Developmental Coordination Disorder (DCD) is a lifelong neurodevelopmental disorder affecting the execution and learning of motor skills to the extent that it severely impacts daily functioning and academic performance. DCD occurs in around 5–6% of school-aged children (American Psychiatric Association, 2022), and is often accompanied by secondary consequences including lower physical activity levels and participation, higher susceptibility to obesity and psychosocial difficulties such as depression or anxiety (Izadi-Najafabadi et al., 2019; Missiuna et al., 2014; Zwicker et al., 2013). In addition, other developmental disorders like Attention Deficit Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD) often co-occur with DCD, complicating both its phenotype and aetiology. So far, the underlying mechanisms of DCD are not well-known, and consequently, current interventions to improve motor learning do not target the mechanisms that contribute to the disorder. More insight into these mechanisms is urgently needed to optimize rehabilitation and therapy for this population.

Optimal motor control and learning requires strong connections between perception and action (Shadmehr et al., 2010; Wolpert, 1997). In cognitive psychology, ideomotor theory states that every motor action is represented by the sensory feedback that it produces, or by its predicted sensory consequences (Prinz, 1997; Shin et al., 2010). Through associative sequence learning, or the repeated experience of performing a motor action and perceiving its sensory consequences, this perception-action link is strengthened (Catmur et al., 2009), and observing someone else's movement automatically triggers an internal motor representation of that movement (Blakemore & Frith, 2005). As a result, we have the tendency to imitate movements that we observe, even if these do not contribute to the task performance, an effect known as ‘automatic imitation’ (Cracco et al., 2018). The externally triggered motor representation of an observed movement can conflict, or compete with, self-generated motor representations during actual movement execution. As a result, automatic imitation has either a facilitating or interfering effect of movement observation on a simultaneously executed movement. This effect is usually measured using stimulus-response compatibility tasks, in which participants are instructed to perform a specific movement (e.g., lifting their index finger) while observing a model performing either the same movement (congruent condition) or a different movement (e.g., lifting their middle finger, incongruent condition) (Brass et al., 2000). In the incongruent condition, when there is a discrepancy between the observed and the executed action, reaction times are typically longer, due to conflicting internal representations of observed and executed movements. This led to the suggestion that automatic imitation can be seen as an index of how strongly perception and action are linked in the brain (Brass & Heyes, 2005). This dynamic is, however, subject to sensorimotor experience and can be attenuated after a short period of training, where movements are paired with observations of movements that do not correspond to those executed (Heyes et al., 2005; Wiggett et al., 2011). This suggests that, in cases like DCD, the motor control and learning deficiencies may manifest in disrupted automatic imitation tendencies.

Indirect evidence supports the idea of reduced automatic imitation effects in individuals with DCD. For example, previous studies suggest an under-activation of the inferior frontal gyrus, during both the execution and the observation of manual fine-motor tasks in children with DCD (Licari et al., 2015; Reynolds, Licari, et al., 2015; Zwicker et al., 2010). Notably, this brain region is a key area in the mirror neuron system, which is thought to underly the automatic imitation effect (Catmur et al., 2009). Also, reduced mu rhythm desynchronization, a proxy of mirror neuron system activation, has been measured during both the execution and the observation of manual motor tasks (Keating et al., 2023; Lust et al., 2019). However, mu (de)synchronization is a debated index for automatic imitation due to significant methodological issues mentioned in the literature (Hobson & Bishop, 2017), challenging the reliability of the findings. Also, while mu suppression is often interpreted as reflecting mirror neuron system activity, it has not been directly assessed in tasks specifically measuring automatic imitation. Taken together, although DCD has been linked to possible mirror neuron system alterations, existing neurophysiological findings remain inconclusive (Reynolds et al., 2015b, 2019), and there is no direct evidence testing automatic imitation in children with DCD.

Interestingly, the phenomenon of automatic imitation, where observing a motor action triggers the performance of that same action, can also serve as a test of the internal modelling deficit hypothesis. This hypothesis has been proposed as a key mechanism underlying motor control difficulties in DCD (Wilson et al., 2017). In a systematic review by Adams et al. (2014), many motor control impairments observed in DCD were linked to deficits in either forward or inverse modelling. Forward modelling involves predicting the sensory outcomes of a movement, while inverse modelling pertains to predicting the motor commands needed to produce a desired action (Wolpert & Kawato, 1998). Effective motor control relies on the integration of both models, and numerous studies in the literature report disruptions to one or both processes in DCD. For example, in reaching tasks, children with DCD appear to have marked difficulty to make rapid online adjustments based on changes in sensory feedback (Hyde & Wilson, 2011a, 2011b), and show marked delay in speeded sequential movements, accompanied by a disruption in eye-hand coordination (Wilmut et al., 2006). These tasks are thought to rely on predictive internal models of action (i.e., efference copy), as the speeded and sequential movements do not allow for use of visual feedback available after or during movement execution, but require predictive control based on an internal representation. Moreover, in tasks that require mental motor representation (i.e., motor imagery), such as hand rotation tasks, children with DCD appear to show broad reductions in performance (Barhoun et al., 2019; Deconinck et al., 2009). Because children with DCD show clear deficits in tasks that rely on predictive control using internal motor representations, the assumption is that these representations are less well-defined, noisy, less accurate, or less accessible (Adams et al., 2014). However, despite these indications, direct tests of internal motor representations in children with DCD are lacking. Therefore, we will study to what extent the inability to internally represent motor action based on observation may underpin deficits in internal modelling.

Despite the increase in studies including (f)MRI and EEG measurements, our understanding of the neural processes underlying motor behaviour in DCD is scarce and often based on small sample sizes (Biotteau et al., 2016). Also, many behavioural studies in DCD suggest that a certain level of behavioural complexity is needed to detect the motor difficulties when compared to typically developing children (Adams et al., 2014; Bhoyroo et al., 2020). Experiments that include both behavioural and neural measures possibly allow for a more sensitive measure of these differences, perhaps also in less complex movements, and will provide more insight into the underlying control mechanisms contributing the observed behaviour. In addition, EEG event-related potential (ERP) analysis provides the unique advantage of capturing real-time high temporal resolution brain mechanisms during the execution of an automatic imitation task, effectively combining neural and behavioural measurements.

Previous research has exposed the neural processes contributing to automatic imitation in stimulus-response compatibility tasks. Using EEG, Deschrijver et al. (2017a) were able to temporally dissociate low-level perceptual and high-level cognitive mechanisms contributing to the effect. A first ERP component that was found was the N190 which is related to the visual processing of body parts (Czekóová et al., 2022; Deschrijver et al., 2017a; Rauchbauer et al., 2018). It was found that incongruent trials elicit larger N190-amplitudes, suggesting that the early visual processing of observed actions is influenced by the motor preparation of the intended action. However, this effect could not be replicated in a follow-up study (Deschrijver et al., 2017b). In addition, the centroparietal late P3 amplitude was investigated and found to be larger for congruent than incongruent trials (Deschrijver et al., 2017a; Deschrijver et al., 2017b). Because the P3 has consistently been associated with self-versus-other related processes and distinguishes the representation of the observed action from that of the intended action (Czekóová et al., 2022; Ferguson et al., 2018; Knyazev, 2013; Vastano et al., 2020), this finding has been interpreted as a stronger need for self-other distinction processes when the observed action resembles the intended action (Deschrijver et al., 2017a; Deschrijver et al., 2017b). Finally, an influence of the observed action on motor preparation was found with the movement-related Readiness Potential (RP), which is a slow negative component observed prior to action (Deschrijver et al., 2017a; Leuthold & Schröter, 2011). Generally, RP is believed to reflect motor preparation, the readiness to execute a motor response (Shibasaki & Hallett, 2006). In automatic imitation, results on RP indicated smaller amplitudes in congruent trials, suggesting that the preparation of own actions was facilitated when the observed hand movement matched the intended movement. Also, in a similar paradigm, Rauchbauer et al. (2018) measured larger amplitudes in congruent trials on the pre-movement positivity, another movement-related ERP component, which is believed to reflect the motor command for movement initiation. Therefore, both motor preparation (RP) and movement initiation (pre-movement positivity) are facilitated by stimulus congruency.

However, it is important to recognize how developmental and individual differences influence these neural patterns. Movement-related ERPs, among which the early slow RP, that resembles the RP that is reported in adults in (Deschrijver et al. 2017a; Deschrijver et al.2017b) show remarkable variations across age and in the context of developmental disorders. For instance, while adults typically demonstrate a negative early slow RP, with central topography, studies in children have sometimes found opposite polarity in self-paced thumb movement tasks (Jarczok et al., 2019; Warren & Karrer, 1984). In a discrete aiming task, Pangelinan et al. (2011) found that this component showed a reduced amplitude in children. Moreover, altered RP patterns have been observed in children with ADHD and adults with high-functioning autism (Deschrijver et al., 2017b; Jarckzok et al., 2019). For example, Deschrijver et al. (2017b) found larger early RP in adults with high-functioning autism in both the congruent and incongruent condition of an automatic imitation task, suggesting difficulties in motor preparation regardless of the simultaneously observed action. In addition, Jarckzok et al. (2019) found amplitudes with positive polarity on the early RP in children with ADHD in a self-paced thumb movement task, and the amplitudes were significantly lower compared to a neurotypical control group. These developmental and disorder-related ERP differences, expressed in both positive and negative polarity in RP, suggest that altered neural mechanisms in action and perception may underlie some of the challenges seen in children with developmental disorders. How these findings translate to motor difficulties as seen in children with DCD has yet to be fully explored.

In conclusion, automatic imitation may give us a useful measure to test internal motor representation in children with DCD, and allows for a detailed mapping and comparison of the underlying neural processes that contribute to these behaviours. Therefore, this study aims to quantify and compare the automatic imitation effect between children with and without DCD. The focus is on children, rather than adults, to better understand the underlying mechanisms of automatic imitation during a crucial developmental window. Automatic imitation is related to social functioning, motor learning, and broader behavioural development. Identification of atypical patterns is important during developmental years, as it may help inform timely interventions before secondary difficulties emerge. While resting-state EEG recordings can be successfully implemented in very young children, performing a behavioural task involves an addition set of abilities (e.g., instruction comprehension, sustained attention, and general task engagement and compliance) that develop throughout childhood. Therefore, and considering the availability of participants of these ages for recruitment, the study focused on children above the age of 9. As sample size is a priority and recruitment of children with DCD is challenging, the age range was extended to include participants up to 16 years old. Due to the proposed problems with internal motor representations in DCD, a reduced automatic imitation effect is expected in this group based on reaction time measurements. In addition, using EEG, this study aims to provide more insight into the neural underpinnings of this difference. In that regard, it is expected that reductions in internal motor representation of both the executed and observed action in DCD would diminish the need for self-versus-other processing, leading to a reduced P3 amplitude in this group. No differences depending on congruency are anticipated related to P3. Children with DCD are not expected to have difficulty in early visual processing, therefore no alterations in N190 are anticipated. Finally, we expect RP to reflect differences in motor preparation between our groups, related to the motor difficulties in DCD.