Hormone replacement therapy and risk of breast cancer: case-control study using QResearch

What is the aim of the study and why is it important?

This study will quantify the risks of breast cancer associated with different types of hormone replacement therapy (HRT) in order to provide women and their clinicians with the best information to inform treatment decisions.
In November 2015, NICE published its first ever guidance on the menopause. Oestrogen depletion associated with menopause can cause irregular periods, hot flushes, night sweats, mood changes, memory loss, vaginal dryness, a lack of interest in sex, headaches, and joint stiffness. Quality of life may be severely affected. Prolonged lack of oestrogen increases risk of osteoporosis.
A central theme to the new NICE guideline is the need to provide patients with information on the short and longer-term risks and benefits of HRT to help women make an informed choice about which treatment to use for menopausal symptoms.
The use of HRT halved following the publication of two large studies which raised concerns about the safety profile of HRT. These were the Women’s Health Initiative in 2002 and the Million Women study in 2003, based on 10-15 year study period and a mean follow up period of between 4 and 5 years. This new NICE guidance, therefore, is likely to result in a resurgence of the use of HRT among women once the guideline is disseminated. Media reports about HRT have not always been accurate, so providing healthcare professionals and women with a robust source of information is vital.
One issue specifically highlighted as a research question in the NICE guidance is how the preparation of HRT affects risks of breast cancer. This is the problem which our proposal will specifically address by the analysis of longitudinal data over a 20-year period. The results are intended to improve the evidence base for HRT to help doctors and patients make better treatment decisions which are better tailored to the individual.

How is the research being done?

This will be a nested case control study within a cohort of community based patients registered with practices in the UK contributing to the QResearch and the Clinical Practice Research Database (CPRD) databases. A case control design is the most efficient study design to answer this question since it uses all available cases and allows detailed analysis of drug exposure and reduces potential bias through matching.

Chief Investigator

Julia Hippisley-Cox,



Location of research

Oxford and Nottingham

Date on which research approved


Project reference ID


Generic ethics approval reference


Are all data accessed are in anonymised form?


Brief summary of the dataset to be released (including any sensitive data)

Women aged 40-79 registered during the study period (1998-2017) with a diagnosis of breast cancer each matched to 5 controls by age, practice and calendar time.
GP data includes diagnoses, risk factors and medication relevant to breast cancer.
Hospital, mortality and cancer registry data includes diagnoses and operations relevant to breast cancer (eg mastectomy; hysterectomy).

Implications and Impact

Fears of breast cancer deter many women from taking HRT even in the presence of debilitating menopausal symptoms. It is essential, therefore, to understand which types of HRT are the safest, not only for the individual women, but also at a population level. Small increases in risk of breast cancer are likely to have a significant impact at a population level given both the frequency of breast cancer and the numbers of women eligible for treatment.
There are many different types of HRT available. These vary by the regimen (unopposed oestrogen; combined cyclical; combined continuous); type of oestrogen (conjugated equine oestrogen, oestradiol) or progesterone (medroxyprogesterone acetate, norethisterone/norgestrel); dose and duration of treatment; and route of delivery (oral, transdermal).
However, there are significant gaps in our knowledge regarding the risk of breast cancer with different types of HRT and different methods of administration. This safety profile information is essential given the large number of women who may now be considered for HRT. Large electronic research databases from primary care (QResearch and CPRD) have detailed information on diagnoses and prescriptions linked to hospital, cancer registration and mortality records. Since these databases have very detailed prescription data from over 20 years for millions of women and tens of millions of HRT prescriptions linked to hard outcomes, they can provide an efficient way to quantify the risks associated with different types of HRT.

Better information for women on the relative and absolute risks of breast cancer associated with different types of HRT will help women make more informed decisions

Funding Source

No external funding

Research Team

Julia Hippisley-Cox, Yana Vinogradova, Carol Coupland

Share this