Emotions and voluntary action: What link in children with autism?

Behavioural Brain Research 251 (2013) 176–187 Contents lists available at SciVerse ScienceDirect Behavioural Brain Research journal homepage: www.elsevier.com/locate/bbr Emotions and voluntary action: What link in children with autism? S. Vernazza-Martin a,∗ , S. Longuet b , J.M. Chamot a , M.J. Orève c a Centre de Recherche sur le Sport et le Mouvement EA 2931, UFR STAPS, Université Paris Ouest Nanterre La Défense, 92000, France Ecole supérieure de biomécanique appliquée à l’ostéopathie, Osteobio, 94230 Cachan, France c Unité d’accueil 4047 Etudes cliniques et innovations thérapeutiques en psychiatrie, Université Versailles Saint Quentin, Centre Hospitalier de Versailles, 177 rue de Versailles, 78157 Le Chesnay Cedex, France b h i g h l i g h t s • • • • • The organization of voluntary movement involves cognitive and automatic processes. This organization depends on the emotion conferred to the goal of the action. A positive emotional valence promotes the cognitive processes in autism. An aversive emotional valence blocks or disturbs it. No emotions effect is observed on the automatic processes. a r t i c l e i n f o Article history: Received 20 November 2012 Received in revised form 23 May 2013 Accepted 25 May 2013 Available online 3 June 2013 Keywords: Emotions Cognitive processes Automatic processes Children with autism Goal directed locomotion a b s t r a c t This research focuses on the impact of emotions – defined as “motivational states” – on the organization of goal directed locomotion in children with autism. Walking toward a goal involves both cognitive processes responsible for movement planning and automatic processes linked to movement programming. To these processes, motivation leading to achieving the goal is added. For some authors, a deficit of planning and/or programming processes is highlighted in autism. Others stand for some impairment of the emotional system. The aim of this research is to link these two viewpoints and to determine if, in children with autism, the organization of locomotion is affected by a positive/aversive emotion conferred to an object to fetch. Twenty-nine children participated in the study (11 children with autism – mean age 122 months; 9 mental age-matched controls – mean age 36 months; and 9 chronological age-matched controls – mean age 122 months). They were instructed to go and get a positive or aversive emotional valence object located straight ahead, at 30◦ to the right or straight ahead then moved at mid-distance to the right. Gait analysis was performed using the Vicon system. The main results suggest that a positive emotional context promotes the cognitive processes involved in movement planning while an aversive emotional context blocks it or disturbs it in children with autism. No emotions effect is observed on movement programming. It is suggested that emotions triggered off and modulated movement planning and that the deficit observed was related to a developmental impairment rather than to a developmental delay. © 2013 Elsevier B.V. All rights reserved. 1. Introduction The human emotional nature is complex and this probably explains the many theories which have been elaborated so far to try and define the emotions in their behavioral, physiological or ∗ Corresponding author at: Université Paris Ouest Nanterre La Défense, UFR STAPS, Bât S. 200 av de la République, 92000 Nanterre, France. Tel.: +33 06 32 23 76 13. E-mail addresses: [email protected] (S. Vernazza-Martin), [email protected] (S. Longuet), [email protected] (J.M. Chamot), [email protected] (M.J. Orève). 0166-4328/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbr.2013.05.049 cognitive and subjective components without consensus on general definition of the term “emotion” [1–4]. It is then crucial to define what we understand by “emotions”. According to Frijda, emotions would correspond to “motivational states”; motivation being responsible for the release, the maintenance and the cessation of an intended behavior as well as the appetitive or aversive value conferred on the goal of the action and/or to the elements of the environment on which this behavior is exerted [5]. Within this framework, the affectivo-motivational models of the behavior indicate that these motivational states induce the approach of positive experiences and the avoidance of aversive experiences [6–8]. In this way, goal directed locomotion is a particularly interesting S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 movement because it is the primary motor behavior used by many living beings. Thus they interact with the environment by moving their whole body in order to either reach a precise point of space or, on the contrary, avoid it. Goal directed locomotion and more generally voluntary movements are organized by the central nervous system in three hierarchical steps [9–11]. The first one is movement planning. It corresponds to cognitive processes allowing the emergence of the act intent, the identification of the goal of the task, the decisionmaking leading to action. Moreover according to Bernstein [12] there would exist, at higher levels of the nervous system, a “higher engram” which would be strongly geometrical and which would represent a very abstract motor image of space (topological and not metric coding). Cognitive processes are therefore involved in the representation of an abstract movement trajectory linked to an estimate of the spatial and postural contexts on the environment and body states. Then the second step corresponds to movement programming. It involves automatic processes corresponding to the motor programs. Not only do they include the movement parameters (amplitude, duration, velocity) but also the kinematic strategies controlling balance throughout the movement, i.e. the balance strategies without which a movement becomes ineffective. They also include the muscles involved both in movement and balance control. These programs profit from automatic regulations conferring on their execution a limited flexibility of expression according to the circumstances and the risks of the environmental conditions. The whole of the genetically wired programs constitutes the basic repertoire available to organize motor behaviors. Movement programming is itself dependent on the movement planning of the higher level which selects, triggers and modulates the motor programs. It can be considered as an interface between movement planning and the third step: movement execution. Thus, the goal oriented movement execution is the objective result of the two previous steps. The effect of emotions on the initiation of gait or during walking is beginning to be studied in the healthy adult. Indeed Stins and Beek [13] show that this effect is especially impacted by negative emotions, thus unpleasant images caused an initial “freezing” response and a tendency to move away from the stimuli during whole-body movements such as voluntary stepping. On another side a study of gait initiation toward pleasant or unpleasant images showed that negative pictures viewing increased reaction time (delay between the beginning of picture viewing and the beginning of the dynamic phenomena on the anteroposterior axis) and decreased the amplitude of the early postural component associated to gait initiation. This was done without a modification of the length of the first step [14]. Lastly, Naugle et al. [15] showed that pollution or mutilation pictures viewing during launched locomotion led to a decrease of length and velocity of the first two steps. Nevertheless, if these studies give us very useful information about the effect of emotions on the initiation and/or execution of a voluntary movement, the emotional value conferred on the goal of this action remains unclear. In fact, in daily life, voluntary movements can be defined as movements which answer the intent linked to the motivation to realize a goal oriented task in a specific context [16]. Initiating a gait or walking toward a picture only corresponds to answering the order to walk toward it, whatever its emotional load. It is different from the realization of a goal directed locomotion which depends on the motivation at the origin of the movement. This motivation depends on the positive or negative emotional value conferred on the goal of this action; “Emotion feelings constitute the primary motivational component of mental operations and overt behaviour” [4]. But what is happening when the movement organization and the emotional system are impaired the way it is in autism? Indeed, autism is classically defined as a disorder characterized by a triad of 177 impairments including social difficulties, communication trouble and restricted interests together with repetitive behavior (American Psychiatric Association, 2000). But some authors highlight a deficit of the executive functions including the whole cognitive processes allowing the control and the execution of finalized activities [17–20] and then movement planning [21,22]. Others point the impairments of the movement itself, assuming a disturbance of the automatic processes linked to a deficient movement programming [22–27]. Others still stand for an impairment of the emotional system. Indeed, children with autism are able to feel the four basic emotions (fear, anger, enjoyment and sadness), to establish links between them and to identify different valence emotions. However, they should be more sensitive to an aversive stimulus. They badly identify their own emotions and feel more often aversive emotions than the children of the same age [28–33]. Moreover, the children with autism have difficulties not only as far as the perception of social-emotional signals is concerned but also with regards to the regulation of their behavior in response to these signals [34,35]. This impairment of the emotional system allows an inability to interact in an emotional way with other people and it would be an original cause of autism [36,37]. If these three components of the voluntary movement organization: emotions, cognitive and automatic processes are clearly impaired in autism, their link has never been studied. Yet, determining if cognitive and/or automatic processes are directly impaired in autism or if their impairment is the result of an impaired emotional system, seems to be fundamental not only for the diagnosis but for the therapy too. In a recent study, we showed that planning was mainly disturbed in an emotionally negative situation when children with autism were walking in a straight line to retrieve a positive or aversive emotional valence object. Movement programming was consequently impaired. Then, this finding suggests that cognitive and/or automatic processes are not directly impaired in autism since they are preserved in an emotionally positive situation [38]. But up to now, it only concerns a motion toward an object located opposite the children. What is happening if the object is not located in front of them or is being shifted while they walk? Thus, the present study emphasizes the link between emotional system and movement planning by analyzing the effect of a positive emotion or an aversive one conferred to an object to fetch, object located in three different positions in space. Each position would induce a particular locomotor trajectory, while the goal of the task remains unchanged (fetching the object). The first corresponds to an object located opposite the children inducing a straight line trajectory (control condition). The second corresponds to an object located sideways inducing, from scratch, a deflected trajectory with respect to the straight line. The last corresponds to an object shifted while walking inducing a reorganization of movement planning in the midst of the action in order to determine the new trajectory with respect to the moved object. 2. Methods This research was the subject of a collaboration between researchers in neurosciences and psychology of the laboratory “Research Centre on Sports and Movements” of the University Paris Ouest Nanterre La Défense and some practitioners of the hospital complex Théophile Roussel, supervised by Dr. Orève, in Montesson, France. 2.1. Participants Eleven children with autism (mean age: 119 months, range 78–158 months) were recruited at the Hospital Complex Théophile Roussel for this experiment. They were selected by an experienced clinician (Hospital MD, specialized in psychiatry), according to the criteria of ICD-10 and evaluated by the Childhood Autism Rating Scale: CARS [39]. Total CARS scores range from 15 to 60, with a minimum score of 30 serving as the cutoff for a 178 S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 diagnosis of autism. So, only children with a total score higher than 30 were selected in our study (CARS average: 41.8 ± 9). The “Psycho-Educative Profile”: PEP-R [40] was used to determine mental age. The mean development age of the selected children was 36 ± 19 months. Children with Rett syndrome, Asperger syndrome, and disintegrative disorders of childhood were excluded. We also recruited eighteen normally developing children from the local community as controls and divided them into two groups: a chronological age-matched group (CA-matched: 9 children, mean age: 116 months, range 81–151 months) and a mental age-matched group (MA-matched: 9 children, mean age 36 months, range 17–81 months). All children participated in the study with their parents’ consent, according to the declaration of Helsinki and the approval of the Ethical Committee. 2.2. Materials and procedure Because of the specific problems which can be encountered during an experiment at a pedopsychiatric center, the experiment was divided into two steps. The children with autism were recorded in the psychomotricity room in the Unit Misès C within the hospital complex Théophile Roussel. Known by all the children, this room allowed preserving a familiar environment. For practical questions, the experiment was made during 15 consecutive days. Each child with autism was accompanied by his “referent”, the specialized educator who was responsible for the child in the hospital. With regard to control children, the absolute experiment replica took place in the university laboratory. The children were accompanied by their parents. In the psychomotricity room as well as in the laboratory, all objects or materials that could have distracted children were removed and the curtains were closed. Participants, individually tested, were instructed to go and get a positive or aversive emotional valence object on a desk 5 m away from themselves in three direction conditions: first, the object was located straight ahead, second, it was positioned at an angle of 30◦ to the right (condition: “30◦ ”) and third, the object was initially placed straight ahead then, when the child reached a middle distance mark on the ground, it was shifted by an experimenter located behind the desk. This shift meant it was moved from 49◦ to the right in order to put it in the same position as the “30◦ ” condition (condition: “moving”) (Fig. 1A). In order to homogenize the experimental conditions and not to disturb the children when the object was shifted, the experimenter stood behind the desk on all trials. The starting position of the locomotion was indicated by a cross. The oral instruction “go and get the object” was given to the child by an unknown experimenter so as not to interfere with the motivation experiment. The two objects were determined for each child by the children themselves and their “referents” or their parents thanks to the Andrews and Whitney’s scale [41] (Fig. 1C). This allows determined the emotional valence conferred to the object as positive versus aversive. Unfortunately the arousal of the emotional valence conferred to the object was not taken into account due to an inability for the studied children with autism to evaluate this aspect. Fig. 1. (A) Experimental setup. The crosses on the desk represent the two possible object locations: straight ahead and from 30◦ to the right. With respect to the initial position (cross between the feet), three theoretical trajectories are studied: straight ahead, deflected from 30◦ to the right with respect to the straight ahead and straight ahead then moved at middle distance (dotted line) from 49◦ to the right in order to locate the object in the same position as the “30◦ ” condition. (B) Body reconstruction from the markers location on the body. (C) Simplified scale with three faces from Andrews and Whitney. Children had to point out the face corresponding to their feelings in the presence of the selected object. The chosen object was located completely to the left for aversive object or completely to the right for positive objects. S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 60 30 18 60 30 18 4M–5F 1–5 6–7 8–9 Mean ± sd CA-matched MA-matched 36 ± 20 116 ± 23 5M–6F Mean ± sd Children with autism M: male; F: female. A significant difference between the realized and successful trials (Z test) is observed in children with autism only when the emotional valence conferred to the object is aversive (Z = 2.04, p < 0.05). 30 15 9 30 15 9 30 15 9 30 15 9 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 10 5 3 60 60 30 30 30 30 10 10 10 10 10 10 10 10 10 10 30 0 1 30 15 0 14 3 15 1 15 30 15 24 30 16 18 16 15 15 18 15 10 5 8 10 5 6 5 5 5 6 5 10 5 8 10 5 6 5 5 5 6 5 10 5 6 10 5 6 4 5 5 5 5 10 0 1 10 5 0 4 1 5 0 5 10 5 8 10 5 6 6 5 5 6 5 10 5 8 10 5 6 6 5 5 6 5 10 5 7 10 5 4 3 2 5 6 5 10 0 0 10 5 0 5 1 5 0 5 30 15 24 30 16 18 16 15 15 18 15 30 14 20 30 16 16 11 12 15 17 15 O− 10 0 0 10 5 0 5 1 5 1 5 10 4 7 10 6 6 4 5 5 6 5 10 5 8 10 6 6 5 5 5 6 5 10 10 60 30 48 60 32 36 32 30 30 36 30 Total R Total S O− Total R Total S Total R O+ O− S trials O+ O− R trials O+ O− O+ O+ O+ S trials R trials S trials R trials O+ 30◦ Straight ahead 10 5 8 10 6 6 5 5 5 6 5 31 56.5 47.5 30.5 34.5 40 42 43.5 33 53.5 48 42 ± 9 27 14 15 78 53 33 48 33 38 15 37 36 ± 19 86 158 138 118 153 104 78 120 120 106 123 119 ± 25 M M M M M M F F F M M An analysis of variance (ANOVA) with repeated measures was used to assess the effect of the factor GROUP (CA-matched/DA-matched/children with autism), the factor VALENCE (positive vs. aversive) and the factor DIRECTION (straight ahead/30◦ /moving). This was taken into account for each parameter except for the trajectory analysis. In this last case, the analysis took 1 2 3 4 5 6 7 8 9 10 11 Mean ± sd 2.4. Statistical analyses Mental age (month) (dx2 + dy2 ) where dx: amplitude on the anteroposterior axis and dy: ampli- Chronological age (month) d= tude on the lateral axis. The amplitude, the duration and the average velocity of the stride were normalized according to the height of the child. - The balance strategies used during walking. They corresponded to the intersegmental coordination between the head, the shoulders and the hip in the lateral plane and were obtained with the head, shoulder and hip anchoring index in the lateral plane [43,44]. This index allowed comparing the stabilization of a given segment with respect both to the external space and to the underlying anatomical segment. It was calculated for each trial with the following formula: IA = (Sd Rel2 − Sd Abs2 )/(Sd Rel2 + Sd Abs2 ) where Sd Abs was the absolute standard deviation and Sd Rel was the corresponding standard deviation of the angular distribution about the lateral axis of the segment under investigation with respect to the underlying anatomical segment. The AI values were then normalized by the use of transformed Z. For any given experimental condition, a positive AI value of a given segment indicated a stabilization in space and so the use of an “articulated” balance strategy. A negative value of the AI indicated a stabilization on the underlying anatomical segment and so the use of a “bloc” strategy. An AI value close to 0 indicated a lack of stabilization. Gender 2.3.2. Movement programming was studied through: - The gait parameters: the amplitude, the duration and the average velocity of the stride (parameters between two heel contacts of the same foot determined by the heel marker). The amplitude calculation took into account a potential movement of the foot according to the lateral axis according the following formula: CARS score d2 /N; where d was the Euclidean distance between the real and theoretical trajectories coordinates, and N is the number of points in the trajectory) [42]. So, results were represented in centimeters and in terms of absolute value. - The distance between the last position of the child (last position of the sacrum [L5] which better represented the whole body position) and the object with respect to the anteroposterior axis and the lateral axis. The distance was represented in centimeters. Table 1 Chronological age, mental age, CARS score, and number of realized (R) and successful (S) trials.  O− 2.3.1. Movement planning was studied through: - Goal achievement, i.e. the percentage of successful trials, a trial being considered as achieved when the child catches or touches the object. - The mean deviation between the real children trajectory (L5: sacrum segment trajectory) and the theoretical trajectory corresponding – for each experimental condition – to the most direct trajectory between the initial position of the child and the object location (see Fig. 1A). The distance between the straight line and the real trajectory was calculated thanks to the root mean square  error (RMSE): mean O− Moving 2.3. Coding and preliminary analysis sums between both trajectories were added (RMSE = 60 14 21 60 31 16 25 15 30 18 30 Total S Having no instruction concerning which of the hands would take the object, children were allowed to get the object either with their right hand or their left one. When fetching the object, some children with autism were initially accompanied by their “referent”, of course without the latter directing or touching them in any way. Children were free to go everywhere in the room, no predetermined trajectory was fixed, and they were free not to answer the instructions. The locomotion analysis was performed using an automatic motion analyzer (VICON system) with six cameras having a sampling frequency of 200 Hz. The children wore a bodysuit on which 22 spherical retroflective markers (14 mm in diameter) were taped on the children’s back. They were placed as follows: head (3 markers), thoracic vertebras T7 and T12, lombar vertebra L5 (sacrum), right and left acromion, elbow, iliac spine, trochanter, tibial plate, heel, external malleolus, and fifth metatarsal joint (Fig. 1B). Up to six sets of ten trials (one trial in each experimental condition: positive and aversive valence in the three directions) were performed (so it meant a total of sixty trials). The trials were randomized inside a set, in order to limit the learning and anticipating phenomenon. CA-matched controls carried out all the 60 trials. Concerning MA-matched controls, their concentration and attention ability being sometimes limited, the number of recorded trials varied between 30 and 60. As for the children with autism, only two children could carry out the 60 trials; the others stopped everything after 30 trials. For 3 children, only the trials with the positive emotional valence objects were recorded, these children having refused to walk in the negative emotional context. Moreover, according to the trials and independently from the emotional valence conferred to the object, some children with autism turned around (so hiding markers from cameras), went out of the camerafields, ran or did not tolerate the markers on their bodysuit and removed them. Because of that, the total number of analyzed trials varies between 14 and 60 (see Table 1). 179 S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 180 Table 2 Percentage of trials achieved under each condition. Children with autism O+ straight ahead O− straight ahead O+ 30◦ O− 30◦ O+ moving O− moving S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 100.0 80.0 87.5 100.0 100.0 100.0 80.0 100.0 100.0 100.0 100.0 100.0 0.0 0.0 100.0 0.0 80.0 100.0 100.0 20.0 16.7 100.0 100.0 100.0 75.0 100.0 100.0 100.0 80.0 100.0 100.0 83.3 100.0 100.0 0.0 12.5 100.0 0.0 100.0 80.0 100.0 20.0 0.0 100.0 100.0 100.0 87.5 100.0 66.7 100.0 50.0 100.0 40.0 100.0 100.0 100.0 0.0 0.0 100.0 0.0 100.0 83.3 100.0 20.0 0.0 100.0 Mean sd 95.2 8.4 56.1 47.4 94.4 9.8 55.7 47.8 85.8 22.7 54.8 49.2 CA-matched MA-matched 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 place in two steps. The first consisted in analyzing the mean deviation trajectory in the “straight ahead” and “30◦ ” conditions. Then the analysis concerned the “moving” condition after the shift of the object. The latter analysis allowed studying specifically the new “in line” movement planning. According to Cohen [45] the effect size was specified by the partial eta squared (2p ). The Bonferroni tests were used as all pairwise multiple comparison procedures and the significance level was set at p < 0.05. The aim of this article being to evaluate the effects of the valence conferred to an object to reach according to its location in space in children with autism compared with chronological age-matched (CA-matched) and mental age-matched (MA-matched) control children, the results of the multiple comparison procedures between the two control children groups will not be presented here. The ANOVA was completed by a correlation study between the emotional valence conferred to the object and goal achievement in each experimental condition (R). The trials when the children ran or refused to do the task were excluded from the analysis except for the reaching of the goal (number of successful trials). 3. Results 3.1. Movement planning 3.1.1. Successful trials Table 2 summarizes the percentage of trials achieved under each condition by children with autism. All control children (CA-matched and MA-matched) having achieved the goal, their Fig. 2. (A) Superposition of the children’s trajectories (sacrum trajectories L5) of all trials for two representative control children (an MA-matched one and a CA-matched one) and one child with autism toward the positive and the aversive object in the straight ahead condition (left part) and 30◦ condition (right part). Note that children with autism showed a scattered and deviant trajectory especially in the aversive emotional context, while the goal was not reached. (B) Mean deviation between the real trajectory of the children and the theoretical trajectory (straight ahead and 30◦ conditions mixed up) in the positive emotional context (white bar chart) and the aversive (black bar chart). **p < 0.01; ***p < 0.001. S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 181 Fig. 3. (A) Superposition of the children’s trajectories (sacrum trajectories L5) of all trials for two representative control children (an MA-matched one and a CA-matched one) and one child with autism toward the positive and the aversive object in the “moving” condition. (B) Mean deviation between the children’s theoretical trajectory and their real trajectory after the object’s shift of the three groups of children in the positive and aversive emotional context. percentage correspond to 100% whatever the studied condition. We can first note that the mean percentages of trials achieved by children with autism strongly decreased when an aversive emotional valence was conferred to the object whatever its location and secondly that this decrease was accompanied by a strong increase of the variability of these percentages (bold, grayed out values). Indeed, the individual results indicate that if some children with autism reached the goal in all cases, others did not attain it systematically or never reached it. The ANOVA indicates a GROUP effect and a VALENCE effect on goal achievement (respectively F(2,26) = 8.9, p < 0.01, 2p = 0.41 and F(2,26) = 5.4, p < 0.05, 2p = 0.17) with a interaction between the GROUP and the VALENCE (F(2,26) = 5.8; p < 0.01, 2p = 0.31). The post test highlights first that children with autism less often reached the goal than the two groups of control children only when the emotional valence conferred to the object was negative whatever the object location (p < 0.01). It also underlines the fact that children with autism less often reached the goal when the emotional valence conferred to the object was negative than when it was positive (p < 0.001). 3.1.2. Trajectory in the “straight ahead” and “30◦ ” conditions The global locomotion trajectory is represented in Fig. 2A by the L5 trajectory. This figure shows an example of the global locomotion trajectory in two representative control children (a MA-matched one and a CA-matched one) and a child with autism who did not systematically reached the goal when the object was loaded with an aversive emotional valence. The ANOVA on the mean deviation between the children’s theoretical trajectory and their real trajectory highlights a DIRECTION effect (F(1,23) = 5.14; p < 0.05, 2p = 0.18), all the children deflecting their trajectory more in the “30◦ ” condition than in the “straight ahead” one. Moreover, as we can see in Fig. 2B, a GROUP and a VALENCE effects are found with an interaction between these two factors. According to the posttest we can note that these global statistical effects are only due to the children with autism, who deflected their trajectory more 182 S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 when the object to reach was associated with an aversive emotional valence. 3.1.3. Trajectory in the “moving” condition Fig. 3A shows an example of global locomotion trajectories in two representative control children (a MA-matched one and a CA-matched one) and in one child with autism who did not achieve the goal in the “moving” condition when the emotion was aversive. The trajectories of this child correspond to three representative behaviors observed in children with autism when locomotion was activated but did not reach its goal. Either the locomotion was stopped before reaching the goal or the trajectory was deflected in order to avoid the object. A combination of both could also occur. Fig. 3B shows the mean deviation between the children’s theoretical trajectory and their real trajectory after the object’s shift according to the global factors effects given by the statistical analysis. On this figure, it is visible that the trajectories of the children with autism were more deflected than those of the control children (MA-matched and CA-matched) in the two valence conditions (p < 0.001). We can also notice that their diversion was higher when the emotional valence conferred to the object was aversive with respect to a positive emotional valence (p < 0.001). 3.1.4. Last position of the children with respect to the object Fig. 4 shows the distance between all children groups and the object on the anteroposterior axis and on the lateral axis at the end of the trajectory (respectively Fig. 4A and B). Concerning the anteroposterior axis, the ANOVA shows a global GROUP effect (F(2,23) = 4.73; p < 0.05, 2p = 0.19), a DIRECTION effect (F(2,46) = 28.17; p < 0.001, 2p = 0.55) but no VALENCE effect despite a strong statistical tendency (F(1,23) = 4.15; p = 0.053, 2p = 0.15), with an interaction between the three factors (F(4,46) = 2.91; p < 0.05, 2p = 0.33). As it is visible on Fig. 4, this simple tendency can be explained by a variability of this last position which strongly increases in children with autism especially in an aversive emotional context. According to the post test, it is interesting to note that the statistical effects are only due to the last position of the children with autism in the “straight ahead and aversive” condition. This position was significantly more distant from the goal than in all the other conditions (p < 0.001) and also more distant from it than the last position of the controls in all the other conditions (p < 0.01). Concerning the lateral axis, the ANOVA indicates a global GROUP effect (F(2,23) = 11.44; p < 0.001, 2p = 0.50): children with autism stopped their movement laterally, far from the goal with respect to both control groups. Nevertheless a statistical tendency appears, concerning the VALENCE effect (F(1,23) = 3.23, p = 0.085, 2p = 0.12) and the interaction between GROUP and VALENCE (F(2,23) = 3.13, p = 0.062, 2p = 0.21). The post test indicates that this gap between children with autism and the object to reach was increased with respect to both control groups when the object was associated with an aversive emotional valence (p < 0.01). We can point out that the valence effect inside the group of children with autism is also close to the statistical significance level (p = 0.052). This is probably due to the important standard deviation which occurred when the object was associated with an aversive emotional valence. 3.2. Movement programming 3.2.1. Stride parameters Fig. 5 shows the parameters of the stride according to the global factors effects given by the ANOVA. According to the post test, children with autism performed a shorter stride than CAmatched controls (p < 0.001), a longer stride than MA-matched Fig. 4. Distance from the control children and children with autism to the object according to the experimental conditions in the antero-posterior axis (A) and the lateral axis (B) at the end of the trajectory. This value was obtained by measuring the distance between the object and the final position of the sacrum (L5) when the children stopped walking. controls (p < 0.001) and then displayed a weaker mean velocity than both control groups (respectively p < 0.01 and p < 0.05). Their stride amplitude decreasing in the “moving” condition with respect to the “30◦ ” condition (p < 0.05) appropriately explains the DIRECTION effect obtained in a global way. 3.2.2. Balance strategy Fig. 6 shows normalized anchoring indexes of the head, shoulder and hip for all children groups in each experimental condition according to the global factors effects given by the ANOVA. The post-test first allows us to say that the normalized anchoring index of the shoulder decreased for all children in the “moving” condition with respect to the “30◦ ” condition whatever the emotional valence conferred to the object. This indicates a less effective stabilization of the shoulder on space whenever the object was shifted. Secondly, the hip anchoring index of the children with autism was weaker than the one of the CA-matched controls, especially in the “30◦ ” and “moving” conditions (p < 0.01 and p < 0.05, respectively). This fact indicates a less effective stabilization of the hip on space in children with autism as soon as the object was not S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 183 Fig. 5. Amplitude (at the top), duration (at the center) and mean velocity (bottom part) of the stride according to the ANOVA (right part). Note that there is no VALENCE effect on all these parameters. located straight ahead anymore. Nevertheless, the head, shoulder and hip anchoring indexes remain significant and positive (p < 0.001) whatever the experimental condition in spite of these statistical effects. This indicates a stabilization of all the segments studied on space for all children in all conditions and so the use of an “articulated” balance strategy at the three segmental levels by all children. 4. Discussion The aim of the present study was to determine whether locomotion planning and/or locomotion programming are affected in children with autism by a positive or aversive emotion conferred to an object to fetch, according to its position in space. Our results concerning movement programming, i.e. gait parameters and balance strategy analysis highlight two main points. The first one is that there is no valence effect on these parameters. Then, movement programming being automatic, it would be independent from emotions. The second point is that some differences are observed between children with autism and the control groups, independently from the emotions’ effect. Indeed, children with autism produced a smaller and slower stride than CA-matched controls. These results support a number of authors who have described an “unusual gait” and “disturbances in walking” [23,25,27,46–51]. Concerning balance strategy, children with autism used the same “articulated” strategy as all control children, i.e. a stabilization of the head, shoulders and hip on space, as it is classically described in the literature for healthy children [52]. In agreement with our previous results [22,38], children with autism were thus able to coordinate both the propulsion of their body and their lateral stability in the same way as the control children. What was found was a less effective stabilization on the shoulders 184 S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 Fig. 6. Normalized anchoring index and standard error of the head (at the top), the shoulder (at the center) and the hip (at the bottom) in the positive (right part) and aversive (left part) emotional context, all directions mixed. We can note that – remaining positive in all cases for all children – the balance strategy used corresponded to an “articulated strategy”. in space when the object was shifted or a less effective stabilization on the hip in space as soon as the object was not located straight ahead. This did not endanger balance control during walking. All these results seem to indicate that the emotional system does not intervene directly at the level of locomotion programming. The main problem encountered comes from a slight dysfunction in the accurate control of the automatic locomotor program rather than from a disturbance of the motor program itself. This would especially come up when the object is not located straight ahead. Besides these slight locomotion programming disturbances, the main results observed in the present study concern locomotion planning. They are in line with the executive dysfunction found in autism [19,20,53–55] and with the results obtained by Maurer and Damasio [25] concerning plans or strategies for adapting to changing environmental contingencies. Nevertheless these results seem to indicate that the main deficit at the planning movement level appears through an impaired emotional system and more specifically through a deficit in the aversive emotion management. Let us consider goal achievement. Our results indicate that there is no difference between children with autism and the control children when the emotional valence conferred to the object is positive whatever the object location. The children with autism reached the goal of the action as often as control children. Then, children with autism were able to plan a voluntary movement according to an instruction by clearly establishing the goal of the action and making the decision to act independently from the object location. S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 Moreover in agreement with Baron-Cohen and Rieffe et al., they were able to feel positive emotions. On the other hand, an aversive emotional context produces a more varied range of behaviors. Some children with autism reached the goal but for some others the aversive emotional load blocked the movement completely, these children refusing to walk and to take the object. For the others the locomotion was activated but did not reach its goal. In this case, locomotion was stopped before reaching the goal and/or the trajectory was deflected in order to avoid the object. As a result, the mean percentage of trials achieved decreased sharply and became more variable for children with autism when an aversive emotional valence was conferred to the object. It is of interest to note that the children with Autism who did not reach the goal when the objects were aversive were also the children with the higher CARS scores. At the same time MA-matched control children always reached the goal of the task independently from the emotional valence conferred to the object. This result highlights the severity of the pathology as an indicator of goal achievement. This also showed some developmental impairment rather than a developmental delay. Then, the present study allows completing the results of Rieffe et al. [31]: children with autism felt more often aversive emotions than the children of the same biological age, but also than younger children too. Moreover, this allows generalizing the hypothesis according to which the impairment of the emotional system would produce an inability to interact in an emotional way with other people, see [37,56] for review. Indeed, the emotionrelated difficulties in autism would not merely be a facet of social impairments but in a more global way would produce an inability to interact with the environment, through a voluntary movement planning which would be not correctly triggered and/or organized in an aversive emotional context. The lack or the decrease of the motivation directed to the goal linked to an impaired management of aversive emotions could explain why the goal was not reached when an aversive emotional valence was conferred to the object. Now let us return to the trajectory analysis. The non-effect of a positive emotional valence on the body trajectory in “straight ahead” and “30◦ ” conditions indicates that the planning trajectory is correct in children with autism in a positive emotional context. Nevertheless, it is interesting to focus on the “moving” condition which involves a new movement planning “in line”. In this condition the trajectory of children with autism became deflected after the object shift with respect to all control children whatever the emotional valence conferred to the object. Then, a new planning “in line” seemed to be difficult for children with autism whatever the emotional context. Two explanations can be given. The first one is that their new trajectory planning process was prolonged with respect to the control children. Children with autism waited for a longer time before changing their trajectory, this delay moving their trajectory away from the new theoretical trajectory. The second explanation – which can be associated with the first one – is that the new trajectory was more irregular and variable than those of control children. In this case, the object shift would introduce some variability or some interference in the new trajectory planning “in line” independently from the emotional context. These explanations are in keeping with the hypothesis of Gepner and Feron [57] granting that dynamic sensory information are too fast to be perceived in real-time by people with autism, producing an impairment of their spatio-temporal processing and leading to a disorder of the understanding of language and emotions, a deficit of the imitative and executive functions and in particular a delay in the motor answers. On another hand, our results indicate that the trajectory’s diversion was higher in an aversive emotional context compared to a positive one in children with autism . This led us to think that the impairment of the spatio-temporal data processing would be increased in an aversive emotional 185 context and that emotions would contribute to maximize the smoothness of the trajectory in the frame of the hypothesis of kinematic parameters optimal control at the origin of trajectory planning [58]. Added to some spatio-temporal data processing impairment, this optimal control of the trajectory would be affected in children with autism too. Then, all these results seem to indicate that the emotional system would have two different functions according to two types of cognitive processes involved in movement planning. The first type would correspond to the cognitive processes allowing together the emergence of the act intent, the identification of the goal of the task and the decision-making to act. It would not be necessary to call for them once again when the object is moved during the locomotion. Indeed, the intent to act and the goal of the task remained unchanged, the decision to act being already made. It would be strongly linked to the emotional system which would trigger it. In this way, an aversive emotional context would be able to completely block these processes or disturb them enough to prevent goal achievement in children with autism independently from the object location. Our results indicating that children with autism refused to perform the task, stopping their walk before goal achievement or avoiding the object in an aversive emotional context but not in a positive one are consistent with this hypothesis. The second type would correspond to cognitive processes involved in the abstract trajectory planning linked to an estimate of the spatial and postural contexts on the environment and body states. When the object to reach is static, an aversive emotional context would interfere with these cognitive processes that would introduce some diversion and variability of the trajectory, more particularly when the object is located straight ahead. When the object is moved during locomotion, these processes would be involved in the new trajectory planning “in line” again and would involve a spatio-temporal processing of the dynamic sensory information. This data processing would be impaired in autism leading to a trajectory’s diversion whatever the emotional context. But an aversive emotional context would emphasize this impairment leading to a higher trajectory’s diversion compared to what happened with a positive one. In all cases, an aversive emotional context increases the trajectory’s diversion and/or its variability. This indicates that the function of the emotional system would be to modulate these cognitive processes in order to obtain a smooth and accurate trajectory. To conclude, the present study shows a close link between the sensori-motor system and the emotional system, emotions being considered as motivational states. Our results seem to emphasize a difficulty to manage aversive emotions in children with autism having two main effects on movement planning. The first one is a blocking or a disturbance of the cognitive processes at the origin of the emergence of the act intent, the identification of the goal of the task and the decision-making to act. The second one is a disturbance of the cognitive processes modulation leading to an accurate and smooth trajectory. But our results emphasize the correct behavior of these children in a positive emotional context too. The hospital staff who took part in this study were very surprised to see “their” children walking and reaching an object according to an instruction, changing their direction according to the object location and repeating this action several times. In this way, using an affectivo-motivational model based on the biphasic theory – the approach toward a pleasant stimulus or situation and the avoidance of an aversive – could improve our knowledge on the perceptual and motor skills in children with autism and explain some avoidances to interact observed in these children. The sensorimotor follow-up care for patients should take into account the good effect of positive emotion on the voluntary movement in order to encourage children with autism to interact more. S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 186 Conflict of interest statement All authors disclose any potential sources of conflict of interest. Acknowledgments We would particularly like to thank the French Charity, “Fondation de France” for its financial support and the hospital staff of the Théophile Roussel Hospital Center for their involvement in the study. References [1] Frijda NH. The emotions. New York: Cambridge University Press; 1986. [2] Luminet O. Psychologie des émotions: Confrontation et Evitement. Bruxelles: DeBoeck Université; 2002. [3] Kleinginna PR, Kleinginna AM. A categorized list of emotion definitions with suggestions for a consensual definition. Motivation and Emotion 1981;5:345–79. [4] Izard CE. Emotion theory and research: highlights, unanswered questions, and emerging issues. Annual Review of Psychology 2009;60:1–25. [5] Coquery JM, Marins D. Motivation. In: Le Ny JFMS, editor. Grand Dictionnaire de la Psychologie. Paris: Larousse; 1991. p. 480. [6] Elliot AJ, Thrash TM. Approach-avoidance motivation in personality: approach and avoidance temperaments and goals. Journal of Personality and Social Psychology 2002;82:804–18. [7] Lane RD, Reiman EM, Bradley MM, Lang PJ, Ahern GL, Davidson RJ, et al. Neuroanatomical correlates of pleasant and unpleasant emotion. Neuropsychologia 1997;35:1437–44. [8] Lang PJ, Bradley MM, Cuthbert BN. Motivated attention: affect, activation, and action. In: Lang Rfsmtb PJ, editor. Attention and orienting: sensory and motivational processes. Hillsdale, NJ: Erlbaum; 1997. p. 97–135. [9] Allen GI, Tsukahara N. Cerebrocerebellar communication systems. Physiological Reviews 1974;54:957–1006. [10] Paillard J. Les niveaux sensori-moteur et cognitif du contrôle de l’action. In: Laurent M, Therme P, editors. Recherches en Activités Physiques et Sportives. Aix-Marseille: Publication du Centre de Recherches de l’UEREPS, Université; 1985. p. 147–63. [11] Paillard J. Apraxia and the neurophysiology of motor control. Philosophical Transactions of the Royal Society of London, Series B: Biological Sciences 1982;298:111–34. [12] Bernstein N. The problem of the interrelation of co-ordination and localization. In: Whiting HTA, editor. Human motor actions Bernstein reassessed. Amsterdam: North-Holland; 1984. p. 77–119. [13] Stins JF, Beek PJ. Organization of voluntary stepping in response to emotioninducing pictures. Gait and Posture 2011;34:164–8. [14] Gelat T, Coudrat L, Le Pellec A. Gait initiation is affected during emotional conflict. Neuroscience Letters 2011;497:64–7. [15] Naugle KM, Joyner J, Hass CJ, Janelle CM. Emotional influences on locomotor behavior. Journal of Biomechanics 2010;43:3099–103. [16] Massion J. Cerveau et motricité. Fonctions sensori-motrices. Paris: Presses Universitaire de France; 1997. [17] Hugues C. L’autisme, cinquante années après Kanner. In: Aussilloux C, LivoisPetersen MF, editors. Théories cognitives de l’autisme. Ramonville; 1994. p. 33–42. [18] Ozonoff S. Executive functions in autism. In: Schopler E, Mesibov GB, editors. Learning and cognition in autism. New York: Plenum Press; 1995. [19] Russell J. Autism as an executive disorder. Oxford: Oxford University Press; 1997. [20] Russell J, Hill EL. Action-monitoring and intention reporting in children with autism. Journal of Child Psychology and Psychiatry and Allied Disciplines 2001;42:317–28. [21] Longuet S, Ferrel-Chapus C, Oreve MJ, Chamot JM, Vernazza-Martin S. Emotion, intent and voluntary movement in children with autism. An example: the goal directed locomotion. Journal of Autism and Developmental Disorders 2012;42:1446–58. [22] Vernazza-Martin S, Martin N, Vernazza A, Lepellec-Muller A, Rufo M, Massion J, et al. Goal directed locomotion and balance control in autistic children. Journal of Autism and Developmental Disorders 2005;35:91–102. [23] Damasio AR, Maurer RG. A neurological model for childhood autism. Archives of Neurology 1978;35:777–86. [24] Kohen-Raz R, Volkmar FR, Cohen DJ. Postural control in children with autism. Journal of Autism and Developmental Disorders 1992;22:419–32. [25] Maurer RG, Damasio AR. Childhood autism from the point of view of behavioral neurology. Journal of Autism and Developmental Disorders 1982;12:195– 205. [26] Nobile M, Perego P, Piccinini L, Mani E, Rossi A, Bellina M, et al. Further evidence of complex motor dysfunction in drug naive children with autism using automatic motion analysis of gait. Autism 2011;15: 263–83. [27] Vilensky JA, Damasio AR, Maurer RG. Gait disturbances in patients with autistic behavior: a preliminary study. Archives of Neurology 1981;38: 646–9. [28] Ben Shalom D, Mostofsky SH, Hazlett RL, Goldberg MC, Landa RJ, Faran Y, et al. Normal physiological emotions but differences in expression of conscious feelings in children with high-functioning autism. Journal of Autism and Developmental Disorders 2006;36:395–400. [29] Berthoz S, Hill EL. The validity of using self-reports to assess emotion regulation abilities in adults with autism spectrum disorder. European Psychiatry 2005;20:291–8. [30] Hill E, Berthoz S, Frith U. Brief report: cognitive processing of own emotions in individuals with autistic spectrum disorder and in their relatives. Journal of Autism and Developmental Disorders 2004;34:229–35. [31] Rieffe C, Meerum Terwogt M, Kotronopoulou K. Awareness of single and multiple emotions in high-functioning children with autism. Journal of Autism and Developmental Disorders 2007;37:455–65. [32] Rieffe C, Meerum Terwogt M, Stockmann L. Understanding atypical emotions among children with autism. Journal of Autism and Developmental Disorders 2000;30:195–203. [33] Baron-Cohen S. Do people with autism understand what causes emotion? Child Development 1991;62:385–95. [34] Loveland KA. Toward an eco-logical theory of autism. In: Burack JATC, Yirmiya N, Zelazo PR, editors. The development of autism: perspectives from theory and research. Mahwah, NJ: Lawrence Erlbaum Associates; 2001. p. 17–37. [35] Loveland KA. Social-emotional impairment and self-regulation in autism spectrum disorders. In: Nadel J, Muir D, editors. Emotional development. Oxford: Oxford University Press; 2005. p. 365–82. [36] Hobson RP. The autistic child’s appraisal of expressions of emotion: a further study. Journal of Child Psychology and Psychiatry and Allied Disciplines 1986;27:671–80. [37] Hobson RP. The autistic child’s appraisal of expressions of emotion. Journal of Child Psychology and Psychiatry and Allied Disciplines 1986;27: 321–42. [38] Longuet S, Ferrel-Chapus C, Oreve MJ, Chamot JM, Vernazza-Martin S. Emotion, intent and voluntary movement in children with autism. An example: the goal directed locomotion. Journal of Autism and Developmental Disorders 2012;42(7):1446–58. [39] Schopler E, Reichler RJ, DeVellis RF, Daly K. Toward objective classification of childhood autism: Childhood Autism Rating Scale (CARS). Journal of Autism and Developmental Disorders 1980;10:91–103. [40] Schopler E, Reichler RJ. Le profil psycho-Educatif. Masson; 1988 [Paris]. [41] Andrews FM, Withey SB. Social indicators of well-being: Americans’ perceptions of life quality. Plenum Press; 1976 [New York]. [42] Ferrel C, Bard C, Fleury M. Coordination in childhood: modifications of visuomotor representations in 6- to 11-year-old children. Experimental Brain Research 2001;138:313–21. [43] Assaiante C. Development of locomotor balance control in healthy children. Neuroscience and Biobehavioral Reviews 1998;22:527–32. [44] Assaiante C, Amblard B. Ontogenesis of head stabilization in space during locomotion in children: influence of visual cues. Experimental Brain Research 1993;93:499–515. [45] Cohen J. Statistical power analysis for the behavior sciences. 2nd ed. New Jersey: Hillsdale; 1988. [46] Teitelbaum P, Teitelbaum O, Nye J, Fryman J, Maurer RG. Movement analysis in infancy may be useful for early diagnosis of autism. Proceedings of the National Academy of Sciences of the United States of America 1998;95: 13982–7. [47] Jeste SS. The neurology of autism spectrum disorders. Current Opinion in Neurology 2011;24:132–9. [48] Cattaneo L, Fabbri-Destro M, Boria S, Pieraccini C, Monti A, Cossu G, et al. Impairment of actions chains in autism and its possible role in intention understanding. Proceedings of the National Academy of Sciences of the United States of America 2007;104:17825–30. [49] Nayate A, Bradshaw JL, Rinehart NJ. Autism and Asperger’s disorder: are they movement disorders involving the cerebellum and/or basal ganglia? Brain Research Bulletin 2005;67:327–34. [50] Rinehart NJ, Tonge BJ, Iansek R, McGinley J, Brereton AV, Enticott PG, et al. Gait function in newly diagnosed children with autism: cerebellar and basal ganglia related motor disorder. Developmental Medicine and Child Neurology 2006;48:819–24. [51] Jansiewicz EM, Goldberg MC, Newschaffer CJ, Denckla MB, Landa R, Mostofsky SH. Motor signs distinguish children with high functioning autism and Asperger’s syndrome from controls. Journal of Autism and Developmental Disorders 2006;36:613–21. [52] Assaiante C, Mallau S, Viel S, Jover M, Schmitz C. Development of postural control in healthy children: a functional approach. Neural Plasticity 2005;12:109–18 [discussion 263–72]. [53] Plumet MH, Hugues C, Tardif C, Mouren-Siméoni MC. L’hypothèse d’un déficit des fonctions exécutives dans l’autisme. In: Nadel J, Rogé B, editors. Psychologie française. Autisme: l’option biologique 1 Recherche, Presses universitaires de Grenoble; 1998. p. 157–67. [54] Ozonoff S, Pennington BF, Rogers SJ. Executive function deficits in high-functioning autistic individuals: relationship to theory of mind. Journal of Child Psychology and Psychiatry and Allied Disciplines 1991;32: 1081–105. S. Vernazza-Martin et al. / Behavioural Brain Research 251 (2013) 176–187 [55] Hill EL. Executive dysfunction in autism. Trends in Cognitive Sciences 2004;8:26–32. [56] Gaigg SB. The interplay between emotion and cognition in autism spectrum disorder: implications for developmental theory. Frontiers in Integrative Neuroscience 2012;6:113. 187 [57] Gepner B, Feron F. Autism: a world changing too fast for a mis-wired brain? Neuroscience and Biobehavioral Reviews 2009;33:1227–42. [58] Hogan N. An organizing principle for a class of voluntary movements. Journal of Neuroscience 1984;4:2745–54.