Relationship between psychiatric symptoms and regional cerebral blood flow in patients with mild Alzheimer's disease


Dr Keisuke Nakajima MD, Department of Psychiatry, Kitasato University School of Medicine, 2-1-1 Asamizodai, Sagamihara, Kanagawa 228-8520, Japan. Email:


Background:  Behavioral and psychological symptoms of dementia (BPSD) are frequently observed in patients with dementia and often cause serious problems. However, the cause of BPSD has not yet been elucidated. Moreover, the precise evaluation of BPSD in mild dementia has not been studied in any great detail. In the present study, we investigated the relationship between psychiatric symptoms and regional cerebral blood flow (rCBF) in patients with mild Alzheimer's disease (AD).

Methods:  The present study included 47 patients (20 men and 27 women) who were diagnosed with mild AD. Mean patient age was 72.8 ± 8.2 years. Single photon emission computed tomography (SPECT) with 99mTc-ethyl cysteinate dimer (99mTc-ECD) was performed in all patients. The SPECT data were analyzed using a three-dimensional stereotactic region of interest template, which evaluated CBF in 24 segments. Psychiatric symptoms were evaluated in patients using the Brief Psychiatric Rating Scale. Each psychiatric symptom was designated as ‘symptom present’ in cases in which the BPRS item score was more than 3. We compared 10 segments of rCBF in symptom-present patients with those in symptom-absent patients.

Results:  Motor retardation was the most common psychiatric symptom (36.2%), followed by depression (19.1%), anxiety (17.0%), emotional withdrawal (17.0%), and somatic concern (14.9%). Alzheimer's disease patients with motor retardation exhibited a tendency towards lower rCBF in seven segments (left callosomarginal, bilateral parietal, bilateral angular, and bilateral temporal). However, no specific tendency was observed in depression, anxiety, and somatic concern.

Conclusions:  In the present study, we observed a tendency for decreased brain perfusion in patients with motor retardation. Further studies are necessary to confirm that this trend contributes to the appearance of psychiatric symptoms in patients with mild AD.


The number of dementia patients is predicted to increase due to an increase in the elderly population. In dementia patients, various psychiatric symptoms and behavioral problems (behavioral and psychological symptoms of dementia1 (BPSD)) develop concomitantly, as well do major symptoms of impairment of cognition and memory. Although BPSD often threaten the quality of life of patients and could contribute to the physical and mental burden of the caregivers of such patients,2 the causality and mechanism(s) underlying the development of the psychiatric symptoms are not fully understood, despite the various investigations that have been conducted.

Recently, studies on various psychiatric diseases and cerebral dysfunction have been conducted using analysis of functional images, particularly changes in cerebral blood flow (CBF) analyzed on images obtained using single photon emission computed tomography (SPECT). Although a detailed setting for the Region of Interest (ROI) is essential for precise measurement of regional CBF (rCBF), manual methods are likely to generate errors, causing considerable differences in study results. Takeuchi et al. were successful in setting a fixed universal ROI group (template) on an anatomical standard image obtained by Statistical Parametric Mapping (SPM99)3 and achieved precise and simple measurements of rCBF, eliminating data spread in the process of analysis, as well as between observers and subjects, using an analysis program with the template, namely the three-dimentional stereotactic ROI template (3DSRT).4

There have been a few studies that have investigated the mechanism underlying psychiatric symptoms in mild Alzheimer's disease (AD) from the perspective of CBF, but no such studies have been conducted in Japan. It is known that the severity of BPSD is associated with the severity of dementia.5 However, Shimabukuro et al. reported no significant difference between the frequency of several psychiatric symptoms, such as hallucinations, delusions, and anxiety, in mild and moderate AD.6 These findings suggest that the timing of and the mechanisms responsible for the appearance of BPSD remain unclear and that an investigation of BPSD and mild AD from the perspective of CBF may contribute to an understanding of the mechanism underlying the appearance of BPSD.

It is known that frontal lobe damage may cause decreased spontaneity, inactivity, and disinhibition.7,8 In a recent study, it was found that frontal brain dysfunction may underlie depression in both cerebrovascular disease and neurodegenerative disorders.9 A variety of emotional disorders are commonly associated with the temporal lobe10 and various psychiatric symptoms, such as alteration of mood, obsessive thinking, and hallucinations, often appear in patients with temporal lobe epilepsy.11 In a study of AD patients, Kotrla et al. reported that those patients with delusions exhibited hypoperfusion of the left frontal lobe relative to the right frontal lobe, whereas those with hallucinations exhibited hypoperfusion in the parietal lobe.12 Therefore, in the present study we investigated the relationship between the presence of psychiatric symptoms, in association with mild AD, and rCBF in frontal, temporal, and parietal lobes, measured with 3DSRT, to obtain information regarding the mechanism underlying the appearance of BPSD.


Sixty-five patients were diagnosed with probable AD, based on the diagnostic criteria of the National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA),13 of 122 patients who visited the dementia discrimination course for the early detection of dementia at Kitasato University East Hospital from March 2005 to March 2007. In the present study, we defined mild AD as a simultaneous score of <1 point on the Clinical Dementia Rating (CDR) scale14 and >16 points on the Mini-Mental Scale Examination (MMSE).15 Forty-seven patients who met these criteria were included in the study. Written informed consent was obtained from each patient and the study was performed after approval had been granted by the Ethics Committee in Kitasato University, School of Medicine.

Of the 47 subjects included in the study, 20 were men and 27 were women. The mean (± SD) age of all subjects was 72.8 ± 8.2 years and the mean length of time from onset of AD was 1.7 ± 1.0 years. The MMSE was conducted for each patient to evaluate their cognitive dysfunction. The mean MMSE score in the 47 patients was 20.9 ± 3.3. Nine patients had been treated previously with donepezil and seven patients had been treated with antidepressants. Thirty-one patients had been treated for physical complications, such as hypertension or diabetes.

The Brief Psychiatric Rating Scale (BPRS)16 was also used to determine the presence of BPSD. When the rating was determined as a score of ≥3 for each item of the BPRS, the psychiatric symptom was considered to be present; patients were categorized as either ‘symptom present’ or ‘symptom absent’ for each item of the BPRS. Although disorientation could be included in BPRS as a subitem, it was excluded from analysis in the present study because it is a major symptom of AD. In addition, the Neuropsychiatric Inventory (NPI)17 is often used in the evaluation of BPSD; however, in the present study we used the BPRS to assess patients because we included patients with mild symptoms that were not able to be judged using the NPI.

Single photon emission computed tomography with 99mTc-ethyl cysteinate dimer (99mTc-ECD) was performed and CBF was measured quantitatively with the Patlak Plot18 method for all subjects. In addition, SPECT images were analyzed with 3DSRT Ver. 3 (Daiichi Radioisotope Laboratories, Tokyo, Japan) to measure rCBF. In the present study, we selected bilateral callosomarginal, precentral, parietal, angular, and temporal segments to represent rCBF in the frontal, temporal, and parietal lobes. Student's t-test with Bonfferoni correction was used to assess differences in mean rCBF values between the two groups, those with and without psychiatric symptoms (the frequency of which was >10%). P < 0.005 was considered significant.

Patients with a high frequency of BPSD were again divided into groups based on the severity of BPRS (symptoms absent, mild, moderate, or moderately severe) and differences in mean rCBF between these groups were examined. One-way analysis of variance (anova) was used to compare the 10 segments in which rCBF was measured in these groups. If the anova indicated a significant difference, Scheffé's post hoc test was applied. Again, P < 0.05 was considered significant.

We also examined correlations between mean rCBF values and patient age or the duration of AD using Pearson's bivariate correlation coefficient test. In addition, differences in mean rCBF according to gender were examined using Student's t-test with Bonferroni correction (with P < 0.005 indicating significance). If there was a significant difference in these factors, the correlation between these factors and the severity of BPSD was examined using Spearman's correlation coefficient test to establish risk factors for BPSD.

All statistical analyses were performed using SPSS15.0 J for Windows (SSPS, Chicago, IL, USA).


The frequency of appearance of BPSD in patients in the present study is given in Table 1. Of the psychiatric symptoms, motor retardation occurred most frequently (36.2%), followed by depression (19.1%), anxiety (17.0%), emotional withdrawal (17.0%), and then somatic concern (14.9%; Table 1). Although moderately severe levels of somatic concern were observed in a few patients, most patients rated the degree of BPSD as mild.

Table 1.  Frequency and levels of psychiatric symptoms in Alzheimer's disease patients
 FrequencyBPRS rating (n)
n(%)MildModerateModerately severe
  1. BPRS, Brief Psychiatric Rating Scale.

Motor retardation17(36.2)116 
Emotional withdrawal8(17.0)71 
Somatic concern7(14.9)241

There were no significant differences in rCBF between the symptom-present and -absent patient groups for any segment investigated. However, in the group with symptoms of motor retardation, there was a tendency for lower rCBF in the left callosomarginal, bilateral parietal, bilateral angular, and bilateral temporal segments (Table 2). We then investigated the differences in rCBF between three groups, namely those patients with mild and moderate motor retardation and those without. Regional CBF was significantly decreased in the right temporal region of the group with moderate symptoms compared with the group without symptoms.

Table 2.  Comparison of regional cerebral blood flow in Alzheimer's disease patients with and without motor retardation
 Regional cerebral blood flow (mL/100 g per min)P
AD + motor retardationAD − motor retardation
(n = 17)(n = 30)
  1. Data are the mean ± SD.

  2. AD, Alzheimer's disease.

 Right34.6 ± 4.536.6 ± 3.40.09
 Left34.9 ± 4.137.2 ± 3.50.048
 Right36.3 ± 4.437.7 ± 4.00.279
 Left37.2 ± 4.939.4 ± 4.20.111
 Right34.6 ± 3.737.1 ± 4.50.048
 Left33.9 ± 3.436.6 ± 4.00.023
 Right36.8 ± 4.039.8 ± 5.40.049
 Left35.2 ± 3.638.5 ± 4.90.019
 Right33.2 ± 3.835.8 ± 3.70.027
 Left32.9 ± 3.535.3 ± 3.90.044

No specific differences were found in CBF between symptom-positive and symptom-negative groups for any segments in depression, anxiety, or somatic concern (data not shown).

In the present study, there was no significant correlation between the duration of AD and rCBF, nor was there any significant difference in mean rCBF between men and women. The age of patients exhibited a weak negative correlation with rCBF in widespread regions of the brain, except the bilateral angular segment, and we found a weak positive correlation between the age of patients and the severity of motor retardation (r = 0.362; P = 0.012).


In the present study, using 3DSRT, it was found that there was a tendency towards reduced rCBF in the left callosomarginal, bilateral temporal and parietal segments in mild AD patients with motor retardation compared with rCBF in patients without symptoms. In addition, rCBF was decreased in the right temporal region of patients with moderate motor retardation compared with the symptom-free group. Age was also correlated with a widespread reduction of rCBF and the severity of motor retardation. Depression, anxiety, emotional withdrawal, and somatic concern did not correlate significantly with rCBF in either the symptom-present or -free groups.

Motor retardation is considered an expression of apathy, which is one of the peripheral symptoms of AD. Robert et al. reported a significant decrease in CBF in the right anterior cingulate gyrus of AD patients lacking initiative and interest (MMSE score 22.8 ± 3.0).19 Benoit et al. reported negative correlations between initiative and CBF in the right anterior cingulate gyrus, between interest and CBF in the right middle orbital frontal gyrus, and between apathy and CBF in the upper left prefrontal cortical laterodorsal side in AD patients (MMSE score 22.8 ± 9.3).20 However, according to Lanctot et al., although decreased CBF was observed in the right orbital frontal cortex and left anterior cingulate gyrus of AD patients with apathy (MMSE score 22.3 ± 5.1), CBF was increased in the bilateral hippocampus, left medial superior gyrus, and right, middle, and left medial temporal cortex.21 Regarding the relationship between sugar metabolism evaluated using positron emission tomography and apathy, there have been some reports demonstrating decreased activity in the bilateral anterior cingulate gyrus and bilateral thalamus of patients with apathy.22 Although a definitive answer has not yet been obtained to explain the relationship between apathy and changes in CBF in particular regions, previous studies19–22 have demonstrated decreased CBF in the anterior cingulate gyrus and frontal cortex of patients with apathy, which partially matches the results of the present study. Moreover, we also found a correlation between age and widespread reductions in rCBF, as well as a correlation between age and the severity of motor retardation. These findings suggest that associations exist between age, reductions in rCBF, and the appearance of motor retardation. Further studies with increased sample size are necessary to elucidate the relationships between reductions in rCBF, aging, and the appearance of apathy in AD patients.

The relationship between rCBF and depression has been reported in some publications. Liao et al. reported significantly decreased rCBF in the bilateral anterior and posterior cingulate gyrus and the precuneus in mild and moderate AD patients with depression (MMSE score 13.3 ± 4.0), and rCBF in the right anterior cingulate gyrus exhibited a negative correlation with patients' scores on the Hamilton Depression Rating Scale.23 However, we were not able to demonstrate a correlation between rCBF in any specific region and the presence of depression. We believe that the following reasons possibly explain the differences between the studies. The complication rate of depression in AD patients has been reported to be approximately 40–50%.24 In the present study, the complication rate of depression was 19.1%, which is relatively low, and the severity of the symptoms, assessed using the BPRS score, was mild. This is because the chief complaint of our patients was memory loss and they had attended the dementia discrimination course hoping for a precise diagnosis. Thus, because patients with serious depression are likely to be recognized by their family or caregivers, if the symptoms are apparent prior to a visit to our dementia discrimination course, treatment for depression is likely to occur first, followed by a visit to our course after improvement of these symptoms. According to previous investigations, it is a well-known fact that treatment of depression improves decreased CBF.25 Therefore, a statistically significant difference in rCBF could not be detected in the present study because the number of patients with comorbid depression was small and the severity of the depression was mild. To our knowledge, there have been no reports indicating that anxiety and somatic concern in AD patients are directly related to rCBF. Further studies with a larger number of subjects may elucidate the relationships between rCBF and the appearance of depression, anxiety, and somatic concern.

There are two clear limitations of the present study. First, the overall sample size was small and the number of BPSD patients was very small. Second, the two groups that were divided based on the presence or absence of psychiatric symptoms were not matched for age and sex. Therefore, the possibility that the study was not of sufficient size to detect differences between two groups should be considered. Regarding statistical analysis, previous studies26,27 using 3DSRT have used t-tests or Mann–Whitney U-tests to detect significant differences for each segment, which may result in a probability problem that all results of t-tests are simultaneously correct, known as type I error based on multiplicity. Therefore, we measured rCBF of the brain in 10 segments and tested for differences using Student's t-test with Bonfferonni correction to reduce the probability of a type I error. However, unexamined region(s) of the brain may be associated with psychiatric symptoms. Thus, all segments of 3DSRT should be investigated in a sufficiently large sample size using appropriate methods in the near future.


This work was supported by a grant from the Ministry of Health, Labour and Welfare on Nervous and Mental Disorders 17A-5. The authors thank Dr Katsutoshi Tanaka (Department of Occupational Mental Health, Graduate School of Medical Science, Kitasato University) for his statistical advice.