Exp Brain Res (1998) 121:255±262
Springer-Verlag 1998
RESEARCH ARTICLE
Yvon Brenire ´ Blandine Bril
Development of postural control of gravity forces in children during the first 5 years of walking
Received: 2 June 1997 / Accepted: 22 December 1997
Abstract The aim of this work was to propose developmental indexes relative to the control of balance and gravity forces, using force-plate data, for children in their first 5 years of independent walking. The first part of this paper is devoted to the definition of an index to quantify postural capacity during walking. Based on the assumption that the vertical acceleration of center of mass (CM) reflects the capacity of muscular forces between the head-arms-trunk and the stance leg segments to control the external forces, the value of the CM vertical acceleration at foot contact is proposed as a developmental index of the postural capacity of the child to control gravitational forces. This index was analyzed longitudinally in five children, over the course of eight experimental sessions. The children were examined during their first 5 years of independent walking (for a total of 457 step sequences). The covariation between the CM vertical acceleration at foot contact and the gait velocity was considered as a second index characterizing the development of coordination between the postural and dynamic requirements of body progression. From these indexes it was established that the postural capacity needed just to control balance with the leg muscles was not attained before 4±5 years of independent walking, i.e., at about 5±6 years of age. Key words Child gait ´ Posturokinetic capacity ´ Balance control ´ Hip forces ´ Developmental indexes
Introduction The development of walking in children is in keeping with the theory of self-organizing systems and infant mo-
)
Y. Brenire ( ) Laboratoire de Physiologie du Mouvement, ER-CNRS 124, UniversitØ Paris-sud, F-91405 Orsay, France e-mail:
[email protected], Fax: +33-169-855219 B. Bril EHESS, Apprentissage, Cognition et Contexte, 54 Boulevard Raspail, F-75270 Paris Cedex 06, France
tor development (Thelen 1986), which considers locomotor skill development as resulting from a multidimensional phenomenon, consistent with the dynamic approach. In this context, the development of gait is largely multidimensional, especially considering the different posturodynamic requirements of gait, which can be of a locomotor, postural, or anticipatory nature. However, the development of the locomotor and postural components do not necessarily coincide. Through kinematic techniques, Sutherland et al. (1988) found that children of 3 years already displayed five determinants of mature gait, while Beck et al. (1981), using a force plate, showed that mature patterns were reached at 5 years of age. From a longitudinal study, Bril and Brenire (1992, 1993) established a two-phase process of gait acquisition, the first phase being devoted to the learning of gait postural requirements from 3 to 6 months after the onset of independent walking, with a second phase, lasting several years, being devoted to fine tuning of gait. The development of a plantigrade gait has been subject to different interpretations; resulting from a central development (Forssberg 1985) or from postural adaptations (Thelen et al. 1992). By 7 years of age, the adult muscle activation pattern is complete (Woolllacott and Jensen 1996), while head control and head-trunk coordination (Assaiante and Amblard 1992), integration of body parameters (Brenire et al. 1989; Ledebt and Brenire 1994) and development of anticipatory postural adjustments (Hirschfeld and Forssberg 1992; Ledebt et al. 1998) necessitated at least 8 years of walking experience. Analyses of mechanical energy transfer have been used to study the development of gait velocity with respect to an ªoptimal speedº in children 2±12 years of age (Cavagna et al. 1983) or with respect to other optimality criteria, according to the self-optimization principle (Jeng et al. 1996). Thus the idea that greater energy exchange implies that less mechanical energy is required from the muscle tendon (Willems et al. 1995) is based on a reasonable assumption. However, this assertion implicitly assumes that subjects potentially have the muscular capacity to generate the movements and control the
256
posture implied in the best ªenergy managementº solution that they can adopt. What is true in adults is not necessarily true in children; in particular, one of the most constraining postural requirements associated with gait is the control of external forces such as gravity forces from hip muscles, which requires a high level of strength in these muscles, principally during the one-leg stance phases. Some of the studies which have analyzed this particular involvement of hip muscles in the upper body balance control (McKinnon and Winter 1993) showed the great intensity of musculotendinous forces, which could reach up to 6±8 times the body weight at certain points in the one-leg stance phases (Frankel and Nordin 1980; Paul 1980). The short lever arms of the hip muscles are responsible for the great intensity of these forces, and it is reasonable to assume that children do not have the innate capacity to develop such forces once they begin to walk. Thus the hip, which is the interface between the locomotor movements and the head-arms-trunk (HAT) postural control is a particularly sensitive joint for body balance during gait, and its kinematics reflect the global capacity of leg muscles to compensate for the effects of gravity forces, i.e., the body weight. Thus, from this analysis of postural requirements of gait which concern more specifically the balance control in the frontal plane, the aim of this work was to study the development of the ability, in children, to satisfy these requirements. This complex posturokinetic problem involves analyzing the development of the capacity of hip muscles to control the effects of gravitational forces during gait for balance maintenance in the frontal plane. These intersegmental forces are neither directly perceptible nor measurable via the usual techniques. Given the methodological and biomechanical difficulties of calculating and analyzing the leg muscular forces in young children (Schneider et al. 1990), this developmental analysis is twofold: first, it proposes a simplified method for analyzing this postural capacity from a significant parameter of the hip kinematics, and, second, it establishes the developmental course of this parameter. The simplified method is based on the balance (or imbalance) between the external forces and the internal forces reflected by the vertical acceleration of the center of mass (CM) and assumes that the CM acceleration can be representative of the hip acceleration, at least at certain points in the gait cycle. Analytically, during gait, the CM vertical acceleration z 00G can be calculated from external forces, i.e., body weight (mg) and vertical ground reaction (Rz), as established from the dynamics equation: mz 00G
Rz ÿ mg But, unlike a passive material system, for this particular and active system, z 00G always results from vertical accelerations of all the body segments, i.e., from all the musculoskeletal forces involved in movement. Thus Rz reflects the ability or inability of the musculoskeletal system to compensate for the body weight. In particular, when standing still at the beginning of a sequence of steps,
z 00G =0 means that the entire musculoskeletal system is just compensating for the body weight. During gait the sign of z 00G reflects, at the time being considered, the deficit (if it is negative) or surplus (if it is positive) in the postural capacity of musculoskeletal system. However, this global dynamic aspect does not indicate whether or not balance is necessary at the considered time or which muscles are involved in this control. Thus a preliminary developmental study showed that, after 6 months of independent walking, toddlers always displayed negative values of the vertical CM acceleration at foot contact, while adults displayed positive values. It concluded that the child can walk, despite lacking the strength required to maintain balance, but that he is ªwalking by fallingº (Brenire and Bril 1988). From this comparison of vertical CM acceleration between young and mature walkers, the question is when does the child develop the same values as adults? Growth by itself does not facilitate the development of muscular capacities, because the body weight increases by a ratio of 1±6 between the onset and the end of growth. Thus the muscles of the stance leg not only have to compensate for the hypothetical deficit in this postural capacity but also have to compensate for this weight increase. Biomechanically, from the previous relation, written as z 00G = (Rz ± mg)/m, we see that if z 00G displays the same value, negative or positive, during a developmental period while the body weight, mg, has increased to become m©g (with m©>m), this means that the musculoskeletal system has necessarily developed greater strengths; however, this increase in muscular forces only serves to compensate for the effects of the increase in body weight. But if z 00G , negative or positive, increases during a developmental period, the development of muscular strength during the same period not only compensates for this increase in weight but also reduces the previous deficit in the posturokinetic capacity. In other words, the CM vertical acceleration appears as a developmental index of the global postural capacity to control balance or gravitational forces during gait. However, to use the CM vertical acceleration more specifically as a developmental index of the capacity of hip muscles with respect to the postural requirements of gait, it was necessary to find at what point in the gait cycle the CM vertical acceleration is the most representative of this hip vertical acceleration. Foot contact, which is considered to be the most unstable period of the entire cycle (Yang et al. 1990), i.e., the most constraining for the hip muscles, has been chosen as the reference time to compare these local (at the hip) and global accelerations. The comparison between hip and CM accelerations was achieved via the methodology, using accelerometers and a force plate, already used in adults (Brenire and Dietrich 1992). This time was easy to detect from a force plate, reproducible from one step to another and from one child to another, and on the whole particularly well suited for longitudinal analysis.
257
Materials and methods Experimental procedures All measurements were made from the force plate used in previous studies on gait in children. The force plate is an equilateral triangle with 2-m sides, suspended by its apexes from nine transducers and large enough for the child to make 3±7 steps (Fig. 1A) depending on the age. The force plate measured the three components of ground reaction, R, and the instantaneous position of the point of application of the vertical component of this force, P (data sampling frequency, 200 Hz). For this study only the vertical component, Rz, and the coordinate xP of P along the progression axis were considered. The precision of force determination was less than 5 N and less than 5 mm for the xP± measure. A preliminary test using a vertical accelerometer (ENTRAN type EGCS 5 g) fixed on a plate attached to the pelvic girdle, with the accelerometer placed at the hip level (Fig. 2), as in a previous experiment in adults (Brenire and Dietrich 1992), allowed comparison of the vertical hip acceleration and the vertical CM acceleration (Fig. 1B). The accelerometric test and the comparison between vertical hip and CM accelerations were performed on six children of 7±8 years of age who executed a total of 24 step sequences. Subjects Five children (four boys, one girl) were examined longitudinally from the onset of independent walking (IW; see Table 1), i.e., in the week during which children were able to make 5±10 independent steps (experimental session I), and during the first 5 years of IW, respectively, between 1 and 2 months of IW (experimental session II), at 3 months (experimental session III), at 5 months (experimental session IV), between 6 and 8 months (experimental session V), between 15 and 21 months (experimental session VI), between 28 and 34 months (experimental session VII), between 42 and 46 months (experimental session VIII), and between 54 and 60 months (experimental session IX). Three adults constituted a target group for the same parameters; they executed a total of 30 step sequences at different velocities on the force plate. The mean values and standard deviations of the same gait parameters were calculated.
Fig. 1A, B Force plate and mechanical traces. A Force plate and walkway (Oxyz fixed ground reference, G child's center of gravity, m® aG dynamic resultant vector, W subject's weight, R resultant vector of ground forces, P center of foot pressure). B Comparison of vertical hip acceleration and CM acceleration (z 00H hip vertical acceleration, z 00G CM vertical acceleration; z 00FC values of z 00G at FC, XP CP anteroposterior coordinate, t time, FC successive foot contacts)
accelerometers
Protocol Each child performed 20 sequences of steps on the force plate at each session. Depending on the age of the child, each sequence included 3± 7 steps on the force plate and several more on the walkway beyond it. Before each walking sequence, the child was standing still, barefoot. The child then walked, at their own pace, toward their mother standing at the far side of the force plate. Selection of movements was carried out as indicated in a previous study (Brenire et al. 1989): only sequences of steps executed naturally by the child in the forward direction which led to a steady-state velocity were analyzed. Gait developmental parameters
Fig. 2 Experimental device for recording of the vertical hip acceleration from accelerometers
The force plate data were used for the following parameters (Fig. 1B): A ´ mz 00G , which is directly calculated from the force plate data, Rz, by applying the direct dynamics principle (with m, child mass):
Results
mz 00G
Rz ÿ mg
Characterization of developmental parameters z 00FC ,
The CM vertical acceleration at successive foot contacts (FC), is calculated for each sequence of steps as the mean value of three consecutive values of z 00FC during a whole gait cycle. B. The instantaneous coordinates of anteroposterior coordinates (CP), xP, which are a force plate data. It is used to determine the FC and to calculate the step frequency, the step length, L, as the variation of xP between two FC and the mean progression velocity, v=Lf.
Comparison between hip vertical acceleration and CM vertical acceleration Figure 1B presents the superimposed curves, in a child, of vertical hip acceleration and CM vertical acceleration during a step sequence, obtained from the different exper-
258
Fig. 3 Developmental traces of CM vertical accelerations during a gait cycle, in the same child at different ages compared with a trace from an adult. (z 00G CM vertical acceleration directly calculated from the force plate recordings at different ages, z 00FC value of z 00G at foot contact)
imental devices, i.e., the vertical accelerometer at the hip and the force plate. The correlation between the values of vertical CM and hip acceleration measured at t=FC was very significant (r=0.72; P
