A large scale study of all former trials of breast cancer risks with hormone replacement therapy shows a definite link between a risk of cancer and length of use. At the same time two cautionary notes have been sounded about the lack of evidence o

The Centers for Disease Control in Atlanta, which analyzed some 16 studies of the effect of HRT on breast cancer risk, concluded that the risk of cancer increased forall types of women using any type of HRT with every year of use.

The study claimed no risk for women using the drugs for less than five years except those who had ever used another form of oestrogen or were premenopausal.

Women using HRT who hadn't undergone menopause doubled the risk of developing breast cancer.

"Oestrogen use increased the relative risk of breast cancer irrespective of family history of breast cancer, parity or history of benign breast disease," said the study, although the harmfuleffects of HRT was enhanced among women with a family history of breast cancer or those who'd had first babies late.

Among women with a family history of breast cancer, those who had at any point used oestrogen replacement were three and a half times more at risk than those who had not.

Another recent study put the risk at 190 cases per

100, 000 women on the drug every year, or almost 1in 500.

Despite these risks, the CDC study attempted to argue, at least HRT reduces the risk of heart disease a conclusion that at least two doctors writing elsewhere say has no basis in anything but "meta analysis," where projections have come up with the figures supposedly showing that HRT halves the risk of heart disease.

"Oral oestrogens are drugs. Thus, a statement recommending hormone replacement as a method of heart disease prevention for postmenopausal women would seem to warrant a clinical trial, as required for other drugs," wrote Dr Elizabeth Barrett-Connor of the University of California San Diego. "Failing any randomized clinical trial, a large number of women in this and other developed countries may be prescribed extended oestrogen replacement to prevent cardiovascular disease, but without the supporting data usually required for the widespread use of any other drug recommended for disease prevention."

Or as Jan V. Vandenbroucke put it in the Lancet, "The responsibility of encouraging liberal use of hormone replacement therapy in tens of millions of women worldwide lies entirely with meta analysts of purely observational epidemiological research."

!AJAMA, 10 and 17 April, The Lancet, 6 April 1991.