1.1. Developmental coordination disorder

Developmental coordination disorder (DCD) is a neurodevelopmental disorder that affects motor skills development in children and influences their ability to perform multiple activities of daily living [1, 2]. According to the Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition [3], DCD is characterized by difficulties in movement acquisition and execution. Diagnostic criteria include motor abilities that are significantly inferior to those expected given the individual chronological age and learning opportunities, and impairments significantly interfering with daily activities, academic and/or professional accomplishments and hobbies, which emerge during a child’s early development and are not better explained by intellectual disability, neurological condition affecting movement or visual impairment. Motor difficulties can be manifested as clumsiness, slowness and inaccuracy in movement execution.

Significant signs of DCD usually emerge during school years, but motor difficulties associated with DCD are developmental and tend to persist into adolescence [4] and adulthood [5]. Its prevalence is estimated between 5 and 6% in children aged 5 to 11 years old [3]. More boys than girls are affected, with reported ratios of men to women ranging from 1.9:1 to 7.3:1 [6, 7]. Great variability exists within countries, with prevalence among school-age children ranging from 1.8% in the United-Kingdom [7], 8% in Canada and 19% in Greece [8]. DCD is present across all cultural, ethnic and socioeconomic groups, but daily living activities and functional impacts of the disorder may vary within these different groups. Thus, these variables should be considered at the time of diagnosis.

DCD has been formerly referred to as clumsy child syndrome, and today the terms specific developmental disorder of motor function and dyspraxia can also be used to describe DCD [3, 9]. The terms DCD and dyspraxia (or developmental dyspraxia) are the most frequently used terms in the literature. DCD focusses on the observable symptoms, on their functional impacts and manifestations in daily living, and its diagnosis often belongs to occupational therapists or to physiotherapists [10, 11]. As for dyspraxia, the term is used interchangeably with DCD, and is widely used and diagnosed in neuropsychology [10–13]. Dyspraxia is often defined as difficulties with organizing, planning, executing and coordinating movement, which leads to impairments in acquisition of complex movements and of movement sequences [10, 12, 14]. It relies on a developmental conception of the brain, implying a motor cognition disorder that includes visual and spatial processing problems [11]. Children with dyspraxia and DCD typically show visuoperceptual, visuospatial, visuoconstructives and visuomotor deficits [12, 15, 16], further impacting school achievement and daily living.

Several theoretical models of dyspraxia have been proposed over the years to provide a coherent view of its core elements. The one proposed by Vaivre-Douret and her team [13] offers a thorough definition as it stipulates that dyspraxia involves a deficit of both execution of voluntary gesture and planning/programming of movement. According to this model, when a child intends to perform an action, he first has to formulate a plan to execute that movement, that considers perceptual information, then to build a mental representation of the movement by coding spatiotemporal parameters, and finally to execute the movement while correcting it according to sensorimotor feedback [13, 14]. By conducting neuropsychological, neuro-psychomotor and neuro-visual assessments in children with dyspraxia, the authors showed that the planning and programming processes were the core problems in this disorder, and that the execution mechanisms were disturbed only when dyspraxia was comorbid with other neuropsychological disorders [13]. Therefore, the question arises as to whether children with dyspraxia (or DCD) present only planning impairments on motor tasks, or broader planning and executive functioning deficits.

1.2. Neurodevelopmental comorbidities

It seems that children who only have DCD are the exception rather than the rule, comorbid disorders being present in most of the cases [17–19]. Attention deficit/hyperactivity disorder (ADHD) is the most frequent comorbid condition, occurring approximately in 50% of the cases [3, 4, 6, 20, 21]. Besides, some authors report that DCD and ADHD overlap in their symptoms, as children with ADHD frequently demonstrate motor difficulties [22], and that executive dysfunctions and slow processing speed, which are typical features of ADHD, are often present in children with DCD [20]. It is thus unclear which difficulties are inherent to which disorder, but ADHD and DCD must still be considered as separate disorders since their core deficits are distinct [23]. Learning disorders, including developmental dyslexia (DDL) [24, 25], specific language impairment (SLI) [17, 26, 27], behavioral problems [17, 28] and autism spectrum disorder (ASD) [6, 29] are also common in children with DCD, concomitance between these disorders and DCD reaching up to 30 to 50%. Since deficits in executive and attentional functioning are generally part of these disorders [30–32], it seems essential to consider their presence and potential influence on executive and attentional capacities in children with DCD.

1.3. Executive and attentional functions

Executive functions are generally conceptualized as a set of general high order control processes [33] working together [34] to direct and manage cognitive, emotional and behavioral functions, especially during active problem solving [35]. A variety of cognitive abilities falls under the umbrella of executive functions, including working memory, inhibition, cognitive flexibility/shifting, goal-setting, planning, organization, self-regulation and fluency [33–37]. To this day, however, it is still unclear exactly how many components define executive functions and under which terms they should be grouped [38]. Indeed, several models of executive functions have been proposed, integrating different components and establishing interactions between them. One of the most integrative and complete models was proposed by Diamond [36]. According to this model, working memory is defined as the ability to hold information in mind for a short period of time while mentally manipulating it to execute a task, and it includes verbal and visuospatial subcomponents. Inhibitory control is described as the ability to control attention, behavior, thoughts and emotions to override a dominant, automatic or prepotent response. The author also explains that inhibitory control is divided into two subcomponents: interference control, combining cognitive and attentional inhibition, and response inhibition, defined as behavioral inhibition. It also encompasses self-regulation, which includes attentional and response inhibition, while focusing primarily on emotional control and regulation. According to Diamond [36], working memory and inhibitory control support each other: working memory allows one to maintain their goals in their mind along with what they should or should not do, and inhibitory control allows one to stay focused on the important working memory content by inhibiting distractors. When working together, these two executive functions allow cognitive flexibility, defined as one’s ability to see things from another perspective and to shift between tasks. According to this model, fluency skills are part of cognitive flexibility, since one must be able to shift between different mind sets to be fluent in generating various ideas. Higher-level executive functions are also underpinned by working memory, inhibitory control and cognitive flexibility in this model: reasoning and problem-solving, which are synonymous with fluid intelligence, and planning skills [36].

Attentional functions can be divided into three central components [39, 40]: alertness, selective attention, and divided attention. Alertness allows individuals to adopt and sustain a state of vigilance and is typically involved in long, monotonous tasks. It contrasts with the second component, selective attention, which is the ability to selectively process a specific kind of information while ignoring distractors and involves filtering mechanisms. The third subfunction of attention is divided attention and concerns the notion of sharing mental processing capacities, which appears necessary when performing two or more tasks simultaneously [39, 40]. These attentional functions are also closely related to executive functions, particularly to inhibitory control and working memory; inhibition capacities allow attentional orienting and control [36], while the allocation of attentional resources to specific information stream will dictate the ones accessing working memory [41].

Executive dysfunction leads to resistance to change, incapacity to modify non-optimal behaviors, impulsivity and risk-taking, social difficulties, avolition and learning difficulties [34, 36, 42]. For this reason, executive functions are essential to children’s optimal daily functioning and their future quality of life, in their affective, social as well as academic spheres [11, 36].

1.4. State of knowledge

Several studies have reported impairments in executive and attentional functions in children with DCD [21, 43–45]. While some authors report a generalized executive dysfunction in this population [20], others report deficits in specific domains of attentional or executive functioning [45, 46], which can be more severe in nonverbal modalities [47]. Consequently, and according to the theoretical model of dyspraxia previously discussed, the question arises as to whether the impairments are only present in visuospatial/nonverbal modalities and on planning tasks, and thus are more linked to the primary deficits found in DCD, or whether difficulties can be found across a broader range of executive functions. This systematic review of the literature aims to answer this question, while considering the influence of comorbid neurodevelopmental disorders combined with DCD on attentional and executive profiles found in these children. This review is an important step in the acquisition and gathering of supplemental evidence regarding the attentional and executive deficits found in DCD. Indeed, consolidating the results found in previous research appears relevant, especially given that several studies based their conclusions on small samples.

On a clinical level, the purpose of this review is to shed light on impairments that can be a part of DCD rather than being explained by the comorbid disorders often observed in this disorder, and to better identify the needs of children with DCD considering their potentially complex clinical picture, in order to optimize their assessment and the interventions carried out. This systematic review is the first to focus exclusively on the cognitive profile associated with DCD, and more specifically on the attentional and executive functions in young individuals with this disorder, while explicitly considering the possible influence of co-occurring disorders.

2.1. Search strategy and selection criteria

A systematic literature research was performed in PubMed/Medline and PsycINFO databases, including articles published between January 1980 and May 2020. Considering that pediatrics neuropsychology was poorly developed before 1980, it was chosen as the earliest publication date. Due to the multiple ways of conceptualizing attentional and executive functions in terms of their components, and to be as inclusive as possible, our research was made using all terms previously mentioned that are used to describe these functions. Thereby, research was conducted in English with the following keywords: (1) “developmental coordination disorder” OR “dyspraxi*” OR “motor skills disorder” OR “specific developmental disorder of motor function” OR “clumsiness” OR “clumsy child syndrome”; (2) AND “executive function*” OR “goal setting” OR “set-shifting” OR “shifting” OR “switching” OR “flexibility” OR “planning” OR “inhibit*” OR “working memory” OR “organis*” OR “organiz*” OR “self-regulation” OR “fluency” OR “attention*”; (3) AND “child*” OR “adolescen*” OR “teen*” OR “youth” OR “schoolchild*” OR “preschool*”. Publications referenced in the included articles were also screened to find additional articles.

Studies were included if (1) their participants were children or adolescents (17 years of age or younger; studies including older participants were excluded), (2) they had a group of participants with a diagnosis of DCD made by a health care professional, by a score at or below the 5^th percentile on the Movement Assessment Battery for Children (MABC first or second edition), as this score indicates a significant movement difficulty [49], or by DSM (IV, IV-TR or 5) criteria for DCD combined with a movement ability measure, in which case a total score at or below the 15^th percentile on the MABC was accepted, (3) their participants did not explicitly have any medical condition that could affect their motor or cognitive abilities, (4) they measured one or more attentional or executive functions using performance tests, (5) they used normative data of standardized measures or a control group comprising healthy individuals for results comparison, (6) they were published in French or English in a peer-reviewed journal and (7) they had an empirical research design.

Studies were first selected according to their title. Second, abstracts were read by two authors (CL and MPL) and studies that did not meet the eligibility criteria listed above were excluded. Then the same two authors screened full articles independently to ensure that all eligibility criteria were met. Their disagreements were discussed to reach a consensus and, whenever necessary, another author (SL) settled. The remaining articles were entirely read by the first author.

2.2. Data extraction

Data extraction from articles that met the selection criteria was made by the two first authors. Information was organized in an spreadsheet and included: title of the article, authors, year of publication, journal in which it was published, aims of the study, groups, their origin and samples size, gender and age of the subjects assessed (mean, standard deviation and range, when available), country in which the study took place, inclusion and exclusion criteria, presence of comorbid disorders in the sample and their nature, information about assessment of motor functions and confounding variables, cognitive functions assessed and tasks used, statistical analysis, results, limitations and commentaries about the paper. Subsequently, the relevant information was analyzed and summarized in Table 1: the authors and year of publication, the sample size, the gender of the participants and the number of males, the mean age of the participants, the presence of comorbid disorders, the attentional or executive function studied, the task(s) used to evaluate the function, and a summary of the results. Components of attentional and/or executive functioning assessed in each included study were determined according to what the study purported to measure. When the same score on a task was reported to assess more than one component of attentional or executive functioning, or to measure different components in different studies, results were reported only in relation to the component it was the most associated with, according to the Compendium of Neuropsychological Tests [50] or the task’s reference. When several scores were available on the same task and associated with different components of attentional or executive functioning, information provided by each score was considered as a measure of its respective component.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Studies included in the systematic review.

https://doi.org/10.1371/journal.pone.0252043.t001

2.3. Quality assessment of included studies

Quality assessment of studies included in this review was conducted using a checklist we developed based on the Newcastle-Ottawa Quality Assessment Scale (NOS) [51]. The NOS consists of three domains: “selection of subjects”, “comparability of subjects” and “outcome”, each domain including two to three items. Although the use of the NOS has been established in meta-analyses [52], it was first developed for case-control and cohort studies. Therefore, its use was not suitable for the present study. Since we could not find a tool giving standardized criteria for assessing the quality of neuropsychological and behavioral studies, we adapted the NOS based on the methods used by Wu et al. [53–55] and Caçola et al. [56], that were inspired by the NOS and the PRISMA standards. We developed three to four quality items for each domain of the NOS (selection, comparability and outcome). The result is a 10-item checklist, including: inclusion/exclusion criteria and samples source (for the selection domain), comparability of samples regarding age, gender and IQ (for the comparability domain), and description of outcome measures, adequacy of outcome analysis and discussion (for the outcome domain; see Appendix 1 for the complete checklist). Items could be answered by “yes” or “no”, and quality level of evidence was rated as high (8 “yes” or more), medium (6–7 “yes”) or low (5 “yes” or less). The quality assessment was carried out by the first two authors independently and any discrepancy was discussed until they reached a decision by consensus.

3.1. Characteristics of included studies

A total of 1088 articles were identified through databases; 708 remained after removing duplicates. Of these, twenty articles (2.8%) had to be discussed between two authors to reach a consensus regarding their inclusion or exclusion, and a third had to settle for five (0.7%) of them. As a result, 41 studies were included in this systematic literature review (see flow diagram in Fig 1), representing 1,097 children and adolescents with DCD (or developmental dyspraxia; DD) with or without comorbid disorders, 16 with dyspraxia following preterm birth, 106 with motor difficulties (MD) without DCD, 51 comparison subjects with neurodevelopmental conditions other than DCD, and 1,213 typically developing (TD) controls, for a total of 2,766 subjects. MD groups were composed of children that had been identified with motor difficulties by scoring below the 16^th percentile on the MABC-2 [47, 57–59] or had a standardized score under 80 on the McCarron Assessment of Neuromuscular Developmental [60], without having a diagnosis of DCD [61]. Some articles used the same tasks in the same sample of participants with DCD, and therefore, whether it was mentioned in the article [47, 58] or obvious (common authors, close in time, same sample size, characteristics and inclusion/exclusion criteria) [62–64], their samples were considered as one and their participants were counted only once in the reported total participants. The age range for all 41 studies combined is 3 to 17 years old and the mean is 9.46 years old. Results and characteristics of studies included in this review are presented in Table 1. The quality ratings ranged between 5 and 10 out of 10. Table 2 provides the details regarding quality assessment of all included studies. Seven studies (17.1%) needed to be discussed between two authors to reach a consensus about their quality level of evidence. It was finally rated as high in 18 studies, medium in 21 studies and low in two studies. The two studies that were rated as low quality were nevertheless included, because they still met the eligibility criteria established by authors. However, they are identified in the sections in which they are discussed in order to nuance their contribution.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flow chart illustrating articles identification and selection process.

https://doi.org/10.1371/journal.pone.0252043.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Quality assessment of studies included in the systematic review.

https://doi.org/10.1371/journal.pone.0252043.t002

No study explicitly assessing organization, self-regulation or goal-setting processes and meeting eligibility criteria established in this review was found. Overall, included studies explored nine dimensions of cognition: three attentional functions (alertness and sustained attention, selective attention, and divided attention) and six executive functions (inhibitory control, working memory, planning, cognitive flexibility, fluency and general executive functioning). A study was considered to discuss general executive functioning when the task they used intentionally assessed multiple executive components, by using an ecological task or a standardized task that did not allow isolating one component of executive functioning. Results are presented in terms of these nine cognitive domains. When several components of attentional and executive functioning were assessed in the same study, the results for each component were reported in their respective section. When it was mentioned in the study, the different modalities (auditory/verbal or visual/nonverbal) used to measure each cognitive function are discussed separately (when available, the modality is identified in Table 1). When it was not possible to identify the modalities, the general cognitive function of interest is discussed. An exhaustive summary of attentional and executive functions assessed and their modalities, tasks used, sample size and age group in which they were respectively assessed, along with studies providing results about each cognitive domain, is provided in Appendix 2.

3.2. Attentional functions

3.3. Executive functions

3.4. Developmental considerations

The effect of age on cognitive functions has been studied in visual sustained attention, response inhibition, working memory, planning, cognitive flexibility and fluency abilities.

Regarding visual sustained attention, in a cross-sectional study, de Castelnau et al. [70] found that this capacity improved with age in both their sample of children with DCD and their TD group, but the discrepancy between the two groups remained, from 8–9 years old to 12–13 years old. These results suggest that difficulties in sustained attention persist with age in children with DCD.

Four studies considered the effect of years passing on response inhibition capacities. Three studies using a cross-sectional design found that nonverbal response inhibition capacities were better in older participants with DCD than in younger children with the disorder [70, 77, 78], so that these capacities in children with DCD approached those of TD children as they got older [77, 78]. The other study, using a longitudinal design, found no improvement in nonverbal response inhibition in a DCD sample after a two-year follow-up, and the differences originally found between their DCD group and their TD group persisted two years later [57]. These results indicate that the impairment found in nonverbal response inhibition in children might be due to a delay in the development of this cognitive function rather than a primary deficit, but given the heterogeneity of results across studies, more research is necessary to confirm this hypothesis. Regarding verbal inhibition capacities, Bernardi et al. [57] reported no significant change after two years in their DCD sample, although there was no deficit in DCD at both time points.

Additionally, among the functions that they found to be impaired in children with DCD in their longitudinal study, Bernardi et al. [57] reported a significant improvement of visuospatial working memory, nonverbal planning, nonverbal cognitive flexibility and nonverbal and verbal fluency over time in children with DCD. They also mention that improvements in their TD group were similar, so that the deficit found in these domains in children with DCD, in comparison to TD children, persisted after a two-year follow-up [57].

In summary, attentional and executive functions seem to improve with age in children with DCD, as well as in TD controls, but studies found that discrepancies between performance of children with DCD and that of TD children persist over time on most executive tasks.

4.1. Summary of findings

The 41 articles included in this systematic review, totalizing 1,097 children and adolescents with DCD, repeatedly reported significant impairment of nonverbal response inhibition, attentional inhibition, visuospatial and verbal working memory, planning, nonverbal fluency and general executive functioning in children with DCD. Studies assessing cognitive flexibility are divided on the presence or magnitude of impairments: one study, with a relatively small sample size (n = 17 DCD), found a deficit only on a nonverbal measure, while another study, with 23 subjects with DCD, did not find any significant impairment. Similarly, half of the studies measuring verbal response inhibition found some impairment in this area, although completion time was the only aspect affected in one study. Regarding studies exploring alertness and sustained attention, selective attention, divided attention and verbal fluency, most found no deficit in these areas in children with DCD. No studies investigating organization and self-regulation skills were included in this systematic review as none met the inclusion criteria.

In light of these results, there is no evidence of an obvious attentional deficit in children with DCD, but executive functions are widely impaired. Namely, difficulties are not limited to domains directly associated with core impairments of the disorder (e.g. nonverbal planning), but extend to other areas of executive processes, including nonverbal inhibitory control, nonverbal and verbal working memory, and general planning abilities, as well as on ecological tasks integrating multiple components of executive functions. In summary, executive functions are more impaired on a nonverbal/visuospatial modality than on a verbal modality, and results suggest that a relatively broad executive deficit is present in children with DCD.

A few studies also explored the possible differences between a group of children with DCD and a group of children with MD without DCD. As previously described, these groups were composed of children whose motor difficulties have been objectified, but who did not meet formal diagnostic criteria for DCD. It is interesting to note that in most studies, no significant difference was noted between MD and DCD groups in terms of executive impairments. This raises the possibility that executive difficulties are tightly linked with motor impairments, irrespective of severity. Consequently, the findings and conclusions about DCD highlighted in this review could also apply to a large portion of children with measurable motor difficulties that do not have a clinical diagnosis of DCD.

4.2. Influence of comorbid disorders

Among the cognitive functions that were noted as being impaired in most studies, the presence of comorbid disorders did not seem to have a determinant influence on inhibitory control (both response and attentional inhibition), visuospatial working memory, planning, cognitive flexibility, nonverbal fluency and general executive functioning capacities. Thus, the weaknesses described in these domains do not appear to be attributable to co-occurring disorders, but rather suggest that cognitive impairments are integrant elements of DCD. This is unexpected since DCD is primarily a motor disorder, and the definition does not include executive functioning problems [1–3]. However, not all studies reviewed here addressed the issue of comorbidities in the same way: some studies based their exclusion criteria only on comorbid disorders that had been diagnosed in children, while others measured and confirmed their presence in children composing their sample. In most studies in which response inhibition, visuospatial working memory, planning, cognitive flexibility, nonverbal fluency and general executive functioning were assessed, children with co-occurring disorders were excluded based on the presence of previously diagnosed disorders, without these diagnoses being confirmed by the research teams. However, the presence of children’s difficulties linked to comorbid disorders was confirmed by authors in most studies investigating attentional inhibition. Therefore, it is particularly clear that deficits in attentional inhibition are part of DCD. However, since most studies did not confirm or refute the presence of co-occurring disorders, it is possible that some children in DCD samples nonetheless suffered from additional neurodevelopmental disorders that were not documented. Considering this, additional studies using large sample size and thorough screening methods to assess the presence of comorbid disorders are needed to conclude on the definite influence of comorbid conditions on the cognitive profile of children with DCD.

Despite the imprecision of screening and sometimes inclusions and exclusion criteria in some studies, it seems likely that comorbid conditions exert at least some influence on verbal working memory capacities. Indeed, among the eight samples in which verbal working memory difficulties were found, all comorbid disorders were excluded from four samples [59, 61, 79, 91], and two other samples were free of ADHD, behavioral problems or ASD [26, 63]. It is thus unlikely that the impairments reported in these studies are due to these co-occurring neurodevelopmental disorders, but it does not exclude the possible influence of language or learning difficulties in half of the studies [20, 26, 63, 90]. Interestingly, among the three samples in which verbal working memory abilities were preserved, two were pure DCD samples and one was free of comorbid ADHD or SLI. In summary, when at least the co-occurrence of SLI is excluded, impairments in verbal working memory are less probable in children with DCD. Therefore, deficits in this area might not be part of the DCD profile per se, but may appear when co-occurring difficulties, especially in terms of language, are present.

It appears that difficulties observed on measures of visuospatial and verbal working memory are frequent but not specific to children with DCD, since they are not significantly different than those found in children with neurodevelopmental disorders that do not necessarily involve a motoric component, including ADHD, SLI, RELD, ASD, and learning difficulties. For this reason, it is possible that impairments in these domains could reflect a general sign of an atypical cognitive development, without being specific to a given disorder, nor systematically present in children with neurodevelopmental conditions.

4.3. Brain correlates

Studies on the neural correlates of DCD contribute to our understanding of the results included in the present review. Several neuroimaging studies have revealed implications of corpus callosum, basal ganglia, inferior parietal cortex, thalamus and cerebellum in DCD, and of connections between these structures [44, 65, 94, 95]. Abnormalities have also been noted in frontal cortex functioning and in white matter maturation and composition in individuals with DCD [44, 95–97]. Furthermore, deficits in functional activation during a motor response inhibition task have been found in the right pre- and postcentral gyri, as well as in the left superior and middle frontal gyri [81]. However, the authors specified that these deficits were only present in children with co-occurring DCD and ADHD. Given the plurality of brain areas involved in DCD, the hypothesis of an “Atypical Brain Development” (ABD) has been proposed. This view stipulates that brain abnormalities underlying the cognitive and motor deficits observed in various neurodevelopmental disorders are rather diffuse than localized [19, 98]. ABD is not a disorder per se, but rather a suboptimal developmental process that manifests itself in different forms, such as motor, attentional and/or reading disorders [19, 99]. Given the strong overlap between symptoms of DCD and ADHD [19], the umbrella term “Deficit in Attention, Motor control and Perception” (DAMP) was formerly proposed to define a combination of DCD and ADHD [6, 100]. The authors insisted on the importance of a term acknowledging both attentional and motor control problems in clinical practice, stipulating that these symptoms have strong common background factors and that the prognosis of children with DAMP was poorer than that of children with either DCD or ADHD [100].

Several authors have also hypothesized that children with DCD could have an automatization deficit, related to a cerebellar dysfunction [19]. Although possibly greater and more obvious in children with DCD than in children with ADHD and/or DDL, this deficit does not appear to be specific to the DCD population, but is on the contrary common to most developmental disorders [19, 101]. It could therefore explain the high prevalence of comorbidities among neurodevelopmental disorders and the non-specificity of the cognitive impairments found amongst them [98, 102].

The fact that the frontal cortex, basal ganglia, cerebellum and parietal cortex are especially implicated in attentional and executive functioning [4, 38, 41], and that these brain structures do not operate optimally in individuals with DCD, as previously described, might explain some of the attentional and executive impairments compiled in the present review. Additionally, an atypical hemispheric lateralization for attention and inhibitory functions has been shown in children with DCD, along with a reduced efficiency of cerebral network involved in inhibitory control [44]. Evoked potential studies have also reported that children with DCD allow fewer resources than TD controls for spatial locations comparison and response retrieval and selection [68], and that they may have less mature anticipatory and executive processes, reduced interhemispheric and cognitive-to-motor transfer speeds as well as an atypical neural activity associated with attentional control [85, 88, 89]. Without providing correlational or causal evidence, the set of cerebral abnormalities observed in DCD, which involves key structures robustly involved in the corticostriatal function and the frontal cortex, the presence of impairments in the executive and attentional domain in DCD is not unexpected.

5.1. Strengths and limitations of the current review

This systematic review of the literature has several strengths. On a methodological level, the facts that we used a broad variety of research terms for executive functions and that we included studies using different types of performance tasks, from experimental tasks to standardized neuropsychological tests, allow our review to be exhaustive, inclusive, and thus more representative of the attentional and executive functions in children and adolescents with DCD. The inclusion of studies using more ecological tasks is also a strength since our results contribute to a better understanding and representation of these children’s daily difficulties. In addition, this review is the first to explicitly discuss the influence of concomitant disorders on attentional and executive processes in DCD. The fact that we included studies that used DCD samples with comorbid disorders, as well as the fact that the included studies recruited their DCD sample from various sources, increases the likelihood that the conclusions drawn are ecologically valid and generalizable to the population of children with DCD.

Regarding methodological limitations, the lack of search and inclusion of grey literature and nonpublished studies may bring a publication bias. Furthermore, a standardized quality assessment tool for neuropsychological studies could not be found, so we developed our own quality assessment checklist based on the NOS and the PRISMA standards. By this mean, this review raised the risk of including studies that provide inadequate quality of evidence. Although we explicitly mentioned in the relevant sections when the quality of a study was low, a common and important limitation of several studies was the small sample size, which entails that even when grouped together, the number of participants remained relatively limited. This is especially relevant for results regarding auditive alertness/sustained attention, divided attention, comparisons between verbal and nonverbal planning and between verbal and nonverbal cognitive flexibility, and verbal fluency. Similarly, some of the attentional and executive components discussed in this review have only been the topic of a small number of studies, particularly selective attention, divided attention, cognitive flexibility, fluency and general executive functioning. Thereby, we must be cautious when considering the conclusions drawn regarding these domains, and further studies are necessary to confirm the interpretations made of these findings. Another important limitation of studies included in this review is the inconsistency regarding the clarity of exclusion criteria of several DCD samples. As the presence of comorbid disorders was not always rigorously documented nor confirmed, it is possible that comorbid disorders may play a larger role in the results highlighted in this review than we believe.

Other factors are to be considered when interpreting the conclusions of this systematic review. Firstly, we presented our results according to nine attentional and executive components to reduce the heterogeneity of the construct evaluated in each section. However, on a conceptual level, attentional and executive functions, although dissociable, are also united and interrelated [33, 36, 41, 103]. Similarly, all tasks require the implication of more than a single cognitive function, and in certain cases, it might be difficult to isolate one attentional or executive component. As a matter of fact, some studies using the same assessment tools reported their results in relation to different cognitive domains. Hence, the results described might be partially due to impairment in other cognitive areas than those purportedly assessed. Also, several tasks used to measure cognitive functions require a motor component. Children with DCD could perform more poorly on these tasks not because of an attentional or executive impairment, but because of their primary motor deficit associated with their disorder. Secondly, no study explicitly assessing organization, self-regulation or goal-setting processes were included since none met the eligibility criteria established in this review and because few tasks assessing these components of executive functions exist. Therefore, this systematic review does not allow discussing the integrity of these executive functioning components. Thirdly, even though eligibility criteria allowed including studies with subjects up to 17 years old, the most often studied ages were 8 to 12 years old. Given the considerable development of executive functions through childhood and adolescence [36, 38, 104], it would be incorrect to generalize the results of this review to younger children and older adolescents. Fourthly, it was previously mentioned that impairments found in the domains of visuospatial and verbal working memory were not specific to children with DCD, since children with ADHD or ASD shared these difficulties. Yet children with ADHD and ASD also often share sensorimotor deficits with children with DCD [23, 24, 105]. Given the overlap between symptoms that seems to be present, there is a limitation to the comparisons that can be made between these clinical groups. Finally, as in most neurodevelopmental disorders, there is great heterogeneity among children with DCD [19, 26, 65, 91]. The lack of details regarding the severity and predominant type of impairments could contribute to the variability of results noted in certain cognitive areas.

5.2. Directions for future research

Additional studies using larger samples and replicating the results of these studies are needed, especially in domains of alertness/sustained attention, divided attention, verbal and nonverbal planning, verbal and nonverbal cognitive flexibility, and verbal fluency. Likewise, a small number of studies have explored selective attention, divided attention, cognitive flexibility, fluency and general executive functioning in children in DCD, and more research on these cognitive functions is also needed to confirm the results highlighted in this review. Since no studies discussing organization, self-regulation and goal-setting skills and meeting our eligibility criteria were found, more rigorous studies exploring these executive functions are needed. Future studies should also focus on children under 8 years old and over 12 years old. Considering the development of attentional and executive functions during childhood and adolescence, and the variability of developmental trajectories among the different components of executive functioning [104], cross-sectional and longitudinal studies comparing different age groups and following children through several years would be important in order to determine whether children with DCD present a lasting executive deficit or whether their difficulties are due to a developmental delay that catches up with time. In addition, to improve results generalization and comparison between studies, future research should avoid selecting their sample only in specialized clinical settings in order to obtain samples that are more representative of the population of children with DCD. Furthermore, given the fact that co-occurring disorders may have an influence on some executive functions, especially on verbal working memory, future studies should document and verify the presence of comorbid conditions in their sample. More studies should also compare attentional and executive functioning profiles of children with DCD without any other developmental or neurological problem, with that of children with other neurodevelopmental disorders and of children with multiple diagnoses to explore the possible distinction between these groups. Lastly, future research should allow better understanding of the overlap between executive functions and motor skills. To do so, the choice of tasks should be judicious and make possible a better comparison of children’s performance on tasks requiring motor and visuospatial skills with their performance on tasks requiring no such abilities. In addition, given the overlap between DCD, ADHD and ASD previously discussed, a dimensional approach could be useful to go beyond group comparisons and allow a better understanding of co-occurring motor and cognitive difficulties. With more studies using larger samples, a meta-analytic approach would be relevant and could control for such confounding variables.