Caregiver characteristics are associated with neuropsychiatric symptoms of dementia
Sink KM, Covinsky KE, Barnes DE, Newcomer RJ, Yaffe K;
Commented by , 23 Jun 2006
Background
Behavioural and psychological symptoms in dementia (BPSD) affect nearly all patients with dementia at some point, and are associated with poor outcome, including increased caregiver burden, caregiver depression, and earlier institutionalisation. It is not known if BPSD is driven by the (personality deterioration of the) dementing process alone, or if there are particular caregiver characteristics or behaviours that trigger or exacerbate BPSD.
Aim
To determine whether caregiver characteristics are independently associated with BPSD, after correcting for patient characteristics.
Method
Subjects were drawn from the Medicare Alzheimer’s Disease Demonstration and Evaluation Study, an intervention study aimed at reducing depression in dementia caregivers. Cross-sectional data from this study was analysed secondarily. Subjects were community-dwelling volunteers with established dementia from 8 USA cities, and their caregivers.
Data was collected on caregiver demographics, caregiver depression and burden; patient demographics, and severity of dementia activities of daily living (ADL) and BPSD. Multivariate linear regression was used to analyse any association between presence of BPSD and caregiver characteristics.
Results
Data was available on 5,788 subjects with dementia and their caregivers.
Patient demographics – patients had average age 79 years, 60% were female, mean Mini Mental State Examination score was 14. 75% were dependent in 1 – 4 ADLs, and the mean number of neuropsychiatric symptoms was 4.8 (+/- 2.8 standard deviation). The commonest BPSD were repetitive questioning (67 %), wandering (59 %), delusions (51 %), psychomotor agitation
(46 %), anger (46 %) and sleep disturbance (46 %).
Caregiver demographics – caregivers had mean age 64 years (±15 years standard deviation), 71% were women (32% wives, 31% daughters or daughter-in-laws). The majority lived with the patient and reported good health.
Multivariate analysis – after adjusting for patient age, sex, dementia type, MMSE score and ADL, caregivers who were younger, less educated, more burdened, or depressed, reported significantly more BPSD. Caregiver characteristics accounted for twice the variance of patient characteristics.
Dr Seymour's comments
Much has been written about the adverse impact of high expressed emotion in families of people with schizophrenia. Is there an analogous situation with dementia caregivers, i.e., do stressed caregivers inadvertently induce BPSD in the demented person they are caring for? This study, with high numbers of participants that broadly represent patients with moderate-severe dementia in the community, lends weight to this hypothesis.
The main finding was that particular characteristics of the caregiver including education, burden and depression, were associated with higher reported rates of BPSD.
The methodological limitations of this paper are that it was a cross-sectional analysis of existing data, and the direction of association cannot necessarily be assumed between BPSD and caregiver characteristics.
The literature suggests that group interventions aimed at caregiver education have only been modestly effective. The significance of this study for clinicians is that account needs to be taken of the dynamics between patient and caregiver in managing BPSD: caregiver interventions need to be individually tailored rather than relying simplistically on group education.