Infant Behavior & Development 37 (2014) 76–85

Contents lists available at ScienceDirect

Infant Behavior and Development

The impact of object carriage on independent locomotion Diane Marie J. Mangalindan ∗ , Mark A. Schmuckler ∗ , Shelly-Anne Li University of Toronto Scarborough, Canada

a r t i c l e

i n f o

Article history: Received 3 April 2013 Received in revised form 25 September 2013 Accepted 24 December 2013 Available online 23 January 2014 Keywords: Object carriage Walking skill Arm guard position Gait parameter Arm posture

a b s t r a c t The current study examined whether carrying objects in one’s hands influenced different parameters associated with independent locomotion. Specifically, 14- and 24-month-olds walked in a straight path under four conditions of object carriage – no object (control), one object carried in one hand (one object-one hand), two objects carried in each of the hands (two objects-two hands), and one object carried in both hands simultaneously (one objecttwo hands). Although carrying objects failed to influence a variety of kinematic parameters of gait, it did affect children’s arm postures, with children adopting less mature arm positions when carrying objects. Finally, arm position was related to walking skill, but only for older children when they were not carrying objects. These findings indicate that although a relation does exist between arm positions and gait parameters, this relation is easily disrupted by carrying loads, even small ones. © 2014 Elsevier Inc. All rights reserved.

Learning to move efficiently around one’s world certainly has significant survival value, but it also presents numerous challenges for young learners. As an example, the advent of independent, upright walking increases one’s locomotor efficiency in that it allows children to go from one location to another while leaving their hands free to do whatever they wish. This efficiency of movement comes with a cost, however, in that it poses a wealth of challenges for young walkers, including maintaining one’s balance at each step, given that one’s centre of mass continually moves outside of one’s base of support (Garciaguirre, Adolph, & Shrout, 2007). Maintaining balance is especially difficult due to children’s disproportionate body dimensions (i.e., being top-heavy, Palmer, 1994) and inexperience with the dynamic disequilibrium (Bril & Brenière, 1992) brought about by standing on two feet. Regardless, by the end of the first year, children accomplish this feat. What makes this development possible? According to Bril and Brenière (1992), learning to walk is a two-stage process wherein the crucial improvement is in balance control. The first stage, which occurs within four to five months of independent walking, involves integrating all the different gait elements (e.g., step width, walking velocity), allowing them to work together to support the body during forward propulsion. This achievement is followed by the second stage, which involves a fine tuning of these parameters with respect to changing environmental demands. In other words, children learn to put all the important elements into play, and once they get the hang of it, they then adjust to varying situations. Accordingly, practice is vital in this process. The role of walking experience in locomotor development is reflective of a child’s opportunities for learning (Schmuckler, 1996) and has been examined under a variety of situations, including crossing barriers (Kingsnorth & Schmuckler, 2000; Schmuckler, 1996, 2013b), climbing stairs and slopes (Adolph, 1995; Adolph, Eppler, & Gibson, 1993), gap crossing (Zwart, Ledebt, Fong, de Vries, & Savelsbergh, 2005), and carrying loads (Garciaguirre et al., 2007; Vereijken, Pederson, & Størksen, 2009). For instance, Kingsnorth and Schmuckler (2000) examined the factors that influenced children’s abilities to cross over barriers and found that for 14- to 30-month olds, from a range of anthropomorphic, skill, and experiential variables, walking experience best predicted children’s performance. Similarly, Zwart et al. (2005) found that the best predictor of gap

∗ Corresponding authors. E-mail addresses: [email protected] (D.M.J. Mangalindan), [email protected] (M.A. Schmuckler). 0163-6383/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.infbeh.2013.12.008

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

77

crossing in toddlers was walking experience, and Vereijken et al. (2009) found that children’s abilities to compensate for different loads placed on their bodies were related to walking experience. These studies underscore the value of experience, and suggest that with more opportunities, children learn to better deal with the challenges that accompany locomotion. One obvious indication of such learning is observable in improvements in children’s walking skill, with previous work demonstrating that walking experience is correlated with walking skill (Adolph, 1995; Bril & Brenière, 1992; Burnett & Johnson, 1971). Walking skill involves an inter-play of several variables, and can be reliably quantified into a number of different components that are generally related to one another (Adolph, 1995), such as stride length, step width, dynamic base, and so on (Adolph, 1995; Bril & Brenière, 1992; Kingsnorth & Schmuckler, 2000; Ledebt, 2000). Overall, such studies have demonstrated that walking skill, as assessed by parameters such as these, significantly predicts performance across a range of locomotor tasks, including the ability to traverse up and down slopes (Adolph, 1995). Locomotor abilities do not develop independently of the growth of other motor components, however. Ledebt (2000), for example, has demonstrated a relation between such kinematic parameters of walking skill and motor control in one’s upper torso. Specifically, Ledebt observed that changes in the breadth of base were closely related to changes in arm posture, such that as step width became more adult like (i.e., narrower distance between two feet), arm positions also became more mature (i.e., gradually lowered, with arms swinging in opposite phase). Ledebt (2000) argued that although the hands are no longer required for actual forward movement (as opposed to, say, crawling), the arms still play a role (via their position) in controlling posture and balance when walking. Along these lines, arm position (or “arm guard” as it is called) is characterized by a combination of specific rotations in the shoulders, the degree to which the elbows are bent, and the position of the entire arm. Reciprocal arm swing marks the most mature form of arm position, and is important for walking in that it minimizes the vertical displacement of the centre of mass by counterbalancing trunk rotation as a person steps forward. Thus, reciprocal arm swing increases postural stability and stepping efficiency (Elftman, 1939; Murray, Kory, & Bernard, 1967) and mimics adults’ arm position when walking. However, when learning to walk, infants do not initially adopt reciprocal arm swing. Instead, Burnett and Johnson (1971) found that when first walking, children employ a high guard arm posture, described as externally rotated at the shoulder and flexed at the elbow, to help maintain their balance. This high guard position represents, then, the least mature (or adult-like) arm position. Children adopt this position for only a short period of time (Kubo & Ulrich, 2006; Ledebt, 2000), however, and quickly learn to adjust their arm positions throughout childhood, achieving reciprocal arm swing by 18 months, and using it systematically by 42 months (Ledebt & Bril, 2000; Sutherland, Olshen, Cooper, & Woo, 1980). Interestingly, Burnett and Johnson (1971) also observed that children will return to less mature arm positions to cope with situations that impede their balance; thus, situations that impede postural control affect arm guard position. One common real-world situation that might potentially influence posture occurs when children carry objects with their hands. Carrying objects can impose changes on children’s posture because it could not only potentially change children’s centre of mass (Garciaguirre et al., 2007), but it also competes with the use of the arms for balance. Object carriage is actually quite a natural task for children. Karasik, Adolph, Tamis-LeMonda, and Zuckerman (2011) examined spontaneous carrying of objects in the home, and found that children frequently carried objects from one location to another, and they did so even before the advent of independent walking. For instance, crawling infants would put objects in their mouths or carry objects under their arms. However, walking infants largely transported objects using their hands, with the most preferred objects they carried being small toys. These authors also found that walking experience facilitated object transport, with more experienced children being better able to adjust their posture when carrying. Interestingly, Karasik et al. (2011) also found that less experienced walkers actually benefited (i.e., fell less) when carrying objects, suggesting that hand occupancy helped early walkers to be more vigilant in controlling their balance. Carrying objects is also of interest in that it creates a naturalistic situation of load carrying for children. Previous work on load carrying has focused mainly on how children adapt when external loads (i.e., weights) are literally strapped onto their bodies. This work has revealed that loads impact young infants’ walking, causing changes in gait parameters and walking disruptions such as falls and missteps (Garciaguirre et al., 2007; Kistner, Fiebert, & Roach, 2012; Vereijken et al., 2009). With experience, however, children are able to make the necessary postural adjustments (e.g., leaning against the load, Garciaguirre et al., 2007) and footfall modifications (e.g., walking slower, making smaller steps, Adolph & Avolio, 2000; Garciaguirre et al., 2007; Vereijken et al., 2009) when carrying loads. Furthermore, such interlimb changes may be due to ongoing changes in the central nervous system (Gesell, 1939). This idea was supported by Corbetta and Bojczyk (2002) who examined whether changes in reaching patterns co-occurred with the onset of independent walking. Specifically, the researchers assessed whether infants reverted to two-handed reaching at the onset of walking and resumed adaptive uni-modal reaching when they were better able to control balance while walking. Examining infants longitudinally in posture and locomotion tasks, as well as in reaching tasks, the researchers found that infants’ two-handed reaching increased at the onset of walking, but adaptive reaching occurred after infants became more stable in upright locomotion. Because children were learning upright balance which required greater upper arm coupling, children also opted for arm coupling in reaching for objects. The researchers argued that such changes reflect ongoing neuromotor reorganizations. In the present study, instead of having external loads strapped onto children’s bodies, the children were asked to carry objects that varied the load distribution on their hands, and thus, systematically changed the dynamics of the hand posture and movements. Given the work on load carriage, it is reasonable to wonder what impact carrying objects will have on different components of children’s locomotion. Specifically, the present study examined how object carrying would impact: (a) walking skill (i.e., gait parameters), (b) arm guard position, and (c) the relation between walking skill and arm guard

78

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

position. First, based on the work on load carrying, it was expected that object carrying will influence children’s gait parameters. Carrying objects would lead children to employ less mature walking skills. With more walking experience, children should be better able to make the appropriate gait modifications (Garciaguirre et al., 2007; Karasik et al., 2011), demonstrating less of an influence of object carriage. Second, based on the work on arm posture, it was expected that carrying objects would influence the type of arm position children employ when walking. Carrying objects would cause young toddlers to show less mature arm guard positions. Similar to walking skill, increased walking experience should lead children to better able to adapt to object carriage, making the appropriate postural adjustments (Garciaguirre et al., 2007; Karasik et al., 2011) with their arms. And finally, although the findings of a co-development between gait parameters (e.g., step width) and arm position (Ledebt, 2000) exist, there were no clear predictions for the relation between walking skill and arm guard position as a function of load carriage. 1. Methods 1.1. Participants The final sample of participants consisted of sixteen 14-month-old (M age = 14.1, SD = 0.6, 5 male, 11 female) and sixteen 24-month-old (M age = 23.8, SD = 1.0, 10 male, 6 female) children. An additional 13 children were tested (nine 14-month-olds, and four 24-month-olds), but data from these children were not included due to tiredness (n = 1), fussiness (n = 1), crying (n = 3), no interest (n = 2), and not wanting to walk carrying toys (n = 6). These age groups were selected based on previous work suggesting a developmental transition in the paradigm employed1 (Kingsnorth & Schmuckler, 2000). Children were recruited from a list maintained by the Laboratory for Infant Studies located at the University of Toronto Scarborough. The families of these children were all from in or around the Scarborough, Ontario community. All children received a toy and a certificate for participating in this study. Detailed information regarding ethnic background and socioeconomic status were not collected. 1.2. Conditions and design The children completed four conditions which varied according to object carriage. Each condition consisted of three trials (i.e., three repetitions of the trial) in total. In the no objects condition, the child did not carry an object, leaving both hands unoccupied. In the one object-one hand condition, the child carried a single toy in their preferred hand. In the two objects-two hands condition, the child carried one toy in each hand. Finally, in the one object-two hands condition, the child carried one big toy using both hands. For each child, the no object condition was conducted first. This condition was conceptually considered as a control, and employed to obtain a baseline of typical walking skill and arm guard position. This condition was then followed by the remaining three “experimental” conditions, with the order of these conditions randomized for each child. The four conditions were run in a blocked fashion, with children completing up to three trials in each condition. 1.3. Experimental apparatus and procedure The experiment took place in a long hallway. A Sony digital video camera (DCR-TRV103) recorded the entire experimental session. This camera was located in the middle of the hallway, and thus provided a sagittal view of children as they walked from one end of the hallway to the other. On each trial a 6.0 m (length) × 1.5 m (width) strip of white paper was placed in the hallway. The child’s parent sat at one end of the paper, and the experimenter and the child sat at the other end. Prior to walking to the parent, a 10.2 cm × 3.8 cm piece of cardboard was taped to the bottoms of the child’s shoes. This piece of cardboard contained two inkpads. The first inkpad was a triangular piece of moleskin (3.0 cm × 2.5 cm × 3.0 cm) that was affixed such that its tip touched the midpoint of the top edge of the card. The second inkpad was a square piece of moleskin (2.5 cm × 2.5 cm) that was affixed such that it was aligned with the bottom edge, at the centre of the card. Washable ink was applied to the inkpads with a paintbrush immediately before the cards were attached to the shoes. Once the card was in place, the experimenter placed the child at the end of the sheet and encouraged the child to walk towards the parent. This process was repeated on every trial. If a trial was not completed successfully (e.g., the child stopped midway, turned back, walked off), a new piece of paper was positioned and the trial was rerun. The majority (69%) of children completed all 12 trials. The remaining 31% of children (six 14-month-olds and four 24-month-olds) completed between 10 and 11 trials. Children carried different toys in the different conditions. Although every effort was made to make the toys consistent for each condition for each child, some children refused to carry some toys, and some children became bored of using the same toy for the entire experiment. Accordingly, to maintain children’s interest, different toys were used during the course of the study when needed. Overall, the toys employed were light, weighing an average of 1.33 lbs. The actual toys used included two toy blocks (each weighed 1 lb), two stuffed animal (Barney and a stuffed dog; each weighed 2 lbs), one puzzle

1 Compared to children with less walking experience, children with more walking experience have been found to display more mature and adult-like motoric behaviours, in both walking skills (i.e., gait parameters) and arm guard positions when asked to walk on a flat surface.

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

79

toy (weighed 2 lbs) shaped like a football (which could be dismantled, and each piece could also be used and weighed less than 1 lb), and two cars (each weighed 1 lb). In the one object-one hand condition, only one smaller object (e.g., one car) was employed. In the two objects-two hands condition, only identically sized objects (e.g., two cars, two plastic balls) were used. And in the one object-two hands condition, a bigger-sized toy (e.g., stuffed animal, puzzle football), which required using both hands, was employed. 1.4. Dependent measures Walking experience. Parents provided information about their children’s onset of upright locomotion. Upright locomotion was defined as the age at which children could stand on their own, without any help, and walk 10 steps across a room (Schmuckler, 1996, 2013b). From this information, walking experience was derived by subtracting the date of onset from the child’s actual age at the time of the experiment. Walking skill. Walking skill was quantified using information from children’s inked footprint sequences (Adolph, 1995; Adolph et al., 1993; Kingsnorth & Schmuckler, 2000). This analysis focused on the middle portion of each child’s traverse down the pathway, given that it is this section of a continuous gait cycle at which children hit their most natural stride (Brenière, Bril, & Fontaine, 1989). For every trial, a minimum of three footprints per leg were examined for all children. For each footprint, XY coordinates of the tip of the toe print and the midpoint of the heel print were obtained using a transparent grid placed over the footprint sequence. The resolution of the grid was 0.25 cm. Using the XY footprint coordinates, a variety of skill related measures were calculated, including stride length, step length, step width, foot rotation, and dynamic base. Stride length is the distance between heel prints of the same foot. Step length is the distance between the heel strike of one foot and the heel strike of the other foot. Step width is a measure of the lateral distance between the heel strike of one foot and the heel strike of the other foot. Foot rotation involves the angle of a child’s toe relative to their heel (i.e., toe-in or toe-out); children who are more skilled at walking point their feet straight ahead (Adolph, 1995). Finally, dynamic base measures the “angle between the stride of one foot and the step on the other foot” (Adolph, 1995, p. 741); this angle can be obtained from looking at three step sequences. Dynamic base reflects how well children control their path of progression, with improvements in walking skill indicated by dynamic base angles approaching 180◦ (Adolph, 1995). Arm guard position. Arm guard position was assessed from videotaped records of the trials based on categories by Ledebt (2000); these categories are described in Table 1. For each trial, the main position (i.e., position held the longest during the trial) held by the children was determined and was given a score ranging from 1 to 5, which represented different arm guard positions. These scores were then averaged across trials to derive a single score for each condition. In the one object-one hand condition, the occupied hand was coded and analyzed. For both walking skill and arm posture measures, a primary rater coded all trials, and a secondary rater coded 60 randomly selected trials (31%) from each age group. Reliabilities were high for both arm posture (r = .87, p < .0001) and walking skill parameters (average r = .95, all p’s < .0001). 2. Results Preliminary analysis examined if walking experience varied as a function of age by conducting a Univariate Analysis of Variance (ANOVA) with the between-subjects factor of Age (14-month-olds versus 24-month-olds). The test yielded a significant effect, F(1, 30) = 139.95, MSE = 4.32, p < .0001, 2p = .82. In particular, 24-month old children (M = 11.02, SE = 0.52) have significantly more walking experience than 14-month old children (M = 2.33, SE = 0.52). Thus, subsequent analyses used Age as reflective of walking experience. The primary analysis for this study involved examining the arm position scores and the various walking skill measures as a function of the different object carriage conditions and children’s age. Accordingly, the arm position codes were analyzed in a two-way ANOVA, with the within-subjects factor of Condition (no object, one object-one hand, two objects-two hands, and one object-two hands), and the between-subjects factor of Age (14-month-olds versus 24-month-olds), employing the mean coding score, averaged across the multiple trials within each condition. Trial itself was not included as a variable in this analysis because of the occurrence of different numbers of trials for the conditions for some of the children.2 Fig. 1 presents the mean arm positions scores as a function of Age and Condition. The ANOVA on these data revealed a significant main effect for Age, F(1, 30) = 6.02, MSE = 0.35, p < .05, 2p = .17, with older children producing more mature arm movements (M = 2.55, SE = 0.08) than younger children (M = 2.07, SE = 0.08). The main effect for Condition was also significant, F(3, 90) = 26.56, MSE = 0.43, p < .001, 2p = .47. Subsequent independent-samples t tests revealed that none of the object carriage conditions differed from one another, but that all of these conditions differed from the no object condition, with children adopting a less mature arm position when carrying toys compared to when they did not (all p’s .20). Finally, there were no significant Age × Condition interactions for these variables (all F’s < 1.5, all p’s > .20). A final set of analyses examined the relation between arm position and each of the walking skill measures for the different experimental conditions, first combining across the two age groups, and then second within each of the age groups individually. For these analyses, arm positions and walking skill measures were aggregated across the multiple trials in each condition, although comparable findings were observed when the data were averaged across trials. The results of these analyses appear in Table 2. In the no object condition, for the data aggregated across age, arm postures were correlated (to a greater or lesser extent) with walking skill for all measures – with greater stride lengths and step lengths, smaller step widths, and smaller foot rotations and larger dynamic bases, associated with more mature arm postures. It is also clear, however, that this effect was primarily driven by the older children, with stride length, step length, foot rotation, and dynamic base all producing significant (or nearly significant) correlations for the older children. In contrast, for the younger children only step width was significantly correlated with arm posture. A different pattern emerges for the three experimental conditions, however. Both across age and within the individual age groups, arm postures were primarily unrelated to walking skill measures (see Table 2). Therefore, the relationship disappeared when the older children were carrying objects. Close examination of Table 2 does reveal an interesting trend towards a relation between arm posture and walking skills for the one object-one hand condition, with significant (or nearly so) correlations for stride length, step length, and dynamic base, with this pattern again more prominent (albeit insignificant) for the older age group compared to their younger counterparts. In contrast, there do not appear to be any relations for the remaining conditions. What is most intriguing about the pattern for the object carriage conditions is the obvious distinction between having one versus two hands engaged in carrying objects. This finding is speculative and requires additional empirical support before much could be made of this result.

3. Discussion The goal of this study was to examine the impact of object carriage on both arm posture and walking skill, as well as the relation between walking skill and arm position. Overall, this study found that object carriage influenced arm position, with children adopting less mature arm postures when carrying objects, as opposed to a non-carrying control condition.

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

81

Fig. 1. Mean (and standard error) arm guard positions as a function of the object carriage conditions for 14-month-old and 24-month-old infants. Note that the significance indicators refer to tests conducted on arm guard scores averaged across age group.

Table 2 Correlations between arm guard positions and walking skill, as a function of Age (aggregated across age, 14-month-olds only, 24-month-olds only), and Condition (no object, one object-one hand, two objects-two hands, one object-two hands). Condition

Walking skill measures Stride length

No object Across age 14 Months 24 Months

.31** .12 .35*

Step width −.20A −.31* .02

Step length .33** .10 .37*

Foot rotation

Dynamic base

−.18C .03 −.31*

.31** .24 .28B .24* .00 .24

One object-one hand Across age 14 Months 24 Months

.20* −.22 .24

−.11 −.16 .08

.20A −.23 .24

−.12 −.21 .09

Two objects-two hands Across age 14 Months 24 Months

.00 −.32* .16

−.12 .01 −.12

.01 −.32* .16

−.08 .04 −.11

.02 −.21 .16

One object-two hands Across age 14 Months 24 Months

−.05 −.14 −.07

.04 .11 −.03

−.02 −.09 −.06

.02 .11 −.13

−.09 −.17 −.06

* ** A B C

p < .05. p < .01. p < .06. p < .07. p < .08.

82

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

Fig. 2. Means (and standard errors) for the (A) stride lengths, (B) step lengths, (C) step widths, (D) foot rotations, and (E) dynamic bases, as a function of Age and Condition. Statistical tests shown are for the main effect of Age.

In contrast, this study failed to observe any impact of object carriage on individual walking skill measures. Finally and converging with previous work by Ledebt (2000), this study found that arm posture and walking skill were related, with more mature arm postures co-occurring in development with more mature walking skill patterns. However, this last finding is only true when children were not carrying anything. Of course, the result that object carriage influenced arm coordination also converges with previous work (Burnett & Johnson, 1971) in which children were observed to return to less mature arm positions to compensate for carrying objects. What is interesting in the current study is that the objects themselves were not especially difficult to carry in terms of weight, ease of grasping, and so on – a situation that reflects children’s natural selection for object carriage (Karasik et al., 2011). Thus, the changes that occurred were not due to the load per se (this issue will be further discussed subsequently), but rather likely due to the impact of carrying on inter-limb coordination. Moreover, it is also clear that the impact of object carrying on arm guard was not due to children simply being less able to engage in reciprocal arm swing (the most mature arm guard position) during carriage. First, although the one object-two hand

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

83

condition does make it impossible to produce reciprocal arm swing because both arms/hands must be in a fixed, forward position, the decrease in arm position sophistication was also observed in the remaining two object carriage conditions. Because the arms do not need to be in a fixed position, both conditions (one object-one hand & two objects-two hands) neither constrain arm position nor arm movement during carriage. Second, the decrease in arm guard scores when carrying objects were observed for both older and younger infants. Although this drop for the older children did, in some instances, represent a shift from reciprocal arm swing to earlier arm postures, the younger children also showed a decrease in their arm guard score. As a group, however, these children rarely demonstrated reciprocal arm swing, even in the no object condition. Thus, this effect was not due to the objects simply constraining (and potentially eliminating) the possible arm positions that could have occurred. One could point out that because children were asked to walk towards a parent while carrying the object, the children could have intended to give the object to the parent, with this intention then driving an arm position consistent with such a behaviour – essentially, a middle guard position. Thus, it might have been this intention that caused the change in arm guard, and not the act of carrying per se. Although an intriguing alternative explanation of these results, there are reasons to believe that such an explanation, even if partly true, is clearly not the entire story. Most fundamentally, the current data do not support this hypothesis in that, for both 14- and 24-month old children, the frequency of adopting the middle guard arm position occurred equally as often during carriage versus non-carriage conditions. And so, there is no real compelling evidence to suggest that this position was adopted preferentially to facilitate giving the toy to their parent or guardian. Regardless, however, the general question of the impact on carrying and walking of the functional intentions (and subsequent consequences) of the experimental task is intriguing, and one worthy of consideration. Future studies could directly compare children’s arm guard positions under conditions in which children are asked to bring an object to their parents, and conditions in which there is not actual giving task (e.g., asking to move an object from one end of the hallway to another to build a pile) to determine the impact of goal intentionality on arm position and gait. In contrast to the findings on arm posture, object carriage failed to influence the kinematics of walking at either age. This result is interesting from a few vantage points. On the one hand, this finding conflicts somewhat with other work in which carrying objects (in the form of loads) did influence walking skill (Garciaguirre et al., 2007) in terms of changes in gait parameters, overt gait disruptions (e.g., trips, double-steps, falls), and postural adjustments (Kistner et al., 2012). It is important to remember though, that the objects carried in this earlier work consisted of loads that were 15% of body weight, and were placed either symmetrically or asymmetrically on children. Still, if one characterizes the task of carrying an object in one’s hand or hands as a type of load manipulation, then one would expect to see some form of disruption of gait in this study. On the other hand, the lack of an impact of object carriage on walking skill is understandable in a few ways. As just highlighted, one critical difference between the current study and earlier work is that the objects in this study were quite light; on this level, then, they did not represent significant loads to be carried by children. In this study, the average body weight was 23.75 lbs for the 14-month-olds, and 29.06 lbs for the 24-month-olds; accordingly, the toy represented about 4.2% and 3.4% of body weight for the two age groups, respectively. Although arm weight was not explicitly measured in this study, it is typically considered to be about 5% of body weight (Clauser, McConville, & Young, 1969; Dempster, 1955; Drillis, Contini, & Bluestein, 1964). Therefore, the toys were approximately equal to, or slightly less than, the arm weight of the children. Given that previous work (Cotallorda et al., 2003; Garciaguirre et al., 2007; Hong & Brueggemann, 2000) has found that it is only with substantial loads that posture and walking patterns change, it is thus not surprising that the current loads did not influence gait. It would be interesting to follow up this study by having toddlers carry heavier objects in their hands, ones representing a more significant proportion of their body weight. Such manipulations would not only provide a more comparable comparison to earlier work, but also facilitate further exploration of the impact of symmetrical versus asymmetrical object carriage conditions, a manipulation that has been found to be important in earlier research (Adolph & Avolio, 2000; Charteris, Scott, & Nottrodt, 1989; Lloyd & Cookie, 2000; Vereijken et al., 2009; but see Garciaguirre et al., 2007, for conflicting results). This is especially interesting given the trend found in the current study suggesting a difference in one versus two object carriage conditions (i.e., a relation found between arm position and walking skills for only the one object-one hand condition in contrast to the other carriage conditions). It is also possible that any negative impact on gait kinematics created by carrying objects in this study was actually counteracted by increased vigilance to walking in the carriage conditions. Such a result would be very much in keeping with recent work by Karasik et al. (2011), who found that children fell less when they were carrying objects, as opposed to non-carrying situations. According to these authors, carrying objects helped children to better able to make appropriate adjustments to postural factors during walking. In the current study, such increased attention would have ultimately produced better walking, which would have then counteracted any negative influences on gait parameters associated with object carriage per se. The idea that carrying objects may have led to increased vigilance to walking in this experiment leads to the interesting question of what would happen to walking skill if one were to carry objects in situations requiring even more heightened attention to one’s locomotion. One possible arena for examining this question involves looking at visually-guided locomotion contexts (see Schmuckler, 2013a, for a review), such as when one is moving around obstacles (Schmuckler, 1990, 1993; Schmuckler & Gibson, 1989) or walking over barriers (Kingsnorth & Schmuckler, 2000; Schmuckler, 1996, 2013b). For example, previous work (Schmuckler, 1990, 1993) observed that having to control one’s locomotion by walking down a

84

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

narrow path, or guiding one’s self around obstacles, influenced spatial parameters of gait (e.g., step lengths) but not temporal parameters of gait (e.g., walking speed). It would then be interesting to extend these investigations by examining the impact of object carriage on gait kinematics across a range of visual-guidance tasks. The confluence of these constraints (object carriage while avoiding obstacles and/or stepping over barriers) could have a range of consequences, including impacting gait parameters and the ability to cross barriers (i.e., altering thresholds for successful crossing). Finally, this study found that arm posture and walking skill were related, but only for older children when they were not carrying objects. In contrast, the younger children failed to show any relation between arm posture and walking skill measures. The result is convergent with work by Ledebt (2000), who found that step width was related to arm postures for experienced walkers. The current study extends these earlier results by demonstrating a relation between a host of additional gait kinematics, such as stride length, step length, foot rotation, and dynamic base, and arm posture. The finding that the relation between walking skill and arm postures for older children disappeared when carrying an object is arguably the most surprising result of this project. The difference as a function of condition implies that, although not evident in overt changes in gait per se, carrying objects nevertheless did influence older children’s walking on a less obvious level. One possibility in this regard is that integrating the demands of typical walking (i.e., coordinating arm and leg motions) is still a relatively difficult task, even for older children. Accordingly, any external disturbance of this integration, such as would be produced by having to carry an object, requires a simplification of this integration process, accomplished by employing a less demanding, less motorically-mature locomotor style. Put differently, having to carry an object increased the degrees of freedom required for older children in this task, and as such, they needed to reduce these degrees of freedom. In this case, this reduction occurred by “freezing” their arm postures, reducing the amount of movement that occurred, and so, adopting less mature arm guard positions. This explanation also converges with the observed age differences in this result. Comparison of the various levels of arm postures (see Table 1) with the average arm posture scores (see Fig. 1) reveals that it was only the older children in the no object condition whose arm movements began to exhibit the characteristic back and forth movement of arms while walking. In contrast, although the younger children did show more advanced arm postures in the no object condition (relative to the three carriage conditions), the arm postures in all of these condition were generally more fixed and did not require simultaneous arm and leg coordination. Thus, the younger children were already operating at a level of reduced degrees of freedom, thereby removing any potential relation between gait and arm posture sophistication. Of course, it is still surprising that carrying an object did not force the younger children to reduce the degrees of freedom of their behaviour as well, thus producing some observable impact of carrying on gait. However, it is also quite possible that the less experienced walkers simply hit floor in their performance (with their scores clustering towards the lower end of the scale). It would be informative then to utilize other independent locomotion paradigms that would allow such an examination, such as tasks in which there are distinct measures of success (e.g., barrier or gap crossing). Finally, it is also worth noting that, although not significant, there did appear to be a hint of a relation between arm posture and walking skill in the one object-one hand condition. This trend is intriguing in that, of the three carriage conditions, this is the only case in which both of the hands were not occupied. Thus, if this trend does in fact represent a true result (a finding that awaits further experimental evidence), this would support the explanation already suggested – that object carrying impacts arm inter-limb coordination and arm swing in older children. In other words, when arms are restrained in one way or another, children are less able to simultaneously control their arm and leg movements when walking. But because one arm is unoccupied in the one object-one hand condition, children may have been able to somewhat make the appropriate body coordination. In conclusion, the present study examined the impact of object carriage on two different components of walking – arm guard position and walking skill parameters. In general, these findings reveal that with more walking experience, children learn to make the appropriate adjustments to deal with the various difficulties and conflicts that arise when moving around their world. Such work is of general interest in delineating how children solve one of the most complex, and ultimately critical, problems they face as developing organisms – how to navigate a clutter world successfully and move objects to different locations, all the while maintaining their posture and equilibrium. Investigations such as these, then, both delineate crucial development achievements and hopefully shed light on fundamental developmental processes. Of course, they do so only gradually and incrementally, much like independent walking – one step at a time. Author notes This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada to Mark A. Schmuckler. Portions of this work were presented at the biennial meetings of the Society for Research in Child Development, Montreal, PQ, March-April 2011. References Adolph, K. E. (1995). Psychophysical assessment of toddlers’ ability to cope with slopes. Journal of Experimental Psychology: Human Perception and Performance: 21., 734–750. Adolph, K. E., & Avolio, A. M. (2000). Walking infants adapt locomotion to changing body dimensions. Journal of Experimental Psychology: Human Perception and Performance: 26., 1148–1166.

D.M.J. Mangalindan et al. / Infant Behavior & Development 37 (2014) 76–85

85

Adolph, K. E., Eppler, M. A., & Gibson, E. (1993). Crawling versus walking infants’ perception of affordances for locomotion over sloping surfaces. Child Development: 64., 1158–1174. Baker, B. O., Hardyck, C. D., & Petrinovich, L. F. (1966). Weak measurements vs strong statistics: An empirical critique of S. S. Steven’s proscriptions on statististics. Educational and Psychological Measurement: 26., 291–309. Brenière, Y., Bril, B., & Fontaine, R. (1989). Analysis of the transition from upright stance to steady state of locomotion in children with under 200 days of autonomous walking. Journal of Motor Behavior: 21., 20–37. Bril, B., & Brenière, Y. (1992). Postural requirements and progression velocity in young walkers. Journal of Motor Behavior: 24., 105–116. Burnett, C. N., & Johnson, E. W. (1971). Development of gait in childhood. Part II. Developmental Medicine and Child Neurology: 13., 207–215. Charteris, J., Scott, P. A., & Nottrodt, J. W. (1989). Metabolic and kinematic responses of African women headload carriers under controlled conditions of load and speed. Ergonomics: 32., 1539–1550. Clauser, C. E., McConville, J. T., & Young, J. W. (1969). Weight, volume, and center of mass of segments of the human body. Ohio: Wright-Patterson Air Force Base. Corbetta, D., & Bojczyk, K. E. (2002). Infants return to two-handed reaching when they are learning to walk. Journal of Motor Behavior: 34., 83–95. Cotallorda, J. A., Rahmani, A. B., Diop, M. A., Gautheron, V. A., Ebermeyer, E. A., & Belli, A. A. (2003). Influence of school bag on gait kinetics. Journal of Orthopedics B: 12., 1119–1129. Dempster, W. T. (1955). Space requirements of the seated operator: Geometrical, kinematic, and mechanical aspects of the body with special reference to the limbs. Ohio: Wright-Patterson Air Force Base. Drillis, R., Contini, R., & Bluestein, M. (1964). Body segment parameters: A survey of measurement techniques. Artificial Limbs: 25., 44–66. Elftman, H. (1939). The function of the arms in walking. Human Biology: 11., 529–535. Garciaguirre, J. S., Adolph, K. E., & Shrout, P. E. (2007). Baby carriage: Infants walking with loads. Child Development: 78., 664–680. Gesell, A. (1939). Reciprocal interweaving in neuromotor development. Journal of Comparative Neurology: 70., 161–180. Hong, Y., & Brueggemann, G. (2000). Changes in gait patterns in 10-year-old boys with increasing loads when walking on a treadmill. Gait and Posture: 11., 254–259. Karasik, L. B., Adolph, K. E., Tamis-LeMonda, C. S., & Zuckerman, A. L. (2011). Carry on: Spontaneous object carrying in 13-month-old crawling and walking infants. Developmental Psychology: 82., 1199–1209. Kingsnorth, S., & Schmuckler, M. A. (2000). Walking skill versus walking experience as a predictor of barrier crossing in toddlers. Infant Behavior and Development: 23., 331–350. Kistner, F., Fiebert, I., & Roach, K. (2012). Effect of backpack load carriage on cervical posture in primary schoolchildren. Work: 41., 99–108. Kubo, M., & Ulrich, B. (2006). A biomechanical analysis of ‘high guard’ position of arms during walking in toddlers. Infant Behavior and Development: 29., 509–517. Ledebt, A. (2000). Changes in arm posture during the early acquisition of walking. Infant Behavior and Development: 23., 79–89. Ledebt, A., & Bril, B. (2000). Acquisition of upper body stability during walking in toddlers. Developmental Psychobiology: 36., 311–324. Lloyd, R., & Cookie, C. B. (2000). The oxygen consumption associated with unloaded walking and load carriage using two different backpack designs. European Journal of Applied Physiology: 81., 492–496. Lord, F. (1953). On the statistical treatment of football numbers. American Psychologist: 8., 750–751. Murray, M. P., Kory, R. C., & Bernard, E. J. (1967). Patterns of sagittal rotation of the upper limbs in walking: A study of normal men during free and fast speed walking. Physical Therapy: 47., 272–284. Neter, J., Wasserman, W., & Kutner, M. H. (1990). Applied linear statistical models. Boston, MA: Irwin. Palmer, C. E. (1994). Studies of the center of gravity in the human body. Child Development: 15., 99–163. Schmuckler, M. A. (1990). Issues in the development of postural control. In H. Block, & B. I. Bertenthal (Eds.), Sensory-motor organizations and development in infancy and early childhood (pp. 231–236). Dordrecht: Kluwer Academic Publishers. Schmuckler, M. A. (1993). Perception-action coupling in infancy. In G. J. P. Savelsbergh (Ed.), The development of coordination in infancy (pp. 137–173). North-Holland: Elsevier. Schmuckler, M. A. (1996). Development of visually-guided locomotion: Barrier crossing by toddlers. Ecological Psychology: 8., 209–236. Schmuckler, M. A. (2013a). Perceptual-motor relations in obvious and non-obvious domains: A history and review. In P. D. Zelazo (Ed.), Handbook of developmental psychology (pp. 237–270). Oxford, UK: Oxford University Press. Schmuckler, M. A. (2013b). Perseveration in barrier crossing. Journal of Experimental Psychology: Human Perception and Performance,. Advance online publication. doi: 10.1037/a0031119. Schmuckler, M. A., & Gibson, E. J. (1989). The effect of imposed optical flow on guided locomotion in young walkers. British Journal of Developmental Psychology: 7., 193–206. Stiger, T. R., Kosinski, A. S., Barnhart, H. X., & Kleinbaum, D. G. (1998). Anova for repeated ordinal data with small sample size? A comparison of anova, manova, wls and gee methods by simulation. Communications in Statistics - Simulation and Computation: 27., 357–375. Sutherland, D. H., Olshen, R., Cooper, L., & Woo, S. L. (1980). The development of mature gait. Journal of Bone and Joint Surgery - American Volume: 62., 336–353. Velleman, P. F., & Wilkinson, L. (1993). Nominal, ordinal, interval, and ratio typologies are misleading. The American Statistician: 47., 65–72. Vereijken, B., Pederson, A. V., & Størksen, J. H. (2009). Early independent walking: A longitudinal study of load perturbation effects. Developmental Psychobiology: 51., 374–383. Zwart, R., Ledebt, A., Fong, B. F., de Vries, H., & Savelsbergh, G. J. P. (2005). The affordance of gap crossing in toddlers. Infant Behavior and Development: 28., 145–154.