This project is about understanding who is at risk of dementia. In particular, we are thinking about two overlapping risk factors: prescription medications and delirium (worsening short-term confusion when unwell). We know that certain medications, known as ‘anticholinergics’, are linked with developing dementia in later life. We can also identify dementia risk in older adults by short episodes of increased confusion, known as delirium. Sometimes anticholinergic medications can also cause delirium.
Older adults may be prescribed anticholinergics for their beneficial effects, for example treating depression, allergies or incontinence. It can be difficult to weigh these benefits against the potential risk of dementia.
Our study has two key aims. The first is to understand whether long-term use of anticholinergics also raises the chance of developing delirium or delirium progressing to dementia. The second is to categorise any characteristics that leave older adults at a higher risk of anticholinergic medications’ potential harms. Using electronic health records, like GP records, we can answer these questions by looking to see if patients already taking anticholinergics have more positive or negative effects. This will allow doctors and care providers to better understand which individuals should stop these medications to avoid causing harm.
Background:
Anticholinergic medication use is known to be associated with incident dementia and delirium. Incident delirium can represent a herald event for subsequent dementia. Whether anticholinergic medication use could mediate the relationship between delirium and incident dementia is not understood. Higher frailty can predispose individuals to harms resulting from medication use via altered intrinsic drug processing, drug-drug and drug-disease interactions. It is plausible observed associations between anticholinergic medication use, delirium and dementia could be modified by frailty. Understanding these relationships could provide insight into the risk-benefit profile of these commonly used medications.
Objectives:
To utilise Electronic Health Records (EHR) to determine associations between anticholinergic medication use, delirium and dementia, among individuals with differing levels of frailty.
Methods & Data Analysis:
We will describe the use of anticholinergic medications among older adults grouped by frailty level. We will then compare associations with anticholinergic use and delirium, dementia, and mortality.
Part I: We will produce a descriptive analysis of anticholinergic medication use in adults aged 65 >= within the study period. These data will be stratified by frailty using the electronic frailty index.
Part II: We will pair anticholinergic medication users with matched controls, creating disease risk scores for both dementia and delirium to adjust for potential confounding.
Part III: We will use logistic regression models to estimate associations between anticholinergic medication use, delirium and dementia; stratified by frailty. This will include nested analysis on a subsample of individuals with delirium, using dementia as an outcome.
Our primary outcomes are: Incident dementia; Incident delirium.
Incident dementia will be measured by the domains dementia diagnosis, dementia specific prescription, and dementia monitoring activity. We will not include individuals with rare disease-specific dementias thought unrelated to anticholinergic medication use (i.e. alcoholic dementia, dementia in Pick’s or Motor Neurone Disease, Creutzfeldt Jakob’s Disease, Huntington’s and HIV). Read Codes will be evaluated in CPRD and ICD-10 codes in HES and ONS mortality data
The domains will be defined as:
Dementia diagnosis: Dementia; Alzheimer’s; Vascular dementia; Dementia with Lewy Bodies/Parkinson’s; Delirium with dementia; Mixed (Alzheimer’s Vascular) Dementia. Though dementia will be stratified into its subtypes in sensitivity analyses, the majority of dementia diagnoses are recorded as non-specific dementia and therefore will be analysed together.
Dementia-specific prescription: memantine; galantamine; donepezil; rivastigmine.
Dementia monitoring activity: Dementia monitoring; dementia tests; dementia referrals.
Incident delirium will be measured by the domain delirium diagnosis, and delirium related activity.
The domains will be defined as:
Delirium diagnosis: Delirium; Subacute delirium; Delirium with dementia.
Delirium related activity: Confusion; Acute brain syndrome (ABS); Acute organic reaction; Organic brain syndrome (OBS); Acute psycho-organic syndrome; Acute confusional state; subacute confusional state; psychosis associated with intensive care; intensive care psychosis.
A full code list for these outcomes is attached as Appendix A.
Daniel Davis - Chief Investigator - University College London ( UCL )
Mark Rawle - Corresponding Applicant - University College London ( UCL )
Arturo Gonzalez-Izquierdo - Collaborator - University College London ( UCL )
Daniel Davis - Collaborator - University College London ( UCL )
Muhammad Qummer ul Arfeen - Collaborator - University College London ( UCL )
Wallis Lau - Collaborator - University College London ( UCL )
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation