Motor Resonance Mechanisms during Action Imitation in Depression

Transcription

1 Motor Resonance Mechanisms during Action Imitation in Deression Djamila Bennabi, Nicolas Carvalho, Ambra Bisio, Emmanuel Haffen, Thierry Pozzo ABSTRACT Objectives Major deressive disorder has been associated with imairments in social cognition. However, studies exloring the rocessing of social information focused on facial discrimination. The aim of this study was to better characterize the sensorimotor mechanisms underlying motor resonance in deressed atients. Method Twenty-three right-handed atients meeting DSM-IV criteria for uniolar deression were comared to 14 matched healthy controls. In a simle imitation aradigm, the kinematic features of movements in natural condition were comared to those of motions erformed after the observation of a moving dot. Reaction time and ointing velocity were considered to evaluate if the motor erformance was contaminated by the observed stimulus. Results Patient s velocity varied in agreement with dot velocity, roving that they were able to extract the correct information from the stimuli and use it to lan their resonses. Deressed atients actions, as well as healthy controls, were influenced by the dot velocity, suggesting that motor resonance mechanisms are not revented by deression. In contrast, only atients had anticiatory motor resonse and started moving before the end of the stimulus motion. Conclusion Our findings suggest that motor resonance mechanisms are not altered by the disease. However, deressed atients exhibit a secific deficit in motor inhibition in the selection of motor resonses. Keywords Uniolar deression, Major deressive disorder, Psychomotor retardation, Motor resonance, Automatic imitation Introduction Major deressive disorder (MDD) is exected to become a leading cause of disability in the western world, roducing the second largest disease burden by the year 2020 [1]. Eisodes of deression, marked by core deressive emotional and hysical symtoms, are also accomanied by difficulties in social functioning and interersonal relationshis. Imairments in social cognition constitute a substantial comonent of the ersonal, economic and social burden of the disease, reducing time to relase and subsequently increasing suicide risk [2]. However, difficulties in concetualizing and assessing social resonsivity and emathic- 1 Centre Hositalier Universitaire de Besancon, France. 2 IIT@UniFe Center for Translational Neurohysiology of Seech and Communication, Istituto Italiano di Tecnologia, Via Fossato di Mortara, 17-19, Ferrara, Italy. 3 INSERM U Cognition, Action et Plasticité Sensorimotrice, Université de Bourgogne, 21078, Dijon, France. Author for corresondence: Djamila Bennabi, Centre Hositalier Universitaire de Besancon, France djamila.bennabi@univ-fcomte.fr /Neurosychiatry Neurosychiatry (London) (2018) 8(5), ISSN e- ISSN

2 Djamila Bennabi related abilities have imeded rogress towards successful targeted treatment and intervention strategies. Researches seeking to understand the nature of imairments in the rocessing of social information have used rocedures focusing on facial discrimination or metacognitive caabilities. A general emotion recognition deficit in face ercetion have been consistently documented, along with a mood-congruent rocessing bias with hyo-activation towards hay facial exressions and hyer-activation towards sad or negative facial exressions [3,4]. Deficits are not restricted to the visual domain, as anomalies in recognition and identification of emotional rosody has been reorted as well [5]. Additionally, weaknesses in theory of mind (ToM) and metacognition have been observed in ethological studies and emotion research [6,7]. Social functioning requires oerations of increasing level of comlexity deendent on the link between ercetion and action, from early ercetion to high-level self-reflection. If ToM and ersective taking reresent higher-order facets of social skills, motor resonance [8] has been roosed as a basic neural substrate for social information rocessing, in addition to the concet of embodied cognition. Resonance henomenon rovides an imlicit and rereflexive way to establish inter-individual interactions through the mirroring of actions, ostures, gesture or other behaviors. Studies by Rizzolatti [8] and Fadiga [9] have shown that ercetion of a subject s action automatically triggers the observer s corresonding internal reresentations of that action [9,10]. This shared reresentation automatically activates motor areas of the brain leading to the rearation and execution of a motor resonse with features similar to the observed action. This state-matching reaction has been related to the simulation theory and the fronto-arietal network of the human mirroring system [11]. Imitations, and its inhibition during daily life activities, reresent a secific case of action ercetion couling that lays a major adative role by facilitating the learning of new motor skills. Imitative behaviours manifest as the automatic tendency or the intention to reroduce the roerties of the observed motion, and are thought to occur via direct lower-level visuomotor maing [12]. From a clinical ersective, characterizing the sensorimotor mechanisms underlying resonance may otentially reresent a romising and innovative manner to investigate ercetion action couling and emathy-related abilities in neurosychiatric oulations. Observable signs of resonance can be diminished in sychiatric disorders such as autism [13], schizohrenia and ersonality disorders such as sychoathy [14-18]. Assessing emathic abilities with resonance tasks can reveal sontaneous behavioral reactions reflecting daily-life social skills indeendently of articiants intellectual abilities, since rocessing at this level is characterized by being fast, imlicit and domain secific [19,20]. In the current study, we evaluated the sensorimotor mechanisms underlying motor resonance by emloying a simle imitation aradigm. To reduce the imact that higher-order rocesses have on imitation we used a non-human agent model to control social attention. Our general aim was to examine whether the sensorimotor mechanisms underlying motor resonance are affected by MDD. Materials and Method Particiants Twenty-three atients (MDD, 17 females, 6 males, mean 61.8 ± 16.3 years) meeting Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria for uniolar deression and 11healthy controls (HC, 9 females, 2 males, mean 57.6 ± 11.2 years), matched for age, sex and education, articiated in this study. MDD were included into the study if their score was more than 25 on the Montgomery Asberg Deression Rating Scale (MADRS) [21]. Exclusion criteria were: biolar deression, sychotic features, neurological disease, severe organic disease and intake of first-generation antisychotics (FGA). Every atient received an antideressant medication with escitaloram in a constant dosage (10-20 mg/day) over four weeks rior to the exeriment. Concomitant medication with benzodiazeines was ermitted. All articiants gave written informed consent to articiate in the study. Research rotocol was aroved by the Committee of Protection of Persons (CPP-Est-II), and was conducted in accordance with the rincile laid down by the declaration of Helsinski. Psychiatric assessment MDD atients comleted the MADRS and the Saletriere Retardation Rating Scale (SRRS) [22] to determine deression severity and the clinical severity of retardation. All atients were severely deressed (mean scores MADRS: 31,3 ±7,5) and showed a marked degree of retardation (27,95 ± 10,6) Neurosychiatry (London) (2018) 8(5)

3 Motor Resonance Mechanisms during Action Imitation in Deression Research Article Cognitive assessment A trained, licensed neurosychologist conducted a comlete neurosychological test battery in the MDD grou. This battery is designed to rovide a broad assessment of functioning in five cognitive domains: (1) Global cognitive function: Mini Mental State Examination (MMSE) (2) Attention/Processing Seed: Trail Making Test- Part A (TMT A), Crossing Off Test (COT); (3) Executive functions: Trail Making Test-Part B (TMT B), Isaacs Set Test (IST) ; (4) Verbal Memory: Memory Imairment Screen (MIS) ; (5) Language: Picture naming test (DO 30). Movement tasks The task is a modified version of that described in revious studies of our grou [20,23,24]. All articiants were seated in a darkened room in front of a large rear rojection screen (170*230) laced 10 cm beyond the end of articiants extended arm. The visual stimuli were backrojected onto the dislay screen with a videorojector laced behind the screen and connected to a PC. The rojected visual stimulation was generated using MatLab Pyschotoolbox. One assive infrared reflective marker (diameter=20 mm) was alied onto a fingerti of the articiants right hand and arm movements were recorded using an otoelectronic system (SMART CAPTURE) with six cameras recording movements at a samling frequency of 120 Hz. The device was calibrated for each articiant at the beginning of the exerimental sessions. Each articiant erformed a ointing movement exeriment (PM) and a movement observation exeriment (MO). These tasks were resented in two searate blocks, the PM exeriment before the MO. Pointing movement exeriment (PM): this task was aimed at measuring articiants natural ointing movements. The kinematic data served as a baseline to be comared with arm kinematics after motion observation. A green cross aeared on the screen to indicate the starting osition. After 3 s, the cross disaeared and two vertically aligned light blue dots (3.2 cm in diameter with a 51 cm ga between them) were dislayed for 3 s. One of the two dots relaced the green cross and the other one was the target for the movement (Figure 1a). Particiants erformed uwards movements with their right arm in an extended osition from the given starting osition to the target dot at sontaneous natural velocity. The ointing movement was reeated five times, and movement accuracy was not emhasized. Movement observation exeriment: In this exeriment, a green cross was dislayed to indicate the movement s starting osition. After 3 s, the green cross was relaced by a light blue dot (3.2 cm in diameter). The dot ket this osition for 1.5 s, and then started to move vertically uwards with a biological kinematic, covering 51 cm of sace. Dot motions differed in mean velocity (V): slow (S=0.28 m/s), medium (M=0.43 m/s), and fast (F=0.52 m/s) (Figure 1b). Stimulus velocities were randomized. Particiants accomlished two tyes of tasks, imlicit (i) and exlicit (e), which differed only for their instruction. In MOi, articiants were asked to oint the green cross, and then to watch Figure 1: Sequence of visual stimuli (a) In ointing movement exeriment (PM) a green cross aeared on the screen to indicate the starting osition. After 3 s, the cross disaeared and two vertically aligned light blue dots (3.2 cm in diameter with a 51 cm ga between them) were dislayed for 3 s. (b) In movement observation exeriment (MO) a green cross was dislayed to indicate the movement s starting osition. After 3 s, the green cross was relaced by a light blue dot (3.2 cm in diameter). The dot ket this osition for 1.5 s, and then started to move vertically uwards, covering 51 cm of sace with 3 different velocities. 1505

4 Djamila Bennabi the dot s movement, wait until the dot reach it s final, visible osition, and finally oint towards this osition. In MOe, the instructions were similar but articiants were requested to imitate the stimulus motion velocity. There were 6 relications in each dot motion velocity resulting a total of 18 trials. The imlicit task receded the exlicit one to revent contamination of the imlicit movement by the exlicit instruction. For both tasks, the beginning of the exeriment was receded by a training hase, which ended when the articiant understood the task and correctly accomlished all the exeriment at least twice. Moreover, each articiant received verbal feedback from the exerimenter during the testing rocedure in order to eliminate any confusion about their aim. Data treatment Data rocessing: Data was low-ass filtered at 5 Hz using a 2nd order Butterworth filter. To define the onset and offset of the movement, we chose a threshold corresonding to 10% of the maximum value of the movement velocity rofile. Data analysis: Analyses were erformed using MatLab software. In PM, data recorded were the reaction (RT, the time elased between the aearance of the two dots and arm s movement onset), the duration (DUR) of the movements, the mean velocity (Vmean), the maximum velocity (Vmax), the normalized jerk (the rate of change in acceleration) and the time eak velocity (TPV, the ratio between the acceleration hase duration and the total movement duration). In MO, the RT (i.e., the difference in time between the end of dot motion and the onset of articiant s ointing movement) and the Vmean were analysed. Statistical analysis A Shairo-Wilk test was erformed to assess the normality of data. The equality of variance was controlled by a Fisher-Snedecor test. Categorical variables were comared using the chi-square test or the Fisher s exact test (if the samle size was less than 5). The Greenhouse-Geisser correction was used when shericity was not assumed. In ointing movement exeriment the kinematic variables obtained during the erformances of the two grous (deressed vs. controls) were statistically comared by means of one-way ANOVAs. In Movement observation exeriment, two mixed-design ANOVAs with grou (MDD vs. HC) as between subject factor and velocity (slow; medium; fast) as within subject factor were alied on reaction time and mean velocity obtained in MOi and MOe. Concerning reaction time, in order to test the ability to inhibit the motor resonse, we comuted how many resonses were characterized by RT<0 for each stimulus velocity. The sloe of the linear fits was rimarily used to evaluate the degree of influence of the stimuli motions onto the movement execution (sloe=1 means erfect reroduction of the stimulus mean velocity). For this reason, in both MOi and MOe the sloes of the regression lines obtained for each articiant in the two grous were statistically comared using a one-way ANOVA (Grou, as between subject factor). In addition, each set of sloe values was comared with a hyothetical non-contaminated behaviour (horizontal line, sloe=0) using two aired t-tests. To directly assess the effect of the instruction given by the exerimenter, a mixed-design ANOVA with grou (MDD vs HC) as between subject factor, velocity (slow; medium; fast) and instruction (imlicit vs. exlicit) as within subject factor was alied on RT and Vmean values. Further, MOi and MOe sloe values for the two grous were comared by mean of an ANOVA with grou (MDD vs. HC), as between subject factor, and instruction (imlicit vs. exlicit) as within subject factor. Differences between both grous in the PM, MOi and MOe tasks were evaluated using ANCOVAs including grou (deressed, healthy control) as a factor, and neurosychological battery scores as covariates. Correlational analyses were erformed to examine the relationshi between kinematical variables in all tasks and clinical scores of deression and sychomotor retardation. The significance alha level was fixed to Effect sizes were measured by artial Eta squared (η 2 ) with small, medium and large effects defined as 0.01, 0.06 and 0.14 resectively. All comutations were erformed using Stata Software release 10.1 (StataCor, College Station, TX). Results Pointing movements exeriment The PM erformances are resented for both grous in Table 1. The RT was not increased 1506 Neurosychiatry (London) (2018) 8(5)

5 Motor Resonance Mechanisms during Action Imitation in Deression Research Article in MDD atients in comarison with the HC grou. MDD atients had lower Vmean and Vmax than HC. The DUR of arm movement was significantly higher for MDD atients comared to HC. The jerk was significantly higher for MDD atients comared to HC. There was no significant difference between the two grous for the TPV. A comlete descrition of the statistical results is rovided in Table 1. Movement observation exeriment Reaction time: The MO erformances are resented for both grous in Table 2. Mixeddesign ANOVA on RT in MOi task had not shown differences between the MDD and HC grous [F (1,32) =2.29, =0.14, η 2 =0.06 ], and not significant effect of the velocity factor [F (2,64) =0.15, =0.86, η 2 <0.05]. In order to identify anticiatory rocesses, the number of resonses characterized by RT<0 was comuted. MDD atients started before stimulus movement stoed only for slow Vd (t(44)= -2.11, < 0.05) (Figure 2). In MOe exeriment, MDD atients and HC had a similar RT [F(1,32)=1.41, =0.24, η 2 <0.05 ] and there was no significant effect of the Vd [F(2,64)=1.53, =0.22, η 2 <0.05] (Figure 2). Mean velocity: The result of the mixed-design ANOVA on MOi task showed that MDD atients had significantly lower V mean than HC [F (1,32)=21.44, <0.001, η 2 =0.4]. There was also a significant effect of the interaction between grou and stimulus velocity [F (2,64)=6.18, <0.01, η 2 =0.16]. Post hoc comarisons showed that the difference between MDD atients and HC was significant for medium and fast velocities (always <0.05) and was not significant for slow velocity (>0.05). The one-way ANOVA comaring the sloes values of the MDD atients (Mean=0.16, S.D=0.28) and HC (Mean=0.27, S.D=0.18) showed that erformances were equally influenced by the observed motion [F (1,32)=1.36, =0.25, η 2 =0.04]. Moreover, the comarison with a hyothetical non-imitative behavior (horizontal line, sloe=0) showed that MDD atients and HC differed significantly to this hyothetical line (resectively deressed: t(44)=2.81, <0.01; controls : t(20)=4.90, <0.001). Hence, MDD atients and HC were automatically influenced by the stimulus velocities (Figure 3). In MOe task, the statistical analysis showed that HC had significantly lower Vmean than MDD [F(1,32)=12.61, <0.01, η 2 =0.28]. Further, there was a significant effect of the stimulus velocity [F(2.64)=20.63, <0.001, η 2 =0.39]. Post hoc values showed significant difference for Table1: Kinematic arameters values of MDD and HC movements in ointing movement task. PM arameters MDD (N=23) HC (N=11) F(1,32) η 2 RT (s) 0.46 ± ± N.S. <0.01 DUR (s) 0.82 ± ± < Vmean (m/s) 0.6 ± ± < Vmax (m/s) 1.18 ± ± < Jerk ± < Tv 0.41 ± ± N.S. <0.01 Legend: MDD: Major Deressive Disorder; HC: Healthy Controls; RT: Reaction Time; DUR: Duration; Vmean: mean velocity; Vmax: maximum velocity; TPV: Time Peak Velocity, η 2 : Partial eta squared; NS: Not Significant. Values given as mean ± S.D. Table 2: Kinematic arameters values of MDD and HC movements in imlicit and exlicit movement observation tasks. MOi MOe MDD (N=23) HC (N=11) MDD (N=23) HC (N=11) RT (s) Slow ± ± 0.2 N.S ± ± 0.2 N.S. Medium ± ± 0.1 N.S ± ± 0.2 N.S. Fast 0.03 ± ± 0.1 N.S ± ± 0.2 N.S. Vmean (m/s) Slow 0.54 ± ± 0.2 N.S ± ± 0.1 NS Medium 0.55 ± ± 0.1 < ± ± 0.1 NS Fast 0.59 ± ± 0.1 < ± ± 0.1 <0.01 Legend: MDD: Major Deressive Disorder; HC: Healthy Controls; Moi: imlicit movement observation; MOe: exlicit movement observation; RT: Reaction Time; Vmean: mean velocity; NS: Not Significant. Values given as mean ± S.D. 1507

6 Djamila Bennabi!"#$%#&'(!"#$%#)'*+,-. Figure 2: Particiants reaction time (RT) as function of the dot velocity (Vd) in PM, MOi and MOe. Horizontal blue and red continuous lines indicate articiants RT values in natural condition (PM). The circles indicate articiants RT values when observing dot motion (Vd). The circle below the x-axis corresonding to the starting movement before the stimulus stoed. Blue and Red elements refer to MDD and HC grous data, resectively. The error bars refer to the standard errors. Figure 3: Particiants mean velocity (Vmean) as function of the dot velocity (Vd) in MOi and MOe. Red colours reresent MDD atients movements and blue colours the HC. Circle indicates MOi Vmean values and triangle indicate MOe Vmean values when observing dot motion as function of the stimuli velocity (x-axis). The error bars refer to the standard errors. The y=x grey line indicates the theoretical erfect imitation of the stimulus velocity. The circle above this line imlies overestimation of the stimulus velocity. The histogram reresents sloe values (y-axis) and statistics. ** indicates a statistically significant effect (<.01) of the task factor (i vs. e), regardless of the stimuli resented. fast velocity between the two grous (<0.01) but not for slow (>0.05) and medium (>0.05) velocities. A one-way ANOVA comaring the sloes values of the MDD atients (Mean=0.52, S.D=0.38) and HC (Mean=1.19, S.D=0.43) showed that the ability to imitate the observed motion were different and significantly increase in HC [F(1,32)=21.26, <0.001, η 2 =0.39] (Figure 3). Moreover, each set of sloe values was comared with a hyothetical non-imitative behavior (horizontal line, sloe=0) using two aired t-test and MDD atients or HC differed significantly of this hyothetical line (resectively deressed: t(44)=2.81, <0.01; controls : t(20)=4.90, <0.001). Imlicit versus exlicit movement observation exeriment Reaction time: MDD atients and HC had similar RT values [F(1,32)=2.94, =0.09, η 2 =0.08]. There was no significant effect of velocity [F(2,64)=0.25, =0.77, η 2 <0.01], instruction [F(1,32)=0.006, =0.93, η 2 <0.001], and no interaction between velocity and instruction [F(2,64)=0.14, =0.86, η 2 < 0.01] Neurosychiatry (London) (2018) 8(5)

7 Motor Resonance Mechanisms during Action Imitation in Deression Research Article Mean velocity: MDD atients had significantly lower V than HC [F (1,32)=21.47, <0.001, η 2 =0.4]. There was a significant effect of velocity [F(2,64)=14.03, <0.001, η 2 =0.3], instruction [F(1,32)=6.56, <0.05, η 2 =0.17], and an interaction between velocity and instruction [F(2,64)=14.92, <0.001, η 2 =0.32]. Post hoc showed significant effect of instruction on slow and medium velocities (resectively <0.001; <0.01) in MDD grou. There was no significant effect for fast velocity (>0.05). In HC, instruction had only significant effect for slow velocity (<0.01) and no for medium (>0.05) and fast (>0.05) velocities. At last, in order to evaluate if imitation was imroved in MOe, the sloe values of the linear regression models in MOi and MOe were statistically comared. The results of the mixeddesign ANOVA revealed that the imitative erformances significantly imroved in MOe in the two oulations [F (1,32)=8.61, <0.01, η 2 =0.21]. Relationshis between kinematical variables and clinical factors Correlational analyses were erformed to examine the relationshi between kinematical variables in all tasks and clinical scores of deression and sychomotor retardation. No significant correlations were found. Discussion The resent study investigated for the first time the sensorimotor mechanisms underlying motor resonance in atients with severe uniolar major deression. Imitation and inhibition abilities were evaluated by means of the automatic (imlicit) and voluntary (exlicit) imitation aradigm during which articiant erformed an arm ointing movement. Our main findings revealed (1) a global slowness of movements in each exerimental condition, (2) an intact ability to voluntarily imitate and to be imlicitly influenced by the stimulus velocity in atients with MDD comared with a control grou of healthy adults. These results suggest that lowlevel resonance rocess is not affected by uniolar deression. Differently, a motor inhibitory deficiency aeared in deressed atients. Kinematic features of the articiants ointing movement in natural condition (PM) In natural condition (i.e., subjects asked to oint toward a satial target), the mean RT of MDD atients and HC were similar. Longer RT values have been reviously reorted in serial choice task with a high attention demand [25] but not systematically with simle reaction time rocedures. In line with these observations, it has been roosed that deressive retardation does not affect all the stes of central nervous system information treatment, but was limited to the comonents of resonse-selection and motor-adjustment [26]. Moreover, antideressant treatments may have imroved RT before the exeriment, without imart a arallel reduction in the severity of deression. Beneficial effects of antideressants on temoral measures of motor function such as RT have been reorted reviously with conventional tricyclic antideressants or SSRI [27]. However, in the absence of longitudinal data, we cannot determine whether antideressants have lead to a reduction in the severity of the motor imairment. MDD atients showed increased duration and showed more jerky arm movement (i.e. with higher eak accelerations being measured and less efficient movement aths) relative to healthy articiants. We suggest that this finding reflects MDD atients necessity to continuously monitor the ongoing action, and their deendency to the sensory feedback information during erformance execution. To exlore the ossibility of a eriheral motor deficiency, the use of control tasks that involves a balistic movement would be useful in future researches. These behavioural motor abnormalities observed in our deressed atients are similar to those reorted in other athological conditions such as Alzheimer s disease and Parkinson disease. Concerning Alzheimer s disease, revious studies showed that atients were not able to maintain the initial motor lan throughout its course, as indicated by the increased movements duration and jerk with resect to healthy aged-matched articiants. Regarding Parkinson disease, revious studies have drawn similarities between bradyhrenia in deressed atients and bradykinesia in Parkinson disease, secifically in self-initiated movement in reliance to external or internal cues, or in rograming the velocity of movement [28-30]. Irregular atterns of velocity in deressed atients may stem from functional dysfunction of a brain network including the basal ganglia, the sensorimotor cortices and the sulementary motor area, essential for movement sequencing [31]. Sohisticated studies, including brain 1509

8 Djamila Bennabi imaging methods and using a motor aradigm, would be needed to confirm this hyothesis. Pointing movements of MDD atients were contaminated by the stimulus velocity as well as those of HC In imlicit task, HC ointing velocities varied as a consequence of the stimuli velocities, suggesting an intact ability to automatically match the erceived kinematics with brain action reresentation, as already shown in revious studies on both young and elderly healthy articiants [20,24]. As exected, imitation erformance was imroved when HC was exlicitly instructed to reroduce the stimulus velocity. Similarly, deressed atients behaviour was imlicitly influenced by the dislay velocity, and their imitative erformances were significantly imroved in MOe during voluntary imitation. Thus, our findings indicated that deression did not alter automatic imitation and the associated mechanisms of imlicit motion recognition. This result indicates that visual rocessing automatically induces related motor resonses and rovides evidence that the sensorimotor resonance mechanisms (i.e. ercetion action matching) underlying automatic imitation are not affected by the athology. Literature on motor imitation demonstrated that motion observation and imitation induce simultaneous activation of fronto-arietal mirror neuron system [12]. Fadiga [9] roosed that these brain areas would give rise to the resonance mechanism that directly and imlicitly mas a kinematic descrition of the observed action onto an internal motor reresentation of the same action [9]. Therefore, the reservation of automatic imitation in MDD imlies the functioning of a circuit which involves the mirror neuron system, the suerior temoral sulcus [32] and the sensorimotor areas for movement roduction. However, evidence from functional magnetic resonance imaging (fmri) and ositron emission tomograhy (PET) identified the involvement of frontal and arietal regions in deression [33]. Another hyothesis could be the resence of mechanisms imlying intact brain regions taking over the functions to comensate for functional alterations of cortical and subcortical motor regions [34]. Uncontrolled initiation of deressed atients motor resonse while observing a moving stimulus In MO, deressed atients tended to start before the stimulus reached his final osition, in contrast to healthy subjects who were able to wait until the stimulus stoed before starting their movement. In fact, the mere resence of the stimulus was sufficient to trigger the action indicating the deendence on the visual stimulus. This uncontrolled motion initiation would indicate deressed atients deficiency to voluntarily inhibit resonse roduction and could be underinned by inadequate functioning of the inhibitory mechanisms. Our observation corroborates and exands to the sensorimotor domain reviously described deficits in resonse monitoring in MDD while rocessing neutral or emotional information [35,36]. Inhibition describes an active rocess that temers unwanted stimuli (external or internal) that comete for rocessing resources in the context of a limited caacity system and reresent a key mechanism in the regulation of emotion [37]. Direct behavioral evidence indicates that individuals with major deression have greater deficits than healthy subjects in inhibition task such as the Stroo test [38,39] the Sto Signal task [40] and the antisaccade tasks [41]. These studies suggest that sychomotor retardation cannot fully account for imaired erformance on inhibition tasks and highlight the need to distinguish between basic motor slowing and more elaborated motor rocesses involved in secific inhibitory disorder [42]. A systematic distinction between these rocesses is critical to better characterize the kind of cognitive-motor rocesses that are altered in deression. From a clinical ersective, inhibition deficit could lay a fundamental role in the clinic exression of MDD, leading to behavioral trouble such as suicidal behaviour, and may hel to redicts adverse outcomes of deression [43,44]. Additionally, brain imaging data indicates that the anterior art of the frontal lobe may be critical for inhibiting automatic imitation [45]. Inhibition deficits in MDD could underlie alterations of the frontal network and its role behavioural regulation. A ossible limitation of this study, which is common to many exeriments in the field, is related to the effect of sychotroic medication. Pharmacological treatments can contribute to imrove sychomotor functioning, but may also have disrutive effects, causing sedation or imairment in sychomotor and cognitive function Neurosychiatry (London) (2018) 8(5)

9 Motor Resonance Mechanisms during Action Imitation in Deression Research Article Conclusion Motor imitation reresents a owerful biological resource for cognitive develoment and social interaction. Notably, low-level sensory-motor matching mechanisms that work on movement lanning reresent the basis for higher levels of social interaction. Our atients erformances suggest that the resonance mechanisms at the basis of social cognition are reserved during uniolar deression. Beyond movement retardation, deressed atients exhibited a secific deficit in motor inhibition in the selection of motor resonses that may underlie the resence of altered social cognition abilities. Indeed, deficits in cognitive control are related to the use of maladative emotion regulation strategies and have been associated with social dysfunction resonsible of social interaction skills degradation (i.e. roblems in exressing themselves clearly to others, inaroriately and excessively self-disclosing information, esecially if it is negatively toned) [46]. The alteration of this mechanism is a relevant finding for hysical and cognitive interventions in deression. Future studies should examine the relationshi shared between motor resonances, quality of life or social and interersonal functioning in MDD. Additionally, research investigating the trait or state deendency of action monitoring in severely deressed atients would be highly relevant to hel clarify the long-term clinical and functional outcomes of this oulation.. The combination of transcranial magnetic stimulation (TMS) with brain imaging studies reresents a owerful method of analysis of neural underinnings of action observation in this oulation. This research did not receive any secific grant from funding agencies in the ublic, commercial, or not-for-rofit sectors. References 1. Murray CJL, Vos T, Lozano R, et al. Disabilityadjusted life years (DALYs) for 291 diseases and injuries in 21 regions, : a systematic analysis for the Global Burden of Disease Study Lancet 380 (9859), (2012). 2. May AM, Klonsky ED, Klein DN. Predicting future suicide attemts among deressed suicide ideators: a 10-year longitudinal study. J. Psychiatr. Res 46(7), (2012). 3. Münkler P, Rothkirch M, Dalati Y, et al. Biased recognition of facial affect in atients with major deressive disorder reflects clinical state. PLoSONE 10(1), e (2015). 4. Schneider D, Regenbogen C, Kellermann T, et al. Emathic behavioral and hysiological resonses to dynamic stimuli in deression. Psychiatry. Res 200(2-3), (2012). 5. Schlif S, Batra A, Walter G, et al. Judgment of emotional information exressed by rosody and semantics in atients with uniolar deression. Front. Psychol 4(1), 461 (2013). 6. Ladegaard N, Larsen ER, Videbech P, et al. Higher-order social cognition in first-eisode major deression. Psychiatry. Res 216(1), (2014). 7. Ladegaard N, Lysaker PH, Larsen ER, et al. A comarison of caacities for social cognition and metacognition in first eisode and rolonged deression. Psychiatry. Res 220(3), (2014). 8. Rizzolatti G, Fadiga L, Fogassi L, Gallese V. Resonance behaviors and mirror neurons. Arch. Ital. Biol 137(2-3), (1999). 9. Fadiga L, Fogassi L, Pavesi G, et al. Motor facilitation during action observation: a magnetic stimulation study. J. Neurohysiol 73(6), (1995). 10. Rizzolatti G, Fogassi L, Gallese V. Neurohysiological mechanisms underlying the understanding and imitation of action. Nat. Rev. Neurosci 2(1), (2001). 11. Jeannerod M. Neural simulation of action: a unifying mechanism for motor cognition. Neuroimage 14(1-2), (2001). 12. Iacoboni M. Imitation, emathy, and mirror neurons. Annu Rev Psychol 60(2), (2009). 13. Daretto M, Davies MS, Pfeifer JH, et al. Understanding emotions in others: mirror neuron dysfunction in children with autism sectrum disorders. Nat. Neurosci 9(1), (2006). 14. Hagenmuller F, Rössler W, Endrass J, et al. Imaired resonance in offenders with sychoathic traits. Neurosychiatr 26(2), (2012). 15. Haker H, Rössler W. Emathy in schizohrenia: imaired resonance. Eur. Arch. Psychiatry. Clin. Neurosci 259(6), (2009). 16. Helt MS, Eigsti IM, Snyder PJ, et al. contagious yawning in autistic and tyical develoment. Child. Dev 81(5), (2010). 17. Lhermitte F. [Imitation and utilization behavior in major deressive states]. Bull. Acad. Natl. Med 177(6), (1993). 18. Lhermitte F, Pillon B, Serdaru M. Human autonomy and the frontal lobes. Part I: Imitation and utilization behavior: a neurosychological study of 75 atients. Ann. Neurol 19(4), (1986). 19. Adolhs R. Concetual challenges and directions for social neuroscience. Neuron 65(6), (2010). 20. Bisio A, Casteran M, Ballay Y, et al. Motor resonance mechanisms are reserved in Alzheimer s disease atients. Neuroscience 222(1), (2012). 21. Montgomery SA, Asberg M. A new deression scale designed to be sensitive to change. Br. J. Psychiatry 134(4), (1979). 22. Jouvent R, Frechette D, Binoux F, et al. [Retardation in deressive states : elaboration of a quantitative rating scale (author s transl)]. Encehale 6(1), (1980). 23. Bisio A, Casteran M, Ballay Y, et al. Voluntary Imitation in Alzheimer s disease Patients. Front. Aging. Neurosci 8(1), 48 (2016). 24. Bisio A, Stucchi N, Jacono M, et al. Automatic versus voluntary motor imitation: effect of visual context and stimulus velocity. PLoSONE 5(1), e13506 (2010). 25. Bennabi D, Vandel P, Paaxanthis C, et al. Psychomotor retardation in deression: a systematic review of diagnostic, athohysiologic, and theraeutic imlications. Biomed. Res. Int 2013(1), (2013). 26. Bonin-Guillaume S, Hasbroucq T, Blin O. [Psychomotor retardation associated to 1511

10 Djamila Bennabi deression differs from that of normal aging]. Psychol. Neurosychiatr. Vieil 6(2), (2008). 27. Taylor BP, Bruder GE, Stewart JW, et al. sychomotor slowing as a redictor of fluoxetine nonresonse in deressed outatients. Am. J. Psychiatry 163(1), (2006). 28. Caligiuri MP, Ellwanger J. Motor and cognitive asects of motor retardation in deression. J. Affect. Disord 57(1-3), (2000). 29. Rogers D, Lees AJ, Smith E, et al. Bradyhrenia in Parkinson s disease and sychomotor retardation in deressive illness. An exerimental study. Brain 110(Pt 3), (1987). 30. Rogers MA, Bradshaw JL, Phillis JG, et al. Parkinsonian motor characteristics in uniolar major deression. J. Clin. Ex. Neurosychol 22(2), (2000). 31. Walther S, Höfle O, Federsiel A, et al. Neural correlates of disbalanced motor control in major deression. J. Affect. Disord 136(1-2), (2012). 32. Iacoboni M, Woods RP, Brass M, et al. Cortical mechanisms of human imitation. Science 286(5449), (1999). 33. Shen T, Li C, Wang B, et al. Increased Cognition Connectivity Network in Major Deression Disorder: A fmri Study. Psychiatry Investig 12(2), (2015). 34. Bracht T, Federsiel A, Schnell S, et al. Cortico- Cortical White Matter Motor Pathway Microstructure Is Related to Psychomotor Retardation in Major Deressive Disorder. PlosONE 7(7), e52238 (2012). 35. Leänen JM, Milders M, Bell JS, et al. Deression biases the recognition of emotionally neutral faces. Psychiatry Res 128(2), (2004). 36. Suslow T, Junghanns K, Arolt V. Detection of facial exressions of emotions in deression. Percet. Mot. Skills 92(3), (2001). htts://doi.org/ /ms Hasher L, Stoltzfus ER, Zacks RT, et al. Age and inhibition. J. Ex. Psychol. Learn. Mem. Cogn 17(1), (1991). 38. Koss E, Ober BA, Delis DC, et al. The Stroo color-word test: indicator of dementia severity. Int. J. Neurosci 24(1), (1984). 39. Sass SM, Heller W, Fisher JE, et al. Electrohysiological evidence of the time course of attentional bias in non-atients reorting symtoms of deression with and without co-occurring anxiety. Front. Psychol 5(1), 301 (2014). 40. Amieva H, Lafont S, Auriacombe S, et al. Inhibitory breakdown and dementia of the Alzheimer tye: a general henomenon?. J. Clin. Ex. Neurosychol 24(4), (2002). 41. Carvalho N, Laurent E, Noiret N, et al. Eye Movement in Uniolar and Biolar Deression: A Systematic Review of the Literature. Front. Psychol 6(1), 1809 (2015). 42. Snyder HR, Banich MT, Munakata Y. All cometition is not alike: neural mechanisms for resolving underdetermined and reotent cometition. J. Cogn. Neurosci 26(11), (2014). 43. Alexooulos GS, Kiosses DN, Heo M, et al. Executive dysfunction and the course of geriatric deression. Biol. Psychiatry 58(3), (2005). 44. Potter GG, Kittinger JD, Wagner HR, et al. Prefrontal neurosychological redictors of treatment remission in late-life deression. Neurosychoharmacology 29(1), (2004). 45. Cai W, Ryali S, Chen T, et al. Dissociable roles of right inferior frontal cortex and anterior insula in inhibitory control: evidence from intrinsic and task-related functional arcellation, connectivity, and resonse rofile analyses across multile datasets. J. Neurosci 34(44), (2014). 46. Tse WS, Bond AJ. The Imact of Deression on Social Skills: A Review. J. Nerv. Ment. Dis 192(4), 260 (2004) Neurosychiatry (London) (2018) 8(5)