Schizophrenia Research 168 (2015) 174–179

Contents lists available at ScienceDirect

Schizophrenia Research

journal homepage: www.elsevier.com/locate/schres

Mirror deficit in schizophrenia: Evidence from repetition suppression

Nicole Möhring, Christina Shen, Eric Hahn, Thi Minh Tam Ta, Michael Dettling, Andres H. Neuhaus ⁎

Department of Psychiatry, Charité University Medicine Berlin, Germany article info abstract

Article history: Background: Schizophrenia is associated with impaired cognition, especially cognition in social contexts. The mir- Received 13 February 2015 ror neuron system (MNS) serves as an important neuronal basis for social cognitive skills; however, previous in- Received in revised form 7 July 2015 vestigations on the integrity of MNS function in schizophrenia remain approximate. Accepted 19 July 2015 Methods: We employed a repetition suppression paradigm that allows for measuring neuronal responses to ges- Available online 29 July 2015 ture observation and gesture execution. Cross-modal repetition suppression, i.e., adaptation between observe/ execute and execute/observe conditions, was defined as the decisive experimental condition characterizing Keywords: Hand gestures the unique sensori-motor properties of mirror . Event-related potentials (ERPs) were assessed in 15 Event-related potential schizophrenia patients and 15 matched controls. Repetition suppression Results: We isolated an ERP signature of specific adaptation effects to identical hand gestures. Of critical importance, Adaptation this ERP signature indicated intact intra-modal adaptive pattern, i.e., observe/observe and execute/execute, of system comparable magnitude between groups, but deficient cross-modal adaptation, i.e., observe/execute and execute/ Social cognition observe, in schizophrenia patients. Conclusion: Our data provide robust evidence that pure perception and execution of hand gestures are relatively intact in schizophrenia. In contrast, visuo-motor transformation processes mediated by the MNS seem to be spe- cifically disturbed in schizophrenia. These results unambiguously demonstrate MNS deficits in schizophrenia and extend our understanding of the neuronal bases of social dysfunction in this disorder. © 2015 Elsevier B.V. All rights reserved.

1. Introduction 2004). Based on this sensori-motor property, an internal motor repre- sentation of the observed action (e.g., posture/gesture) is generated, Impaired social cognition is a common feature of schizophrenia as- linked with a corresponding affective state, or optionally modulated by sociated with functional outcome (Couture et al., 2006; Smith et al., higher order cognitive processes. Through this automatic mirroring 2014). Deficits of social interaction have been related to mentalizing/ mechanism, social interaction is facilitated by enabling individuals to theory-of-mind abilities and empathy, both of which are substantial understand behavior and intentions of others and thus to imitate or to for social interaction (Corbera et al., 2013; Derntl et al., 2012; Martin share emotions (Iacoboni, 2005; Rizzolatti and Sinigaglia, 2007). Follow- et al., 2014). These abilities, in turn, rely on basic social cognitive skills ing that rationale, it is thought that psychiatric disorders that present including perception of socially relevant cues like facial expressions with impaired social cognition are likely to exhibit a deficient MNS and gestures. In contrast to the large literature on face processing in (Buccino and Amore, 2008; Haker and Rössler, 2009). schizophrenia, only few studies focused on processing of gestures and Direct evidence of neurons with mirror properties is possible body postures. These studies clearly demonstrate that schizophrenia through invasive measurements only (Mukamel et al., 2010). For com- patients are less accurate in interpreting (Bucci et al., 2008; Thoma prehensible reasons, there are no systematic studies applying invasive et al., 2014) and imitating gestures and body postures (Matthews methods for assessing the integrity of the MNS in psychiatric disorders. et al., 2013; Walther et al., 2013a, 2013b). In healthy subjects, mainly functional magnetic resonance imaging To successfully process gestures and body postures, the recipient (fMRI) was used to show topographically overlapping activations dur- needs to understand both movement and meaning, which neuronally ing action observation and tasks (Molenberghs et al., 2012). intersect in the mirror neuron system (MNS). Mirror neurons are active Using this non-invasive approach, a recent study of Thakkar et al. during both observing and executing an action (Rizzolatti and Craighero, (2014) found an altered activation pattern in relevant MNS nodes in schizophrenia. Contrarily, Horan et al. (2014a) reported decreased self-reported empathy in correlation with activity of the inferior frontal ⁎ Corresponding author at: Department of Psychiatry, Charité University Medicine, Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany. gyrus (IFG), but failed to show reduced activity within this classical mir- E-mail address: [email protected] (A.H. Neuhaus). ror neuron area in schizophrenia patients. Besides these inconsistencies,

http://dx.doi.org/10.1016/j.schres.2015.07.035 0920-9964/© 2015 Elsevier B.V. All rights reserved. N. Möhring et al. / Schizophrenia Research 168 (2015) 174–179 175 demonstrating overlapping activity, can only provide indirect and pre- All patients were recruited from the outpatient clinic of the Depart- liminary evidence, because this approach cannot exclude spatially over- ment of Psychiatry, Charité University Medicine Berlin, Campus Benja- lapping, but distinct neuronal populations (Dinstein et al., 2008). Here, min Franklin. They met DSM-IV criteria and had no psychiatric the use of repetition suppression (RS) paradigms constitutes a major disorder other than schizophrenia and nicotine abuse/dependence. Ex- advance towards unambiguously and non-invasively characterizing clusion criteria were current drug abuse and history of severe medical MNS function (Chong et al., 2008; Kilner et al., 2009). or neurological disorder including a history of electroconvulsive therapy. RS describes the reduction of neuronal activity in response to repeated Mean duration of illness was 141.23 ± 72.5 months and mean number presentation of the same stimulus, e.g., in a paired stimulus design (Grill- of episodes was 2.0 ± 1.1. All patients received atypical antipsychotics Spector et al., 2006). Comparing neuronal response amplitudes to repeti- with a mean chlorpromazine equivalent of 442.04 ± 407.9 mg/d. None tions with non-repetitions then allows for deciding whether the involved of the patients suffered from extrapyramidal motor side effects due to neuronal population can be considered sensitive to the repeated stimulus antipsychotic medication within the last 6 months. Clinical symptom se- feature. In the mirror neuron context, intra-modal RS effects indicate verity was assessed with the Positive And Negative Syndrome Scale sensory (observation/observation) or motor (execution/execution) prop- (PANSS) for schizophrenia: positive symptoms 15.57 (±4.2); negative erties, while cross-modal RS (observation/execution; execution/observa- symptoms 20.43 (±4.9); and general psychopathology 31.29 (±5.5). tion) indicates sensori-motor properties that unambiguously characterize Control subjects were recruited via newspaper advertisements. They mirror neurons (Möhring et al., 2014a), although, as described above, this were screened for mental and physical health by a board certified psy- approach cannot prove the existence of mirror neurons in a strict sense. chiatrist and were excluded when meeting the criteria of psychiatric So far, the RS paradigm has not been adopted for MNS studies in disorders according to DSM-IV as determined by semi-structured clini- schizophrenia, which is mainly investigated via transcranial magnetic cal interviews. Moreover, family history of psychiatric illness, medical or stimulation to test excitability (Enticott et al., 2008; neurological disorders, and current intake of psychotropic drugs led to Mehta et al., 2014b) and via mu rhythm suppression in electroenceph- the exclusion of the study. alography (EEG; Horan et al., 2014b; Mitra et al., 2014; Singh et al., All participants completed a multiple choice vocabulary test (MWT; 2011). These studies partially contradict each other, e.g., in terms of de- Lehrl et al., 1995) and the German performance testing system (LPS; creased (Mitra et al., 2014; Singh et al., 2011) versus increased mu sup- Horn, 1983) to estimate verbal and non-verbal intelligence, respective- pression in schizophrenia (Horan et al., 2014b), and, even more ly. The Interpersonal Reactivity Inventory (IRI; Davis, 1983) was applied important, they investigated surrogates of motor cortex function that for assessing empathic ability. All participants had normal or corrected- is – as an effector organ – associated with, but not part of the MNS to-normal vision and where right-handed according to the Edinburgh (Rizzolatti and Craighero, 2004). Handedness Inventory (EHI; Oldfield, 1971). The study protocol was ap- Here, we applied a cross-modal adaptation protocol to investigate proved by the ethics committee of the Charité University Medicine Ber- MNS activity in schizophrenia for the first time. Adopting the study pro- lin, and the study was conducted in accordance with the Declaration of tocol of Dinstein et al. (2007), our participants observed or executed Helsinki and its amendments. All subjects gave written informed con- gestures of the rock–paper–scissors game. According to our previous sent before participating and received monetary reimbursement for normative study (Möhring et al., 2014b), we expected to find evidence their efforts. of reduced mirror neuron activity in mid-latency event-related poten- tials (ERPs) as expressed by reduced or absent cross-modal RS effects. 2.2. Experimental design Specifically, we focused our analysis on the N190 and the P2 compo- nents that have been shown to be sensitive to repetitions in our norma- The experiment was carried out in a windowless, dimly lit, electro- tive study using the same paradigm (Möhring et al., 2014b). magnetically shielded, and sound attenuated room. Participants were asked to take a seat in a comfortable chair in front of the screen and to direct their gaze towards the monitor. Standardized instructions were 2. Material and methods given verbally by the experimenter and visually on the screen. In a train- ing phase comprising 10 trials, participants practiced the task to ensure 2.1. Subjects that they followed instructions correctly. During the whole experimen- tal session, subjects were visually monitored by the experimenter Fifteen medicated patients with diagnosis of schizophrenia (11 men, through a window from a neighboring room to control for accurate ac- 4women)andfifteen healthy controls participated in the study. Groups tion execution. were matched for sex and age (±2 years). Demographic and clinical Participants were instructed to passively observe static images of a data of all participants are summarized in Table 1. hand forming gestures of the popular rock–paper–scissors game (obser- vation condition) and to actively execute respective hand gestures as soon as imperative stimuli depicting rock, paper, or scissors were Table 1 displayed (execution condition). Stimuli were presented on a 24 in. Demographic and clinical data. monitor with a viewing distance of approximately 60 cm and a visual Schizophrenia Controls p angle of approximately 15 × 10° for the outer stimulus contour using (N = 15) (N = 15) Presentation (Neurobehavioral Systems, Albany, CA). On a light gray Age (years) 35.60 (7.7) 35.40 (7.9) .964 background, three naturalistic photographs of a right male hand forming Age range (years) 27–57 25–58 – rock, paper, or scissors symbols were displayed in the observation con- Education (years) 14.67 (4.6) 17.75 (3.8) .054 dition and three realistic pictures of a rock, paper, or scissors served as IQ Verbal IQ 102.80 (15.1) 102.83 (7.5) .994 imperative stimuli in the execution condition. Stimulus duration was Non-verbal IQ 109.07 (11.4) 112.75 (6.7) .291 2000 ms. Stimulus presentation was organized in pairs (S1 = adapter Laterality index 90.67 (14.4) 90.67 (14.4) 1.000 stimulus; S2 = test stimulus) with an inter-stimulus interval (ISI) of Interpersonal reactivity index 500 ms that was identified as optimal for eliciting maximal RS effects Perspective taking 17.53 (3.8) 19.46 (3.0) .136 Fantasizing 16.33 (4.7) 14.77 (3.1) .295 (Harris and Nakayama, 2007; Kuehl et al., 2013). Stimulus pairs were Empathic concern 20.33 (2.3) 19.85 (3.8) .706 evenly distributed across intra-modal, i.e., purely sensory (observe/ob- Personal distress 14.40 (5.4) 11.85 (5.7) .216 serve) or motor (execute/execute) repetitions, and cross-modal trials, All values are mean values with standard deviation in parenthesis. Between-group differ- i.e., repetitions across modalities (observe/execute or execute/observe). ences were assessed by t-tests for independent samples. IQ, intelligence quotient. Stimulus pairs showed either identical hand figures/objects (categorized 176 N. Möhring et al. / Schizophrenia Research 168 (2015) 174–179 as repetition trials) or different hand figures/objects (categorized as study, i.e., N190 and P2 (Möhring et al., 2014b). For each ERP compo- non-repetition trials). Thus, combining all experimental conditions nent, the latency window was set around the mean corresponding (repetition/non-repetition × rock/paper/scissors × intra-modal/cross- peak of the grand average, separately for the observation and execution modal × observation/execution), our paradigm consisted of 24 different condition. To define regions of interest for further analyses, current den- types of trials. Inter-trial intervals pseudo-randomly varied between sity scalp maps were used to select those electrodes that were located in 3000 ms and 4000 ms. Fig. 1 gives an overview of the task. the center of gravity of each ERP component. The N190 component was In total, 72 stimuli were presented per block, consisting of 24 repeti- scored at electrodes P7, P8, TP7, and TP8 as the most negative peak with- tion trials and 12 non-repetition trials. In total, five blocks were presented, in 60 ms around the mean corresponding peak of the grand average, i.e., each lasting 5 min. The duration of the experiment was thus approxi- from 156 ms to 216 ms (observation condition) and from 150 ms to mately 25 min excluding the training phase and short breaks between 210 ms (execution condition). The P2 component displayed a broader blocks. positivity and was thus quantified at electrodes PO3, PO4, PO5, PO6, PO7, PO8, P3, P4, P5, P6, P7, and P8 within 80 ms around the mean cor- 2.3. EEG data acquisition and ERP analysis responding peak of the grand average, i.e., from 216 ms to 296 ms while observing and from 224 ms to 304 ms while executing. EEG was recorded with a 64-channel DC amplifier (Advanced Neuro Technology, Enschede, The Netherlands) with a sampling rate of 512 Hz using an elastic cap equipped with 64 average referenced Ag–AgCl elec- 2.4. Statistical analysis trodes according to the extended International 10/10 system and a ground electrode positioned on the forehead. Additionally, bipolar elec- SPSS for Windows version 21.0 (IBM, Armonk, NY) was used for sta- trodes were placed on the outer canthus of the left eye to monitor eye tistical analysis. In a first step, electrodes of interest within each hemi- movements as well as on the Musculus interosseus between the 2nd sphere were averaged for the N190 and P2 ERP components to avoid and 3rd fingers of the right hand to control for muscle activity during ac- circular analyses as outlined by Kriegeskorte et al. (2009). In a second tion execution. Electrode impedances were kept below 5 kΩ. step, repeated measures analyses of variance (ANOVAs) were per- Brain Vision Analyzer 2.03 (Brain Products, Munich, Germany) was formed separately for both ERP components. Initial ANOVAs included used for offline analysis. Raw data were notch-filtered at 50 Hz and ‘repetition’ (adapter vs. repeated test stimulus vs. non-repeated test Butterworth-filtered at 0.1 Hz high-pass and 20 Hz low-pass with 24 stimulus), ‘hand figure’ (rock vs. paper vs. scissors), ‘modality’ (intra- dB/octave. Ocular artifacts were corrected with independent component modal vs. cross-modal), ‘action’ (observation vs. execution) and ‘hemi- analysis (Jung et al., 2000). After down-sampling to 500 Hz, data was seg- sphere’ (left vs. right) as within-subject factors and ‘group’ (schizophre- mented to a length of 1100 ms starting 100 ms before and ending nia vs. control) as between-subject factor, respectively. In a next step, 1000 ms after stimulus onset. Segmentation was performed for each ex- the factor ‘figure’ was removed from analysis after it was confirmed perimental condition (S1/S2; repetition/non-repetition; hand figure/ob- that it neither significantly contributed to any main effect nor any inter- ject; modality; and action), resulting in 48 different segments for each action. Thus, final ANOVAs were conducted with a 3 (repetition) × 2 EEG file. Hand action conditions were entered into analysis only if EMG (modality) × 2 (action) × 2 (hemisphere) × 2 (group) design. For all amplitude was N50 μV within 100 to 1000 ms following an imperative ob- tests, Mauchly's test ascertained that the sphericity assumption was ject stimulus. In the next steps, data were baseline corrected and seg- not violated. Partial eta squared (η2) served as an estimator of effect ments contaminated by artifacts (≥80 μV at any electrode except EMG) size, i.e., the proportion of data variance accounted for by the statistical were removed. Finally, averages were constructed for each experimental model. For all significant main effects and interactions, post hoc t-tests condition. were performed with Bonferroni correction by multiplying the specific In exploratory butterfly plots (data not shown), adapter stimuli were p value with the number of comparisons in that specific test. Correlation averaged separately for observation and execution conditions in con- analysis was done using Pearson correlation. The alpha level was set at trols. Only those ERP components were selected for subsequent analyses p b .05 for all tests. Values are given as mean values with standard devi- that were associated with cortical gesture processing in our previous ation in parenthesis, if not explicitly stated otherwise.

Intra-modal trials Cross-modal trials

Observe/ObserveExecute/Execute Observe/Execute Execute/Observe

Repetition

S1 S2 S1 S2 S1 S2 S1 S2 ISI ITI

Non- Repetition

S1 S2 S1 S2 S1 S2 S1 S2 ISI ITI

Fig. 1. Experimental design and stimuli. Trials always consisted of adapter (S1) and test stimuli (S2) that were presented for 2000 ms each with an inter-stimulus interval (ISI) of 500 ms. Each trial was followed by an inter-trial interval (ITI) that pseudo-randomly varied between 3000 and 4000 ms. Static images of a hand forming rock, paper, or scissors gestures served as passive stimuli (observe condition). Images of rock, paper, or scissors objects served as imperative stimuli where participants had to execute the corresponding hand gesture (execute condition). Stimulus pairs were categorized as intra-modal trials (observe/observe or execute/execute) or cross-modal trials (observe/execute or execute/observe). Experimental trials were further categorized as repetitions (same hand gesture or object) or non-repetitions (different hand gestures or objects). N. Möhring et al. / Schizophrenia Research 168 (2015) 174–179 177

3. Results p = .008; schizophrenia: p b .001). In contrast to the homogeneous re- sults of intra-modal trial analysis, we observed a specific dissociation of 3.1. N190 neuronal responses in cross-modal trials between groups. In controls, P2 amplitudes in response to repeated test stimuli (3.81 ± 1.3 μV) Repeated measures ANOVA of the N190 component revealed signif- were significantly reduced compared to adapter stimuli (4.57 ± 2 icant main effects of ‘repetition’ (F2,56 = 4.837; p = .012; η = .147) 1.8 μV; T14 = −3.222; p = .006). In schizophrenia patients, however, 2 and ‘action’ (F1,28 = 30.915; p b .001; η =.525). no significant differences between adapters (4.97 ± 2.5 μV), repetitions The factor ‘repetition’ was driven by a specific adaptation effect, (5.02 ± 2.8 μV), and non-repetitions (5.21 ± 2.8 μV) were found, thus where adapter stimuli (−3.25 ± 1.4 μV) elicited higher N190 ampli- indicating a failure of neuronal adaptation processes in cross-modal tri- tudes than repeated test stimuli (−2.72 ± 1.4 μV; T29 = −2.900; als in schizophrenia. An independent t-test including amplitude differ- p = .007). Similarly, non-repetitions (−3.13 ± 1.3 μV) evoked higher ences of adapter stimuli minus repeated test stimuli in cross-modal

N190 amplitudes than repetitions (T29 = −2.924; p = .007). Our trials confirmed a significant difference of cross-modal adaptation mag- data hence demonstrate an adaptation effect due to identical stimulus nitude between groups (controls: 0.76 ± 0.9 μV; schizophrenia: repetition. Subsequent analysis of the main effect ‘action’ showed that −.05 ± 1.2 μV; T28 = 2.111; p = .044). No correlation was found be- passively observing static images of hand gestures (−3.59 ± 1.6 μV) tween P2 measures and IRI or PANSS scores. lead to higher N190 amplitudes than actively executing hand gestures in response to corresponding object stimuli (−2.48 ± 1.1 μV; 4. Discussion

T29 = −5.637; p b .001). This is consistent with greater sensitivity of the N190 to body parts than to inanimate objects. The present study aimed at investigating neurophysiological effects of RS in response to hand gesture observation and execution in schizo- 3.2. P2 phrenia. Both schizophrenia patients and control participants exhibited specific adaptation effects in intra-modal trials. This robust RS during ob- Omnibus ANOVA of the P2 component revealed a significant main serving hand gesture stimuli is well in line with our previous findings 2 effect of ‘repetition’ (F2,56 = 12.253; p b .001; η = .304) and significant (Möhring et al., 2014b). Specific isolation of the neuronal signature and 2 interactions of ‘repetition ∗ group’ (F2,56 = 4.204; p = .020; η = .131) temporal dynamics of MNS activity was based on cross-modal adaptation 2 as well as ‘repetition ∗ modality’ (F2,56 =9.232;pb .001; η = .248). effects of parietal gesture processing ERPs that should adaptively respond The main effect of ‘repetition’ was based on significantly suppressed to cross-modal repetitions, i.e., when an action is first observed and then P2 responses to repeated test stimuli (4.09 ± 2.1 μV) compared with executed and vice versa, to allow for inferring MNS activity. In controls, adapters (4.83 ± 2.1 μV; T29 = −3.986; p b .001) and with non- this cross-modal RS effect was demonstrated, whereas this effect was ab- repeated test stimuli (4.74 ± 2.4 μV; T29 = −5.148; p b .001). Paired sent in schizophrenia patients, consistent with a deficient MNS. t-tests demonstrated that the ‘repetition ∗ group’ interaction relied on From a functional perspective, our data is in good agreement with a strong RS effect of the P2 in the control group, with significantly the majority of MNS studies in schizophrenia in general (Mehta et al., lower amplitudes in response to repeated test stimuli (3.55 ± 1.4 μV) 2014a) and with findings of previous electrophysiological studies that compared with both adapters (4.62 ± 1.7 μV; T14 = −4.742; reported a motor cortex deficit in terms of excitability and mu wave p b .001) and non-repetitions (4.08 ± 1.8 μV; T14 = −3.158; p = suppression in schizophrenia in particular (Enticott et al., 2008; Mehta .007). This effect corresponds to a specific adaptation due to identical et al., 2014b; Mitra et al., 2014; Singh et al., 2011). Moreover, instead stimulus repetition that was absent in schizophrenia (see Fig. 2). of drawing inferences from surrogates of motor cortex function – such Most important within the scope of our study, we obtained a signif- as mu rhythm – about the integrity of the MNS, we provide unambigu- icant interaction of ‘repetition ∗ modality’. Paired t-tests indicated intact ous evidence of a deficient MNS in schizophrenia by directly assessing specific adaptation effects in intra-modal trials in both groups, as evi- sensori-motor properties that characterize mirror neurons. Our data im- denced by comparisons of repetitions with both adapters (controls: plicate that disturbed social interaction in schizophrenia is independent p b .001; schizophrenia: p = .024) and non-repetitions (controls: of perception, but attributable to deficient visuo-motor transformation

[µV] Controls[µV] Schizophrenia A 6 P2 B 6 P2

5 5

4 4

3 Repetition 3 Repetition suppression suppression

2 2

1 1

100 200 300 400 500 600 700 800 900 [ms] 100 200 300 400 500 600 700 800 900 [ms] -1 -1

-2 Adapter (S1) -2 Adapter (S1) Repetition (S2) Repetition (S2) -3 Non-Repetition (S2) -3 Non-Repetition (S2)

Fig. 2. Grand averages of the component at electrode position PO3, PO4, PO5, PO6, PO7, PO8, P3, P4, P5, P6, P7, and P8 stratified for the significant main effect of ‘repetition’,i.e.,adapt- er stimuli (black line), repeated test stimuli (red line) and non-repeated test stimuli (blue line). Time frame of the P2 ERP component is highlighted in gray. (A) P2 amplitude modulation in controls. Repetition stimuli lead to significantly reduced amplitudes in the P2 time frame. P2 amplitudes in response to repetitions compared with responses to non-repetitions differ sig- nificantly, which indicates a specific adaptation effect. (B) P2 amplitude modulation in schizophrenia patients. Adapters and repetitions do not show significant amplitude differences in the P2 time frame. Significant P2 amplitude differences between repetitions and non-repetitions indicate an irregular repetition enhancement. 178 N. Möhring et al. / Schizophrenia Research 168 (2015) 174–179 processes in the parietal MNS node (Dinstein et al., 2007; Iacoboni et al., Conflict of interest fl fi 1999; Möhring et al., 2014b). Functionally, our results can thus be nicely There is no con ict of interest, nancial or otherwise, related to this work for any of the authors. linked to disturbances of the frontoparietal network and of IPL function in schizophrenia that has been linked to impaired sensory integration Acknowledgments and body image in schizophrenia (Torrey, 2007; Tu et al., 2013). The authors wish to thank Emily Brandt for assistance in the early phase of this study Temporal dynamics indicates relatively early and therefore auto- and all participants for their efforts. matic deficits in schizophrenia, which is well in line with the automatic nature of visuo-motor transformation processes mediated by the MNS References (Heyes, 2011). Especially strategic games like the rock-paper-scissors game applied in this study are thought to elicit automatic imitation Belot, M., Crawford, V.P., Heyes, C., 2013. Players of Matching Pennies automatically imi- tate opponents' gestures against strong incentives. Proc. Natl. Acad. Sci. U. S. A. 110 (Belot et al., 2013; Cook et al., 2012), which is corroborated and further (8), 2763–2768. specified by the ERP modulation reflecting MNS activity in the present Bucci, S., Startup, M., Wynn, P., Baker, A., Lewin, T.J., 2008. Referential delusions of com- – study. Crucially and of higher clinical relevance, our data imply that de- munication and interpretations of gestures. Psychiatry Res. 158 (1), 27 34. fi Buccino, G., Amore, M., 2008. Mirror neurons and the understanding of behavioural cient social interaction in schizophrenia is rather an automatic process symptoms in psychiatric disorders. Curr. Opin. Psychiatry 21 (3), 281–285. than under voluntary control, which may impact on clinical interven- Chong, T.T.J., Cunnington, R., Williams, M.A., Kanwisher, N., Mattingley, J.B., 2008. fMRI ad- tion strategies by implying effectiveness of training rather than therapy. aptation reveals mirror neurons in human inferior parietal cortex. Curr. Biol. 18 (20), 1576–1580. Our results also nicely complement a recent fMRI study on MNS integri- Cook, R., Bird, G., Lünser, G., Huck, S., Heyes, C., 2012. Automatic imitation in a strategic ty in schizophrenia that reported inferior parietal hypoactivation in the context: players of rock–paper–scissors imitate opponents' gestures. Proc. R. Soc. B schizophrenia group (Thakkar et al., 2014)byconfirming those results 279, 780–786. fi Corbera, S., Wexler, B.E., Ikezawa, S., Bell, M.D., 2013. Factor structure of social cognition in and additionally specifying the temporal course of MNS de cit in the schizophrenia: is empathy preserved? Schizophr. Res. Treat. 409205. cortical information processing cascade. Couture, S.M., Penn, D.L., Roberts, D.L., 2006. The functional significance of social cognition Some limitations have to be acknowledged. Regarding the paradigm, in schizophrenia: a review. Schizophr. Bull. 32 (Suppl. 1), 44–63. Davis, M.H., 1983. Measuring individual differences in empathy: evidence for a multidi- the probability of a repetition trial was two times higher than the prob- mensional approach. J. Pers. Soc. Psychol. 44 (1), 113–126. ability of a non-repetition trial, thus inducing an inherent expectation Derntl, B., Finkelmeyer, A., Voss, B., Eickhoff, S.B., Kellermann, T., Schneider, F., Habel, U., for repetitions. Summerfield et al. (2008) have shown that expectation 2012. Neural correlates of the core facets of empathy in schizophrenia. Schizophr. – – affects repetition suppression. However, when specifically investigating Res. 136 (1 3), 70 81. Dinstein, I., Hassoun, U., Rubin, N., Heeger, D.J., 2007. Brain areas selective for both ob- ERPs, it was also found that expectation neither affected all electrodes served and executed movements. J. Neurophysiol. 98 (3), 1415–1427. nor did it abolish RS effects (Summerfield et al., 2011). Still, a balanced Dinstein, I., Gardner, J.L., Jazayeri, M., Heeger, D.J., 2008. Executed and observed move- ments have different distributed representation in human aIPS. J. Neurosci. 28 (44), task design would have been favorable to rule out the impact of expec- – fi fi 11231 11239. tation with greater con dence. Furthermore, we did not speci cally Enticott, P.G., Hoy, K.E., Herring, S.E., Johnston, P.J., Daskalakis, Z.J., Fitzgerald, P.B., 2008. control for attention, which is a possible confound, because attention Reduced motor facilitation during action observation in schizophrenia: a mirror neu- deficits are highly prevalent in schizophrenia (Shen et al., 2014). Instead ron deficit? Schizophr. Res. 102 (1–3), 116–121. Grill-Spector, K., Henson, R., Martin, A., 2006. Repetition and the brain: neural models of of a behavioral task, we controlled attention indirectly by monitoring stimulus-specific effects. Trends Cogn. Sci. 10 (1), 14–23. muscle activity of the right hand with EMG. Last, although logical infer- Haker, H., Rössler, W., 2009. Empathy in schizophrenia: impaired resonance. Eur. Arch. ence seems to allow for concluding that cross-modal sensori-motor rep- Psychiatry Clin. Neurosci. 259 (6), 352–361. Harris, A., Nakayama, K., 2007. Rapid face-selective adaptation of an early extrastriate etition suppression taps into MNS function, our paradigm cannot component in MEG. Cereb. Cortex 17 (1), 63–70. directly prove the existence of mirror neurons in humans or their dys- Heyes, C., 2011. Automatic imitation. Psychol. Bull. 137 (3), 463–483. function in schizophrenia, which – in a strict sense – can only be Horan, W.P., Iacoboni, M., Cross, K.A., Korb, A., Lee, J., Nori, P., Quintana, J., Wynn, J.K., Green, M.F., 2014a. Self-reported empathy and neural activity during action imitation achieved by invasive single-neuron recordings. and observation in schizophrenia. Neuroimage Clin. 5, 100–108. In conclusion, this is the first study applying a gestural RS para- Horan, W.P., Pineda, J.A., Wynn, J.K., Iacoboni, M., Green, M.F., 2014b. Some markers of digm to infer sensori-motor properties of mirror neurons in schizo- mirroring appear intact in schizophrenia: evidence from mu suppression. Cogn. Af- – phrenia. The data demonstrate differential neuronal responses fect. Behav. Neurosci. 14 (3), 1049 1060. Horn, W., 1983. Leistungsprüfsystem – Handanweisung. 2nd ed. Hogrefe, Göttingen. consistent with mirror neuron activation in healthy subjects and a Iacoboni, M., 2005. Neural mechanisms of imitation. Curr. Opin. Neurobiol. 15 (6), clear deficit in schizophrenia. This finding contributes to a better un- 632–637. Iacoboni, M., Woods, R.P., Brass, M., Bekkering, H., Mazziotta, J.C., Rizzolatti, G., 1999. Cor- derstanding of a neuronal mechanism that is likely to contribute to – fi tical mechanisms of human imitation. Science 286 (5449), 2526 2528. social cognitive de cits in schizophrenia. Intact intra-modal adapta- Jung, T.P., Makeig, S., Humphries, C., Lee, T.W., McKeown, M.J., Iragui, V., Sejnowski, T.J., tion indicates intact perception of gestural stimuli in schizophrenia; 2000. Removing electroencephalographic artifacts by blind source separation. Psy- consequently, deficits of non-verbal communication may rather not chophysiology 37 (2), 163–178. Kilner, J.M., Neal, A., Weiskopf, N., Friston, K.J., Frith, C.D., 2009. Evidence of mirror neu- arise from the perceptual level. Instead, our data in accord with pre- rons in human . J. Neurosci. 29 (32), 10153–10159. vious studies strongly suggests a failure to establish perception- Kriegeskorte, N., Simmons, W.K., Bellgowan, P.S.F., Baker, C.I., 2009. Circular analysis in action links and visuo-motor transformation processes within the systems neuroscience: the dangers of double dipping. Nat. Neurosci. 12 (5), 535–540. Kuehl, L.K., Brandt, E.S.L., Hahn, E., Dettling, M., Neuhaus, A.H., 2013. Exploring the time parietal portion of the MNS. Subsequent research questions, espe- course of repetition suppression: a preliminary study. Int. J. Psychophysiol. 87 cially to which degree this MNS deficit in schizophrenia is longitudi- (2), 183–188. nally correlated with functional outcome and how it may be Lehrl, S., Triebig, G., Fischer, B., 1995. Multiple choice vocabulary test MWT as a valid and short test to estimate premorbid intelligence. Acta Neurol. Scand. 91 (5), 335–345. therapeutically alleviated, remain fascinating topics for future Martin, A.K., Robinson, G., Dzafic, I., Reutens, D., Mowry, B., 2014. Theory of mind and the studies. social brain: implications for understanding the genetic basis of schizophrenia. Genes BrainBehav.13(1),104–117. Matthews, N., Gold, B.J., Sekuler, R., Park, S., 2013. Gesture imitation in schizophrenia. Role of funding source Schizophr. Bull. 39 (1), 94–101. This work was done without funding. Mehta, U.M., Thirthalli, J., Aneelraj, D., Jadhav, P., Gangadhar, B.N., Keshavan, M.S., 2014a. Mirror neuron dysfunction in schizophrenia and its functional implications: a sys- tematic review. Schizophr. Res. 160 (1–3), 9–19. Mehta, U.M., Thirthalli, J., Basavaraju, R., Gangadhar, B.N., Pascual-Leone, A., 2014b. Re- Contributors duced mirror neuron activity in schizophrenia and its association with theory of The study was conceived and designed by authors AHN and NM. Data were assessed mind deficits: evidence from a transcranial magnetic stimulation study. Schizophr. by authors NM and CS. ERP analysis was done by authors NM, EH, and TT. Statistical anal- Bull. 40 (5), 1083–1094. ysis was done by authors NM and MD. The first manuscript version was written by author Mitra, S., Nizamie, S.H., Goyal, N., Tikka, S.K., 2014. Mu-wave activity in schizophrenia: ev- NM. All authors contributed to the preparation of final manuscript and take responsibility idence of a dysfunctional mirror neuron system from an Indian study. Indian for its content. J. Psychol. Med. 36 (3), 276–281. N. Möhring et al. / Schizophrenia Research 168 (2015) 174–179 179

Möhring, N., Brandt, E.S., Mohr, B., Pulvermüller, F., Neuhaus, A.H., 2014a. ERP adaptation Summerfield, C., Trittschuh, E.H., Monti, J.M., Mesulam, M.M., Egner, T., 2008. Neural rep- provides direct evidence for early mirror neuron activation in the inferior parietal etition suppression reflects fulfilled perceptual expectations. Nat. Neurosci. 11 (9), lobule. Int. J. Psychophysiol. 94 (1), 76–83. 1004–1006. Möhring, N., Shen, C., Neuhaus, A.H., 2014b. Spatiotemporal dynamics of early cortical Summerfield, C., Wyart, V., Johnen, V.M., de Gardelle, V., 2011. Human scalp electroen- gesture processing. Neuroimage 99, 42–49. cephalography reveals that repetition suppression varies with expectation. Front. Molenberghs, P., Cunnington, R., Mattingley, J.B., 2012. Brain regions with mirror properties: a Hum. Neurosci. 5, 67. meta-analysis of 125 human fMRI studies. Neurosci. Biobehav. Rev. 36 (1), 341–349. Thakkar, K.N., Peterman, J.S., Park, S., 2014. Altered brain activation during action imita- Mukamel, R., Ekstrom, A.D., Kaplan, J., Iacoboni, M., Fried, I., 2010. Single-neuron responses tion and observation in schizophrenia: a translational approach to investigating social in humans during execution and observation of actions. Curr. Biol. 20 (8), 750–756. dysfunction in schizophrenia. Am. J. Psychiatry 171 (5), 539–548. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory. Thoma, P., Soria Bauser, D., Norra, C., Brüne, M., Juckel, G., Suchan, B., 2014. Do you see Neuropsychologia 9 (1), 97–113. what I feel?—electrophysiological correlates of emotional face and body perception Rizzolatti, G., Craighero, L., 2004. The mirror-neuron system. Annu. Rev. Neurosci. 27, in schizophrenia. Clin. Neurophysiol. 125 (6), 1152–1163. 169–192. Torrey, E.F., 2007. Schizophrenia and the . Schizophr. Res. 97 (1–3), Rizzolatti, G., Sinigaglia, C., 2007. Mirror neurons and motor intentionality. Funct. Neurol. 215–225. 22 (4), 205–210. Tu, P.C., Lee, Y.C., Chen, Y.S., Li, C.T., Su, T.P., 2013. Schizophrenia and the brain's control Shen, C., Popescu, F.C., Hahn, E., Ta, T.T.M., Dettling, M., Neuhaus, A.H., 2014. network: aberrant within- and between-network connectivity of the frontoparietal Neurocognitive pattern analysis reveals classificatory hierarchy of attention deficits network in schizophrenia. Schizophr. Res. 147 (2–3), 339–347. in schizophrenia. Schizophr. Bull. 40 (4), 878–885. Walther, S., Vanbellingen, T., Müri, R., Strik, W., Bohlhalter, S., 2013a. Impaired panto- Singh, F., Pineda, J., Cadenhead, K.S., 2011. Association of impaired EEG mu wave suppres- mime in schizophrenia: association with function. Cortex 49 (2), sion, negative symptoms and social functioning in biological motion processing in 520–527. first episode of psychosis. Schizophr. Res. 130 (1–3), 182–186. Walther, S., Vanbellingen, T., Müri, R., Strik, W., Bohlhalter, S., 2013b. Impaired gesture Smith, M.J., Schroeder, M.P., Abram, S.V., Goldman, M.B., Parrish, T.B., Wang, X., Derntl, B., performance in schizophrenia: particular vulnerability of meaningless pantomimes. Habel, U., Decety, J., Reilly, J.L., Csernansky, J.G., Breiter, H., 2014. Alterations in brain ac- Neuropsychologia 51 (13), 2674–2678. tivation during cognitive empathy are related to social functioning in schizophrenia. Schizophr. Bull. 41 (1), 12–14.