Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161 Received: August 22, 2014 Accepted: October 16, 2014 Published online: January 7, 2015
© 2015 S. Karger AG, Basel 1420–8008/15/0394–0194$39.50/0 www.karger.com/dem
Original Research Article
Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease Davide Quaranta Maria Gabriella Vita Alessandra Bizzarro Carlo Masullo Chiara Piccininni Guido Gainotti Camillo Marra Research Center for Neuropsychology, Institute of Neurology, Catholic University, Rome, Italy
Key Words Alzheimer’s disease · Psychosis · Delusions · Delusional misidentifications · Hallucinations · Neuropsychological assessment
Dr. Davide Quaranta Research Center for Neuropsychology, Institute of Neurology, Catholic University Largo A. Gemelli 8 IT–00168 Rome (Italy) E-Mail davide.quaranta @ rm.unicatt.it
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Abstract Aims: To investigate the relationship between psychotic symptoms and cognitive impairment in Alzheimer’s disease (AD). Methods: A total of 108 subjects affected by AD were subdivided into subjects without delusions (ND), subjects with paranoid delusions (PD), subjects with delusional misidentifications (DM), subjects with both DM and PD (DM+PD), subjects with visual hallucinations (v-HALL), and subjects without visual hallucinations (N-HALL). Results: PD and ND subjects performed similarly on neuropsychological tests, while DM patients performed significantly worse than PD and ND patients. v-HALL patients performed worse than N-HALL patients on memory, visuospatial, and executive functions. As for behavioral features, DM and v-HALL subjects reported higher scores on the abnormal motor behavior subscale of the neuropsychiatric inventory (NPI); PD subjects reported higher scores on the disinhibition subscale of the NPI. The severity of PD was predicted by the severity of disinhibition (B = 0.514; p = 0.016) but not by neuropsychological performances. The severity of DM was predicted by age (B = 0.099; p = 0.048) and MMSE (B = –0.233; p = 0.001). The severity of v-HALL was predicted by age (B = 0.052; p = 0.037) and scores on an immediate visual memory task (B = –0.135; p = 0.007). Conclusions: The occurrence of PD may require the relative sparing of cognitive functions and be favored by frontal lobe dysfunction, while DM is associated with the overall level of cognitive impairment. Finally, v-HALL are associated with the impairment © 2015 S. Karger AG, Basel of visuospatial abilities.
195
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Psychotic symptoms are a common feature of Alzheimer’s disease (AD), affecting about 30–40% of patients [1]. The presence of psychotic symptoms in AD has great clinical relevance, as they are one of the main causes of caregiver’s distress, institutionalization [2], and a risk factor for reduced survival [3]. Over the last two decades, growing evidence has supported the hypothesis that the presence of psychotic symptoms may configure an independent subtype of AD (AD + psychosis, AD-P), characterized by a specific neurobiological and genetic substrate [4]. Classical descriptions of psychosis in AD identified three main groups of symptoms, namely paranoid delusions (PD), delusional misidentifications (DM), and hallucinations (HALL) [5–7]. The clinical reliability of this subdivision was well confirmed by subsequent factor analysis studies conducted on subjects affected by AD-P [8]. Besides its clinical relevance, AD-P may also be an intriguing model as it may offer the possibility to investigate the connection between specific cognitive disturbances and the genesis of specific phenomena belonging to the core of the psychotic symptoms. The occurrence of AD-P as a whole has been associated with the progression of cognitive impairment, as assessed by general cognitive measures, such as the Mini Mental State Examination (MMSE) [9–13], with rare exceptions [14]. Several studies have compared the neuropsychological profile of AD-P subjects with that of nonpsychotic AD patients. Jeste et al. [15] have reported an association of AD-P with worse performances on measures of executive functions (Dementia Rating Scale subtests verbal fluency, conceptualization, and concentration, as well as the WAIS-R subtest analogies). These findings were confirmed, from a different point of view, by Swanberg et al. [16] who reported an association between executive dysfunction and AD-P. Hopkins and Libon [17] assessed the neuropsychological differences between psychotic and nonpsychotic subjects affected by AD or vascular dementia; their results indicate that subjects with delusions obtained better performances on a naming test, with lower scores on simple executive tests derived from the Wechsler Memory Scale. Paulsen et al. [18] assessed the relationship between frontal behavioral abnormalities and AD-P in subjects with severe AD, reporting a higher degree of disinhibition in psychotic subjects. This evidence supports the hypothesis that AD-P may be associated with frontal lobe dysfunction, as apparently confirmed by SPET, PET, and MRI studies [19–22]. On the basis of these results, it is conceivable to consider frontal lobe dysfunction at least as a supporting factor for the occurrence of AD-P when it is viewed as a unitary phenomenon. Nevertheless, the assessment of neuropsychological correlates of AD-P considered as a whole could be theoretically misleading, as the heterogeneous phenomenology of psychotic disturbances may reflect differences in their clinical, neuroanatomical, and cognitive substrates. To bypass this limitation, some studies considered HALL and delusions in their relationship with cognitive functioning separately. Becker et al. [23] found an association between the incidence of HALL and a more severe cognitive impairment, in particular in spatial orientation. Wilson et al. [24] reported subjects with visual and/or auditory HALL to have a faster decline in memory, visuoconstructive abilities, and language (repetition and naming). A previous factorial analysis on an ad hoc scale showed that the cluster represented by ‘phantom boarder sign/anthropomorphic visual hallucinations’ was associated with more pronounced alterations of visuoperceptual abilities, while ‘image misidentifications and auditory hallucinations’ were associated with a reduction in abstract reasoning. On the other hand, the cluster including delusion of abandonment, theft, and reference was not associated with specific neuropsychological deficits [8]. AD subjects affected by DM have been reported to display more severe cognitive impairment, mainly involving visuospatial functioning and verbal fluency, whereas subjects affected by isolated persecutory delusions did not differ significantly from delusion-free subjects [25].
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Introduction
196
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
The cognitive and behavioral profile of subjects affected by AD-P has not been completely clarified yet, probably because different methods have been used in previous studies to classify these patients. These inconsistencies may explain the different results obtained in studies considering psychosis in AD as a whole from those considering specific clusters of psychotic phenomena. Since the sources of specific psychotic symptoms could rely on specific cognitive disorders and the neuronal network beneath them, in the present study, we aim to explore the neuropsychological and behavioral features associated with PD, DM, and HALL in subjects affected by AD. On the basis of data reported in the neuropsychological literature and for pathophysiological reasons, the following predictions were made. (1) Patients with PD should be less cognitively impaired than those belonging to other psychopathological groups. The rationale of this prediction is that delusions constitute a wrong but systematic organization of thoughts and beliefs and that relatively spared cognitive functions are necessary to build a similar system. This prediction is based on data reporting patients with PD [8, 25], so as not to differ significantly from delusion-free subjects. (2) Patients with DM should also show a severe cognitive impairment, which could be particularly evident on executive tasks. This prediction stems from the fact that both in patients with focal brain damage [26] and in AD patients [27, 28], DM are usually associated with a prevalent impairment of the right frontal areas. (3) Patients with visual HALL (v-HALL) should show a more severe cognitive impairment, which could be particularly important in visuospatial tasks. The rationale of this prediction is that, contrary to auditory HALL, which are part of and reinforce delusional thoughts [29], v-HALL are frequent in delirium [30]. Furthermore, in patients with Lewy body dementia (LBD), v-HALL frequently co-occur with visuoperceptive and visuoconstructive disturbances [31]. Data consistent with this prediction can be found in previous studies [8, 23, 24]. Methods
Neuropsychiatric Evaluation Each subject underwent the 12-item version of the Neuropsychiatric Inventory (NPI) [33, 34], a caregiver-based clinical interview exploring 12 behavioral domains (delusions, HALL, agitation/aggression, depression, anxiety, euphoria, apathy, disinhibition, irritability, abnormal motor behavior, sleep, and appetite disturbances).
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Subjects A total of 108 consecutive patients were recruited from those referred to the Neuropsychology Unit of Policlinico Gemelli, Rome, Italy. All of them fulfilled the criteria for probable AD [32]. Diagnoses were formulated by two neurologists with a specific expertise in the field of dementia (C.M. and C.M.), and reviewed by a third expert (G.G.) who was blinded to the assigned diagnosis. Each subject underwent a complete medical and neurological examination. Functional (HMPAO-SPECT) and anatomical (MRI or CT scan) neuroimaging were performed on all subjects. Only subjects at their first clinical evaluation were enrolled, and behavioral and neuropsychological data were collected during the first visit, prior to any therapeutic prescription; this approach was chosen in order to bypass the possible effect of anti-dementia drugs (anticholinesterase drugs, memantine) on psychotic symptoms. Further inclusion criteria were the age range of 50–90 years, an MMSE score ≥10, and a Clinical Dementia Rating Scale score ≥1. Exclusion criteria were the absence of a caregiver informed enough to provide accurate clinical information and the assumption of antipsychotic medication in the month prior to our evaluation. This restriction was adopted in order to avoid an underdiagnosis of psychotic phenomena. Furthermore, subjects with symptoms compatible with an LBD diagnosis (evident Parkinsonism, significant fluctuations, vegetative dysfunction) were excluded. According to the criteria of the Local Ethics Committee and the Helsinki Declaration, all AD patients and/ or their proxy caregivers gave their informed consent to be examined and to undergo a full neuropsychological examination.
197
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Psychotic phenomena were coded either as dichotomous variables (present/absent) or according to the individual score (frequency × severity) of each symptom. Symptoms reported on the subscale delusions were grouped as follows: (a) PD, including delusions of theft, infidelity, abandonment, persecution, and (b) DM, including the belief that a family member is someone else, that one’s home is not one’s home, that images on TV or photographs are actually ongoing in the house, and the belief in the presence of ‘phantom boarders’. A total score was obtained for each category, summing up individual scores reported for each symptom (e.g. a subject obtaining a score of 2 on the delusion of theft, 3 on delusion of infidelity, and 1 on delusion of abandonment would obtain a total score of 6 on PD). It must be recognized that the definition of DM (particularly Capgras syndrome) adopted by classical psychopathology is usually more stringent than the one we used, as it requires a mismatch between what a person looks like and who the patient thinks he is, due to an associated persecutory ideation. That is, a person who at a first view could be identified as a family member is thought to be someone else, namely a specific persecutor. Nevertheless, current literature is quite homogeneous about the use of the more permissive definition of DM that we applied [7]. HALL were classified according to their sensory modality in ‘visual’, ‘auditory’, and ‘other modalities’. According to the hypothesis of possible different substrates for v-HALL and auditory HALL, total scores for the two conditions were considered separately. Neuropsychological Examination All participants underwent the MMSE and the following neuropsychological tests: Rey’s Auditory Verbal Learning Test (RAVLT) [35], Rey-Osterrieth’s Complex Figure (ROCF) – copy and recall [36], Raven’s Progressive Colored Matrices (RPM) [35], digit and visuospatial span [37], immediate visual memory [38], phonological verbal fluency [35], semantic verbal fluency [39], object naming [40], Stroop’s test (short version) [41], copy of object without and with landmarks [35], and the Multiple Features Targets Cancellation (MFTC), a test of visual search [42]. Statistical Analysis Most of the collected data displayed a nonnormal distribution as assessed by Shapiro-Wilk’s test. Thus, statistical analyses were conducted using nonparametric techniques. Frequency comparisons were conducted by means of the nonparametric χ2 test, with Yates’ continuity correction and Fisher’s exact test when required. Group comparisons for continuous variables were conducted by means of the Kruskal-Wallis or the Mann-Whitney U test as indicated. PD, DM, and HALL univariate linear regression analyses were conducted setting the total PD, DM, and HALL scores as dependent variables and single clinical and neuropsychological variables as possible predictors. Variables displaying a statistical significance in univariate models (p < 0.05) were entered into multiple-variable linear regression models assuming the total scores obtained in each symptom category as dependent variables, and controlling for age, educational level, illness duration, and MMSE score. Furthermore, since different types of psychotic symptoms are often present in the same subject, the regression model for each category was also controlled for the total scores obtained on the other categories (e.g. the regression model for PD was controlled for DM, and HALL total scores). The multiple-variable analyses were conducted using a backward stepwise technique.
The study sample included 108 AD patients (69 women; 64%), with a mean age of 73.3 years (SD: 7.62; range: 52–90) and a mean education of 8.0 years (SD: 4.59). The mean illness duration at observation was 42.6 months (SD: 23.3; range: 18–84) and the mean MMSE score was 17.2 (SD: 5.78; range: 10–24). In total, 48 (46%) subjects were affected by at least one psychotic symptom. Among the AD patients with psychotic symptoms, 42 (38.9%) displayed delusional symptoms, 16 (15.2% of the total sample) were affected by isolated PD, 11 (10.5%) by DM, and 15 (14.3%) by a combination of at least one paranoid and one misidentification symptom (DM+PD). Of the study sample, 23 (21.9%) patients displayed HALL. In most cases hallucinatory symptoms
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Results
198
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Table 1. Individual frequencies of psychotic symptoms
Symptom
n (%)
PD Persecution Stealing Jealousy Abandonment DM ‘A family member is someone else’ ‘One’s home is not one’s home’ ‘Images on TV are ongoing in the house’ ‘Phantom boarders’ HALL v-HALL Isolated auditory Other modalitiesa
10 (9.5) 25 (23.8) 5 (4.8) 10 (9.5) 12 (11.4) 16 (15.2) 6 (5.7) 10 (9.5) 21 (20) 1 (0.9) 3 (2.9)
The total percentage is >100, as several subjects were affected by more than one symptom. a Other modalities include olfactory (n = 2) and tactile/haptic HALL (n = 1).
Group Comparisons among Patients Showing No Delusions or Some Form of Delusional Symptoms Table 2 shows the differences in the neuropsychological and behavioral profiles of the four groups of patients. The groups differed significantly as for illness duration and MMSE score as well as in several neuropsychological measures. In general, DM subjects showed lower performances than the other groups, whereas PD subjects had better performances than the other groups to the same extent as ND subjects. On post hoc comparisons, PD patients did not differ significantly from ND patients, with the exception of digit span forward, in which they performed better (Mann-Whitney U test; p = 0.005). PD subjects were significantly younger than DM patients (p = 0.043), whereas their illness duration was shorter (p = 0.019). On neuropsychological testing, PD subjects performed better than DM patients on MMSE (p = 0.001), RAVLT immediate recall (p = 0.001) and recognition accuracy (p = 0.029), digit span backward (p = 0.045), spatial span forward (p = 0.008), RPM (p = 0.016), phonological verbal fluency (p = 0.009), and semantic verbal fluency (p = 0.001). The cognitive performance of PD patients was also generally better than that of DM+PD patients; statistical significance was reached for RAVLT immediate recall (p = 0.037) and spatial span backward (p = 0.002).
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
were associated with DM [5/11 (45%) patients] or DM+PD [9/15 (60%) patients]; more rarely, they were reported in association with PD [3/16 (19%) patients]. Only 6/66 (9%) subjects without delusional symptoms were affected by HALL (χ2 = 22.03; p < 0.001). There were no differences in the distribution of delusional symptoms with respect to gender (χ2 = 2.855; p = 0.827). Data on the prevalence of individual psychotic phenomena are presented in table 1. The vast majority of subjects with HALL (21/23, 91%) were affected by v-HALL. Vocalizations or verbal expression directed toward the subject of the HALL were reported in 10 subjects; nevertheless, in none of these cases the patients seemed to experience simultaneous auditory HALL. Since the subjects affected by auditory or other HALL constituted a very small group, further analyses were carried out only on v-HALL.
199
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Table 2. Cognitive and behavioral profile of subjects subdivided on the basis of delusional symptoms
Age, years Education, years Illness duration, months MMSE score Neuropsychological assessment RAVLT Immediate recall Delayed recall Recognition (accuracy) Recognition (false alarms) ROCF Copy Delayed recall RPM Immediate visual memory Digit span forward Digit span backward Spatial span forward Spatial span backward Phonological verbal fluency Semantic verbal fluency Object naming Copy of figures Copy of figures with landmarks MFCT Accuracy False alarms Time of execution Stroop’s test Interference errors Interference time NPI Agitation Depression Anxiety Euphoria/dysphoria Apathy Disinhibition Irritability Abnormal motor behavior Sleep disturbances Appetite/alimentary disturbances Total score
ND (n = 66)
PD (n = 16)
DM (n = 11)
DM+PD (n = 15)
p*
72.24 (7.21) 8.03 (4.73) 39.78 (17.95) 18.95 (5.33)
70.56 (9.46) 8.87 (4.33) 39.03 (15.49) 18.31 (4.53)
78.18 (6.37)†, ‡ 8.45 (4.76) 64.54 (30.14)†, ‡ 12.09 (3.27)†, ‡
76.00 (6.51) 6.73 (4.45) 36.67 (17.45)§ 14.47 (4.52)†
0.056 0.429 0.030
© 2015 S. Karger AG, Basel 1420–8008/15/0394–0194$39.50/0 www.karger.com/dem
Original Research Article
Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease Davide Quaranta Maria Gabriella Vita Alessandra Bizzarro Carlo Masullo Chiara Piccininni Guido Gainotti Camillo Marra Research Center for Neuropsychology, Institute of Neurology, Catholic University, Rome, Italy
Key Words Alzheimer’s disease · Psychosis · Delusions · Delusional misidentifications · Hallucinations · Neuropsychological assessment
Dr. Davide Quaranta Research Center for Neuropsychology, Institute of Neurology, Catholic University Largo A. Gemelli 8 IT–00168 Rome (Italy) E-Mail davide.quaranta @ rm.unicatt.it
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Abstract Aims: To investigate the relationship between psychotic symptoms and cognitive impairment in Alzheimer’s disease (AD). Methods: A total of 108 subjects affected by AD were subdivided into subjects without delusions (ND), subjects with paranoid delusions (PD), subjects with delusional misidentifications (DM), subjects with both DM and PD (DM+PD), subjects with visual hallucinations (v-HALL), and subjects without visual hallucinations (N-HALL). Results: PD and ND subjects performed similarly on neuropsychological tests, while DM patients performed significantly worse than PD and ND patients. v-HALL patients performed worse than N-HALL patients on memory, visuospatial, and executive functions. As for behavioral features, DM and v-HALL subjects reported higher scores on the abnormal motor behavior subscale of the neuropsychiatric inventory (NPI); PD subjects reported higher scores on the disinhibition subscale of the NPI. The severity of PD was predicted by the severity of disinhibition (B = 0.514; p = 0.016) but not by neuropsychological performances. The severity of DM was predicted by age (B = 0.099; p = 0.048) and MMSE (B = –0.233; p = 0.001). The severity of v-HALL was predicted by age (B = 0.052; p = 0.037) and scores on an immediate visual memory task (B = –0.135; p = 0.007). Conclusions: The occurrence of PD may require the relative sparing of cognitive functions and be favored by frontal lobe dysfunction, while DM is associated with the overall level of cognitive impairment. Finally, v-HALL are associated with the impairment © 2015 S. Karger AG, Basel of visuospatial abilities.
195
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Psychotic symptoms are a common feature of Alzheimer’s disease (AD), affecting about 30–40% of patients [1]. The presence of psychotic symptoms in AD has great clinical relevance, as they are one of the main causes of caregiver’s distress, institutionalization [2], and a risk factor for reduced survival [3]. Over the last two decades, growing evidence has supported the hypothesis that the presence of psychotic symptoms may configure an independent subtype of AD (AD + psychosis, AD-P), characterized by a specific neurobiological and genetic substrate [4]. Classical descriptions of psychosis in AD identified three main groups of symptoms, namely paranoid delusions (PD), delusional misidentifications (DM), and hallucinations (HALL) [5–7]. The clinical reliability of this subdivision was well confirmed by subsequent factor analysis studies conducted on subjects affected by AD-P [8]. Besides its clinical relevance, AD-P may also be an intriguing model as it may offer the possibility to investigate the connection between specific cognitive disturbances and the genesis of specific phenomena belonging to the core of the psychotic symptoms. The occurrence of AD-P as a whole has been associated with the progression of cognitive impairment, as assessed by general cognitive measures, such as the Mini Mental State Examination (MMSE) [9–13], with rare exceptions [14]. Several studies have compared the neuropsychological profile of AD-P subjects with that of nonpsychotic AD patients. Jeste et al. [15] have reported an association of AD-P with worse performances on measures of executive functions (Dementia Rating Scale subtests verbal fluency, conceptualization, and concentration, as well as the WAIS-R subtest analogies). These findings were confirmed, from a different point of view, by Swanberg et al. [16] who reported an association between executive dysfunction and AD-P. Hopkins and Libon [17] assessed the neuropsychological differences between psychotic and nonpsychotic subjects affected by AD or vascular dementia; their results indicate that subjects with delusions obtained better performances on a naming test, with lower scores on simple executive tests derived from the Wechsler Memory Scale. Paulsen et al. [18] assessed the relationship between frontal behavioral abnormalities and AD-P in subjects with severe AD, reporting a higher degree of disinhibition in psychotic subjects. This evidence supports the hypothesis that AD-P may be associated with frontal lobe dysfunction, as apparently confirmed by SPET, PET, and MRI studies [19–22]. On the basis of these results, it is conceivable to consider frontal lobe dysfunction at least as a supporting factor for the occurrence of AD-P when it is viewed as a unitary phenomenon. Nevertheless, the assessment of neuropsychological correlates of AD-P considered as a whole could be theoretically misleading, as the heterogeneous phenomenology of psychotic disturbances may reflect differences in their clinical, neuroanatomical, and cognitive substrates. To bypass this limitation, some studies considered HALL and delusions in their relationship with cognitive functioning separately. Becker et al. [23] found an association between the incidence of HALL and a more severe cognitive impairment, in particular in spatial orientation. Wilson et al. [24] reported subjects with visual and/or auditory HALL to have a faster decline in memory, visuoconstructive abilities, and language (repetition and naming). A previous factorial analysis on an ad hoc scale showed that the cluster represented by ‘phantom boarder sign/anthropomorphic visual hallucinations’ was associated with more pronounced alterations of visuoperceptual abilities, while ‘image misidentifications and auditory hallucinations’ were associated with a reduction in abstract reasoning. On the other hand, the cluster including delusion of abandonment, theft, and reference was not associated with specific neuropsychological deficits [8]. AD subjects affected by DM have been reported to display more severe cognitive impairment, mainly involving visuospatial functioning and verbal fluency, whereas subjects affected by isolated persecutory delusions did not differ significantly from delusion-free subjects [25].
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Introduction
196
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
The cognitive and behavioral profile of subjects affected by AD-P has not been completely clarified yet, probably because different methods have been used in previous studies to classify these patients. These inconsistencies may explain the different results obtained in studies considering psychosis in AD as a whole from those considering specific clusters of psychotic phenomena. Since the sources of specific psychotic symptoms could rely on specific cognitive disorders and the neuronal network beneath them, in the present study, we aim to explore the neuropsychological and behavioral features associated with PD, DM, and HALL in subjects affected by AD. On the basis of data reported in the neuropsychological literature and for pathophysiological reasons, the following predictions were made. (1) Patients with PD should be less cognitively impaired than those belonging to other psychopathological groups. The rationale of this prediction is that delusions constitute a wrong but systematic organization of thoughts and beliefs and that relatively spared cognitive functions are necessary to build a similar system. This prediction is based on data reporting patients with PD [8, 25], so as not to differ significantly from delusion-free subjects. (2) Patients with DM should also show a severe cognitive impairment, which could be particularly evident on executive tasks. This prediction stems from the fact that both in patients with focal brain damage [26] and in AD patients [27, 28], DM are usually associated with a prevalent impairment of the right frontal areas. (3) Patients with visual HALL (v-HALL) should show a more severe cognitive impairment, which could be particularly important in visuospatial tasks. The rationale of this prediction is that, contrary to auditory HALL, which are part of and reinforce delusional thoughts [29], v-HALL are frequent in delirium [30]. Furthermore, in patients with Lewy body dementia (LBD), v-HALL frequently co-occur with visuoperceptive and visuoconstructive disturbances [31]. Data consistent with this prediction can be found in previous studies [8, 23, 24]. Methods
Neuropsychiatric Evaluation Each subject underwent the 12-item version of the Neuropsychiatric Inventory (NPI) [33, 34], a caregiver-based clinical interview exploring 12 behavioral domains (delusions, HALL, agitation/aggression, depression, anxiety, euphoria, apathy, disinhibition, irritability, abnormal motor behavior, sleep, and appetite disturbances).
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Subjects A total of 108 consecutive patients were recruited from those referred to the Neuropsychology Unit of Policlinico Gemelli, Rome, Italy. All of them fulfilled the criteria for probable AD [32]. Diagnoses were formulated by two neurologists with a specific expertise in the field of dementia (C.M. and C.M.), and reviewed by a third expert (G.G.) who was blinded to the assigned diagnosis. Each subject underwent a complete medical and neurological examination. Functional (HMPAO-SPECT) and anatomical (MRI or CT scan) neuroimaging were performed on all subjects. Only subjects at their first clinical evaluation were enrolled, and behavioral and neuropsychological data were collected during the first visit, prior to any therapeutic prescription; this approach was chosen in order to bypass the possible effect of anti-dementia drugs (anticholinesterase drugs, memantine) on psychotic symptoms. Further inclusion criteria were the age range of 50–90 years, an MMSE score ≥10, and a Clinical Dementia Rating Scale score ≥1. Exclusion criteria were the absence of a caregiver informed enough to provide accurate clinical information and the assumption of antipsychotic medication in the month prior to our evaluation. This restriction was adopted in order to avoid an underdiagnosis of psychotic phenomena. Furthermore, subjects with symptoms compatible with an LBD diagnosis (evident Parkinsonism, significant fluctuations, vegetative dysfunction) were excluded. According to the criteria of the Local Ethics Committee and the Helsinki Declaration, all AD patients and/ or their proxy caregivers gave their informed consent to be examined and to undergo a full neuropsychological examination.
197
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Psychotic phenomena were coded either as dichotomous variables (present/absent) or according to the individual score (frequency × severity) of each symptom. Symptoms reported on the subscale delusions were grouped as follows: (a) PD, including delusions of theft, infidelity, abandonment, persecution, and (b) DM, including the belief that a family member is someone else, that one’s home is not one’s home, that images on TV or photographs are actually ongoing in the house, and the belief in the presence of ‘phantom boarders’. A total score was obtained for each category, summing up individual scores reported for each symptom (e.g. a subject obtaining a score of 2 on the delusion of theft, 3 on delusion of infidelity, and 1 on delusion of abandonment would obtain a total score of 6 on PD). It must be recognized that the definition of DM (particularly Capgras syndrome) adopted by classical psychopathology is usually more stringent than the one we used, as it requires a mismatch between what a person looks like and who the patient thinks he is, due to an associated persecutory ideation. That is, a person who at a first view could be identified as a family member is thought to be someone else, namely a specific persecutor. Nevertheless, current literature is quite homogeneous about the use of the more permissive definition of DM that we applied [7]. HALL were classified according to their sensory modality in ‘visual’, ‘auditory’, and ‘other modalities’. According to the hypothesis of possible different substrates for v-HALL and auditory HALL, total scores for the two conditions were considered separately. Neuropsychological Examination All participants underwent the MMSE and the following neuropsychological tests: Rey’s Auditory Verbal Learning Test (RAVLT) [35], Rey-Osterrieth’s Complex Figure (ROCF) – copy and recall [36], Raven’s Progressive Colored Matrices (RPM) [35], digit and visuospatial span [37], immediate visual memory [38], phonological verbal fluency [35], semantic verbal fluency [39], object naming [40], Stroop’s test (short version) [41], copy of object without and with landmarks [35], and the Multiple Features Targets Cancellation (MFTC), a test of visual search [42]. Statistical Analysis Most of the collected data displayed a nonnormal distribution as assessed by Shapiro-Wilk’s test. Thus, statistical analyses were conducted using nonparametric techniques. Frequency comparisons were conducted by means of the nonparametric χ2 test, with Yates’ continuity correction and Fisher’s exact test when required. Group comparisons for continuous variables were conducted by means of the Kruskal-Wallis or the Mann-Whitney U test as indicated. PD, DM, and HALL univariate linear regression analyses were conducted setting the total PD, DM, and HALL scores as dependent variables and single clinical and neuropsychological variables as possible predictors. Variables displaying a statistical significance in univariate models (p < 0.05) were entered into multiple-variable linear regression models assuming the total scores obtained in each symptom category as dependent variables, and controlling for age, educational level, illness duration, and MMSE score. Furthermore, since different types of psychotic symptoms are often present in the same subject, the regression model for each category was also controlled for the total scores obtained on the other categories (e.g. the regression model for PD was controlled for DM, and HALL total scores). The multiple-variable analyses were conducted using a backward stepwise technique.
The study sample included 108 AD patients (69 women; 64%), with a mean age of 73.3 years (SD: 7.62; range: 52–90) and a mean education of 8.0 years (SD: 4.59). The mean illness duration at observation was 42.6 months (SD: 23.3; range: 18–84) and the mean MMSE score was 17.2 (SD: 5.78; range: 10–24). In total, 48 (46%) subjects were affected by at least one psychotic symptom. Among the AD patients with psychotic symptoms, 42 (38.9%) displayed delusional symptoms, 16 (15.2% of the total sample) were affected by isolated PD, 11 (10.5%) by DM, and 15 (14.3%) by a combination of at least one paranoid and one misidentification symptom (DM+PD). Of the study sample, 23 (21.9%) patients displayed HALL. In most cases hallucinatory symptoms
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
Results
198
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Table 1. Individual frequencies of psychotic symptoms
Symptom
n (%)
PD Persecution Stealing Jealousy Abandonment DM ‘A family member is someone else’ ‘One’s home is not one’s home’ ‘Images on TV are ongoing in the house’ ‘Phantom boarders’ HALL v-HALL Isolated auditory Other modalitiesa
10 (9.5) 25 (23.8) 5 (4.8) 10 (9.5) 12 (11.4) 16 (15.2) 6 (5.7) 10 (9.5) 21 (20) 1 (0.9) 3 (2.9)
The total percentage is >100, as several subjects were affected by more than one symptom. a Other modalities include olfactory (n = 2) and tactile/haptic HALL (n = 1).
Group Comparisons among Patients Showing No Delusions or Some Form of Delusional Symptoms Table 2 shows the differences in the neuropsychological and behavioral profiles of the four groups of patients. The groups differed significantly as for illness duration and MMSE score as well as in several neuropsychological measures. In general, DM subjects showed lower performances than the other groups, whereas PD subjects had better performances than the other groups to the same extent as ND subjects. On post hoc comparisons, PD patients did not differ significantly from ND patients, with the exception of digit span forward, in which they performed better (Mann-Whitney U test; p = 0.005). PD subjects were significantly younger than DM patients (p = 0.043), whereas their illness duration was shorter (p = 0.019). On neuropsychological testing, PD subjects performed better than DM patients on MMSE (p = 0.001), RAVLT immediate recall (p = 0.001) and recognition accuracy (p = 0.029), digit span backward (p = 0.045), spatial span forward (p = 0.008), RPM (p = 0.016), phonological verbal fluency (p = 0.009), and semantic verbal fluency (p = 0.001). The cognitive performance of PD patients was also generally better than that of DM+PD patients; statistical significance was reached for RAVLT immediate recall (p = 0.037) and spatial span backward (p = 0.002).
Downloaded by: University of Illinois at Chicago 128.248.155.225 - 5/4/2015 12:24:53 AM
were associated with DM [5/11 (45%) patients] or DM+PD [9/15 (60%) patients]; more rarely, they were reported in association with PD [3/16 (19%) patients]. Only 6/66 (9%) subjects without delusional symptoms were affected by HALL (χ2 = 22.03; p < 0.001). There were no differences in the distribution of delusional symptoms with respect to gender (χ2 = 2.855; p = 0.827). Data on the prevalence of individual psychotic phenomena are presented in table 1. The vast majority of subjects with HALL (21/23, 91%) were affected by v-HALL. Vocalizations or verbal expression directed toward the subject of the HALL were reported in 10 subjects; nevertheless, in none of these cases the patients seemed to experience simultaneous auditory HALL. Since the subjects affected by auditory or other HALL constituted a very small group, further analyses were carried out only on v-HALL.
199
Dement Geriatr Cogn Disord 2015;39:194–206 DOI: 10.1159/000369161
© 2015 S. Karger AG, Basel www.karger.com/dem
Quaranta et al.: Cognitive and Behavioral Determinants of Psychotic Symptoms in Alzheimer’s Disease
Table 2. Cognitive and behavioral profile of subjects subdivided on the basis of delusional symptoms
Age, years Education, years Illness duration, months MMSE score Neuropsychological assessment RAVLT Immediate recall Delayed recall Recognition (accuracy) Recognition (false alarms) ROCF Copy Delayed recall RPM Immediate visual memory Digit span forward Digit span backward Spatial span forward Spatial span backward Phonological verbal fluency Semantic verbal fluency Object naming Copy of figures Copy of figures with landmarks MFCT Accuracy False alarms Time of execution Stroop’s test Interference errors Interference time NPI Agitation Depression Anxiety Euphoria/dysphoria Apathy Disinhibition Irritability Abnormal motor behavior Sleep disturbances Appetite/alimentary disturbances Total score
ND (n = 66)
PD (n = 16)
DM (n = 11)
DM+PD (n = 15)
p*
72.24 (7.21) 8.03 (4.73) 39.78 (17.95) 18.95 (5.33)
70.56 (9.46) 8.87 (4.33) 39.03 (15.49) 18.31 (4.53)
78.18 (6.37)†, ‡ 8.45 (4.76) 64.54 (30.14)†, ‡ 12.09 (3.27)†, ‡
76.00 (6.51) 6.73 (4.45) 36.67 (17.45)§ 14.47 (4.52)†
0.056 0.429 0.030
