Theory of mind in Parkinson's disease and related basal ganglia disorders: A systematic review



Theory of mind (ToM), defined as the ability to infer other people's mental states, is a crucial prerequisite of human social interaction and a major topic of interest in the recent neuroscientific research. It has been proposed that ToM is mediated by a complex neuroanatomical network that includes the medial prefrontal cortex, the anterior gyrus cinguli, the sulcus temporalis superior, the temporal poles, and the amygdala. Various neurological and psychiatric diseases are accompanied by aspects of dysfunctional ToM processing. In this review, the association between basal ganglia, involved in the organization of complex cognitive and emotional behavior, and ToM processing is discussed. The purpose of this review is to provide an overview of research on ToM abilities in basal ganglia disorders, especially Parkinson's Disease and related disorders. © 2009 Movement Disorder Society


Social cognitive1 and affective neuroscience2 are rapidly growing fields of research. One topic that has attracted vast interest recently is “Theory of Mind” (ToM), defined as the ability to infer other persons' mental states, including their beliefs, desires, and intentions. ToM is helpful in explaining and predicting other people's behavior.3 As a central aspect of social cognition, ToM can be regarded as an essential prerequisite for successful human social interaction.1

Three decades ago the term “Theory of Mind“ was first introduced by Premack and Woodruff,4 two researchers who investigated mentalizing abilities in macaques. For a longer period of time, the concept had predominantly inspired the field of developmental psychology.5 ToM has been thoroughly studied in autism spectrum disorders as these have long been characterized by difficulties with social interaction and are currently regarded as a set of disorders in which a lack of ToM ability is a core deficit.6 Recently, neuroscientists have begun to investigate acquired ToM deficits in various neurological and psychiatric diseases. As such, ToM dysfunctions have been described in patients with schizophrenia,7 depression,8 frontotemporal dementia,9 epilepsy,10 and focal lesions of the frontal lobe,11 amongst others. The investigation of ToM deficits in basal ganglia disorders has begun recently.12–14 This article offers arguments in support of the hypothesis that ToM dysfunctions may be a frequent consequence of Parkinson's disease and other basal ganglia disorders and it provides a comprehensive overview of recent data from this field of research.


ToM is regarded as a complex neuropsychological function,7, 15–17 partly associated with other cognitive functions. Recently, subcomponents of ToM have been defined, suggesting a dissociation between “cognitive ToM” and “affective ToM.” Although cognitive ToM is thought to require cognitive understanding of the difference between the speaker's knowledge and that of the listener (knowledge about beliefs), affective ToM is believed to additionally require an empathic appreciation of the listener's emotional state (knowledge about emotions).7 The systematic investigation of these subcomponents has just begun, suggesting that both subcomponents can be impaired independently from each other.7, 11, 12 Two mechanisms have been proposed to explain the process of taking another person's perspective, and it is possible that these mechanisms correspond to the two postulated subcomponents of ToM.11, 16 “Simulation Theory” assumes that the mental states of others are understood by “simulating” those mental states via one's own resonant states and thus “slipping into the other person's shoes”.8 In contrast, the “theory theory” posits that the mental states of others are inferred rationally by a knowledge system that is independent from one's own mental states.19 It has been suggested that cognitive ToM may primarily constitute a cognitive process corresponding to the “theory theory,” while simulation may be the underlying mechanism for affective ToM.16, 20 However, it is possible that these mechanisms, instead of being strictly separated, can be processed in parallel, with one or the other being stronger in different situations, depending on the type of ToM process. Remarkably, alterations in emotional processing are described as consequences of basal ganglia dysfunction.21 Emotional processing may interact widely with ToM, especially its affective subcomponent. Furthermore, emotion recognition can be seen as a basic skill within the context of the comprehensive ToM ability.

As a crucial prerequisite for social human interaction, ToM can be seen as an ability highly relevant for various types of social situations. Decoding and attributing mental states as intentions and emotions of a counterpart enables and deepens social relationships by facilitating appreciative and sympathetic communication. Even the understanding of humor and sarcasm, as sophisticated communication skills in human cultural development, depends on ToM abilities. Furthermore, ToM is involved in the experience of complex social emotions such as embarrassment, shame or sympathy, emotions that cannot be perceived without referring to the mental state of a counterpart. Thus, impairments of ToM in neurological and psychiatric disorders may have vast impact on a patient's life.


ToM is thought to be mediated by a complex neuroanatomical network that includes the medial prefrontal cortex, the superior temporal sulcus region and the temporoparietal junction, the temporal pole and the amygdala1, 3 (Fig. 1). Recently, Abu-Akel22 proposed that temporal structures are more important in processing ToM, and as such they mediate ToM competence, while prefrontal structures are especially relevant for performance as it allows one to apply ToM in specific situations. Furthermore, some studies suggest that different parts of the prefrontal cortex are engaged in affective as opposed to cognitive ToM. Evidence for the special role of the ventromedial prefrontal cortex (VMPFC) in affective ToM comes from lesion11, 23 and neuroimaging studies.15, 17 In cognitive ToM, it has been suggested that additional lateral, and specifically dorsolateral, parts of the prefrontal cortex (DLPFC) are recruited.17, 24, 25 Both ventromedial and dorsolateral parts of the prefrontal cortex are closely related to subcortical structures. These “frontostriatal circuitries” connect the basal ganglia via the thalamus to areas of the prefrontal cortex (Fig. 2), and thus are involved in affective and cognitive aspects of behavior.26 In addition, basal ganglia's contributions to ToM are supported by the “neurochemical hypothesis of ToM.” Abu-Akel, who proposed this hypothesis,27 discusses ToM deficits of patients with schizophrenia as a consequence of their broad dysfunction within the dopaminergic system, which is mainly innervated by basal ganglia structures. However, the contributions of neurotransmitters,25 in particular dopamine12, 27 to ToM are recently attracting vast research interest. Interestingly, evidence from animal studies also suggests the involvement of dopamine in social behavior, i.e., pair bonding and attachment.28 Another focus of interest is oxytocin, a neuropeptid known to be involved in social bonding and affiliation29, 30 as well as trust building in humans.31 Remarkably, oxytocin enhances social cognition,32 e.g., ToM.33 The interaction of dopamine and oxytocin within the mesolimbic system has been conceptualized as a process underlying complex social behavior.34, 35

Figure 1.

Key neuroanatomical structures for ToM processing. Red: ventromedial prefrontal cortex; green: superior temporal sulcus region; blue: amygdala (subcortical); light blue: temporal pole.

Figure 2.

Frontostriatal-prefrontal circuitry. blue: dorsolateral prefrontal-striatal-circuitry; red: orbitofrontal-limbic-striatal circuitry. DLPFC: dorsolateral-prefrontal cortex; OFC: orbitofrontal cortex; NCC: nucleus accumbens; GPi: globus pallidus interna; GPe: globus pallidus externa; SNpr: substantia nigra pars reticulata; SNpc: substantia nigra pars compacta; VTA: ventral tegmental area.


In this review, we compassed the most prevalent basal ganglia disorders including Parkinson's disease, atypical Parkinson syndromes (multiple system atrophy, progressive supranuclear palsy, corticobasal ganglionic degeneration, Lewy body disease), restless legs syndrome and Huntington's disease. To identify studies that examine ToM functions in those disorders, we performed a systematic literature search in the databases Pubmed, Psyndex, and Psychinfo using a search strategy including the keywords “Theory of Mind,” “social cognition,” “empathy,” “mindreading,” “mentalizing,” or “perspective taking” in combination with “Parkinson,” “multiple system atrophy” or “MSA,” “progressive supranuclear palsy” or “PSP,” “corticobasal ganglionic degeneration” or “CBD,” “Lewy body,” “restless legs” or “RLS” or “Huntington.” We also checked the reference lists of each relevant study that resulted from this search for further appropriate articles. There was no restriction on year of publication or language of the study. Articles published through March 2009 were included. 16 articles were identified as relevant. Table 1 provides a detailed summary of the studies reported.

Table 1. Included papers on theory of mind in basal ganglia disorders
 ReferenceStudy samples*ToM functions assessed**Used paradigmsMain findingsRemarks
  • *

    Only study samples in which ToM or related abilities were assessed and which are reviewed are listed.

  • **

    Terms as used by the respective authors.

  • PD, Parkinson's disease; jPD, juvenile Parkinsonism; CBD, corticobasal degeneration; PSP, progressive supranuclear palsy; FTD, frontotemporal dementia; SD, semantic dementia; PA, progressive nonfluent aphasia; AD, Alzheimer's dementia; DLB, Lewy body dementia; VD, vascular dementia; PP, patients with primary psychiatric diagnoses who are includes as clinical control group; RLS, restless legs syndrome; HD, Huntington; HC, healthy controls.

Parkin-son's DiseasePéron et al.1227 PDAffective and cognitive ToMReading the Mind in the Eyes TestPD patients with advanced progression of disease suffered from cognitive ToM deficits, whereas affective ToM abilities seem sustainedAuthors suggest that ToM dysfunctions occur only in advanced stages of PD, while early PD have sustained ToM abilities. The absence of dopaminergic treatment's effect on ToM performance in early PD patients was interpreted as independence of ToM abilities from the dopaminergic system.
  17 early PDFaux pas storiesEarly PD patients were assessed before and after dopaminergic medication, showing no difference in their ToM performance
  26 HC  
 Bodden et al. (unpublished data)21 PDAffective and cognitive ToMYoni taskDeficits of patients with PD in affective and cognitive ToM abilitiesAuthors suggest that affective ToM dysfunctions might be associated with dysfunctions of the fronto-striatal limbic loop, while cognitive ToM impairments might be associated with alterations of the fronto-striatal-dorsolateral loop
 21 HCReading the Mind in the Eyes TestAffective ToM deficits were associated with reduced health- related quality of life
 Yoshimura et al.3611 jPDEmotion recognitionVideotapesPatients with jPD show no impairment in facial expressions decodingIt remains debatable if ToM abilities were investigated.
 16 HCPhotographsRelatively spared social cognition in jPD is attributed to a more severe dopaminergic depletion in the nigrostriatal system than in the mesocorticolimbic system, the latter of which might be associated with ToM
 Mimura et al.3718 PDMindreadingReading the Mind in the Eyes TestDeficits of patients with PD in the ToM taskAuthors suggest that deficits of mindreading and decision making might be associated with dysfunctions in the orbitofrontal and anterior cingulate circuitry, while executive dysfunctions are associated with alterations of the dorsolateral prefrontal circuit
 20 HCCorrelation of ToM task and decision making (Iowa gambling task) but not between ToM task and tasks for executive functions
 Berg et al.3826 PD“Making inferences” about characters in stories“BeSS” test battery to assess language impairments including subtest “making inferences”Patients with PD had deficits to infer intentions of story characters 
 26 HCA subgroup of mildly demented PD showed the most deficits in “making inferences,” “recreating sentences,” “understanding metaphors,” and “understanding ambiguities”
 Glozman et al.3960 PDEmotional MemoryEmotional “expressive” wordsIn contrast to controls, PD did not show any difference in memory of expressive and neutral wordsIt remains unclear whether ToM abilities were investigated.
 10 HCIdentification of emotional statesPictures with emotional statesPD scored lower in recognition of emotional states
 McNamara and Durso4020 PDPragmatic communication skillsPragmatic protocol including verbal, paralinguistic, and nonverbal behaviorsPatients with PD showed deficits in pragmatic communication abilities which can be conceptualized as ToM functions, e.g., conversational turn-taking or speech acts 
 10 HCfrontal lobe measures correlate with pragmatic dysfunction in PD
 Mengelberg and Siegert4113 PDToMCard-sequencing cartoon task with four conditions: false belief, social situations, mechanical, complex reasoningDeficits of patients with PD in the false belief condition of the card sequencing task and in ToM short stories 
 11 HCShort story taskMore pronounced ToM deficits in the PD group with higher depression scores
 Saltzman et al.1411 PDToMFalse belief storiesDeficits of patients with PD in false belief stories and Spy model taskFirst study that described ToM dysfunction in PD
 8 older HCDroodles cartoon taskSignificant correlations of one ToM task (knower/ guesser task) with executive functions
 9 young HCSpy model task 
  Hide and seek task 
  Knower guesser task 
 Benke et al.4248 PDHumorous cartoonsDetection of humorous cartoonsPatients with PD with cognitive impairment showed dysfunctions in the discovery of pictorial humor and the recognition and production of emotional prosody 
 18 HCProsodyDiscrimination and production of emotional prosodyPatients with PD without cognitive impairment showed deficits in the production of emotional prosody
Atypical Parkinson SyndromesRankin et al.4314 CBDSocial behavior, specified as interpersonal coldness and lack of empathyInterpersonal Measure of Psychopathy (IMP) behavior checklistPatients with FTD and SD showed specific spontaneous social behaviors, which differ from HC, PP, and patients with other dementias.Interpretation of this study is limited as the interesting patient populations with atypical PD were pooled in the “other dementias group”.
 21 PSP
 40 FTD
 21 SD
 13 PA
 37 DLB
 45 AD
 16 VD
 29 VD/AD
 35 PP
 17 HC
 O'Keeffe et al.4411 CBDEmpathy Awareness/ metacognitionMeasure of Empathic TendencyReduced empathy abilities in patients with CBD and PSP 
 10 PSPAwareness ScalesCorrelated dysfunctions in empathy and deficits of metacognitive awareness were reported for the entire sample
 14 FTD  
 20 HC  
 Rankin et al.1315 CBDEmpathy and perspective takingInterpersonal Reactivity Index - subscales “empathy” and “perspective taking”Correlations were found for the entire sample between overall empathy score and gray matter volume in the right temporal pole, the right fusiform gyrus and the right medial inferior frontal region 
 6 PSPPatients with CBD and PSP scored lower than HC in the empathy and perspective taking scales
 30 FTD 
 26 SD 
 8 PA 
 38 AD 
 20 HC 
Huntington's DiseaseHo et al.4575 HDInsight/ metacognitionDysexecutive Questionnaire (DEX) from the Behavioral Assessment of the Dysexecutive Syndrome batteryPatients were able to evaluate others performance quite accurately but failed when evaluating themselves.The authors suggest that patients' anosognosia might contribute to their inappropriate social behavior, i.e., disregarding social norms.
 67 HC (carers)
 Snowden et al.913 HDToMToM Stories ToM deficits were more pronounced in the FTD than HD groupAuthors suggest that social breakdown in FTD and HD may have a different underlying basis and that the frontal neocortex and the striatum contribute differently to social behavior.
 13 FTDHumorous CartoonsQualitative differences in FTD and patients with HD: while patients with FTD failed to make mental state inferences, patients with HD drew inferences but “misconstructed” them with an “eccentric bias”
 18 HCEye gaze task 


Parkinson's disease (PD) is a progressive neurodegenerative disease that is characterized by the cardinal motor symptoms of bradykinesia, resting tremor, rigidity, and impaired postural reflexes. Dopaminergic cell loss within the substantia nigra pars compacta in combination with the occurrence of Lewy bodies is the neuropathological hallmark in PD. Neuropsychological deficits frequently occur in PD even in the early stages of the disease.46 Domains that are typically affected include executive functions, attention, memory, and visuospatial skills.47 Furthermore, different aspects of emotional processing that may have an impact on ToM can be disturbed, such as facial expression decoding48, 49 or prosody processing.50–52

Saltzman et al.14 first described ToM dysfunctions in a sample of patients with PD. Compared to healthy control subjects, their patients demonstrated worse performance in two out of four ToM tasks, including false belief stories and an “online” (“spy model”) ToM task. Replications of this finding that ToM is impaired in PD were reported in studies using non-verbal card-sequencing tasks, false belief and first and second order ToM short stories,41 faux pas recognition stories,12 the “Reading the Mind in the Eyes Test” (RMET)53, 37 and visual material.39 Furthermore, two Japanese review articles on social cognitive deficits in PD are available.54, 55 Remarkably, intact social cognition in a patient sample with juvenile Parkinsonism was reported by Yoshimura et al.36 This might suggest that ToM dysfunctions are related to specific pathologies that differ between PD subtypes. Patients described by Yoshimura had more severe dopamine depletion in the nigrostriatal than the mesocorticolimbic system. This would be in line with the assumption of differential contributions of these systems to ToM performance, highlighting the impact of the mesocorticolimbic system, a structure closely related to the VMPFC.

A systematic investigation of different ToM subcomponents was recently begun.16 Specifically, an investigation by Péron et al.12 involved the study of affective and cognitive ToM abilities in advanced patients with PD and compared their ToM abilities to those of early patients with PD (assessing those patients with and without dopaminergic medication) and to healthy controls, and suggested that only the patients with advanced PD show deficits in cognitive ToM while newly diagnosed patients performed in ToM tasks comparable to controls, with or without dopaminergic medication. Unpublished data from our group suggest that patients with PD who were on average at more advanced Hoehn and Yahr stages than Péron's patients, show impairments in both affective and cognitive ToM abilities.56 We assessed 21 patients with the computerized “Yoni task”7 and the RMET.53 These heterogeneous results can be integrated by taking into account an earlier deterioration of the dorsolateral prefrontal-striatal-circuitry as compared to the frontostriatal-limbic circuitry. An additional dysfunction of the second circuitry that occurs later in the course of the disease can be assumed as physiological basis for the additional deficits in affective ToM.36, 57 Combining the results of Péron et al.12 and our study, these data suggest that ToM dysfunctions occur and are aggravated during the progression of PD. Thus patients with early PD perform comparably to healthy controls while patients with more advanced PD first develop cognitive ToM deficits, which are accompanied by affective ToM deficits in more advanced stages of disease.

Several studies in patients with PD have reported dysfunctions in neuropsychological paradigms that could be assigned to ToM, though this term was not used by the authors. For example, Benke et al.42 described impairments in patients with PD with moderate cognitive deficits in the production and recognition of emotional prosody and the discovery of pictorial humor. These concepts can be seen as contributing to ToM (prosody) or even presuppose ToM (detection of humor). High level language deficits, such as difficulties in “making inferences” about short story characters' intentions, were reported by Berg et al.38 Finally, McNamara and Durso40 delineated reductions in pragmatic language abilities in PD. Specifically, their patients exhibited impairments in verbal, paralinguistic, and nonverbal aspects of social as well as conversational skills—functions that may be associated with ToM.

Contradictory results were found regarding the association of ToM and executive functions in PD. While Saltzman et al.14 reported a significant correlation between ToM and executive functions, Mimura et al.37 were unable to confirm this finding, but reported an association between ToM and decision making. It should be emphasized here that ToM, executive functions, and decision making all rely on the integrity of the prefrontal cortex and associated frontostriatal circuitry. The recruited cortical regions and corresponding circuitry only partially differ.26, 57 Although all relevant frontostriatal circuitry are impaired in PD, the extent of dysfunction in each appears to vary across patients and may explain their different cognitive profiles,37, 36 amongst other factors e.g. stage of disease.

Furthermore, it should be noted that ToM dysfunctions seem to be accentuated by depressive symptoms41 as common in PD. In addition, in our patient sample deficits in affective ToM abilities were associated with a considerably reduced health-related quality of life, a finding that highlighted the clinical relevance of the ToM concept.58


Only few studies thus far have investigated ToM or related abilities in patients with atypical Parkinson syndromes, such as multiple system atrophy (MSA), progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), or dementia with Lewy bodies (DLB).

MSA, a rapid progressive neurodegenerative disease with autonomic dysfunction as a prominent symptom is also accompanied by cognitive impairments, i.e., symptoms associated with frontostriatal dysfunctions, while fully developed dementia is an exclusion criterion of the diagnosis.59 Herting et al.59 discuss disturbance of affect, i.e., depressive symptoms in MSA patients associated with disease-inherent dysfunctional frontostriatal circuitry. To the best of our knowledge, studies investigating ToM or related concepts in patients with MSA have not yet been published. PSP, the most common atypical parkinsonian disorder, is a neurodegenerative disease characterized by defects in the vertical ocular gaze, bulbar dysfunction, increased frequency of falling, and akinetic-rigid features. In addition, cognitive impairments, in particular executive dysfunctions associated with alterations within the frontostriatal circuitry, occur.60 Progression to the point of dementia is common. CBD, as a disease with a mid-to late age of onset, is characterized by slowly progressing, unilateral Parkinsonism with dystonia or myoclonus, unresponsiveness to levodopa, and limb apraxia. Patients with CBD often demonstrate impairments in visuospatial processing and visuoconstruction61 in combination with acalculia and dysexecutive symptoms.62 Dysfunctions in emotional facial expression decoding have been described.63, 44 Furthermore, alien limb phenomena often occur and are discussed as being related to “agency,”64 a concept that is also proposed to be associated with ToM.65, 66

Both patients with PSP and CBD were included in a study by Rankin et al.13 who investigated the structural basis of mentalizing using the affective and cognitive subscales of the Interpersonal Reactivity Index (IRI) and structural MRI. Referring to the entire study sample, including patients with frontotemporal lobar degeneration and patients with Alzheimer, correlations between empathy and atrophy in the right-sided temporal pole, anterior fusiform gyrus, and medial inferior frontal cortex were found. Both patients with PSP and CBD had lower scores than healthy control subjects in the affective and cognitive IRI subscales. Unfortunately, statistical details for these patient groups were not provided and thus specific profiles of deficits cannot be derived. However, on a descriptive level patients with CBD and PSP scored lower than HC in both subscales. Another study by Rankin et al.43 investigated spontaneous social behavior in several different disorders including PSP, CBD, and patients with dementia with Lewy bodies. Patients' behavior during a clinical assessment was rated according to the Interpersonal Measure of Psychopathy (IMP), a standardized behavior checklist. Unfortunately, patients with atypical parkinsonian disorder were pooled with Alzheimer's disease patients in the “other dementias group” so that no specific conclusions were provided. Notably, a study by Cordato et al.67 using voxel-based morphometry demonstrated significant reductions in average orbitofrontal and medial frontal gray matter volumes, while no alterations within the dorsolateral prefrontal cortex were reported. These findings suggest that PSP may be associated with a more pronounced dysfunction of affective ToM as compared to cognitive ToM. However, no studies systematically investigating affective and cognitive ToM subcomponents in these patient groups were found in our literature search. Reduced empathy abilities in patients with PSP and CBD, operationalized through the “Measure of Empathic tendency” (MET), were also reported by O'Keeffe et al.44 Interestingly, these deficits were associated with anosognosia, a metacognitive ability that has been conceptualized as a specific deficit in processing ToM,68 and has also been associated with dysfunctions of the orbitofrontal cortex.69

In DLB, Lewy bodies and Lewy-body related pathology are found from the brain stem to the cortex. DLB is commonly characterized by fluctuating attention, executive dysfunctions, a decline in visuospatial functioning, recurrent visual hallucinations, and Parkinsonism. Low dopamine transporter uptake in the basal ganglia appears when functional neuroimaging is used.70 PD and DLB, both disorders of α-synuclein metabolism, have different etiologies and thus debate has ensued as to whether these should be considered variants of “Lewy body spectrum disorders.”71 Recently, the time of onset of cognitive symptoms, in relation to the motor symptoms, has been used as a criterion for differentiating among these diagnoses.72 Given that patients with DLB share neuropathological characteristics with patients with PD including structural changes within the amygdala, it may be speculated that these patients also suffer from similar ToM dysfunctions. However, to our knowledge no studies investigating ToM in DLB have been published.


With a prevalence of 5–15%, restless legs syndrome (RLS) is a common neurological disease that is characterized by a sensory component (paraesthesia till pain, i.e., uncomfortable and unpleasant sensations in the legs) and motor symptoms (urge to move the limbs to reduce the pain). Patients profit from dopaminergic restitution therapy. In addition, it should be noted that the pathophysiology of RLS and the potential impact of dopamine are not yet fully understood.73 Research on neuropsychological deficits in patients with RLS is rare. The impact of sleep pathology on cognition, especially executive functions, has been discussed.74, 75 No studies investigating ToM or associated concepts have been published so far. Preliminary results from our own group suggest the existence of ToM dysfunctions in highly affected patients with RLS. Further investigations are needed to specify the findings.


Huntington's disease (HD) is an inherited, autosomal-dominant neurodegenerative disease that includes basal ganglia dysfunction and is characterized by a progressive deterioration of the nigrostriatum. This pathology accounts for a range of symptoms from movement disturbances and cognitive decline to dementia. Deficits in processing facial emotional expressions, especially the discovery of disgust, have been described.76–78 Others report a wider range of impairment in recognition of emotions, indicating that patients' difficulties are not limited to recognition of disgust but concern the recognition of negative emotions more generally.79, 80 Neuropsychological symptoms, in particular executive dysfunctions, have been ascribed to disruptions of the frontostriatal circuitry.81

Although a lack of empathy and a breakdown in interpersonal relationships in daily life are often prominent clinical features in HD, these symptoms have not yet been studied systematically. Only one study demonstrated that patients with HD are dysfunctional in both verbal and nonverbal ToM tasks.9 Remarkably, these impairments were qualitatively different from those of patients with FTD. While the latter group demonstrated a general lack of ToM ability, patients with HD did draw inferences about mental states, but they “misconstructed” in doing so. Their inferences were biased in an “eccentric” way. Thus, it was suggested that the patients' dysfunctions arise more at an emotional than a cognitive level, indicating that a more differentiated investigation of affective and cognitive ToM in this patient group may be especially fruitful. Another interesting hint comes from Ho et al.45 who assessed the extent of anosognosia in a sample of patients suffering from HD. These authors report patients' difficulties with insight, and suggest that this issue might be partially responsible for patients' inappropriate social behavior, that is, their disregard of social norms.


ToM, an essential ability for successful social interaction, is a rather new neuropsychological concept that has inspired neuroscientific research within the last three decades. However, ToM dysfunctions in basal ganglia disorders have thus far been almost overlooked as a research topic.

The rationale for the hypothesis that ToM dysfunctions may be a frequent consequence of basal ganglia dysfunction is complex. It includes (1) a “neuropsychological argument” referring to the fact that domains associated with ToM are frequently disturbed in basal ganglia disorders; (2) a “neuroanatomical argument pointing to the relevance of the basal ganglia for ToM-related functions of the frontal lobe and limbic structures via frontostriatal circuitry; and (3) a “neurophysiological argument” that accounts for the possible relevance of the dopaminergic system, frequently disturbed in basal ganglia disorders, to ToM.

Neuropsychological Argument

Although the domain-specificity of ToM is currently under discussion,82 the concept has been associated with other cognitive domains.83 Remarkably, the basal ganglia are known to be relevant for these domains.26, 57 The capacities that are most frequently discussed as being linked to ToM are executive functions.14, 84 Although ToM and executive functions can be deteriorated independently from each other, and functional independence has thus been recommended,85 executive functions have been conceptualized as a “co-opted” system, parallel to a “core” ToM system, which is necessary to succeed at least in particular variants of ToM tasks.83 The basal ganglia, via the dorsolateral-prefrontal-striatal-circuitry, are known to be involved in executive functions. Furthermore, patients with PD57 as well as patients with diseases such as PSP,60 HD81 or RLS75 often develop executive dysfunctions. Thus, although the relationship or potential interaction between both concepts is not yet fully understood,14, 83, 84 it can be speculated that executive dysfunctions in basal ganglia disorders might impede ToM performance. Language dysfunctions such as a non-motor sign in basal ganglia disorders are currently under discussion86 and include, for example, deficits in the understanding of emotional prosody.51 The relationship between ToM and language, two concepts which have frequently been shown to be associated, especially in the context of developmental studies87 but which may also deteriorate separately,88 is not yet fully understood. Decoding prosody can be conceptualized as one facet of ToM processes. In some situations, understanding of verbal humor and sarcasm presuppose the capacity to gather different mental states. Similar to executive functions, language might serve as a “co-opted” system of ToM,83 such that language difficulties may account for ToM deficits. Similar to prosody, facial expression decoding can be seen as a crucial contribution to ToM. Decoding of emotional facial expression is disturbed in various basal ganglia disorders, such as PD48, 49 or HD.78–80 Thus, this deficit might also entail ToM deficits.

Neuroanatomical Argument

Based on our current understanding, the substantia nigra plays a pivotal role in the different frontostriatal circuitry hypothesized to mediate ToM functions (Fig. 2). Together with the VTA, the substantia nigra is part of the mesolimbic system, which is crucial in the prediction of reward and reward-based decisions.89 Closely related to prefrontal areas via frontostriatal circuitry, the basal ganglia is involved in various complex behaviors. More specifically, it is involved in emotional aspects of behavior via the orbitofrontal-limbic-striatal circuitry as well as in executive aspects of behavior via the dorsolateral-prefrontal-striatal-circuitry57 (Fig. 2). Both circuitries are mostly organized in a parallel fashion, recurring to different prefrontal areas as well as different parts of the substatia nigra. Presupposing that ToM comprises affective as well as cognitive subprocesses, it might be speculated that the substantia nigra pars compacta, as part of the dorsolateral-prefrontal-striatal-circuitry, is more strongly involved in cognitive aspects of ToM, whereas ventral parts (i.e., substantia nigra pars reticulate) might be primarily associated with affective ToM (Fig. 2). This concept would be in line with findings that patients with PD firstly develop cognitive ToM deficits which are accompanied by impaired affective ToM only in more advanced stages of disease.12, 36, 57 Abu-Akel describes ToM deficits that differ qualitatively, depending on the specific region of dysfunction.22 Similarly, disruptions to both circuitry at different prefrontal areas might induce selective impairment of specific ToM sub-processes, an assumption supported by the findings of different lesion studies.11, 23 The conceptualizations presented may be an oversimplification of the very complex interactions and ramifications of the different circuitry underlying ToM functionality. More research is needed to pinpoint the different pathways and their connections to the different basal ganglia disorders. In addition, it should be mentioned that PD, which is has recently been considered a systemic disease, is accompanied by multiple pathological changes such as widespread occurrence of Lewy bodies, which also populate the amygdala.90 This provides an additional argument as to why affective processing and ToM might be impaired in PD. Further studies are needed to confirm this assumption.

Neurophysiological Argument

Recently, the contributions of specific neurotransmitter systems to ToM have been discussed. In his “neurochemical hypothesis of ToM,” Abu-Akel27 proposes that dopaminergic and serotonergic dysfunction may cause impairments in ToM. He argues that (i) neuroanatomic regions that are critical for ToM, i.e., the prefrontal cortex, the temporoparietal junction and the anterior cingulate cortex, are innervated by the dopaminergic and serotonergic system; (ii) cognitive abilities such as executive functions that contribute to ToM, as well as the language that depends on ToM can deteriorate in the case of abnormalities in neurotransmission within one of these two systems; and (iii) the dopamine system is relevant for signaling predictions about the consequences of future events—an ability closely related to ToM. Further evidence of a possible role of the dopaminergic and serotonergic systems in ToM processing comes from the fact that these transmitter systems are affected in patients with autism spectrum disorders and schizophrenia, both of which are accompanied by ToM dysfunctions.6, 7 With the substantia nigra, the basal ganglia encompass crucial dopaminergic cells. Further evidence accentuating the dopaminergic contributions to ToM, comes from the close relationship between ToM and hallucinations. Hallucinations are known as common side effects of dopaminergic medication treatment in PD. Hallucinations are also common in patients with DLB. Patients with CBD often develop alien hand phenomena. Remarkably, an association between hallucinations and ToM deficits has been suggested, supposing that inaccurate ToM contributes to the development of hallucinations.91 ToM deficits are closely related to paranoid symptoms such as hallucinations in schizophrenia,92 and alterations of the mesolimbic dopamine system have been proposed as a possible pathophysiological basis for hallucinations.93 Even though the complex interaction of dopamine, ToM and psychosis is not yet fully understood, one might assume a reciprocal interference. Though it has only been speculated that these are associated concepts, ToM deficits and hallucinations might share a pathophysiological basis. As the basal ganglia are crucial for dopaminergic innervations and ToM deficits are frequent in basal ganglia disorders, there may be further evidence to confirm the “neurochemical hypothesis of ToM”27 that posits a key role for dopamine in social cognition, especially in its interaction with other peptides such as oxytocin,34, 35 a neuropeptid known to be highly relevant for affiliation as well as social cognition.


Although knowledge in this field is far from being exhaustive and is sometimes contradictory, the data available so far suggest that ToM dysfunctions occur in a variety of basal ganglia disorders. Preliminary results point to the possibility of different profiles in subcomponents of this ability, i.e., affective and cognitive ToM in different disorders as well as in individual patients. Here, the involvement of different components of the frontostriatal circuitries is discussed as a possible neural correlate. Furthermore, different etiologies might correspond to qualitatively different disturbances in ToM functions. As such, an “eccentric bias” in making inferences about mental states that has been described in HD may contrast with a lack of mentalizing ability in other basal ganglia disorders.

Future research should provide more specific profiles of ToM dysfunctions in basal ganglia disorders, including different subcomponents and possible qualitative differences in ToM disabilities. Potentially influential factors, including clinical parameters, neuropsychological functions, and psychiatric symptoms may also be further specified. Conversely, the impact of ToM dysfunctions on other cognitive domains, activities of daily living and quality of life remains an interesting topic. Functional neuroimaging studies may elucidate underlying mechanisms of ToM dysfunction in basal ganglia disorders, possibly demonstrating an association with different impairments of frontostriatal circuitries. The study of PD as a “classical basal ganglia disorder” with dysfunctions in dopaminergic transmission provides an opportunity to investigate the “neurochemical hypothesis” of ToM as outlined by Abu-Akel.27 For example, effects of dopaminergic pharmacotherapy or modulation of dopaminergic transmission by deep brain stimulation may be fruitful. Finally, an adaptation of recent psychotherapeutic approaches such as “mentalization based treatment”94 might be a useful approach for the treatment of patients with basal ganglia disorders who suffer from ToM dysfunctions as well.


We thank Vanessa Dempsey, Department of Neurology, Philipps-University, Marburg, Germany for her assistance in literature search and Dr. Özgür Onur, Department of Neurology, University Hospital Cologne; Cognitive Neurology Section (INM-3), Research Centre Jülich, Germany for providing the brain image used in Figure 1. E. Kalbe was funded in part by the EC-FP6-project DiMI, LSHB-CT-2005-512146.

Author Roles:

Maren Bodden: research project: conception, organization, execution, manuscript: writing of the first draft. Richard Dodel: research project: conception, organization, execution; manuscript: review and critique. Elke Kalbe: research project: conception, organization, execution, manuscript: review and critique.

Financial disclosure: Maren Bodden: employed in Philipps-University Marburg, Germany. Richard Dodel: consultancies for Solvay, Lilly, Octapharma, a member of advisory boards for Solvay, Lilly, Octapharma received honoraria from Merz, Lundbeck, Boehringer Ingelheim received research grants from Rentschler, ZLB Behring, Michael J. Fox Foundation, holds patents on immunization in neurodegenerative disorders, employed in Philipps-University Marburg, Germany. Elke Kalbe: a member of advisory board (Novartis, Germany), a honoraria member of EISAI, Germany, employed in Research Centre Jülich, Germany.