Abstract
The aim of the current study was to present a possible mechanism underpinning echopraxia in schizophrenia. It is proposed that echopraxia occurs in schizophrenia when the mirror neuron system provides a representation to the inferior frontal gyrus (IFG) and the motor cortex (and via the IFG, to the anterior cingulate cortex) and that this potential becomes executed movement, when the disorder is associated with decreased inhibition and increased arousal.
The aim of this paper was to present a possible mechanism underpinning echopraxia in schizophrenia. The proposal is based on a suite of neuroscience findings: the mirror neuron system (MNS), inferior frontal gyrus (IFG) components (premotor cortex and Broca's area), the anterior cingulate cortex (ACC), theory of mind (ToM), inhibition and arousal. Before drawing the proposal together, we need to describe these findings under appropriate headings. This necessity may give the paper a somewhat undulating course.
Echopraxia
Echopraxia is the pathological repetition by imitation of the movements of another person, and echolalia is the pathological repetition by imitation of the speech of another person. Along with waxy flexibility, echopraxia and echolalia have been considered examples of command automatism, wherein commands or suggestions are automatically and uncritically fulfilled [1]. Stengel coined the term ‘echo-reactions’, which is useful but not widely used [2].
Echo-reactions are described in catatonic schizophrenia. Kraepelin estimated that up to 30% of people with schizophrenia suffered this form [3]. Bleuler believed that these symptoms arose from a weakening of associations, which made it impossible for patients to inhibit imitation [4]. Catatonia is currently diagnosed less frequently in the developed countries, and more frequently in developing countries [5]. One recent leading textbook does not even mention echo-reactions under the heading of catatonic behaviour [6]. Whether the incidence of catatonic phenomena has declined in developed countries, or whether there is another explanation such as classification issues [7] is uncertain. But echo-reactions, particularly echopraxia, are encountered in the developed countries, in the ‘functional’ psychoses [8], by the clinician with an interest in the topic. Very recently there has been renewed attention [9], [10], and echo-reactions remain important diagnostic features of catatonia [11].
Chapman and McGhie studied echopraxia in a small cohort of schizophrenic people [12]. They found that all study patients were subjectively aware of their tendency to imitate others. The finding that echopraxia was more common when patients were attempting to communicate with others [12], and was most common when communication became difficult, is highly relevant to our theory.
Echo-reactions continue to be described in ‘organic’ disorders, such as frontal lobe lesions [13], and difficult-to-classify disorders such as Gilles de la Tourette (21–46%) [14] and Latah (75%) [15]. Echolalia is common and echopraxia is occasionally reported in autism [16]. It is acknowledged that psychodynamic explanations of echo-reactions have been advanced [17], [18], but these have received little attention in the last half century. Our proposal is organically based and psychodynamic explanations will not be further explored.
Mirror neuron system
In recent years ‘mirror neurons’ (MN) have been identified. These are unique cells that fire both when the individual performs an action, and when the individual observes the same action being performed by another individual [19]. Mirror cells were initially identified in the IFG, which includes both premotor cortex and Broca's area, and the inferior lobule of the parietal lobe [20–24].
The term ‘mirror neuron system’ is applied to a network of neural structures that become activated when an individual observes a second individual performing certain tasks. The IFG and inferior lobule of the parietal lobe are the primary, perhaps the only, components of the MNS. But the posterior superior temporal sulcus (STS), the motor cortex and temporal poles are functionally linked to these primary components [21], [25–28].
In both observed and executed movement of hand and lip, a progression of neural activity across the cortex has been demonstrated using magnetoencepalography [25], [29]. In both observed and executed movement, and in both hemispheres, cortical activation commenced at the occipital cortex and progressed to the STS, the inferior parietal cortex, the IFG and, finally, to the primary motor cortex.
Buccino et al. performed a functional magnetic resonance imaging (fMRI) study of movements of distinct body parts (hand, foot and mouth) [30]. For both observed and executed movements, the movement of different body parts was associated with different patterns of brain activity. It was concluded that in both the premotor cortex and the parietal lobe, activation foci were somatotopically organized.
Molnar-Szakacs et al. (2006), using fMRI, found that activity in the MNS varied according to the motoric complexity of observed actions [31]. This group then showed (using transcranial magnetic stimulation (TMS) to measure cortical excitability) that the MNS of humans responds to culture-specific gestures [32].
Broca's region
The IFG is composed of a number of fields, which are not yet fully described, and different authors use different terminology. For present purposes the posterior IFG can be considered as composed of two main areas: the inferior premotor area and Broca's area on the left, and the equivalent area on the right.
Broca's area on the left is composed of Brodmann's area 44 (BA 44) and BA 45. A similar region with some different and some similar functions exists on the right [33]. Left hemisphere BA44/45 lesions preferably affect verbal sequencing, while right hemisphere lesions affect non-verbal sequencing.
Classically, Broca's area has been regarded as exclusively concerned with speech [34]. But recent evidence indicates additional roles in cognition and in our understanding of the actions of others. With respect to hand movements, activation of BA 44 may precede activation of BA 4 [45].
Spoken language and hand movements are controlled by a common system or by systems interdependent on each other [36], and are governed by the same rules [37]. This overlapping suggests that hand gestures evolved into language [38].
In Australia, when explaining how to perform a skilled action, we make the joke, ‘It's all in the way you hold your mouth’. Whether mastering handwriting or golf swing, at least in the learning stage there is a tendency to perform some sort of, apparently unrelated, mouth movement. Perhaps, when hand movements are difficult, there is recruitment of nearby cells, including those related to the mouth.
When subjects view and listen to speaking faces, activation of Broca's region is stronger during incongruent than during congruent audiovisual stimuli [39]. We speculate that dysfunction of the MNS is involved in echopraxia, and that increased arousal increases the risk of echopraxia. It would be reasonable to assume that incongruent audiovisual stimuli would cause greater arousal than would congruent audiovisual stimuli.
Motor cortex
The classic view is that the motor cortex consists of two areas: (i) BA 4 and the lateral convexity component of BA 6 (primary or M1); and (ii) supplementary (mesial component of BA 6), and that this ‘motor cortex’ is exclusively involved with motor activity. This view is no longer sustainable. The motor cortex is formed by a mosaic of independent motor areas. These have unique sets of connections with parietal, prefrontal and cingulate cortices. Some motor areas are involved in higher-order aspects of motor control related to motivation, memory, temporal planning of motor behaviour, and other cognitive functions [40].
The primary motor cortex is a component of the MNS [21] and is activated when the human (but not the monkey) subject observes complex movements, such as the use of tools [23].
Importantly, the occurrence of an internal representation of action is insufficient to initiate movement. In the healthy individual, movement occurs only when motivation and certain environmental factors are in place [41].
Anterior cingulate
The medial prefrontal cortex (MPFC) is a large region [42], which includes the ACC, paracingulate cortex and other structures. The MPFC has dense connections with other regions of the prefrontal cortex and midline structures including the thalamus. We pay particular attention here to the ACC, which has strong connections with the (i) motor cortex; (ii) premotor cortex; (iii) supplementary motor area; (iv) brainstem and spinal cord; (v) dorsolateral prefrontal cortex (DLPFC); (vi) other frontal cortices; (vii) limbic system via other cortical regions; and (viii) thalamus.
Direct access of the ACC to the skeletal musculature has been demonstrated by electrical stimulation [43]. Reciprocal connections with the DLPFC indicate a role with cognition, allowing Paus to describe the ACC as a point of convergence of cognitive and motor processes [44]. Allman et al. described the ACC as an interface between emotion and cognition [45].
The ACC receives afferents from the midline thalamus and brainstem, suggesting that it is well supplied by arousal information. It is the main target area of the mesocortical pathways, which originate in the ventral tegmental area. It is especially activated under conflict conditions [46–49].
MPFC lesions involving the cingulate cortex reveal (i) deficits in spontaneous initiation of movement and speech (akinetic mutism in bilateral lesions); and (ii) an inability to suppress externally triggered motor responses (alien hand syndrome [50], [51]). Alien hand syndrome is frequently associated with the lesions that also encroach on the supplementary motor area and the corpus callosum, and it is difficult to define the precise role of the ACC. The point made here is that the ACC has a role in movement.
Our proposal includes that the MNS is involved in echopraxia. The ACC is not considered a component of the MNS, but it has close connections with components of the MNS, and has outflow that can produce movement. The ACC is mentioned in our proposal because it could function as a direct or indirect route to echopraxia. Dysfunction of the ACC is reported in schizophrenia [52]; this could mean that representations from the MNS might be passed on without the usual degree of filtering.
Disinhibition
The mechanisms of the MNS ensure that, when an individual observes an action, corresponding cortical activity (potential movement) occurs in that individual. Our proposal includes that imitation (executed movement) could occur under certain conditions, one of these being decreased nervous system inhibition.
Decreased inhibition has been associated with lesions of the MPFC/ACC [50], [51], DLPFC [53], the orbitofrontal cortex (OFC) [54], and IFG/Broca's area [55], [56].
In schizophrenia abnormal structure or function has been described in each of these areas by at least two groups of researchers: ACC [57–59], DLPFC [60–62], OFC [58], [63], and Broca's area [64–66]. Thus, in schizophrenia the conditions may exist for potential movement to become imitation (echopraxia), through the actions of the MNS and a decreased capacity for inhibition.
Theory of mind
The MNS is proposed as part of the neural basis of ToM [67]. ToM refers to the ability to attribute mental states (such as thoughts, beliefs, desires and intentions to people (others and ourselves). It has been conceptualized as central to empathy [68], [69].
People with schizophrenia do significantly less well than healthy controls on ToM tests [70–72]. These findings suggest disorder of the MNS in schizophrenia.
Schizophrenia: clinical and pathophysiological
The textbook clinical symptoms of schizophrenia (among others) include social withdrawal, reduced capacity for empathy and disorder of the form of thought. In clinical practice, while presenting with flattened affect, such patients often report feeling anxious. Older opinions include that catatonia (often associated with echopraxia) may result from ‘excessive cerebral excitatory processes’ [73]. Patients with schizophrenia often find conversation with others to be ‘stressful’ (particularly when detailed or precise accounts are expected). Thus, schizophrenia may be associated with heightened arousal, which can be increased by interaction with others.
Long-established broad principles relevant to our proposal are that schizophrenia is characterized by (i) frontal lobe pathology [74]; (ii) frontal lobe functional deficits [75], [76]; and (iii) gating mechanisms dysfunction [77], [78].
In a previous section ‘Disinhibition’ we have referred to anatomical and functional deficits in specific regions (ACC [57–59], DLPFC [60–62], OFC [58], [63], and Broca's area [64–66]. Physiological dysfunction of the motor cortex in schizophrenia [79] will be mentioned under the following heading.
Possible mechanism of echopraxia in schizophrenia
It is proposed that echopraxia occurs in schizophrenia when the MNS provides a representation to the IFG and motor cortex (and via the IFG, the ACC), and that this potential becomes an executed movement, when the disorder is associated with decreased inhibition and/or increased arousal.
We have argued that schizophrenia is characterized by subjective elevated arousal and that this is increased when the subject is required to interact with others, and we have identified some pathophysiological findings that could form a basis for decreased nervous system inhibition.
The ACC is a point of convergence of cognitive and motor processes [44] with rich motoric outflow. Dysfunction of the ACC has been reported in schizophrenia [52] and this may allow representations from the MNS to be passed forward with less than the usual degree of filtering.
People with schizophrenia do less well than healthy individuals on ToM tests [70], [72]. This finding can be taken as indirect evidence of dysfunction of the MNS in schizophrenia.
Schürmann et al. studied patients with schizophrenia and their unaffected twins using magnetoencepalography [79]. They found that, while observing movements, relative to their unaffected twins, patients with schizophrenia demonstrated reduced motor cortex reaction.
Discussion
This is a speculative paper, based on available reports. To this point the MNS studies have mainly focused on relatively small movements of the hand, foot and face [30], rather than the larger and more complex movements that are sometime observed in echopraxia. But recent studies of larger movements (arm reaching) have shown that these also activate the MNS, and that they are associated with extended areas of activation [80]. Recent work demonstrating that the MNS codes culture-specific gestures [32] is important, because echopraxia takes the form of the gestures of others.
Learning by copying is common practice (e.g. in sports), suggesting that the MNS has a role in complex behaviour. Wright and Jackson have suggested a link between the MNS and the ability to judge the direction of a served tennis ball [81]. Clinically, there is speculation that ‘action observation therapy’ (watching others perform tasks) may be helpful in stroke rehabilitation [82], [83], and Ertelt et al. have reported some success with this technique in a small study [84].
Our proposal for echopraxia relies on the MNS. From this perspective, execution of echopractic movements could be achieved by the IFG communicating with the motor cortex. Alternatively, execution could be achieved via the IFG communication with the ACC, which has motoric output. It is quite possible, however, that these two pathways could support each other.
Our proposal is that MNS representations, rather than remaining potential, become executed movements. This might suggest that the MNS should be overactive in people with schizophrenia, but the evidence is to the contrary; Schürmann et al. found, in people with schizophrenia, that observed movements were associated with decreased reaction [79]. This is consistent with the hypofrontality theory of schizophrenia. This does not, however, necessarily negate our proposal. We are suggesting that MNS-generated representations are handled pathologically, that the MNS is involved in dysfunction, not necessarily that the MNS is ‘overactive’ and thereby constituting an irresistible force.
Catatonia, of which echopraxia is a frequent feature, responds well to electroconvulsive therapy (ECT). The mechanism of ECT remains to be fully elucidated. Some evidence suggests that for depression, slowing of the post-ECT electroencephalogram is associated with longer recovery [85], but it is not possible to build the available evidence into a comprehensive explanatory/predictive statement.
Work is needed to test this proposal. Comparisons of healthy people and people with schizophrenia with and without echopraxia, using functional imaging, during both observation and executed movement, could begin the process. A pragmatic approach to treatment with repetitive (r)TMS may also be indicated (especially in light of the beneficial effects of ECT). Echopraxia is a symptom, and to most patients not a particularly distressing one (compared to other positive symptoms). Thus, the study of the response of echopraxia to rTMS would need to be part of a larger study. The site and frequency of stimulation would need to be considered, but in the end, various combinations of parameters would need to be applied on an experimental basis.