Internal Model Deficits Underlying Motor Problems in Very Preterm Children?

Internal Model Deficits Underlying Motor Problems in Very Preterm Children?

Cece C. Kooper

Institute of Interdisciplinary Studies, University of Amsterdam 52442RP00Y: Research Project 2

July 17, 2020

Amsterdam UMC, AMC, Department of Neonatology Daily supervisor: Celina. E. Henke, MSc

Examiner: Prof. dr. Jaap Oosterlaan Assessor: Dr. Egbertdina S. Twilhaar

Abstract

Very preterm (VP) birth puts children at-risk for motor impairments. Earlier research suggested that visuomotor (VM) deficits may play a crucial role in the occurrence of motor impairments in VP children. Moreover, it is thought that the executive functions, response inhibition and response re-engagement, are associated with VM performance and together propose underlying internal model deficits in VP children. Therefore, the first aim of this study is to investigate whether VM performance is related to response inhibition as well as response re-engagement in VP children at 5 years corrected age (CA). Secondly, this study aims to investigate whether VM performance predict the individual differences in motor skills. The motor development of 71 VP children was assessed using the second Dutch edition of the Movement Assessment Battery for Children. Then, VM performance, response

inhibition and response re-engagement, were assessed using a touch-screen computerized tracing task. Pearson correlation analyses showed that response inhibition and response re-engagement were both weakly correlated with VM performance (r(62) = .35, p = .004; r(62) = .36, p = .004 respectively). In addition, linear regression analysis showed that VM

performance explained 20.7% (p < .001) of the variance in motor skills in VP children. The findings are in line that internal model deficits might underlie the frequently seen motor impairments in VP children and especially highlight the predictive value of VM deficits on motor impairments. Consequently, the findings may help develop effective strategies for interventions in VP children at 5 years CA.

Keywords: Internal Model Deficit, Motor Impairment, Response Inhibition, Response Re-Engagement, Very Preterm, Visuomotor Deficit

Internal Model Deficits Underlying Motor Problems in Very Preterm Children? According to the World Health Organization, every year around 15 million children worldwide are born preterm (PT; i.e. born before 37 weeks of gestation), and this number is still increasing (World Health Organization [WHO], 2018). In addition, the survival rate of children born very preterm and extremely preterm (VP and EP; i.e. born before 32 and 27 weeks of gestation respectively) have increased noticeably due to advances in neonatal care (Central Bureau of Statistics [CBS], 2019; Saigal & Doyle, 2008). However, the last trimester of pregnancy is the crucial stage of major brain growth and development, which VP children spend outside the womb (de Kieviet et al., 2012; Volpe, 2009). Unfortunately, VP children are susceptible to numerous pathogenic mechanisms such as cerebral ischaemia and

inflammation, affecting their development (Volpe, 2009). Moreover, prematurity is associated with altered brain development of whom children born EP are most at-risk (Alexander et al., 2019). It is well established that PT birth is associated with negative neurodevelopmental outcomes, and that low gestational age (GA) puts children more at-risk (Pascal et al., 2018; Serenius et al., 2016).

A developmental domain in which children born VP often show deficits is the motor domain. The presence of a motor impairment is more than six times higher in children born VP compared to full-term (FT; i.e. born after 37 weeks of gestation) and this impairment is likely to persist throughout childhood (Edwards et al., 2011; De Kieviet et al., 2009; Pascal et al., 2018). Furthermore, Van Hus et al. (2014) showed that VP children at 5 years corrected age (CA; i.e. chronological age reduced by the number of days born before 40 weeks of gestation) with motor impairment showed larger neurodevelopmental impairments such as slow processing speed, compared to VP children free of motor impairment. In addition, visuomotor (VM) deficits in VP children have been consistently found in earlier research, and even seem to persist into adulthood (Geldof et al., 2012). There is growing evidence that

deficits in VM performance may play an important role in the commonly seen motor

impairment in VP children (Bolk et al., 2018; De Kieviet et al., 2009; De Kieviet et al., 2013; Goyen et al., 1998). Therefore, investigating the underlying mechanisms of VM performance in VP children is of high need to create a better understanding and to eventually help create an efficient intervention for the frequently occurring motor impairment.

VM performance is the ability to use visual information to guide one’s motor behaviour (Evensen et al., 2009). Previously, VM performance has been investigated in FT children using internal models that support motor planning and control (Imamizu & Kawato, 2009; Shadmehr et al., 2010). VM performance demands the nervous system to predict and control movements, as traditionally described by Wolpert et al. (1995), integrating efferent and afferent signals to eventually enhance adaptive functioning. By having internal models, predicting the outcome of movements can be processed much quicker in the brain than when a child has to rely on relatively slow sensory-feedback information. Interestingly, previous assessment of VM performance in VP and FT children showed they have similar success in well-structured and predictable motor activities, yet perform worse than FT children in non-structured motor activities (De Kieviet et al., 2013). Thereby, proposing the role of

predictability in motor adaptive functioning to be crucial in VP children. Consequently, in VP children these internal models or representations of intended actions are suggested to be impaired, also referred to as internal model deficits.

To be able to have internal models, the higher-order cognitive processes are of need (i.e. executive functions need to be developed). Yet, executive functioning (EF) has found to be inversely related to GA (Stålnacke et al., 2019). Besides, two meta-analyses found lower EF in children born VP compared to children born FT (Aarnoudse-Moens et al., 2009; Mulder et al., 2009). Neuroimaging research suggested that lower EF in VP children is influenced by connectivity disturbances between posterior brain regions and prefrontal cortices (Skranes et

al., 2009). Furthermore, a meta-analysis showed that the corpus callosum, needed for cerebral connectivity, and the hippocampus, combining input from the entire sensory field, were both smaller in children born VP compared to FT (De Kieviet et al., 2012). Recently, Wheelock et al. (2018) showed cerebral connectivity disturbances in relation to motor skills. VP children showed weaker associations between brain connectivity of motor networks and actual motor skills compared to FT children. Taken together, these studies are supportive of underlying internal model deficits, possibly putting them at-risk for motor impairments.

There is growing consensus on the role of EF in VP children that could help explain lower-level deficits such as motor impairments (Anderson, 2014; Stålnacke et al., 2019). This study focusses on the two core components of EF as proposed by Diamond (2013): 1)

response inhibition and 2) response re-engagement, the latter similar to cognitive flexibility of Diamond (2013). Firstly, response inhibition is defined as the ability to stop an ongoing action while the ultimate goal is to enhance adaptive functioning (Halperin et al., 1994). Secondly, response re-engagement is defined as the ability to shift rapidly to an alternate action (Oosterlaan & Sergeant, 1998). Both mechanisms seem to underlie the ability to change rapidly in a current course of action, similar to the goal of VM performance and need to be examined in VP children. Fortunately, the touch-screen computerized task developed by De Kieviet et al. (2013) to assess VM performance also facilitates the assessment of response inhibition and response re-engagement seperately.

To date, no previous research has investigated the relationship between response inhibition as well as response re-engagement and VM performance, and its predicting value on motor impairments in VP children at 5 years CA. Therefore this study aims to further examine, 1) whether VM performance is related to response inhibition and response re-engagement, and 2) whether VM performance (separately for assessment with and without a well-framed structure) predicts the individual differences in motor skills of VP children.

Firstly, due to the hypothesized internal model deficits, it is expected that VP children with high response inhibition scores will show high VM performance scores. In line, it is expected that children with high re-engagement scores will show high VM performance scores.

Secondly, it is hypothesized that VM performance has positive predictive value on motor skills. Expected is that children with high VM scores (on the structured and more so the unstructured part), reflecting a low VM performance, will show low motor scores. Then, GA is hypothesized to have positive predictive value on motor skills. Expected is that children with high GA will show high motor scores. The results of the current study will help

understand underlying mechanisms of VM performance, and its predictive value on the motor development in VP children at 5 years CA. Accordingly, it could possibly help to create effective strategies for new interventions that might be helpful for VP children.

Methods Participants

A sample of 71 Dutch children born between October 2011 and November 2014, at less than 32 weeks of gestation and/or with a birth weight below 1500 grams, who had been admitted to the Neonatal Intensive Care Unit in the Amsterdam University Medical Centers (location AMC and VUmc) or the University Medical Center Groningen participated in this study. Recruitment occurred via the regular neonatal follow-up program at 5 years CA. Exclusion criteria were 1) an overall Full-Scale Intelligence Quotient score below 70 on the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-III-NL; Hendrikson & Hurks, 2009), 2) severe visual impairment, and/or 3) cerebral palsy (Gross Motor Function

Classification System level II or higher). This study was part of a larger project that has been approved by the Medical Research Ethics Committee of the Amsterdam UMC, location AMC.

Measures

VM Performance and Underlying Mechanisms

The caterpillar tracing task is a touch-screen computerized test to assess multiple aspects of VM performance in children (De Kieviet et al., 2013). The task is based on previous so-called tracing and pursuit tasks that have been widely used and validated for studying VM performance in a variety of clinical populations (Caeyenberghs et al., 2010; Koekkoek et al., 2008; Rommelse et al., 2007; Stirling et al., 2013). The caterpillar tracing task was executed on an electronic device (iPad, version 10.3.3, model MD510NF/A), secured in a specific holder to allow for correct vision and physical support. Children were seated in front of the electronic device and instructed by the trained research assistant in a standardized manner to trace the nose of the caterpillar. Also, they were instructed to only use their index finger of their preferred hand. During this task, the research assistant repeated the instructions when needed. The task contains a structured and unstructured part. Both parts start with a practice trial followed by six test trials of 30 seconds each. The speed level of the caterpillar gradually increases per trial from 40, 60, 80, 100, 120, to 140 pixels/ms, resulting in a systematic increase of workload. The task always starts with the structured part, where the caterpillar moves in a fixed circle (diameter = 14.9 cm, example given in Figure 1C). This is followed by the unstructured part, where the caterpillar moves randomly over the entire screen (example given in Figure 1D).

Figure 1

Screenshots of Four Different Phases in the Caterpillar Tracing Task

Note. A) The caterpillar is ready to start moving and will move when the participant touches the nose. B) After each trial, the caterpillar returns home, and the participant can rest. C) The structured part where the caterpillar is always moving in circles. D) The unstructured part where the caterpillar moves randomly across the entire screen.

Assessment of VM performance in the caterpillar tracing task is separate for the structured and unstructured part, differentiated by the predictability of the structure of the caterpillar. For each trial, discrepancy scores were calculated with the mean distance (in pixels) between the x and y coordinates of the index finger and the x and y coordinates of the nose of the caterpillar (stored every 16.6 ms). Specifically, the formula of the Pythagorean Theorem was used to calculate this distance (expressed in c), c = √a2 _{+ b}2 _{where a stands for}

the difference in x coordinates and b stands for the difference in y coordinates. Then for each participant, the mean discrepancy scores of the six test trials, separately for the structured and unstructured part, were calculated and used as indicators of VM performance. For both measures, lower discrepancy scores indicated a better VM performance. For this study, only the unstructured part of VM performance was used to assess its possible underlying

mechanisms.

Assessment of both response inhibition and response re-engagement have been previously done with the change task based on Logan’s race model, yet this task exerts high cognitive processing demands (Logan & Burkell, 1986; Logan et al., 1984; Oosterlaan & Sergeant, 1998). Fortunately, the caterpillar tracing task allows assessment of response inhibition and response re-engagement without the demand of numerous higher order cognitive skills. Only the unstructured part allowed to calculate response inhibition and response re-engagement scores. Specifically, the average response time (in sec) between changed direction of the caterpillar and changed direction of the index finger towards the caterpillar, was calculated for the six test trials. The mean of the six test trials was then calculated and used as an indicator of response inhibition. The average response time (in sec) of re-tracing the caterpillar’s head (i.e. radius of 0.84 cm from the caterpillar’s nose) after children changed direction towards the caterpillar was calculated for the six trials. The mean was then calculated and used as an indicator of response re-engagement. For both scores, a lower score indicated a better performance.

Motor Skills

The second Dutch edition of the Movement Assessment Battery for Children (M-ABC-II-NL) is a standardized test to assess motor skills in children aged 3 to 16 years (Henderson et al., 2010). It is the most commonly used test for assessing motor performance in children without severe diagnosed impairments and has a good validity and reliability (De

Kieviet et al., 2009; Ellinoudis et al., 2011). The test was assessed by an experienced

children’s physiotherapist or paediatrician at the outpatient clinic. The M-ABC-II-NL consists of eight small tasks which can be categorized into three components, namely 1) balance, 2) manual dexterity, and 3) aiming and catching. The total score is the sum of all three

components and can be interpreted as an overall measure of motor development. For each component and the total score, the points were transformed into age-adjusted standard scores (based on a large normative sample) according to the CA of the participants. A higher

standard score indicates better motor skills, and a standard score below eight (i.e. 1.0 SD below the normative mean) is considered as an indicator of motor impairment (Henderson et al., 2007). Here, the corrected age-adjusted total standard score was used as indicator of motor skills.

Procedure

The data was obtained as part of a larger study (Power Move: a Randomized Waitlist-Controlled Study on a Computerized Motor Intervention Program to Improve Motor Function in Very Preterm Children at 5 Years of Age) and conducted in the participating academic hospitals and participants their home. When children were eligible for participation in the study, the parent(s)/caregiver(s) were informed about the study during the regular neonatal follow-up program by the children’s physiotherapist or paediatrician and received the information letter. One week after visiting the outpatient clinic, the parent(s)/caregiver(s) received a phone call from the researcher to see whether they had any questions and whether they wanted to participate in the Power Move study. If so, a home visit was planned. During the home visit, informed consent was signed by parent(s)/caregiver(s) of the participant and additional tasks on electronic devices, paper, as well as physical tasks were conducted (the used measures are shown in Figure 2). Already during the outpatient clinic, the M-ABC-II-NL was conducted as part of the regular neonatal follow-up care by a children’s

physiotherapist or paediatrician. During the home visit, the participants completed the caterpillar tracing task under the supervision of a research assistant. The participants did not receive any compensation.

Figure 2

Timeline of the Power Move Study

Note. Tasks marked in blue will be used for the analyses. WPPSI-III-NL = Wechsler

Preschool and Primary Scale of Intelligence third Dutch edition; MABC-II-NL = Movement Assessment Battery for Children second Dutch edition; VMI = Visual-Motor Integration; MC = Motor Coordination; VP = Visual Perception.

Statistical Analyses

All analyses were performed using R, version 1.2.5019 (R Core Team, 2017). The level of significance was set to .05, and participants with missing substantial amounts of data were excluded from the analyses. To examine the link between the underlying mechanisms and VM performance, the assumption of normality was tested for all variables using Shapiro Wilk and when violated transformed in Van der Waerden scores (Soloman & Sawilowsky, 2009). Pearson correlation analyses were conducted between the response inhibition scores and VM performance scores (unstructured part) as well as between the response

Secondly, to assess whether VM performance predict the individual differences in motor skills of VP children, multiple linear regression analyses were conducted. To start, all assumptions of linear regression were checked, namely linearity, homoscedasticity, normality of residuals and influential data points. Then, the predicting value of VM performance on motor scores were measured separately for the score in the structured part (Model 1a) and the unstructured part (Model 1b). Next, GA was added to both models, assessing the predicting value of ‘VM performance (structured part) + GA’ (Model 2a) and ‘VM performance (unstructured part) + GA’ (Model 2b) on motor scores. To account for the complexity of the models, the Akaike Information Criterion (AIC) was used to examine the model’s best fit with the fewest predictors, where a lower AIC score is considered best (Navarro, 2019).

Results Participant Characteristics

Due to substantially incomplete data of the caterpillar tracing task seven participants were excluded. Consequently, analyses were performed with 90.1% of all included

participants (n = 64), born between 24 and 33 weeks of gestation. The participant

characteristics and M-ABC-II-NL scores, separately for all subscales, are listed in Table 1. Interestingly, 85.9% of this sample show a score of 1.0 SD below the age-adjusted normative mean on one of the M-ABC-II-NL subscales indicating high rates of motor impairments in the sample. In addition, Table 1 shows parental education of the participants, representing the highest education completed by one of the parent(s)/caregiver(s).

Table 1

Characteristics of the Participants From the Neonatal Period and at Assessment

Variable Mean SD Range

Birth Weight (grams) Gestational Age (days) Age (years : months)

Corrected Age (years : months)

1052 196 5:5 5:2 255 13 0:3 0:3 430-1580 172-234 5:2-6:2 4:11-5:11 MABC (corrected age-adjusted)

Manual Dexterity Balance Skills

Aiming and Catching Total Motor Skills

7.3 8.1 8.6 7.2 2.6 3.3 2.7 2.8 2-13 2-16 4-17 2-14 Number % Boys

Motor Impairment Based on Subscale Motor Impairment Based on Total Scale Highest Parental Education a, b

Low Intermediate High 32 55 38 6 17 37 50.0 85.9 59.4 9.4 26.6 57.8 Note. MABC = Movement Assessment Battery for Children.

a

Parental education was classified according to the Central Bureau of Statistics (2019) in the Netherlands. ‘Low’ indicates primary education or prevocational secondary education; ‘intermediate’ indicates secondary education or middle vocational education; and ‘high’ indicates higher professional education or university.

Underlying Mechanisms of VM Performance

In order to assess the underlying mechanisms of VM performance, the relation between VM scores with 1) response inhibition scores and 2) response re-engagement scores were tested. VM scores (M = 152.6, SD = 16.7) were normally distributed. Response

inhibition scores (M = 0.3, SD = 0.1; W (64) = .67, p < .001) and response re-engagement scores (M = 1.1, SD = 0.5; W(64) = .60, p < .001) were not normally distributed and transformed into Van der Waerden scores. The VM scores were positively, yet weakly correlated with response inhibition scores, r(62) = .35, p = .004. In addition, the VM scores were also positively, but weakly correlated with response re-engagement scores, r(62) = .36, p = .004. Furthermore, Figure 3 shows the weak correlations, separately for both tested

underlying mechanisms of VM performance. To explore whether response inhibition and response re-engagement are partly independent processes a Pearson correlation analysis was conducted between the scores and showed a moderate association, r(62) = .53, p < .001.

Figure 3

The Correlation Between VM Performance and Response Inhibition (Left), as Well as VM Performance and Response Re-Engagement (Right)

Note. R stands for Pearson’s rho and the grey area indicates the confident interval of the black regression line. VM = Visuomotor.

Predictive Value of VM Performance on Motor Skills

To assess whether VM performance (separately for the structured and unstructured part) is associated with the individual differences in motor skills of VP children, the predicting value of VM performance scores and GA was tested using the CA adjusted standard total motor score as dependent variable. To perform the multiple linear regression analyses, the assumption of normality (for the dependent variable), linearity,

homoscedasticity, normality of residuals and influential data points were first checked and did not show violations in the tested models. In Table 2 the results of the multiple linear

regression analyses are presented, including the F ratio, degrees of freedom, the adjusted R squared and AIC separately for each model. Results of Model 1a showed that VM scores in the structured part (B = -0.02, 95% CI [-0.04, -0.0003]) significantly predicted motor scores, accounting for 4.7% of the variation in motor skills. Interestingly, Model 1b showed that VM scores in the unstructured part was a significant predictor (B = -0.08, 95% CI [-0.12, -0.04]), accounting for 20.7% of the variation in motor skills. As a second step, GA was added as a predictor to both Model 1a and 1b to assess the combined predictive value, yet GA did not have significant predictive value in both models.

Exploratory analyses were performed to increase the understanding and GA in days (W(64) = .95, p = .009) was not normally distributed. Spearman correlation analysis showed that GA and motor scores were only weakly associated, r_s(62) = .26, p = .04. Therefore, GA was normalized and added to Model 1a and 1b. However, the results with GA normalized remained similar to Model 2a and Model 2b (see Table 2). To further explore characteristics in the current sample a Pearson correlation analysis was conducted between birth weight and motor scores. Birth weight was normally distributed and showed a weak association with motor skills, r(62) = .29, p = .02. To explore the influence of sex an independent T-test was performed with motor score as dependent variable. The results showed that boys (M = 6.9,

SD = 3.1) and girls (M = 7.4, SD = 2.6) did not significantly differ, t(62) = -0.79, p = .434. Therefore, sex was not added as independent variable to the models.

Table 2

Results of the Linear Multiple Regression Analyses Assessing Whether Numerous Predictors Explain the Variance in Motor Skills

Predictor B F df p adj. R2 _AIC

Model 1a 4.14 [1, 62] .046 .047 316.4 VM performance (structured) -0.018 .046 Model 1b 17.24 [1, 62] < .001 .207 304.7 VM performance (unstructured) -0.080 < .001 Model 2a 4.10 [2, 61] .021 .090 314.4 VM performance (structured) -0.018 .039 Gestational age 0.051 .054 Model 2b VM performance (unstructured) -0.075 9.46 [2, 61] < .001 < .001 .212 305.2 Gestational age 0.029 .243

Note. AIC = Akaike Information Criterion.

Discussion

This study investigated two possible underlying mechanisms of VM performance and whether VM performance predicts the individual differences in motor skills in VP children at 5 years CA. The main finding of this study indicates that VM deficits predict motor

impairments in VP children. In addition, the findings suggest that both response inhibition and response re-engagement are associated with VM performance. The results of this study contribute to the understanding of possible underlying mechanisms of VM performance, and

it supports that internal model deficits might underlie the commonly seen motor impairments in children born VP. Consequently, this improves our understanding of underlying

mechanisms beneficial for effective strategies for the development of interventions and are thereby important for current neonatal follow-up care.

The current study was the first to assess response inhibition and response re-engagement in relation to VM performance in VP children. Both response inhibition and response re-engagement are suggested to be weakly related to VM performance. Although more research is necessary, the findings are supportive of the earlier suggested internal model deficits in children born VP (De Kieviet et al., 2013; Imamizu & Kawato, 2009, Shamedhr et al., 2010; Zago et al., 2009). Besides, both associations extend the role of EF in relation to lower-level deficits in VP children (Cameron et al., 2012). Hence, it emphasizes the

importance of examining higher-order cognitive processes in unravelling lower-level deficits and the possible interrelation.

The second aim of this study was to investigate the predictive value of VM

performance on motor development in VP children at 5 years CA. As expected, the findings indicate that VM performance, regardless of the structural differences in assessment, partly predict the individual differences in motor skills. Therefore, the results of this study add evidence to the growing consensus on the role of VM performance in the motor development of VP children (Bolk et al., 2018; De Kieviet et al., 2009; Goyen et al., 1998). Interestingly, the results show that VM performance (unstructured part) reliant on a constant demand in motor response adaptation is a far better predictor of motor skills than VM performance (structured part) reliant on a singular motor response adaptation. In other words, this suggests that the extent of being able to constantly utilize and generate internal models of motor

the results of this study predict 21.2%, after controlling for GA, of the individual differences in motor development and more predictors are yet to be examined.

This section addresses limitations and strengths of the present study. This study was the first to use the caterpillar tracing task to assess response inhibition and response re-engagement plus examining the association with VM performance. Results show small effect sizes and should be interpreted with extra care. Nevertheless, the touch-screen computerized assessment has many strengths. Importantly, considering the current interest in the motor domain and the possible interrelation with cognitive abilities, assessment of EF should not interfere with testing intelligence (Cunha et al., 2018; Oudgenoeg-Paz et al., 2017; Twilhaar et al., 2018). In this study VM performance, response inhibition and response re-engagement are all highly standardized, have a highly accurate response-time recording, have minimum conflict with testing cognitive abilities, and thereby increase accuracy in scoring. Plus, the results are robust to the adverse effect of a possible language barrier, since language did not play a part in the used measures. However, this study did not control for numerous neonatal care factors that possibly limit the predictive value on the individual differences in motor skills. Against expectations, GA and motor skills only showed a weak association, and GA did not have significant predictive value on motor skills in this study. However, the used motor scores had been age-adjusted according to the CA which could contribute to the unexpected weak association. Fortunately, this study with solely children born premature included a large range in GA and birth weight. Accordingly, the findings are highly informative for children born VP and EP, plus for children with very low birth weight (VLBW). In favour of generalizing the findings, this sample also includes different levels in education of the parent(s)/caregiver(s).

Overall, due to the increased survival rate of children born VP and/or with VLBW, with 85.9% of VP children showing motor problems at 5 years CA, the population of interest

that could benefit from the findings is increasing till date. Consequently, this stresses that neonatal follow-up care is still of high need at 5 years CA. Furthermore, the findings extend the understanding of underlying mechanisms in long-term neurodevelopmental outcomes after prematurity. Possibly efficient intervention strategies should aim to highlight improving adaptive motor functioning to eventually help motor development in VP children.

Subsequently, the findings suggest that the caterpillar tracing task could be a useful tool to examine who is at risk for motor impairments while simultaneously assessing two core components of EF in children in a standardized, short manner with low examiner-burden. Next, the caterpillar tracing task extends efficient objective EF assessment possibilities in children (Baron et al, 2012; Böhm et al., 2010). However, the findings first long for future studies to investigate the reliability and validity of the caterpillar tracing task. Assessment of larger as well as different populations, including typically developing children, will increase the understanding whether the results of this study are applicable for solely VP children. Furthermore, future research should explore the influence of VM performance (unstructured part) on other neurodevelopmental outcomes such as cognitive abilities and school

performance in VP children. In addition, future studies should control for attention problems and hyperactive behaviour that have been associated with VM deficits specifically in VP children compared to FT children at 5 ½ years of age (Böhm et al., 2010). Finally, future research with neuroimaging techniques may help the understanding of the suggested internal model deficits on a neural level and examine the relationship between altered brain

development and VM deficits in VP children.

In conclusion, the results of this study suggest that internal model deficits might underlie the commonly seen motor impairments in children born VP at 5 years CA.

Importantly, this study extends the crucial role of VM deficits, especially during continuous response motor adaptation demands, as a predictor of motor impairments early in

development of children born VP. Finally, it may contribute to provide effective strategies for the development of beneficial interventions in VP children before the demands in school strongly increase.

References

Aarnoudse-Moens, C. S., Smidts, D. P., Oosterlaan, J., Duivenvoorden, H. J., & Weisglas-Kuperus, N. (2009). Executive function in very preterm children at early school age. Journal of Abnormal Child Psychology, 37(7), 981-993.

https://doi.org/10.1007/s10802-009-9327-z

Alexander, B., Kelly, C. E., Adamson, C., Beare, R., Zannino, D., Chen, J., & Anderson, P. J. (2019). Changes in neonatal regional brain volume associated with preterm birth and perinatal factors. Neuroimage, 185, 654-663.

https://doi.org/10.1016/j.neuroimage.2018.07.021

Anderson, P. J. (2014). Neuropsychological outcomes of children born very preterm. In Seminars in Fetal and Neonatal Medicine (19(2), pp. 90-96). WB Saunders: Elsevier. https://doi.org/10.1016/j.siny.2013.11.012

Baron, I. S., Kerns, K. A., Müller, U., Ahronovich, M. D., & Litman, F. R. (2012). Executive functions in extremely low birth weight and late-preterm preschoolers: effects on working memory and response inhibition. Child Neuropsychology, 18(6), 586-599. https://doi.org/10.1080/09297049.2011.631906

Böhm, B., Lundequist, A. & Smedler, A. C. (2010). Visual‐motor and executive functions in children born preterm: The Bender Visual Motor Gestalt Test revisited. Scandinavian Journal of Psychology, 51, 376–384.

https://doi.org/10.1111/j.1467-9450.2010.00818.x

Bolk, J., Farooqi, A., Hafström, M., Åden, U., & Serenius, F. (2018). Developmental coordination disorder and its association with developmental comorbidities at 6.5 years in apparently healthy children born extremely preterm. JAMA Pediatrics, 172(8), 765-774. https://doi.org/10.1001/jamapediatrics.2018.1394

Caeyenberghs, K., Leemans, A., Geurts, M., Taymans, T., Vander Linden, C.,

Smits-Engelsman, B. C. M., & Swinnen, S. P. (2010). Brain-behavior relationships in young traumatic brain injury patients: fractional anisotropy measures are highly correlated with dynamic visuomotor tracking performance. Neuropsychologia, 48(5), 1472- 1482. https://doi.org/10.1016/j.neuropsychologia.2010.01.017

Cameron, C. E., Brock, L. L., Murrah, W. M., Bell, L. H., Worzalla, S. L., Grissmer, D., & Morrison, F. J. (2012). Fine motor skills and executive function both contribute to kindergarten achievement. Child Development, 83(4), 1229-1244.

https://doi.org/10.1111/j.1467-8624.2012.01768.x

Central Bureau of Statistics. (2019). Standaard Onderwijsindeling 2016 - Editie 2018/’19. file:///H:/Downloads/PubSoi2016_ed1819.pdf

Central Bureau of Statistics. (2019, June 6). Doodgeboren kinderen; leeftijd moeder, kenmerken geboorte [Data set].

https://opendata.cbs.nl/statline/#/CBS/nl/dataset/82607NED/table?dl=2FD3C

Cunha, A. B., Babik, I., Ross, S. M., Logan, S. W., Galloway, J. C., Clary, E., & Lobo, M. A. (2018). Prematurity may negatively impact means-end problem solving across the first two years of life. Research in Developmental Disabilities, 81, 24-36.

https://doi.org/10.1016/j.ridd.2018.03.007

Diamond, A. (2013). Executive functions. Annual Review of Psychology, 64, 135-168. https://doi.org/10.1146/annurev-psych-113011-143750

De Kieviet, J. F., Piek, J. P., Aarnoudse-Moens, C. S., & Oosterlaan, J. (2009). Motor development in very preterm and very low-birth-weight children from birth to adolescence: a meta-analysis. JAMA, 302(20), 2235-2242.

De Kieviet, J. F., Zoetebier, L., Van Elburg, R. M., Vermeulen, R. J., & Oosterlaan, J. (2012). Brain development of very preterm and very low‐birthweight children in childhood and adolescence: a meta‐analysis. Developmental Medicine & Child Neurology, 54(4), 313-323. https://doi.org/10.1111/j.1469-8749.2011.04216.x

De Kieviet, J. F., Stoof, C. J., Geldof, C. J., Smits, N., Piek, J. P., Lafeber, H. N., & Oosterlaan, J. (2013). The crucial role of the predictability of motor response in

visuomotor deficits in very preterm children at school age. Developmental Medicine & Child Neurology, 55(7), 624-630. https://doi.org/10.1111/dmcn.12125

Edwards, J., Berube, M., Erlandson, K., Haug, S., Johnstone, H., Meagher, M., & Zwicker, J. G. (2011). Developmental coordination disorder in school-aged children born very preterm and/or at very low birth weight: a systematic review. Journal of

Developmental & Behavioral Pediatrics, 32(9), 678-687. https://doi.org/10.1097/DBP.0b013e31822a396a

Ellinoudis, T., Evaggelinou, C., Kourtessis, T., Konstantinidou, Z., Venetsanou, F., Kambas, A. (2011). Reliability and validity of age band 1 of the Movement Assessment Battery for Children - Second Edition. Research in Developmental Disabilities, 32, 1046-1051. https://doi.org/10.1016/j.ridd.2011.01.035

Evensen, K. A. I., Lindqvist, S., Indredavik, M. S., Skranes, J., Brubakk, A. M., & Vik, T. (2009). Do visual impairments affect risk of motor problems in preterm and term low birth weight adolescents?. European Journal of Paediatric Neurology, 13(1), 47-56. https://doi.org/10.1016/j.ejpn.2008.02.009

Geldof, C. J. A., Van Wassenaer, A. G., De Kieviet, J. F., Kok, J. H., & Oosterlaan, J. (2012). Visual perception and visual-motor integration in very preterm and/or very low birth weight children: a meta-analysis. Research in Developmental Disabilities, 33(2), 726-736. https://doi.org/10.1016/j.ridd.2011.08.025

Goyen, T. A., Lui, K., & Woods, R. (1998). Visual‐motor, visual‐perceptual, and fine motor outcomes in very‐low‐birthweight children at 5 years. Developmental Medicine & Child Neurology, 40(2), 76-81.

https://doi.org/10.1111/j.1469-8749.1998.tb15365.x

Halperin, J. M., McKay, K. E., Matier, K., & Sharma, V. (1994). Attention, response inhibition, and activity level in children: Developmental neuropsychological perspectives. In Advances in Child Neuropsychology (pp. 1-54). Springer. https://doi.org/10.1007/978-1-4612-2608-6_1

Henderson S. E., Sugden D. A., & Barnett A. L. (2007). Movement Assessment Battery for Children (2nd ed.). Harcourt Assessment.

Henderson S. E., Sugden D. A., Barnett A. L., Smits-Engelsman B. C. M. (2010). Movement Assessment Battery for children (2nd ed.), Dutch translation. Pearson.

Hendriksen, J. G. M., & Hurks, P. P. M. (2009). WPPSI-III-NL Wechsler Preschool and Primary Scale of Intelligence; Nederlandse bewerking. Pearson.

http://www.pearsonclinical.nl/wppsi-iii-nl-wechsler-preschool-primary-scale-intelligence

Imamizu, H., & Kawato, M. (2009). Brain mechanisms for predictive control by switching internal models: implications for higher-order cognitive functions. Psychological Research, 73(4), 527-544. https://doi.org/10.1007/s00426-009-0235-1

Koekkoek, S., de Sonneville, L. M., Wolfs, T. F., Licht, R., & Geelen, S. P. (2008). Neurocognitive function profile in HIV-infected school-age children. European Journal of Paediatric Neurology, 12(4), 290-297.

https://doi.org/10.1016/j.ejpn.2007.09.002

Logan, G. D., & Burkell, J. (1986). Dependence and independence in responding to double stimulation: A comparison of stop, change, and dual-task paradigms. Journal of

Experimental Psychology: Human Perception and Performance, 12(4), 549-563. https://doi.org/10.1037/0096-1523.12.4.549

Logan, G. D., Cowan, W. B., & Davis, K. A. (1984). On the ability to inhibit simple and choice reaction time responses: A model and a method. Journal of Experimental Psychology: Human Perception and Performance, 10(2), 276-291.

https://doi.org/10.1037/0096-1523.10.2.276

Mulder, H., Pitchford, N. J., Hagger, M. S., & Marlow, N. (2009). Development of executive function and attention in preterm children: a systematic review. Developmental Neuropsychology, 34(4), 393-421. https://doi.org/10.1080/87565640902964524

Navarro, D. (2019). Chapter 15: Linear regression. In Learning statistics with R: A tutorial for psychology students and other beginners - version 0.6 (pp. 457-496). Adelaide.

https://learningstatisticswithr.com

Oudgenoeg-Paz, O., Mulder, H., Jongmans, M. J., Van der Ham, I. J., & Van der Stigchel, S. (2017). The link between motor and cognitive development in children born preterm and/or with low birth weight: A review of current evidence. Neuroscience &

Biobehavioral Reviews, 80, 382-393. https://doi.org/10.1016/j.neubiorev.2017.06.009 Oosterlaan, J., & Sergeant, J. A. (1998). Response inhibition and response re-engagement in

attention-deficit/hyperactivity disorder, disruptive, anxious and normal children. Behavioural Brain Research, 94(1), 33-43.

https://doi.org/10.1016/S0166-4328(97)00167-8

Pascal, A., Govaert, P., Oostra, A., Naulaers, G., Ortibus, E., & Van den Broeck, C. (2018). Neurodevelopmental outcome in very preterm and very‐low‐birthweight infants born over the past decade: a meta‐analytic review. Developmental Medicine & Child Neurology, 60, 342-355. https://doi.org/10.1111/dmcn.13675

R Core Team. (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/.

Rommelse, N. N., Altink, M. E., De Sonneville, L. M., Buschgens, C. J., Buitelaar, J.,

Oosterlaan, J., & Sergeant, J. A. (2007). Are motor inhibition and cognitive flexibility dead ends in ADHD?. Journal of Abnormal Child Psychology, 35(6), 957-967. https://doi.org/10.1007/s10802-007-9146-z

Saigal, S., & Doyle, L. W. (2008). An overview of mortality and sequelae of preterm birth from infancy to adulthood. The Lancet, 371, 261-269.

https://doi.org/10.1016/S0140-6736(08)60136-1

Serenius, F., Ewald, U., Farooqi, A., Fellman, V., Hafström, M., Hellgren, K., & Strömberg, B. (2016). Neurodevelopmental outcomes among extremely preterm infants 6.5 years after active perinatal care in Sweden. JAMA Pediatrics, 170(10), 954-963.

https://doi.org/10.1001/jamapediatrics.2016.1210

Shadmehr, R., Smith, M. A., & Krakauer, J. W. (2010). Error correction, sensory prediction, and adaptation in motor control. Annual Review of Neuroscience, 33, 89-108.

https://doi.org/10.1146/annurev-neuro-060909-153135

Skranes, J., Lohaugen, G. C., Martinussen, M., Indredavik, M. S., Dale, A. M., Haraldseth, O., & Brubakk, A. M. (2009). White matter abnormalities and executive function in children with very low birth weight. Neuroreport, 20(3), 263-266.

https://doi.org/10.1097/WNR.0b013e32832027fe

Soloman, S. R., & Sawilowsky, S. S. (2009). Impact of rank-based normalizing

transformations on the accuracy of test scores. Journal of Modern Applied Statistical Methods, 8(2), 9. https://doi.org/10.22237/jmasm/1257034080

Stålnacke, J., Lundequist, A., Böhm, B., Forssberg, H., & Smedler, A. C. (2019). A

in children born very or extremely preterm. Child Neuropsychology, 25(3), 318-335. https://doi.org/10.1080/09297049.2018.1477928

Stirling, L. A., Lipsitz, L. A., Qureshi, M., Kelty-Stephen, D. G., Goldberger, A. L., & Costa, M. D. (2013). Use of a tracing task to assess visuomotor performance: effects of age, sex, and handedness. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences, 68(8), 938-945. https://doi.org/10.1093/gerona/glt003

Twilhaar, E. S., Wade, R. M., De Kieviet, J. F., Van Goudoever, J. B., Van Elburg, R. M., & Oosterlaan, J. (2018). Cognitive outcomes of children born extremely or very preterm since the 1990s and associated risk factors: a analysis and

meta-regression. JAMA Pediatrics, 172, 361-367. https://doi.org/10.1001/jamapediatrics.2017.5323

Van Hus, J. W., Potharst, E. S., Jeukens-Visser, M., Kok, J. H., & Van Wassenaer-Leemhuis, A. G. (2014). Motor impairment in very preterm‐ born children: links with other developmental deficits at 5 years of age. Developmental Medicine & Child Neurology, 56, 587-594. https://doi.org/10.1111/dmcn.12295

Volpe, J. J. (2009). Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. The Lancet Neurology, 8, 110-124.

https://doi.org/10.1016/S1474-4422(08)70294-1

Wolpert, D. M., Ghahramani, Z., & Jordan, M. I. (1995). An internal model for sensorimotor integration. Science, 269(5232), 1880-1882. https://doi.org/10.1126/science.7569931 World Health Organization. (2018, February 19). Preterm birth.

https://www.who.int/news-room/fact-sheets/detail/preterm-birth

Zago, M., McIntyre, J., Senot, P., & Lacquaniti, F. (2009) Visuo-motor coordination and internal models for object interception. Experimental Brain Research, 192(4), 571-604. https://doi.org/10.1007/s00221-008-1691-3