Reaching to throw compared to reaching to place: A comparison across individuals with and without Developmental Coordination Disorder

Research in Developmental Disabilities 34 (2013) 174–182

Contents lists available at SciVerse ScienceDirect

Research in Developmental Disabilities

Reaching to throw compared to reaching to place: A comparison across individuals with and without Developmental Coordination Disorder Kate Wilmut *, Maia Byrne, Anna L. Barnett Department of Psychology, Oxford Brookes University, Oxford, OX3 0BP, United Kingdom

A R T I C L E I N F O

A B S T R A C T

Article history: Received 13 June 2012 Received in revised form 20 July 2012 Accepted 20 July 2012 Available online 31 August 2012

When picking up an object, adults show a longer deceleration phase when the onward action has a greater precision requirement. Tailoring action in this way is thought to need forward modelling in order to predict the consequences of movement. Some evidence suggests that young children also tailor reaching in this way; however, how this skill develops in children with Developmental Coordination Disorder (DCD) is unknown. The current study compared the kinematics of reaching to an object when the onward intention was: to place the object on a target (either with high or low precision requirements), to throw the object or to lift the object vertically. Movements of both adults (N = 18) and children (N = 24) with DCD and their age-matched controls were recorded. The typically developing adults discriminated across all action types, the adults with DCD and the typically developing children only across the actions to place and throw and the children with DCD only between the actions to lift and throw. The results demonstrate developmental progression towards fine tuning the planning of reaching in relation to onward intentions. Both adults and children with DCD are able to plan movement using inverse models but this skill is not yet fully developed in early adulthood. ß 2012 Elsevier Ltd. All rights reserved.

Keywords: Developmental Coordination Disorder Reach and grasp Forward modelling Movement planning

1. Introduction Every action that we carry out during our daily lives is composed of a string of movements; drinking tea from a cup consists of reaching out and grasping the cup, lifting the cup, moving it towards our mouth, tipping it to an angle that allows us to drink and not spill the tea, and then placing the cup back down. The way in which we grasp the cup and where we place our hand allows both lifting and drinking; if we grasped the cup handle too far down we would need to re-assess our grip before continuing. Therefore, the grasping movement needs to take into account that the next movement is lifting and the one after drinking. This forward anticipation of movement whereby one movement is tailored to the onward action allows a spatially and temporally optimal movement, minimising fatigue, maximising comfort, lowering energy output and achieving accuracy (Haggard, 1998; Rosenbaum, Vaughan, Barnes, & Jorgensen, 1992). One might imagine that aspects of the movement which are tailored to the onward action are the overt aspects of reachto-grasp such as positioning of the hand and fingers on the object, type of grasp used and pressure of grasp. However, it would seem that this is not all that is planned. Marteniuk, MacKenzie, Jeannerod, Athenes, and Dugas (1987) reported that the onward action also effects the precise kinematics of the initial reach-to-grasp trajectory. They asked participants to either reach and fit an object into a hole or reach and throw the object. Initial demands of the reach component were identical

* Corresponding author. Tel.: +44 01865 483 781. E-mail address: [email protected] (K. Wilmut). 0891-4222/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ridd.2012.07.020

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

175

across conditions. However, during the reach phase an elongated deceleration phase was seen for the fit as compared to the throw action and this resulted in a longer movement when grasping to fit compared to grasping to throw (Marteniuk et al., 1987). The authors suggest that the different precision requirements of the two onward actions could be used to explain this; the higher the precision requirements the higher the movement times and deceleration phase. Armbruster and Spijkers (2006) considered a greater range of actions: lifting, throwing and placing an object. They found that peak aperture of the hand was larger and peak deceleration was higher when the grasp was followed by either a throwing or a placing movement as compared to a lift movement. Unlike Marteniuk et al. (1987) they did not see a difference between reach characteristics for the place and the throw conditions. Findings such as these are not only seen for movement kinematics. For example Ansuini, Giosa, Turella, Altoe, and Castiello (2008) and Ansuini, Santello, Massaccesi, and Castiello (2006) considered the grasp preceding a lift movement, a place in a tight niche or a place in a large niche movement. They found that the degree of endgoal accuracy did affect hand shaping during the approach phase, with a distinctly different hand shape across the two ‘place’ movements. Once again this highlights movement precision as a key factor in the mechanism behind fine tuning a movement to an onward action. These findings point towards the importance of tuning each sub-component of an action to optimise the transition from one component to another. This type of fine tuning has been previously demonstrated in co-articulation studies which have shown that sequences of actions are more than the positioning of action elements in succession (e.g., Rosenbaum, 1991). It has been suggested that this concatenation of actions could be explained in terms of inverse models (Ansuini et al., 2006). Inverse models anticipate the consequences of a motor plan prior to execution (Wolpert & Kawato, 1998) and as such could anticipate which specific reaching movement was needed given a subsequent action (Ansuini et al., 2006). Tuning a reaching movement in this way seems to develop in early life, Claxton, Keen, and McCarty (2003) have demonstrated a difference in the initial grasp movement when 10-month-old infants are asked to throw a ball compared to place it in a tube. When reaching to throw, the infants demonstrated a higher peak velocity compared to when they were reaching to place. Furthermore, by 4 years of age children demonstrate an elongated deceleration period for a ‘grasp to fit’ action compared to a ‘grasp to throw’ action (Chen & Yang, 2007). Interestingly, Chen and Yang (2007) also considered children with Cerebral Palsy (CP; who were able to reach-and-grasp and hold an object with their less affected hand), these children did not seem to engage in forward planning their reach movement and showed no difference in reach kinematics across action types (a similar finding has been demonstrated elsewhere: e.g. Mutsaarts, Steenbergen, & Bekkering, 2006). The authors suggest that this may indicate that children with CP lack the ability to use inverse models to anticipate onward action and therefore, sub-movements are performed sequentially without an integrative plan. An alternative explanation is that the lack of difference may be caused by the limited repertoire of movements in the children with CP. Within the population, approximately 2% of children present with difficulties in the execution and coordination of body movements which cannot be accounted for in terms of an intellectual impairment or identifiable physical or neurological disorder (Lingam, Hunt, Golding, Jongmans, & Emond, 2009). This condition is termed Developmental Coordination Disorder (DCD) and problems manifest in difficulties with fine motor tasks such as tracing, writing and fastening buttons, and/or in gross motor tasks such as jumping, hopping and riding a bike (Sugden & Wright, 1998). Many children with DCD continue to exhibit problems throughout adolescence (Cousins & Smyth, 2003; Losse et al., 1991), however, little is known about how these individuals develop into and throughout adulthood. Despite an increasing number of studies focusing on DCD, very little is known about the underlying cause of the movement problems (see Visser, 2003, for a review). However, several studies have highlighted poor motor planning (e.g. Schoemaker et al., 1994), which in part may be explained by a difficulty in the internal modelling of movement used to predict or anticipate the outcome of action (Maruff, Wilson, Trebilock, & Currie, 1999; Smits-Engelsman, Caeyenberghs, van Rood, & Swinnen, 2007; Williams, Thomas, Maruff, Butson, & Wilson, 2006). To date, no research has considered whether children with DCD are able to tailor an initial reaching movement to the end-goal. Therefore, this aspect of forward anticipation needs consideration in a DCD population. The current study considers four different onward actions: throw, lift and two place actions (one with a high precision and one with a lower precision requirement) across children and adults, typically developing and with DCD. Based on previous research, the four actions chosen should give us the best chance of seeing a tailoring to onward intention effect in the typically developing group and whether precision requirements (as suggested by Marteniuk et al., 1987) can be used to explain the differential kinematics across reach movements for different onward actions. Previous studies have used a range of different objects, from tennis balls to disks, all of which have shown effects of onward action. In this study we chose to use a small cylinder. Given previous findings we would expect both typically developing populations to show some differences in the kinematics of the initial reach movement for the throw as compared to the other actions. This will most likely manifest in a longer deceleration phase for the higher precision movements (place vs. throw). The movement difficulties of children with CP are both quantitatively and qualitatively different to those in DCD. However, both populations have demonstrated a difficulty with using inverse models to forward anticipate movement (Maruff et al., 1999; Mutsaarts et al., 2006; SmitsEngelsman et al., 2007) and so we might expect to see a similar pattern of behaviour in the children with DCD as was seen in the children with Cerebral Palsy (Chen & Yang, 2007), whereby the children with DCD fail to tune an initial reaching movement to the onward action. However, the way in which this develops from childhood to adulthood is unknown. This is the first study to consider the exact nature of tailoring initial kinematics to onward action in a DCD population. As such it will provide a greater insight into the nature of motor planning in this population. This study is also one of the first to examine skill performance across children and adults with DCD. These two novel aspects of the study will help to further our understanding of the mechanisms underlying DCD.

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

176

2. Method 2.1. Participants 2.1.1. Adults Eighteen adults with DCD (mean age: 24 years, age range: 18:01 to 29:11) were recruited to take part in this study. Selecting adults with DCD is complicated given that there is no standardised assessment of motor coordination with norms above 21 years of age. Therefore, we used a range of selection tests which we believe allowed us to recruit adults who did meet the DSMIV criteria for DCD. These adults were recruited through a support group for local families and participants that had previously been tested by us as children. Part of the recruitment involved taking a detailed developmental and medical history; all reported current motor difficulties, none indicated any neurological deficit or comorbid condition that would explain their motor difficulties and all but three of the adults with DCD had received a previous diagnosis of DCD or equivalent. In the absence of a suitable motor test for this age group we elected to use two tests: age band 3 of the Movement Assessment Battery for Children 2nd Edition (MABC-2; Henderson, Sugden, & Barnett, 2007) and the brief form of the Bruininks-Oseretsky Test of Motor Proficiency 2nd Edition (Bruininks & Bruininks, 2005). The inclusion criteria were a score below the 15th percentile on both tests. All DCD participants fell below the 5th percentile on the MABC-2 and the 15th percentile on the BOT-2. In addition, all except one participant completed the self-report Adult Developmental Coordination Disorder checklist (ADC, Kirby, Edwards, Sugden, & Rosenblum, 2010). All met the criteria for significant motor difficulties during childhood, which is necessary for a diagnosis of DCD. The ADC has a separate section on self-reported difficulties as an adult, although this does not focus specifically on the motor domain (Barnett & Wilmut, 2012). Even so, seven of the overall scores classified participants as either ‘at risk of DCD’ or ‘probable DCD’. Given the concern regarding the comorbidity between ADHD and DCD, and the impact that inattention may have on motor skill, the Conner’s Adult Rating Scales (CARS: Conners, Erhardt, & Sparrow, 1999) was used to assess ADHD symptoms. Out of the 18 participants 2 were above average on the hyperactivity and inattention subsets of the CARS. Eighteen gender and age-matched adults were recruited from the Oxfordshire area, none of these participants reported movement difficulties either in child- or adulthood, as judged by the ADC. Due to constraints of both time and space it was only possible to administer the manual dexterity subset of the MABC-2 but all scored above the 25th percentile. Both groups of adults had qualifications above the basic UK secondary school level indicating the absence of a marked intellectual impairment. 2.1.2. Children Twenty-four children with DCD (mean age: 8 years, age range: 8:06 to 11:08) were recruited through the database of individuals with DCD held at Oxford Brookes University. Eighteen of these children had received a formal diagnosis of DCD or equivalent. In order to ensure that both these and the undiagnosed children met the DSM-IV criteria for DCD, motor ability was assessed using the appropriate age-band from the MABC-2. In addition, the MABC-2 checklist (completed by a parent) was used to determine impact on daily living. All children in the DCD group fell below the 5th percentile on the test component of the MABC-2, indicating a significant motor difficulty. Twenty-two children scored below the 5th percentile on the MABC-checklist and 2 between the 15th and 5th percentile, indicating that the motor difficulties had a significant impact on daily living. None of the parents of the children with DCD indicated any neurological deficit or comorbid condition which would explain their childs’ motor difficulties. Inattention and hyperactivity was measured using the Strengths and Difficulties Questionnaire (Goodman, 1997). Thirteen of the children with DCD had hyperactivity and inattention scores which were above average for their age and 6 children had borderline hyperactivity and inattention scores. Twenty-four age (within 6 months) and gender matched typically developing children were recruited from local schools. All of the typically developing (TD) children completed the test component of the MABC-2 and all fell above the 16th percentile. On a measure of receptive vocabulary (British Picture Vocabulary Scales; Dunn, Dunn, & Sewell, 2009) all children (TD and DCD) fell within the normal range for their age, suggesting the absence of a marked intellectual impairment. Participant details are given in Table 1. Table 1 Details of each group, including number of participants, age, gender ratio and scores from motor assessments. Children

Adults Typically developing

DCD

Typically developing

DCD

N Mean age Gender ratio f:m MABC-2 checklist Numbers below 5th and 15th percentiles

18 24 years 11 months 7:11

18 25 years 0 months 7:11

24 9 years 1 month 4:20

24 9 years 1 month 4:20

N/A

N/A

N/A

MABC-2 percentile BOT-2 percentile ADC (average score) Childhood Adulthood

60a N/A

1.7 6.9

66 N/A

22 < 5th 2 < 15th 4.2 N/A

6.6 17.2

23.3 47.6

N/A

N/A

a

TD adults only completed the manual dexterity subset of the MABC-2.

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

177

2.2. Materials Participants sat at a desk on which was placed a wooden cylinder 7 cm in height and 2 cm in diameter. The cylinder was placed on the midline, 0.3 times arm length in front of a start node which was to be grasped between the thumb and index finger of the dominant hand at the start of each trial. In each condition, one of four ‘target’ objects was placed 0.2 total arm length from the cylinder, to the right of the cylinder for left handed participants and to the left for right handed participants (see Fig. 1 for exact locations). ‘Target’ objects all consisted of a 8 cm  8 cm wooden square with: a tight single hole with a diameter of 2 cm (the same diameter as the cylinder, for the tight place action), a loose single hole with a diameter of 4 cm (for the loose place action); a dowel which stood perpendicular to the table to a height of 21 cm (for the lift movement); and a 4 cm deep container measuring 8 cm  8 cm (for the throw movement). A Vicon 3D motion capture system (Oxford Metrics, United Kingdom), consisting of six infra-red cameras and running at 120 Hz, was used to track the movement of five reflective markers (6 mm in diameter) placed on the thumb, index finger, knuckle and wrist of the preferred hand. A fifth marker was placed on top of the cylinder. 2.3. Procedure Participants were instructed first to grasp the start object and then to perform one of four movements. Each movement started with the participant reaching out and grasping the cylinder and then either: performing a tight place (in a hole the same size as the cylinder); performing a loose place (in a hole twice the size of the cylinder); lifting the cylinder; or throwing the cylinder. Adults performed 32 trials (8 trials of each action) and children 24 (6 of each). The order of presentation of actions was pseudo-randomised for each participant. Each action was explained to the participant prior to the start of the trial, for both the ‘tight place’ and ‘loose place’ conditions participants were told to place the cylinder in the target hole. For the ‘lift condition’ participants were instructed to grasp the cylinder and lift it up so that the base was approximately the same height as the top of the dowel. In the ‘throw’ condition participants were asked to throw the cylinder into the container. 2.4. Data analysis Trials were excluded if the wrong action was performed or if data lost, due to occlusion of markers, exceeded 10 consecutive frames. In total this resulted in a loss of 6% of trials from the typically developing adults, 11% from the typically developing children, 9% from the adults with DCD and 8% from the children with DCD. Although both the adults and the children with DCD showed attention difficulties they all completed the task appropriately, this was taken as a sign that attention was directed at the task and so all the individuals with DCD were included. VICON hand movement data was

Position of target square for right handers

0.2 times arm length

Cylinder 0.3 times arm length Start point

Participant

Fig. 1. A schematic illustration of the experimental set-up. The set-up shown is for a right handed participant. The target for each action was a wooden 8 cm  8 cm square, this had a small hole for the tight place action, a larger hole for the loose place action, a dowel raised to 20 cm for the lift action and was a tray for the throw action.

178

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

filtered with an optimised Woltring filter (low pass 12 Hz) and tailored MatLab routines were used for analysis. The start and end of the hand movement was determined from velocity curves, the time point at which velocity departed from zero (>3% max velocity) or returned to zero (<3% max velocity) was identified and checked by eye. The start and end point of the initial reach movement allowed us to calculate duration of movement (ms), proportion of time spent decelerating (start of deceleration period determined by peak velocity) and time taken to reach peak acceleration. In addition the maximum distance between the finger and thumb was taken as maximum grip aperture. The kinematic variables were chosen based on those used in previous research, movement duration (Marteniuk et al., 1987), proportion of time spent decelerating (Armbruster & Spijkers, 2006; Chen & Yang, 2007; Claxton et al., 2003; Marteniuk et al., 1987), maximum grip aperture (Armbruster & Spijkers, 2006). The additional variable, time to peak acceleration was calculated as is has been previously shown to be important in the description of prehensile movements and therefore may show a difference across onward actions (Kritikos, Jackson, & Jackson, 1998). Finally, ‘quality’ of the onward action was quantified, this was only done for place movements as quantifying functionality of the lift or throw movement was not deemed possible. For place movements, the amount of time spent adjusting the position of the cylinder was calculated from the marker on the cylinder. Adjustments of the trajectory were defined as secondary peaks in velocity (zero-crossing of acceleration) during the end deceleration phase. The point at which the first final adjustment occurred was taken as the start of the adjustment period, which then lasted until the end of the movement. 3. Results 3.1. Movement duration Movement duration can be found in Table 2. A three-way ANOVA (group  age  action type) found a main effect of group [F(1,80) = 4.56, p = 0.036, h2 = .037] indicating that the individuals with DCD showed a longer movement duration compared to the typically developing individuals. A main effect of action type [F(3,240) = 11.58, p < .001, h2 = .126] and an interaction between action type and group [F(3,240) = 3.09, p = .028, h2 = .037] was also found. In order to fully explore the interaction between action type and group two-way ANOVA (age  action type) considered each group separately. Both the typically developing individuals and the individuals with DCD demonstrated an effect of action type [F(3,120) = 9.43, p < .001, h2 = .196 and F(3,120) = 5.64, p = 0.001, h2 = .123 respectively]. For the DCD group this difference was between the lift and throw movement, for the TD group the difference was between the lift movement and the two place movements. No other significant comparisons were found (p < .05 with Bonferroni correction). The typically developing group also showed an effect of age [F(1,40) = 5.26, p = .027, h2 = .116], whereby the adults moved faster than the children. 3.2. Proportion of time in deceleration The proportion of time spent in deceleration can be found in Fig. 2, this data was considered using a three-way ANOVA (group  age  action type). A main effect of group was found, indicating that the individuals with DCD spent a higher proportion of movement time decelerating compared to the TD individuals [F(1,80) = 4.41, p = .039, h2 = .052] and was higher for the children compared to the adults [F(1,80) = 16.12, p < .001, h2 = .17]. In addition, an interaction between age and group was found [F(1,80) = 8.80, p = .004, h2 = .099]. In order to further explore the interaction two-way ANOVA (group  action type) was carried out for the adults and children separately. For the adults, a main effect of group was found [F(1,34) = 10.49, p = .003, h2 = .236], indicating that the Table 2 Mean movement duration (ms), maximum grip aperture (mm) and time to peak acceleration (ms) in all four action types. Given for typically developing adults, adults with DCD, typically developing children and children with DCD. Standard deviation is given in brackets. Adults

Movement duration (ms) Tight place Loose place Lift Throw Time to peak acceleration (ms) Tight place Loose place Lift Throw MGA (mm) Tight place Loose place Lift Throw

651 647 727 661

Children DCD

TD (75) (90) (104) (107)

145 (45) 146(48) 190 (68) 175 (46) 81.1 80.9 80.2 79.7

(12.6) (12.0) (15.6) (11.1)

TD

DCD

758 755 812 742

(215) (207) (224) (143)

722 706 792 741

(99) (102) (155) (155)

776 803 810 717

(165) (193) (198) (158)

151 152 173 158

(78) (75) (86) (57)

139 120 188 163

(44) (72) (92) (71)

165 175 162 164

(70) (69) (64) (76)

80.8 81.2 80.6 79.9

(14.6) (13.6) (15.4) (13.2)

77.9 77.4 76.3 73.6

(8.4) (7.7) (9.2) (9.8)

84.9 84.4 84.1 84.4

(11.3) (11.7) (11.1) (11.1)

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

179

Proportion of time decelerating (%)

Proportion of time decelerating 64 62 60 58 56 54 52 50 48 46

Children: TD Adults: TD

Children: DCD Adults: DCD

44 Tight Place Loose place

Lift

Throw

Fig. 2. Graph showing the proportion of time spent decelerating across the four different action types (tight place, loose place, lift and throw). Adults are shown with black lines with square markers, children with grey lines and diamond markers. Typically developing individuals are depicted with solid lines and individuals with DCD with dashed lines. Standard error is displayed.

adults with DCD spent a greater proportion of movement time in deceleration compared to the typically developing adults. A main effect of action type was also found [F(3,102) = 10.16, p < 0.001, h2 = .230]. If we consider the DCD and TD separately this main effect persists for both [F(3,51) = 3.94, p = .013, h2 = .188 and F(3,51) = 8.26, p < .001, h2 = .327 respectively]. Post hoc tests show that for the typically developing adults this difference lies between the tight place, the loose place and the throw action (tight place > loose place = lift > throw). In contrast, for the adults with DCD the difference lies only between the throw action and the other three actions (tight place = loose place = lift > throw) (all, p < .05 with Bonferroni correction). The child groups showed no main effect of group indicting that both child groups spent an equivalent proportion of time decelerating. An interaction between action type and group was found [F(3,102) = 2.45, p = .066, h2 = .051]. The effect of action, spending longer decelerating for the throw action compared to the other three actions, persisted when just comparing the typically developing children [F(3,69) = 6.82, p < .001, h2 = .229], however, no significant effect of action type was seen when considering the children with DCD. 3.3. Time to peak acceleration For the time of peak acceleration, which can be found in Table 2, a three-way ANOVA (group  age  action type) found no main effect of group, however, a main effect of action type [F(3,240) = 5.71, p = .001, h2 = .067] and an interaction between action type and group was found [F(3,240) = 4.27, p = .006, h2 = .051]. In order to explore this interaction two-way ANOVA (group  action type) was considered for each age group separately. The adults showed no effect of group, but did show an effect of action type [F(3,102) = 5.24, p = .002, h2 = .134] which only persisted in the typically developing adults when they were considered separately [F(3,51) = 7.91, p < .001, h2 = .318]. This effect was due to an earlier peak acceleration time for the place action compared to the lift and throw action. The children, again did not show an effect of group but did show an interaction between action type and group [F(3,128) = 4.15, p = .007, h2 = .083]. This effect persisted when just comparing the typically developing children [F(3,69) = 6.05, p = .001, h2 = .208] and again this was due to an earlier peak acceleration time for the place actions compared to the lift or throw action. No significant effect of action type was seen when considering the children with DCD. 3.4. Maximum grip aperture Maximum grip aperture (MGA) was also considered using a three-way ANOVA (group  age  action type), this data can be found in Table 2. No overall main effect of group was found, however, a main effect of action type was found [F(3,240) = 2.77, p = .042, h2 = .033]. Using post hoc tests, it was unclear where this difference lay for MGA (due to the correction rate when using Bonferroni no significant differences were seen across action type). In order to fully explore the developmental effects of both groups, two-way ANOVA (group  action type) was considered for each age group separately. The adults showed no effects in terms of either group or action type. In contrast, the children showed a significant effect of group [F(1,46) = 9.32, p = .004, h2 = .167], whereby the children with DCD used larger MGA’s compared to the typically developing children. No effect of action type was found. 3.5. Movement outcome Finally, we attempted to quantify the outcome of the onward action, this was only done for the two place movements and can be found in Fig. 3. Time spent adjusting was analysed using a 3-way ANOVA (action type, tight and loose  group  age).

180

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

Adjustment time for the two place movements 400 Adults TD Adults DCD Children TD Children DCD

Adjustment time (ms)

350 300 250 200 150 100 50 0 Tight

Loose

Fig. 3. Quality of place movements: mean time spent adjusting just prior to placement (ms). Graph shows all four groups and error bars show standard error of the mean.

A main effect of action type [F(1,80) = 22.16, p < .001, h2 = .217], group [F(1,80) = 13.46, p < .001, h2 = .144] and age [F(1,80) = 12.50, p = .001, h2 = .135] was found. These results demonstrated that the time spent adjusting was longer for a tight versus a loose place action, longer for individuals with DCD compared to TD individuals and longer for children compared to adults. If we compare each group of participants (TD adults, adults with DCD, TD children, children with DCD) we see a significant effect of group [F(1,80) = 8.66, p < .001, h2 = .25], with post hoc tests showing that the typically developing adults show a significantly shorter adjustment time compared to the other three groups (p < .05 with Bonferoni correction). The analysis above looks at how long each group takes to place an object, however, it does not consider whether the kinematics of the initial reach movement changes the functionality of the onward action. If one differentiates between how they grasp the cylinder in a loose place vs. a tight place movement do they show a more functional movement? In order to explore this we calculated the percentage change in the proportion of movement time spent decelerating from a tight place to a loose place and compared this to the average adjustment time across tight and loose place trials. A Pearson correlation found a significant negative relationship for the typically developing adults (r = 0.532, p = 0.023), indicating that as adults showed a greater percentage change in deceleration period from tight to loose place they showed a shorter adjustment time. No significant relationship was seen for the typically developing children, the adults with DCD or the children with DCD. 4. Discussion Developmental Coordination Disorder is characterised by a difficulty in the smooth coordination of fast movements. Hence, it should be no surprise when we see differences in the kinematics of their reaching movements as compared to reaching movements of typically developing individuals. In this study we have seen that the individuals with DCD (adults and children) show longer movement duration and spend more time decelerating as compared to the typically developing individuals (adults and children), similar findings have been demonstrated with pointing paradigms (Wilmut & Wann, 2008). In addition, the typically developing children used smaller maximum grip apertures as compared to the children with DCD. In terms of age differences typically developing individuals showed a clear developmental trajectory, whereby both movement duration and the proportion of time spent decelerating decreased as age increased. In contrast, the individuals with DCD showed no developmental effect in terms of either movement time or proportion of movement spent decelerating. The primary aim of this study was to consider the way in which both typically developing individuals and individuals with DCD tailor an initial reaching movement to an onward action. If we start by just considering the typically developing adults, they demonstrated differential movement kinematics across the four movement types. More specifically, the adults showed a shortened proportion of time spent decelerating for a throw action as compared to a place or lift movement. On the surface this finding reflects that of Marteniuk et al. (1987), however, they considered the time decelerating not the proportion of movement time spent decelerating. In fact the adults in the Marteniuk et al. study showed a decrease in time decelerating which was accompanied by a decrease in movement time. Whether they would have seen a decrease in the proportion of the deceleration period is unclear. Marteniuk et al. (1987) and Ansuini et al. (2006) suggest that it is the different precision requirements of the onward actions which result in the differences in deceleration phase and grasp shape. In this study we considered two onward actions with identical movement requirements aside from the precision element (tight and loose place). The typically developing adults discriminate between these and spend a greater proportion of time decelerating for the high precision action (tight place) as compared to the low precision action (loose place). This supports the idea that it is the precision difference across actions which drives the fine tuning of an initial reach movement. Other kinematic variables

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

181

were also seen to differ across actions, for example: the time of peak acceleration was later for the throw and lift as compared to the two place movements; and movement duration was longer for the lift as compared to the throw movement (in line with Armbruster & Spijkers, 2006). Interestingly, Fleming, Klatzky, & Behrmann (2002) found no systematic effects of the different onward action on the initial reach-to-grasp movement when comparing a lift movement to a posting movement. Given the findings of the current study, and the past research, it would seem that typically developing adults do show differential kinematics in an initial reach movement for specific onward actions but studies which only include onward actions with similar precision requirements may not reveal these differences. When considering the typically developing children, they show a shortened proportion of movement time in deceleration when throwing as compared to the other three actions; however, unlike the typically developing adults, they show no discrimination between the place movements. In terms of discriminating between a throw and a place movement, this is in line with the findings of Chen and Yang (2007) who considered a younger cohort of children. The lack of an effect between the two place movements may suggest that tuning of the length of the deceleration period to onward action is refined between early childhood and adulthood; children do not make subtle distinctions between essentially the same actions albeit with different requirements. Building on the Chen and Yang study, we have also seen that the timing of the peak acceleration is earlier for the two place movements compared to the throw and lift movement and that movement duration was longer for the lift as compared to the throw movement. These effects are in line with what is seen in typically developing adults. Interestingly, in terms of the proportion of time spent decelerating, the adults with DCD show the same pattern as the typically developing children. They show a shortened proportion of movement time in deceleration for the throw action as compared to the other three actions, but they show no distinction between the two place actions. The adults with DCD show no difference in the timing of peak acceleration, but they do show a difference in the timing of the maximum grip aperture and a difference in movement duration. The maximum grip aperture took longer to achieve for the low precision movements such as throw and lift compared to the high precision place movements. Movement duration was longer for the lift as compared to the throw movement. These findings, especially those relating to the proportion of time spent decelerating, suggest that adults with DCD are able to concatenate actions, using a movement plan to link together sub-movements and account for the onward action. If they were not able to do this at all, we would not see any differences in the kinematics of the initial reaching movement across different onward actions. If we accept Ansuini et al.’s (2006) conclusion, that this linking of sub-movements requires the use of inverse models then we are essentially saying that adults with DCD are able to use internal forward modelling, even if it is not as sophisticated as that seen in typically developing adults. Finally we can consider the children with DCD. These children did show a difference in the movement duration of the initial reaching movement, whereby it was longer for a lift action as compared to a throw action. However, no other kinematic variables which described the initial reach movement differed across onward actions. This finding, as for the adults with DCD, does suggest that these children are able to forward anticipate movement, because they have done it in terms of movement duration for a lift as compared to a throw. Taking this finding, alongside the findings in adults with DCD, it would seem that individuals with DCD are able to use inverse models in order to forward anticipate movement. What is unclear at the moment however, is whether this ever reaches the level of sophistication seen in typically developing adults (ability to discriminate between place movements with different precision requirements) and whether it is driven by different underlying mechanism(s). However, we would speculate that if a skill does not reach maturity in early adulthood it is unlikely to show further improvement. This suggests that the control of action in DCD may be different to the typically developing population rather than simply delayed. The final point that needs to be addressed is the quality or functionality of the onward action. If the children and adults with DCD are failing (even if only in some respects) to concatenate movement does this actually result in a less functional movement? To address this point, we quantified movement quality for the place actions only. The typically developing adults demonstrated a relationship between the degree to which they discriminated between the place movements during the initial reach action and the time spent adjusting during the place action; a greater tuning for onward action led to a place movement which required less final adjustment. This effect was not seen in the other groups but does hint at some association between tailoring to an onward action and producing a functional movement. Thus where individuals with DCD fail to tailor initial movements this may result in some of the characteristic movement difficulties. This study has demonstrated that both adults and children with DCD seem able to use inverse modelling to anticipate movement and concatenate action. However, this is not as sophisticated as that seen in either typically developing adults or typically developing children. This less refined motor planning, especially in the children with DCD, was associated with a less functional movement. From these findings we suggest that the underlying forward anticipation of action is different, rather than simply delayed in DCD.

Acknowledgements This study was funded by an ESRC grant awarded to Kate Wilmut (RES-061-25-0472). We would like to thank all of the participants who took part in this study, Ian Wilmut who made the equipment and Prof. John Wann who kindly lent the VICON system.

182

K. Wilmut et al. / Research in Developmental Disabilities 34 (2013) 174–182

References Ansuini, C., Giosa, L., Turella, L., Altoe, G., & Castiello, U. (2008). An object for an action, the same object for other actions: Effects on hand shaping. Experimental Brain Research, 185, 111–119. Ansuini, C., Santello, M., Massaccesi, S., & Castiello, U. (2006). Effects of end-goal on hand shaping. Journal of Neurophysiology, 95, 2456–2465. Armbruster, C., & Spijkers, W. (2006). Movement planning in prehension: Do intended actions influence the initial reach and grasp movement? Motor Control, 10(4), 311–329. Barnett, A., & Wilmut, K. (2012). The assessment of adults with DCD: Do we have the tools? Edinburgh: DCD Research UK. Bruininks, R. H., & Bruininks, D. D. (2005). Bruininks-Oseretsky Test of Motor Proficiency (second edition (BOT-2) brief form). MN, USA: Pearson. Chen, Y.-P., & Yang, T.-F. (2007). Effect of task goals on the reaching patterns of children with Cerebral Palsy. Journal of Motor Behavior, 39(4), 317–324. Claxton, L. J., Keen, R., & McCarty, M. E. (2003). Evidence of motor planning in infant reaching behavior. Psychological Science, 14(4), 354–356. Conners, C. K., Erhardt, D., & Sparrow, M. A. (1999). Conners’ Adult ADHD Rating Scales (CAARS). New York: Multihealth Systems Inc. Cousins, M., & Smyth, M. M. (2003). Developmental coordination impairments in adulthood. Human Movement Science, 22, 433–459. Dunn, L. M., Dunn, D. M., & Sewell, J. (2009). British Picture Vocabulary Scale (3rd edition). London: GL Assessment. Fleming, J., Klatzky, R. L., & Behrmann, M. (2002). Time course of planning for object and action parameters in visually guided manipulation. Visual Cognition, 9(4/ 5), 502–527. Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38, 581–586. Haggard, P. (1998). Planning of action sequences. Acta Psychologica, 99, 201–215. Henderson, S., Sugden, D., & Barnett, A. (2007). Movement Assessment Battery for Children (second edition). Oxford: Pearson. Kirby, A., Edwards, L., Sugden, D., & Rosenblum, S. (2010). The Development and Standarisation of the Adult Developmental Coordination disorders/dyspraxia checklist (ADC). Research in Developmental Disabilities, 31(1), 131–139. Kritikos, A., Jackson, G. M., & Jackson, S. R. (1998). The influence of initial hand posture on the expression of prehension parameters. Experimental Brain Research, 119, 9–16. Lingam, R., Hunt, L., Golding, J., Jongmans, M., & Emond, A. (2009). Prevalence of Developmental Coordination Disorder using the DSM-IV at 7 Years of Age: A UK Population Based Study. Pediatrics, 123, e693–e700. Losse, A., Henderson, S., Elliman, D., Hall, D., Knight, E., & Jongmans, M. (1991). Clumsiness in children—Do they grow out of it? A 10 year follow up study. Developmental Medicine and Child Neurology, 33, 55–68. Marteniuk, R. G., MacKenzie, C. L., Jeannerod, M., Athenes, S., & Dugas, C. (1987). Constraints on human arm movement trajectories. Canadian Journal of Psychology, 41, 365–678. Maruff, P., Wilson, R., Trebilock, M., & Currie, J. (1999). Abnormalities of imagined motor sequences in children with developmental coordination disorder. Neuropsychologia, 37, 1317–1324. Mutsaarts, M., Steenbergen, B., & Bekkering, H. (2006). Anticipatory planning deficits and task context effects in hemiparetic cerebral palsy. Experimental Brain Research, 172, 151–162. Rosenbaum, D. A. (1991). Human motor control. San Diego, CA: Academic Press Inc. Rosenbaum, D. A., Vaughan, J., Barnes, H. J., & Jorgensen, M. J. (1992). Time course of movement planning: Selection of hand grips for object manipulation. Journal of Experimental Psychology: Learning, Memory and Cognition, 18, 1058–1073. Schoemaker, M. M., Schellekens, J. M. H., Kalverboer, A. F., & Kooistra, L. (1994). Pattern drawing by clumsy children: A problem of movement control? In (Series Ed.) & M. L. Simner, W. Hulstijn, & P. Girouard (Vol. Eds.), Monograph of the Association of Forensic Examiners: Vol. 1. Contemporary issues in the forensic, developmental and neurological aspects of handwriting (pp. 45–65). . Smits-Engelsman, B. C. M., Caeyenberghs, K., van Rood, D., & Swinnen, S. P. (2007). Children with Developmental Coordination Disorder and learning disabilities (DCD/LD) show a reduced ability to imagine motor task difficulty. Paper presented at the 7th international conference on developmental coordination disorder. Sugden, D., & Wright, H. C. (1998). Motor coordination disorders in children. London: SAGE Publications. p. 39. Visser, J. (2003). Developmental coordination disorder: A review of research on subtypes and comorbidities. Human Movement Science, 22, 479–493. Williams, J., Thomas, P. R., Maruff, P., Butson, M., & Wilson, P. H. (2006). Motor, visual and egocentric transformations in children with Developmental Coordination Disorder. Child: Care, Health and Development, 32(6), 633–647. Wilmut, K., & Wann, J. P. (2008). The use of predictive information in goal directed action in children with Developmental Coordination Disorder. Experimental Brain Research, 191(4), 403–418. Wolpert, D. M., & Kawato, M. (1998). Multiple paired forward and inverse models for motor control. Neural Networks, 11, 1317–1329.