For the last few years, women have heard conflicting reports about the risks and benefits of hormone replacement therapy (HRT). A new study that analyzes many trials together concludes that HRT can reduce heart attacks by about one-third in women under age 60 but has mixed results for older women.

The statistical analysis of 23 trials with more than 39,000 women found that when women started HRT (estrogen with or without progesterone) in their 50s to relieve such menopausal symptoms as hot flashes, their risk of heart attacks or cardiac death fell by 32 percent compared with those receiving a placebo or no treatment.

This analysis, by researchers from Cornell and Stanford universities, is published in the April issue of the Journal of General Internal Medicine.

For women over the age of 60, however, HRT increased the risk of heart attacks in the first year of treatment, but after two years of treatment HRT began to reduce heart attacks compared with women taking a placebo or no treatment.

The study follows up on a similar 2004 meta-analysis (a study that pools trials together) conducted by the same researchers that probed overall deaths related to age and HRT. That study found 39 percent fewer deaths among women who began HRT before 60, compared with women of the same age given a placebo or no treatment.

"In the earlier study we did not have enough data to evaluate cardiac death, only total death," said astrophysicist Edwin Salpeter, the J.G. White Distinguished Professor of Physical Sciences Emeritus at Cornell, who has more recently focused his attention on medical statistics and led the statistical analysis in the study.

Various reports on the risks and benefits of HRT have generated confusion for women, Salpeter said. One reason is because the Women's Health Initiative (WHI) -- a long-term national health study focused on strategies for preventing heart disease, breast and colorectal cancer, and fractures in postmenopausal women -- had initially reported that HRT was linked to more heart attacks when they analyzed the sample as a whole, in women with a mean age of 63 years, rather than breaking down women's ages after menopause. This is important because younger women may respond differently to HRT than older women. Since then, the WHI researchers have provided data for coronary heart disease events for younger and older women separately.

Salpeter and his colleagues, including his daughter, lead author Shelley Salpeter, a clinical professor of medicine at Stanford's School of Medicine and a physician at Santa Clara Valley Medical Center in San Jose, Calif., included the new WHI data in their larger analysis.

"We pooled the data from the two WHI trials for younger women with all other randomized placebo-controlled trials and found a statistically significant 32 percent reduction in coronary heart disease events for women who start treatment in their 50s," said Shelley Salpeter.

A paper published by the two Salpeters and colleagues in Diabetes Obesity and Metabolism (2005, electronic publication ahead of print) may offer a clue as to why younger and older women may respond to HRT in different ways. That study found that HRT significantly reduces the risk of diabetes compared with placebo but increases the levels of a protein associated with inflammation in the body. This may reduce heart attacks in younger women without heart disease but increase the risks for older women who already have cardiovascular disease.

The most recent study shows that HRT reduces coronary heart disease events in women who start treatment shortly after menopause. Yet, other studies suggest that HRT increases the risk of breast cancer but reduces risk of colorectal cancer. Thus, younger postmenopausal women need to consult their physicians and weigh their medical histories to decide whether HRT is right for them, said Cornell's Salpeter.