SEARCH

SEARCH BY CITATION

Keywords:

  • Alzheimer's disease;
  • Anosognosia Questionnaire for Dementia;
  • behavioral symptoms;
  • cognitive deficits;
  • Japanese patients;
  • reliability;
  • validity

Abstract

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

Abstract  Although a number of studies have examined anosognosia of cognitive deficits in patients with Alzheimer's disease (AD), not much is known about the anosognosia of behavioral symptoms in AD. The aims of the present study were to establish a Japanese version of the Anosognosia Questionnaire–Dementia (AQ-D) and to examine its factor structure, reliability and validity, and to identify the effects of various variables on the AQ-D. Factor structure, internal consistency, test–retest reliability and concurrent validity of the Japanese version of the AQ-D were analyzed. Multiple regression was then done using the results of the AQ-D as dependent variables and entering all relevant predictor variables. Both the internal consistency and the test–retest reliability of the AQ-D were excellent. Factor analysis indicated four factors: anosognosia of basic and instrumental activities of daily living; that of episodic memory and orientation; that of disinhibited behaviors; and that of apathy and depression. The first two factors were regarded as anosognosia of cognitive deficits and were associated with Mini-Mental State Examination scores, while the latter two factors were regarded as anosognosia of behavioral symptoms and were associated with the Neuropsychiatric Inventory (NPI) score. A dissociation between the two domains of anosognosia was confirmed, namely of cognitive deficits and of behavioral symptoms using the Japanese version of the AQ-D. The knowledge that various factors may have different effects on different domains of anosognosia in patients with AD may serve as useful information for clinicians assessing anosognosia in AD.


INTRODUCTION

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

In patients with Alzheimer's disease (AD), all domains of insight are not always lost. Starkstein et al. divided the domains of anosognosia (loss of insight) into two domains: lack of cognitive awareness and lack of behavioral awareness.1 Recently they extended these findings by analyzing four domains of anosognosia in a large series of patients with AD.2 While a number of studies on dementia have examined aspects related to the loss of insight into cognitive deficits, less is known about the loss of insight into behavioral disturbances. Neuropsychiatric symptoms (behavior disturbances) are commonly encountered in AD patients, and may not only increase the distress of caregivers but also affect the patients' own distress levels. Thus, it is important to investigate the clinical significance of anosognosia of behavior disturbances in patients with AD dementia.

However, to the best of our knowledge, a standardized anosognosia questionnaire covering both the cognitive and behavioral domains is not available in Japan. The aim of the present study was therefore to develop the Japanese version of the Anosognosia Questionnaire–Dementia (AQ-D), originally devised by Migliorelli et al.,3 to examine its factor structure, and to confirm its reliability and validity. We also examined the effects of clinical variables, such as cognitive function and the presence of neuropsychiatric symptoms, on these domains of anosognosia in patients with AD. We hypothesized that the two domains, cognitive deficits and behavioral deficits, could also be validated in the Japanese version of the AQ-D.

METHODS

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

Participants

One hundred and forty-three consecutive Japanese outpatients with AD who visited Nagoya City University Hospital were recruited. The inclusion criteria consisted of (i) diagnosis of probable AD according to the National Institute of Neurology and Communicative Disorder and Stroke/Aizheimer's Disease and Related Disorders Association (NINCDS/ADRDA) criteria;4 and (ii) very mild to mild functional severity (grade 0.5–1 on clinical dementia rating [the CDR]5). Patients were excluded if (i) other neurological diseases were present; (ii) there was a previous history of mental illness; (iii) focal brain lesions were seen on magnetic resonance imaging; or (iv) Mini-Mental State Examination (MMSE)6 score was <11. Each patient's caregiver was defined as the principal family member providing day-to-day care.

The study protocol was approved by the Ethics Committee of Nagoya City University Graduate School of Medical Sciences. The examiner explained the study protocol including the study purpose, procedures and so on using simple and straightforward language to both AD patients and their caregivers. The examiner repeatedly explained the protocol to all AD patients and their caregivers as often as required to ensure adequate comprehension of the objective of the study. After a complete description of the study protocol, written informed consent was obtained from all AD patients and their caregivers.

Translation of AQ-D

The AQ-D is a 30-item questionnaire seeking responses from both the patients and their caregivers about the current level of impairment of the patients. With the original authors' permission, we translated the original English version into Japanese. We followed a standard back-translation procedure to ascertain the semantic equivalence of the Japanese version with the original English version.

Procedure

In accordance with the procedure recommended for the original AQ-D, the AD patients and their caregivers completed the questionnaire independently in different rooms. The AQ-D consisted of 30 questions covering the patient's insight of two domains (intellectual functioning and behavior). Each answer was rated as follows: never (0 points), sometimes (1 point), usually (2 points), or always (3 points). Higher scores indicated a more severe impairment of insight. The final score was obtained by subtracting the score on the patient's report version from that of the caregiver's report version. Therefore, a positive final AQ-D score indicated that the caregiver had rated the patient as more impaired than the patient had.

At the time of administering the AQ-D questionnaire, the following tests were also conducted.

(1) Neuropsychiatric Inventory (NPI):7 a semiquantitative assessment based on information provided by the caregiver. This interview consists of 10 behavior-related questions. Hirono et al. confirmed the reliability and validity of the Japanese version of this interview.8

(2) Hyogo Activities of Daily Living Scale (HADL): a scale designed for a detailed and comprehensive assessment of the functional abilities of subjects with AD.9 It consists of 18 questions related to basic and instrumental activities of daily living (ADL). Assessment using this scale is also based on information provided by the caregiver. The reliability and validity of this questionnaire was confirmed by Hirono et al.9

Statistical analysis

We used SPSS 11.0 J for Windows (SPSS, Chicago, IL, USA) for the statistical analysis.

Factor analysis

A principal component factor analysis using varimax rotation was performed on the 30 items of the AQ-D. The models included factors with an eigenvalue >1. An item was considered to load onto a factor if its factor loading score exceeded 0.30. In addition, we calculated the congruence coefficients for the components after varimax rotation between the present results and the Starkstein et al. results.2 The congruence coefficients range from −1 to +1, and values >0.90 are usually considered very high.

Reliability

The reliability of this scale was assessed in two ways. First, the test–retest reliability was assessed among a subset of 85 patient–caregiver pairs 1 month after the initial evaluation, using an analysis of variance intraclass correlation coefficient (anova ICC). In general, an anova ICC >0.7 indicates good reliability. Second, the internal consistency of the scale was estimated by determining Cronbach's alpha coefficients. Cronbach's alpha >0.70 is indicative of a good internal consistency.10

Validity

To examine the concurrent validity, we assessed the correlation between the Japanese version of the AQ-D score and the criteria described by Reed et al.11 One of the authors (N.S.), an experienced psychiatrist, rated the patient's level of insight (4-point scale) without knowledge of the results of the AQ-D score.

Prediction of the AQ-D score

We conducted a multiple regression analysis to explore the correlation between the AQ-D score and cognitive or behavior symptoms and ADL functioning. The dependent variable was the final AQ-D score. The independent variables were the duration of AD (in years), the MMSE score, the NPI score, and the HADL score. We used the three dimensions (mood factor, psychosis factor, and euphoria factor) of the NPI score as demonstrated in our previous study.12

RESULTS

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

Subject characteristics

Table 1 presents the mean scores and standard deviations of the clinical and demographic characteristics. Among the caregivers (n = 143), 58% were spouses, 19% were daughters-in-law, 18.1% were adult children, and 5.8% were other close relatives.

Table 1.  Subject data
 AD patients Mean ± SD (n = 143)Caregivers Mean ± SD (n = 143)
  1. AD, Alzheimer's disease; MMSE, Mini Mental State Examination.

Gender (male/female)59/8456/87
Age (years)72.1 ± 7.361.5 ± 11.2
Education (years)9.6 ± 0.810.4 ± 2.0
Duration of dementia (years)1.9 ± 0.5(–)
MMSE20.1 ± 4.3(–)

Factor analysis

An exploratory principal component analysis reduced the 30 variables to five factors. The five factors explained 64.2% of the variance in the data. The first four factors were retained because of their clinical relevance. However, the fifth factor was composed of only one item. Thus, a four-factor model best fitted the Japanese version of the AQ-D (Table 2). The first factor loaded on the items corresponding to ADL abilities. This factor was termed ‘anosognosia of basic and instrumental ADL’. The second factor included most of the items corresponding to episodic memory and orientation. Therefore, we named this factor anosognosia of episodic memory and orientation. The items in the third factor corresponded to increased irritability, laughing inappropriately. Therefore, the third factor was named anosognosia of disinhibited behavior. The fourth factor included such items as less interest in favorite things, more depressed. Therefore, the fourth factor was named anosognosia of apathy and depression. The factor congruence coefficients for the four factors between the present results and those derived from Starkstein et al.2 were between 0.62 and 0.88 (Table 3). The factor congruence coefficient between the two studies was either substantive or moderate. Korth suggested that congruence coefficients tend to become smaller when larger numbers of factors are extracted,13 as in the present study. Therefore, we consider that the observed factors in the present study were replicated in the original factor model.

Table 2.  Factor analysis for 143 AD patients
Anosognosia characteristicsFactor 1Factor 2Factor 3Factor 4
  1. A, Intellectual Functions; B, Behavior. Bold, loadings ≥0.30.

  2. AD, Alzheimer's disease.

A-21. Problems with feeding oneself0.8060.1480.0590.062
A-14. Problems with communicating with people0.7610.2010.0660.086
A-15. Problems doing mental calculations0.7390.0870.0130.092
A-20. Problems with doing home activities0.6860.3880.0010.117
A-19. Problems with orientation in the house0.6580.3910.0410.088
A-22. Problems with doing clerical work0.6190.3360.1900.218
A-18. Problems with understanding the plot of a movie0.6150.4980.0210.109
A-13. Problems with practicing favorite hobbies0.5860.3100.0690.144
A-10. Problems with handling money0.3890.261−0.0310.287
A-17. Problems with bladder control0.3500.2530.273−0.055
A-16. Problems with remembering shopping lists0.5550.5230.0700.01
A-11. Problems with orientation in your neighborhood0.5110.4890.0580.232
A-6. Problems with understanding the newspaper0.1830.6750.0430.269
A-3. Problems with remembering telephone calls0.1380.6640.0260.179
A-2. Problems with orientation in new places0.2750.6640.0990.055
A-5. Problems with signing your name0.4070.652−0.1040.239
A-1. Problems with remembering dates0.3540.6370.1790.061
A-4. Problems with understanding conversations0.3410.630−0.0810.179
A-8. Problems with remembering where things were left0.1750.6270.2110.116
A-7. Problems with keeping belongings in order0.3890.4700.0310.208
A-12 Problems with remembering appointments0.4200.4700.1260.097
B-27. Laughing inappropriately0.0110.0540.7950.110
B-28. Increased sexual interest0.002−0.0830.7080.096
B-26. More frequent crying episodes0.061−0.0390.6260.181
B-25. More irritable0.1410.2230.5280.297
B-24. More egotistical and self-centered0.2100.1910.3150.641
B-29. Less interest in favorite activities0.0200.1370.5550.633
B-30. More depressed0.0640.2180.4600.608
B-23. More rigid and inflexible about decisions0.1690.2440.2160.543
A-9. Problems with writing0.3660.430−0.0640.190
Table 3.  Factor congruence coefficients for the present and Starkstein et al. studies2
Factor structureFirst factorSecond factorThird factorFourth factor
  1. Vertical: factor structure for the Starkstein et al. study;2 horizontal: factor structure for the present study.

  2. First factor, basic and instrumental ADL; second factor, episodic memory and orientation; third factor, disinhibition; fourth factor, apathy and depression.

  3. bADL, basic activities of daily living; iADL, instrumental activities of daily living.

Factor 1 (iADL)0.621   
Factor 2 (bADL) 0.644  
Factor 3 (depression)   0.833
Factor 4 (disinhibition)  0.655 

Reliability

The alpha coefficients for the patient responses (0.90; 95% confidence interval [CI], 0.88–0.92) and the caregiver responses (0.92; 95%CI, 0.88–0.93) indicated an excellent internal consistency for all 30 items of the AQ-D. The test–retest reliability (n = 85) of the total score after 1 month was excellent, with a good ICC (0.81; 95%CI, 0.72–0.88).

Validity

We observed a statistically significant and substantive correlation between the clinical assessment of anosognosia in AD according to the Reed et al. criteria11 and the final AQ-D score (= −0.71, P < 0.01).

Prediction of the AQ-D score

While multiple regression analysis showed that the MMSE and HADL scores predicted the score for anosognosia of cognitive deficits, the NPI scores (mood, psychosis, and euphoria factors) and the HADL scores predicted the score for anosognosia of behavior symptoms ( Tables 4,5).

Table 4.  Variables predicting score for anosognosia of cognitive deficits (first and second factor scores) on AQ-D
VariablesβtP
  • *

     < 0.01,

  • **

     < 0.001.

  • β, standardized regression coefficient.

  • AQ-D, Japanese version of Anosognosia Questionnaire–Dementia; HADL, Hyogo Activities of Daily Living Scale; NPI, Neuropsychiatric Inventory.

Duration of illness (years)−0.016−0.4190.676
MMSE score−0.329−3.3480.001*
HADL0.6086.119<0.001**
NPI
 Mood factor−0.069−1.6650.098
 Euphoria factor−0.038−0.8830.379
 Psychosis factor−0.055−1.3280.186
Table 5.  Variables predicting score for anosognosia of behavior deficits (third and fourth factor scores) on AQ-D
VariablesβtP
  • *

     < 0.05,

  • **

     < 0.01.

  • β, standardized regression coefficient.

  • AQ-D, Japanese version of Anosognosia Questionnaire–Dementia; HADL, Hyogo Activities of Daily Living Scale; NPI, Neuropsychiatric Inventory.

Duration of illness (years)−0.04−0.6740.501
MMSE score−0.123−0.8180.415
HADL0.4983.2750.001**
NPI
 Mood factor0.1812.870.005**
 Euphoria factor0.1422.1880.03*
 Psychosis factor0.1442.2720.025*

DISCUSSION

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

We confirmed that the Japanese version of the AQ-D had good internal and external reliability andsatisfactory concurrent validity. Four factors of the AQ-D were identified in the present study: basic and instrumental ADL; episodic memory and orientation; disinhibition; and apathy and depression. In addition, while the MMSE and HADL scores predicted the scores for the domains of anosognosia of cognitive deficits on the AQ-D, the NPI score predicted the scores for the domains of anosognosia of behavioral symptoms on the AQ-D. As shown in Table 3, the factor congruence coefficient between the first factor in the present study and that in the Starkstein et al. study2 was 0.621; that between the second factor in the present study and that in the Starkstein et al. study2 was 0.644. The factor congruence coefficient is an index of similarity between factor loadings. Because factor congruence coefficients for the first and second factor between the two studies are modest, we can assume that there was good convergence of the first two factors between the present study and that by Starkstein et al.2 In addition, the factor congruence coefficient for the third and fourth factors between the present and the Starkstein et al. studies2 were either high or modest. Therefore, the present study suggests that the four-factor structure demonstrated by Starkstein et al.2 was replicated in the present sample.

In contrast, we failed to observe the two different domains of basic ADL (bADL) and instrumental ADL (iADL) observed in the Starkstein et al . study.2 Starkstein et al. demonstrated the four-factor structure of anosognosia in a large series of 750 patients with very mild to severe stage AD.2 In that study the cognitive domains could be clustered into two subdomains, namely, bADL and iADL, and behavioral domains could also be clustered into two subdomains, namely, depression, and disinhibition. Unlike the present study, the patient series examined by Starkstein et al. included not only patients with mild AD, but also patients with moderate or severe AD.2 They found a marked difference in the anosognosia score for the bADL domain between the moderate and severe stage. Tekin et al. suggested that while iADL impairment can be an early sign of dementia, bADL ability was maintained until late in the course of AD disease.14 Thus, the subtle difference between the present results of those of Starkstein et al.2 may be related to differences in the clinical severity among the two series of patients. Larger study groups of Japanese patients with all stages of AD will clarify the phenomenon of anosognosia into two different levels of ADL.

Although we named the second factor ‘episodic memory and orientation’, some items in this factor may correspond to the ability to perform bADL tasks. However, we could not term the second factor ‘bADL’ because the items in the second factor (except for ‘problems with signing your name’) did not include items among the bADL domains defined by Starkstein et al.2 Several studies of patients with AD have suggested that ADL ability (particularly iADL ability) was associated with executive dysfunction.15,16 The present data indicate that executive function may be associated with iADL (i.e. doing clerical work, handling money) in the first factor because the activities in these items are complex. However, episodic memory and orientation, rather than executive function, may contribute to the activities of items in the second factor (e.g. remembering the date, remembering where things were left). Therefore, different cognitive mechanisms likely underlie activities associated with ADL between the first and second factors in the present study.

The third and fourth factors identified in the present study were identical to the two factors, anosognosia for depression and for disinhibition, reported by Starkstein et al.2 Unlike their study, we named the third factor ‘apathy and depression’ because apathy and depression may be distinct syndromes in patients with dementia.17 Because the fourth factor included items corresponding to inappropriate activities, we termed this factor as ‘disinhibited behavior’; this factor was identical to the fourth factor identified by Starkstein et al.2

Results from multiple regression support the existence of two domains of anosognosia in subjects with AD. While the score for anosognosia of cognitive deficits was correlated with the MMSE and HADL scores, the presence of neuropsychiatric symptoms identified by the NPI was correlated with the score for anosognosia of behavior deficits. The present study may also provide additional evidence supporting a previous hypothesis that anosognosia is not a consequence of cognitive deficits but is a wider behavioral disorder,1–3 because the present study included patients with very mild or mild stages of AD. Several caveats should be addressed. First, although we confirmed the reliability and validity of the Japanese version of the AQ-D, the information obtained from the caregivers may have been influenced by various factors (i.e. reliability of emotional stress, and cognitive level).2,18 Therefore, information obtained from the caregivers should be used with caution. Second, previous studies have identified deficits associated with specific neuropsychological findings (i.e. frontal lobe function).2,19 Unfortunately we could not conduct comprehensive neuropsychological tests, except for the MMSE.

In conclusion the present study demonstrated that the Japanese version of the AQ-D was both reliable and valid for a large number of patients with AD. More importantly, the present study demonstrated a dissociation between the two domains of anosognosia, cognitive and behavior deficits, on the Japanese version of the AQ-D. We expect the Japanese version of the AQ-D to become a clinically useful standard assessment of anosognosia in patients with AD.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES
  • 1
    Starkstein SE, Sabe L, Chemerinski E, Jason L, Leiguarda R. Two domains of anosognosia in Alzheimer's disease. J. Neurol. Neurosurg. Psychiatry 1996; 61: 485490.
  • 2
    Starkstein SE, Jorge R, Mizrahi R, Robinson RG. A diagnostic formulation for anosognosia in Alzheimer's disease. J. Neurol. Neurosurg. Psychiatry 2006; 77: 719725.
  • 3
    Migliorelli R, Teson A, Sabe L et al. Anosognosia in Alzheimer's disease: A study of associated factors. J. Neuropsychiatry Clin. Neurosci. 1995; 7: 338344.
  • 4
    McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer's disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 1984; 34: 939944.
  • 5
    Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new clinical scale for the staging of dementia. Br. J. Psychiatry 1982; 140: 566572.
  • 6
    Folstein MF, Folstein SE, McHugh PR. ‘Mini-mental state’. A practical method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res. 1975; 12: 189198.
  • 7
    Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, Gornbein J. The Neuropsychiatric Inventory: Comprehensive assessment of psychopathology in dementia. Neurology 1994; 44: 23082314.
  • 8
    Hirono N, Mori E, Ikejiri Y et al. Japanese version of the neuropsychiatric inventory: A scoring system for neuropsychiatric disturbance in dementia patients. No To Shinkei 1997; 49: 266271.
  • 9
    Hirono N, Yamadori A, Mori E, Yamashita H, Tokimasa A. Daily living activities of patients with Alzheimer's disease. Jpn. J. Neuropsychol. 1995; 11: 186195.
  • 10
    Nunnally JC. Psychometrics Theory, 2nd edn. McGraw-Hill, New York, 1978.
  • 11
    Reed BR, Jagust WJ, Coulter L. Anosognosia in Alzheimer's disease: Relationships to depression, cognitive function, and cerebral perfusion. J. Clin. Exp. Neuropsychol. 1993; 15: 231244.
  • 12
    Matsui T, Nakaaki S, Murata Y et al. Determinants of the quality of life in Alzheimer's disease patients as assessed by the Japanese version of the Quality of Life-Alzheimer's disease scale. Dement. Geriatr. Cogn. Disord. 2006; 21: 182191.
  • 13
    Korth B. The distribution of chance congruence coefficients from simulated date. Psychometrika 1975; 40: 361371.
  • 14
    Tekin S, Fairbanks LA, O'Connor S, Rosenberg S, Cummings JL. Activities of daily living in Alzheimer's disease: Neuropsychiatric, cognitive, and medical illness influences. Am. J. Geriatr. Psychiatry 2001; 9: 8186.
  • 15
    Boyle PA, Malloy PF, Salloway S et al. Executive dysfunction and apathy predict functional impairment in Alzheimer disease. Am. J. Geriatr. Psychiatry 2003; 11: 214221.
  • 16
    Stout JC, Wyman MF, Johnson SA, Peavy GM, Salmon DP. Frontal behavioral syndromes and functional status in probable Alzheimer disease. Am. J. Geriatr. Psychiatry 2003; 11: 683686.
  • 17
    Levy ML, Cummings JL, Fairbanks LA et al. Apathy is not depression. J. Neuropsychiatry Clin. Neurosci. 1998; 10: 314319.
  • 18
    Snow AL, Graham DP, Molinari VA et al. Factors affecting deficit awareness in persons with dementia. Dement. Geriatr. Cogn. Disord. 2005; 20: 133139.
  • 19
    Michon A, Deweer B, Agid Y, Dubois B. Relation of anosognosia to frontal lobe dysfunction in Alzheimer's disease. J. Neurol. Neurosurg. Psychiatry 1994; 57: 805809.