You are here

Article Text

Article menu

Download PDFPDF

Neuropsychiatry
Research paper
Deconstructing psychosis and misperception symptoms in Parkinson’s disease
  1. Yoshiyuki Nishio1,
  2. Kayoko Yokoi1,2,
  3. Makoto Uchiyama1,3,
  4. Yasuyuki Mamiya1,
  5. Hiroyuki Watanabe1,
  6. Miyeong Gang1,
  7. Toru Baba1,
  8. Atsushi Takeda4,
  9. Kazumi Hirayama2,
  10. Etsuro Mori1
  1. 1 Department of Behavioral Neurology and Cognitive Neuroscience, Tohoku University School of Medicine, Sendai, Japan
  2. 2 Department of Occupational Therapy, Yamagata Prefectural University of Health Science, Yamagata, Japan
  3. 3 Department of Speech, Language and Hearing Sciences, Niigata University of Health and Welfare, Niigata, Japan
  4. 4 Department of Neurology, Sendai Nishitaga National Hospital, Sendai, Japan
  1. Correspondence to Dr Yoshiyuki Nishio, Department of Behavioral Neurology and Cognitive Neuroscience, Tohoku University Graduate School of Medicine, 2-1, Seiryo-machi, Aoba-ku, Sendai 980-8575, Japan; nishiou{at}med.tohoku.ac.jp

Abstract

Objective Patients with Lewy body disease develop a variety of psychotic and misperception symptoms, including visual hallucinations and delusions, as well as ‘minor hallucinations’, that is, a sense of presence, passage hallucinations and visual illusions. Although these symptoms have been suggested to have common underlying mechanisms, the commonalities and differences among them have not been systematically investigated at the neural level.

Methods Sixty-seven patients with Parkinson’s disease underwent neuropsychological and behavioural assessments, volumetric MRI and 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET). A factor analysis was performed to discover correlations among psychotic and misperception symptoms, other behavioural symptoms and neuropsychological performances. Partial least-squares correlation analysis was used to investigate the relationship between these symptoms and the joint features of MRI and FDG-PET.

Results A sense of presence, passage hallucinations and visual illusions constituted a single behavioural factor (minor hallucinations/illusions). Visual hallucinations formed another behavioural factor along with delusions, depression and fluctuating cognition (psychosis/dysphoria). Three distinct brain–behaviour correlation patterns were identified: (1) posterior cortical atrophy/hypometabolism associated with minor hallucinations/illusions and visuospatial impairment; (2) upper brainstem and thalamic atrophy/hypometabolism associated with psychosis/dysphoria and (3) frontal cortical atrophy/hypometabolism associated with non-visual cognition. No significant differences in neuroimaging findings were identified between patients who had minor hallucinations/illusions alone and patients who also had visual hallucinations.

Conclusions Our findings suggest that combined damage to the upper brainstem/thalamus and the posterior neocortex underlies both minor hallucinations/illusions and visual hallucinations and that the former pathology is more associated with visual hallucinations/frank psychosis and the latter is more associated with minor hallucinations/illusions.

  • delusions
  • minor hallucinations
  • sense of presence
  • visual hallucinations
  • visual illusions
  • Parkinson’s disease

https://doi.org/10.1136/jnnp-2017-315741

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Phenomenologically, visual hallucinations are abnormalities in vision. A variety of vision disorders, ranging from eye diseases to damage to the central visual pathways, are associated with visual hallucinations, suggesting a role for visual impairments in the mechanisms of this symptom.1 Consistent with this view, past research has demonstrated that visuoperceptual impairment is prevalent and associated with the development of visual hallucinations in Lewy body disease (LBD).2 3 From another perspective, visual hallucinations are abnormalities in belief or awareness.4 Most patients with LBD suffering from recurrent and persistent visual hallucinations lack or have diminished insights into their hallucinatory experiences and believe the veridicality of their false perception.5 In addition, non-visual hallucinatory phenomena, such as a sense of presence (a sensation that someone is nearby) and passage hallucinations (a sensation that something is passing sideways), have been documented in association with visual hallucinations in LBD.6 7 Deficient spatial awareness or abnormal central vestibular processing has been implicated in the mechanisms of these types of hallucinations.8 This suggests that similar non-visual abnormalities may underlie visual hallucinations in LBD. In the present study, we attempted to disentangle the complex relationship among a variety of visual and non-visual types of psychotic and misperception symptoms in LBD. To do so, we conducted a detailed phenomenological assessment and multivariate neuroimaging analyses for 67 patients with Parkinson’s disease (PD).

    To investigate the brain–behaviour relationship for psychotic and misperception symptoms in PD, we used a partial least-squares (PLS) correlation analysis with MRI morphometry and 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET). PLS is a multivariate method that enables us to analyse multiple behavioural measures and multimodal neuroimaging data in a single statistical model.9 10 The relationships between behavioural symptoms and lesion locations/dysfunctional regions are complex and do not exhibit a one-to-one correspondence. The rarity of psychotic symptoms in patients with focal brain lesions suggests that some combination and balance of lesions in multiple brain regions may be associated with the emergence of psychosis in patients with neurological disorders.11 In these situations, multivariate approaches for brain–behaviour analyses are advantageous compared with conventional mass-univariate correlational approaches. In addition, a joint analysis of the neuroimaging data from multiple modalities is anticipated to provide information that a single modality analysis does not reveal because each modality provides distinct information about the brain.12

    Materials and methods

    Participants

    We recruited 67 patients with PD and their caregivers. The demographics and clinical profiles of the participants are shown in table 1. PD was diagnosed according to the UK PD Society Brain Bank criteria.13 The inclusion criteria for the patients were as follows: (1) age between 50 and 79 years; (2) age at onset >40 years; (3) best-corrected visual acuity ≥0.4; (4) a score ≥24 on the Mini-Mental State Examination (MMSE) and (5) the Clinical Dementia Rating≥1. We excluded patients with a history of other neurological, psychiatric or severe ocular diseases and patients with MRI evidence of focal brain lesions. In total, 61 patients were taking levodopa and/or dopamine agonists; 5 were taking an anticholinergic drug (trihexyphenidyl) and 16 were taking benzodiazepines. No patient was taking antidepressants or antipsychotics.

    View this table:
    Table 1

    Demographic and clinical profiles of the patients (n=67)

    Motor and cognitive profiles

    Motor symptoms were assessed in the ‘ON’ state using the Unified Parkinson’s Disease Rating Scale (UPDRS) part III. The MMSE was used to assess general cognitive function. Long-term memory and working memory were evaluated using the word list recall task from the Alzheimer’s Disease Assessment Scale14 and the digit span and the spatial span subtests from the Wechsler Memory Scale-Revised, respectively. We used two measures for visuoperceptual/visuospatial function: (1) the sum of the scores on the Shape Detection Screening and Position Discrimination subtests of the Visual Object and Space Perception battery,15 the score on the Object Decision subtest of the Birmingham Object Recognition Battery16 and the scores on the Face Recognition subtests of the Visual Perception Test for Agnosia17; and (2) the correct response score on the overlapping figure identification test.18 The patients’ scores and performances on these measures are summarised in table 1. In the following statistical analyses of behavioural data, we used raw neuropsychological scores instead of age-adjusted scores. The effect of age was considered in the PLS brain–behaviour correlation analysis because age influences both behaviours and neural structures/functions.

    Assessment of behavioural and misperception symptoms

    The Neuropsychiatric Inventory was administered to the patients’ caregivers.19 In this study, we focused on the delusion, hallucination and dysphoria/depression domains. Cognitive fluctuation was assessed using the Cognitive Fluctuation Inventory.20 Visual illusions, a sense of presence and passage hallucinations were assessed using the original questionnaire (see online supplementary material). We classified visual illusions into two subtypes: simple visual illusions, which are visual illusions associated with motion, shape, colours, size and/or distance; and complex visual illusions in which objects are misperceived as different objects.

    Supplementary Material

    Supplementary data

    The scene pareidolia test

    The scene pareidolia test is a test that evokes and measures visual hallucination-like illusions. The administration and scoring methods have been described elsewhere.21 Briefly, the subjects were instructed to point to and describe the objects shown in each of the 25 coloured pictures of visual scenes in as much detail as possible. The subjects’ responses were classified as pareidolic illusions when they identified objects that were not in the pictures. The total number of pareidolic responses was used as an outcome measure.

    Factor analysis for behavioural variables

    In general, neuropsychological tests and behavioural questionnaires are designed according to the symptom domains, which are defined by convention or by a priori hypotheses. However, the actual symptoms may not conform to symptomatic classifications and there may be hidden correlation patterns across the symptom domains. We used a factor analysis to identify these correlations. In this analysis, we considered dopaminergic, cholinergic and GABAergic drugs because of their well-known effects on cognition and behaviour. A principal component analysis with varimax rotation was performed to assess neuropsychological test performance, behavioural symptoms (delusions, hallucinations, depression, fluctuating cognition, simple and complex visual illusions, sense of presence and passage hallucinations) and medication (levodopa equivalent dose, anticholinergic drugs and benzodiazepines). The analysis was performed using the ‘psych’ package for R.22 We determined the number of factors extracted using a scree plot. Variables with absolute factor loading values >0.6 were considered significant.

    Neuroimaging data acquisition and preprocessing

    Both three-dimensional MRI (spatial resolution, 0.98×0.98×1.5 mm3) and FDG-PET (spatial resolution, 3.38×3.38×3.38 mm3 and 1.3×1.3×1.3 mm3 for original and reconstructed images, respectively) were obtained from all participants. Neuroimaging data were preprocessed with SPM12 (http://www.fil.ion.ucl.ac.uk/spm/software/spm12/). Spatial normalisation and tissue segmentation of the MR images were performed using the unified segmentation algorithm. The FDG-PET images were spatially normalised with the parameters obtained from the unified segmentation of the MR images. The obtained grey matter and FDG-PET images were resampled into a voxel size of 2×2×2 mm3. We employed differential smoothing in which the grey matter and FDG-PET images were smoothed with 8 and 9 mm full-width at half-maximum, respectively.23 Binary inclusive brain masks were created using the SPM masking toolbox.24 Details of the neuroimaging methods are described in online supplementary material.

    PLS correlation analysis

    Among the several variants of PLS, we used a PLS correlation analysis to discover the relationship between joint spatial patterns of the grey matter and FDG-PET images and the clinical variables. We used the regular behaviour PLS routine of the PLSGUI software (https://www.rotman-baycrest.on.ca/pls). The clinical variables included age, disease duration, UPDRS part III and factor scores for the top four factors extracted from the factor analysis (ie, psychosis/dysphoria, minor hallucinations/illusions, cognition and visuospatial function; see section ‘Factor analysis of the neuropsychological and behavioural variables’). We did not include the medication-related factors (factors 5 and 6) in the analysis because of their relative small eigenvalues and because their impacts on neuroimaging were expected to be global not regional.

    The mathematical details of the PLS correlation analysis are described elsewhere.9 10 Briefly, the PLS correlation analysis seeks a discrete number of latent variables that effectively describe the brain–behaviour relationship by maximising the covariance between the neuroimaging data and behavioural variables (see online supplementary material). The brain score, which is conceptually analogous to factor score in factor analysis, indicates the degree to which the spatial pattern profile is expressed by an individual subject. The behaviour score indicates the degree to which an individual subject expresses the behaviour weight profile. The statistical significance of each latent variable was determined using a permutation test with 500 permutations and a statistical threshold of p<0.05. The significance of the contribution of each voxel to each latent variable and that of the correlation between the brain score and each behavioural variable was determined using bootstrap resampling with 200 replicates. A given brain voxel significantly contributed to each latent variable when its bootstrap ratio was >2.0. A given behavioural measure was considered to significantly correlate with each brain score if the 95% CI for its correlation coefficient did not cross zero.

    Univariate neuroimaging analysis

    We performed a group-wise comparison using one-way analysis of variance (ANOVA) to supplement the PLS analysis. No nuisance variables were included in the model. For the FDG-PET analysis, proportional scaling was used to control for individual variations in global tracer uptake. The threshold for statistical significance was set to a family-wise error rate <0.05.

    Results

    Frequency of psychotic and misperception symptoms

    The results are summarised in table 1. Fifteen patients had positive scores on the hallucination domain of the Neuropsychiatric Inventory, of which 10 patients exhibited visual hallucinations without auditory hallucinations and 5 exhibited both visual and auditory hallucinations. There were no patients who had auditory hallucinations alone. Visual hallucinations were fully formed (persons and/or animals) in 13 patients and elementary (moving dots or ‘small insect-like’ things) in 2 patients. There was no patient who had both formed and elementary visual hallucinations. Thirty patients exhibited one or more minor hallucinations/illusions (a sense of presence, passage hallucinations and simple and complex visual illusions), of which 11 patients exhibited both visual hallucinations and minor hallucinations/illusions. Six patients had delusions, of which five exhibited both visual hallucinations and minor hallucinations/illusions, and one only exhibited minor hallucinations/illusions. There were no patients who had delusions without hallucinations or illusions.

    Factor analysis of the neuropsychological and behavioural variables

    The factor analysis yielded six factors that accounted for 72% of the total variance in the data (table 2). The first four factors were symptom-related and the other two were associated with medications.

    View this table:
    Table 2

    Results of the factor analysis

    • Factor 1 (psychosis/dysphoria): Hallucinations, delusions, dysphoria/depression and fluctuating cognition constituted this factor.

    • Factor 2 (minor hallucinations/illusions): A sense of presence, passage hallucinations, simple and complex visual illusions and illusory responses on the scene pareidolia test were positively loaded on this factor.

    • Factor 3 (cognition): Scores on the MMSE and digit and tapping spans were positively loaded on this factor.

    • Factor 4 (visuospatial function): The total score of the visuospatial tests and better performance on the overlapping figure test constitute this factor.

    • Factors 5 and 6 were medication-related factors. The use of trihexyphenidyl and benzodiazepines constituted a single factor. The levodopa equivalent dose formed another single factor.

    PLS correlation analysis

    The PLS correlation analysis identified three statistically significant latent variables. The first latent variable accounted for 29.1% of the total covariance (permutation p<0.001), the second accounted for 17.4% (permutation p<0.001) and the third accounted for 12.1% (permutation p=0.014). The spatial maps associated with these latent variables and the correlations between the brain scores and behavioural variables are summarised in figure 1. The first latent variable was associated with a relative preservation of grey matter volume/glucose metabolism in the posterior neocortices, including the bilateral superior temporal/posterior insular, the ventral temporo-occipital, the inferior parietal, the occipito-parietal and the primary visual cortices, as well as the caudate head. Brain scores for both grey matter and FDG-PET images that were associated with this spatial pattern were correlated with a younger age, lower UPDRS motor score, milder minor hallucinations/illusions and better visuospatial function. The spatial map associated with the second latent variable was characterised by a relative preservation of grey matter volume/glucose metabolism in the upper brainstem, cerebellum, thalamus and sensorimotor cortices and by a relative decrease in grey matter volume/glucose metabolism in the posterior parietal cortex. Brain scores for the grey matter and FDG-PET images that were associated with this spatial pattern correlated with a lower UPDRS motor score. The grey matter brain score, but not the FDG-PET brain score, marginally correlated with a lower ‘psychosis/dysphoria’ factor score. The third latent variable was associated with a relative decrease in grey matter volume/glucose metabolism in the medial and lateral frontal cortices and a relative preservation of grey matter volume/glucose metabolism in the cerebellum. Both the grey matter and FDG-PET brain scores that were associated with this spatial pattern correlated with an older age and worse cognitive function.

    Figure 1
    Figure 1

    Results of the partial least-square correlation analysis. The left column shows the spatial patterns associated with the three statistically significant latent variables (LV1–LV3). The threshold for statistical significance was set at a bootstrap ratio greater than or less than 2.0. The right column shows bar graphs of the correlation between the brain scores and behavioural variables. The vertical axis indicates the correlation coefficients. Significant differences are indicated in colours (blue for grey matter images and red for 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET) images). Duration, disease duration; LV, latent variable; UPDRS, Unified Parkinson’s Disease Rating Scale.

    As shown in the behavioural analysis, a majority of patients with visual hallucinations also exhibited minor hallucinations/illusions (table 1), suggesting that visual hallucinations and minor hallucinations/illusions may be associated with similar spatial patterns of pathologies but differ in the severity of neural damage. To examine this hypothesis, we performed group-wise comparisons of the brain scores for 33 patients with no hallucinations, 19 with minor hallucinations/illusions only, 11 with both minor hallucinations/illusions and visual hallucinations and 4 with visual hallucinations only (figure 2). A one-way ANOVA (the four patients with visual hallucinations only were excluded) demonstrated a significant quadratic trend for the grey matter brain scores associated with latent variable 1 (no hallucinations > minor hallucinations/illusions = minor and visual hallucinations; t=2.455; p=0.017) and for the FDG-PET brain scores associated with latent variables 1 (no hallucinations > minor hallucinations/illusions = minor and visual hallucinations; t=−2.681; p=0.0095) and 2 (no hallucinations < minor hallucinations/illusions = minor and visual hallucinations; t=2.242; p=0.029). In summary, we did not observe any quantitative differences in the brain scores between patients with minor hallucinations/illusions only and patients with both minor hallucinations/illusions and visual hallucinations.

    Figure 2
    Figure 2

    Group comparisons of the brain scores among patients without hallucinations, patients with minor hallucinations/illusions only and patients with both minor and visual hallucinations. Statistically significant differences are highlighted with thick lines. Although patients with visual hallucinations only were not included in the statistical analysis, their brain scores are shown as a reference. Brain scores for the grey matter and 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET) images that are associated with the first latent variable (LV1) are larger in patients without hallucinations than in the other groups but are equivalent in patients with minor hallucinations/illusions only and in patients with both minor and visual hallucinations (one-way analysis of variance, t=2.455; p=0.017 for the grey matter images; t=−2.681; p=0.0095 for the FDG-PET images). Brain scores for the FDG-PET images that are associated with the second LV2 are smaller in patients without hallucinations than in the other groups but are equivalent in patients with minor hallucinations/illusions only and in patients with both minor and visual hallucinations (t=2.242; p=0.029).

    Univariate neuroimaging analysis

    Group comparisons with one-way ANOVA were performed among 33 patients with no hallucinations, 19 with minor hallucinations/illusions only and 11 with both minor hallucinations/illusions and visual hallucinations. We examined the linear and two quadratic trends among the groups (no hallucinations > minor hallucinations/illusions > minor hallucinations/illusions and visual hallucinations; no hallucinations > minor hallucinations/illusions = minor and visual hallucinations; no hallucinations = minor hallucinations/illusions > minor and visual hallucinations). Significant group differences in the regional grey matter volume were not observed. A significant quadratic trend of the group differences in the FDG-PET images was observed. Regional glucose metabolism was decreased in the medial occipito-parietal cortex in the patients with minor hallucinations/illusions only and in patients with both minor hallucinations/illusions and visual hallucinations compared with patients without hallucinations (SPM t-contrast of (no hallucinations, minor hallucinations/illusions only, minor and visual hallucinations) = (1,–0.5, −0.5)) (figure 3). The other comparisons ((1, 0, –1) and (0.5, 0.5,–1)) of the FDG-PET images were not significant.

    Figure 3
    Figure 3

    Results of the univariate analysis of the 18F-fluorodeoxyglucose-positron emission tomography images. The regional glucose metabolic rate in the posterior cingulate and the medial occipito-parietal cortices was lower in patients with minor hallucinations/illusions only and patients with both minor and visual hallucinations than in patients without hallucinations. The threshold for statistical significance was set at a family-wise error rate of p<0.05.

    Discussion

    A substantial proportion of patients with PD experience a sense of presence, passage hallucinations and/or visual illusions.7 25 Despite the phenomenological heterogeneity, these symptoms have been grouped under the umbrella concept of ‘minor hallucinations’, which has been implicated as a prodromal state for frank visual hallucinations and delusions. However, the integrity of minor hallucinatory phenomena and their relation to frank psychosis have not been systematically investigated. In the present study, a sense of presence, passage hallucinations and the two subtypes of visual illusions (simple and complex) constituted a single behavioural factor (the ‘minor hallucinations/illusions’ factor), suggesting the validity of the concept of minor hallucinations. Visual hallucinations formed another behavioural factor, along with delusions, dysphoria/depression and fluctuating cognition (the ‘psychosis/dysphoria’ factor). Although minor hallucinations/illusions and visual hallucinations were classified into distinct behavioural factors, most patients developed visual hallucinations in association with minor hallucinations/illusions, and isolated visual hallucinations were relatively rare in our sample. This finding suggests that minor hallucinations/illusions and visual hallucinations are closely related phenomena. At the neural level, minor hallucinations/illusions were significantly correlated with grey matter loss/hypometabolism in the posterior neocortex, whereas visual hallucinations and delusions weakly correlated with reduced grey matter volumes in the upper brainstem, thalamus and motor cortices and a relative preservation of the parietal grey matter volume (figure 1). The group-wise comparisons did not reveal significant differences in the regional grey matter volume/glucose metabolism between patients with minor hallucinations/illusions only and patients with both minor hallucinations/illusions and visual hallucinations (figures 2 and 3). Based on these neuroimaging findings, we suggest that combined damage to the upper brainstem/thalamus and the posterior neocortex is a common underlying mechanism for both minor hallucinations/illusions and visual hallucinations and that the former pathology is more associated with visual hallucinations/frank psychotic symptoms and the latter is more associated with minor hallucinations/illusions.

    In the present study, fluctuating cognition, visual hallucinations and delusions were classified into a single behavioural factor (the ‘psychosis/dysphoria’ factor). Cholinesterase inhibitors concurrently ameliorate all of these symptoms,26 suggesting that subcortical pathology associated with neuromodulator abnormalities underlies both fluctuating cognition and psychosis in LBD. This view is supported by several neuropathological studies that demonstrate the relationship between fluctuating cognition/consciousness and cholinergic and dopaminergic abnormalities in the thalamus in LBD27 28 and by our neuroimaging finding that fluctuating cognition, visual hallucinations and delusions correlated with grey matter reduction in the upper brain stem and thalamus (LV2 in the figure 1). In addition to fluctuating cognition, dysphoria/depression was associated with visual hallucinations and delusions, consistent with previous studies of LBD.29 30 The relationship between psychosis and depression has long been recognised in psychiatry.31 Although depression is not a pivotal cause of psychosis, comorbid depression is associated with a greater severity of hallucinations and delusions in schizophrenia. There are two possible explanations for this relationship. First, non-neural or psychological factors may be involved in the development of hallucinations. This is consistent with our finding that patients with both minor hallucinations/illusions and visual hallucinations exhibited no clear neuroimaging features that were distinct from the features of patients with minor hallucinations/illusions alone. Depression and other types of negative moods may play an important role in the process by which minor hallucinations/illusions evolve into frank visual hallucinations. An alternative explanation is that depression is a psychological reaction to, or consequence from, visual hallucinations and delusions. Long-term observational and/or interventional studies would be required to elucidate a causal relationship between psychosis and depression.

    In this study, we identified two distinctive cognition-related factors, one of which (the ‘cognition’ factor) was associated with frontal lobe grey matter volume/glucose metabolism, and the other (the ‘visuospatial function’ factor) was associated with posterior neocortical grey matter volume/glucose metabolism. In the neuroimaging–behaviour correlation analysis, the ‘psychosis/dysphoria’ factor and ‘minor hallucinations/illusions’ factor were independent of frontal cortex-mediated cognitive function, but the ‘minor hallucinations/illusions’ factor shared neuroimaging correlates with the ‘visuospatial function’ factor. Consistent with our finding, a recent MRI morphometry study demonstrated that grey matter volume in the medial parieto-occipital cortex was decreased in patients with PD who had minor hallucinations compared with those who did not.32 Although the relationship between visual hallucinations and visuospatial impairment has been established in LBD,2 3 the link between visuospatial function and minor hallucinations/illusions has been poorly understood. A sense of presence and related phenomena was observed in patients with epileptic seizures who have a temporo-parieto-occipital junction lesion and when electrical stimulation was directly delivered to the temporo-parietal junction through intracranial electrodes.33 34 Similarly, visual illusions, which are also termed metamorphopsia in the neurology literature, have been observed in patients with epilepsy who have a lesion in the parieto-occipital junction.35 36 Furthermore, the epileptic seizures arising from these cortical regions occasionally induce frank visual hallucinations.37 Based on these findings, spatial perception abnormalities that arise from dysfunction of the posterior association neocortex, such as the temporo-parietal junction and medial parieto-occipital cortex, are involved in the mechanism underlying visual illusions and various types of hallucinatory phenomena. However, a simple correlation between brain hypofunction and cognitive impairment does not provide a satisfactory explanation of the neural mechanisms underlying these symptoms. In fact, hallucinations and other misperception symptoms are rarely observed in patients with cerebrovascular diseases that affect the posterior association neocortex or in other neurodegenerative diseases, such as Alzheimer’s disease, in which neuronal loss preferentially occurs in this cortical region.38 Several lines of evidence suggest that abnormalities in the cholinergic system and/or other neuromodulators are additional key mechanisms. Focal damage to the upper brainstem tegmentum, in which many cell groups containing neuromodulators such as acetylcholine, serotonin and dopamine are densely co-localised, causes visual and other types of hallucinations (peduncular hallucinosis).39 A recent study demonstrated that consciousness disturbance in focal epilepsy is associated with the inhibition of cholinergic neurons in the upper brainstem and basal forebrain.40 Therefore, combined damage to the posterior association neocortex and the cholinergic system may lead to minor hallucinations/illusions and precipitate frank visual hallucinations.

    There are a number of limitations to this study. The first limitation is its cross-sectional design. We must determine the chronological order in which the individual symptoms develop and disappear to fully understand the relationships between minor hallucinations/illusions, frank psychosis and cognitive dysfunction in LBD. Second, the sample size of the study is small, and relatively few patients had frank visual hallucinations and delusions. Further investigations with a larger sample size that include patients with a wider range of severity of the symptoms are necessary to generalise our discussion on the mechanism of hallucinations and illusions in DLB. Third, we may have underestimat the impact of anticholinergic drugs and benzodiazepines because this study included a small number of patients who took these types of drugs. Finally, the neuroimaging modalities used in this study, that is, MRI morphometry and FDG-PET, are insensitive to structural/functional changes in small subcortical structures and unable to capture alterations in dopamine and/or acetylcholine neurotransmission. This may be related to our finding that the brain–behaviour correlation between the ‘psychosis/dysphoria factor’ and subcortical structures (LV2 in figure 1) was not as strong as the correlation between the ‘minor hallucination/illusions’ factor and neocortical structures (LV1 in figure 1).

    References

    1. ↵
      2. Collerton D ,
      3. Perry E ,
      4. McKeith I
      . Why people see things that are not there: a novel Perception and Attention Deficit model for recurrent complex visual hallucinations. Behav Brain Sci 2005;28:737–57. discussion 57-94.doi:10.1017/S0140525X05000130
      OpenUrlPubMedWeb of Science
    2. ↵
      2. Mori E ,
      3. Shimomura T ,
      4. Fujimori M , et al
      . Visuoperceptual impairment in dementia with lewy bodies. Arch Neurol 2000;57:489–93.doi:10.1001/archneur.57.4.489
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      2. Mosimann UP ,
      3. Mather G ,
      4. Wesnes KA , et al
      . Visual perception in Parkinson disease dementia and dementia with Lewy bodies. Neurology 2004;63:2091–6. 63/11/2091 [pii].doi:10.1212/01.WNL.0000145764.70698.4E
      OpenUrlCrossRefPubMed
    4. ↵
      2. Fletcher PC ,
      3. Frith CD
      . Perceiving is believing: a bayesian approach to explaining the positive symptoms of schizophrenia. Nat Rev Neurosci 2009;10:48–58.doi:10.1038/nrn2536
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      2. Ballard C ,
      3. McKeith I ,
      4. Harrison R , et al
      . A detailed phenomenological comparison of complex visual hallucinations in dementia with lewy bodies and Alzheimer’s disease. Int Psychogeriatr 1997;9:381–8.doi:10.1017/S1041610297004523
      OpenUrlCrossRefPubMed
    6. ↵
      2. Nagahama Y ,
      3. Okina T ,
      4. Suzuki N , et al
      . Classification of psychotic symptoms in dementia with Lewy bodies. Am J Geriatr Psychiatry 2007;15:961–7.doi:10.1097/JGP.0b013e3180cc1fdf
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      2. Fénelon G ,
      3. Mahieux F ,
      4. Huon R , et al
      . Hallucinations in Parkinson’s disease: prevalence, phenomenology and risk factors. Brain 2000;123(Pt 4):733–45.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      2. Lopez C ,
      3. Halje P ,
      4. Blanke O
      . Body ownership and embodiment: vestibular and multisensory mechanisms. Neurophysiol Clin 2008;38:149–61.doi:10.1016/j.neucli.2007.12.006
      OpenUrlCrossRefPubMed
    9. ↵
      2. Krishnan A ,
      3. Williams LJ ,
      4. McIntosh AR , et al
      . Partial Least squares (PLS) methods for neuroimaging: a tutorial and review. Neuroimage 2011;56:455–75.doi:10.1016/j.neuroimage.2010.07.034
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      2. McIntosh AR ,
      3. Bookstein FL ,
      4. Haxby JV , et al
      . Spatial pattern analysis of functional brain images using partial least squares. Neuroimage 1996;3(3 Pt 1):143–57.doi:10.1006/nimg.1996.0016
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      2. Cummings JL ,
      3. Mega MS
      . Psychosis, delusions and schizphirenia. Neuropsychiatry and behavioral neuroscience. New York: Oxford University Press, 2003:172–86.
    12. ↵
      2. González-Redondo R ,
      3. García-García D ,
      4. Clavero P , et al
      . Grey matter hypometabolism and atrophy in Parkinson’s disease with cognitive impairment: a two-step process. Brain 2014;137(Pt 8):2356–67.doi:10.1093/brain/awu159
      OpenUrlCrossRefPubMed
    13. ↵
      2. Gibb WR ,
      3. Lees AJ
      . The relevance of the Lewy body to the pathogenesis of Idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 1988;51:745–52.doi:10.1136/jnnp.51.6.745
      OpenUrlAbstract/FREE Full Text
    14. ↵
      2. Mohs RC ,
      3. Rosen WG ,
      4. Davis KL
      . The Alzheimer’s disease assessment scale: an instrument for assessing treatment efficacy. Psychopharmacol Bull 1983;19:448–50.
      OpenUrlPubMedWeb of Science
    15. ↵
      2. Warrington EK ,
      3. James M
      . Visual object and space perception battery. Bury St Edmunds UK: Thames Valley Test Company, 1991.
    16. ↵
      2. Riddoch M ,
      3. Humphreys G
      . Birmingham Object recognition battery. Hove UK: Lawrence Erlbaum, 1992.
    17. ↵
      Dysfunction JSfHB. Visual perception test for Agnosia. Tokyo: Shinko Igaku Shuppan, 1997.
    18. ↵
      2. Ishioka T ,
      3. Hirayama K ,
      4. Hosokai Y , et al
      . Illusory misidentifications and cortical hypometabolism in Parkinson’s disease. Mov Disord 2011;26:837–43.doi:10.1002/mds.23576
      OpenUrlCrossRefPubMed
    19. ↵
      2. Cummings JL ,
      3. Mega M ,
      4. Gray K , et al
      . The Neuropsychiatric Inventory: comprehensive assessment of psychopathology in dementia. Neurology 1994;44:2308–14.doi:10.1212/WNL.44.12.2308
      OpenUrlCrossRefPubMed
    20. ↵
      2. Hashimoto M ,
      3. Manabe Y ,
      4. Mori E , et al
      . Content validity and inter-rater reliability of the cognitive fluctuation inventory. Brain Nerve 2014;66:175–83.
      OpenUrl
    21. ↵
      2. Uchiyama M ,
      3. Nishio Y ,
      4. Yokoi K , et al
      . Pareidolias: complex visual illusions in dementia with lewy bodies. Brain 2012;135(Pt 8):2458–69.doi:10.1093/brain/aws126
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      2. Revelle W
      .The ‘psych’ package for R.(version 1.2.12) http://personality-project.org/r/psych/.2015.
    23. ↵
      2. Villain N ,
      3. Fouquet M ,
      4. Baron JC , et al
      . Sequential relationships between grey matter and white matter atrophy and brain metabolic abnormalities in early Alzheimer’s disease. Brain 2010;133:3301–14.doi:10.1093/brain/awq203
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      2. Ridgway GR ,
      3. Omar R ,
      4. Ourselin S , et al
      . Issues with threshold masking in voxel-based morphometry of atrophied brains. Neuroimage 2009;44:99–111.doi:10.1016/j.neuroimage.2008.08.045
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      2. Pagonabarraga J ,
      3. Martinez-Horta S ,
      4. Fernández de Bobadilla R , et al
      . Minor hallucinations occur in drug-naive Parkinson’s disease patients, even from the premotor phase. Mov Disord 2016;31:45–52.doi:10.1002/mds.26432
      OpenUrlCrossRefPubMed
    26. ↵
      2. Mori E ,
      3. Ikeda M ,
      4. Kosaka K , et al
      . Donepezil for dementia with lewy bodies: a randomized, placebo-controlled trial. Ann Neurol 2012;72:41–52.doi:10.1002/ana.23557
      OpenUrlCrossRefPubMed
    27. ↵
      2. Piggott MA ,
      3. Ballard CG ,
      4. Dickinson HO , et al
      . Thalamic D2 receptors in dementia with lewy bodies, Parkinson’s disease, and Parkinson’s disease dementia. Int J Neuropsychopharmacol 2007;10:231–44.doi:10.1017/S146114570600647X
      OpenUrlCrossRefPubMed
    28. ↵
      2. Pimlott SL ,
      3. Piggott M ,
      4. Ballard C , et al
      . Thalamic nicotinic receptors implicated in disturbed consciousness in dementia with lewy bodies. Neurobiol Dis 2006;21:50–6.doi:10.1016/j.nbd.2005.06.008
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      2. Fénelon G ,
      3. Alves G
      . Epidemiology of psychosis in Parkinson’s disease. J Neurol Sci 2010;289:12–17.doi:10.1016/j.jns.2009.08.014
      OpenUrlCrossRefPubMed
    30. ↵
      2. Kazui H ,
      3. Yoshiyama K ,
      4. Kanemoto H , et al
      . Differences of behavioral and psychological symptoms of Dementia in Disease Severity in four Major dementias. PLoS One 2016;11:e0161092.doi:10.1371/journal.pone.0161092
    31. ↵
      2. Hartley S ,
      3. Barrowclough C ,
      4. Haddock G
      . Anxiety and depression in psychosis: a systematic review of associations with positive psychotic symptoms. Acta Psychiatr Scand 2013;128:327–46.doi:10.1111/acps.12080
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      2. Pagonabarraga J ,
      3. Soriano-Mas C ,
      4. Llebaria G , et al
      . Neural correlates of minor hallucinations in non-demented patients with Parkinson’s disease. Parkinsonism Relat Disord 2014;20:290–6.doi:10.1016/j.parkreldis.2013.11.017
      OpenUrlCrossRefPubMed
    33. ↵
      2. Arzy S ,
      3. Seeck M ,
      4. Ortigue S , et al
      . Induction of an illusory shadow person. Nature 2006;443:287.doi:10.1038/443287a
    34. ↵
      2. Zijlmans M ,
      3. van Eijsden P ,
      4. Ferrier CH , et al
      . Illusory shadow person causing paradoxical gaze deviations during temporal lobe seizures. J Neurol Neurosurg Psychiatry 2009;80:686–8.doi:10.1136/jnnp.2008.154310
      OpenUrlAbstract/FREE Full Text
    35. ↵
      2. Laff R ,
      3. Mesad S ,
      4. Devinsky O
      . Epileptic kinetopsia: ictal illusory motion perception. Neurology 2003;61:1262–4.doi:10.1212/01.WNL.0000090464.03392.9B
      OpenUrlCrossRefPubMed
    36. ↵
      2. Mailo J ,
      3. Tang-Wai R
      . Insight into the precuneus: a novel seizure semiology in a child with epilepsy arising from the right posterior precuneus. Epileptic Disord 2015;17:321–7.doi:10.1684/epd.2015.0759
      OpenUrl
    37. ↵
      2. Salanova V ,
      3. Andermann F ,
      4. Rasmussen T , et al
      . Parietal lobe epilepsy. clinical manifestations and outcome in 82 patients treated surgically between 1929 and 1988. Brain 1995;118(Pt 3):607–27.doi:10.1093/brain/118.3.607
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      2. Cummings JL
      . Toward a molecular neuropsychiatry of neurodegenerative diseases. Ann Neurol 2003;54:147–54.doi:10.1002/ana.10616
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      2. Nishio Y ,
      3. Ishii K ,
      4. Kazui H , et al
      . Frontal-lobe syndrome and psychosis after damage to the brainstem dopaminergic nuclei. J Neurol Sci 2007;260():271–4.doi:10.1016/j.jns.2007.04.021
      OpenUrlCrossRefPubMed
    40. ↵
      2. Motelow JE ,
      3. Li W ,
      4. Zhan Q , et al
      . Decreased subcortical cholinergic arousal in focal seizures. Neuron 2015;85:561–72.doi:10.1016/j.neuron.2014.12.058
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors Conception of research project: YN, KH and EM. Organisation of research project: YN, TB, AT, KH and EM. Execution of research project: YN, KY, MU, YM, TB, AT and KH. Design of analysis: YN. Execution of analysis: YN, HW and MG. Writing of the first draft of the manuscript: YN. Review and critique of the manuscript: EM.

    • Funding This work was supported by a Grant-in-Aid for Scientific Research (KAKENHI) (B) (24390278 to EM) and a Grant-in-Aid for Scientific Research for Young Scientists (KAKANHI) (90451591 to YN).

    • Competing interests None declared.

    • Ethics approval The Tohoku University Hospital ethics committee.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement A limited additional set of data for research purposes is available upon request.

    • Correction notice This paper has been amended since it was published Online First. Owing to a scripting error, some of the publisher names in the references were replaced with ‘BMJ Publishing Group’. This only affected the full text version, not the PDF. We have since corrected these errors and the correct publishers have been inserted into the references.