SEARCH

SEARCH BY CITATION

Keywords:

  • 18F-FDG PET;
  • Alzheimer's disease;
  • amnestic mild cognitive impairment;
  • early detection;
  • WAIS-III;
  • WMS-R

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES

Background:  We encountered eight early amnestic mild cognitive impairment (aMCI) patients (early MCI group) who did not fulfill the diagnostic criteria for aMCI. We compared the scores of neuropsychological examinations as well as the cerebral metabolic rate for glucose consumption (CMRglc) decrease on 18F-FDG PET examination between the early MCI group and 10 aMCI patients (MCI group) or six normal elderly subjects (normal group), to examine whether the current diagnostic criteria can detect early-stage aMCI.

Methods:  The three groups underwent Mini-Mental State Examination (MMSE), Wechsler Adult Intelligence Scale – Third Edition (WAIS-III), Wechsler Memory Scale Revised (WMS-R), magnetic resonance imaging and 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) examinations.

Results:  The early MCI group did not show significant memory impairment of 1.0 SD or other cognitive dysfunctions on neuropsychological examinations, and did not fulfill the diagnostic criteria of aMCI. With one-way anova and Tukey's HSD post-hoc test, the early MCI group showed the highest scores for WAIS-III, whereas the MCI group showed the lowest scores for WMS-R, although there were no significant differences between the early MCI and normal groups. In order to show a discrepancy in scores between WAIS-III and WMS-R, we subtracted the scores of WMS-R from WAIS-III. Consequently, the normal group showed significantly smaller differences in scores than the other groups, although there were no significant differences between the early MCI and MCI groups. 18F-FDG PET recognized a CMRglc decrease in the posterior cingulate gyrus and/or part of the parietotemporal area in both the MCI and early MCI groups, of which the extent and magnitude were weaker in the early MCI group. The normal group did not show a significant CMRglc.

Conclusions:  The early MCI group should be included in aMCI not only based on the discrepancy between intelligence and memory scores, but also based on the 18F-FDG PET findings. The combination of these examinations would make it possible to diagnose early-stage aMCI.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES

The detection of mild cognitive impairment (MCI) is important for an early diagnosis of dementia.1–3 According to the current clinical diagnostic criteria,4,5 amnestic MCI (aMCI), a type of MCI that has a strong possibility of progressing to Alzheimer's disease (AD), is defined as not being associated with aging or dementia, and presents with memory complaints, memory impairment and normal functional activities. aMCI is characterized by a Clinical Dementia Rating (CDR)6 of 0.5 or Global Deterioration Scale (GDS)7 of 3. Therefore, the presence of memory impairment with objective assessment is essential for a diagnosis of aMCI. Since Petersen et al.8 reported that aMCI patients show a decline by an average of 1.5 SD in a memory task, a number of studies9–11 have defined aMCI based on a memory decline of 1.0 or 1.5 SD from the average score of normal elderly subjects.

However, progression on aging to aMCI or AD is gradual and continuous. For this reason, it is possible that early-stage aMCI cannot be diagnosed as aMCI when its diagnosis is made based on a significant memory impairment of 1.0 or 1.5 SD.

In addition, memory impairment is generally assessed by taking the age and educational level into consideration.8 In Japan, however, the educational level might not be important for an assessment of memory impairment, because most Japanese people undergo a relatively high-level education compared with people in foreign countries with a considerable variation in educational levels. Some recent studies assessed memory impairment related not only to the age and educational level, but also intelligence.5,12 Lange and Chelune13 presented a diagnostic method for AD taking the Wechsler Adult Intelligence Scale – Third Edition (WAIS-III)14 score into consideration. Although scores in the Wechsler Memory Scale third edition (WMS-III)15 and the General Ability Index16,17 in WAIS-III were used in this method, they have not yet been standardized in the Japanese version.

Recently, we have encountered eight patients who consulted our memory clinic with a complaint of memory impairment, of which their memory was not significantly impaired but had mildly declined compared with their youth based on objective assessment. They showed a decreased cerebral metabolic rate of glucose (CMRglc) similar to that of aMCI on using 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) examination. In the present study, we compared these eight patients with aMCI patients and normal elderly subjects while controlling for age and educational level, and showed a discrepancy in scores of memory and intelligence using Wechsler Memory Scale Revised (WMS-R)18 and WAIS-III, respectively, in order to examine whether the current clinical diagnostic criteria of aMCI4,5 can detect early-stage aMCI.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES

Eight elderly patients (three woman and five men; early MCI group) consulted our memory clinic at Juntendo Tokyo Koto Geriatric Medical Center in 2008 (Table 1). They and/or their family complained of memory impairment (e.g. they often forget the names of people they had known for a long time and where things were). Most of them were able to do complicated work and travel long distances to unknown places, and their impaired memory did not usually cause serious problems in their daily lives. Their CDR score was between 0 and 0.5, and their GDS score was 2.

Table 1.  Clinical and neuropsychological features of the early mild cognitive impairment group
CaseGenderAge (years)No. years of educationMMSEWAIS-IIIWMS-R
VIQPIQFIQVCGMDR
  1. All patients in the early mild cognitive impairment group scored more than 85 in both WAIS-III14 and WMS-R,18 whereas WMS-R score is lower than 85 with 1.00 SD and lower than 75 with 1.65 SD for the definition of amnestic mild cognitive impairment. DR, Delayed Recall; GM, General Memory; FIQ, Full Scale IQ; MMSE, Mini-Mental State Examination; PIQ, Performance IQ; VC, Verbal Comprehension; VIQ, Verbal IQ; WAIS-III, Wechsler Adult Intelligence Scale – Third Edition, WMS-R, Wechsler Memory Scale Revised.

E1Male6616271161231211129194
E2Male671630137123134120110104
E3Female691629123120124116116106
E4Male7012261161061131129490
E5Male711630132112126131103107
E6Male73162913411913013110792
E7Female741130130125130129103110
E8Female741428129134134118103110
Mean70.514.628.6127.1120.3126.5121.1103.4101.6
SD3.12.11.58.08.47.18.18.18.3

In addition, 10 aMCI patients (seven women and three men; MCI group) were selected from our memory clinic, and six normal elderly subjects (four women and two men; normal group) were selected from volunteers while controlling for age and years of education (Table 2). The MCI group complained of memory impairment, and showed a CDR of 0.5 and GDS of 3, whereas the normal group did not complain of cognitive dysfunction including memory impairment, and showed a CDR of 0 and GDS of 1 or 2.

Table 2.  Comparison of clinical and neuropsychological features between the early mild cognitive impairment group and normal or mild cognitive impairment group.
GroupNormalEarly MCIMCIANOVAPost-hoc test
Female : male (n)4:23:57:3
MeanSDMeanSDMeanSD
  1. * P < 0.05, ** P < 0.01. Tukey's HSD was used for the post-hoc test (P < 0.05). ‘A, B’ means there is no significant difference between A's and B's score, and ‘C > D’ means C's is significantly higher than D's score. DR, Delayed Recall; E, early mild cognitive impairment group, General Memory; FIQ, Full Scale IQ; MCI, mild cognitive impairment group; MMSE, Mini-Mental State Examination; N, normal group; PIQ, Performance IQ; VC, Verbal Comprehension; VIQ, Verbal IQ; WAIS-III, Wechsler Adult Intelligence Scale – Third Edition, WMS-R, Wechsler Memory Scale Revised.

Age (years)68.34.770.53.168.85.5NS 
No. years of education14.01.814.62.113.82.1NS 
MMSE29.21.628.61.526.41.6**N, E > M
WAIS-III        
 VIQ113.310.2127.18.0113.911.4*E > N, M
 PIQ107.79.5120.38.4105.88.7**E > N, M
 FIQ112.210.5126.57.1111.410.5**E > N, M
 VC107.37.6121.18.1107.612.2*E > N, M
WMS-R        
 GM109.212.9103.48.177.510.1**N, E > M
 DR109.711.4101.68.360.77.3**N, E > M
Discrepancy        
 VIQ minus GM4.27.423.87.136.411.2**M > E > N
 FIQ minus GM3.05.923.17.333.911.7**M, E > N
 VC minus GM−1.87.617.89.330.113.0**M, E > N
 VIQ minus DR3.75.525.58.353.211.3**M > E > N
 FIQ minus DR2.54.724.96.750.711.7**M > E > N
 VC minus DR−2.34.619.59.646.913.3**M > E > N

The three groups underwent neuropsychological, magnetic resonance imaging (MRI) and 18F-FDG PET examinations, which are a routine set in our memory clinic. For neuropsychological examinations, Mini-Mental State Examination (MMSE),19 WAIS-III, and WMS-R were carried out.

Quantitative data were analyzed using one-way anova and Tukey's HSD post-hoc test. Nominal data were analyzed by the χ2-test. In all tests, the null hypothesis was rejected at a significance level of P < 0.01 or P < 0.05.

Neither patients nor subjects had any medical or psychiatric conditions that could potentially impact on cognitive function such as stroke, epilepsy, parkinsonism, schizophrenia or mood disorder. All diagnoses were carried out by a psychiatrist specializing in dementia, and neuropsychological examinations were carried out by a clinical psychologist.

The research ethics committee of Juntendo Tokyo Koto Geriatric Medical Center approved the present study, and all patients and subjects gave informed consent.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES

The clinical and neuropsychological features (age, years of education, and scores of MMSE, WAIS-III, and WMS-R) of the early MCI group are presented in Table 1, and are compared with those of the normal and MCI groups in Table 2. One-way anovas involving each feature showed no significant differences in age and years of education; however, they showed significant differences in several neuropsychological examination scores between the three groups. The early MCI group did not show significant memory impairment of 1.0 SD nor other cognitive dysfunctions on neuropsychological examinations, and did not fulfill the clinical diagnostic criteria of aMCI (Table 1). The MCI group showed significant memory impairment of 1.0 SD on neuropsychological examinations, fulfilling the clinical diagnostic criteria of aMCI,4,5 whereas the normal group had cognitive function scores above the average (Table 2). With Tukey's HSD post-hoc test, the early MCI group showed the highest scores for Verbal IQ (VIQ), Performance IQ (PIQ), Full Scale IQ (FIQ), and Verbal Comprehension (VC) in WAIS-III. In contrast, the MCI group showed the lowest scores for General Memory (GM) and Delayed Recall (DR) in WMS-R, although there were no significant differences between the early MCI and normal groups. All patients in the early MCI group scored more than 85 in both WAIS-III14 and WMS-R,18 whereas WMS-R score was lower than 85 with 1.00 SD and lower than 75 with 1.65 SD for the definition of aMCI.

In order to show a discrepancy in scores between WAIS-III and WMS-R, we subtracted GM or DR in WMS-R from VIQ, FIQ, or VC in WAIS-III (Table 2). Any discrepancy in scores was generally the highest in the MCI group and the lowest in the normal group. One way anovas on all discrepancies in scores showed significant differences among the three groups. Tukey's HSD post-hoc test showed that the normal group scored significantly lower in FIQ-GM and VC-GM than the other groups, although there were no significant differences between the MCI and early MCI groups. For the other four discrepancies in scores, there were significant differences among the three groups. When the accuracy of differentiating a discrepancy in scores in the early MCI group from those in the normal group was examined by the ROC curve, all discrepancies showed a high-level accuracy.

MRI showed no significant brain atrophy and/or vascular change based on age in the normal and Early MCI groups. In contrast, it showed mild atrophy of the hippocampus without significant vascular change in the MCI group. In order to evaluate the regional CMRglc decrease in 18F-FDG PET, the CMRglc in each patient was compared with the age-matched normative database using 3D-SSP analysis,20 and the CMRglc decrease was expressed as a Z-score and superimposed on the 3D-SSP maps. Consequently, a CMRglc decrease was recognized in the posterior cingulate gyrus and/or part of the parietotemporal area in both the MCI and early MCI groups, of which the extent and magnitude were weaker in the early MCI group (Fig. 1). The normal group did not show a significant CMRglc decrease.

image

Figure 1. 18F-fluorodeoxyglucose positron emission tomography images using 3D-SSP analysis in the normal, early mild cognitive impairment (MCI) and MCI groups. N1-6, normal group; E1-8, early MCI group; M1-10, MCI group.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES

In the present study, the early MCI group met the clinical diagnostic criteria of aMCI,4,5 presented memory complaints recognized by themselves and/or their families, were not demented and showed normal functional activities. Although no significantly objective memory impairment was detected with WMS-R, there was a significant discrepancy between scores of WAIS-III and WMS-R, similar to the MCI group. Furthermore, a CMRglc decrease in the posterior cingulate gyrus and/or part of the parietotemporal area was observed on 18F-FDG PET examination, similar to the MCI group. These findings suggest that the early MCI group might be differentiated from the normal group, which shows a normal memory decline as a result of aging, and that the early MCI group might correspond to early-stage aMCI, although it does not fulfill the clinical diagnostic criteria of aMCI. It should be noted that the early MCI group might have been more sensitive to memory decline than normal elderly subjects, because it showed the highest intelligence scores among the three groups and more discrepancy in memory and intelligence scores than the normal group.

In the current diagnostic criteria of aMCI, it is recommended to assess memory impairment with memory examinations including WMS-R, as well as to consider age and educational level.4,5 In the present study, the age and educational level were controlled among the three groups. However, WMS-R scores of the early MCI group were average and did not show a decline of more than 1.0 SD. These findings show that the early MCI group cannot be diagnosed with aMCI when only the age and educational level are considered.

The early MCI group showed high scores for WAIS-III, but significantly lower WMS-R than WAIS-III scores, although there was little difference between WAIS-III and WMS-R scores in the normal group. There were several discrepancies between WAIS-III and WMS-R scores in the early MCI group, and some of these discrepancies showed high-level accuracy for differentiating the early MCI from the normal group. These findings suggest that early-stage aMCI might be diagnosed by comparing memory scores of WMS-R with intelligence scores of WAIS-III, when memory scores of WMS-R do not show significant impairment in consideration of age and educational level. However, we could not determine which discrepancies are particularly useful for diagnosing early-stage aMCI in the present study. Further studies are necessary to determine useful discrepancies by carrying out large random sampling involving normal elderly subjects and aMCI patients, and setting a cut-off point for identifying early-stage aMCI with high-level accuracy.

Assessment of premorbid cognitive function is needed to identify cognitive impairment. However, there are few opportunities to assess premorbid memory function. As shown in the present study, there is a risk of missing early-stage aMCI if only a memory decline of 1.0 or 1.5 SD from the average score of normal elderly subjects is used for a diagnosis of aMCI. The scores of VIQ and VC do not easily decline, even in the presence of organic brain syndromes, such as AD;21,22 therefore, these scores can be reliable for the assessment of a premorbid memory function that might be assessable using an intelligence examination such as WAIS-III.13,23

The current diagnostic criteria of aMCI have another risk of including normal elderly subjects with aMCI patients. Such an error might occur when normal elderly subjects show a lower educational level or intelligence,12 and might be avoidable by using intelligence examinations such as WAIS-III. Although memory examinations such as WMS-R are useful for diagnosing aMCI, both memory and intelligence examinations are necessary to facilitate a specific diagnosis.

The early MCI group should be included in aMCI not only based on the discrepancy between intelligence and memory scores, but also based on the 18F-FDG PET findings. A number of studies24–27 have shown the utility of 18F-FDG PET for diagnosing aMCI. The early MCI group also showed a decreasing CMRglc pattern similar to that shown by previous 18F-FDG PET studies of aMCI. In Japan, there are few medical institutions where 18F-FDG PET examination in addition to neuropsychological examination and clinical consultation can be utilized to differentiate aMCI patients from normal elderly subjects or AD patients. The combination of these examinations would make it possible to diagnose early-stage aMCI, as shown in the present study.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES

This study was supported in part by the research center of Health and Sports Science, Juntendo University School of Medicine, and The Ogasawara Foundation for the Promotion of Science and Engineering.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. REFERENCES