What are the additional risks of dementia in women taking hormone replacement therapy?

Talk Code: 
Yana Vinogradova
Tom Dening, Julia Hippisley-Cox, Lauren Taylor, Michael Moore, Carol Coupland
Author institutions: 
University of Nottingham, University of Oxford, University of Southampton


Dementia is a progressive condition with major consequences for affected individuals, their families and their carers. It is most common in older people, and affects 1 in 14 over the age of 65. Some factors such as smoking, alcohol and obesity are known to increase the risk of developing dementia, but there is a lack of information about the effects of other factors, such as commonly used medications. For example, hormone replacement therapy (HRT) is increasingly used by women with severe menopausal symptoms, but how exposure to these hormones affects a patient’s risk of developing dementia is still unclear. All previous studies have been relatively small short-term or have not accounted for important confounding variables, and the most recent NICE guideline on menopause (2015) stressed the need to acquire more information on HRT risks and benefits with respect to developing dementia. This study focused on risks of incident dementia associated with different types of HRT, using routinely collected primary care data linked to hospital and mortality records and representative of the general female population of the UK.


Two nested case-control studies used the UK primary care databases, QResearch and CPRD, both linked to Hospital Episode Statistics and Office for National Statistics Mortality data. Overall between 1998 and 2020, 118,501 women aged 55 and older with a record of dementia or use of an anti-dementia drug, were matched by birth year and practice to up to 5 controls, alive and registered at the time of case diagnosis (index date). Exposure to HRT was taken from prescription records, excluding those within the 3 years prior to the index date. Risks for different types of HRT and duration of use were analysed using conditional logistic regression, and adjusted for ethnicity, body-mass index, smoking, alcohol consumption, social deprivation, co-morbidities and other drugs. Analyses were run separately for each database and the results combined using meta-analysis techniques.


Overall, 16,291 (13.7%) cases and 68,726 (13.8%) controls were ever exposed to HRT. No associations were found between different types or durations of HRT use and overall dementia risk. However, a subgroup analysis of women diagnosed with Alzheimer’s disease demonstrated an increased risk associated with oestrogen-progestogen therapy (odds ratio 1.08, 95% CI: 1.03-1.12), particularly for longer exposures (5-10 years: OR 1.11, 95% CI: 1.04-1.20; 10 years or more: OR 1.19, 95% CI: 1.06-1.33).


This is the largest single study providing population-based risk estimates. It showed no increased risks in short-term HRT users, but found a small increase in risk of developing Alzheimer’s disease for long-term users. The findings should assist doctors and patients considering HRT treatments, especially longer-term use. The results will also inform future national and international guidance on best practice.

Submitted by: 
Yana Vinogradova
Funding acknowledgement: 
The study was funded by School for Primary Care Research NIHR.