A million women get hormonal. (Research Initiatives).

In recent years, one million has come to be regarded as the critical mass for a show of support for social change, as evident in the many groups and marches whose names include that amount. In the case of the Million Women Study, however, the descriptor is not only apt, it's somewhat modest. In this British study 1.3 million postmenopausal women are participating in an unprecedented investigation of hormone replacement therapy (HRT) and its health effects. "There are very few epidemiological studies that include more than about 20,000 people," says Valerie Beral, co--principal investigator of the study, which is being coordinated by the Imperial Cancer Research Fund's Cancer Epidemiology Unit at the University of Oxford. "We have, over about a three-year period [1997-2000], recruited about 20,000 every month--it's like doing a standard epidemiological study every month for three years," she says.

The study is, in a word, ambitious. A primary goal of the Million Women Study is to investigate the breast cancer risk associated with HRT. Although many epidemiologic data exist on the link between breast cancer and HRT, results have been inconclusive about whether different types of HRT have different effects. Researchers with the Million Women Study aim to resolve that problem.

The researchers recruited women through the National Health Service Breast Screening Programme, which offers routine breast cancer screening for women in the United Kingdom aged 50-64. Through detailed questionnaires, the researchers collected data on HRT use and other variables such as diet, health and reproductive history, and alcohol and tobacco use. "As the largest study of its kind ever conducted, the Million Women Study has the potential to address a large number of research questions," says Emily Banks, a senior epidemiologist at the University of Oxford and the study's other co--principal investigator. "From the very earliest stages of recruitment, the enthusiasm women had for the project was fantastic--the questionnaires were very carefully filled in, there were huge numbers of women keen to take part, and we had lots of letters and calls praising the study and asking the sort of questions the study is designed to answer," she says. Preliminary results from the first 121,000 respondents--primarily an overview of the study itself--were published in 1999 in Breast Cancer Research. Beral says further results will be made available as soon as possible.

But whether the Million Women Study will resolve the question of the role of HRT in breast cancer is open for debate. Despite efforts to guard against biases that have hampered previous studies--for example, HRT users being generally healthier than nonusers--Garnet Anderson, a biostatistician at the Women's Health Initiative Coordinating Center at the Fred Hutchinson Cancer Research Center in Seattle, Washington, says some biases cannot be prevented. "The many observational studies that have looked at the association of hormones and breast cancer ... have left us confused because of inherent weaknesses in such studies," explains Anderson. "It is hard to know how yet another observational study, though admittedly a well-powered one, can clarify these areas."

Banks and Beral counter that they are taking into account known trouble spots in observational studies, factors such as self-selection bias and recall bias. They are planning studies to see whether these two issues will affect their results; the potential impact is currently unknown. Although the Million Women Study has the power, statistically speaking, to uncover small effects related to HRT, such effects could potentially be masked by uncertainty attributable to confounding factors.

Still, Anderson anticipates that the Million Women Study may shed new light on other exposures and some of the rarer hormonally related diseases such as colorectal, ovarian, and endometrial cancers. "When you have a million women and you can capture their health events over several years' time, you'll be able to see things that the smaller research programs could never dream of looking at," she says.