In the UK, stroke affects one in five men and one in six women by the age of 75. Memory loss and cognitive changes are common feared complications of stroke and contribute substantially to disability. However, it is not clear why particular groups of patients develop problems with cognition after stroke. It is hypothesised that ethnicity might play an important role, perhaps because some dementia risk factors such as obesity are more common among certain ethnic populations. With the proportion of elderly individuals from Black, Asian and other ethnic minorities (BAME) set to increase over the next few years, this may have important implications for health and care.
In this study, we aim to explore whether adult stroke survivors of non-White ethnicity are more likely to develop dementia than stroke survivors of White ethnicity. To do this, we will use anonymised electronic health records from primary care linked to hospital records to compare dementia rates between different ethnic groups after stroke. Improved understanding of the effect of ethnicity on post-stroke dementia risk could help healthcare professionals to target preventive interventions e.g. against obesity at the ethnic groups who are at highest risk and thereby reduce the burden of post-stroke dementia.
Cognitive problems contribute substantially to the disability experienced after stroke and result in a major health and societal burden. Both short- and long-term cognitive changes occur after stroke, but reasons for the development of post-stroke dementia remain unclear. Despite evidence suggesting ethnicity is a risk factor for dementia, the role of ethnicity in post-stroke dementia has been poorly described globally, and to our knowledge, the incidence of post-stroke dementia by ethnicity has not been described in a UK setting.
Here we aim to investigate the association between ethnicity and incident dementia in a population cohort of adult stroke survivors using electronic health records (EHR). Using routinely collected EHR data will increase power and generalisability of findings as well as overcoming some methodological difficulties that may hamper traditional cohort studies of post-stroke dementia such as ascertainment bias. We will carry out a multivariable Cox regression analysis to compare incidence rates of post-stroke dementia among White and non-White ethnic groups in time periods after stroke, controlling for potential socio-demographic and clinical confounding factors. In secondary analyses, we will use a categorical exposure variable for ethnicity, subdividing non-White individuals into Black, Asian and Mixed.
Outcomes will be incident dementia, categorised into early post-stroke dementia (from three months to one year post stroke)[8] and later post-stroke dementia (from one to five years post stroke). The first three months post stroke will be excluded due to the risk of misclassification of reversible cognitive changes associated with stroke as dementia.
Although it is likely that post-stroke dementia will have a primarily vascular origin, dementia often has a mixed pathology and specific diagnosis is infrequently recorded. We will therefore use a broad definition that incorporates codes for vascular dementia, Alzheimers disease and unspecified dementia. These are:
ICD-10 codes
F00* Dementia in Alzheimer disease
F01 Vascular dementia
F03 Unspecified dementia
We will also obtain a list of codes classified under F02* Dementia in diseases classified elsewhere, which refer to dementia associated with known conditions such as Pick disease, CJD, Huntingtons disease, Parkinsons disease, HIV and other chronic conditions. In sensitivity analysis we will investigate the effect of including dementia cases based on these codes. We will use the mapping tool available at http://dementiapartnerships.com/resource/dementia-read-codes/ to map these groups of ICD-10 codes to Read codes. This list will be reconciled with previous codelists used in a CPRD study of dementia as well as published dementia Read codelists from previous studies e.g. https://clinicalcodes.rss.mhs.man.ac.uk/medcodes/article/25/codelist/re… then reviewed clinically by Professor Smeeth. The provisional code lists used to generate feasibility counts is shown in appendix 2.
Harriet Forbes - Chief Investigator - London School of Hygiene & Tropical Medicine ( LSHTM )
Harriet Forbes - Corresponding Applicant - London School of Hygiene & Tropical Medicine ( LSHTM )
Charlotte Warren-Gash - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Gharti Magar - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Liam Smeeth - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Neil Pearce - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Rohini Mathur - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
HES Admitted Patient Care;Practice Level Index of Multiple Deprivation