Paper 3 Marcha Niños

8/18/2019 Paper 3 Marcha Niños

1/13

doi:10.1152/jn.00262.200697:2790-2801, 2007. First published 24 January 2007; J NeurophysiolNadia Dominici, Yuri P. Ivanenko and Francesco LacquanitiAdaptation to Load ChangesControl of Foot Trajectory in Walking Toddlers:

You might find this additional info useful...

64 articles, 25 of which can be accessed free at:This article cites

http://jn.physiology.org/content/97/4/2790.full.html#ref-list-1

6 other HighWire hosted articles, the first 5 are:This article has been cited by

[PDF][Full Text][Abstract], April 1, 2008; 99 (4): 1890-1898. J Neuroph ysiol

Y. P. Ivanenko, A. d'Avella, R. E. Poppele and F. LacquanitiOn the Origin of Planar Covariation of Elevation Angles During Human Locomotion

[PDF][Full Text][Abstract]

2009; 101 (3): 1419-1429. J Neuroph ysiolFrancesco LacquanitiNadia Dominici, Elena Daprati, Daniele Nico, Germana Cappellini, Yuri P. Ivanenko andElongation: Evidence for a Locomotor Body Schema?Changes in the Limb Kinematics and Walking-Distance Estimation After Shank

[PDF][Full Text][Abstract], February , 2010; 103 (2): 746-760. J Neuroph ysiol

LacquanitiGermana Cappellini, Yuri P. Ivanenko, Nadia Dominici, Richard E. Poppele and FrancescoMotor Patterns During Walking on a Slippery Walkway

[PDF][Full Text][Abstract]

, March , 2010; 103 (3): 1673-1684. J Neuroph ysiolFrancesco LacquanitiNadia Dominici, Yuri P. Ivanenko, Germana Cappellini, Maria Luisa Zampagni andSurfacesKinematic Strategies in Newly Walking Toddlers Stepping Over Different Support

[PDF][Full Text][Abstract], August 11, 2010; 30 (32): 10720-10726. J. Neurosci.

Matthias D. Ziegler, Hui Zhong, Roland R. Roy and V. Reggie EdgertonWhy Variability Facilitates Spinal Learning

including high resolution figures, can be found at:Updated information and services

http://jn.physiology.org/content/97/4/2790.full.html

can be found at: Journal of NeurophysiologyaboutAdditional material and information

http://www.the-aps.org/publications/jn

This infomation is current as of August 16, 2011.

American Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our website at http://www.the-aps.org/.(monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2007 by the

publishes original articles on the function of the nervous system. It is published 12 times a year Journal of Neurophysiology

http://jn.physiology.org/content/97/4/2790.full.html#ref-list-1http://jn.physiology.org/content/99/4/1890.full.pdfhttp://jn.physiology.org/content/99/4/1890.full.htmlhttp://jn.physiology.org/content/99/4/1890.abstract.htmlhttp://jn.physiology.org/content/99/4/1890.full.htmlhttp://jn.physiology.org/content/99/4/1890.full.pdfhttp://jn.physiology.org/content/99/4/1890.abstract.htmlhttp://jn.physiology.org/content/99/4/1890.full.htmlhttp://jn.physiology.org/content/101/3/1419.full.pdfhttp://jn.physiology.org/content/101/3/1419.full.htmlhttp://jn.physiology.org/content/101/3/1419.abstract.htmlhttp://jn.physiology.org/content/101/3/1419.full.htmlhttp://jn.physiology.org/content/101/3/1419.full.pdfhttp://jn.physiology.org/content/101/3/1419.abstract.htmlhttp://jn.physiology.org/content/101/3/1419.full.htmlhttp://jn.physiology.org/content/103/2/746.full.pdfhttp://jn.physiology.org/content/103/2/746.full.htmlhttp://jn.physiology.org/content/103/2/746.abstract.htmlhttp://jn.physiology.org/content/103/2/746.full.htmlhttp://jn.physiology.org/content/103/2/746.full.pdfhttp://jn.physiology.org/content/103/2/746.abstract.htmlhttp://jn.physiology.org/content/103/2/746.full.htmlhttp://jn.physiology.org/content/103/3/1673.full.pdfhttp://jn.physiology.org/content/103/3/1673.full.htmlhttp://jn.physiology.org/content/103/3/1673.abstract.htmlhttp://jn.physiology.org/content/103/3/1673.full.htmlhttp://jn.physiology.org/content/103/3/1673.full.pdfhttp://jn.physiology.org/content/103/3/1673.abstract.htmlhttp://jn.physiology.org/content/103/3/1673.full.htmlhttp://www.jneurosci.org/content/30/32/10720.full.pdfhttp://www.jneurosci.org/content/30/32/10720.full.htmlhttp://www.jneurosci.org/content/30/32/10720.abstract.htmlhttp://www.jneurosci.org/content/30/32/10720.full.htmlhttp://www.jneurosci.org/content/30/32/10720.full.pdfhttp://www.jneurosci.org/content/30/32/10720.abstract.htmlhttp://www.jneurosci.org/content/30/32/10720.full.htmlhttp://jn.physiology.org/content/97/4/2790.full.htmlhttp://jn.physiology.org/content/97/4/2790.full.htmlhttp://jn.physiology.org/content/99/4/1890.full.pdfhttp://jn.physiology.org/content/99/4/1890.full.htmlhttp://jn.physiology.org/content/99/4/1890.abstract.htmlhttp://jn.physiology.org/content/101/3/1419.full.pdfhttp://jn.physiology.org/content/101/3/1419.full.htmlhttp://jn.physiology.org/content/101/3/1419.abstract.htmlhttp://jn.physiology.org/content/103/2/746.full.pdfhttp://jn.physiology.org/content/103/2/746.full.htmlhttp://jn.physiology.org/content/103/2/746.abstract.htmlhttp://jn.physiology.org/content/103/3/1673.full.pdfhttp://jn.physiology.org/content/103/3/1673.full.htmlhttp://jn.physiology.org/content/103/3/1673.abstract.htmlhttp://www.jneurosci.org/content/30/32/10720.full.pdfhttp://www.jneurosci.org/content/30/32/10720.full.htmlhttp://www.jneurosci.org/content/30/32/10720.abstract.htmlhttp://jn.physiology.org/content/97/4/2790.full.html#ref-list-1


2/13

Control of Foot Trajectory in Walking Toddlers: Adaptation to Load Changes

Nadia Dominici,1,2 Yuri P. Ivanenko,1 and Francesco Lacquaniti1,2,3

1 Department of Neuromotor Physiology, Santa Lucia Foundation; and 2 Department of Neuroscience and 3Centre of Space Bio-medicine, University of Rome Tor Vergata, Rome, Italy

Submitted 10 March 2006; accepted in final form 20 January 2007

Dominici N, Ivanenko YP, Lacquaniti F. Control of foot trajec-tory in walking toddlers: adaptation to load changes. J Neuro- physi ol 97: 2790–2801, 2007. First published January 24, 2007;doi:10.1152/jn.00262.2006. On earth, body weight is an inherentconstraint, and accordingly, load-regulating mechanisms play animportant role in terrestrial locomotion. How do toddlers deal withthe effects of their full body weight when faced with the task of independent upright locomotion for the first time? Here we studiedthe effect of load variation on walking in 12 toddlers during theirfirst unsupported steps, 15 older children (1.3–5 yr), and 10 adults.To simulate various levels of body weight, an experimenter held

the trunk of the subject with both hands and supplied an approx-imately constant vertical force during stepping on a force platform.During unsupported stepping, the shape of the foot path in toddlers(typically single-peak toe trajectory) was different from that of adults and older children (double-peak trajectory). In contrast toadults and older children, who showed only limited changes inkinematic coordination, the “reduced-gravity” condition consider-ably affected the shape of the foot path in toddlers: they tended tomake a high lift and forward foot overshoot at the end of swing. Inaddition, stepping at high levels of body unloading was character-ized by a significant change in the initial direction of foot motionduring early swing. Intermediate walkers (1.5–5 mo after walkingonset) showed only partial improvement in foot trajectory charac-teristics. The results suggest that, at the onset of walking, changesin vertical body loads are not compensated accurately by the

kinematic controllers; compensation necessitates a few months of independent walking experience.

I N T R O D U C T I O N

What is the role of body weight for the control of locomo-tion? Foot–support interactions and load-regulating mecha-nisms play a crucial role in shaping patterned motor outputduring stepping (Duysens et al. 2000). Transient loading of thelimb enhances the activity in antigravity muscles during stanceand delays the onset of the next flexion (Duysens and Pearson1980). Furthermore, the safe trajectory of the foot with mini-mal toe clearance at midswing is a precise endpoint control

task that is under multisegmental motor control in both thestance and swing limbs (Bernstein 1967; Winter 1992). Inadults, minimal contact forces are sufficient for accurate foottrajectory control because the required variation of the param-eters that describe shape and variability of the foot path is verylimited despite drastic changes in limb kinetics with bodyunloading (Ivanenko et al. 2002). These findings argue againsta simple controller that commands a fixed change of limbkinematics for a fixed change of force. Thus the basic kine-matic control mechanisms operating in normal walking (Lac-

quaniti et al. 1999, 2002) also apply to walking with a highlevel of body weight support (BWS) in adults, suggesting thatthe kinematic locomotor program compensates well for bodyweight.

Previous studies have shown that infants have the ability toadapt to environmental changes. Infants, when supported, canstep with a variety of speeds, in different directions, and withappropriate response to external perturbations long before thetime of the onset of independent walking (Forssberg 1985;Lam et al. 2003a; Lamb and Yang 2000; Pang and Yang 2001;

Thelen 1986; Thelen and Cooke 1987; Thelen et al. 1987;Yang et al. 1998, 2005; Zelazo 1983), suggesting that thepattern generation network and ability to adapt to environmen-tal changes are present in the human locomotor system beforeindependent walking. Nevertheless, initial stepping in toddlershas also some specific features. For instance, global gaitfeatures, such as the planar law of intersegmental coordination(Cheron et al. 2001a,b) and the pendulum mechanism of walking (Holt et al. 2006; Ivanenko et al. 2004), are stillimmature at the beginning of independent walking. A toddler’sability to walk on uneven terrain or slopes is very limited(Adolph 1997) and an appropriate control of foot motion is alsolacking: infants show a larger foot clearance, an unusual singlepeak trajectory (Ivanenko et al. 2005), and flat-footed contactwith the ground (Forssberg 1985).

Young infants (2 mo) can support only 20–40% of theirbody weight on one limb (Forssberg 1985). In addition, slim-mer babies with smaller bones and smaller muscle mass arebelieved to begin crawling and walking sooner than chubbier,more top-heavy infants (Adolph 1997; Garn 1966; Jensen2005; Payne and Isaacs 2005; Shirley 1931; Thelen et al.1984). In toddlers, given limited muscle strength and limitedexperience with their body weight and leg movements in anupright posture against gravity, body heaviness may representa challenging factor for stepping. To obtain further insightsinto the foot–support interactions and the intrinsic dynamics of the neuromotor system at the walking onset, we examined how

toddlers adapt to a change in the amount of load on their legs.Trunk support represents a common experimental strategy toprovide equilibrium and thus facilitate locomotor movementsin infants. Moreover, one might expect that infants are welladapted to trunk support and body unloading before the onsetof independent walking, but unsupported stepping represents a“new” condition for them. However, other than the occasional“stick-diagram” or individual trajectory plots at uncontrolledBWS levels, a comprehensive study of limb kinematics intoddlers walking at different gravitational loads has not yet

Address for reprint requests and other correspondence: Y. P. Ivanenko,Dept. of Neuromotor Physiology, IRCCS Fondazione Santa Lucia, viaArdeatina 306, 00179 Rome, Italy (E-mail: [email protected]).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement ”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

J Neurophysiol 97: 2790–2801, 2007.First published January 24, 2007; doi:10.1152/jn.00262.2006.

2790 0022-3077/07 $8.00 Copyright © 2007 The American Physiological Society www.jn.org


3/13

appeared in the literature. Few previous studies have dealt withadaptation to limb loading (Schmuckler 1993; Thelen et al.1984) and unexpected unloading (e.g., removal of the extraweight strapped around the lower limb, Lam et al. 2003b) to

the additional body mass (Adolph and Avolio 2000), or withspontaneous leg movements in different postures (supine, an-gled and vertical), i.e., in different gravitational contexts ininfants (Jensen et al. 1994). Such studies have reported mainlylocal or transient modifications in muscle activity, joint rota-tions, or step length characteristics.

In our previous work (Ivanenko et al. 2005), we used trunk support as a means toward reducing the effects of posturalinstability in walking toddlers and found no significant changesin lower limb kinematics. In that study, we reported only theeffect of postural instability, and to this end, we suppliedlimited vertical force (typically 20% of the body weight).However, trunk stabilization (postural stability) and trunk un-

loading (weight bearing) may have separate consequences onstepping. Here, we report stepping with and without trunk support in toddlers who were just beginning to walk indepen-dently, and the amount of body weight support was varied fromtrial to trial to cover a wide range of values (0–90% of fullbody weight). Gait improvements during the first 1–3 mo showa very high rate of change, which is followed by a less steepcurve of maturation (Adolph et al. 2003; Hallemans et al. 2005,2006; Holt et al. 2006; Ivanenko et al. 2007; Sutherland et al.1988). We hypothesized that appropriate adaptation to bodyweight changes may also occur within the first several monthsafter the onset of independent walking. To this end, we alsocompared toddlers’ gaits with walking in older children and inadults at different levels of BWS. To analyze kinematic pat-

terns, we used methods developed earlier for adult gait (Bian-chi et al. 1998; Borghese et al. 1996; Ivanenko et al. 2002).

M E T H O D S

Subjects

We recorded surface locomotion in 12 toddlers (6 males, 6 females,11–15 mo of age), 6 intermediate walkers (4 males, 2 females, 1.5–5mo after the walking onset), 9 children 2–5 yr old (7 males, 2 females,9–50 mo after the walking onset), and 10 healthy adults [5 femalesand 5 males, 31 7 (SD) yr; Table 1]). We used the same develop-mental groups (Table 1) as in our previous paper (Ivanenko et al.2004). Five of our toddlers were recorded a second time 0.1–2 yr afterthe first recording session (3 of our toddlers were the same as in our

previous paper about the effect of posture stabilization, Ivanenko et al.2005; but now we report their locomotor response to a full range of BWS levels). Informed consent was obtained from all the adults andfrom the parents of the children. The experimental procedures were

approved by the ethics committee of the Santa Lucia Institute andconformed with the Declaration of Helsinki. The laboratory settingand experimental procedures were adapted to the children such thatthe minimal risks were equal or lower to that of walking at home. Botha parent and an experimenter remained astride the younger children toprevent them from hurting themselves during falls. For the toddlers,daily recording sessions were programmed around the parents’ ex-pectation of the very first day of independent walking, until unsup-ported locomotion was recorded.

Walking conditions

WALKING WITHOUT SUPPORT. For the recording of the very firststeps, one parent initially held the toddler by hand. Then, the parentstarted to move forward, leaving the toddler’s hand and encouraging

her or him to walk unsupported on the floor. For each subject, about10 trials were recorded under similar conditions. Short trials (3 min,depending on endurance and tolerance) were recorded with rest breaksin between. Only sequences of steps executed naturally by the toddler(e.g., no stop between steps) and while looking forward were retainedto avoid initiation and braking phases and head movements caused bylooking in other directions. Older children and adults were monitoredwhile they walked at natural, freely chosen speeds. Typically, weanalyzed two to five consecutive step cycles in each trial, for toddlers,older children, and adults.

WALKING AT DIFFERENT LEVELS OF BODY WEIGHT SUPPORT. In thetoddlers we studied the effect of changing the load on the lower limbkinematics. To this end, after recording gait under standard condi-tions, some trials were recorded while an experimenter (or a parent)

held the trunk of the child under his/her arms with both hands andsupplied an approximately constant vertical force during steppingattempts (Fig. 1, A and B). The amount of the external BWS wasvaried from trial to trial to cover a wide range of values. The sameprocedure of body unloading was performed in older children duringoverground walking.

In adults, the experiments with different BWS levels (0, 35, 50, 75,and 95% of body weight) were carried out on a treadmill at 3 km/h.BWS was obtained by suspending the subjects in a parachute harness(Reha, BONMED) connected to a pneumatic device that applied acontrolled upward force at the waist (WARD system, Gazzani et al.2000). As a result, each supporting limb experienced a simulatedreduction of gravity proportional to the applied force, while theswinging limb experienced 1g. The overall constant error in theapplied force and dynamic force fluctuations monitored by a load cell

TABLE 1. Developmental group characteristics

N Age, Months Limb Length, cm Mass. kgNumber of Recorded

Unsupported StepsNumber of Steps

With BWS

Toddlers (within 1 wk of unaided walkingexperience) 12 12.9 1.2 32.1 4.1 10.4 1.0 32 19 34 18

Intermediate walkers(1.5–5 mo afteronset of walking) 6 17.4 1.1 37.2 2.5 11.7 0.6 16 11 11 3

Children 2–5 yr(between 9 and 50mo walking experience) 9 46.7 16.9 43.7 7.6 15.6 3.6 29 14 21 10

Adults 10 (5)* 31 7 (yrs) (34 6) 83.7 3.5 (81.1 2.2) 64.6 15.0 (53.5 6.1) 26 6 (12 3) 76 17 (11 5)

Values are means SD. The length of the lower limb was measured as thigh plus shank length. *Overground walking with manual body weight support.

2791FOOT TRAJECTORY IN TODDLERS

J Neurophysiol • VOL 97 • APRIL 2007 • www.jn.org


4/13

was 5% of the body weight. All adults were also recorded duringoveground walking at a natural speed. In addition, five smaller-weight(43–58 kg) adults were recorded using the same manual unloadingprotocol during overground walking as in children.

Data recording

Bilateral kinematics of locomotion was recorded at a digitizing rateof 100 Hz by means of the VICON-612 motion analysis system(Oxford, UK). The positions of selected points on the body wererecorded by attaching passive infrared reflective markers (diameter,

1.4 cm) to the skin overlying the following bony landmarks on bothsides of the body: gleno-humeral joint (GH), the tubercle of theanterosuperior iliac crest (IL), greater trochanter (GT), lateral femurepicondyle (LE), lateral malleolus (LM), and fifth metatarso-phalan-geal joint (VM). Particular care was taken to place the VM marker inthe same position (on the lateral aspect of the 5th metatarso-phalan-geal joint) in all subjects. Individual deviations of the VM markerposition relative to the actual fifth metatarso-phalangeal joint (becausethe diameter of the marker was 1.4 cm and it was attached to the skinon a short 1-cm height stem) were measured before the recordings andcorrected afterward with software.

The ground reaction forces under both feet (Fig. 1 B) were recordedat 1,000 Hz by a force platform (0.9 0.6 m; Kistler 9287B, Zurich,Switzerland). Toddlers generally performed two to three steps on theforce platform in each trial.

Data analysis

Deviations of gait trajectory relative to the x -direction of therecording system were corrected by rotating the xz axes by the angleof drift computed between start and end of the trajectory. The bodywas modeled as an interconnected chain of rigid segments: GH-IL forthe trunk, IL-GT for the pelvis, GT-LE for the thigh, LE-LM for the

shank, and LM-VM for the foot. The main limb axis was defined asGT-LM. The elevation angle of each segment corresponds to the anglebetween the segment projected on the sagittal plane and the vertical(positive in the forward direction, i.e., when the distal marker fallsanterior to the proximal one).

Walking speed was measured by computing the mean velocity of the horizontal IL marker movement. The length of the lower limb (L)was measured as thigh (GT-LE) plus shank (LE-LM) length.

In adults, gait cycle duration was defined as the time intervalbetween two successive maxima of the elevation angle of the mainlimb axis of the same limb and stance phase as the time intervalbetween the maximum and minimum values of the same angle(Borghese et al. 1996). Thus a gait cycle (stride) referred to a cyclicmovement of one leg and equaled two steps. When children stepped

on the force platforms, these kinematic criteria were verified bycomparison with foot strike and lift-off measured from the changes of the vertical force around a fixed threshold. In general, the differencebetween the time events measured from kinematics and kinetics was3%. However, the kinematic criterion sometimes produced a sig-nificant error in the identification of stance onset in toddlers if therewas an unusual forward foot overshoot at the end of swing (cf.Forssberg 1985). In such cases, foot contact was determined using arelative amplitude criterion for the vertical displacement of the VMmarker (when it was elevated to 7% of the limb length from the floor).

The amount of the external body weight support in children andadults during overground walking was estimated as the percentreduction of the mean vertical force on the platform (Fig. 1 B).

To verify whether the trunk angle in the sagittal plane varied withweight support, we measured the mean angle of the long axis of the

trunk with respect to the vertical axis. The long axis of the trunk wasdefined by connecting the midpoint of the two (left and right) ILmarkers with the midpoint of the two GH markers.

Intersegmental coordination

Intersegmental coordination was evaluated in position space aspreviously described (Bianchi et al. 1998; Borghese et al. 1996). Inadults, the temporal changes of the elevation angles at the thigh,shank, and foot covary during walking. When these angles are plottedin three dimensions (3D), they describe a path that can be least-squares fitted to a plane over each gait cycle. In adults, changing BWSresults in limited changes of the kinematic coordination (Ivanenko etal. 2002). Here, we studied this issue of intersegmental coordination

(the gait loop and its associated plane) in children. To this end, wecomputed the covariance matrix of the ensemble of time-varyingelevation angles (after subtraction of their mean value) over each gaitcycle. The three eigenvectors u1–u3, rank ordered on the basis of thecorresponding eigenvalues, correspond to the orthogonal directions of maximum variance in the sample scatter. For each eigenvector, theparameters uit, uis, and uif correspond to the direction cosines with thepositive semi-axis of the thigh, shank, and foot angular coordinates,respectively. The first two eigenvectors u1 and u2 lie on the best-fittingplane of angular covariation. The third eigenvector (u3) is the normalto the plane and thus defines the plane orientation. To quantify therotation of the plane with BWS, we analyzed the u3t parameter (thedirection cosine of the normal to the plane with the axis of thighelevation), which was found to vary systematically with BWS intoddlers (see RESULTS).

FIG. 1. An example of unsupported and supported walking. A: unilateralsagittal stick diagrams (of a single step) when the toddler walked unsupported(left ) and when an experimenter held the trunk of the child with both hands andsupplied an approximately constant vertical force during stepping attempts(middle and right ). B: ground reaction forces under both feet were recordedwith a force platform. Dotted horizontal lines indicate body weight. Amount of external body weight support was estimated as percent reduction of mean

vertical force on the platform (F y).

2792 N. DOMINICI, Y. P. IVANENKO, AND F. LACQUANITI



5/13

Foot trajectory

The shape of the endpoint path was compared by computing thevertical excursion of the VM marker (during swing) and correlating itwith the corresponding ensemble average in adults (using Pearsoncorrelation coefficient). VM trajectories were time-normalized overthe swing phase duration. They were analyzed both in space and

relative to the instantaneous hip (GT marker) position. To show thedynamics of foot motion during swing, we also calculated the instan-taneous pitch angle of the VM velocity vector in the sagittal plane. Ineach gait cycle, the foot moves back and forth relative to the body. Toemphasize the differences in the initial direction of foot motion duringearly swing across subjects, we computed the mean pitch angle of VMmotion during the first 30% of swing (beginning from the momentwhen the VM marker reversed direction from backward to forward).

Foot trajectory variability

Foot-trajectory spatial variability in the sagittal plane was quanti-fied in terms of spatial density and normalized tolerance area of VM,computed over the swing phase (Ivanenko et al. 2002, 2005). Theseindices describe the integrated variability of foot path, includingvariability in both the vertical and horizontal directions. To comparesubjects of different heights, VM trajectories (relative to the instan-taneous position of GT) were first scaled by the limb length (inproportion to the mean limb length of adults) and resampled in thespace domain by means of linear interpolation of the x,y time series(1.5 mm steps) over all gait cycles. All steps (typically 7–15) fromdifferent trials under the same walking conditions were pooled to-gether for this analysis.

Spatial density was calculated as the number of points falling in1 1 cm2 cells of the spatial grid divided by the number of stepcycles. Normalized tolerance area was derived as follows. The meanlength of foot trajectory over the swing phase across all steps wascalculated from the corresponding path integral. For every interval

corresponding to 10% of the maximal horizontal excursion, wecomputed the 2D 95% tolerance ellipsis of the points within theinterval. The typical number of points in each interval ranged from500 to 1,200 (depending on the number of gait cycles). The areas of all tolerance ellipses were summed and normalized by the meanlength of foot trajectory. This normalized area provides an estimate of the mean area covered by the points along 1 cm of path. A greatertolerance area indicates greater variability.

Age-related changes

The time-course of changes of kinematic and kinetic parameters asa function of age was fitted by an exponential function, y aet / b, where y is the specific parameter under study, t is the time sinceonset of unsupported walking, is the time constant, and a and b aretwo weighting constants.

Statistics

Statistical analyses (Student’s t -test) were used when appropriate.Reported results were considered significant for P 0.05. Statisticson correlation coefficients was performed on the normally distributed,Z-transformed values. The algebra of angles (circular variables) issomewhat different from the rules governing other (linear) quantities.Circular statistics (Watson’s U 2 test) were used to compare the initialdirections of foot motion during swing across subjects. Sphericalstatistics on directional data (Batschelet 1981) were used to charac-terize the mean orientation of the normal to the covariation plane (seeabove) and its variability across steps. To assess the variabilitydirectly, we calculated the angular SD (called spherical angulardispersion) of the normal to the plane.

R E S U L T S

General characteristics of foot trajectory in children

During unsupported walking, foot trajectory characteristicsdiffered systematically in toddlers compared with those inadults and older children. The dominant template of footmotion is shown in Fig. 2 A (left ). All toddlers typically movedthe leg in such a way that foot lift (VM y) had only onemaximum at midswing. This behavior was entirely in contrastto that observed in the typical adult gait, which was character-ized by prominent foot lift in early swing, a minimum footclearance during midswing, and another separate toe lift in the

end of swing. As a result, the correlation coefficient betweenthe time series of the VM y during swing in toddlers and thecorresponding ensemble average in adults was typically nega-tive (0.51 0.19). The spatial variability of the endpoint(foot) path in the sagittal plane was considerably greater thanin the adults (see also Ivanenko et al. 2005): the normalizedtolerance area (scaled to the mean limb length of adults) was

FIG. 2. Changes of the foot trajectorycharacteristics (shape and variability) withage. A: stick diagrams (of a single step) andcorresponding foot trajectories in 1 toddler, 2

older children, and in 1 representative adult. B: correlation coefficient between VM y data inchildren and corresponding ensemble averagein adults during swing. C : normalized 5thmetatarso-phalangeal joint (VM) tolerancearea during swing. All data refer to over-ground walking at natural speed. Values foradult group were obtained by computing mean(SD) from pooled data obtained for walkingat natural, freely chosen speed (on average,3.8 0.4 km/h). Time-course of changes withage was fitted by an exponential function (r 0.92 and 0.86 in B and C , respectively). iscorresponding time constant.




6/13

20.5 6.5 cm2 /cm in toddlers and 3.8 1.9 cm2 /cm in adults(P 0.0001, Student’s unpaired t -test).

Age-related changes of the kinematic parameters related tothe foot trajectory control are presented in Fig. 2, B and C, forthe whole group of subjects during overground walking at anatural speed. Both the correlation coefficient between VM y inchildren and VM

y in adults (Fig. 2 B) and the foot path variability

(normalized VM tolerance area; Fig.2C ) changed rapidly over thefirst few months of independent walking experience. Changeswith age were fitted by an exponential function (see METHODS).The time constant was fast both for the correlation with theensemble average in adults ( 2.5 mo) and for the changes of the VM spatial variability across steps ( 1.9 mo).

Stepping at different BWS levels

To study the effect of BWS, an experimenter or a parentfirmly held the trunk of the child with both hands providing anapproximately constant vertical force while the child stepped.

Supporters were instructed to avoid influencing the toddlers’forward motion. However, we were unable to completelycontrol for the possibility that some aspects of stepping wereinfluenced by external forces generated by the supporter on thetoddler’s forward motion. Nevertheless, during supported step-ping, the mean walking speed and stride length (1.6 0.5km/h and 0.44 0.08 m at 40–90% BWS) were similar tothose during unsupported stepping (1.5 0.4 km/h and 0.35 0.06 m; P 0.05, paired t -test). The position of the arms wasalso similar because the upper extremities in newly walkingtoddlers are typically held away from the body (Sutherland etal. 1980). The mean walking speed in nine older children (2–5yr), supported in the same way as the toddlers, also did notchange significantly (3.0 0.5 km/h at 0 BWS and 3.1 0.4

km/h at 40–90% BWS).The walking speed freely chosen by toddlers was in general

lower than in adults or in older children when expressed indynamically equivalent terms—using the Froude number(Fr V 2 / gL, where V is the average speed of locomotion, g isthe acceleration of gravity, and L is the limb length). TheFroude number is a dimension-less parameter suitable for thecomparison of locomotion in subjects of different size walkingat different speeds, under the assumption of the gravity-relatedpendulum mechanism of movement (Cavagna et al. 1983). Fr was 0.07 0.04 in our toddlers (across all conditions) versus0.18 0.05 in older children and 0.19 0.04 in adults duringoverground walking. The mean walking speed in toddlers

(1.5 km/h) corresponds to 2.5 km/h in adults when nor-malized using the Froude number (i.e., to the square root of L).Therefore, although the general characteristics of foot motionin adults do not depend significantly on the walking speed(Ivanenko et al. 2002), for the sake of comparison, we recordedtreadmill locomotion with BWS in adults at a comparableequivalent speed (3 km/h).

It is worth noting that the general characteristics of the footpath in adults did not differ significantly during overgroundwalking and walking on the treadmill at 0 BWS except that thefoot path variability was slightly larger during overgroundwalking: VM tolerance area was 3.8 1.9 cm2 /cm duringoverground walking and 2.7 1.1 cm2 /cm during treadmillwalking at 3 km/h. This difference was likely caused by some

variability in the walking speed (and consequently in the stridelength) across overground trials.

Effect of BWS on intersegmental coordination

Figure 3 A plots averaged thigh, shank, and foot elevationangles and corresponding gait loops in toddlers, older children

(2–5 yr), and adults at two levels of BWS. In adults and olderchildren, temporal changes of the elevation angles of lowerlimb segments covaried along a plane, describing a character-istic loop over each stride. Paths progress in time in thecounter-clockwise direction, lift-off and touch-down corre-sponding approximately to the top and bottom of the loop,respectively (Fig. 3 A). Planarity was quantified by the percent-age of variance accounted for by the third eigenvector (PV3) of the data covariance matrix: the closer PV3 is to 0, the smallerthe deviation from planarity. Planar covariation was obeyed atall tested BWS levels (on average, PV3 was 1.0 0.2%),although the 3D gait loop became slightly narrower withincreasing BWS (e.g., at 95% BWS, in 4 of the 10 adults PV2

was

3%), consistent with our previous study; (Ivanenko et al.2002; PV1 increase and PV2 decrease; Fig. 3 B, right ). The firsttwo eigenvectors u1 and u2 of the covariance matrix lie on thebest-fitting plane of angular covariance, whereas the thirdeigenvector (u3) is the normal to the plane and is characterizedby the direction cosines plotted in Fig. 3 B (bottom panels).

The gait loop and its associated plane depend both on theamplitude and phase of the limb segment oscillations. Theplots of Fig. 3 A show the basic stereotypy of the patterns of intersegmental covariations obtained at different BWS levels,but they hide subtle yet important trends with BWS changes. Inall subjects, the plane of angular covariation rotated withincreasing BWS (U3t monotonic decrement; Fig. 3 B).

In toddlers, at 0 BWS, the gait loop departed significantly

from planarity and the mature pattern. Although the plane(PV1 PV2 PV3) resembled that of the adults, PV3 wassignificantly higher in toddlers (3.1 0.8%) than in adults(0.9 0.2%, P 0.001 Student’s unpaired t -test), in agree-ment with previous findings (Cheron et al. 2001b; Ivanenko etal. 2004). Moreover, the step-by-step variability of plane ori-entation (estimated as the angular dispersion of the planenormal) was considerably higher in toddlers (15.8 7.3°) thanin adults (2.9 1.0°). In addition, because the amplitude of thigh movements was relatively higher with respect to that of shank and foot movements in toddlers (Fig. 3 A), the gait loopwas less elongated than in adults, as shown by the largercontribution of the second eigenvector (PV2), although with

increasing BWS, the gait loop in toddlers (Fig. 3 A) tended to bemore elongated (PV2 decreased; Fig. 3 B, left ).

Effect of BWS on foot trajectory

BWS did not influence appreciably the shape of the endpointpath in adults (Fig. 4) in agreement with our previous study(Ivanenko et al. 2002). At 95% BWS (nearly complete unload-ing), adult subjects were still able to step on the treadmill,typically with only forefoot loading during stance (no heelcontact). However, the correlation coefficient between the timeseries of the VM y during swing at 0 and 95% BWS was stillvery high (0.94 0.03). The path and its variability wererelatively conserved from 0 to 95% BWS. Finally, the addi-




7/13

tional kinematic data from smaller-weight adults (n 5, Table1) during overground walking with manual unloading were notsignificantly different from those obtained during walking on atreadmill when suspending the subjects in a parachute harnessconnected to a pneumatic device that applied a controlledupward force. The correlation coefficient between the timeseries of the VM y during treadmill and overground walking wasvery high (0.97 0.02 at 0 BWS and 0.95 0.03 at 40–90%BWS), and the maximal foot lift during swing and the relativeposition of the main peak were also similar at all BWS levels,independent of the unloading procedure (Fig. 4, bottom).

In contrast, the reduced “gravity” considerably affected theshape of the foot path in toddlers: all toddlers tended to make

an unusual high foot lift and forward foot overshoot (a kicking-like movement) at the end of swing at high levels of bodyunloading (Fig. 4 A, right ). For instance, at 0 BWS, the meanpeak position of the VM y was at midswing, whereas at 40–90%BWS, it was significantly shifted forward (Fig. 4 B; P 0.001,paired t -test). This behavior (forward foot overshoot) persistedover all consecutive steps performed at high BWS levels inone trial (typically 2–5 steps). It is worth noting that thesame unloading procedure that was used in toddlers did notchange significantly the shape of the foot path in olderchildren (Fig. 4).

Variability inherent in the toddlers’ stepping pattern mayhave limited our ability to draw definitive conclusions about

FIG. 3. Effect of body weight support (BWS) on intersegmental coordination. A: averaged thigh, shank, and foot elevation angles and corresponding gait loopsplotted in 3D in toddlers, older children (2–5 yr), and adults at 2 levels of BWS (0 and 75%). Data are plotted vs. normalized gait cycle. Data for adults referto treadmill walking at 3 km/h. A loop is obtained by plotting thigh waveform vs. shank and foot waveforms (after mean values subtraction). Paths progress intime in counter-clockwise direction, foot strike and lift-off corresponding approximately to the top and bottom of the loop, respectively. B: characteristics of gaitloops in toddlers, older children, and adults. Interpolation planes result from orthogonal planar regression. First eigenvector (u1) is aligned with the long axisof gait loops, 2nd eigenvector (u2) is aligned with the short axis, and 3rd eigenvector (u3) is normal to the plane. Planar covariance for thigh-shank-foot loopsis assessed by percentage of variance accounted for by eigenvectors u1, u2, and u3 (PV1, PV2, and PV3, respectively). PV1 and PV2 are related to the global formof 3D gait loop, and PV3 provides an index of planarity of the loop (0% corresponds to an ideal plane). Normal to covariance plane is characterized by directioncosines of plane normal (u3) with positive semi-axis of thigh, shank, and foot angular coordinates, respectively (u3t, u3s, and u3f ). Each point corresponds to 1individual stride. Changes with BWS are fitted by a linear function.




8/13

the exact dependence of the foot trajectory on the BWS level(Fig. 4). To estimate the level at which changes in the forwardfoot overshoot become significant, we compared the relativepeak position (Fig. 4 B) obtained at 0 BWS and under weightsupported conditions within each 20% interval of BWS: wemoved the BWS window every 10% (10–30%, then 20–40%,etc.) until the difference with the 0 BWS condition becamesignificant (using t -test). Such procedure revealed significantchanges with respect to the unsupported walking conditionbeginning from the 30% BWS values.

Figure 5 A shows the average template of the foot trajectoryand the dynamics of the pitch () angle in toddlers, olderchildren (2–5 yr), and adults. In the time domain (Fig. 5 B), atall levels of BWS, the vertical excursion of the foot in toddlers

tended to have only one maximum at midswing, and thecorrelation coefficient between the time-series of the VM yduring swing in toddlers and the corresponding ensembleaverage in adults remained negative at 40–90% BWS(0.72 0.10). However, there were significant changes inthe hip angle during swing initiation at high levels of BWS(Fig. 5C ): the hip tended to be slightly extended (by 7°) intoddlers, whereas it was similar in older children and adultsduring body unloading. The knee angle tended to be somewhatflexed during stance-to-swing transition in toddlers at highlevels of BWS (Fig. 5C ).

The amplitude of the vertical hip (GTy) oscillations through-out the gait cycle did not change significantly with BWS(2.8 0.6 cm at 0 BWS and 3.2 0.8 cm at 40–90% BWS,

FIG. 4. Effect of BWS on shape of footpath. A: examples in 3 toddlers, 1 older child,and 1 adult. Trajectories over 6–10 steps weresuperimposed in space relative to mean VM xposition during each stride; adult data werealso corrected for treadmill belt speed. Notethat path shapes are roughly comparable in all

support conditions in the adult and in the olderchild, whereas these both differ from those forthe toddlers. B: maximal foot (VM) lift andrelative position of main peak. In adults, weanalyzed the 1st peak of VM path (just afterlift-off) because it was usually higher than the2nd peak at the end of swing. Changes withBWS are fitted by a linear function in adultsand older children and by a 2nd-order polyno-mial function in toddlers. Values for adultgroup were obtained both for overgroundwalking and for walking on a treadmill at 3km/h. L, limb length.




9/13

P 0.05) making it unlikely that the increased trunk oscilla-tions could govern larger foot movements in space. In addition,the mean trunk orientation in toddlers tended to tilt forwardduring body unloading (BWS 40%; by 3.1 5.6°, P 0.05,one-tailed t -test), although in older children (2–5 yr) theamount of forward tilt was similar (6.1 4.8°). We alsoanalyzed the foot path relative to the instantaneous hip posi-tion. Again, the distinctive forward foot overshoot was presentin toddlers at high levels of BWS (Fig. 6 A), and stepping atthese levels was characterized by an apparent change in theinitial VM direction during early swing. The VM tolerancearea also remained high (19.8 3.7 cm2 /cm with 40 –90%body support and 20.5 6.5 cm2 /cm without). In toddlers, the

pitch angle of this directional vector (mean circular SD,relative to the horizontal) was 32 9° at 0 BWS and 0 4°at 40–90% BWS, whereas in older children it was 0 7 and5 6° and in adults it was 4 6 and 1 6°, respectively.At 0 BWS, this angle (32 9°) was significantly differentfrom that of adults and older children (P 0.001, Watson’s U 2

test), indicating a high initial foot elevation during swing intoddlers, whereas at high levels of BWS, the difference wasminor (Fig. 6 B).

Figure 7 shows the effect of BWS on the shape of the footpath in all groups of subjects (Table 1). Older children (2–5 yr)displayed quite mature foot trajectory characteristics, whereasintermediate walkers (1.5–5 mo after the onset of walking)

FIG. 5. A: average VM trajectories in space in toddlers, older children, and adults during walking at 0 and at high levels of BWS. Plots are anisotropic, verticalscale being expanded relative to horizontal scale. Data for the adult group were obtained from smaller-weight adults ( n 5, Table 1) during overground walkingwith manual BWS. Bottom panels show instantaneous pitch angle of VM path. LO, lift off; TD, touch down. B: time-course of average vertical foot (VM y)displacement and average hip, knee, and ankle angles. Black line corresponds to walking at high levels of BWS. Solid line surrounded by gray shading representsunsupported stepping curve and SE, respectively. Vertical dotted lines correspond to stance-to-swing transition. C : changes in joint angles (SD) at moment of lift off at high levels of BWS with respect to unsupported stepping. Asterisks denote significant differences ( P 0.05).




10/13

showed only partial improvement, supporting the conclusionthat a few months of walking experience is the minimumpractice duration for the beginning of an effective (adult-like)foot trajectory control.

D I S C U S S I O N

Our study supports the common notion that toddlers are ableto adapt to environmental changes and they can step at differ-ent levels of BWS. The intersegmental coordination (rotation

of the covariance plane; Fig. 3 B) showed similar monotonicchanges with BWS in all subjects, suggesting similar ampli-tude and phase regulation of segment motion with body un-loading. We also confirmed our previous findings (Ivanenko etal. 2005) that trunk stabilization (low BWS levels) did notaffect significantly limb motion. Our main finding in this study,however, was that, in contrast to adults and older children, foottrajectory characteristics in toddlers showed systematicchanges with limb unloading. This difference was exemplifiedby the higher foot lift, forward foot overshoot (Figs. 4 and 5 A)

FIG. 6. Foot trajectories relative to instantaneous hip position (swing phase). A: example of 1 toddler, the same child recorded 3 yr later, and 1 representativeadult. Top to bottom: stick diagrams of 1 cycle relative to instantaneous hip position and VM spatial density plots. Initial direction of VM path (1st 30% of swing)is indicated by an arrow. Spatial density of VM path was integrated over swing phase (across 10–15 steps) performed at different BWS. Density of each cellwas depicted graphically by means of a color scale (empty cells were excluded): the lower the density (toward blue in color-cued scale), the greater the variability.Plots are anisotropic, vertical scale being expanded relative to horizontal scale. B: mean initial directions of VM path for toddlers, older children, and adults forstepping at 0 and high levels of BWS.

FIG. 7. Effect of BWS on the shape of foot path (swing phase) in all groups of subjects. A: correlation coefficient between VM y data in children andcorresponding ensemble average in adults. B: relative position of main peak of VMy (same parameter as in Fig. 4 B). C : mean initial directions of VM path. IWK,intermediate walkers (see Table 1). Asterisks denote significant differences (P 0.05).




11/13

and remarkable change in the initial foot direction (Fig. 6) withbody weight unloading. These performance changes with levelof unloading were observed in all toddlers who were justbeginning to walk independently. Intermediate walkers (1.5–5mo after walking onset) showed only partial improvement infoot trajectory characteristics (Fig. 7).

Methodological considerations

Sensory signals interact with central rhythm-generating cen-ters in a complex manner (Pearson 1995). Afferent activityhelps to shape the motor patterns, control phase-transitions,and reinforce ongoing activity. In stance, load-detecting affer-ents, both from muscles and skin, facilitate the center for thegeneration of extension and inhibit that for flexion. In toddlers,the foot shape and ossification of the soft bones of the feet arestill immature at the age of 1 yr (Bertsch et al. 2004). Duringlimb loading, a variety of receptors can be activated, such asGolgi tendon organs, cutaneous receptors of the foot, and

spindles from stretched muscles (Duysens et al. 2000). Anumber of cutaneous reflexes participate in the fine control of foot positioning in animals (Guertin et al. 1995; Schouenborgand Weng 1994) and humans (Aniss et al. 1992; Van Wezel etal. 1997; Yang and Stein 1990; Zehr and Stein 1999). Eventhough the spatio-temporal distribution of load on the soles of the feet may be generally important in the regulationof locomotion, we have previously found that it does notappreciably affect the endpoint trajectory constraints in adults(Ivanenko et al. 2002): in the limit cases of 95% BWS and100% BWS with surrogate contact, detection of ground contactat the forefoot was sufficient to trigger the central program foraccurate endpoint control in adults. These findings suggest thatan idiosyncratic distribution of load on the soles (Bertsch et al.

2004; Hallemans et al. 2004) and nonplantigrade stepping intoddlers were not the primary reason for the observed adaptivedifference between toddlers and adults (Fig. 5).

One could argue that the enlarged vertical foot displace-ments at high levels of BWS in toddlers (Fig. 4) were a simpleresult of changes in the trunk motion influenced by an exper-imenter who supported the child. However, this is unlikelybecause of the following reasons. First, the same unloadingprocedure in older children and adults did not affect signifi-cantly the foot path (e.g., Figs. 4–6). Second, the amplitude of the vertical hip oscillations in toddlers throughout the gaitcycle did not change significantly with BWS. Also, the meantrunk angle varied similarly with weight support between the

groups of children. Finally, the foot path relative to the instan-taneous hip position showed similar behavior (Fig. 6) as inspace (Fig. 5 A). In addition, when holding the trunk under thearms, i.e., at low levels of trunk support (BWS 20%), legand foot movements were similar to those during unsupportedstepping (Ivanenko et al. 2005), making unlikely the contribu-tion of simple sensory contact of the hands with the child’strunk. As a final point, the observed phenomenon (Figs. 4 and6) was not caused by the aftereffect of transient load changes(Lam et al. 2003b) because it persisted over all (from 2 to 5)consecutive steps performed at high levels of BWS. Thereforewe believe that the unusual forward foot overshoot in toddlerswas caused by the nonspecific reduction of gravitational loadduring stance.

Effect of body weight unloading on foot trajectory control

in toddlers

What is special in toddler gait? When children start to walk independently, their gait is characterized by considerable trunk oscillations revealing postural instability (Assaiante et al.1993; Bril and Brenière 1993; Roncesvalles et al. 2001), wide

swinging arms (Sutherland et al. 1980), high interstep variabil-ity (Clark et al. 1988), and immature foot trajectory character-istics and intersegmental coordination (Cheron et al. 2001a,b;Ivanenko et al. 2004, 2005). In particular, early steppingtypically manifests itself in shorter step lengths, disorderedvertical hip displacements, nonplantigrade gait (lack of theheel-to-toe roll-over pattern), and a single peak foot lift duringswing. One may argue that this evolutionary adopted pattern(nonplantigrade gait with a high foot lift) is beneficial fortoddlers as an optimal starting point strategy (Ivanenko et al.2007) to avoid stumbling and falls and for reducing the effectof involuntary foot drag and the lack of dorsiflexor activity,which has been observed at the stance-to-swing transition

during treadmill stepping in young infants (Yang et al. 2004).The phenomenon may possibly reflect flexor-biased patterngeneration (Yakovenko et al. 2005) in infants that is replacedby mature control with walking experience. For instance, theflexion reflex and the placing reflex seem to be importantcomponents of infant’s stepping (Forssberg 1985; Zelazo et al.1972). An early form of locomotor-like alternating kicking andstepping movements is present at birth and even prenatally inhumans (Thelen 1981; Zelazo 1983). Hip flexion dominates inan early period of infancy during supported neonatal stepping(4 wk after birth), throughout the first year up to the begin-ning of independent walking (Forssberg 1985; Lamb and Yang2000; Okamoto et al. 2003; Pang and Yang 2001; Thelen 1981;Yang et al. 1998). The relative hyperflexion in hip and knee is

slightly reduced compared with the neonatal stepping but it isstill present before the onset of independent walking (Forss-berg 1985). Indeed, the amplitude of thigh movements isrelatively higher with respect to that of shank and foot move-ments, resulting in the wider gait loop in toddlers and by thelarger contribution of the second eigenvector (PV2 in Fig. 4)(see also Cheron et al. 2001a,b). Moreover, several aspects of infant stepping, including the vertical movement of the hip joint and of the foot, the shape of the gait loop, and the burstsof EMG activity on foot placements, all suggest that toddlersimplement a mixed locomotor strategy, combining forwardprogression with elements of stepping in place (Ivanenko et al.2005). This hypothesis is also consonant with the finding that

toddlers had not yet learned to use transverse pelvic rotation inleg-pelvis coordination to increase speed of walking (Holt et al.2006).

In addition to the idiosyncratic foot trajectory characteristicsduring unsupported stepping, our data indicate that toddlersadapt differently than adults and older children to variations inload (Figs. 5 A, 6, and 7). At low levels of BWS (30%), foottrajectory characteristics were similar to those during unaidedstepping, in agreement with our previous results (Ivanenko etal. 2005). However, at higher levels of BWS, the “flexor-biased” (or stepping-in-place) component of swing initiation(Figs. 5 A and 6) decreased significantly with limb unloadingresulting in alternating swing movements of the legs with anovershoot at the end of swing. If one accepts the terminology




12/13

of motor primitives or stereotypies in infants (Thelen 1981),there was a switch from the stepping-in-place motor compo-nent during unsupported walking to the kicking componentduring walking at high levels of BWS.

We do not know the exact nature or transmission mecha-nisms of the information that shapes the foot trajectory and itsvariation with foot unloading in toddlers. It has been hypoth-esized that the infant population provides a good subject poolfor studying the afferent control of walking in the human,before cerebral influences are fully developed (Yang et al.1998), and some of the adaptive mechanisms we found mayreside in subcortical structures. For instance, spinal cats cannotcompensate for cutaneous denervation, whereas intact catswith normal supraspinal control can easily adapt to the lack of load-related cutaneous information (Bouyer and Rossignol2003). Therefore we cannot exclude the possibility that theobserved phenomenon reflects undeveloped supraspinal con-trol of foot placement in toddlers because the descendingpathways (Kinney et al. 1988; Paus et al. 1999; Yang and

Gorassini 2006) and central conduction delays (Eyre et al.1991) are immature at the beginning of unsupported walking.Ankle plantarflexor activity is an important contributor to

the initiation of the swing phase at the end of stance (Neptuneet al. 2001), and its immaturity can possibly reorganize thelocomotor program by compensating with hip flexors or per-forming shorter steps, as happens in patients (Nadeau et al.1999) and in the elderly (Winter et al. 1990). In toddlers,virtually no power is generated at the ankle joint beforefoot-off (Forssberg 1985; Hallemans et al. 2005). The argu-ment can be made, however, that limited muscle strength maynot be the main factor for body weight perception and forwardpropulsion. For instance, when lifting oneself out of the waterafter a swim or bath, one perceives the body as being very

heavy, and it can be initially very difficult to walk, even thoughthe absolute strength of the muscles is not affected at all. Thisimplies that body weight bearing requires appropriate posturaland coordinative behavior. As for the toddler stepping, even athigh levels of body weight support (requiring much less plan-tarflexor activity for propulsion), toddlers still move their legsdifferently from adults and older children (with a differentdynamics and an overshoot of foot motion; Figs. 4 and 5 A).Finally, toddlers are able to carry the additional heavy weights(Adolph and Avolio 2000). We instead suggest that the toddleridiosyncratic pattern may be a result of immature foot–supportinteractions and a lack of the appropriate pendulum behavior of the limbs and of the center-of-body-mass, rather than muscular

underdevelopment. The transition to a pendulum pattern occurswithin a few months of independent walking experience (Iv-anenko et al. 2004) and is accompanied by rapid improvementsin stability (Assaiante et al. 1993; Bril and Brenière 1993;Goodman et al. 2000; Roncesvalles et al. 2001) and other gaitcharacteristics (Fig. 2, see also Cheron et al. 2001b; Hallemanset al. 2006; Ivanenko et al. 2005; Sutherland et al. 1988).Experience with body mass propulsion requires producingactive forces in the form of escapement pulses that mayactivate and complement the pendulum or spring dynamics(Holt et al. 2006), given appropriate (unsupported gait) input(Ivanenko et al. 2007). In fact, walking practice is a muchstronger predictor of posture and gait improvement in newlywalking toddlers than age or body size (Adolph et al. 2003;

Holt et al. 2006; Ivanenko et al. 2004, 2005; Roncesvalles et al.2001; Sundermier et al. 2001; Sutherland et al. 1988).

A C K N O W L E D G M E N T S

We thank Dr. W. Miller for helpful comments on the manuscript and V.Sabia for the toddler drawing (Fig. 1 A).

G R A N T S

The financial support of Italian Health Ministry, Italian University Ministry(PRIN and FIRB projects), and Italian Space Agency is gratefully acknowl-edged.

R E F E R E N C E S

Adolph KE. Learning in the development of infant locomotion. Monogr Soc Res Child Dev 62: 1–158, 1997.

Adolph KE, Avolio AM. Walking infants adapt locomotion to changing bodydimensions. J Exp Psychol Hum Percept Perform 26: 1148–1166, 2000.

Adolph KE, Vereijken B, Shrout PE. What changes in infant walking andwhy. Child Dev 74: 475–497, 2003.

Aniss AM, Gandevia SC, Burke D. Reflex responses in active muscleselicited by stimulation of low- threshold afferents from the human foot.

J Neurophysiol 67: 1375–1384, 1992.

Assaiante C, Thomachot B, Aurenty R. Hip stabilization and lateral balancecontrol in toddlers during the first four months of autonomous walking. Neuroreport 4: 875–878, 1993.

Batschelet E. Circular Statistics in Biology. New York: Academic Press,1981.

Bernstein N. The Co-ordination and Regulation of Movements. Oxford, UK:Pergamon Press, 1967.

Bertsch C, Unger H, Winkelmann W, Rosenbaum D. Evaluation of earlywalking patterns from plantar pressure distribution measurements. First yearresults of 42 children. Gait Posture 19: 235–242, 2004.

Bianchi L, Angelini D, Orani GP, Lacquaniti F. Kinematic co-ordination inhuman gait: relation to mechanical energy cost. J Neurophysiol 79: 2155–2170, 1998.

Borghese NA, Bianchi L, Lacquaniti F. Kinematic determinants of humanlocomotion. J Physiol 494: 863–879, 1996.

Bouyer LJ, Rossignol S. Contribution of cutaneous inputs from the hindpawto the control of locomotion. II. Spinal cats. J Neurophysiol 90: 3640–3653,

2003.Bril B, Brenière Y. Posture and independent locomotion in early childhood:

learning to walk or learning dynamic postural control? In: The Development of Coordination in Infancy, edited by Savelsbergh GJP. Amsterdam:Elsevier, 1993, p. 337–358.

Cavagna GA, Franzetti P, Fuchimoto T. The mechanics of walking inchildren. J Physiol 343: 323–339, 1983.

Cheron G, Bengoetxea A, Bouillot E, Lacquaniti F, Dan B. Early emer-gence of temporal co-ordination of lower limb segments elevation angles inhuman locomotion. Neurosci Lett 308: 123–127, 2001a.

Cheron G, Bouillot E, Dan B, Bengoetxea A, Draye JP, Lacquaniti F.

Development of a kinematic coordination pattern in toddler locomotion:planar covariation. Exp Brain Res 137: 455–466, 2001b.

Clark JE, Whitall J, Phillips SJ. Human interlimb coordination: the first 6months of independent walking. Dev Psychobiol 21: 445–456, 1988.

Duysens J, Clarac F, Cruse H. Load-regulating mechanisms in gait and

posture: comparative aspects. Physiol Rev 80: 83–133, 2000.Duysens J, Pearson KG. Inhibition of flexor burst generation by loading ankleextensor muscles in walking cats. Brain Res 187: 321–332, 1980.

Eyre JA, Miller S, Ramesh V. Constancy of central conduction delays duringdevelopment in man: investigation of motor and somatosensory pathways.

J Physiol 434: 441–452, 1991.Forssberg H. Ontogeny of human locomotor control. I. Infant stepping,

supported locomotion and transition to independent locomotion. Exp Brain Res 57: 480–493, 1985.

Garn SM. De Genetica Medica, part II. Rome: Gregor Mendel Institute, 1966,p. 415–434.

Gazzani F, Fadda A, Torre M, Macellari V. WARD: a pneumatic system forbody weight relief in gait rehabilitation. IEEE Trans Rehabil Eng 8:506–513, 2000.

Goodman L, Riley MA, Mitra S, Turvey MT. Advantages of rhythmicmovements at resonance: minimal active degrees of freedom, minimal noise,and maximal predictability. J Motor Behav 32: 3–8, 2000.




13/13

Guertin P, Angel MJ, Perreault MC, McCrea DA. Ankle extensor group Iafferents excite extensors throughout the hindlimb during fictive locomotionin the cat. J Physiol 487: 197–209, 1995.

Hallemans A, D’Aout K, De Clercq D, Aerts P. Pressure distributionpatterns under the feet of new walkers: the first two months of independentwalking. Foot Ankle Int 24: 444–453, 2004.

Hallemans A, De Clercq D, Otten B, Aerts P. 3D joint dynamics of walkingin toddlers. A cross-sectional study spanning the first rapid development

phase of walking. Gait Posture 22: 107–118, 2005.Hallemans A, De Clercq D, Van Dongen S, Aerts P. Changes in foot-function parameters during the first 5 months after the onset of independentwalking: a longitudinal follow-up study. Gait Posture 23: 142–148, 2006.

Holt KG, Saltzman E, Ho CL, Kubo M, Ulrich BD. Discovery of thependulum and spring dynamics in the early stages of walking. J Motor

Behav 38: 206–218, 2006.Ivanenko YP, Dominici N, Cappellini G, Dan B, Cheron G, Lacquaniti F.

Development of pendulum mechanism and kinematic coordination from thefirst unsupported steps in toddlers. J Exp Biol 207: 3797–3810, 2004.

Ivanenko YP, Dominici N, Cappellini G, Lacquaniti F. Kinematics in newlywalking toddlers does not depend upon postural stability. J Neurophysiol 94:754–763, 2005.

Ivanenko YP, Dominici N, Lacquaniti F. Development of independentwalking in toddlers. Exerc Sport Sci Rev In press.

Ivanenko YP, Grasso R, Macellari V, Lacquaniti F. Control of foottrajectory in human locomotion: role of ground contact forces in simulated

reduced gravity. J Neurophysiol 87: 3070–3089, 2002.Jensen JL. The puzzles of motor development: how the study of developmen-

tal biomechanics contributes to the puzzle solutions. Infant Child Dev 14:501–511, 2005.

Jensen JL, Schneider K, Ulrich BD, Zernicke RF, Thelen E. Adaptivedynamics of the leg movement patterns of human infants. I. The effects of posture on spontaneous kicking. J Motor Behav 26: 303–312, 1994.

Kinney HC, Brody BA, Kloman AS, Gilles FH. Sequence of central nervoussystem myelination in human infancy. J Neuropathol Exp Neurol 47:217–234, 1988.

Lacquaniti F, Grasso R, Zago M. Motor patterns in walking. News PhysiolSci 14: 168–174, 1999.

Lacquaniti F, Ivanenko YP, Zago M. Kinematic control of walking. Arch Ital Biol 140: 263–272, 2002.

Lam T, Wolstenholme C, van der Linden M, Pang MY, Yang JF.

Stumbling corrective responses during treadmill-elicited stepping in human

infants. J Physiol 553: 319–331, 2003a.Lam T, Wolstenholme C, Yang JF. How do infants adapt to loading of thelimb during the swing phase of stepping? J Neurophysiol 89: 1920–1928,2003b.

Lamb T, Yang JF. Could different directions of infant stepping be controlledby the same locomotor central pattern generator? J Neurophysiol 83:2814–2824, 2000.

McCrory J, Schwass J, Connell R, Cavanagh P. In-shoe force measure-ments from locomotion in simulated zero gravity during parabolic flight.Clin Biomech (Bristol, Avon) 12: S7, 1997.

Nadeau S, Gravel D, Arsenault AB, Bourbonnais D. Plantarflexor weaknessas a limiting factor of gait speed in stroke subjects and the compensating roleof hip flexors. Clin Biomech (Bristol, Avon) 14: 125–135, 1999.

Neptune RR, Kautz SA, Zajac FE. Contributions of the individual ankleplantar flexors to support, forward progression and swing initiation duringwalking. J Biomech 34: 1387–1398, 2001.

Okamoto T, Okamoto K, Andrew PD. Electromyographic developmental

changes in one individual from newborn stepping to mature walking. Gait Posture 17: 18–27, 2003.

Pang MY, Yang JF. Interlimb co-ordination in human infant stepping. J Physiol 533: 617–625, 2001.

Paus T, Zijdenbos A, Worsley K, Collins DL, Blumenthal J, Giedd JN,Rapoport JL, Evans AC. Structural maturation of neural pathways inchildren and adolescents: in vivo study. Science 238: 1908–1911, 1999.

Payne V, Isaacs LD. Human Motor Development: A Lifespan Approach (6th

ed.). Blacklick, OH: McGraw Hill, 2005.Pearson KG. Proprioceptive regulation of locomotion. Curr Opin Neurobiol

5: 786–791, 1995.Roncesvalles MNC, Woollacott MH, Jensen JL. The development of com-

pensatory stepping skills in children. J Motor Behav 32: 100–111, 2000.Roncesvalles MNC, Woollacott MH, Jensen JL. The development of lower

extremity kinetics for balance control in infants and young children. J Motor

Behav 33: 180–192, 2001.Schmuckler MA. Perception-action coupling in infancy. In: The Development

of Coordination in Infancy, edited by Savelsburgh GJP. Amsterdam:Elsevier Science, 1993, p. 137–173.

Schouenborg J, Weng HR. Sensorimotor transformation in a spinal motorsystem. Exp Brain Res 100: 170–174, 1994.

Shirley MM. The First Two Years: A Study of Twenty-Five Babies. Postural

and Locomotor Development. Minneapolis: University of Minnesota Press,

1931, vol. 1.Sundermier L, Woollacott M, Roncesvalles N, Jensen J. The development

of balance control in children: comparisons of EMG and kinetic variablesand chronological and developmental groupings. Exp Brain Res 136: 340–350, 2001.

Sutherland D, Olshen R, Biden E, Wyatt M. The development of maturewalking. In: Clinics in Developmental Medicine. Philadelphia, PA: Mac-keith Press, 1988, p. 1–271.

Sutherland DH, Olshen R, Cooper L, Woo SL. The development of maturegait. J Bone Joint Surg 62: 336–353, 1980.

Thelen E. Kicking, rocking, and waving: contextual analysis of rhythmicalstereotypies in normal human infants. Anim Behav 29: 3–11, 1981.

Thelen E. Treadmill-elicited stepping in seven-month-old infants. Child Dev57: 1498–1506, 1986.

Thelen E, Cooke DW. Relationship between new-born stepping and laterwalking: a new interpretation. Dev Med Child Neurol 29: 380–393, 1987.

Thelen E, Fisher DM, Ridley-Johnson R. The relationship between physicalgrowth and newborn reflex. Infant Behav Devel 7: 479–493, 1984.

Thelen E, Ulrich B, Niles D. Bilateral coordination in human infants: steppingon a split-belt treadmill. J Exp Psychol Hum Percept Perform 13: 405–410,1987.

Van Wezel BM, Ottenhoff FA, Duysens J. Dynamic control of location-specific information in tactile cutaneous reflexes from the foot during humanwalking. J Neurosci 17: 3804–3814, 1997.

Winter DA. Foot trajectory in human gait: a precise and multifactorial motorcontrol task. Phys Ther 72: 45–53, 1992.Winter DA, Patla AE, Frank JS, Walt SE. Biomechanical walking pattern

changes in the fit and healthy elderly. Phys Ther 70: 340–347, 1990.Yakovenko S, McCrea DA, Stecina K, Prochazka A. Control of locomotor

cycle durations. J Neurophysiol 94: 1057–1065, 2005.Yang JF, Gorassini M. Spinal and brain control of human walking: implica-

tions for retraining of walking. Neuroscientist 12: 379–389, 2006.Yang JF, Lam T, Pang MYC, Lamont E, Musselman K. Infant stepping: a

window to the behaviour of the human pattern generator for walking. Can J Physiol Pharmacol 82: 662–674, 2004.

Yang JF, Lamont EV, Pang MY. Split-belt treadmill stepping in infantssuggests autonomous pattern generators for the left and right leg in humans.

J Neurosci 25: 6869–6876, 2005.Yang JF, Stein RB. Phase-dependent reflex reversal in human leg muscles

during walking. J Neurophysiol 63: 1109–1117, 1990.Yang JF, Stephens MJ, Vishram R. Infant stepping: a method to study the

sensory control of human walking. J Physiol 507: 927–937, 1998.Zehr EP, Stein RB. What functions do reflexes serve during human locomo-

tion? Prog Neurobiol 58: 185–205, 1999.Zelazo PR. The development of walking: new findings and old assumptions.

J Motor Behav 15: 99–137, 1983.Zelazo PR, Zelazo NA, Kolb S. “Walking” in the newborn. Science 176:

314–315, 1972.


J Neurophysiol • VOL 97 • APRIL 2007 • www jn org

Paper 3 Marcha Niños

Documents

Transcript of Paper 3 Marcha Niños