Behavioral changes are common in patients with Alzheimer's disease (AD) and have a major impact on their quality of life and social affiliations.1,2 The underlying mechanism for these behavioral symptoms in AD is not well understood. The apolipoprotein E epsilon 4 (APOE E4) allele is a known risk factor in late-onset AD. The APOE E4 allele has been assumed to be involved in the pathomechanism of AD, and thus the association between the E4 allele and the clinical expression, including psychiatric manifestations, of AD has been hypothesized. Several studies have examined the involvement of this allele in common mental symptoms, including delusions, hallucinations, and depression, with conflicting results.3—9 Some of these conflicts may be accounted for by methodological factors, including different definitions of behavioral changes, the lack of statistical power with small sample sizes, and inappropriate application of statistical analysis without controlling the possible confounding factors. In the present study, taking these concerns into consideration, we analyzed the relationship between the APOE E4 allele and neuropsychiatric manifestations in a considerably large cohort by using a well-established comprehensive tool for the assessment of behavioral abnormalities.
All procedures of this study strictly followed the Clinical Study Guidelines of the Ethics Committee of Hyogo Institute for Aging Brain and Cognitive Disorders (HI-ABCD), Himeji, Japan, in 1993, and were approved by the Internal Review Board. After a complete description of all procedures of this study, written informed consent was obtained from patients or their relatives.
The subjects of this study were 175 of a consecutive series of 279 Japanese patients with sporadic AD who were given a short-term admission to the infirmary of the HI-ABCD, a research-oriented hospital for dementia, for examination between August 1995 and September 1997. All patients were examined by both neurologists and psychiatrists and were given routine laboratory tests and standard neuropsychological examinations. In addition, electroencephalography, magnetic resonance (MR) imaging of the brain, MR angiography of the neck and head, and cerebral perfusion/metabolism studies by positron emission tomography or single-photon emission computed tomography were done. All results were incorporated in the diagnosis. The inclusion criteria were those of both the DSM-IV10 for dementia of the Alzheimer's type and the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association11 for probable AD. One hundred four patients were excluded from this study in accordance with established exclusion criteria. These consisted of complications of other neurological diseases (n=33) or physical illnesses (n=12), evidence of parkinsonism (n=6) or focal brain lesions on MRI (n=3), the presence of severe hearing or vision impairments (n=2), a history of mental illness or substance abuse before the onset of dementia (n=8), the absence of reliable informants (n=28) such as living-with family or caregivers, or the inability to obtain informed consent (n=12).
The included patients were 146 women and 29 men; the age at examination (mean±SD) was 71.9±8.1 years, and the mean educational attainment was 9.3±2.5 years. The functional severity was 0.5 (very mild) in 20 patients, 1 (mild) in 100 patients, 2 (moderate) in 43 patients, and 3 (severe) in 12 patients as determined by the Clinical Dementia Rating Scale (CDR).12 The age at onset and the duration of the disease were ascertained through an interview with the primary caregiver. Age at onset was defined as the age of the first appearance of symptoms of sufficient severity to interfere with social or occupational functioning, and the duration was defined as the time in months between the onset and the admission.13 The mean age at symptom onset and duration of symptoms were 69.1±8.5 years and 33.6±20.7 months, respectively. The mean value of the Mini-Mental State Examination14 was 19.3±4.6.
Assessment of Psychiatric Status
We assessed each patient's behavioral changes semiquantitatively during an interview with the caregiver by using the Japanese version15 of the Neuropsychiatric Inventory (NPI).16 Satisfactory reliability and validity have been established for the Japanese version.15 A single physician was employed in the assessment. In the NPI, the following 10 behavioral and psychotic changes in dementia were rated on the basis of the patient's condition in the previous month before the interview: delusions, hallucinations, depression (dysphoria), anxiety, agitation and aggression, disinhibition, euphoria, irritability and lability, apathy, and aberrant motor activity. According to the defined criteria, the severity of each manifestation was classified into 3 grades (from 1 to 3) and frequency of each manifestation was classified into 4 grades (from 1 to 4). The NPI score (severity×frequency) was calculated for each manifestation. Behavioral changes were also rated on a present/absent basis.
Determination of APOE Genotype
The detailed method for APOE genotyping is described elsewhere.17 In brief, genomic DNA was extracted from whole blood samples by the phenolchloroform method and was amplified by the polymerase chain reaction (PCR) as described by Wenham et al.18 The PCR products were digested with 10 units of Hha I for 5 hours at 37° C. The DNA fragments were electrophoresed for 5 hours at 60 mA through a 15% nondenaturing polyacrylamide gel. The gel was stained with ethidium bromide and photographed under ultraviolet light. The genotypes were determined by the size of DNA fragments.
Because a dose effect of the APOE E4 alleles has been shown to affect AD risk,19—21 we tested the relationship between the number of E4 alleles and the NPI score of each behavioral change. We used Spearman rank correlation coefficients because the NPI scores were not normally distributed. To examine the effects of E4 alleles on the presence of each behavioral change, we also used a multiple logistic analysis, with the dichotomy (presence or absence) of each manifestation as a dependent variable and the number of E4 alleles as an independent variable. Each analysis was repeated initially without controlling the effect of potential confounding variables, and secondarily by enrolling age at onset, sex, duration of symptoms, education level, and CDR as severity of dementia into the model. In order to reduce the chance of a statistical type II error, we adopted a nonconservative alpha level of 0.05 despite the use of multiple comparisons.
The APOE genotyping was E2/E3 in 6 patients, E2/E4 in 2 patients, E3/E3 in 61 patients, E3/E4 in 88 patients, and E4/E4 in 18 patients. Thus, 67 patients had no E4 allele, 90 had one E4 allele, and 18 had two E4 alleles. The frequency and the mean value for the NPI score of each neuropsychiatric manifestation are summarized in T1. The most common abnormal behavioral manifestation was apathy, followed by aberrant motor behavior, delusions, dysphoria, and irritability. The least common was euphoria.
The correlation analysis between the number of APOE E4 alleles and each NPI score revealed no significant association (T2). Even after the effects of age at onset, sex, duration of disease, education level, and CDR severity of dementia were controlled, the results remain unchanged. The multiple logistic regression analysis (T3) also failed to show any association between the number of APOE E4 alleles and the presence of any behavioral changes, either before or after controlling the possible confounding factors. For euphoria, controlling these factors was not possible because the frequency of positive cases was too low.
The NPI results obtained from the Japanese AD patients demonstrated that apathy was the most common symptom and that disinhibition, hallucinations, and euphoria were rare. This finding is consistent with the findings of other studies in the United States, Italy, and Mexico.22 Anxiety and agitation in our subjects were not uncommon, but they were one-third less frequent than they were in the United States.2,22 On the other hand, the distribution of the APOE genotype demonstrated in this study was similar to the distributions reported in other races and countries.17,23
The major finding in the present study is that the number of APOE E4 alleles was not associated with any of the behavioral changes in Japanese probable AD patients in terms of prevalence or severity. Ramachandran et al.3 reported an association of the APOE E4 allele with psychosis and depression. Murphy et al.4 reported that the APOE E4 allele is associated with greater behavioral disturbances in AD patients. However, they did not investigate individual behavioral symptoms. On the other hand, studies that have analyzed the effect of the APOE E4 allele on delusions, hallucinations, and depression in larger cohorts while controlling confounding factors failed to demonstrate the association of the APOE E4 allele with either delusions,5—8 hallucinations,5—9 or depression.5—7,9 Together with these earlier studies, the present study clearly demonstrated that the APOE E4 allele has no effect on the manifestation of delusions, hallucinations, or depression in AD. Although AD patients manifest other types of behavioral changes, few studies have addressed the effect of the APOE E4 allele. Our results demonstrated no effect of APOE E4 alleles on any symptoms including agitation, anxiety, euphoria, apathy, disinhibition, irritability, or aberrant motor behavior, and are consistent with a study of Holmes et al.7 that reported no significant association with noncognitive symptoms including aggression, wandering, and stereotyped behaviors. We confirmed in the Japanese patients that the neuropsychiatric manifestation in AD is not associated with the APOE genotype.
The APOE E4 allele did not associate with behavioral changes in AD. The APOE E4 allele is an important risk factor for developing AD and may affect the pathological changes.23,24 However, previous epidemiological studies have indicated that the profile and rate of cognitive impairment7,25—29 and survival rates30,31 are not different according to the APOE genotype once persons have a diagnosis of AD. Moreover, cerebral glucose metabolism studies did not find any APOE genotype—related differences in patients with the full dementia syndrome of AD.32,33 Our findings are compatible with these studies and do not support the hypothesis that neuropsychiatric manifestation of AD is different in patients with the APOE E4 allele.
The authors thank Hajime Kitagaki, M.D., and Kazunari Ishii, M.D., of the Division of Neuroimaging Research, for their help in various parts of the study, and Yoko Takatsuki, B.A., R.S.T., and Akitsugu Tokimasa, R.O.T., of the Neurorehabilitation Service, for their technical assistance.