge, no studies have investigated breast-feeding and risk associated with different types of breast cancer. Furthermore, no study has investigated the association between breastfeeding, types of breast cancer and survival yet.

"Our findings need be followed by studies on survival to see if these more aggressive breast tumours actually lead to a higher death rate or not, because we do know that breast cancers that do not have aggressive characteristics can also have high rates of mortality if they are diagnosed late. This is something that we intend to study next."

Dr Butt and her colleagues examined data collected prospectively from a group of 17,035 women in The Malm?¶ Diet and Cancer Study. They evaluated 622 cases of breast cancer for a range of factors that indicated how aggressive the tumours were (e.g. invasiveness, tumour size, axillary lymph node status, HER2 status, Ki67, which is an indicator for tumour proliferation, etc). They analysed the duration of breast feeding for each child, total amount of time a woman had breast-fed, and the average time of breast-feeding per child; the average duration of breast feeding was divided into four groups: less than 2.2 months, less than four months, four months or more, and 6.2 months or more.

Dr Butt said: "We found a statistically significant risk of grade III tumours in women with an average time of breast-feeding of 6.2 months or more. The risk of tumours expressing higher levels of Ki67 was also significantly associated with longer duration of breast-feeding. We concluded that long duration of breast-feeding was associated with more unfavourable types of breast cancer."

She stressed that these findings should not discourage women from breast-feeding as there were several strong studies that showed that breast-feeding could reduce a woman's overall risk of breast cancer, and that longer breast-feeding times were good for both mother and baby.

"The most important thing would be to identify women with a higher risk of aggressive types of breast cancer and offer them intensified screening, in order to identify their tumours early."

She said the study was an epidemiological one that could show risk associations but not causes. "The biological mechanisms behind this are still to be identified. What is known is that breast-feeding reduces the number of ovulatory menstrual cycles over a lifetime, thereby reducing the impact of hormone levels present during normal menstrual cycles and, in particular, reducing the progesterone exposure. This may explain the finding in previous studies of a reduced risk of breast cancer in women who had breast-fed. However, breast-feeding stimulates the production of prolactin, a hormone that has been reported to have tumour-promoting effects. But the relation between breastfeeding, prolactin and breast cancer is complex and not fully understood."

Source The University of Western Australia