Comprehensive Evaluation of Psychotic Features and Their Clinical Correlates in Early Parkinson’s Disease

INTRODUCTION

Parkinson’s disease (PD)-associated psychosis is one of the most debilitating non-motor features and has been linked to multiple adverse outcomes. Ravina et al. (2007) proposed that the presence of illusions, false sense of presence, hallucinations or delusions occurring after the onset of PD, with a duration of one month that could not be attributed to another cause of parkinsonism or primary psychiatric disorder summarize the key components of the entity of psychosis in PD [1]. Since then, visual and auditory hallucinations in PD have been identified as the most prevalent manifestations of PD-psychosis in advanced stages of PD [2]. In contrast, hallucinations of other sensory modalities, minor hallucinations (presence hallucinations, passage hallucinations, and illusions) and delusional ideas have been relatively understudied. Minor psychotic phenomena have recently gained more attention, as they have been reported to be the most common type of hallucination in early stages of PD, even predating the onset of motor symptoms [3].

Several cognitive, non-motor and motor factors have been proposed to relate to psychotic features in moderate and late stages of PD. However, the clinical comorbidities that could be implicated in the exhibition of psychotic manifestations in earlier stages of PD are still under investigation [4].

The impact of, even minor, psychotic phenomena on the course of the disease is substantial, as they can lead to decreased quality of life and significant patient and caregiver distress. Therefore, the need of an early screening of psychotic symptoms becomes crucial [5]. The main objective of the current study was to evaluate a wide spectrum of psychotic symptoms in the early PD population, by using detailed psychiatric measures, and to identify clinical features that contribute to their occurrence.

MATERIALS AND METHODS

RESULTS

Spectrum of psychotic symptoms in patients with PD and healthy controls

Based on both CAARMS interview and 10PDQ scale, psychotic features were present in 39% (27 out of 70) of patients with PD and 4% (3 out of 74) of healthy controls [χ² = 26.0, df = 1, p < 0.001].

Information about modality and quality of psychotic manifestations was retrieved from the individual items of both 10PDQ and CAARMS scales. In healthy controls, auditory disturbances were more frequent followed by minor psychotic phenomena, visual and olfactory hallucinations. Illusions, tactile hallucinations, and delusional ideation were not observed in controls.

In the PD group, there was a predominance of passage hallucinations followed by auditory phenomena, illusions and formed visual hallucinations, olfactory and tactile abnormalities, delusional ideation and, finally, sense of presence (Fig. 1). Among all modalities of psychotic symptoms, the presence of minor hallucinations including passage hallucinations and illusions was significantly higher in patients with PD compared to healthy individuals, even after controlling for age and sex between the two study groups (χ² = 7.4, df = 1, p = 0.008 and χ² = 5.5, df = 1, p = 0.025, respectively, Supplementary Material).

OPEN IN VIEWER

Fig. 1

Psychotic features in patients with PD and healthy controls. The bar chart illustrates the distribution of several modalities of psychotic manifestations in healthy controls and patients with PD. In healthy subjects, auditory disturbances were more frequent followed by minor psychotic phenomena, visual and olfactory hallucinations. In the PD group, there was a predominance of passage hallucinations followed by auditory phenomena, illusions and formed visual hallucinations, olfactory and tactile abnormalities, delusional ideation and, finally, sense of presence. The presence of passage hallucinations and illusions was significantly higher in patients with PD, compared to healthy individuals.

DISCUSSION

The main objective of this study was to evaluate the prevalence of psychotic features in early PD using detailed measures. More than one third of patients with early PD reported at least one type of psychotic symptom, compared to healthy individuals. The prevalence of passage hallucinations and illusions reached a significant difference between controls and PD patients. Furthermore, motor examination and PD medication did not differ between patients with and without early-onset PD psychotic features. The presence of non-motor symptoms, RBD and depressive features seem to mildly increase the probability of exhibiting PD-associated psychotic manifestations. Frontal dysfunction marginally did not significantly correlate to psychotic features in this early PD population.

The novelty of our study was that a detailed psychiatric analysis by two separate scales was used to identify psychotic features in early PD. On the one hand, the implementation of CAARMS interview enabled clinicians to comprehensively probe into the nature of psychotic phenomena, given the open-ended structure of the questions that require further elaboration on examinees’ responses regarding severity and the frequency. Ambiguous answers were excluded, according to CAARMS global rating scale, in order to identify only frank psychotic phenomena. Moreover, CAARMS was the first instrument designed to identify imminent development of first-episode psychotic disorder, through an extensive exploration of subclinical psychopathology [18]. It maps, thus, the psychopathological symptoms of subclinical intensity, below the sensitivity threshold of other clinical scales, such as Brief Psychiatric Rating Scale (BPRS) and the Positive and Negative Symptom Scale (PANSS) that have been used in other PD clinical studies [19, 20]. On the other hand, 10PDQ scale is specifically designed for PD-associated psychosis and has been developed based on previous non-validated versions that have been implemented in clinical trials. [21, 22]. It explores the severity of the most common psychotic symptoms in PD, including sense of presence, visual, auditory, and olfactory hallucinations and delusional ideation of infidelity or persecution, in a 4-point scale with an excellent inter-rater and significant intra-rater reliability [17]. The worst psychotic experience is evaluated in terms of its duration, insight, affective consequences, and actions. In contradiction to the CAARMS interview, that documents self-distress, 10PDQ scale includes patients’ and family input to account for the poor insight many patients possess. Despite different methodological designs, an association was observed between 10PDQ scores and CAARMS scores addressing perceptual abnormalities (Supplementary Material). This finding implies that psychotic disturbances evaluated by CAARMS interview tend to correspond to the hallucinatory categories of the 10PDQ scale. Therefore, CAARMS assessment may provide a useful instrument for monitoring minor or major psychotic symptoms, especially of hallucinatory nature, in this sample of early PD patients, although its design and psychometric features address first-episode psychotic disorder individuals.

Until present, no “ideal” scale for PD psychosis has been identified [5]. According to the critique of MDS Task Force on Rating Scales in PD, several scales were designed specifically for the PD population, but are still inadequate to capture the entire phenomenology of psychotic features, such as MDS-UPDRS Part I scale [8]. Other psychosis scales derived from psychiatric research, for instance BPRS and PANSS scales, have been implemented, although several items relevant to schizophrenia are less useful to PD psychosis [19, 20]. Besides, these tools could be cumbersome for everyday clinical practice, due to their complex form. There is still lack of a valid questionnaire addressing psychotic manifestations in early PD and emphasizing minor hallucinations. Our attempt to combine an open-ended interview with another scale better at cataloging specific features, enabled us to identify a significantly high prevalence of psychotic features in the early period of PD.

One of the primary outcomes of our study was that perceptual abnormalities, especially minor, were dominant in the early PD population. This finding is in accordance with recent literature [23]. Early reports in the pre-levodopa period have described patients with illusions and visual hallucinations [24], even in the absence of prominent cognitive impairment. Several cohorts of early PD population, such as PRIAMO and PPMI studies, have observed an increasing prevalence of psychotic features over a 2-year period from 3% to 10% [25, 26]. In another study of “honeymoon” period of PD, Erro et al. (2016) identified a significant rise of prevalence of psychotic symptoms over a 4-year evaluation from 1.4% to 13.5% for hallucinations and 12.2% for delusions [27]. In our series, psychotic symptoms were present in 39% of patients with PD. According to literature, the highest prevalence of psychotic manifestations during the pre-levodopa era was 42%, including mainly minor phenomena in a sample of 50 drug-naïve PD patients [3]. Among these studies, there is a considerable discrepancy in point-prevalence of PD-associated psychosis due to methodological issues. Yet, all these data emphasize that psychotic symptoms can manifest early in the course of the disease.

A range of possible motor and non-motor factors that could be related to psychotic experiences was investigated in our PD sample. Patients with psychotic symptoms were older and had an older age of symptom onset compared to individuals without psychosis. Moreover, consistent with previous findings, performance in several motor measurements was similar between psychotic and non-psychotic early PD patients [3, 28–30]. It has been previously reported that visual hallucinations are observed in patients with greater motor severity of disease, even though this was generally in participants with more advanced PD than in our study (mean disease’s duration:1.7 years, HY stage≤2) [4, 31]. Furthermore, no predominance of akinetic-rigid type in patients with psychotic symptoms was detected in our sample. It seems that by splitting PD into tremor dominant or non-tremor dominant (akinetic-rigid or PIGD), the full spectrum of clinical manifestations in PD is not well-captured. Multi-domain classifications including motor, cognitive and psychiatric measures have shown that this classic tremor/PIGD dichotomy has been a less robust and sufficient instrument to predict clinically meaningful outcomes such as mortality, dementia, or PD-associated psychosis [32, 33]. Taken together, severity and type of motor symptoms do not appear to associate with psychotic symptoms, at least at the early period of PD.

There was no significant difference in proportion of participants on dopamine agonist, levodopa, or MAO inhibitors therapy between those with and without psychotic features. Even though individuals presenting psychotic manifestations had higher total LEDD, this difference marginally did not reach statistical significance. In clinical practice, PD psychotic features often emerge in the setting of augmentation of dopaminergic therapy, especially dopamine agonists, and subside with dose reduction or discontinuation [34]; this can occur even in non-demented PD cases, as an “isolated delusional syndrome” [35]. However, PD psychotic features have been recognized even in unmedicated patients [3, 36]. Our findings are in line with contemporary view of the role of medication in PD psychosis, as a “modifier rather than a necessary feature for the diagnosis of PD psychosis”, according to the NINDS-NIMH work group [1].

In our series, the major factors associated with PD psychotic features were non-motor symptoms including RBD. Prior studies have shown that autonomic symptom burden could be an independent risk factor for future hallucinations [37]. In addition, RBD has been widely associated with visual hallucinations and has been characterized as a predictor of future hallucinations [38]. One possible link between RBD and psychotic features in PD might be based on the cholinergic dysfunction observed in both conditions [39]. Reduced cholinergic nucleus 4 (Ch4) density has been recently associated with future psychotic symptoms in early PD subjects [37]. Another possible common basis of RBD and psychotic experiences in PD could be the dysfunction of prefrontal networks. Gan and colleagues (2021) have observed weaker positive couplings between visual network, default mode network (DMN), DMN and basal ganglia network and within DMN in patients with RBD-PD, while functional connectivity changes within the DMN and aberrant connectivity between posterior areas of DMN and visual-processing areas have been reported in patients with PD and minor hallucinations [40, 41]. Finally, reduced grey matter volume in hippocampal and parahippocampal regions, thalamus and limbic regions might be considered as a common substrate in PD patients with RBD and those with visual hallucinations [39].

Furthermore, the role of cognitive impairment in the presence of psychotic features was investigated in our sample. Despite the fact that total mean MoCA scores were relatively similar between patients with and without psychotic manifestations, by analyzing the individual items of MoCA scale, poorer performance in executive, visuospatial, naming and abstraction domains was revealed in patients with PD-associated psychotic symptoms (p < 0.05). This is in accordance with recent literature findings [42]. Visual hallucinations have been associated with multidomain cognitive dysfunction in PD including deficits in executive function [43], sustained attention [44] and visuospatial-perceptual functions [45]. Further, poor frontal lobe function has been identified in some cases of patients with psychosis in early PD [46, 47]. Neuroimaging findings on greater atrophy across frontal and parietal cortices further support our findings [34, 48]. A common idea for determining the role of executive and frontal functions in the emergence of psychotic features in early PD is the disruption in the processing of information across complex attentional networks including dorsal, ventral attentional and DMN networks [49]. Montagnese and colleagues (2023) reported mild language and naming difficulties in patients with visual hallucinations, after detailed examination of a mixture of language-related cognitive domains [50]. The early identification of this characteristic pattern of cognitive impairment in patients with psychotic manifestations, especially in the early stages of the disease, is of great significance to plan prevention therapies in the future and conduct stratified and optimal management of the ongoing neuropsychiatric symptoms.

Regarding other neuropsychiatric symptoms, depressive features mildly increased the probability of expressing psychotic features in our sample. Psychiatric symptoms in the depressive domain have been frequently recorded in PD with psychotic features and have been suggested to predispose to the development of psychotic experiences [4, 51]. These neurocognitive and neuropsychiatric changes could be grouped in a unified “serotoninergic” clinical subtype of psychosis, as Factor and colleagues (2017) have described in a novel classification scheme based on the main neurotransmitter implicated in each patient group (acetylcholine, serotonin, dopamine) [52]. More specifically, this “serotoninergic” type of psychosis refers to psychosis unrelated to global cognitive decline, in which hallucinatory phenomena (most commonly minor hallucinations) are more likely to occur in the context of depression or anxiety. The name of this class of psychosis reveals the underlying pathological mechanisms that bring psychosis and depression in proximity and implicates changes in the serotoninergic systems. Serotonin may contribute to the pathogenesis of complex visuals hallucinations probably via involvement of the serotonin 2 receptor, as clozapine, primavanserin, and other atypical antipsychotics effectively reduce the occurrence of visual abnormalities [53]. Upregulation of the 5HT2A receptors seems to occur in the prefrontal, visual, and temporal cortex areas only in PD patients with psychosis and the resulting enhanced activity at upregulated 5HT2A receptors in the temporal cortex and visual pathways could precipitate the onset of visual hallucinatory experiences [54]. Therefore, depressive and psychotic features in non-demented early PD population could share similar underlying mechanisms of predominantly brainstem and midbrain pathology.

We acknowledge that our study presents some limitations which need to be addressed carefully. Firstly, due to the cross-sectional design of the study, the identification of a temporal and causal relationship between several clinical factors and psychosis could not be precisely evaluated. In addition, a small proportion of the sample might be contaminated by Lewy body dementia cases as hallucinations early in the course of the disease should alert the clinician towards this diagnosis. However, since the diagnosis of PD always preceded the onset of psychotic features in this sample and none of the included patients had dementia at the time of evaluation, this bias should be negligible. Another limitation of the current study is related to the small size of our sample. In particular, patients were grouped together in the PDP group, regardless of whether they had minor type symptoms or more genuine psychotic phenomena. This meant that potential associations with specific symptoms (minor hallucinations/visual hallucinations rather than delusional ideation) and specific predictors could not be examined due to power concerns. Even though the recruitment of both PD patients and controls was based on a two-center basis, the representativeness of the current sample is not ideal. Yet, observer and self-selection bias were eliminated as much as possible. Finally, clinicians were not blind to clinical status (patients-controls), therefore potential expectational bias should be taken into consideration, in terms of magnifying differences in psychotic tendency in patients, compared to controls.

Overall, our study provides evidence that a variety of psychotic features, especially those of minor nature, are common in early PD. A comprehensive psychiatric evaluation including the CAARMS interview and the 10PDQ scale enabled the clinicians to monitor even subtle perceptual abnormalities and delusional ideas. Undoubtedly, a simple, sensitive, and valid screening tool addressing specifically minor phenomena could be part of the “clinical toolbox” in order to identify psychotic symptoms early. Further research in this area could examine whether the presence of early mild psychotic phenomena carries a prognostic significance for the development of more severe psychotic manifestations or possibly cognitive decline later in the disease course.