Sage Journals: Discover world-class research

Abstract

Background

Most structured visual hallucinations (VH) in Parkinson's disease (PD) involve animate-social objects, yet current theories fail to account for the prominent social component of VH in PD.

Objective

To study social perception in PD patients with VH in a behavioral task and its relationship with social traits such as perceived social isolation and anthropomorphism (tendency to ascribe human-like characteristics to non-human stimuli).

Methods

In this online web-based study, 28 PD with visual hallucinations (PD-VH), 55 PD patients without hallucinations (PD-nH), and 45 age-matched healthy controls (HC) performed a visual social task (human numerosity estimation), a control task, and filled an anthropomorphism and a loneliness questionnaire.

Results

Our data reveal a deficit in social visual perception characterized by a larger overestimation bias in human numerosity estimation in PD-VH versus control PD-nH and HC. Moreover, PD-VH had higher social traits of anthropomorphism and loneliness versus control PD-nH and HC and the overestimation bias was absent for non-human control stimuli.

Conclusions

These data describe a stronger social visual deficit and higher social traits in PD patients with VH, suggesting that neurodegenerative changes in PD-VH predominantly affect structures involved in social visual perception.

Plain language summary

Most structured visual hallucinations (VH) in Parkinson's disease (PD) involve animate-social objects, yet current theories fail to account for the prominent social component of VH in PD. Here, we conducted an online web-based study that 170 PD patients and 45 age-matched healthy controls (HC) carried out remotely from their homes, investigating (1) social perception in PD patients with VH (PD-VH) using a behavioral visual social task (human numerosity estimation) and (2) its relationship with social traits such as perceived social isolation and anthropomorphism (tendency to ascribe human-like characteristics to non-human stimuli). Our data reveals a deficit in social visual perception (larger overestimation bias in human numerosity estimation, absent for non-human control stimuli) and higher social traits (anthropomorphism and loneliness) in PD-VH versus PD patients without hallucinations and HC, suggesting that neurodegenerative changes in PD-VH predominantly affect functions and structures involved in social visual perception.

Keywords

neurodegeneration web-based assessment psychosis

Introduction

Although Parkinson's disease (PD) is traditionally defined as a movement disorder, PD affects numerous non-motor circuits leading to a variety of non-motor symptoms.¹ Hallucinations are one such frequent non-motor symptom,^2–5 with 30–50% of PD patient experiencing them regularly.^3,6,7 The prevalence of hallucinations has been found to increase >50% in advanced stages of PD^8,9 and often includes structured visual hallucinations (VH).¹⁰ VH almost invariably consist of humans or animals that are perceived centrally, often in dim light and are usually not frightening.¹¹ VH are among the most clinically relevant and debilitating non-motor symptoms and are associated with a more severe form of the disease characterized by a more rapid cognitive decline and dementia,^6,12–17 earlier home placement,^18–20 and a higher risk of mortality.^18,20

VH in PD have been linked with dopaminergic treatment^3,21 as well as visual dysfunction (for review see¹¹). Neurally, VH and visual deficits have been associated with altered structures in the visual pathways and retinal pathology.²² Neuroimaging studies also described impairments in posterior visual-perceptual brain networks^23–25 and attention networks,^26,27 as well as altered integration of bottom-up visual signals with top-down cognitive processes (i.e., dysfunctional integration between fronto-parietal and visual brain networks^28–31).

Structured VH in PD almost always involve animate-social objects, that is humans or animals (>80%),^3,4,32 whereas VH of inanimate objects are more rarely reported (<20%).³ However, current theories fail to account for the prominent social component of VH in PD. Even if impairments in higher visual perception have been observed,^33–39 to date no selective disturbance in the perception of social visual stimuli (versus control visual stimuli) has been reported. Humans are highly reliant on social interaction and its absence has been argued to induce significant physical and psychological distress and social hypervigilance.⁴⁰ Moreover, hallucinations in general^41,42 and in patients with psychosis⁴³ and with Alzheimer's disease have been linked to social traits of perceived social isolation.⁴⁴ However, social isolation has not been investigated in PD patients with VH nor in relation to social perception. Finally, PD patients have impairments in social cognition, such as inferring person's intentions, emotions and mental states (theory of mind),^45–47 and these latter functions have been related to the trait of anthropomorphism,⁴⁸ which is the tendency to ascribe human-like characteristics to non-human stimuli.⁴⁹ It has recently been argued that anthropomorphism^50–52 is associated with specific hallucinations⁵³; however, this has not been tested in patients with PD.

Recent work in human neuroscience has described specialized visual mechanisms⁵⁴ and brain regions involved in the processing of animate stimuli, including humans, such as the fusiform face area and the extrastriate body area,^55–57 underlining the biological importance of this function. Are structured VH in PD that predominantly involve social stimuli^3,4,32 associated with a deficit in the perception of social stimuli, resulting from interference with social visual processes?

Extending existing visual numerosity paradigms for simple visual stimuli^58,59 to the social domain, we recently developed a paradigm in which we briefly showed small groups of humans on a computer screen (range: 5 to 8 humans) to participants and asked them to judge the number of humans (Numerosity Estimation of Humans, NEH).⁶⁰ Using this NEH task, we showed that PD patients with minor hallucinations (i.e., presence hallucinations, passage hallucinations, visual illusions) showed a systematic deficit in perceiving groups of human stimuli, characterized by perceiving more people than were visually presented (NEH overestimation bias). This overestimation bias was not found in PD patients without minor hallucinations and, critically, was absent when using non-human inanimate objects (instead of the human stimuli) in a separate control task.

Here, we used a home-based online NEH task and tested the same group of 170 patients with PD, but investigated the links between VH, NEH performance, and social traits (perceived social isolation, anthropomorphism). In particular, we investigated whether PD patients with VH (PD-VH) (1) show a deficit in NEH characterized by an overestimation bias, whether (2) this NEH bias differs from healthy age-matched controls (HC), as well as from PD patients without any hallucinations (PD-nH), and whether (3) the bias was absent for non-human control objects (Numerosity Estimation Objects, NEO). To further investigate whether VH in PD are associated with alterations in the social domain, we tested whether (4) PD-VH and NEH are associated with two social traits (anthropomorphism, loneliness).^61,62

Methods

The data was acquired as part of an online web-based experiment,^60,63 carried out from August 2021 to June 2022. In this online web-based experiment, participants filled in some socio-demographic information, answered a questionnaire on alteration of perception corresponding to the frequency of hallucination occurrence in daily life, followed by a screen calibration procedure, the numerosity task, an Anthropomorphism Questionnaire (AQ), and a loneliness question (Figure 1). The experiment was available in French and English. The experiment could be performed on a computer or a tablet. Participants were instructed to perform the experiment while being alone in the room.

Figure 1.

Experimental protocol. In this online web-based experiment, participants first filled socio-demographic information and then answered a questionnaire on alteration of perception corresponding to the frequency of hallucination occurrence in daily life. Next, they performed a screen calibration procedure, then the numerosity task, and finally the anthropomorphism and loneliness questionnaires. For further detail, see the method section.

Study population

170 patients with PD (93 women, 77 men; age ranging from 42 to 79 years, mean ± SD age = 65.4 ± 7.83 years; PD duration ranging from 1 month to 25.6 years, mean ± SD PD duration = 6.44 ± 5.19 years) and 45 HC (27 women, 18 men; age ranging from 40 to 81 years, mean ± SD age = 62.1 ± 11 years) took part in this study. Of the 170 patients with PD and 45 HC, 83 patients with PD and 38 HC were kept for the analysis. Participants who had a very low refresh rate (less than 20 Hertz) or resolution (less than 800px on both axis) were excluded from the analysis (8 PD and 4 HC), as well as participants who reported visual disturbances that could negatively impact the task (12 PD and 3 HC) (Supplemental Material 1). Patients with PD who reported minor hallucinations (presence hallucination, passage hallucination or visual illusions), but not well-structured visual hallucinations were also excluded from the current analysis (n = 67). Based on obtained responses, patients were separated into a PD-VH group (PD patients with visual hallucinations; n = 28) and a PD-nH group (PD patients without hallucinations; n = 55). The demographic and clinical characteristics of PD-VH, PD-nH and HC are reported in Table 1. All participants consented to voluntarily participate in the study, prior to the beginning of the experiment. All data collected were anonymous. This study was considered as falling outside of the scope of the Swiss legislation regulating research on human subjects, so that the need for local ethics committee approval was waived (Commission Cantonale d'Ethique de la Recherche sur l'Être Humain – CCER, Switzerland – Req-2021-00378).

Table 1.

Clinical variables. Mean and standard deviation for clinical-demographic variables are shown (PD-VH, PD-nH, HC).

	PD-VH (N = 28)	PD-nH (N = 55)	HC (N = 38)	p
Gender	14 (M)	27 (M)	15 (M)	0.60 (χ²)
Age (y)	65.36 ± 8.21	65.11 ± 8.14	61.53 ± 10.89	0.05
PD duration (y)	7.03 ± 5.22	6.18 ± 5.26	N.A.	0.49
LEDD (mg/day)	402.56 ± 360.12 (N = 25)	276.56 ± 317.20 (N = 53)	N.A.	0.12

PD-VH: Parkinson's disease patients with visual hallucinations; PD-nH: Parkinson's disease patients with no hallucination; HC: healthy controls; LEDD: Levodopa Equivalent Daily Dose. LEDD was calculated as a sum of the conversion of each parkinsonian medication to Levodopa Equivalent Dose^64–66.

Questionnaires

The socio-demographic information to fill at the beginning of the experiment included gender, age, country, time, visual disturbances, PD diagnostic (if positive, date of diagnostic, onset side, current medication with daily dosage and time of last Levodopa intake). A self-assessment questionnaire about frequency of specific hallucinations in daily life (passage hallucination, presence hallucination, visual illusion, visual hallucination) was given (5-item Likert scale; Supplemental Material 2). The AQ⁶¹ contains 20 items assessing anthropomorphic beliefs and behaviors in childhood and adulthood, and has demonstrated good internal consistency (Cronbach α = 0.87 (95% CI = [0.84; 0.90]) in the current study). Loneliness was assessed with the following question: “How often do you feel lonely?” (7-item Likert scale (0 – Not at all; 6 – Very much so)).

Numerosity task

The numerosity estimation task was divided into two parts, an online human numerosity estimation (NEH) task and an online object numerosity estimation (NEO) task, in randomized order across participants.⁶⁰ Each task contained 40 trials, 10 trials per numerosity (ranging from 5 to 8), also randomized order across participants. There was a total of 10 different stimuli configuration possible per numerosity (ranging from 5 to 8). The stimuli were 1280px width by 720px height pictures and measured onscreen approximately 35.2 × 19.8 cm on the computer version and 17.6 × 9.9 cm on the tablet version. The stimuli were displayed for 250 ms. The configuration of the visual stimuli are described in Albert et al.⁶⁰ Trials with an evident mistake in reporting (answer ≤ 3 or answer > 50) or excessive response time (> 15 s) were excluded (in PD: 2.71% of NEH trials (90 of 3320 trials) and 3.10% of NEO (107 of 3320); in HC: 0.72% of NEH trials (11 of 1520 trials) and 0.46% of NEO (7 of 1520)). Prior to the numerosity estimation task, a screen calibration procedure allowing scaling the stimuli similarly for all participants was done, as described in Albert et al.⁶⁰

Statistical analysis

Statistical analysis were performed in R,⁶⁷ with packages lme4⁶⁸ and lmerTest⁶⁹ for linear models, package emmeans⁷⁰ for estimation of marginal means, and package effectsize⁷¹ for estimation of Cohen d effect size.

A linear mixed-effects model with population group (PD-VH, PD-nH, or HC), presented numerosity and type of stimuli (virtual human agents and control objects) as fixed effects, and random intercept for each subject was performed on the numerosity estimation data. Associations between questionnaires scores (anthropomorphism and loneliness) and group (PD-VH, PD-nH, or HC) were assessed with linear mixed-effects models, controlling for age and gender. The pairwise correlations between numerosity estimation, anthropomorphism and loneliness were assessed with linear models. The intensity of the correlations was assessed with Kendall's Tau. The significance of fixed effects was estimated with the likelihood Ratio Test. Post-hoc analysis was performed on the significant interactions and main effects and corresponded in pairwise comparisons using independent-samples t-tests, with reported p-values corrected for multiple comparisons with the Tukey correction.

Technical characteristics (calibration scale factor, refresh rate, resolution) were assessed with Levene's test for equality of variance and t-test for equality of means.

Technical setup

The online experiment was developed in-house using javascript and the jsPsych library⁷² (on client's side: javascript, html, css; on server's side: https server in node.js, nginx as a reverse proxy, running in two docker containers) and was hosted on an EPFL server in a dedicated Virtual Machine (1xvCPU, 1GB RAM, 40GB HDD) in demilitarized zone. Data was saved and stored on EPFL's servers.

Results

Structured visual hallucinations in patients with Parkinson's disease are associated with higher overestimation in human numerosity estimation task

We observed a general NEH overestimation in all three groups, but critically, this bias was larger in PD-VH group compared to PD-nH (t(125) = 4.05; p < 0.001; effect size = 0.63 (95% confidence interval = [0.32; 0.94])) and HC (t(125) = 3.96; p < 0.001; effect size = 0.67 (95% confidence interval = [0.33; 1.00])) (post-hoc analysis of the significant interaction between type of stimuli and group (F(2, 9327) = 34.89; p < 0.001); Supplemental Material 3). Thus, on average PD-VH patients reported seeing 2.4 more humans than were shown on the screen (see Figure 2(a), Supplemental Material 3). This overestimation bias was significantly present for all numerosities respectively, and increased with the number of humans shown. That is, PD-VH patients reported on average 1.5 more when 5 humans were shown (t(27) = 6.47; p < 0.001), 1.9 more with 6 humans (t(27) = 7.40; p < 0.001), 2.5 more with 7 humans (t(27) = 6.92; p < 0.001), and 3.5 more with 8 humans (t(27) = 6.45; p < 0.001). This NEH overestimation was larger in PD-VH group compared to HC for all numerosities (all p < 0.05), and compared to PD-nH for numerosities 6, 7 and 8 (all p < 0.05; for numerosity 5: p = 0.17) (post-hoc analysis of the significant interaction between presented numerosity, type of stimuli and group (F(6, 9327) = 2.94; p = 0.008); Supplemental Material 3). The difference between PD-nH and HC was not significant (t(125) = 0.22; p = 0.97; effect size = 0.05 (95% confidence interval = [-0.25; 0.31])).

Figure 2.

Online human and object numerosity estimation tasks. (a) Performance is shown in all three participant groups in the online human numerosity estimation task for PD-VH (red), PD-nH (brown) and HC (blue) separately. Each dot indicates the mean of an individual participant in the human numerosity estimation task. The dots with the bars indicate the mixed-effects linear regression between PD-VH (red), PD-nH (brown) and HC (blue). The occurrence of VH significantly modulates online human numerosity estimation task, with PD patients who reported well-structured visual hallucinations showing stronger online human numerosity overestimation, as compared to PD patients without hallucinations and healthy controls. (b) Performance is shown for the same participants, but in the online object numerosity estimation task. No statistical difference in online object numerosity estimation task was observed between groups. ***p ≤ 0.001. PD: Parkinson's disease; PD-VH: Parkinson's disease patients with visual hallucinations; PD-nH: Parkinson's disease patients with no hallucination; HC: healthy controls.

Moreover, an overestimation bias was also present in NEO, to a lesser extent than in NEH, but, critically, this bias was not significantly different in PD-VH group compared to PD-nH (t(125) = 1.32; p = 0.39; effect size = 0.21 (95% confidence interval = [-0.10; 0.52])) nor with HC (t(125) = 1.74; p = 0.20; effect size = 0.29 (95% confidence interval = [-0.04; 0.62])) (post hoc analysis of the significant interaction between type of stimuli and group (F(2, 9327) = 34.89; p < 0.001); Figure 2(b); Supplemental Material 3). Thus, PD-VH were more precise when estimating non-human control stimuli than when estimating human stimuli and, on average, only reported seeing 0.8 more objects than were shown on the screen. This PD-VH NEO overestimation was not significantly different for all numerosities compared to PD-nH (all p > 0.1) and HC (all p > 0.25) (post hoc analysis of the significant interaction between presented numerosity, type of stimuli and group (F(6, 9327) = 2.94; p = 0.008); Supplemental Material 3). The difference between PD-nH and HC was neither significant (t(125) = 0.60; p = 0.82; effect size = 0.09 (95% confidence interval = [-0.20; 0.37])).

Additional analysis revealed no statistical differences in response times between groups (PD-VH, PD-nH and HC; F(1118) = 2.73; p = 0.07; no main effect nor interaction; Supplemental Material 4).

Structured visual hallucinations in patients with Parkinson's disease are associated with higher anthropomorphism and subjective loneliness

Next we analyzed anthropomorphism across groups and found that anthropomorphism ratings differed across the three groups (significant main effect of group: F(2, 116) = 4.50; p = 0.013) (see Figure 3(a)). PD-VH had a higher anthropomorphism score compared to PD-nH (t(116) = 2.96; p = 0.010; effect size = 0.69 (95% confidence interval = [0.22; 1.16])) and HC (t(116) = 2.45; p = 0.042; effect size = 0.62 (95% confidence interval = [0.11; 1.12])) (Supplemental Material 5). The difference between PD-nH and HC was not significant (t(116) = 0.33; p = 0.94; effect size = 0.07 (95% confidence interval = [-0.50; 0.36])). Perceived loneliness also significantly differed between the three groups (significant main effect of group: F(2, 116) = 4.64; p = 0.012) (see Figure 3(b)), being highest in PD-VH and higher compared to PD-nH (t(116) = 2.89; p = 0.013; effect size = 0.67 (95% confidence interval = [0.20; 1.14])) and HC (t(116) = 2.76; p = 0.018; effect size = 0.70 (95% confidence interval = [0.19; 1.21])) (Supplemental Material 6).

Figure 3.

Anthropomorphism and loneliness traits. (a) Anthropomorphism score is shown for PD-VH (red), PD-nH (brown) and HC (blue) separately. Each dot indicates the anthropomorphism total score mean estimate. The dots with the bar on the left sides indicate the mixed-effects linear regression. The occurrence of hallucinations significantly modulates anthropomorphism, with PD patients who reported well-structured visual hallucinations showing higher anthropomorphism as compared to PD patients without hallucinations and healthy controls. (b) Loneliness is shown for the same three groups. The occurrence of hallucinations significantly modulates loneliness, with PD patients who reported well-structured visual hallucinations feeling lonelier as compared to PD patients without hallucinations and healthy controls. (c) Higher feelings of loneliness were associated with higher anthropomorphism scores. Confidence interval represents standard error bounds. Error bar represents 95% confidence interval. *p ≤ 0.05. PD: Parkinson's disease; PD-VH: Parkinson's disease patients with visual hallucinations; PD-nH: Parkinson's disease patients with no hallucination; HC: healthy controls.

There was a significant relationship between anthropomorphism and perceived loneliness in the overall sample (F(1, 119) = 16.59; p < 0.001) (see Figure 3(c)), with higher feeling of loneliness being associated with higher anthropomorphism (r_t= 0.22; p = 0.001). There was no significant relationship between NEH overestimation bias and anthropomorphism (F(1, 119) = 0.12; p = 0.73), nor between NEH overestimation bias and loneliness (F(1, 119) = 0.69; p = 0.41) (Supplemental Material 7), suggesting NEH overestimation being an independent mechanism from anthropomorphism and loneliness.

Technological online findings

There were no significant differences in any of the many gathered technical characteristics of the devices used between PD-VH, PD-nH and HC, including resolution (F(2, 118) = 0.071, p = 0.93), refresh rate (F(2, 118) = 1.10, p = 0.34), and calibration scale factor (F(2, 118) = 1.19, p = 0.31) (see Supplemental Material 8), showing that the differences between PD-VH, PD-nH and HC cannot be accounted for by technical confounds.

Discussion

Using a recently described online task to investigate social perception we investigated a large group of PD patients and HC, and asked them to estimate the number of humans (or non-human control objects) shown on the screen of a computer device. The present data reveal (1) a large overestimation bias for visual human stimuli, that was (2) larger in PD-VH as compared to control PD-nH and HC. This NEH bias was (3) absent for non-human control stimuli (objects), demonstrating that the visual deficit of PD-VH patients is a visual-perceptual deficit for social human stimuli and that the NEH overestimation bias is an implicit digital online marker for VH. As PD-VH patients also (4) showed higher social traits of anthropomorphism and loneliness, these data collectively suggest that neurodegenerative changes in PD-VH predominantly affect brain mechanisms of social perception.

Animate versus inanimate visual perception

A large array of visual pathomechanisms has been proposed for structured VH in PD.^{4,11,13,73,74} Reduced visual acuity, impaired contrast sensitivity, impaired color perception, as well as impaired object and face recognition have been reported in PD^73,75–77 and are more common or severe in PD patients with VH.^11,24,33,35 Moreover, cortical dysfunctions in PD patients with VH have been linked to the ventral (occipitotemporal) and the dorsal (occipitoparietal) visual pathways^13,74,78 and atrophy in primary visual cortex and in occipitotemporal and posterior parietal regions (i.e., ventral and dorsal visual streams) were reported in PD patients with VH. Despite the importance and consistency of these findings about altered visual structures in PD, they do not account for the observation that ∼80% of VH involve social visual stimuli: patients hallucinate humans or animals much more commonly than non-social objects.^3,4,32,79

The present data allow to behaviorally link this social aspect of VH to a social perceptual impairment in numerosity perception of visual humans, by revealing that the NEH bias was larger in PD-VH versus control PD-nH and HC and absent for non-human control objects. Extending previous observations of visual deficits to the social domain, we argue that VH and the NEH bias result from neurodegeneration of neural systems that are specialized for the visual perception and recognition of social stimuli. Extensive work in human neuroscience has described visual mechanisms in ventral temporo-occipital cortex that are highly specialized for the processing of social stimuli (for review see⁸⁰), such as the human face (i.e., fusiform face area, FFA^81–83) and the human body (extrastriate body area, EBA⁸⁴), forming a neural system for visual animacy detection.^85–87 Behavioral findings also suggest that the human perceptual system is ‘hyper-sensitive’ in detecting conspecifics, for evolutionary adaptive reasons^54,88,89; this was mirrored in the specificity found in the present NEH task, compared to performance with non-human inanimate control stimuli in the NEO task. We further point out the large magnitude of this human overestimation effect, with an average amplitude of perceiving 2.4 more humans than participants were actually shown Figure 2(a). Based on these findings we propose that an impairment of brain mechanisms involved in social perception in PD-VH not only leads to the structured VH of PD (that may consist of small groups of humans, as tested here), but also leads to overestimation of visual humans during the NEH task, as reported here in a large group of PD patients with hallucinations. Future in-laboratory studies need to corroborate these online behavioral findings, and should also acquire neuroimaging data, allowing to investigate the involvement of social visual cortex in VH related to NEH.⁸⁰

Importantly, social perception has been shown to rely on neural mechanisms extending beyond posterior visual areas, with involvement of the prefrontal and premotor cortices, for example in social cognition,⁹⁰ biological motion perception⁹¹ and action understanding and imitation.⁹² Also, electrical stimulation of the prefrontal cortex has been shown to induce complex VH including the perception of people,⁹³ suggesting that disruptions in these higher-level areas may also contribute to VH in PD. Additionally, VH in PD have been proposed to originate from dysfunction of the default mode network (DMN)^29,94 and frontal attentional network.⁹⁵ Our interpretation aligns with these models and suggests that DMN dysfunction may impact prefrontal or posterior social visual networks. The involvement of these brain areas and networks in NEH and VH in PD also seems of particular relevance, as recent brain imaging work revealed involvement of premotor and prefrontal regions in PD patients with minor hallucinations that generally precede VH.⁹⁶

Overall, we showed that the overestimation bias is an implicit marker for VH in PD (present study), for PH in PD (Albert et al.⁶⁰, study 2), and that a robotic procedure is able to induce PH experimentally in PD with PH.⁹⁶ However, it has not been shown whether an overestimation bias in NEH in PD patients with hallucinations would also reflect robot-induced PH. Future work is needed to test this.

Anthropomorphism and loneliness

Our data also reveal that PD-VH report higher feelings of loneliness as compared to PD-nH and HC. An association between hallucination, neurodegenerative disease and social isolation has previously only been shown for patients with Alzheimer's disease.⁴⁴ PD patients (n = 1527) with loneliness have been shown to experience greater symptom severity than those without,⁹⁷ and PD patients (n = 559) report higher feeling of loneliness than HC.⁹⁸ Moreover, it has been argued that increased loneliness in PD is caused by motor impairments and related lower quality of life and may lead to withdrawal and social isolation.⁹⁸ The present data show that especially PD-VH patients also have higher anthropomorphism scores as compared to PD-nH and HC and that both social traits (anthropomorphism and perceived social isolation) positively correlate with each other. Anthropomorphism is the tendency to ascribe human-like characteristics, motivations, intentions, or emotions to non-human agents⁹⁹ and has been argued to depend on sociality motivation among other factors.¹⁰⁰ The higher tendency to anthropomorphize ambiguous stimuli in PD-VH patients may thus reflect compensatory social mechanisms to ameliorate adverse social outcomes (e.g., loneliness) associated with PD and VH and patients’ progressive disconnection from their social environment.^101,102 Anthropomorphism in PD may also relate to pareidolias.⁴ Future studies should investigate how pareidolias and experimental approaches to pareidolia^103,104 relate to social traits and the present social visual task (NEH). This is also in line with the social deafferentation hypothesis of loneliness.^41,42 However, the present data found no correlation between the magnitude of the NEH overestimation bias and the strength of either social trait (anthropomorphism, loneliness). As loneliness was quantified with a single item in this online study, future studies may want to administer a more extensive questionnaire assessing several dimensions of loneliness, and also test PD patients on NEH tasks not online, but in the research laboratory.

To conclude, we argue that the predominantly anthropomorphic character of spontaneous VH,^4,79,105,106 the NEH overestimation bias, the stronger tendency to anthropomorphize and the higher loneliness in PD-VH are all indicative of an impairment of social processes. These data show that PD-VH suffer from social perceptual impairments, likely related to neurodegeneration that predominantly affects structures involved in social visual perception (and related social cognitive functions such as theory of mind that may also predominate in PD-VH versus PD-nH and HC).

Limitations

There are several limitations of our study. First, the data were acquired through an anonymized home-based online measurement, with limited information regarding the patients’ symptoms and other clinical variables. For example, strong motor impairments and reduced mobility could lead to the withdrawal from social activities, which could influence the loneliness score. Moreover, cognitive functions of our participants and cognitive (and motor) impairments could have affected performance in the NEH task. Finally, we cannot exclude that family members of caregivers somehow assisted the participants in task performance. Future clinical research carried out within a clinic or laboratory should explore this NEH overestimation bias marker jointly with detailed neurological, neuropsychological (e.g., Montreal Cognitive Assessment) and other clinical (e.g., Unified Parkinson's Disease Rating Scale) and socio-demographical (e.g., marital status, questionnaire assessing involvement in different types of social activities) evaluations. Second, while using a 7-item Likert scale to answer, the feeling of loneliness was assessed through a single self-report question, which limits its interpretation. Further research would benefit from a more complete questionnaire measuring different dimensions of loneliness. However, we also note that in PD, it would be particularly important to disentangle the mechanisms underlying the feeling of loneliness (social trait) from loneliness related to physical-social isolation resulting from poor mobility in PD.⁹⁸ This is particularly important as the exact directionality of the relationship between experiencing hallucinations and social isolation (related to advanced disease) currently remains unclear.^44,107,108 Third, while we found an association between anthropomorphism and loneliness, arguing for anthropomorphism as a social outcome of loneliness, this relationship is not straightforward, and the reverse relationship is not to be ruled out. Further research investigating the association between different dimensions of loneliness such as social disconnection and anthropomorphism would shed light on directionality.

Supplemental Material

sj-docx-1-pkn-10.1177_1877718X251336196 - Supplemental material for Visual hallucinations in Parkinson's disease are associated with deficits in social perception

Supplemental material, sj-docx-1-pkn-10.1177_1877718X251336196 for Visual hallucinations in Parkinson's disease are associated with deficits in social perception by Louis Albert, Neza Vehar, Jevita Potheegadoo, Fosco Bernasconi and Olaf Blanke in Journal of Parkinson's Disease

Footnotes

Acknowledgments

The authors thank all patients with Parkinson's disease and healthy controls for their participation,as well as Parkinson Schweiz,Parkinson's UK and Association France Parkinson in their help of recruiting patients with PD.

ORCID iDs

Louis Albert

Neza Vehar

Jevita Potheegadoo

Fosco Bernasconi

Olaf Blanke

Ethical considerations

All data collected were anonymous. This study was considered as falling outside of the scope of the swiss legislation regulating research on human subjects,so that the need for local ethics committee approval was waived (Commission Cantonale d'Ethique de la Recherche sur l'Être Humain – CCER,Switzerland – Req-2021-00378).

Consent to participate

All participants consented to voluntarily participate in the study,prior to the beginning of the experiment.

Consent for publication

Not applicable

Author contributions/CRediT

Louis Albert: Conceptualization;Methodology;Software;Validation;Formal analysis;Investigation;Resources;Data curation;Writing - original draft;Writing - review & editing;Visualization;Project administration. Neza Vehar: Conceptualization;Writing - original draft;Writing - review & editing. Jevita Potheegadoo: Resources;Writing - review & editing;Funding acquisition. Fosco Bernasconi: Conceptualization;Methodology;Validation;Writing - review & editing;Supervision;Funding acquisition. Olaf Blanke: Conceptualization;Methodology;Resources;Writing - original draft;Writing - review & editing;Supervision;Project administration;Funding acquisition.

Funding

This research was supported by the Swiss National Science Foundation (grants number SNF 320030_188798 and SNF CRSII5_189989) to O.B.;two generous donors advised by CARIGEST SA (Fondazione Teofilo Rossi di Montelera e di Premuda and a second one wishing to remain anonymous) to O.B.;Bertarelli Foundation to O.B.;Empiris Foundation to O.B;Wyss Center for Bio and Neuro Engineering (Lighthouse project: Non-invasive Neuromodulation of Subcortical Structures) to O.B.

Conflicting interests

L.A.,O.B.,F.B.,and J.P. are inventors on patent WO 2024/161264 A1 held by the Swiss Federal Institute (EPFL) that covers a solution for measuring and quantifying an impairment in numerosity estimation in human subjects.

Data availability

The main data supporting the results in this study are available within the paper and its supplemental material. The source data have been deposited in

Supplemental material

Supplemental material for this article is available online.

References

Chaudhuri

Healy

Schapira

. Non-motor symptoms of Parkinson’s disease: diagnosis and management. Lancet Neurol 2006; 5: 235–245.

Fénelon

. Psychosis in Parkinson’s disease: phenomenology, frequency, risk factors, and current understanding of pathophysiologic mechanisms. CNS Spectr 2008; 13: 18–25.

Fénelon

Mahieux

Huon

, et al. Hallucinations in Parkinson’s disease: prevalence, phenomenology and risk factors. Brain 2000; 123: 733–745.

Ffytche

Creese

Politis

, et al. The psychosis spectrum in Parkinson disease. Nat Rev Neurol 2017; 13: 81–95.

Lenka

Hegde

Jhunjhunwala

, et al. Interactions of visual hallucinations, rapid eye movement sleep behavior disorder and cognitive impairment in Parkinson’s disease: a review. Parkinsonism Relat Disord 2016; 22: 1–8.

Bernasconi

Pagonabarraga

Bejr-Kasem

, et al. Theta oscillations and minor hallucinations in Parkinson’s disease reveal decrease in frontal lobe functions and later cognitive decline. Nat Ment Health 2023; 1: 477–488.

Lenka

Pagonabarraga

Pal

, et al. Minor hallucinations in Parkinson disease: a subtle symptom with major clinical implications. Neurology 2019; 93: 259–266.

Aarsland

Batzu

Halliday

, et al. Parkinson disease-associated cognitive impairment. Nat Rev Dis Primer 2021; 7: 47.

Levin

Hasan

Höglinger

. Psychosis in Parkinson’s disease: identification, prevention and treatment. J Neural Transm (Vienna) 2016; 123: 45–50.

10.

Barnes

David

. Visual hallucinations in Parkinson’s disease: a review and phenomenological survey. J Neurol Neurosurg Psychiatry 2001; 70: 727–733.

11.

Weil

Schrag

Warren

, et al. Visual dysfunction in Parkinson’s disease. Brain 2016; 139: 2827–2843.

12.

Aarsland

Hutchinson

Larsen

. Cognitive, psychiatric and motor response to galantamine in Parkinson’s disease with dementia. Int J Geriatr Psychiatry 2003; 18: 937–941.

13.

Diederich

Fénelon

Stebbins

, et al. Hallucinations in Parkinson disease. Nat Rev Neurol 2009; 5: 331–342.

14.

Forsaa

Larsen

Wentzel-Larsen

, et al. What predicts mortality in Parkinson disease?: a prospective population-based long-term study. Neurology 2010; 75: 1270–1276.

15.

Galvin

Pollack

Morris

. Clinical phenotype of Parkinson disease dementia. Neurology 2006; 67: 1605–1611.

16.

Marinus

Zhu

Marras

, et al. Risk factors for non-motor symptoms in Parkinson’s disease. Lancet Neurol 2018; 17: 559–568.

17.

McDermott

Marder

, et al. Incidence of and risk factors for cognitive impairment in an early Parkinson disease clinical trial cohort. Neurology 2009; 73: 1469–1477.

18.

Aarsland

Larsen

Tandberg

, et al. Predictors of nursing home placement in Parkinson’s disease: a population-based, prospective study. J Am Geriatr Soc 2000; 48: 938–942.

19.

Goetz

Emre

Dubois

. Parkinson’s disease dementia: definitions, guidelines, and research perspectives in diagnosis. Ann Neurol 2008; 64: S81–S92.

20.

Gonzalez

Dalen

Maple-Grødem

, et al. Parkinson’s disease clinical milestones and mortality. NPJ Parkinsons Dis 2022; 8: 58.

21.

Ravina

Marder

Fernandez

, et al. Diagnostic criteria for psychosis in Parkinson’s disease: report of an NINDS, NIMH work group. Mov Disord 2007; 22: 1061–1068.

22.

Lee

J-Y

Martin-Bastida

Murueta-Goyena

, et al. Multimodal brain and retinal imaging of dopaminergic degeneration in Parkinson disease. Nat Rev Neurol 2022; 18: 203–220.

23.

Goldman

Stebbins

Dinh

, et al. Visuoperceptive region atrophy independent of cognitive status in patients with Parkinson’s disease with hallucinations. Brain 2014; 137: 849–859.

24.

Ramirez-Ruiz

Junque

Marti

M-J

, et al. Cognitive changes in Parkinson’s disease patients with visual hallucinations. Dement Geriatr Cogn Disord 2007; 23: 281–288.

25.

Watanabe

Senda

Kato

, et al. Cortical and subcortical brain atrophy in Parkinson’s disease with visual hallucination. Mov Disord 2013; 28: 1732–1736.

26.

Meppelink

Koerts

Borg

, et al. Visual object recognition and attention in Parkinson’s disease patients with visual hallucinations. Mov Disord 2008; 23: 1906–1912.

27.

Ballard

Aarsland

McKeith

, et al. Fluctuations in attention: PD dementia vs DLB with parkinsonism. Neurology 2002; 59: 1714–1720.

28.

Muller

Shine

Halliday

, et al. Visual hallucinations in Parkinson’s disease: theoretical models. Mov Disord 2014; 29: 1591–1598.

29.

Shine

Halliday

Naismith

, et al.

Visual misperceptions and hallucinations in Parkinson’s disease: dysfunction of attentional control networks?

Mov Disord 2011; 26: 2154–2159.

30.

Shine

Halliday

Gilat

, et al. The role of dysfunctional attentional control networks in visual misperceptions in Parkinson’s disease. Hum Brain Mapp 2014; 35: 2206–2219.

31.

Goetz

Vaughan

Goldman

, et al. I finally see what you see: Parkinson’s disease visual hallucinations captured with functional neuroimaging. Mov Disord 2014; 29: 115–117.

32.

Mosimann

Rowan

Partington

, et al. Characteristics of visual hallucinations in Parkinson disease dementia and dementia with Lewy bodies. Am J Geriatr Psychiatry 2006; 14: 153–160.

33.

Barnes

Boubert

Harris

, et al. Reality monitoring and visual hallucinations in Parkinson’s disease. Neuropsychologia 2003; 41: 565–574.

34.

Ramírez-Ruiz

Junqué

Martí

M-J

, et al. Neuropsychological deficits in Parkinson’s disease patients with visual hallucinations. Mov Disord 2006; 21: 1483–1487.

35.

Gallagher

Parkkinen

O’Sullivan

, et al. Testing an aetiological model of visual hallucinations in Parkinson’s disease. Brain 2011; 134: 3299–3309.

36.

Coundouris

Adams

Grainger

, et al. Social perceptual function in Parkinson’s disease: a meta-analysis. Neurosci Biobehav Rev 2019; 104: 255–267.

37.

Jaywant

Shiffrar

Roy

, et al. Impaired perception of biological motion in Parkinson’s disease. Neuropsychology 2016; 30: 720–730.

38.

Kloeters

Hartmann

Pundmann

, et al. Impaired perception of human movements in Parkinson’s disease. Behav Brain Res 2017; 317: 88–94.

39.

Liu

Cao

, et al. Dopaminergic modulation of biological motion perception in patients with Parkinson’s disease. Sci Rep 2017; 7: 10159.

40.

Hawkley

Cacioppo

. Loneliness matters: a theoretical and empirical review of consequences and mechanisms. Ann Behav Med 2010; 40: 218–227.

41.

Hoffman

. A social deafferentation hypothesis for induction of active schizophrenia. Schizophr Bull 2007; 33: 1066–1070.

42.

Brederoo

de Boer

Linszen

MMJ

, et al. Social deafferentation and the relation between loneliness and hallucinations. Schizophr Bull 2023; 49: S25–S32.

43.

Michalska da Rocha

Rhodes

Vasilopoulou

, et al. Loneliness in psychosis: a meta-analytical review. Schizophr Bull 2018; 44: 114–125.

44.

El Haj

Jardri

Larøi

, et al. Hallucinations, loneliness, and social isolation in Alzheimer’s disease. Cogn Neuropsychiatry 2016; 21: 1–13.

45.

Bodden

Dodel

Kalbe

. Theory of mind in Parkinson’s disease and related basal ganglia disorders: a systematic review. Mov Disord 2010; 25: 13–27.

46.

Coundouris

Adams

Henry

. Empathy and theory of mind in Parkinson’s disease: a meta-analysis. Neurosci Biobehav Rev 2020; 109: 92–102.

47.

Saltzman

Strauss

Hunter

, et al. Theory of mind and executive functions in normal human aging and Parkinson’s disease. J Int Neuropsychol Soc 2000; 6: 781–788.

48.

Cullen

Kanai

Bahrami

, et al. Individual differences in anthropomorphic attributions and human brain structure. Soc Cogn Affect Neurosci 2014; 9: 1276–1280.

49.

Epley

Waytz

Akalis

, et al. When we need a human: motivational determinants of anthropomorphism. Soc Cogn 2008; 26: 143–155.

50.

Guthrie

. Faces in the clouds: A new theory of religion. Oxford University Press, 1995.

51.

Barrett

Johnson

. The role of control in attributing intentional agency to inanimate objects. J Cogn Cult 2003; 3: 208–217.

52.

Barrett

Lanman

. The science of religious beliefs. Religion 2008; 38: 109–124.

53.

Vehar

Potheegadoo

Blanke

. Linking agent detection of invisible presences to the self: relevance for religious and spiritual experiences. Front Behav Neurosci 2022; 16: 952736.

54.

Papeo

Stein

Soto-Faraco

. The two-body inversion effect. Psychol Sci 2017; 28: 369–379.

55.

McCarthy

Puce

Gore

, et al. Face-specific processing in the human fusiform gyrus. J Cogn Neurosci 1997; 9: 605–610.

56.

Peelen

Downing

. Selectivity for the human body in the fusiform gyrus. J Neurophysiol 2005; 93: 603–608.

57.

Astafiev

Stanley

Shulman

, et al. Extrastriate body area in human occipital cortex responds to the performance of motor actions. Nat Neurosci 2004; 7: 542–548.

58.

Jevons

. The power of numerical discrimination. Nature 1871; 3: 281–282.

59.

Kaufman

Lord

Reese

, et al. The discrimination of visual number. Am J Psychol 1949; 62: 498–525.

60.

Albert

Potheegadoo

Herbelin

, et al. Numerosity estimation of virtual humans as a digital-robotic marker for hallucinations in Parkinson’s disease. Nat Commun 2024; 15: 1905.

61.

Neave

Jackson

Saxton

, et al. The influence of anthropomorphic tendencies on human hoarding behaviours. Personal Individ Differ 2015; 72: 214–219.

62.

Prenger

MTM

Madray

Van Hedger

, et al. Social symptoms of Parkinson’s disease. Parkinsons Dis 2020; 2020: e8846544.

63.

Albert

Potheegadoo

Herbelin

, et al. Digital-robotic markers for hallucinations in Parkinson’s disease. bioRxiv 2023. DOI: https://doi.org/10.1101/2023.06.14.544929.

64.

Nyholm

Jost

. An updated calculator for determining levodopa-equivalent dose. Neurol Res Pract 2021; 3: 58.

65.

Schade

Mollenhauer

Trenkwalder

. Levodopa equivalent dose conversion factors: an updated proposal including opicapone and safinamide. Mov Disord Clin Pract 2020; 7: 343–345.

66.

Tomlinson

Stowe

Patel

, et al. Systematic review of levodopa dose equivalency reporting in Parkinson’s disease. Mov Disord 2010; 25: 2649–2653.

67.

R Core Team. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing, 2022. https://www.R-project.org/

68.

Bates

Mächler

Bolker

, et al. Fitting linear mixed-effects models using lme4. J Stat Softw 2015; 67: 1–48.

69.

Kuznetsova

Brockhoff

Christensen

RHB

. Lmertest package: tests in linear mixed effects models. J Stat Softw 2017; 82: 1–26.

70.

Lenth

Bolker

Buerkner

, et al. emmeans: Estimated marginal means, aka least-squares means, 2024. https://cran.r-project.org/web/packages/emmeans/index.html .

71.

Ben-Shachar

Lüdecke

Makowski

. Effectsize: estimation of effect size indices and standardized parameters. J Open Source Softw 2020; 5: 2815.

72.

de Leeuw

. Jspsych: a JavaScript library for creating behavioral experiments in a Web browser. Behav Res Methods 2015; 47: 1–12.

73.

Archibald

Clarke

Mosimann

, et al. The retina in Parkinson’s disease. Brain 2009; 132: 1128–1145.

74.

Collerton

Barnes

Diederich

, et al. Understanding visual hallucinations: a new synthesis. Neurosci Biobehav Rev 2023; 150: 105208.

75.

Archibald

Clarke

Mosimann

, et al. Visual symptoms in Parkinson’s disease and Parkinson’s disease dementia. Mov Disord 2011; 26: 2387–2395.

76.

Holroyd

Currie

Wooten

. Prospective study of hallucinations and delusions in Parkinson’s disease. J Neurol Neurosurg Psychiatry 2001; 70: 734–738.

77.

Matsui

Udaka

Tamura

, et al. Impaired visual acuity as a risk factor for visual hallucinations in Parkinson’s disease. J Geriatr Psychiatry Neurol 2006; 19: 36–40.

78.

Pagonabarraga

Bejr-Kasem

Martinez-Horta

, et al. Parkinson disease psychosis: from phenomenology to neurobiological mechanisms. Nat Rev Neurol 2024; 20: 135–150.

79.

Boubert

Barnes

. Phenomenology of visual hallucinations and their relationship to cognitive profile in Parkinson’s disease patients: preliminary observations. Sage Open 2015; 5: 2158244015585827.

80.

Peelen

Downing

. The neural basis of visual body perception. Nat Rev Neurosci 2007; 8: 636–648.

81.

Fan

Wang

Shao

, et al. The bottom-up and top-down processing of faces in the human occipitotemporal cortex. eLife 2020; 9: e48764.

82.

Kanwisher

McDermott

Chun

. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J Neurosci 1997; 17: 4302–4311.

83.

Kanwisher

Yovel

. The fusiform face area: a cortical region specialized for the perception of faces. Philos Trans R Soc B Biol Sci 2006; 361: 2109–2128.

84.

Downing

Jiang

Shuman

, et al. A cortical area selective for visual processing of the human body. Science 2001; 293: 2470–2473.

85.

Chao

Haxby

Martin

. Attribute-based neural substrates in temporal cortex for perceiving and knowing about objects. Nat Neurosci 1999; 2: 913–919.

86.

Konkle

Caramazza

. Tripartite organization of the ventral stream by animacy and object size. J Neurosci 2013; 33: 10235–10242.

87.

Grill-Spector

Weiner

. The functional architecture of the ventral temporal cortex and its role in categorization. Nat Rev Neurosci 2014; 15: 536–548.

88.

New

Cosmides

Tooby

. Category-specific attention for animals reflects ancestral priorities, not expertise. Proc Natl Acad Sci U S A 2007; 104: 16598–16603.

89.

Stein

Sterzer

Peelen

. Privileged detection of conspecifics: evidence from inversion effects during continuous flash suppression. Cognition 2012; 125: 64–79.

90.

Amodio

Frith

. Meeting of minds: the medial frontal cortex and social cognition. Nat Rev Neurosci 2006; 7: 268–277.

91.

Saygin

Wilson

Hagler

, et al. Point-light biological motion perception activates human premotor cortex. J Neurosci 2004; 24: 6181–6188.

92.

Rizzolatti

Fogassi

Gallese

. Motor and cognitive functions of the ventral premotor cortex. Curr Opin Neurobiol 2002; 12: 149–154.

93.

Blanke

Landis

Seeck

. Electrical cortical stimulation of the human prefrontal cortex evokes complex visual hallucinations. Epilepsy Behav 2000; 1: 356–361.

94.

Franciotti

Delli Pizzi

Perfetti

, et al. Default mode network links to visual hallucinations: a comparison between Parkinson’s disease and multiple system atrophy. Mov Disord 2015; 30: 1237–1247.

95.

Onofrj

Espay

Bonanni

, et al. Hallucinations, somatic-functional disorders of PD-DLB as expressions of thalamic dysfunction. Mov Disord 2019; 34: 1100–1111.

96.

Bernasconi

Blondiaux

Potheegadoo

, et al. Robot-induced hallucinations in Parkinson’s disease depend on altered sensorimotor processing in fronto-temporal network. Sci Transl Med 2021; 13: eabc8362.

97.

Subramanian

Farahnik

Mischley

. Synergy of pandemics-social isolation is associated with worsened Parkinson severity and quality of life. NPJ Parkinsons Dis 2020; 6: 28.

98.

Prell

Schönenberg

Heimrich

. The impact of loneliness on quality of life in people with Parkinson’s disease: results from the Survey of Health, Ageing and Retirement in Europe. Front Med (Lausanne) 2023; 10: 1183289.

99.

Epley

Waytz

Cacioppo

. On seeing human: a three-factor theory of anthropomorphism. Psychol Rev 2007; 114: 864–886.

100.

Shin

Kim

. My computer is more thoughtful than you: loneliness, anthropomorphism and dehumanization. Curr Psychol 2020; 39: 445–453.

101.

Bartz

Tchalova

Fenerci

. Reminders of social connection can attenuate anthropomorphism: a replication and extension of Epley, Akalis, Waytz, and Cacioppo (2008). Psychol Sci 2016; 27: 1644–1650.

102.

Epley

Akalis

Waytz

, et al. Creating social connection through inferential reproduction: loneliness and perceived agency in gadgets, gods, and greyhounds. Psychol Sci 2008; 19: 114–120.

103.

Mamiya

Nishio

Watanabe

, et al. The Pareidolia Test: a simple neuropsychological test measuring visual hallucination-like illusions. PLoS One 2016; 11: e0154713.

104.

Uchiyama

Nishio

Yokoi

, et al. Pareidolia in Parkinson’s disease without dementia: a positron emission tomography study. Parkinsonism Relat Disord 2015; 21: 603–609.

105.

Greene

Cote

Fahn

. Treatment of drug-induced psychosis in Parkinson’s disease with clozapine. Adv Neurol 1993; 60: 703–706.

106.

Urwyler

Nef

Müri

, et al. Visual hallucinations in eye disease and Lewy body disease. Am J Geriatr Psychiatry 2016; 24: 350–358.

107.

Schlier

Winkler

Jaya

, et al. Fluctuations in hallucination spectrum experiences co-vary with social defeat but not with social deafferentation. A 3-week daily assessment study. Cogn Ther Res 2018; 42: 92–102.

108.

Brown

Freeman

Loe

, et al. Paranoia and unusual sensory experiences in Parkinson’s disease. Aging Ment Health 2025: 1–16. DOI: https://doi.org/10.1080/13607863.2025.2450258.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.80 MB

0.00 MB