- Email: [email protected]
Development of proximal arm muscle control during reaching in young infants: From variation to selection
Infant Behavior & Development 33 (2010) 30–38
Contents lists available at ScienceDirect
Infant Behavior and Development
Development of proximal arm muscle control during reaching in young infants: From variation to selection Hanneke Bakker a , Victorine B. de Graaf-Peters a , Leo A. van Eykern a , Bert Otten b , Mijna Hadders-Algra a,∗ a
Department Paediatrics – Developmental Neurology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands Institute of Human Movement Sciences, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
b
a r t i c l e
i n f o
Article history: Received 1 September 2008 Received in revised form 8 April 2009 Accepted 23 October 2009 Keywords: Motor development Infants Reaching EMG Variation Kinematics
a b s t r a c t Reaching movements are initiated by activity of the prime mover, i.e. the first activated arm muscle. We aimed to investigate the relationship between prime mover activity and kinematics of reaching in typically developing (TD) infants in supine and sitting position. Fourteen infants were assessed at 4 and 6 months during reaching in supine and supported sitting. Kinematics and EMG-activity of deltoid, pectoralis major, biceps (BB) and triceps brachii were recorded. Kinematic analysis focused on number of movement units (MUs) and transport MU (MU with longest duration). Prime mover use was variable, but at 6 months a dominance of BB emerged in both testing conditions. Kinematics were also variable, but with increasing age the number of MU decreased and the relative proportion of the transport MU increased. BB as prime mover at 6 months was related to a larger transport MU. Conclusion: Between 4 and 6 months BB prime mover dominance emerges which is related to relatively efficient reaching characteristics. © 2009 Elsevier Inc. All rights reserved.
1. Introduction Reaching seems a simple daily activity. But, on close inspection the task of reaching towards an object is far from simple: the multi-jointed upper limb has to be directed towards the object while simultaneously the position of hand and fingers has to be prepared for the pick-up of the object (Jeannerod, 1990). Therefore it is not surprising that the development of reaching runs a protracted course which extends into adolescence (Kuhtz-Buschbeck, Stolze, Jöhnk, Boczek-Funcke, & Illert, 1998). Our knowledge on typical development of reaching in early infancy is still limited. Nevertheless such information is a prerequisite for the understanding of deviant development. What we do know is the following. Immediately after birth the infant mainly produces spontaneous non-goal-directed activity (Hadders-Algra, 2004). Nevertheless, the neonate has some capacity for goal-directed arm activity (Hopkins & Prechtl, 1984; Van der Meer, Van der Weel, and Lee, 1995). From 3 months onwards the infant starts to produce more types of goal-directed arm activity (Thelen et al., 1993; Van der Fits, Klip, Van Eykern, & Hadders-Algra, 1999), in particular when an interesting object is present. In response to the presence of an object the infant may put the hands into the mouth or it may produce oscillating movements of the extended arm in the direction
∗ Corresponding author. Tel.: +31 50 3614247; fax: +31 50 3619158. E-mail address: [email protected] (M. Hadders-Algra). 0163-6383/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.infbeh.2009.10.006
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
31
of the object or into another direction (Van der Fits, Klip, et al., 1999). With increasing age, arm movements are more often directed towards the object and from 4 to 5 months onwards they generally end in successful grasping, to become virtually always successful at 6 months (Van der Fits, Klip, et al., 1999). Initially reaching movements are characterized by large variation: variation in movement velocity and in the so-called movement units (De Graaf-Peters, Bakker, Van Eykern, Otten, & Hadders-Algra, 2007; Fallang, Saugstad, & Hadders-Algra, 2000; Thelen, Corbetta, & Spencer, 1996). MUs are submovements of reaching, which are determined with the help of peaks in the velocity profile of the hand (Von Hofsten, 1991). At 4 months of age reaching movements consist of about 4 MUs. With increasing age the variation in the number of MUs and the number itself decrease to about 3 MUs at 6 months (Carvalho, Tudella, Caljouw, & Savelsbergh, 2008; De Graaf-Peters et al., 2007; Fallang et al., 2000). It is first in adolescence, i.e. around the age of 12 years, that the majority of reaching movements has the adult configuration of 1 MU (Kuhtz-Buschbeck et al., 1998). In parallel to the decrease in MUs an increase of the relative length and duration of the first MU, the so-called transport MU, occurs so that it gradually covers a large proportion of the approach to the target (De Graaf-Peters et al., 2007; Fallang et al., 2000; Kuhtz-Buschbeck et al., 1998). The performance of reaching is related to postural control and body position. At the age of 4–5 months reaches are more successful and have a shorter duration in sitting position than in supine (Carvalho et al., 2008; Out, Van Soest, Savelsbergh, & Hopkins, 1998; Savelsbergh & Van der Kamp, 1994). Conflicting data have been reported whether these positions also affect other kinematic characteristics of early reaching behaviour, such as the number of MUs or the straightness of the movement path (Carvalho et al., 2008; Out, Savelsbergh, Van Soest, & Hopkins, 1997; Out et al., 1998). The onset of a reaching movement is mediated by activity of one of the proximal arm or shoulder muscles, in particular by activation of the biceps brachii muscle or the deltoid muscle. The muscle activated first is called the ‘prime mover’. Adult persons select in a specific situation a specific prime mover. For example, in a task of pointing at self-paced speed in standing position some adults use fast, ballistic arm movements during which the deltoid muscle is the prime mover, whereas others use slower arm movements including some elbow flexion during which the biceps brachii is the prime mover (Van der Fits, Klip, Van Eykern, & Hadders-Algra, 1998). Young infants variably use the biceps brachii, triceps brachii, deltoid, trapezius or pectoralis major muscles as prime mover (Thelen & Spencer, 1998). No data are available on the development of prime mover use during infancy, nor is it known whether developmental changes in prime mover use are related to changes in the kinematics of reaching or whether it is affected by infant position. Guided by the ideas of the Neuronal Group Selection Theory (Edelman, 1989; Hadders-Algra, 2000) we hypothesized that variability in prime mover use at early age, i.e. in the period during which reaching rapidly grows into a consistently successful event, changes into preference for a specific muscle as prime mover. If this is the case, we expect that the selection of a specific prime mover is associated with improved kinematics of the reaching arm movement, i.e. reaches consisting of less MUs or reaches with proportionally large transport MUs. We hypothesize that the developmental change may be observed in supine and sitting position. The aim of the present study was to test these hypotheses. We investigated in typically developing infants (1) the development of prime mover use during reaching at the ages of 4 and 6 months, where we paid specific attention to variation and selection, (2) the relation between prime mover use and kinematical characteristics of reaching at 4 and 6 months and (3) whether the effect of position, i.e. lying supine or sitting in an infant chair, affects prime mover use and the relation between prime mover use and kinematical features of reaching. The two different conditions differ from each other in terms of postural challenge – the challenge being larger in the sitting than in the supine position (Van der Fits, Klip, et al., 1999) – and in challenge to produce a successful reach. The latter challenge is larger in supine, when the arm behaves as a simple inverted pendulum, than in sitting where arm behaviour can be described in terms of a simple pendulum (Out et al., 1997). A way to deal with a larger mechanical challenge could be a larger degree of antagonistic co-activation (Out et al., 1997). Therefore we also assessed the degree of co-activation between biceps and triceps brachii. In order to address the above questions reaching movements in supine and sitting position of 14 typically developing infants were studied longitudinally at the ages of 4 and 6 months with the help of simultaneous EMG recordings of arm and shoulder muscles and kinematical recordings. We used testing situations which resembled daily life as much as possible, i.e. we did not use extra straps or devices to support the head to assist the child’s control of posture. Previously we reported about the development of postural control and its relation with the kinematics of reaching of these children (De Graaf-Peters et al., 2007); in the present article the focus is on the development of the prime mover in relation to the kinematics of reaching.
2. Methods 2.1. Participants Fourteen typically developing infants participated in this longitudinal study. The infants were tested at the ages of 4 and 6 months. Adequate recordings at the age of 4 months were obtained in 6 boys and 7 girls and in 6 boys and 6 girls at 6 months. The infants’ postmenstrual age at birth varied between 38 and 42 weeks (median: 39 weeks); birth weight between 2930 g and 4280 g (mean: 3582 g, SD: 468 g). The infants were recruited from amongst acquaintances of the investigators. The parents of the infants gave informed consent. The study was approved by the Medical Ethics Committee of the University Medical Center Groningen.
32
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
Fig. 1. Schematic drawing of testing conditions: lying supine (left) and sitting supported in infant chair (right). The chair had a back rest and a semicircular horizontal bar in front.
2.2. Procedure Reaching was assessed in two positions: lying supine and sitting upright with support in an infant chair (Fig. 1). No additional straps or specific head supporting devices were used. The infants were tested while wearing diapers only. Testing order of the two positions was determined randomly. Reaching movements were elicited by presenting small, attractive toys in the midline at arm length distance. A set of toys was used consisting of rings and small figures with sizes varying between 7 cm × 7 cm × 0.5 cm and 5 cm × 3 cm × 2 cm. Toys were only presented when the infant was alert and not crying. We aimed at recording at least ten reaching movements with the right arm in each position, but when the infant became fussy or tired the session was shortened. In order to confirm neurological integrity a standardized neurological examination (Touwen Infant Neurological Examination (TINE); Hadders-Algra, Heineman, Bos, & Middelburg, 2009) was carried out after each reaching session. 2.3. EMG, kinematical and video recordings EMG-activity was measured with surface electrodes on the following arm and shoulder muscles on the right side of the body: deltoid muscle (DE) pectoralis major (PM), biceps brachii (BB), and triceps brachii (TB). In the remainder of the paper these muscles are referred to as arm muscles. In addition, EMG-activity was recorded of postural muscles (Fig. 2). The EMG data were recorded continuously, digitized and fed into a personal computer with the help of POLY, a software program for long-lasting polygraphic recordings (sampling rate 500 Hz; Inspector Research Systems, Amsterdam, The Netherlands).
Fig. 2. Velocity profile of a successful reach of a sitting infant aged 6 months. The combination of one acceleration and one deceleration forms one MU; the trial consists of 3 MU.
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
33
The entire reaching session was video recorded with a split-screen allowing for a simultaneous lateral and frontal view of the infant. The video recordings were time-coupled to the EMG registration. The registration took about 0.5 h. Simultaneous with the EMG, arm movements were recorded kinematically with an ELITE system (BTS, Milan, Italy) in a two-camera configuration at a sampling frequency of 50 Hz. A reflective marker was placed on the right wrist of the infant, i.e. on the styloid process of the radius. The kinematical recording was discontinuous; it started just before the toy was presented and lasted for 10 s. The periods of kinematical sampling were indicated on the EMG recording.
2.4. Data analyses The video registration was used for multiple purposes. First, the video recordings were used to determine whether the infant was in an appropriate behavioural and attentional state during a specific trial. Second, the video was used for a classification of the behaviour of arm movements during toy presentation. Four different movements were distinguished: (a) pre-reaching arm movements, which were defined as arm movements occurring in response to toy presentation, which were not directed towards the toy (Trevarthen, 1984), (b) reaching movements: arm movement in the direction of the toy which did not end in toy contact (‘reach’), (c) reaching movements which did end in toy contact but not in grasping of the toy (‘reach and touch’), and (d) reaching movements which resulted in grasping of the toy (‘grasp’). The movements were categorized into movements with the right arm (right arm was dominating the reaching task) or left arm (left arm was dominating the reaching task). The results are restricted to the behaviour of the right arm. Reaches were classified as successful when they either ended in toy contact or grasping. Third, the video analysis served the selection of reaching movements with the right arm, as EMG and kinematical data were only available for the right arm. EMG and kinematical data-analyses were restricted to movements with ended in toy contact or grasping and to trials with direction-specific postural activity (De Graaf-Peters et al., 2007). Direction-specific postural activity means primary recruitment of postural muscles on the side of the body opposing body sway. For instance, reaching, which induces a forward sway of the body, is accompanied by direction specific postural activity in the dorsal postural muscles (Forssberg & Hirschfeld, 1994). The number of trials which fulfilled the behavioural criteria and which had an appropriate EMG and kinematical recording is displayed in Table 1. A computer algorithm was used for the EMG analyses to detect phasic muscle activity. The algorithm used a derivative of the root mean square of a full rectified signal (200 ms moving window), and marked significant deviations from a fixed detection level. The detection level was based upon a long-term (3.7 s) mean baseline activity. EMG bursts were detected when the activity exceeded the detection level for at least 50 ms (Van der Fits et al., 1998). The EMG bursts were discriminated from heart activity. The prime mover was defined as the first arm muscle which showed phasic activity around reaching onset. For biceps brachii and triceps brachii we also determined the presence of antagonistic co-activation. Co-activation was considered to be present when the antagonist showed phasic activity within 50 ms after the onset of the agonist. For each infant, each condition and each age the following parameters were calculated: (1) percentage of trials during which a specific arm muscle was used as a prime mover. (2) The preference prime mover, which was arbitrarily defined as the arm muscle being the prime mover in at least 50% of the trials. (3) The latencies of recruitment of the other arm muscles, defined as the time interval between the onset of the prime mover and the onset of activity in any other arm muscle. For each infant, age, and position median latency values were calculated. Offline kinematical analysis was carried out with the help of the software package DataMonster (E. Otten, Institute of Human Movement Sciences, University of Groningen). The data were filtered using a low-pass filter of 6 Hz (zero time-lag filter). Arm movements were analysed in two dimensions. The onset of arm movement was defined as the moment at which the velocity of the wrist increased ≥5% of peak velocity, the end of the movement was defined as the moment at which wrist velocity decreased to ≤5% of peak velocity. In the kinematical analysis, only trials with a clearly demarcated start and stop were included. The following parameters were used to describe the reaching movements: (1) the number of MU per trial. A MU consisted of one acceleration and one deceleration in the velocity profile of the wrist marker (Fig. 3). (2) The total duration of the reaching movement. (3) The relative duration of the first MU (i.e. the transport MU) in relation to total duration of the arm movement. (4) Maximum reaching velocity. Median values of the parameters were calculated for each infant, each condition, and each age.
Table 1 Number of trials analysed per individual for EMG and kinematical recordings. Age
EMG recordings
Kinematical recordings
Supine
4 Months 6 Months
Sitting
Supine
Sitting
n
Med
Range
n
Med
Range
n
Med
Range
n
Med
Range
13 12
11 13
3–20 10–15
12 12
9 12
4–13 7–17
9 10
4 5
3–9 3–12
12 12
5 6
3–10 4–11
n = number of infants and med = median value.
34
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
Fig. 3. Developmental trajectories of prime mover preference. Each line represents the development of one infant. BB: biceps brachii, DE: deltoid muscle, PM: pectoralis major, and Non: no specific preference for prime mover. Effect of age on prime mover preference, Wilcoxon; supine: p < 0.05, sitting: p < 0.01.
2.5. Statistics Statistical analysis was performed using the computer package SPSS (version 12.1). Analyses were performed on infant level this means that for the EMG latencies and the kinematical data first median values of the reaches of each infant were calculated on the basis of which further analyses were performed, i.e. the effect of age and position was based on these median values. The non-parametric Wilcoxon test was used to evaluate the effect of age and position. Due to data-loss in the kinematical recordings – a well known problem in infant research (Van der Fits, Otten, Klip, Van Eykern, & HaddersAlgra, 1999) – and thereby loss of pairs of data per infant, it was not possible to use the Wilcoxon test for the analyses of the kinematical data and the analyses of the relationships between the EMG and reaching data. We therefore decided to use the Mann–Whitney U test. Throughout the analyses, differences with p-values < 0.05 were considered statistically significant.
3. Results 3.1. Reaching and grasping in supine and sitting position At 4 months of age 11 out of 13 infants were able to reach and touch or grasp the toy successfully at least once, 11 infants produced successful reaching in supine, five in sitting position. Position seemed to affect the type of reaching movements at 4 months: in supine the infants showed more often successful reaching movements and in sitting more often pre-reaching movements (Table 2). The effect of position did not however reach statistical significance. At 6 months the number of successful reaches had increased and that of pre-reaching movements decreased (Table 2). At this age position did have a significant effect on the success of reaching: reaching was more often successful in supine than in sitting (Table 2; Mann–Whitney U test: p < 0.05). Preliminary data analyses indicated that there were no differences in prime mover preference and kinematical characteristics between movements classified as ‘reach and touch movements’ and successful grasping movements. For this reason the data of these movements were pooled.
Table 2 Types of arm movement in supine and sitting position at 4 and 6 months of age. Grasp
Reach and touch
Reach
Pre-reach
4 Months
Supine (n = 11) Sitting (n = 10)
21 (0–82) 8 (0–40)
0 (0–53) 7 (0–100)
19 (0–86) 7 (0–20)
6 Months
Supine (n = 12) Sitting (n = 12)
94** (67–100) 67*,† (15–100)
6 (0–23) 0 (0–15)
0 (0–13) 0 (0–18)
38 (0–100) 66 (0–100) 0** (0–8) 27**,†† (0–55)
n = number of infants. The data indicate median percentages per group, with minimum and maximum values between brackets. Effect of age: with increasing age infants made less pre-reaching movement (Wilcoxon—**supine: p < 0.01, sitting: p < 0.01) and more successful grasping movements (Wilcoxon—**supine: p < 0.01, *sitting: p < 0.05). Effect of position: in sitting the infants made more pre-reaching movements (Mann–Whitney U test: † p < 0.05); in supine the infants made more successful reaching movements (†† p < 0.01; Note: the low number of paired data precluded the use of the Wilcoxon test).
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
35
Fig. 4. Typical example of EMG-activity of arm and postural muscles during successful reaching in an infant aged 6 months during sitting. The upper four channels represent arm muscle activity, phasic activity is denoted by the horizontal underlining. Here BB is the first arm muscle activated, hence is the prime mover. Postural activity is direction specific, i.e. consists of NE and TE recruitment. DE: deltoid muscle, PM: pectoralis major, BB: biceps brachii, TB: triceps brachii, ST: sternocleidomastoid, NE: neck extensor, RA: rectus abdominis, TE: thoracal extensor, and LE: lumbar extensor.
3.2. Development of prime mover preference Fig. 4 shows a typical example of EMG-activity during reaching. At 4 months the identity of the prime mover showed substantial intra- and interindividual variation. Despite the variation, most infants had a preference for a specific prime mover: in supine three infants used preferentially BB, two DE, four PM and two infants did not have a specific prime mover preference. In sitting position all infants had preference prime mover; four infants had a preference for BB, four for DE and one for PM (Fig. 4). At 6 months all infants had a preference prime mover within their variable repertoire of prime movers in both conditions. The identity of prime mover still varied, but most infants used BB as prime mover. This was true for the supine and sitting position (Fig. 4). The increase in the use of BB as preference prime mover between 4 and 6 months was statistically significant (Wilcoxon—supine; p < 0.05, sitting; p < 0.01). The increase could not be attributed to the fact that infants produced more trials which could be analysed at 6 months than at 4 months. Co-activation of BB and TB was virtually restricted to reaching movements during which BB was the prime mover. The frequency of BB–TB co-activation in these trials was however largely variable. It varied across age, condition and infant from 0% to 100% of trials during which BB was the prime mover (median values: 4 months supine 33%, sitting 70%, 6 months supine 38%, sitting 67%). The degree of co-activation between the supine and sitting condition was not statistically significant. 3.3. Development of kinematical characteristics Also the kinematical data were characterized by variation. At 4 months reaching movements consisted of 3.5 and 4 MU (median values, sitting and supine position, respectively). The number of MU decreased with increasing age: in sitting at 6 months reaches consisted of 2.5 MU (difference with 4 months, Wilcoxon—p = 0.02), in supine at 6 months of 3 MU (difference with 4 months: n.s.; Fig. 5). The number of MU was not affected by position. The relative duration of the transport unit at 4 months in supine and sitting position is 30% of the total duration of the reaching movement. The same was true for reaching movements in supine at 6 months. But in the sitting position the relative duration of the transport unit increased significantly with increasing age to 38% (Fig. 5, Wilcoxon—p = 0.04). No significant difference was found between the relative duration of the transport unit between the supine and sitting position. The total duration of the reaching movement at 4 months in supine was about 0.55 s, in sitting about 0.65 s. Movement duration decreased with increasing age: at 6 months till about 0.50 s in both conditions (Wilcoxon, decrease supine: n.s., sitting p < 0.05) (Fig. 6). 3.4. Relationship between prime mover preference and kinematical characteristics At the age of 4 months prime mover preference was not related to the kinematics of reaching; this was true for both positions. At 6 months the relative duration of the transport unit was related to prime mover preference. The transport unit had a longer duration in infants who used BB as preferred prime mover than in infants who used DE or PM as prime mover (Fig. 7, Mann–Whitney U test: supine; p < 0.01, sitting; p < 0.01). The other kinematical parameters of reaching were not related to prime mover preference.
36
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
Fig. 5. Development of kinematical characteristics in supine and sitting position, i.e. number of MU and relative duration of transport unit at 4 and 6 months. Bold horizontal lines indicate median values; the boxes represent interquartile ranges and the vertical lines complete ranges. Age effect of number of MU, Wilcoxon; sitting: p = 0.02. Age effect of relative duration of transport unit, Wilcoxon; sitting: p = 0.04.
Fig. 6. Development of duration of reaching movements in supine and sitting position. Bold horizontal lines indicate median values; the boxes represent interquartile ranges and the vertical lines complete ranges. Age effect of total reaching duration, Wilcoxon; sitting: p < 0.05.
4. Discussion The present study indicates that the improvement in success of reaching with increasing age is paralleled by selection of the biceps brachii as prime mover, a selection which – in turn – is related to a longer duration of the transport MU. The selection process is independent of the infant’s position during reaching. Before discussing the neurodevelopmental significance of our findings, we would like to address an important methodological point, namely that we restricted our analysis to trials with direction-specific postural activity. This means that we selected reaching movements which are relatively good organized, as the presence of direction-specific postural activity is related to more success of reaching, and reaching movements with less MU and a longer duration of the transport MU (De Graaf-Peters et al., 2007). We decided to restrict ourselves to this selection in order to avoid confounding with the effect of direction-specificity. Another methodological point which deserves attention is our testing condition. In order to facilitate EMG- and kinematic recording infants were assessed while being undressed. For practical reasons a diaper was worn. It is possible that the diaper may have provided some postural support, but if so, this was true for both testing conditions. The sitting condition we
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
37
Fig. 7. Development of relation between prime mover and kinematical characteristics. Number of MU and relative duration of transport unit in groups of infants with a specific prime mover preference at 6 months in supine (left panels) and sitting (right panels) position. Bold horizontal lines indicate median values; boxes represent interquartile ranges; vertical lines complete ranges. Effect of prime mover preference on relative duration of transport unit, Mann–Whitney U test: p < 0.02.
used differed largely from the sitting condition used in previous studies (Carvalho et al., 2008; Out et al., 1997, 1998; Savelsbergh & Van der Kamp, 1994). In the previous studies the infants received firm postural support in the sitting position whereas in the present study postural support in sitting was reduced to a minimum. This means that in previous studies the difference in postural challenge between supine and sitting was little if any, leaving the infant in the sitting situation with the mechanical advantage of the inverted pendulum condition (Out et al., 1997). The advantage was reflected in better performance of reaching in sitting than in supine. In the present study, the mechanical advantage for the reaching arm in the sitting condition was entirely overruled by the disadvantage of the large postural challenge. As a result the infants in the present study produced more successful reaches in supine than in the sitting position. Our study showed that infants showed substantial within- and between-subject variation in use of prime mover. Nevertheless, it was also clear that most infants already at the age of 4 months had a preference for a specific prime mover. At this age prime mover preference showed substantial between-subject variation. Major part of between-subject variation had disappeared at the age of 6 months, when most infants used the biceps brachii as prime mover. Selection of biceps brachii as prime mover presumably is related to the way infants in general performs reaching movements: this usually includes some degree of elbow flexion (cf. Van der Fits et al., 1998). Similar developmental changes have been reported for postural adjustments during reaching: at 4 months some degree of individual preference for a specific postural adjustment is present but with large between-subject variation; at 6 months a general preference for a specific postural pattern, i.e. the pattern in which all direction-specific muscles are recruited, exists (De Graaf-Peters et al., 2007). These developmental changes fit well to the ideas of the Neuronal Group Selection Theory (Edelman, 1989; Hadders-Algra, 2000). According to NGST motor development is characterized by two phases of variability. During the phase of primary variability the infant displays variable motor behaviour, but is not able to adapt motor behaviour to the specifics of the situation. During the phase of secondary variability the infant gradually learns to select from its repertoire of motor strategies the best motor solution fitting to the situation. The data suggest that infants enter the phase of secondary variability for prime mover use and postural adjustment between 4 and 6 months. The concurrence of both events seems logical as both deal with control of proximal muscles, in which the corticospinal system plays a less prominent role than in the control of distal muscles (Turton and Lemon, 1999; Hadders-Algra, 2000). The group preference at 6 months for biceps brachii as prime mover suggests that the selection for biceps brachii is associated with some benefit. Indeed, our findings support this notion, as the selection of the biceps brachii as prime mover was related to a longer duration of the transport MU. Selection of biceps brachii was related to the duration of the transport MU but not to the number of MUs. This might imply that the transport MU is more dependent on proximal and postural muscle control than the number of MUs (Jeannerod, 1990). Interestingly, adult persons do not have such a strict preference for biceps brachii as prime mover: about half of adult persons use biceps brachii as prime mover and the other half the deltoid muscle (Van der Fits et al., 1998). This means that development of prime mover selection has not been completed at 6 months of age. Further studies could shed light on the age at which the adult stage of prime mover selection is reached.
38
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
The use of biceps brachii as prime mover was associated in a variable way with triceps brachii co-activation. In other words, antagonistic co-activation was an inconsistent finding and did not depend on position or age. This might imply that at early age variation and exploration are more important determinants of motor control than antagonistic co-activation (cf. Hadders-Algra, Brogen, & Forssberg, 1996). It is noteworthy that the present study showed that success of reaching in early infancy is greater in supine than in supported sitting, this in contrary to former studies as is mentioned in the introduction. An explanation for more successful reaching in supine might be the difference in postural challenge. The supine situation offers a large base of support including stable support of the head. This was not the case in our sitting condition; the infants were supported at the back but not at the head, hence considerably more postural control was required than in the supine situation. In former the studies which showed more successful reaching in sitting, the infants received head support, thereby significantly reducing the task’s postural demands. The difference in postural challenge between the two testing situations of the present study also may explain why developmental changes in reaching behaviour could be observed more clearly in the sitting position. In conclusion our data indicate that the phase of secondary variability of the proximal control of reaching movements emerges between 4 and 6 months: at 6 months a collective preference for the use of biceps brachii as prime mover develops which is associated with more efficient reaching movements. References Carvalho, R. P., Tudella, E., Caljouw, S. R. M., & Savelsbergh, G. J. P. (2008). Early control of reaching: Effects of experience and body orientation. Infant Behavior and Development, 31, 23–33. De Graaf-Peters, V. B., Bakker, H., Van Eykern, L. A., Otten, E., & Hadders-Algra, M. (2007). Postural adjustments and reaching in 4- and 6-months-old infants: An EMG and kinematical study. Experimental Brain Research, 181, 647–656. Edelman, G. M. (1989). Neural Darwinism. The theory of neuronal group selection. Oxford: Oxford University Press. Fallang, B., Saugstad, O. D., & Hadders-Algra, M. (2000). Goal directed reaching and postural control in supine position in healthy infants. Behavioural Brain Research, 115, 9–18. Forssberg, H., & Hirschfeld, H. (1994). Postural adjustments in sitting humans following external perturbations: Muscle activity and kinematics. Experimental Brain Research, 97, 515–527. Hadders-Algra, M. (2004). General movements: A window for early identification of children at high risk of developmental disorders. Journal of Pediatrics, 145, S12–18. Hadders-Algra, M. (2000). The Neuronal Group Selection Theory: A framework to explain variation in normal motor development. Developmental Medicine and Child Neurology, 42, 566–572. Hadders-Algra, M., Brogren, E., & Forssberg, H. (1996). Training affects the development of postural adjustments in sitting infants. Journal of Physiology, 493, 289–298. Hadders-Algra, M., Heineman, K. R., Bos, A. F., & Middelburg, K. J. (2009). The assessment of minor neurological dysfunction using the Touwen Infant Neurological Examination. Developmental Medicine and Child Neurology, Epub ahead of print. Hopkins, B., & Prechtl, H. F. R. (1984). A qualitative approach to the development of movements during early infancy. In H. F. R. Prechtl (Ed.), Clin Dev Med No. 94 Continuity of neural functions form prenatal to postnatal life. Oxford: Blackwell Scientific Publications, pp. 179–197. Jeannerod, M. (1990). The neural and behavioural organization of goal-directed movements. Oxford: Clarendon Press. Kuhtz-Buschbeck, J. P., Stolze, H., Jöhnk, K., Boczek-Funcke, A., & Illert, M. (1998). Development of prehension movements in children: A kinematic study. Experimental Brain Research, 122, 424–432. Out, L., Van Soest, A. J., Savelsbergh, G. J. P., & Hopkins, B. (1998). The effect of posture on early reaching movements. Journal of Motor Behavior, 30, 260–272. Out, L., Savelsbergh, G. J. P., Van Soest, A. J., & Hopkins, B. (1997). Influence of mechanical factors on movement units in infant reaching. Human Movement Science, 16, 733–748. Savelsbergh, G. J. P., & Van der Kamp, J. (1994). The effect of body orientation to gravity on early infant reaching. Journal of Experimental Child Psychology, 58, 510–528. Thelen, E., & Spencer, J. P. (1998). Postural control during reaching in young infants: A dynamic systems approach. Neuroscience Biobehavioral Reviews, 22, 507–514. Thelen, E., Corbetta, D., Kamm, K., Spencer, J. P., Schneider, K., & Zernicke, R. F. (1993). The transition to reaching: Mapping intention and intrinsic dynamics. Child Development, 64, 1058–1098. Thelen, E., Corbetta, D., & Spencer, J. P. (1996). Development of reaching during the first year: Role of movement speed. Journal of Experimental Psychology: Human Perception and Performance, 22, 1059–1076. Trevarthen, C. (1984). How control of movements develops. In H. T. A. Whiting (Ed.), Human motor actions. Bernstein reassessed. Amsterdam: Elsevier, pp. 223–261. Turton, A., & Lemon, R. N. (1999). The contribution of fast corticospinal input to the voluntary activation of proximal muscles in normal subjects and in stroke patients. Experimental Brain Research, 29, 559–572. Van der Fits, I. B. M., Klip, A. W. J., Van Eykern, L. A., & Hadders-Algra, M. (1998). Postural adjustments accompanying fast pointing movements in standing, sitting and lying adults. Experimental Brain Research, 120, 202–216. Van der Fits, I. B. M., Klip, A. W. J., Van Eykern, L. A., & Hadders-Algra, M. (1999). Postural adjustments during spontaneous and goal-directed arm movements in the first half year of life. Behavioural Brain Research, 106, 75–90. Van der Fits, I. B. M., Otten, E., Klip, A. W. J., Van Eykern, L. A., & Hadders-Algra, M. (1999). The development of postural adjustments during reaching in 6–18-month-old infants: Evidence for two transitions. Experimental Brain Research, 126, 517–528. Van der Meer, A. L. H., Van der Weel, F. R., & Lee, D. N. (1995). The functional significance of arm movements in neonates. Science, 267, 693–695. Von Hofsten, C. (1991). Structuring of early movements: A longitudinally study. Journal of Motor Behavior, 23, 280–292.
Contents lists available at ScienceDirect
Infant Behavior and Development
Development of proximal arm muscle control during reaching in young infants: From variation to selection Hanneke Bakker a , Victorine B. de Graaf-Peters a , Leo A. van Eykern a , Bert Otten b , Mijna Hadders-Algra a,∗ a
Department Paediatrics – Developmental Neurology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands Institute of Human Movement Sciences, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
b
a r t i c l e
i n f o
Article history: Received 1 September 2008 Received in revised form 8 April 2009 Accepted 23 October 2009 Keywords: Motor development Infants Reaching EMG Variation Kinematics
a b s t r a c t Reaching movements are initiated by activity of the prime mover, i.e. the first activated arm muscle. We aimed to investigate the relationship between prime mover activity and kinematics of reaching in typically developing (TD) infants in supine and sitting position. Fourteen infants were assessed at 4 and 6 months during reaching in supine and supported sitting. Kinematics and EMG-activity of deltoid, pectoralis major, biceps (BB) and triceps brachii were recorded. Kinematic analysis focused on number of movement units (MUs) and transport MU (MU with longest duration). Prime mover use was variable, but at 6 months a dominance of BB emerged in both testing conditions. Kinematics were also variable, but with increasing age the number of MU decreased and the relative proportion of the transport MU increased. BB as prime mover at 6 months was related to a larger transport MU. Conclusion: Between 4 and 6 months BB prime mover dominance emerges which is related to relatively efficient reaching characteristics. © 2009 Elsevier Inc. All rights reserved.
1. Introduction Reaching seems a simple daily activity. But, on close inspection the task of reaching towards an object is far from simple: the multi-jointed upper limb has to be directed towards the object while simultaneously the position of hand and fingers has to be prepared for the pick-up of the object (Jeannerod, 1990). Therefore it is not surprising that the development of reaching runs a protracted course which extends into adolescence (Kuhtz-Buschbeck, Stolze, Jöhnk, Boczek-Funcke, & Illert, 1998). Our knowledge on typical development of reaching in early infancy is still limited. Nevertheless such information is a prerequisite for the understanding of deviant development. What we do know is the following. Immediately after birth the infant mainly produces spontaneous non-goal-directed activity (Hadders-Algra, 2004). Nevertheless, the neonate has some capacity for goal-directed arm activity (Hopkins & Prechtl, 1984; Van der Meer, Van der Weel, and Lee, 1995). From 3 months onwards the infant starts to produce more types of goal-directed arm activity (Thelen et al., 1993; Van der Fits, Klip, Van Eykern, & Hadders-Algra, 1999), in particular when an interesting object is present. In response to the presence of an object the infant may put the hands into the mouth or it may produce oscillating movements of the extended arm in the direction
∗ Corresponding author. Tel.: +31 50 3614247; fax: +31 50 3619158. E-mail address: [email protected] (M. Hadders-Algra). 0163-6383/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.infbeh.2009.10.006
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
31
of the object or into another direction (Van der Fits, Klip, et al., 1999). With increasing age, arm movements are more often directed towards the object and from 4 to 5 months onwards they generally end in successful grasping, to become virtually always successful at 6 months (Van der Fits, Klip, et al., 1999). Initially reaching movements are characterized by large variation: variation in movement velocity and in the so-called movement units (De Graaf-Peters, Bakker, Van Eykern, Otten, & Hadders-Algra, 2007; Fallang, Saugstad, & Hadders-Algra, 2000; Thelen, Corbetta, & Spencer, 1996). MUs are submovements of reaching, which are determined with the help of peaks in the velocity profile of the hand (Von Hofsten, 1991). At 4 months of age reaching movements consist of about 4 MUs. With increasing age the variation in the number of MUs and the number itself decrease to about 3 MUs at 6 months (Carvalho, Tudella, Caljouw, & Savelsbergh, 2008; De Graaf-Peters et al., 2007; Fallang et al., 2000). It is first in adolescence, i.e. around the age of 12 years, that the majority of reaching movements has the adult configuration of 1 MU (Kuhtz-Buschbeck et al., 1998). In parallel to the decrease in MUs an increase of the relative length and duration of the first MU, the so-called transport MU, occurs so that it gradually covers a large proportion of the approach to the target (De Graaf-Peters et al., 2007; Fallang et al., 2000; Kuhtz-Buschbeck et al., 1998). The performance of reaching is related to postural control and body position. At the age of 4–5 months reaches are more successful and have a shorter duration in sitting position than in supine (Carvalho et al., 2008; Out, Van Soest, Savelsbergh, & Hopkins, 1998; Savelsbergh & Van der Kamp, 1994). Conflicting data have been reported whether these positions also affect other kinematic characteristics of early reaching behaviour, such as the number of MUs or the straightness of the movement path (Carvalho et al., 2008; Out, Savelsbergh, Van Soest, & Hopkins, 1997; Out et al., 1998). The onset of a reaching movement is mediated by activity of one of the proximal arm or shoulder muscles, in particular by activation of the biceps brachii muscle or the deltoid muscle. The muscle activated first is called the ‘prime mover’. Adult persons select in a specific situation a specific prime mover. For example, in a task of pointing at self-paced speed in standing position some adults use fast, ballistic arm movements during which the deltoid muscle is the prime mover, whereas others use slower arm movements including some elbow flexion during which the biceps brachii is the prime mover (Van der Fits, Klip, Van Eykern, & Hadders-Algra, 1998). Young infants variably use the biceps brachii, triceps brachii, deltoid, trapezius or pectoralis major muscles as prime mover (Thelen & Spencer, 1998). No data are available on the development of prime mover use during infancy, nor is it known whether developmental changes in prime mover use are related to changes in the kinematics of reaching or whether it is affected by infant position. Guided by the ideas of the Neuronal Group Selection Theory (Edelman, 1989; Hadders-Algra, 2000) we hypothesized that variability in prime mover use at early age, i.e. in the period during which reaching rapidly grows into a consistently successful event, changes into preference for a specific muscle as prime mover. If this is the case, we expect that the selection of a specific prime mover is associated with improved kinematics of the reaching arm movement, i.e. reaches consisting of less MUs or reaches with proportionally large transport MUs. We hypothesize that the developmental change may be observed in supine and sitting position. The aim of the present study was to test these hypotheses. We investigated in typically developing infants (1) the development of prime mover use during reaching at the ages of 4 and 6 months, where we paid specific attention to variation and selection, (2) the relation between prime mover use and kinematical characteristics of reaching at 4 and 6 months and (3) whether the effect of position, i.e. lying supine or sitting in an infant chair, affects prime mover use and the relation between prime mover use and kinematical features of reaching. The two different conditions differ from each other in terms of postural challenge – the challenge being larger in the sitting than in the supine position (Van der Fits, Klip, et al., 1999) – and in challenge to produce a successful reach. The latter challenge is larger in supine, when the arm behaves as a simple inverted pendulum, than in sitting where arm behaviour can be described in terms of a simple pendulum (Out et al., 1997). A way to deal with a larger mechanical challenge could be a larger degree of antagonistic co-activation (Out et al., 1997). Therefore we also assessed the degree of co-activation between biceps and triceps brachii. In order to address the above questions reaching movements in supine and sitting position of 14 typically developing infants were studied longitudinally at the ages of 4 and 6 months with the help of simultaneous EMG recordings of arm and shoulder muscles and kinematical recordings. We used testing situations which resembled daily life as much as possible, i.e. we did not use extra straps or devices to support the head to assist the child’s control of posture. Previously we reported about the development of postural control and its relation with the kinematics of reaching of these children (De Graaf-Peters et al., 2007); in the present article the focus is on the development of the prime mover in relation to the kinematics of reaching.
2. Methods 2.1. Participants Fourteen typically developing infants participated in this longitudinal study. The infants were tested at the ages of 4 and 6 months. Adequate recordings at the age of 4 months were obtained in 6 boys and 7 girls and in 6 boys and 6 girls at 6 months. The infants’ postmenstrual age at birth varied between 38 and 42 weeks (median: 39 weeks); birth weight between 2930 g and 4280 g (mean: 3582 g, SD: 468 g). The infants were recruited from amongst acquaintances of the investigators. The parents of the infants gave informed consent. The study was approved by the Medical Ethics Committee of the University Medical Center Groningen.
32
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
Fig. 1. Schematic drawing of testing conditions: lying supine (left) and sitting supported in infant chair (right). The chair had a back rest and a semicircular horizontal bar in front.
2.2. Procedure Reaching was assessed in two positions: lying supine and sitting upright with support in an infant chair (Fig. 1). No additional straps or specific head supporting devices were used. The infants were tested while wearing diapers only. Testing order of the two positions was determined randomly. Reaching movements were elicited by presenting small, attractive toys in the midline at arm length distance. A set of toys was used consisting of rings and small figures with sizes varying between 7 cm × 7 cm × 0.5 cm and 5 cm × 3 cm × 2 cm. Toys were only presented when the infant was alert and not crying. We aimed at recording at least ten reaching movements with the right arm in each position, but when the infant became fussy or tired the session was shortened. In order to confirm neurological integrity a standardized neurological examination (Touwen Infant Neurological Examination (TINE); Hadders-Algra, Heineman, Bos, & Middelburg, 2009) was carried out after each reaching session. 2.3. EMG, kinematical and video recordings EMG-activity was measured with surface electrodes on the following arm and shoulder muscles on the right side of the body: deltoid muscle (DE) pectoralis major (PM), biceps brachii (BB), and triceps brachii (TB). In the remainder of the paper these muscles are referred to as arm muscles. In addition, EMG-activity was recorded of postural muscles (Fig. 2). The EMG data were recorded continuously, digitized and fed into a personal computer with the help of POLY, a software program for long-lasting polygraphic recordings (sampling rate 500 Hz; Inspector Research Systems, Amsterdam, The Netherlands).
Fig. 2. Velocity profile of a successful reach of a sitting infant aged 6 months. The combination of one acceleration and one deceleration forms one MU; the trial consists of 3 MU.
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
33
The entire reaching session was video recorded with a split-screen allowing for a simultaneous lateral and frontal view of the infant. The video recordings were time-coupled to the EMG registration. The registration took about 0.5 h. Simultaneous with the EMG, arm movements were recorded kinematically with an ELITE system (BTS, Milan, Italy) in a two-camera configuration at a sampling frequency of 50 Hz. A reflective marker was placed on the right wrist of the infant, i.e. on the styloid process of the radius. The kinematical recording was discontinuous; it started just before the toy was presented and lasted for 10 s. The periods of kinematical sampling were indicated on the EMG recording.
2.4. Data analyses The video registration was used for multiple purposes. First, the video recordings were used to determine whether the infant was in an appropriate behavioural and attentional state during a specific trial. Second, the video was used for a classification of the behaviour of arm movements during toy presentation. Four different movements were distinguished: (a) pre-reaching arm movements, which were defined as arm movements occurring in response to toy presentation, which were not directed towards the toy (Trevarthen, 1984), (b) reaching movements: arm movement in the direction of the toy which did not end in toy contact (‘reach’), (c) reaching movements which did end in toy contact but not in grasping of the toy (‘reach and touch’), and (d) reaching movements which resulted in grasping of the toy (‘grasp’). The movements were categorized into movements with the right arm (right arm was dominating the reaching task) or left arm (left arm was dominating the reaching task). The results are restricted to the behaviour of the right arm. Reaches were classified as successful when they either ended in toy contact or grasping. Third, the video analysis served the selection of reaching movements with the right arm, as EMG and kinematical data were only available for the right arm. EMG and kinematical data-analyses were restricted to movements with ended in toy contact or grasping and to trials with direction-specific postural activity (De Graaf-Peters et al., 2007). Direction-specific postural activity means primary recruitment of postural muscles on the side of the body opposing body sway. For instance, reaching, which induces a forward sway of the body, is accompanied by direction specific postural activity in the dorsal postural muscles (Forssberg & Hirschfeld, 1994). The number of trials which fulfilled the behavioural criteria and which had an appropriate EMG and kinematical recording is displayed in Table 1. A computer algorithm was used for the EMG analyses to detect phasic muscle activity. The algorithm used a derivative of the root mean square of a full rectified signal (200 ms moving window), and marked significant deviations from a fixed detection level. The detection level was based upon a long-term (3.7 s) mean baseline activity. EMG bursts were detected when the activity exceeded the detection level for at least 50 ms (Van der Fits et al., 1998). The EMG bursts were discriminated from heart activity. The prime mover was defined as the first arm muscle which showed phasic activity around reaching onset. For biceps brachii and triceps brachii we also determined the presence of antagonistic co-activation. Co-activation was considered to be present when the antagonist showed phasic activity within 50 ms after the onset of the agonist. For each infant, each condition and each age the following parameters were calculated: (1) percentage of trials during which a specific arm muscle was used as a prime mover. (2) The preference prime mover, which was arbitrarily defined as the arm muscle being the prime mover in at least 50% of the trials. (3) The latencies of recruitment of the other arm muscles, defined as the time interval between the onset of the prime mover and the onset of activity in any other arm muscle. For each infant, age, and position median latency values were calculated. Offline kinematical analysis was carried out with the help of the software package DataMonster (E. Otten, Institute of Human Movement Sciences, University of Groningen). The data were filtered using a low-pass filter of 6 Hz (zero time-lag filter). Arm movements were analysed in two dimensions. The onset of arm movement was defined as the moment at which the velocity of the wrist increased ≥5% of peak velocity, the end of the movement was defined as the moment at which wrist velocity decreased to ≤5% of peak velocity. In the kinematical analysis, only trials with a clearly demarcated start and stop were included. The following parameters were used to describe the reaching movements: (1) the number of MU per trial. A MU consisted of one acceleration and one deceleration in the velocity profile of the wrist marker (Fig. 3). (2) The total duration of the reaching movement. (3) The relative duration of the first MU (i.e. the transport MU) in relation to total duration of the arm movement. (4) Maximum reaching velocity. Median values of the parameters were calculated for each infant, each condition, and each age.
Table 1 Number of trials analysed per individual for EMG and kinematical recordings. Age
EMG recordings
Kinematical recordings
Supine
4 Months 6 Months
Sitting
Supine
Sitting
n
Med
Range
n
Med
Range
n
Med
Range
n
Med
Range
13 12
11 13
3–20 10–15
12 12
9 12
4–13 7–17
9 10
4 5
3–9 3–12
12 12
5 6
3–10 4–11
n = number of infants and med = median value.
34
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
Fig. 3. Developmental trajectories of prime mover preference. Each line represents the development of one infant. BB: biceps brachii, DE: deltoid muscle, PM: pectoralis major, and Non: no specific preference for prime mover. Effect of age on prime mover preference, Wilcoxon; supine: p < 0.05, sitting: p < 0.01.
2.5. Statistics Statistical analysis was performed using the computer package SPSS (version 12.1). Analyses were performed on infant level this means that for the EMG latencies and the kinematical data first median values of the reaches of each infant were calculated on the basis of which further analyses were performed, i.e. the effect of age and position was based on these median values. The non-parametric Wilcoxon test was used to evaluate the effect of age and position. Due to data-loss in the kinematical recordings – a well known problem in infant research (Van der Fits, Otten, Klip, Van Eykern, & HaddersAlgra, 1999) – and thereby loss of pairs of data per infant, it was not possible to use the Wilcoxon test for the analyses of the kinematical data and the analyses of the relationships between the EMG and reaching data. We therefore decided to use the Mann–Whitney U test. Throughout the analyses, differences with p-values < 0.05 were considered statistically significant.
3. Results 3.1. Reaching and grasping in supine and sitting position At 4 months of age 11 out of 13 infants were able to reach and touch or grasp the toy successfully at least once, 11 infants produced successful reaching in supine, five in sitting position. Position seemed to affect the type of reaching movements at 4 months: in supine the infants showed more often successful reaching movements and in sitting more often pre-reaching movements (Table 2). The effect of position did not however reach statistical significance. At 6 months the number of successful reaches had increased and that of pre-reaching movements decreased (Table 2). At this age position did have a significant effect on the success of reaching: reaching was more often successful in supine than in sitting (Table 2; Mann–Whitney U test: p < 0.05). Preliminary data analyses indicated that there were no differences in prime mover preference and kinematical characteristics between movements classified as ‘reach and touch movements’ and successful grasping movements. For this reason the data of these movements were pooled.
Table 2 Types of arm movement in supine and sitting position at 4 and 6 months of age. Grasp
Reach and touch
Reach
Pre-reach
4 Months
Supine (n = 11) Sitting (n = 10)
21 (0–82) 8 (0–40)
0 (0–53) 7 (0–100)
19 (0–86) 7 (0–20)
6 Months
Supine (n = 12) Sitting (n = 12)
94** (67–100) 67*,† (15–100)
6 (0–23) 0 (0–15)
0 (0–13) 0 (0–18)
38 (0–100) 66 (0–100) 0** (0–8) 27**,†† (0–55)
n = number of infants. The data indicate median percentages per group, with minimum and maximum values between brackets. Effect of age: with increasing age infants made less pre-reaching movement (Wilcoxon—**supine: p < 0.01, sitting: p < 0.01) and more successful grasping movements (Wilcoxon—**supine: p < 0.01, *sitting: p < 0.05). Effect of position: in sitting the infants made more pre-reaching movements (Mann–Whitney U test: † p < 0.05); in supine the infants made more successful reaching movements (†† p < 0.01; Note: the low number of paired data precluded the use of the Wilcoxon test).
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
35
Fig. 4. Typical example of EMG-activity of arm and postural muscles during successful reaching in an infant aged 6 months during sitting. The upper four channels represent arm muscle activity, phasic activity is denoted by the horizontal underlining. Here BB is the first arm muscle activated, hence is the prime mover. Postural activity is direction specific, i.e. consists of NE and TE recruitment. DE: deltoid muscle, PM: pectoralis major, BB: biceps brachii, TB: triceps brachii, ST: sternocleidomastoid, NE: neck extensor, RA: rectus abdominis, TE: thoracal extensor, and LE: lumbar extensor.
3.2. Development of prime mover preference Fig. 4 shows a typical example of EMG-activity during reaching. At 4 months the identity of the prime mover showed substantial intra- and interindividual variation. Despite the variation, most infants had a preference for a specific prime mover: in supine three infants used preferentially BB, two DE, four PM and two infants did not have a specific prime mover preference. In sitting position all infants had preference prime mover; four infants had a preference for BB, four for DE and one for PM (Fig. 4). At 6 months all infants had a preference prime mover within their variable repertoire of prime movers in both conditions. The identity of prime mover still varied, but most infants used BB as prime mover. This was true for the supine and sitting position (Fig. 4). The increase in the use of BB as preference prime mover between 4 and 6 months was statistically significant (Wilcoxon—supine; p < 0.05, sitting; p < 0.01). The increase could not be attributed to the fact that infants produced more trials which could be analysed at 6 months than at 4 months. Co-activation of BB and TB was virtually restricted to reaching movements during which BB was the prime mover. The frequency of BB–TB co-activation in these trials was however largely variable. It varied across age, condition and infant from 0% to 100% of trials during which BB was the prime mover (median values: 4 months supine 33%, sitting 70%, 6 months supine 38%, sitting 67%). The degree of co-activation between the supine and sitting condition was not statistically significant. 3.3. Development of kinematical characteristics Also the kinematical data were characterized by variation. At 4 months reaching movements consisted of 3.5 and 4 MU (median values, sitting and supine position, respectively). The number of MU decreased with increasing age: in sitting at 6 months reaches consisted of 2.5 MU (difference with 4 months, Wilcoxon—p = 0.02), in supine at 6 months of 3 MU (difference with 4 months: n.s.; Fig. 5). The number of MU was not affected by position. The relative duration of the transport unit at 4 months in supine and sitting position is 30% of the total duration of the reaching movement. The same was true for reaching movements in supine at 6 months. But in the sitting position the relative duration of the transport unit increased significantly with increasing age to 38% (Fig. 5, Wilcoxon—p = 0.04). No significant difference was found between the relative duration of the transport unit between the supine and sitting position. The total duration of the reaching movement at 4 months in supine was about 0.55 s, in sitting about 0.65 s. Movement duration decreased with increasing age: at 6 months till about 0.50 s in both conditions (Wilcoxon, decrease supine: n.s., sitting p < 0.05) (Fig. 6). 3.4. Relationship between prime mover preference and kinematical characteristics At the age of 4 months prime mover preference was not related to the kinematics of reaching; this was true for both positions. At 6 months the relative duration of the transport unit was related to prime mover preference. The transport unit had a longer duration in infants who used BB as preferred prime mover than in infants who used DE or PM as prime mover (Fig. 7, Mann–Whitney U test: supine; p < 0.01, sitting; p < 0.01). The other kinematical parameters of reaching were not related to prime mover preference.
36
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
Fig. 5. Development of kinematical characteristics in supine and sitting position, i.e. number of MU and relative duration of transport unit at 4 and 6 months. Bold horizontal lines indicate median values; the boxes represent interquartile ranges and the vertical lines complete ranges. Age effect of number of MU, Wilcoxon; sitting: p = 0.02. Age effect of relative duration of transport unit, Wilcoxon; sitting: p = 0.04.
Fig. 6. Development of duration of reaching movements in supine and sitting position. Bold horizontal lines indicate median values; the boxes represent interquartile ranges and the vertical lines complete ranges. Age effect of total reaching duration, Wilcoxon; sitting: p < 0.05.
4. Discussion The present study indicates that the improvement in success of reaching with increasing age is paralleled by selection of the biceps brachii as prime mover, a selection which – in turn – is related to a longer duration of the transport MU. The selection process is independent of the infant’s position during reaching. Before discussing the neurodevelopmental significance of our findings, we would like to address an important methodological point, namely that we restricted our analysis to trials with direction-specific postural activity. This means that we selected reaching movements which are relatively good organized, as the presence of direction-specific postural activity is related to more success of reaching, and reaching movements with less MU and a longer duration of the transport MU (De Graaf-Peters et al., 2007). We decided to restrict ourselves to this selection in order to avoid confounding with the effect of direction-specificity. Another methodological point which deserves attention is our testing condition. In order to facilitate EMG- and kinematic recording infants were assessed while being undressed. For practical reasons a diaper was worn. It is possible that the diaper may have provided some postural support, but if so, this was true for both testing conditions. The sitting condition we
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
37
Fig. 7. Development of relation between prime mover and kinematical characteristics. Number of MU and relative duration of transport unit in groups of infants with a specific prime mover preference at 6 months in supine (left panels) and sitting (right panels) position. Bold horizontal lines indicate median values; boxes represent interquartile ranges; vertical lines complete ranges. Effect of prime mover preference on relative duration of transport unit, Mann–Whitney U test: p < 0.02.
used differed largely from the sitting condition used in previous studies (Carvalho et al., 2008; Out et al., 1997, 1998; Savelsbergh & Van der Kamp, 1994). In the previous studies the infants received firm postural support in the sitting position whereas in the present study postural support in sitting was reduced to a minimum. This means that in previous studies the difference in postural challenge between supine and sitting was little if any, leaving the infant in the sitting situation with the mechanical advantage of the inverted pendulum condition (Out et al., 1997). The advantage was reflected in better performance of reaching in sitting than in supine. In the present study, the mechanical advantage for the reaching arm in the sitting condition was entirely overruled by the disadvantage of the large postural challenge. As a result the infants in the present study produced more successful reaches in supine than in the sitting position. Our study showed that infants showed substantial within- and between-subject variation in use of prime mover. Nevertheless, it was also clear that most infants already at the age of 4 months had a preference for a specific prime mover. At this age prime mover preference showed substantial between-subject variation. Major part of between-subject variation had disappeared at the age of 6 months, when most infants used the biceps brachii as prime mover. Selection of biceps brachii as prime mover presumably is related to the way infants in general performs reaching movements: this usually includes some degree of elbow flexion (cf. Van der Fits et al., 1998). Similar developmental changes have been reported for postural adjustments during reaching: at 4 months some degree of individual preference for a specific postural adjustment is present but with large between-subject variation; at 6 months a general preference for a specific postural pattern, i.e. the pattern in which all direction-specific muscles are recruited, exists (De Graaf-Peters et al., 2007). These developmental changes fit well to the ideas of the Neuronal Group Selection Theory (Edelman, 1989; Hadders-Algra, 2000). According to NGST motor development is characterized by two phases of variability. During the phase of primary variability the infant displays variable motor behaviour, but is not able to adapt motor behaviour to the specifics of the situation. During the phase of secondary variability the infant gradually learns to select from its repertoire of motor strategies the best motor solution fitting to the situation. The data suggest that infants enter the phase of secondary variability for prime mover use and postural adjustment between 4 and 6 months. The concurrence of both events seems logical as both deal with control of proximal muscles, in which the corticospinal system plays a less prominent role than in the control of distal muscles (Turton and Lemon, 1999; Hadders-Algra, 2000). The group preference at 6 months for biceps brachii as prime mover suggests that the selection for biceps brachii is associated with some benefit. Indeed, our findings support this notion, as the selection of the biceps brachii as prime mover was related to a longer duration of the transport MU. Selection of biceps brachii was related to the duration of the transport MU but not to the number of MUs. This might imply that the transport MU is more dependent on proximal and postural muscle control than the number of MUs (Jeannerod, 1990). Interestingly, adult persons do not have such a strict preference for biceps brachii as prime mover: about half of adult persons use biceps brachii as prime mover and the other half the deltoid muscle (Van der Fits et al., 1998). This means that development of prime mover selection has not been completed at 6 months of age. Further studies could shed light on the age at which the adult stage of prime mover selection is reached.
38
H. Bakker et al. / Infant Behavior & Development 33 (2010) 30–38
The use of biceps brachii as prime mover was associated in a variable way with triceps brachii co-activation. In other words, antagonistic co-activation was an inconsistent finding and did not depend on position or age. This might imply that at early age variation and exploration are more important determinants of motor control than antagonistic co-activation (cf. Hadders-Algra, Brogen, & Forssberg, 1996). It is noteworthy that the present study showed that success of reaching in early infancy is greater in supine than in supported sitting, this in contrary to former studies as is mentioned in the introduction. An explanation for more successful reaching in supine might be the difference in postural challenge. The supine situation offers a large base of support including stable support of the head. This was not the case in our sitting condition; the infants were supported at the back but not at the head, hence considerably more postural control was required than in the supine situation. In former the studies which showed more successful reaching in sitting, the infants received head support, thereby significantly reducing the task’s postural demands. The difference in postural challenge between the two testing situations of the present study also may explain why developmental changes in reaching behaviour could be observed more clearly in the sitting position. In conclusion our data indicate that the phase of secondary variability of the proximal control of reaching movements emerges between 4 and 6 months: at 6 months a collective preference for the use of biceps brachii as prime mover develops which is associated with more efficient reaching movements. References Carvalho, R. P., Tudella, E., Caljouw, S. R. M., & Savelsbergh, G. J. P. (2008). Early control of reaching: Effects of experience and body orientation. Infant Behavior and Development, 31, 23–33. De Graaf-Peters, V. B., Bakker, H., Van Eykern, L. A., Otten, E., & Hadders-Algra, M. (2007). Postural adjustments and reaching in 4- and 6-months-old infants: An EMG and kinematical study. Experimental Brain Research, 181, 647–656. Edelman, G. M. (1989). Neural Darwinism. The theory of neuronal group selection. Oxford: Oxford University Press. Fallang, B., Saugstad, O. D., & Hadders-Algra, M. (2000). Goal directed reaching and postural control in supine position in healthy infants. Behavioural Brain Research, 115, 9–18. Forssberg, H., & Hirschfeld, H. (1994). Postural adjustments in sitting humans following external perturbations: Muscle activity and kinematics. Experimental Brain Research, 97, 515–527. Hadders-Algra, M. (2004). General movements: A window for early identification of children at high risk of developmental disorders. Journal of Pediatrics, 145, S12–18. Hadders-Algra, M. (2000). The Neuronal Group Selection Theory: A framework to explain variation in normal motor development. Developmental Medicine and Child Neurology, 42, 566–572. Hadders-Algra, M., Brogren, E., & Forssberg, H. (1996). Training affects the development of postural adjustments in sitting infants. Journal of Physiology, 493, 289–298. Hadders-Algra, M., Heineman, K. R., Bos, A. F., & Middelburg, K. J. (2009). The assessment of minor neurological dysfunction using the Touwen Infant Neurological Examination. Developmental Medicine and Child Neurology, Epub ahead of print. Hopkins, B., & Prechtl, H. F. R. (1984). A qualitative approach to the development of movements during early infancy. In H. F. R. Prechtl (Ed.), Clin Dev Med No. 94 Continuity of neural functions form prenatal to postnatal life. Oxford: Blackwell Scientific Publications, pp. 179–197. Jeannerod, M. (1990). The neural and behavioural organization of goal-directed movements. Oxford: Clarendon Press. Kuhtz-Buschbeck, J. P., Stolze, H., Jöhnk, K., Boczek-Funcke, A., & Illert, M. (1998). Development of prehension movements in children: A kinematic study. Experimental Brain Research, 122, 424–432. Out, L., Van Soest, A. J., Savelsbergh, G. J. P., & Hopkins, B. (1998). The effect of posture on early reaching movements. Journal of Motor Behavior, 30, 260–272. Out, L., Savelsbergh, G. J. P., Van Soest, A. J., & Hopkins, B. (1997). Influence of mechanical factors on movement units in infant reaching. Human Movement Science, 16, 733–748. Savelsbergh, G. J. P., & Van der Kamp, J. (1994). The effect of body orientation to gravity on early infant reaching. Journal of Experimental Child Psychology, 58, 510–528. Thelen, E., & Spencer, J. P. (1998). Postural control during reaching in young infants: A dynamic systems approach. Neuroscience Biobehavioral Reviews, 22, 507–514. Thelen, E., Corbetta, D., Kamm, K., Spencer, J. P., Schneider, K., & Zernicke, R. F. (1993). The transition to reaching: Mapping intention and intrinsic dynamics. Child Development, 64, 1058–1098. Thelen, E., Corbetta, D., & Spencer, J. P. (1996). Development of reaching during the first year: Role of movement speed. Journal of Experimental Psychology: Human Perception and Performance, 22, 1059–1076. Trevarthen, C. (1984). How control of movements develops. In H. T. A. Whiting (Ed.), Human motor actions. Bernstein reassessed. Amsterdam: Elsevier, pp. 223–261. Turton, A., & Lemon, R. N. (1999). The contribution of fast corticospinal input to the voluntary activation of proximal muscles in normal subjects and in stroke patients. Experimental Brain Research, 29, 559–572. Van der Fits, I. B. M., Klip, A. W. J., Van Eykern, L. A., & Hadders-Algra, M. (1998). Postural adjustments accompanying fast pointing movements in standing, sitting and lying adults. Experimental Brain Research, 120, 202–216. Van der Fits, I. B. M., Klip, A. W. J., Van Eykern, L. A., & Hadders-Algra, M. (1999). Postural adjustments during spontaneous and goal-directed arm movements in the first half year of life. Behavioural Brain Research, 106, 75–90. Van der Fits, I. B. M., Otten, E., Klip, A. W. J., Van Eykern, L. A., & Hadders-Algra, M. (1999). The development of postural adjustments during reaching in 6–18-month-old infants: Evidence for two transitions. Experimental Brain Research, 126, 517–528. Van der Meer, A. L. H., Van der Weel, F. R., & Lee, D. N. (1995). The functional significance of arm movements in neonates. Science, 267, 693–695. Von Hofsten, C. (1991). Structuring of early movements: A longitudinally study. Journal of Motor Behavior, 23, 280–292.