Hormone replacement therapy, the treatment used by millions of post-menopausal women, may not be harmful after all.

HRT may actually reduce, rather than increase, the risk of heart disease in women who use it to prevent osteoporosis, say US researchers.

The therapy was at the centre of a health scare five years ago when an American study linked it to an increased risk of heart attacks, strokes and breast cancer.

At the time health chiefs advised HRT should be taken for as short a time as possible and in Britain it was no longer recommended as a drug to prevent bone loss in women with a family history of osteoporosis.

However, a study published yesterday in the Journal of the American Medical Association found no increased risk of heart attack for women in their 50s taking HRT.

The Women's Health Initiative Study discovered that any additional risks may apply only to older women.

Women in their 60s and 70s still experiencing the symptoms of the menopause such as hot flushes and night sweats were at an increased risk of heart attacks and strokes even if they were not taking HRT.

The study concludes: "Women who initiated hormone therapy closer to menopause tended to have reduced coronary heart disease risk compared with the increase in coronary heart disease risk among women more distant from menopause."

However, it said the trend "did not meet our criterion for statistical significance".

Dr John Stevenson, an HRT expert from London's Royal Brompton Hospital, speaking to the British Menopause Society, said the findings were a "U-turn of dramatic proportions".

He said: "We are astonished that a study which made such a claim for the dangers of HRT is now showing just the opposite. It is an affront to science, adding insult to injury to the thousands of women who abandoned HRT as a result."

Women had been advised for many years that HRT after menopause might reduce their risk of heart disease and various aspects of aging.

However, in 2002 the Women's Health Initiative study found that women undergoing HRT had a slightly increased risk of breast cancer, heart disease, stroke, and Alzheimer's disease sufficient to justify stopping the study.

After these results were reported, the number of prescriptions written for two brands of oestrogen - Premarin and PremPro - dropped almost in half in the US as many women discontinued HRT altogether.

This drop in prescriptions was followed by large and successively greater fals in new breast cancer diagnoses at six months, one year, and 18 months after that announcement, for a cumulative 15% drop by the end of 2003.

Surprisingly, no similar drop in Canada's breast cancer rates was observed during the same period, though prescriptions of PremPro and Premarin were reduced in Canada as well. Studies are now under way to determine if the drop is related to the reduced use of HRT.

In December last year research by experts from the University of Texas said a sharp drop in breast cancer cases could have been caused by women abandoning HRT.