- Top of page
- Conflicts of interest declaration
- Description of authors' roles
Aim: Alzheimer's disease (AD) is characterized by cognitive symptoms and behavioral symptoms, and their association is inconsistent. The aim of this study was to investigate the relationship between cognitive function and the changes in behavioral and psychological symptoms of dementia (BPSD) in patients with AD.
Methods: A total of 101 patients with probable AD were enrolled (57 women and 44 men, mean age 77.6 ± 7.7 years). The Category Verbal Fluency Test (CVFT), the Mini-Mental State Examination (MMSE), the Constructional Praxis Test, the Delayed Word Recall Test, the Clinical Dementia Rating Scale, and the Neuropsychiatry Inventory (NPI) were administered at baseline. The NPI was reassessed with a median follow-up duration of 10 months (range 6–18 months). The change in the NPI scores was defined as the end-point score of the NPI minus the initial one. The associations between the changes in NPI total score, its four subdomains (hyperactivity, psychosis, affection, and apathy), and cognitive function were examined using multivariate linear models. The results were adjusted for confounders including demographics, baseline NPI, and duration of follow up.
Results: The mean MMSE was 18.6 ± 5.6, the CVFT score was 7.1 ± 3.9, and the NPI score was 10.9 ± 13.8. Regression analyses found that the CVFT score (β = −0.32, P = 0.004) was significantly associated with the change in NPI score, but not the MMSE, the Delayed Word Recall score, or the Constructional Praxis score. The CVFT score was significantly associated with changes in the psychosis subdomain (β = −0.34, P = 0.001), but not the other subdomains.
Conclusions: Our study showed that CVFT was predictive of the changes in behavior disturbance in patients with AD, particularly in the psychosis domain.
ALZHEIMER'S DISEASE (AD) is characterized by impairment in memory, visuospatial functions, language, executive function, and neuropsychiatric symptoms.1 Behavioural and psychological symptoms of dementia (BPSD)2 are being increasingly recognized as an important aspect of dementia because of the impacts on both patients and their caregivers, premature institutionalization, and increased costs of care.3–5
Several studies have been conducted to determine the predictive factors of behavioral disturbances in patients with AD. Baseline behavioral disturbance, baseline stage of dementia, use of support services, early age of onset, extrapyramidal symptoms, and apolipoprotein e4 allele have been reported to be associated with BPSD among patients with AD.6,7 It is also evident that there is a link between cognitive impairment and behavioral symptoms in patients with AD.8 In recent years, cross-sectional studies suggest that executive dysfunction in dementia is associated with both functional limitation and disturbed behavior.9–11 One current cohort study has reported that early executive function impairment is a predictor of behavior disturbance at a later time in demented patients indicated by comprehensive neuropsychological battery, including category verbal fluency, similarities, ideational fluency and visual reasoning.12 Understanding the changes in BPSD would be helpful in clinical practice and caregivers' counseling,13 however, follow-up studies of BPSD are relatively scarce.
A brief and easy-to-perform neuropsychological measure is more practical to use on the patient first showing in the outpatient clinic setting. The Category Verbal Fluency Test (CVFT), a simple instrument for cognitive function evaluation, has been used to assess semantic memory, executive function and language.14,15 CVFT is considered as an appropriate screening tool for AD, questionable dementia and mild cognitive impairment.16,17 It has been shown to associate with functional decline,18 to differentiate between dementia subtypes,19 and to predict mortality in patients with AD.20 The performance of CVFT in dementia is associated with both functional impairment and disturbed behavior in cross-sectional studies.21
The aim of this study was to examine the patients with diagnosis of AD in a memory outpatient clinic setting to determine whether baseline Mini-Mental State Examination (MMSE) or CVFT was related to the changes in behavioral disturbances at a follow-up study.
- Top of page
- Conflicts of interest declaration
- Description of authors' roles
The present study showed that poor performance on baseline CVFT was independently associated with a more rapid progression in overall neuropsychiatric manifestations at follow up, particularly in the psychosis domain among patients with AD. However, initial MMSE scores, the Delayed Word Recall test scores, and CERAD Constructional Praxis scores were not associated with the change of behavioral disturbance. The MMSE is considered as a tool for general cognitive function evaluation, whereas, the Verbal Fluency Test, is considered as a tool for semantic memory and executive function. Our results indicated that poor performance in these cognitive parts may be more specific and related to changes in behavioral disturbance rather than global cognition, verbal memory and visuospatial function.
Some tests for executive function were also useful for predicting the progression of BPSD. Our finding is similar to a recent study among mild demented patients (mean MMSE score 22.5), which suggested that very early frontal executive impairment, tested by the revised Cambridge Cognitive Examination Executive Function score, but not global cognitive function, was related to subsequent behavior disturbance.12 Another study reported that patients with executive dysfunction tested by a series of six cancellation tasks and the time to complete the first three mazes from the expanded Alzheimer's Disease Assessment Scale–Cognitive Subscale had more frequent symptoms of psychosis than those without.33 As the mean MMSE scores demonstrate (mean MMSE score 18.6), our memory clinic sample was relatively moderately cognitively impaired at initial assessment with a high prevalence rate (83%) of behavioral and psychological symptoms, which is compatible with most previous studies because BPSD is common in patients with AD.32 Our study supports the theory that executive dysfunction has a negative influence on the neuropsychiatric symptoms in AD, and extends the finding from patients with very early AD to those in later stages of the disease.
Many factor analytic methods have been used to explore sub-syndromes contained within the NPI due to heterogeneity of the BPSD symptoms in patients with AD.34 In the subdomain analysis, CVFT at baseline could only predict the changes in psychosis. Some reports demonstrated the correlations between frontal lobe function and human behavior,35 and the dysfunction involved in distinct parallel circuits within frontal–subcortical circuits may be the reason for this difference. The lesion of dorsolateral prefrontal circuit involved mainly in executive function is related with the performances of verbal fluency,35,36 and psychosis in AD patients has also been reported to be associated with the same circuit in previous studies.37 Consistent with previous studies on dementia, our findings may indicate that the psychosis domain contributed substantially to the association between BPSD and executive function.38 In addition, the studies on patients with psychosis symptoms also revealed the association between psychosis and verbal fluency and showed that severity of delusions was related to semantic fluency in schizophrenia and bipolar disorders.39 The deficits in CVFT and the worsening of psychosis in AD may share closer pathophysiological processes than the changes in other subdomains of behavioral disturbance.
We did not find any associations between baseline behavioral symptoms and baseline cognitive function results, which supports prior reports that cognitive and behavioral disturbances are either independent dimensions or that they have a weak linkage.33,40 Nevertheless, we found associations between baseline CVFT score and the changes in behavioral symptoms. A possible explanation of these findings is that the correlations between baseline behavioral symptoms and baseline cognition are different from those between changes in behavior and cognition. However, reasons for this finding are still unclear and need further research.
Limitations are addressed here when interpreting the results. Because few cognitive function tests were administered, this study cannot provide a comprehensive cognitive assessment of the change in behavioral symptoms and cognitive function across the full range of cognitive function. Although we tried to control most of the confounding factors, it might not be enough. The sample size is modest with a relatively short duration of follow up, which might not be long enough to evaluate whether any changes occurred with time. The neuropsychological assessment evaluated the prevalence within one prior month, without taking into account the duration of fluctuation over a longer period from baseline. Our study only enrolled subjects with AD, which may limit its application to other types of dementia.
The CVFT has been recommended as a useful instrument for AD among multidimensional investigations.16,18–20 Our study found that the CVFT was related to the change in BPSD and can be used as a simple and practical tool in clinical practice.