Approved research

Modelling dementia risk in real-world clinical and population data

Lay summary

Managing certain lifestyle factors such as high blood pressure, cholesterol and obesity can help prevent nearly a third of dementia cases worldwide. To enable these prevention efforts, recent research studies have developed 'dementia risk assessment scores', which predict a person's chance of developing dementia based on their lifestyle and genetic history. Dementia risk scores could be used to identify high-risk people for treatments to delay or prevent dementia. However, before they can be used to guide clinical decisions, risk scores must be thoroughly tested to make sure they are accurate. In this project, I propose to use large population databases to test how well risk scores can predict who will develop dementia, with the aim of translating such scores from being used research settings to clinical care. Using the UK Clinical Records Interactive Search (CRIS), I will use anonymised health records of people who currently have a diagnosis of dementia. These health records are also linked with the UK Biobank Study, which provides cognitive, genetic and lifestyle health information from when patients were initially healthy, between 40-65 years old. Using this real-world data I will test if dementia risk scores calculated in mid-life can accurately predict who will develop dementia in later life. The project will be conducted over a span of 30 months. The results of this project will inform dementia risk screenings, and enable us to more efficiently target at-risk people for interventions or clinical trials.