The relationship between the diagnosis method of neuronal dysfunction (DIMENSION) and brain pathology in the early stages of Alzheimer's disease

Authors


Correspondence: Mr Minoru Kouzuki MA, Department of Biological Regulation, School of Health Sciences, Faculty of Medicine, Tottori University, 86 Nishimachi, Yonago 683-8503, Japan. Email: minoru_kouzuki@yahoo.co.jp

Abstract

Objectives

To examine whether the diagnosis method of neuronal dysfunction (DIMENSION), a new electroencephalogram (EEG) analysis method, reflected pathological changes in the early stages of Alzheimer's disease (AD), we conducted a comparative study of cerebrospinal fluid markers and single-photon emission computed tomography.

Methods

Subjects cincluded 32 patients in the early stages of AD with a Mini-Mental State Examination score ≥24 (14 men, 18 women; mean age, 77.3 ± 9.2 years). Cerebrospinal fluid samples were collected from AD patients, and cerebrospinal fluid levels of phosphorylated tau protein (p-tau) 181 and amyloid β (Aβ) 42 were measured with sandwich ELISA. EEG recordings were performed for 5 min with the subjects awake in a resting state with their eyes closed. Then, the mean value of the EEG alpha dipolarity (Dα) and the standard deviation of the EEG alpha dipolarity (Dσ) were calculated with DIMENSION. Single-photon emission computed tomography analyses were also performed for comparison with DIMENSION measures.

Results

Patients with parietal hypoperfusion had significantly increasing p-tau181, decreasing Dα, and increasing Dσ. In addition, there was a negative correlation between Dα and p-tau181, p-tau181/Aβ42, and a positive correlation between Dσ and p-tau181/Aβ42.

Conclusion

Dα and Dσ were related to cerebral hypoperfusion and p-tau181/Aβ42. DIMENSION was able to detect changes in the early-stage Alzheimer's brain, suggesting that it is possibility as a useful examination for early-stage AD with a difficult discrimination in clinical conditions. Moreover, EEG measurement is a quick and easy diagnostic test and is useful for repeated examinations.

Introduction

Alzheimer's disease (AD) is one of the neurodegenerative disorders that most commonly occur in the elderly. About 10% of the population over 65 years of age in Japan suffers from dementia, and half of them have AD.[1, 2] A neuropathological feature of AD is deposition of the senile plaques, neurofibrillary tangle, and neuronal deficiency. The main component of senile plaques is amyloid β (Aβ) protein, and that of neurofibrillary tangles is hyperphosphorylated tau protein (p-tau).

The decrease of Aβ42 and the increase of total tau, p-tau in the cerebrospinal fluid (CSF) and/or their ratios (total tau/Aβ, p-tau/Aβ) are used as biological tests with high reliability as diagnostic markers.[3-8] </CHypoperfusion in the posterior cingulate and reduction in oxygen metabolism and sugar metabolism can be confirmed with single-photon emission computed tomography (SPECT) and positron emission tomography in early-stage AD patients.[9-12] Moreover, regional cerebral blood flow (rCBF) and sugar metabolism in the temporal lobe and parietal lobe decrease with disease progression.[13] These tests are suitable for a definite diagnosis of AD but are not suitable as screening tests for early detection.

EEG is a record of the electrical potential distribution on the scalp and the activity of the neuronal cells in the brain. Previous studies have shown various EEG findings to distinguish AD. AD patients are characterized by higher delta (0–4 Hz) and lower posterior alpha (8–13 Hz),[14-16] as well as by a marked reduction of the synchronization likelihood at both frontoparietal and interhemispheric electrodes.[17, 18] Furthermore, it was discovered that progressive atrophy of the hippocampus correlates with decreased cortical alpha power in mild cognitive impairment and AD conditions.[19]

Recently, Musha et al. developed an EEG analysis method called the diagnosis method of neuronal dysfunction (DIMENSION).[20] DIMENSION models the random electric current dipole distribution of the cortical alpha component by functional depression of neuronal cells. This method has the potential to become a powerful tool for the interpretation of electrophysiological brain activity. It has been reported that DIMENSION's sensitivity and specificity in distinguishing AD from healthy elderly adults were both 90%.[20] However, early-stage AD has not yet been adequately examined with this method, and comparison with the pathologies that characterize AD has not been performed either.

The ability of previous EEG methods, such as spectral power and coherence measures,[21, 22] to accurately detect very early-stage AD and its progression was problematic, as the results were often difficult to interpret; it was not clear whether EEG corresponded to changes in processes within the brain. By comparing results from SPECT analysis and CSF markers in early-stage AD, the present study investigated whether DIMENSION analysis reflects changes in brain processes related to AD.

Material and Methods

Patients and samples

The research protocol was explained to patients and/or their relatives, and informed consent for their participation was obtained. The study design was approved by the ethics committee of Tottori University (Yonago, Japan). All patients were examined by a dementia specialist.

Profiles of all subjects are shown in Table 1. Subjects included 32 patients in the early stages of AD with Mini-Mental State Examination scores ≥24 (14 men, 18 women; mean age, 77.3 ± 9.2 years).[23] AD diagnosis was done according to the DSM-IV and National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association criteria.[24, 25] All diagnoses were made after examinations of the patients' medical histories and family histories, an evaluation of physical and neurological examinations, routine laboratory tests, neuropsychological tests, and magnetic resonance imaging or computed tomography.

Table 1. Patient profiles
 Patients (M/W) (n)Age, mean ± SD (years)MMSE, mean ± SD
  1. M, men; MMSE, Mini-Mental State Examination; W, women.
Posterior cingulate hypoperfusion11 (2/9)76.9 ± 6.426.9 ± 2.2
Temporal lobe hypoperfusion10 (6/4)76.7 ± 12.126.2 ± 2.1
Parietal hypoperfusion11 (6/5)78.3 ± 8.325.3 ± 2.0
Total patients32 (14/18)77.3 ± 9.226.2 ± 2.2

CSF samples were collected by lumbar puncture at EEG measurement. Collected CSF samples were stored immediately at −80°C until use.

ELISA for p-tau181 and Aβ42 in CSF

The p-tau181 and Aβ42 levels in CSF were measured with a commercially available human p-tau181 sandwich ELISA kit (Innogenetics, Ghent, Belgium) and Aβ42 sandwich ELISA kit (Immuno-Biological, Gunma, Japan), respectively, according to the manufacturers' instructions.

EEG recordings

EEG subject recordings were performed for 5 min with the subjects awake in a resting state. EEG recordings were performed from 21 sites on the scalp arranged according to the international 10–20 system (Fp1, Fp2, F3, F4, F7, F8, Fz, T3, T4, T5, T6, C3, C4, Cz, P3, P4, Pz, O1, O2), with a reference electrode on the right earlobe. EEG was recorded with a Digital Bio-Amplifier 5200 (NF Electronic Corporation, Yokohama, Japan) at a sampling rate of 200 Hz with analogue band-pass filtering in the frequency range of 0.5–250 Hz (time constant 0.3 s). There was no visible difference between EEG waves (diffuse slow-wave activity, paroxysmal activity, sharp-wave activity, background activity, focal abnormalities) in individual patients as shown in Figure 1.

Figure 1.

Example of EEG of AD patient in early stages of AD. There was no visible change in EEG, specifically in any artefacts or paroxysmal discharges. AD, Alzheimer's disease; EEG, electroencephalogram.

EEG data analysis

The analysis of the EEG data used DIMENSION developed by Musha et al.[20] DIMENSION is a computer simulation that calculates dipolarity (D), an index that measure the loss of uniformity of an observed scalp EEG alpha potential distribution. We computed the subject's D, defined as D = √1 – ((uobsudip)2)av/(u2obs)av, where uobs and udip are observed and dipole-generated potentials at an electrode site and (x)av denotes the average value of x over all the electrode sites.[26] Cortical lesions result in randomly oriented dipoles, which lower the dipolarity (uniformity) of the scalp potential distribution.[27] The increasingly non-uniform distribution of peak alpha EEG components are caused by impaired neuronal activities. When the amplitude of the alpha component reaches the maximum, the D value also reaches the maximum. The D value of the alpha component shows almost periodic variation. We selected only peak values and averaged them to obtain the mean alpha dipolarity (Dα) and its standard deviation (Dσ). Peak D values smaller than 0.85 were not included because they were largely influenced by instrumentation noise.

SPECT method

Ethyl cysteinate dimer (ECD) is a lipophilic compound that crosses the blood–brain barrier and adheres to the active moieties in the brain. Each subject received a 600-MBq intravenous injection of 99mTc-ethyl cysteinate dimer just prior to the SPECT. Ten minutes after this injection, brain SPECT images were recorded.

SPECT images for all patients were anatomically standardized with an original 99mTc-ethyl cysteinate dimer template using the easy Z-score imaging system.[28-30] A Z-score map for the SPECT image was obtained by comparing it with SPECT images from an age-matched database of normal brains. The mean and standard deviation for each voxel were obtained after anatomical standardization and voxel normalization to global mean values with the following equation: Z-score = ([Control mean] − [Individual value])/(Control SD).

Based on SPECT image analysis using the easy Z-score imaging system and visual inspection, we found decreased rCBF in patients with early stages of AD. The subjects were classified into three groups: (i) posterior cingulate hypoperfusion group; (ii) temporal lobe hypoperfusion group; and (iii) parietal hypoperfusion group (Table 1). In a very early stage of AD, decreasing rCBF in the posterior cingulate has been reported based on SPECT.[11, 12] Therefore, there were some patients with decreasing posterior cingulate blood flow in the temporal lobe hypoperfusion group and the parietal hypoperfusion group. There were no significant differences in age or neuropsychological tests among the three groups (Table 1).

Statistical analysis

The statistical analysis was performed using the Mann–Whitney U-test. The correlations of Dα and Dσ with p-tau181, Aβ42, and p-tau181/Aβ42 were examined with the Pearson product-moment correlation coefficient. Differences where P < 0.05 were considered statistically significant.

Results

Comparison between CSF markers and rCBF

The levels of Aβ42 and p-tau181 in the three groups are shown in Figure 2. The three groups were classified according to the sites with decreased rCBF – posterior cingulate hypoperfusion, temporal lobe hypoperfusion and parietal hypoperfusion. Levels of Aβ42 did not significantly differ among the three groups, but decreasing levels were observed, with the posterior cingulate hypoperfusion group being the highest, followed by the temporal lobe hypoperfusion group and then the parietal hypoperfusion group. P-tau181 levels were significantly increased in the parietal hypoperfusion group compared with the other two groups (P < 0.05).

Figure 2.

Levels of cerebrospinal fluid markers. Measurement of (a) Aβ42 and (b) p-tau181 in the three groups. The level of p-tau181 is significantly higher in the parietal hypoperfusion group compared with the other two groups. Data are mean ± SD. *P < 0.05, **P < 0.01 (Mann-Whitney U test). Aβ, amyloid β; p-tau, phosphorylated tau protein.

Comparison of rCBF with Dα and Dσ

The results of Dα and Dσ in the three groups are shown in Figure 3. Musha et al. reported that Dα exhibits a decreasing tendency as AD progresses while Dσ exhibits an increasing tendency.[20] Dα significantly decreased in the parietal hypoperfusion group compared with the other two groups (P < 0.05). Dσ significantly increased in the parietal hypoperfusion group compared with the other two groups (P < 0.05).

Figure 3.

Comparison of regional cerebral blood flow with Dα, and Dσ. Comparison of (a) Dα and (b) Dσ among the three groups. In the parietal hypoperfusion group, (a) Dα is significantly decreased and (b) Dσ is significantly increased compared with the other two groups. Data are mean ± SD. *P < 0.05, **P < 0.01 (Mann–Whitney U test). Dα, mean value of the EEG alpha dipolarity; Dσ, standard deviation of the EEG alpha dipolarity; EEG, electroencephalogram.

Comparison of CSF markers (p-tau181, Aβ42) with Dα and Dσ

The results of comparisons between p-tau181 with Dα and Dσ are shown in Figure 4a and b. There was a negative correlation between Dα and p-tau181 (r = −0.457, P < 0.01). It was not a statistically significant correlation, though there was a tendency for a positive correlation between Dσ and p-tau181 (r = 0.300, not statistically significant).

Figure 4.

Comparison of regional cerebrospinal fluid markers with Dα and Dσ. Comparison between (a) Dα and p-tau181, (b) Dσ and p-tau181, (c) Dα and Aβ42, and (d) Dσ and Aβ42. A negative correlation was observed only between Dα and p-tau181. Aβ, amyloid β; Dα, mean value of the EEG alpha dipolarity; Dσ, standard deviation of the EEG alpha dipolarity; EEG, electroencephalogram; p-tau, phosphorylated tau protein.

The results of the comparison of Aβ42 with Dα and Dσ are shown in Figure 4c and d. There were no significant correlations between Aβ42 and Dα or Dσ.

Comparison of p-tau181/Aβ42 with Dα and Dσ

The results of comparisons of p-tau181/Aβ42 with Dα and Dσ are shown in Figure 5. The ratio of p-tau181/Aβ42 is used as a diagnostic marker with greater reliability than either p-tau181 or Aβ42 alone.[8] There was a significant negative correlation between Dα and p-tau181/Aβ42 (r = −0.541, P < 0.01). Furthermore, a significant positive correlation was observed between Dσ and p-tau181/Aβ42 (r = 0.391, P < 0.05).

Figure 5.

Comparison of p-tau181/Aβ42b with Dα and Dσ. Comparison between (a) Dα and p-tau181/Aβ42 and (b) Dσ and p-tau181/Aβ42. There was a negative correlation between Dα and p-tau181/Aβ42 and a positive correlation between Dσ and p-tau181/Aβ42. Aβ, amyloid β; Dα, mean value of the EEG alpha dipolarity; Dσ, standard deviation of the EEG alpha dipolarity; EEG, electroencephalogram; p-tau, phosphorylated tau protein.

Discussion

In this study, it was confirmed that, in DIMENSION, the indices Dα and Dσ are related to the pathological changes indicative of AD in early-stage AD patients (i.e. those with a Mini-Mental State Examination score ≥24).

Aβ42 did not significantly differ among the three groups, but it showed a decreasing tendency, with the highest levels in the posterior cingulate hypoperfusion group, followed by the temporal lobe hypoperfusion group, and the lowest levels in the parietal hypoperfusion group. It was thought that the deposition of Aβ increased in the parietal hypoperfusion group. In contrast, p-tau181 was significantly increased in the parietal hypoperfusion group as compared with the other two groups. In a previous study, it was reported that the total tau and the p-tau were related to left parietal hypoperfusion, though a correlation was not demonstrated between Aβ and SPECT.[31, 32] Similarly, a significant increase in p-tau181 was seen in the parietal hypoperfusion group in the present study, although there was no significant correlation between Aβ42 and rCBF. It has been suggested that Aβ may reach a plateau from a rapid phase in the process of cortical degeneration, as shown in a previous positron emission tomography imaging study.[33] These results suggest that decreased rCBF may be related to the progress of AD (decreasing Aβ42 and increasing p-tau181).

Dα and Dσ significantly correlated with the index of p-tau181/Aβ42 (r = −0.541 and r = 0.391, respectively). P-tau181/Aβ42 is a diagnostic marker with high reliability in AD.[8] Our results show that DIMENSION analysis reflects the pathology of AD. Moreover, there was a correlation between Dα and p-tau181 (r = −0.457). It is thought that DIMENSION is strongly related to changes in p-tau181, which is involved in neuronal death. There is corresponding evidence in the report from Musha et al. that decreased electrophysiological activity in the cerebral cortex in AD (i.e. decreased EEG alpha dipolarity) reflects neuronal injury.[20] Neuronal death may be one of the causes of the abnormalities in EEG.

There was a significant increase of Dα and a significant decrease of Dσ in the parietal hypoperfusion group. Abnormality of the EEG rhythms, especially frequency of the slow wave, in dementia has been associated with altered rCBF and cognitive function.[34, 35] In this study, there was no visible difference in EEG (diffuse slow-wave activity), but Dα and Dσ changed with decreased rCBF. This suggests the possibility that abnormality in Dα and Dσ can be found in the progression of AD, in which the rCBF has degraded significantly without clinical signs.

In this study, Dα and Dσ in the early stages of AD (i.e. those with Mini-Mental State Examination score ≥24) are related to p-tau181/Aβ42 and cerebral hypoperfusion. Therefore, DIMENSION analysis was able to detect the pathological changes in the brain in early-stage AD patients. Although the sensitivity is inferior to SPECT and the CSF markers, it is relatively easy, non-invasive and useful as a repeat test. One of the limitations of the present study was the time of day at which the EEG measurement was taken. There is huge circadian variation in EEG, such as the difference in drowsiness between the morning and afternoon. In this study, although the measurement time was not constant, all patients were examined during the day. It has been reported that abnormality on EEG in AD also reflects the influence of defects in acetylcholine and that improvement in EEG can be observed following treatment.[36-38] As a result of the neuronal deficiency, an acetylcholine deficit occurs. Because the reduction in EEG alpha dipolarity reflects neuronal injury, DIMENSION's ability to evaluate the effects of therapeutic treatments should also be considered. In the future, it will be necessary to examine the changes of Dα and Dσ according to the progression of AD symptoms and the effects of therapeutic treatments.

Acknowledgments

We thank all the patients who participated in our study as well as their families. We also thank the doctors at Shinsei Hospital (Kurayoshi, Japan) who cooperated in this study.

Ancillary