Million Women Study

The Million Women Study is the largest study of women's health in the world. In 1996-2001, a quarter of UK females then aged 50-64 years (1.3 million women) joined the study. Study participants have provided details about their lifestyle and health and given signed consent for follow-up. Since then their health has been followed mainly through electronic linkage to routinely collected NHS records.

The aim of the study is to provide reliable information on potentially modifiable causes of common and serious illnesses, to help improve individual and public health. The study is jointly funded by Cancer Research UK, mostly for studies of cancer, and the Medical Research Council, mostly for studies of other conditions.

We have described links between smoking, alcohol, obesity, and use of hormone replacement therapy and women's risk of developing various diseases, recently reporting that the harms of smoking and the benefits of stopping were greater in women than previously thought.

This proposal is for research on dementia, osteoporosis, stroke and other severe disabling conditions that become increasingly common as women age. With this large study, detailed lifestyle information provided by women, and some 15 years follow-up through linkage to NHS medical records, we are uniquely placed to answer some major outstanding questions about what causes some of the serious conditions associated with ageing, and how they might be prevented.

We have already been able to link the study cohort to NHS records on death, cancer incidence, and hospital admissions. In the next 5 years we will enhance the linkage with newly-available information on medications prescribed and on conditions diagnosed in general practice.

Dementia is one of the most serious causes of disability as women age. There is surprisingly little good information about what people can do in middle-age to help prevent dementia occurring in later life. For example the effects of factors like smoking, alcohol, exercise and diet are still uncertain. Because dementia develops slowly over many years, and in the early stage can affect what people do, we need to study lifestyle factors over a long period before dementia is diagnosed. We will be able to follow large enough numbers of women for long enough to be able to provide reliable results, which other studies have not been able to do. With some 20,000 women in our study expected to develop dementia by 2017, we will also be able to study differences in risk factors for different types of dementia, such as Alzheimer's disease and vascular dementia.

Osteoporosis, where bones become weak and liable to fracture, is increasingly common as women age. Osteoporotic fractures of the hip and other bones are major causes of pain, disability and death. Bisphosphonate drugs, such as alendronate, are commonly prescribed to prevent fractures and their use is increasing rapidly. In recent years controversy about the balance between the benefits and risks of long-term use of these drugs has been ignited by reports of potentially serious side effects, which we plan to investigate.

We will also examine whether lifestyle factors in middle age, including diet and exercise, help prevent fractures occurring as women grow older.

This research will be made possible by combining the health information women have so generously given us with the excellent medical records available for the UK population through the NHS.

In 1996-2001 we recruited a cohort of 1.3 million women, a quarter of the female UK population then aged 50-64 years. About 99% have been followed for more than 10 years for hospital admissions and deaths. Mean age is now 70 years and conditions associated with ageing, such as dementia and osteoporosis, are becoming increasingly common. We plan to focus on these emerging health problems, following the cohort for a further 5 years by electronic linkage to NHS records.

The increasing availability of electronic routinely collected NHS records has revolutionized questions that can be addressed in cohort studies. Only recently was it possible to obtain linked NHS records for cause-specific hospital admissions in England. Linkage to NHS primary care prescription and diagnostic data has just become available.

For women with a primary care or hospital record of dementia, we will confirm diagnoses and classify them by phenotype, where possible. We will then examine associations with postulated risk factors, taking care to avoid 'reverse causation bias' by focusing on the 15,000 women first diagnosed with dementia 10+ years after reporting risk factors.

Bisphosphonates are increasingly used for osteoporosis, but surprisingly little robust information exists about their safety. Recent reports of serious pathological lesions of the bone and other possible long-term side effects are stimulating debate about the balance between beneficial and adverse effects in users. Randomized trials have been too small, with too little follow-up, to evaluate such effects. About 10% of study participants have used bisphosphonates, permitting reliable assessment of their net effect, overall and by duration of use.

The uniqueness and timeliness of the research planned derives mainly from the size and maturity of the cohort, the wealth of prospectively collected personal data, and the virtually complete follow-up via electronic NHS records.

The public will be the main beneficiaries of this research, both directly through better information about factors affecting their risk of developing common and serious diseases of middle age and of old age, and indirectly through our contribution to evidence upon which public health policy and guidelines are based.

Our research also contributes to better public understanding of science and engagement in research, particularly since most of our work is done as an integral part of the National Health Service.

Researchers in the UK and internationally, both in academic institutions and in the commercial sector, benefit from new knowledge on disease aetiology that we report and from the new methodologies we have pioneered. For example, we were among the first to use electronic methods for data-capture in epidemiological research, and are now initiating new research by harmonizing routinely collected NHS data from many sources, including hospital and primary care data collected in different databases in England and Scotland. We are regularly approached by other researchers about how to achieve such linkages.

The curation and availability of our data for future research will contribute to scientific research for decades to come.

Through excellence in research, we are helping to train future leaders in epidemiological and public health research. We hope to continue contributing to the research standing of UK universities and of the NHS and hence to funding and other economic benefits. The contribution that our study, and in particular the PI, has made to research and to public health has been recognised by the award of several major honours and prizes in recent years.

We hope that society as a whole will continue to benefit from our contribution to the improvement of public health, with the social and economic benefits this should bring.