What have 30 years of tracking a million women taught us about women’s health?

Back in the early 1990s, hormone replacement therapy (HRT) was becoming hugely popular for treating menopausal symptoms thanks to claims it could improve women’s wellbeing and life expectancy. By the mid-1990s, evidence was beginning to emerge that HRT could increase breast cancer risk, but there was a lack of definitive data on the association between various types of HRT and cancer incidence.  

In search of some answers, the late Professor Dame Valerie Beral and her team in the Oxford Cancer Epidemiology Unit (now part of Oxford Population Health) decided to launch the biggest study of women’s health ever undertaken. Beral knew that to find reliable links between risk factors and disease, she would need to look at as many women as possible over a number of years (what’s known as a prospective study). A million was a bold target, but it was possible thanks to the unique way in which participants were recruited.  

Beral worked with the National Health Service Breast Screening Programme to send out recruitment questionnaires alongside breast screening invitations for women in their 50s and early 60s. Initially, there was a concern that this might put women off coming for screening, but a pilot study showed that this wasn’t the case. By making the process so simple, 66 screening centres recruited 1.3 million women over five years from 1996 to 2001 – a staggering one in every four women aged 50-64 across the UK.   

Funding for the study initially came from the Imperial Cancer Research Fund (now Cancer Research UK), and later from the Medical Research Council and the Health and Safety Executive.  

The sheer size of the Million Women Study made it incredibly powerful for picking apart the influences of everything, from medical history to behaviours, on health and disease. Initially, the study focused on HRT in relation to breast cancer, but it later expanded to become a vital window on women’s health in general.  

Women who took part in the study had reached adulthood in the 1960s and had very different lifestyles to women of previous generations. Besides taking HRT in midlife, many of them had smoked and used oral contraceptives as teenagers and young adults. Little was known about the long-term effects of these behaviours.  

Importantly, the data that women provided was linked to their NHS records, so the study team could follow them into the future. Beral and her research team also went back to the participants every three to five years to ask them for further information about a variety of topics, including their diet, habits and experiences of ageing.

So, how does HRT affect women’s risk of cancer?

After crunching through all the data, Beral and her team – notably statistician Professor Gillian Reeves (now the Director of the Cancer Epidemiology Unit, and co-principal investigator of the Million Women Study) – published their findings in a landmark paper in 2003. The results showed that women taking HRT were at an increased risk of developing breast cancer. This risk was significantly larger for women taking combined HRT containing the hormones oestrogen and progestogen, compared to oestrogen-only.  

Further results from the Million Women Study revealed that HRT can slightly increase the risk of ovarian cancer and that oestrogen-only HRT increases the risk of endometrial (womb) cancer in post-menopausal women who have not had a hysterectomy. Taken together, breast, ovarian and womb cancer add up to 40% of all female cancers in the UK, affecting many tens of thousands of women. Over a five-year period, for every thousand women not taking HRT, around 19 of them will develop one of these diseases. For 1,000 women taking oestrogen-only HRT this goes up to 26, and up to 34 for every 1,000 women taking combined HRT.  

The team’s findings made headlines in the UK and around the world and dramatically changed attitudes towards HRT. They also fit with those of similar smaller studies, including a clinical trial of HRT by the Women’s Health Initiative in the US.  

The growing body of evidence around HRT and cancer risk led to changes in prescription guidelines and the information provided to doctors and patients. Women were advised to take HRT for as short a time as possible to treat menopause symptoms while those at high risk of breast cancer were encouraged to consider alternatives to HRT. As sales of HRT dropped in the US, UK and Europe, the incidence of breast cancer in women over of the age of 50 also decreased.  

In 2019, the Million Women Study contributed to another major report that used data from 58 studies around the world to determine the long-term effects of taking HRT on breast cancer risk. The findings suggested that the increased risk of breast cancer for women taking HRT persisted for more than 10 years after they stopped taking it. This long-term effect meant that the risks were approximately double what had previously been thought. The authors estimated that for women of average weight, taking combined HRT (oestrogen and progestogen) for five years increased the incidence of breast cancer by about one additional case in every 50 users. The amount of risk that persisted after stopping HRT for more than 10 years depended on the duration of use.  

In recent years, HRT use has been increasing again amidst greater awareness and public discussion of menopause and perimenopause. While such discussions are welcome, women should take into account medical advice about the benefits and risks of taking different types of HRT for different time periods, and consider alternatives if they are at high risk for breast cancer.

What other insights into women’s health has the study revealed?

Besides the findings on HRT and cancer risk, the Million Women Study has produced many other important results linking women’s diet, habits, environment and socioeconomic status to their risk of different diseases, including heart disease, motor neurone disease and various cancers. As the women in the study have grown older, more findings have focused on health issues associated with ageing, such as hip fractures, cataracts, and dementia. 

Some of the study’s key findings over the years include: 

• Women who smoke are three times more likely to die prematurely than women who have never smoked, the equivalent to losing 11 years of life, on average. 

• Women who drink more than 14 drinks per week have a three times higher risk of liver cirrhosis. Alcohol intake is also associated with an increased risk of several types of cancer. 

• Women with disabilities are less likely than other women to participate in cancer screening (36% less likely for breast cancer screening and 25% less likely for bowel cancer screening). 

• Using the contraceptive pill for 10 years cuts a woman’s risk of womb cancer by half. This is thought to have prevented around 400,000 cases of womb cancer worldwide over 50 years. 

• Mobile phone use does not increase brain cancer risk. Two studies found no evidence that using mobile phone under usual conditions increased women’s risk of developing brain tumours.  

• Women are at a substantially increased risk of venous thromboembolism for up to 12 weeks after an operation.  

• Women who are obese in midlife have a 21% greater risk of dementia in later life compared to women with a health body mass index (BMI).  

Millions more

Data from the Million Women Study are often combined with data from other large prospective studies to unpick the complex interactions between genes, environment, lifestyle and the risk of many diseases.  

For example, the largest ever study of non-meat diets and cancer risk combined data from nine studies, including the Million Women Study, the UK Biobank and EPIC-Oxford, to conclude that vegetarian diets are associated with a lower risk of several cancers.  

Another study that combined data from the Million Women Study, UK Biobank and EPIC-Oxford examined whether night shift work could increase women’s risk of breast cancer. The researchers found no association, helping to settle a question that had gone unresolved for several decades.  

A recent study that used data from the Million Women Study, UK Biobank and China Kadoorie Biobank showed that diabetes is linked to a higher risk of several cancers, particularly cancers of the liver, pancreas, and bladder. The link with diabetes was strongest for liver cancer and was greater in people who were overweight or who drank more alcohol. 

The Million Women Study continues to serve as a vital resource for questions about women’s health that are yet to be answered. The age of the participants and the size of the study means that researchers can now investigate conditions of older age, such as stroke, disability and dementia, on a scale that was not previously possible.  

The Million Women Study and all the research projects that Oxford Population Health scientists are involved in are only possible thanks to the participants who freely provide their time and information. We are grateful to each and every one of them.

Updated 29 April 2026.