Exp Brain Res (2010) 204:21–32 DOI 10.1007/s00221-010-2287-2
RESEARCH ARTICLE
Developmental aspects of pluriarticular movement control Isabelle Mackrous • Luc Proteau
Received: 1 September 2009 / Accepted: 3 May 2010 / Published online: 30 May 2010 Ó Springer-Verlag 2010
Abstract Precise pluriarticular movement control is required to perform straight and smooth out-and-back movements. Our goal was to determine whether children perform out-and-back movements as accurately as adults do in the presence and absence of visual feedback. To reach our goal, 36 children aged between 6 and 12 years, and 12 young adults, performed an out-and-back movement in a normalvision condition and in a target-only condition. Reversal angle and overlapping error were taken to represent the ability of children to control pluriarticular movement. The results showed that adults exhibited sharper movement reversal than the three children groups did, but only for eccentric targets relative to their midline. This suggests that pluriarticular movement control improved across the course of development for eccentric regions of the workspace. Visual feedback did not result in sharper movement reversal even when relatively large errors were noted (eccentric targets in children). This underlines the relatively minor role of visual feedback for interjoint coordination when proprioception is intact. Finally, we observed that directional variability was smaller at the 100-ms mark for the back than for the out portion of the movement, suggesting that movement-planning processes appear less variable when based on dynamic rather than static afferent information. Keywords Movement control Interjoint coordination Manual aiming Proprioception Visual feedback Children I. Mackrous L. Proteau (&) De´partement de kine´siologie, Universite´ de Montre´al, C.P. 6128, Succursale Centre Ville, Montreal H3C 3J7, Canada e-mail:
[email protected] I. Mackrous e-mail:
[email protected]
From the first spontaneous aiming movement in infants to the accurate goal-directed movements of adults, the processes underlying movement dynamics and movement kinematics evolve tremendously. In reaching, mature kinematic patterns are characterized by the emergence of invariant features of hand trajectory (straight-line trajectory, movement peak velocity scaled to movement extent, and bell-shaped velocity profile; Fetters and Todd 1987; von Hofsten 1979, 1991; Konczak et al. 1995; Mathew and Cook 1990). Invariant kinematic features are mirrored by changes in the control of movement dynamics. For instance, Konczak et al. (1995) demonstrated that the emergence of adult-like hand kinematics in 9-month-old infants resulted from a better control and a better use of external (i.e. gravitational) and reactive forces (i.e. transmitted torques) acting on the hand during movement execution. Efficient control of movement dynamics results from the central nervous system (CNS) anticipating the consequences of biomechanical factors, such as external and reactive forces, and coordinating the motor commands accordingly (Konczak et al. 1995; Sainburg et al. 1993, 1995). In pluriarticular movements, the motion of any limb segment produces forces at other mechanically coupled segments, which must be anticipated by the CNS to ensure optimal performance (Gordon et al. 1994a; Flanagan and Lolley 2001; Sainburg et al. 2003; Sober and Sabes 2003). For example, in out-and-back planar reaching movements, the CNS needs to coordinate motion of the elbow with that of the shoulder while accounting for the transmitted torques between these two segments at movement reversal. Sainburg et al. (1995) showed that sharp movement reversal and good overlapping of the out and the back portions of the movement required a tight coupling in the occurrence of movement reversal of the shoulder and the
123
22
elbow (see also Ghez and Sainburg 1995; Sainburg et al. 1993 for a similar observation). Our first goal was to gain insight into the developmental aspect of how the CNS coordinates movement reversal in out-and-back reaching movements by determining whether movement reversal and overlapping errors show a developmental trend. In healthy adults, Sainburg et al. (1993) showed that the presence of visual feedback did not enhance the sharpness of movement reversal, which led the authors to conclude that visual feedback does not play an important role, if any, in interjoint coordination. However, visual feedback increased sharpness of movement reversal in deafferented patients for whom proprioceptive feedback was severely degraded. This observation suggests that visual feedback might play an important role for interjoint coordination when the processing of proprioceptive information is not mature and/or when visual feedback is the preferred source of afferent information for movement control. In that regard, using an upper limb matching task (for example, the participant must indicate whether the position of one arm matches that of his or her other arm), Elliott et al. (1988) reported a linear improvement in matching performance for children aged between 5 and 10 years, whereas Goble et al. (2005) reported an increase in performance of participants aged between 8–10 and 16–18 years. Moreover, in a manual aiming task, it has been repeatedly proposed that 7–8-year-old children relied more heavily on visual feedback than younger or older children for movement control (Chicoine et al. 1992; Ferrel-Chapus et al. 2002; Hay 1979; Orliaguet 1985, 1986). Taken together, these two lines of evidence suggest that visual feedback might play a more important role for interjoint coordination in children than it does in adults. Our second goal was to determine whether and how visual feedback enhances performance of out-and back reaching movements in children. To reach our goals, healthy children and young adults performed overlapping out-and-back movements from a common starting base. All participants performed the task in both a normal-vision and target-only condition (only the target is visible during movement execution). Because Sainburg et al. (1995; see also Sainburg et al. 1993) have already shown that a deficit in interjoint coordination resulted in an increase in reversal error and difficulty in overlapping the out and back portions of the movement, we chose these two kinematic variables to evaluate interjoint coordination in adults and in children. If interjoint coordination is not as well developed in children as it is in adults, a larger reversal angle and lesser overlapping of the out and the back movements should be observed more in children than in adults. In addition, if children show a sharper reversal angle and a better overlapping of the out and back trajectories when visual feedback is available as opposed to when it is not, it would
123
Exp Brain Res (2010) 204:21–32
suggest that vision facilitates interjoint coordination in children.
Method Participants Thirty-six children (n = 12 for 6–7, 8–9, and 10–12 years old) and 12 adults (aged between 20 and 25 years) took part in this experiment. They all reported normal or corrected-to-normal vision. They took part in a single 30-min experimental session. Adults were paid $10 CDN for their time; children received a toyshop’s gift card for the same amount. The Health Sciences Ethics Committee of the Universite´ de Montre´al had approved this study. Task and apparatus Participants performed out-and-back movements with a computer-mouse-like device from a fixed starting position located close to the body toward a target located away from the body. They were instructed to perform straight and uncorrected movements (i.e., no ‘‘stop and go’’) and to overlap the out and back portions of their movements as accurately as possible. The apparatus is illustrated in Fig. 1a. It consisted of a table, a computer screen, a mirror, a headrest, and a twodegree-of-freedom manipulandum. Participants sat close to the table so that their body rested along the table leading edge. The computer screen (Mitsubishi, Color Pro Diamond, 37 inches) was mounted on a ceiling support positioned directly over the table; the computer screen was oriented parallel to the surface of the table. Its image was reflected on a mirror placed directly beneath it and parallel to the tabletop. The mirror was located at midpoint between the computer screen and the mirror, which were 37 cm apart. A headrest was affixed to the proximal side of the computer screen. The participants’ chair was adjusted in height so that their forehead could rest comfortably on the headrest. Participants could not see their hands and arms for the duration of the experiment, but the information presented on the computer screen (cursor, starting base, and target) was reflected on the mirror and was visible to the participant. The tabletop was covered by a piece of Plexiglas. The manipulandum consisted of two pieces of rigid Plexiglas (43 cm) joined at one end by an axle. One free end of the manipulandum was fitted with a second axle encased in a stationary base affixed to the tabletop. The other free end of the manipulandum was fitted with a small vertical shaft (length: 3 cm, radius: 1 cm), i.e. the stylus, which could be easily gripped by the participant. The starting base
Exp Brain Res (2010) 204:21–32
23
Fig. 1 a Experimental set up. b The reversal angle was calculated as the angle formed by the position of the cursor 100 ms prior to movement reversal, at movement reversal, and 100 ms after movement reversal. The area between the out and back portions of the movement was used to compute the overlapping error. c Orientation variability of the back portion of the movement was calculated in
relation to a vector joining the location of movement reversal and the starting base (100 ms: 100 ms after movement reversal; vel.: peak velocity; dec.: peak deceleration; end: movement endpoint occurred when the cursor was within 1 mm of the piece of Plexiglas defining the starting base)
consisted of a thin strip of Plexiglas glued to the tabletop. It was parallel to the leading edge of the table and had a small indentation on one of its faces. The indentation was located directly in line with the lateral center of the computer screen and the participant’s midline. Each axle of the manipulandum was fitted with a 13-bit optical shaft encoder (US Digital, model S2-2048, sampled at 500 Hz, angular accuracy of 0.0439°), which enabled us to track the displacement of the stylus online and to illustrate it using a 1:1 ratio on the computer screen. Moving the stylus away from the body in the frontal and sagittal planes resulted in an identical displacement of the cursor on the computer screen. The bottom of the stylus and the optical encoder located at the junction of the two arms of the manipulandum were covered with a thin piece of Plexiglas. Lubricating the working surface at the beginning of each experimental session enabled participants to smoothly displace the stylus.
smooth single-motion movements (out and back) with an easily identifiable reversal point. When movements were completed outside the target-movement-time bandwidth, the experimenter reminded the participant of the targetmovement time. A movement-time bandwidth was used to reduce the possibility of different speed-accuracy tradeoffs between the age groups and experimental conditions (Fitts 1954). Thus, differences in performance would mainly be expressed on the spatial components of the movements. The experimental session began with familiarization trials aimed at three targets (six trials per target) located to the right of the participant’s midline and performed in a normalvision condition. At the end of this phase, all participants understood what was expected of them. Because the targets used in this phase were located in the participants’ right hemifield (i.e., outside the experimental workspace), there should have been no, or very little, transfer of learning for the targets used in the following two experimental phases (Gandolfo et al. 1996; Sainburg et al. 1999). In the experimental phases, the targets of interest were located in line with the participants’ midline (0° target) and 35° to their left (-35° target), both at 150 mm from the starting base. These targets were presented randomly, and no knowledge of result (KR) was provided. To maintain the participant’s motivation and ensure calibration of proprioceptive feedback, knowledge of results (KR, illustration of the out-and-back movement on the computer screen) was given every six trials. For the KR trials, we used a third target located 35° to the right of the participants’ midline and at 150 mm from the starting base. Until movement initiation, participants could see the cursor they had to move resting on the starting base and the target to be reached. Participants first performed the task in the target-only condition for which vision of the cursor was blanked at movement onset. Then, participants performed
Procedures Participants used their right dominant hand (self-declared). The targets were located in line with the participants’ midline (0° target) and 35° to their left (-35° target), both at 150 mm from the starting base. The cursor (red, 3 mm in diameter) and the targets (black, 6 mm in diameter) were presented on a white background. Participants were asked to initiate their movement as they pleased following presentation of a target but were required to complete the movement in a movement time ranging between 880 and 1,120 ms (1,000 ms ± 12%; Chicoine et al. 1992; Lhuisset and Proteau 2004a). Movement initiation was detected when the stylus was moved by 2 mm, whereas movement end occurred when the stylus was not displaced by more than 2 mm for a period of 150 ms. The procedure used to define movement endpoint encouraged participants to produce
123
24
the task in the normal-vision condition for which the cursor remained visible throughout movement execution. In both the target-only and normal-vision conditions, participants performed 17 trials toward each of the 0° and the -35° targets and six trials toward the ?35° target. Data reduction The tangential displacement data of the stylus over time for the out and back portions of the movement were first smoothed using a second-order recursive Butterworth filter with a cut-off frequency of 10 Hz. The filtered data were then numerically differentiated once using a central finite technique to obtain the velocity profile of the movement, a second time to obtain the acceleration profile, and a third time to obtain a jerk profile. From these profiles, we determined when and where key kinematic landmarks occurred for each portion of the movement: peak acceleration, peak velocity, peak deceleration, and movement reversal (out) or movement endpoint (back). To facilitate reading of this article, the statistical analyses that were computed are defined at the beginning of each subsection of the results presentation. Geisser– Greenhouse correction was applied when Epsilon value was less than 1. All significant main effects involving more than two means were broken down using Dunn’s technique. Significant interactions were broken down by computing simple main effects, which were followed by Dunn’s post hoc comparisons when they involved more than two means. All effects are reported at p \ .05 (adjusted for the number of comparisons). Dependent variables For the out portion of the movement, direction accuracy was considered as the angular difference between a reference vector (joining the starting base and the target) and the vector defined by the starting base and the cursor at movement reversal. Movement length was the vector joining the starting base and the cursor at movement reversal. Two kinematic measures of performance were computed (Krakauer et al. 1999; Sainburg et al. 1993, 1995): sharpness of reversal and overlapping of the out and back portions of the movement. Sharpness of reversal was represented by the angle between the out and back portions of the movement (Fig. 1b). Locations of the cursor 100 ms before reversal and 100 ms after reversal were used to calculate this angle. Overlapping error was the size of the area between the out and back portions of the movement (Fig. 1b). To normalize for movement length, the overlapping error was divided by the vector length of the out portion of the movement. Measures of within-participant variability were also computed for these dependent variables.
123
Exp Brain Res (2010) 204:21–32
In addition, within-participant variability of movement orientation was computed at four kinematic landmarks for both the out and the back portions of the movements (100 ms after movement initiation [out] or reversal [back]), peak velocity, peak deceleration, and movement reversal (out) or movement endpoint (back). Movement endpoint was defined as the location of the stylus when it came within 1 mm of the Plexiglas strip encasing the starting base (see Fig. 1c). For the back movement, deviation from a new reference vector joining movement reversal point and the starting base was used to compute within-participant variability.
Results Spatial accuracy and variability of the out portion of the movement Figure 2a illustrates out-and-back movements performed in the normal-vision and the target-only conditions by a typical participant in each age group. As illustrated in Fig. 2b, adults and children performed continuous movements as revealed by the smooth bell-shaped velocity profiles observed during the out and back portions of the movement. Figure 3 illustrates the results of all children aged 6–7 and of all adults and show that there was relatively small between-participant variability in the observed movement patterns. Because sharpness of movement reversal might be influenced by target location [Sainburg et al. (1995) showed that intersegmental torque at movement reversal is influenced by target location], we determined first whether the location of movement reversal for each target was approximately the same for the different groups and experimental conditions. To this end, spatial accuracy (direction constant error [i.e. signed error] and movement length) and variability data (direction and length variable errors [i.e. within-participant variability]) of the out portion of the movement were contrasted between age groups and visual feedback conditions. The data of interest were individually submitted to an ANOVA contrasting 4 age groups (children aged 6–7, 8–9, and 10–12, and adults) 9 2 visual feedback conditions (target-only and normalvision condition) 9 2 targets (-35° and 0°) with repeated measurements on the last two factors. The location of movement reversal showed small orientation (\2.5°) and vector length error (\15 mm) for all age groups, who did not significantly differ from one another (see Lhuisset and Proteau 2004a, 2004b for similar observations). However, in the target-only condition, movements ended to the left of the -35° target and to the right of the 0° target (-0.87° and 0.59°, respectively), whereas no such
Exp Brain Res (2010) 204:21–32
25
Fig. 2 a Examples of out and back trajectories for all age groups. Note the sharp movement reversal and good overlapping of out and back trajectories for all age groups for the 0° target, but only for adults for the -35° target. b Out-and-back movements were performed smoothly for all age groups as revealed by smooth and bell-shaped velocity profiles
Fig. 3 a Mean trajectory (thick line) and standard deviation (shadowed area) representing the out-and-back movements separately for the youngest children group (6–7 years old) and for the adults. Note the larger variability for the targetonly condition in comparison with the normal-vision condition. b Mean velocity profile (thick line) and standard deviation (shadowed area) for the youngest children group and for the adults
difference was noted in the normal-vision condition (-0.72° and -0.70°, respectively), F (2, 88) = 9.3. Similarly, in the target-only condition, we observed shorter movements for the -35° target than for the 0° target (153 vs. 163 mm, respectively), whereas no difference was noted in the normal-vision condition (157 vs. 160 mm), F (1, 44) = 30.4.
Movement length variability was significantly larger for 8–9 year-old children than it was for adults in the target-only condition, whereas in the normal-vision condition, 6–7year-old children were more variable than adults were, F (3, 44) = 3.7. Direction variability did not differ significantly between age groups, p = 0.19.
123
26
Reversal and overlapping To determine whether movement reversal and overlapping differed as a function of age and visual feedback, we submitted the reversal angle and the normalized overlapping error data, along with their within-participant variability, to an ANOVA contrasting 4 age groups (children aged 6–7, 8–9, and 10–12, and adults) 9 2 visual feedback conditions (target-only and normal-vision condition) 9 2 targets (-35° and 0°) with repeated measurements on the last two factors. As illustrated in Fig. 4a, no significant age group difference was noted for the 0° target (p [ 0.92). However, for the -35° target, adults had a sharper reversal angle than did all groups of children, who did not significantly differ from one another (p [ 0.90). This is supported by a significant Age group 9 Target interaction, F (3, 44) = 3.4, p = 0.03. Visual feedback had no significant impact on this dependent variable (normal-vision: 9.5°; target-only: 10.7°), F (1, 44) = 3.4, p = 0.07. No significant difference in overlapping error was observed between age groups for the 0° target (p [ 0.06, see Fig. 4b). However, adults had a significantly smaller overlapping error for the -35° target than did all groups of children, who did not significantly differ from one another (p [ .21). This is supported by a significant Age group 9 Target interaction F (3, 44) = 6.9. For all age groups, visual feedback permitted participants to significantly reduce their overlapping error in comparison to the targetonly condition (Fig. 4b), F (1, 44) = 38.2. To better illustrate how visual feedback permitted 6–7-year-old
Fig. 4 a Movement reversal angle. b Normalized overlapping error. c Withinparticipant variability in movement reversal angle and d in normalized overlapping error area for all age groups as a function of the visual feedback condition and target location. Note the larger error and variability for children when compared to adults when aiming at the -35° target. Note also the similar pattern of results for movements performed in the normal-vision and the targetonly conditions
123
Exp Brain Res (2010) 204:21–32
children, and adults, to reduce the overlapping error, Fig. 5 shows deviation of the back portion of the movement from a straight line joining the location of movement reversal to the starting base. For the adults, we observed very little difference between the normal-vision and target-only conditions up to approximately 30% of movement time. After that, deviations from the reference vector increased somewhat in the target-only condition. This increase was larger for the -35° target than it was for the 0° target. This clearly indicates that visual feedback did not permit adults to plan a better back portion of their movement; however, it permitted them to correct their movement online more effectively than proprioceptive feedback alone. For children, right from movement reversal, deviation from the reference vector was slightly larger in the target-only condition than it was in the normal-vision condition. It remained so until movement endpoint. This larger deviation was more pronounced for the -35° target than it was for the 0° target. Thus, children were not as accurate as adults in planning the back portion of the movement and slightly less so in the target-only than in the normal-vision condition. Concerning this last point, the difference between the normal-vision and the target-only condition reached 1 mm approximately at 25% of relative movement time. Reversal angle variability and normalized overlapping variability were significantly larger for children than they were for adults. This was supported by a main effect of age groups, F (3, 44) = 5.3 and F (3, 44) = 9.1, respectively. Children groups did not differ significantly from one another on both these dependent variables, p [ 0.75.
Exp Brain Res (2010) 204:21–32
Fig. 5 Deviation of the back movement in reference to a vector joining the location of movement reversal and the starting base (i.e. target for the back movement)
In addition, reversal angle and normalized overlapping variability were significantly larger for the -35° target than they were for the 0° target, F (1, 44) = 60.2 and F (1, 44) = 63.4, respectively. This is illustrated in Fig. 4c, d, respectively. Finally, performing the task in normal vision resulted in a significantly smaller normalized overlapping variability than it did in the target-only condition (Fig. 4d), F (1, 44) = 33.0.
27
do in dealing with larger intersegmental torques as opposed to differences in planning and executing movement reversal in different regions of the workspace (i.e. along one’s midline vs. angled movements). For planar movements similar to those used in the present study, Sainburg et al. (1995) have already shown that the flexion interaction torque at movement reversal is larger for the -35° target than it is for the 0° target, whereas this flexion torque is larger for the 0° target than it is for a target located 35° to the right of the participant’s midline (hereafter, the ?35° target). Therefore, if children have more difficulty than adults do in dealing with larger intersegmental torques, we should observe larger reversal and overlapping error for children aiming at the -35° target but not the ?35° target. On the contrary, if children have more difficulty than adults do in planning and executing a sharp reversal for movements aimed at eccentric targets as opposed to those located along their midline, adults should outperform children at both the -35° and ?35° targets. To test these hypotheses, we had 6-year-old children and adults perform the same task as in the main experiment but toward targets located at -35°, 0°, and ?35°. Movement reversal was significantly sharper for the 0° target than it was for both the -35° and ?35° targets for children (20.8°, 7°, and 21.2° for the -35°, 0°, and ?35° targets, respectively) and adults (11.5°, 6°, and 14.5°, respectively). Similarly, overlapping error was significantly smaller for the 0° target than it was for both the -35° and ?35° targets for children (8.5, 4.3, and 8.0 for the -35°, 0°, and ?35° targets, respectively) and adults (4.0, 3.2, and 5.0, respectively). However, for both dependent variables, target differences were larger for children than they were for adults, and no large difference was noted between the -35° and ?35° targets (see Fig. 6). These results clearly indicate that children are more apt at planning and executing movement reversal for movement performed along their midline than they are for movement in other regions of the workspace.
Supplementary analyses Variability of the out-and-back movement trajectories For the -35° target, movement reversal was not as sharp for children as it was for adults, and larger overlapping errors were noted more for children than they were for adults. This difference could indicate that children had difficulty planning and executing a sharp reversal for movements initially aimed at eccentric targets relative to their midline, or it could have a biomechanical underpinning. Specifically, it could be argued that because movements performed along one’s midline required less intervention from the shoulder when compared to aiming at the -35° target, the interaction torque at movement reversal was smaller for the 0° than it was for the -35° target. Thus, it could be that the larger errors noted for children reflect the greater difficulty they have than adults
Finally, it has been shown that the directional variability of a series of movements aimed at the same target is large early after movement initiation (peak tangential acceleration) and then decreases as movement unfolds. For example, in Mackrous and Proteau (2007), it went from 7.0° at peak acceleration to 4.06° at peak velocity, 3.25° at peak deceleration, and 2.85° at movement endpoint. The large variability observed soon after movement initiation suggests that movement planning is based on approximations concerning the location of one’s hand and that of the target in the workspace, mechanical constraints, and especially, the state of the motor system (motor pathways, motoneuron pools, motor units, etc.). Because of these approximations
123
28
Fig. 6 Typical trajectories for out-and-back movements performed by 6–7-year-old children and by adults when aiming at eccentric targets for which a large (-35°) or a small (?35°) interaction torque is created at movement reversal. Note the similarity of results for the two targets. Note also the sharper movement reversal and tighter overlapping of out and back trajectories for adults when compared to children
Fig. 7 Orientation variability in the target-only and normalvision conditions at key landmarks for the out and back portions of the movements for all age groups. Note the smaller variability 100 ms after movement reversal when compared to that after movement initiation (v: peak velocity; d: peak deceleration; e: movement reversal [out] or endpoint [back]). Note also the decrease in variability for all age groups as movement unfolded
123
Exp Brain Res (2010) 204:21–32
and the variability inherent to all biological systems, the CNS has developed mechanisms to quickly update motor commands. If, as suggested above, uncertainties about the initial state of the effector are smaller once the movement is underway, we should observe a smaller variability soon after initiation of the back portion of the movement than we do of the out portion of the movement. Orientation variability data were submitted to an ANOVA contrasting 4 age groups (children aged 6–7, 8-9, and 10–12, and adults) 9 2 directions (out and back movements) 9 2 visual conditions (target-only and normalvision condition) 9 4 landmarks (100 ms after movement initiation/reversal, peak velocity, peak deceleration, and movement reversal/endpoint) 9 2 targets (-35° and 0°) with repeated measurements on the last four factors. We did not compute a similar analysis on the extent component of the task because the back portion of the movement ended on a physical stop. Figure 7 illustrates that, for all age groups, orientation variability was significantly smaller in the normal-vision condition than it was in the target-only condition, F (1, 44) = 78.8. In addition, at the 100 ms mark, orientation
Exp Brain Res (2010) 204:21–32
variability was significantly larger for the -35° target than it was for the 0° target. This difference decreased as movement unfolded, F (3, 132) = 190.8. Orientation variability of both the out and back portions of the movement was significantly larger for the two younger groups of children than it was for the adults 100 ms after movement initiation and at peak velocity, regardless of the portion (out or back) of the movement. Children groups did not differ significantly from one another. At peak deceleration and at movement reversal/ endpoint, only the youngest children remained more variable than the adults did. This is supported by a significant age groups 9 landmarks interaction, F (9, 132) = 3.8. The most interesting finding of this analysis revealed a significantly larger variability at the 100-ms mark for the out portion when compared to the back portion of the movement. In the target-only condition, the difference in movement variability between the out and back portions of the movement remained significant at all landmarks. In normal vision, this difference in variability between the out and back portions of the movement decreased but remained significant at peak velocity but not at peak deceleration and at movement endpoint. This is supported by a significant visual feedback x direction x landmarks, F (3, 132) = 7.8. The decrease in variability noted between the out and back portions of the movement observed at 100 ms was not caused by smaller peak acceleration for the back portion of the movement. Peak acceleration values were submitted to an ANOVA contrasting 4 age groups (children aged 6–7, 8–9, and 10–12, and adults) 9 2 directions (out and back movements) 9 2 visual conditions (target-only and normalvision condition) 9 2 targets (-35° and 0°) with repeated measurements on the last three factors. The results revealed that peak acceleration was larger for the back portion of the movement than it was for the out portion for the 6–7-yearold children (out: 3,768 mm/s2 vs. back: 4,373 mm/s2) and the 10–12-year-old children (out: 3,603 mm/s2 vs. back: 3,993 mm/s2). No significant difference was observed for the 8–9-year-old children (out: 3,439 mm/s2 vs. back: 3,586 mm/s2) and the adults (out: 3,824 mm/s2 vs. back: 3,570 mm/s2). This is supported by a significant age group x direction interaction, F (3, 44) = 3.9. Movement time Movement time was submitted to an ANOVA contrasting 4 age groups (children 6–7, 8–9, and 10–12, and adults) 9 2 visual conditions (target-only and normal-vision condition) 9 2 targets (-35° and 0°) with repeated measurements on the last two factors. Mean movement-time data are reported in Table 1. For all children, movement time was significantly shorter in the target-only condition than it was in the
29 Table 1 Movement time (standard deviation) for all age groups as a function of visual feedback and target location (ms) Target-only
Normal-vision
-35°
0°
-35°
0°
6-7 years old
915 (47)
919 (39)
990 (61)
910 (38)
8-9 years old
903 (31)
920 (33)
1,001 (47)
970 (48)
10–12 years old Adults
929 (19) 925 (14)
902 (24) 918 (20)
1,003 (25) 949 (27)
929 (33) 972 (33)
normal-vision condition for the -35° target, but not for the 0° target (see Table 1). We did not observe any significant movement-time difference for the adults. This is supported by a significant age groups x visual condition 9 target interaction, F (3, 44) = 3.6.
Discussion The fluidity of movement required for picking up a glass of milk filled to the brim and bringing it to your lips without spilling it requires fine pluriarticular movement control. In this report, we studied the developmental aspect of pluriarticular movement control by having children and adults perform overlapping out-and-back movements. For the back portion of the movement to overlap the out portion perfectly, the CNS must accurately coordinate the motion of the elbow with that of the shoulder at movement reversal. We also wanted to determine whether and how visual feedback enhances performance of out-and-back reaching movements in children. Reversal and overlapping Adults had significantly sharper movement reversal and smaller overlapping error than children did for the -35° target, but not for a target located in line with their midline. Because visual information about the starting base and target location was available at all times, our results could not be explained by age difference with regard to perception of eccentric target location (Contreras-Vidal 2006; Pellizzer and Hauert 1996). This position is also supported by the observation that the accuracy and variability of the out movement was very similar across age groups. If younger participants had more difficulty than adults did in locating their hand position on the starting base and/or the out target in the workspace, they should have been less accurate and/or more variable than adults for the out portion of the task when aiming at the -35° target, which was not the case. In addition, movement accuracy and variability of the out portion of the movement did not differ between age groups performing the task in the target-only
123
30
condition. This suggests that to ensure spatial accuracy when reaching away from the midline, the processing of proprioception is no less accurate in children than it is in adults. Finally, because this difference between children and adults was similar for movement reversal that created large (-35°) or small (?35°) intersegmental torques (Sainburg et al. 1995), it appears that the larger error observed for children for the eccentric targets is not related to this biomechanical factor. Rather, although children did not perform the task as well as adults did for the -35° target, movement reversal remained much sharper than that reported by Sainburg et al. (1995) for deafferented participants (reversal error of 53° and 143°, respectively, for each one of two participants). This suggests that, similar to adults, children used proprioceptive feedback during both movement planning and movement execution to coordinate the motion of their shoulder and elbow. However, a new important finding of the present study is that the mechanisms underlying the control of upper limb dynamics do not appear to develop uniformly across the workspace. It appears that these mechanisms are more precise and are more quickly developed in children for movements performed along or close to their midline than they are for angled movements. This would explain why, for all age groups, we observed sharper reversal and smaller overlapping error for the 0° target than for the -35° targets. In addition, it would explain why these errors are larger for eccentric targets for children than they are for adults and why this increase in error gradually decreased as children grew older. The observation of similar reversal angle and overlapping error for the -35° and ?35° targets (Fig. 6) also supports this hypothesis. Visual feedback Concerning our second goal, visual feedback did not permit children or adults to have significantly sharper movement reversal than when performing the task in the target-only condition. Thus, visual feedback did not permit adults or children to plan and control movement reversal more effectively. For adults, this result supported the finding of Sainburg et al. (1993) that visual feedback does not facilitate interjoint coordination when intact proprioceptive feedback is available. The present study extends this finding to children as young as 6–7 years old. This observation is not surprising since the children’s performance (movement reversal) was at par with that of adults, as it was for the 0° target. However, observing the same result for the -35° target, for which the children’s performance was not as effective as the adults’, provides an even stronger support to the claim that vision is unnecessary for the control of movement dynamics (Krakauer et al. 1999; Franklin et al. 2007).
123
Exp Brain Res (2010) 204:21–32
Nonetheless, we showed that visual feedback permitted participants to overlap the out portion of their movement sooner and more completely than they did in the targetonly condition. This overlapping difference became apparent soon after movement reversal (i.e. 25–30% of relative movement time), which concurs with Mackrous and Proteau (2007), who showed that visual information can be used to adjust movement trajectory early after movement initiation (see also Grierson and Elliott 2008; Masson and Proteau 1997 for confirming evidence). To our knowledge, this is the first time that such early corrections are reported for children as young as 6–7 years old. Dynamic versus static sensory information In goal-directed movements performed in both normalvision and no-vision conditions, Proteau et al. observed that, once movement variability is normalized for movement length (coefficients of variability), this normalized variability quickly decreased between peak acceleration and peak velocity (Be´dard and Proteau 2005; Lhuisset and Proteau 2004a; see also Robin et al. 2005). They argued that the large variability observed at peak acceleration resulted from movement planning being based on approximations concerning the initial state of the motor system (motoneuron pool, motor pathway, motor unit, biomechanical constraints, etc.) but that dynamic information available during movement execution provided more accurate information about the state of the system, which explains the large decrease in variability at peak velocity. The results of the present study supported this proposition. Direction variability was significantly larger 100 ms after movement initiation for the out portion of the movement (movement planning was based on approximations) than it was for the back portion, which was likely based on the processing of dynamic information available during the out portion of the movement. Children were as efficient as adults in reducing initial direction variability as movement unfolded (see also Lhuisset and Proteau 2004a). This is true for both the out and back portions of the movements and indicates that even the youngest children who participated in the present study could modulate their movement online to compensate for initial error/noise in movement planning and execution processes. Proprioception: multiple developmental trends The results of previous studies indicate that in an upper limb matching task, the accuracy of proprioceptive feedback increased from early childhood to late adolescence, whereas studies in tendon vibration (Hay and Redon 1997) and prismatic displacement (Orliaguet 1985) both show
Exp Brain Res (2010) 204:21–32
increased reliance on proprioceptive feedback for 5-yearold children in comparison with older children in manual aiming. In reaching for grasp along one’s midline (Schneiberg et al. 2002), endpoint-trajectory straightness, smoothness, and variability attained adult levels in children aged 6–8, while evidence of increased variability in interjoint coordination patterns, presumably based on the processing of proprioception, persisted in children as old as 11 years. In the present study, the immature interjoint coordination pattern reported by Schneiberg et al. (2002) did not lead to a larger and more variable reversal angle for movements performed along one’s midline. However, for the -35° target, the CNS was not yet able to process proprioception as efficiently as it could for adults in coordinating motion of the elbow with that of the shoulder while accounting for the transmitted torques between these two segments at movement reversal. The larger variability in movement reversal observed for the -35° target might indicate that the CNS is searching an optimal solution (Thelen 1993). Taken together, these results clearly indicate that there is no unique developmental trend for defining how proprioception is processed for different tasks.
Conclusion In conclusion, the results of the present study reveal that pluriarticular movement control is as efficient for children as it is for adults for movement performed along one’s midline, whereas it appears to improve across the course of development for eccentric regions of the workspace. Visual feedback does not play a significant role in sharp movement reversal in participants with intact proprioception, even when relatively large errors are noted (eccentric targets in children). This suggests that interjoint coordination is largely, if not exclusively, based on the processing of proprioceptive feedback. However, the larger reversal error variability observed for all groups of children when compared with adults suggests that the ability to use proprioceptive information is not entirely mature by the age of 12. Finally, movement-planning processes appear less variable when based on dynamic rather than static afferent information. Acknowledgments This work was supported by a Discovery grant (L.P.) provided by the Natural Sciences and Engineering Research Council of Canada.
References Be´dard P, Proteau L (2005) Movement planning of video and of manual aiming movements. Spat Vis 18:275–296
31 Chicoine A-J, Lassonde M, Proteau L (1992) Developmental aspects of sensorimotor integration. Dev Neuropsychol 8:381–394 Contreras-Vidal JL (2006) Development of forward models for hand localization and movement control in 6- to 10-year-old children. Hum Mov Sci 25:634–645 Elliott JM, Connolly KJ, Doyle AJ (1988) Development of kinaesthetic sensitivity and motor performance in children. Dev Med Child Neurol 30:80–92 Ferrel-Chapus C, Hay L, Olivier I, Bard C, Fleury M (2002) Visuomanual coordination in childhood: adaptation to visual distortion. Exp Brain Res 144:506–517 Fetters L, Todd J (1987) Quantitative assessment of infant reaching movements. J Mot Behav 19:147–166 Fitts PM (1954) The information capacity of the human motor system in controlling the amplitude of movement. J Exp Psychol 47:381–391 Flanagan JR, Lolley S (2001) The inertial anisotropy of the arm is accurately predicted during movement planning. J Neurosci 21:1361–1369 Franklin DW, So U, Burdet E, Kawato M (2007) Visual feedback is not necessary for the learning of novel dynamics. PLoS ONE 2:e1336 Gandolfo F, Mussa-Ivaldi FA, Bizzi E (1996) Motor learning by field approximation. PNAS 93:3843–3846 Ghez C, Sainburg R (1995) Proprioceptive control of interjoint coordination. Can J Physiol Pharmacol 73:273–284 Goble DJ, Lewis CA, Hurvitz EA, Brown SH (2005) Development of upper limb proprioceptive accuracy in children and adolescents. Hum Mov Sci 24:155–170 Gordon J, Ghilardi MF, Cooper SE, Ghez C (1994) Accuracy of planar reaching movements. Exp Brain Res 99:112–130 Grierson LEM, Elliott D (2008) Kinematic analysis of goal-directed aims made against early and late perturbations: an investigation of the relative influence of two online control processes. Hum Mov Sci 27:839–856 Hay L (1979) Spatial-temporal analysis of movements in children: motor programs versus feedback in the development of reaching. J Mot Behav 11:189–200 Hay L, Redon C (1997) The control of goal-directed movements in children: role of proprioceptive muscle afferents. Hum Mov Sci 16:433–451 Konczak J, Borutta M, Topka H, Dichgans J (1995) The development of goal-directed reaching in infants: hand trajectory formation and joint torque control. Exp Brain Res 106:156–168 Krakauer JW, Ghilardi MF, Ghez C (1999) Independent learning of internal models for kinematic and dynamic control of reaching. Nat Neurosci 2:1026–1031 Lhuisset L, Proteau L (2004a) Planning and control of straight-ahead and angled planar movements in adults and young children. Can J Exp Psychol 58:245–258 Lhuisset L, Proteau L (2004b) Visual control of manual aiming movements in 6- to 10-year-old children and adults. J Mot Behav 36:161–172 Mackrous I, Proteau L (2007) Specificity of practice results from differences in movement planning strategies. Exp Brain Res 183:181–193 Masson G, Proteau L (1997) Visual perception modifies goal-directed movement control: supporting evidence from a visual perturbation paradigm. Q J Exp Psychol 50:726–741 Mathew A, Cook M (1990) The control of reaching movements by young infants. Child Dev 61:1238–1257 Orliaguet J-P (1985) Dominance visuelle ou proprioceptive lors de la perception de la position d’un bras chez l’enfant. Cah Psychol Cogn 5:609–618 Orliaguet J-P (1986) Facteurs spatiaux et dominance sensorielle lors de la perception de la position de la main chez l’enfant. Psychol Francaise 31:23–27
123
32 Pellizzer G, Hauert CA (1996) Visuo-manual aiming movements in 6to 10-year-old children: evidence for an asymmetric and asynchronous development of information processes. Brain Cogn 30:175–193 Robin C, Toussaint L, Blandin Y, Proteau L (2005) Specificity of learning in a video-aiming task: modifying the salience of dynamic visual cues. J Mot Behav 37:367–376 Sainburg RL, Poizner H, Ghez C (1993) Loss of proprioception produces deficits in interjoint coordination. J Neurophysiol 70:2136–2147 Sainburg RL, Ghilardi MF, Poizner H, Ghez C (1995) Control of limb dynamics in normal subjects and patients without proprioception. J Neurophysiol 73:820–835 Sainburg RL, Ghez C, Kalakanis D (1999) Intersegmental dynamics are controlled by sequential anticipatory, error correction, and postural mechanisms. J Neurophysiol 81:1045–1056
123
Exp Brain Res (2010) 204:21–32 Sainburg RL, Lateiner JE, Latash ML, Bagesteiro LB (2003) Effects of altering initial position on movement direction and extent. J Neurophysiol 89:401–415 Schneiberg S, Sveistrup H, McFadyen B, McKinley P, Levin M (2002) The development of coordination for reach-to-grasp movements in children. Exp Brain Res 146:142–154 Sober SJ, Sabes PN (2003) Multisensory integration during motor planning. J Neurosci 23:6982–6992 Thelen E (1993) Timing and developmental dynamics in the acquisition of early motor skills. In: Turkewitz G, Devenny DA (eds) Developmental time and timing. Lawrence Erlbaum Associates, Hillsdale, p 266 von Hofsten C (1979) Development of visually guided reaching: the approach phase. J Hum Mov Stud 5:160–178 von Hofsten C (1991) Structuring of early reaching movements: a longitudinal study. J Mot Behav 23:280–292
