ARTICLE IN PRESS

Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200

www.elsevier.com/locate/neubiorev

Review

Development of postural control in typically developing children and

children with cerebral palsy: Possibilities for intervention?
Victorine B. de Graaf-Peters, Cornill H. Blauw-Hospers, Tineke Dirks, Hanneke Bakker,
Arie F. Bos, Mijna Hadders-Algra
Department Neurology and Developmental Neurology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

Abstract

The basic level of postural control is functionally active from early infancy onwards: young infants possess a repertoire of direction-
specific postural adjustments. Whether or not direction-specific adjustments are used depends on the child’s age and the nature of the
postural task. The second level of control emerges after 3 months: children start to develop the capacity to adapt postural activity to
environmental constraints. But the adult form of postural adaptation first emerges after adolescence.
Children with cerebral palsy (CP) in general have the ability to generate direction-specific adjustments, but they show a delayed
development in the capacity to recruit direction-specific adjustments in tasks with a mild postural challenge. Children with CP virtually
always have difficulties in the adaptation of direction-specific activity.
The limited data available on the effect of intervention on postural development suggest that intervention involving active trial and
error experience may accelerate postural development in typically developing infants and may improve postural control in children with
or at high risk for a developmental motor disorder.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Postural control; EMG; Development; Cerebral palsy

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1191
2. Basic principals in the organization of postural adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1192
2.1. Developmental changes in the first level of control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1192
2.2. Developmental changes in the second level of control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1193
2.3. Postural control in children with CP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1194
2.4. The effect of intervention on development of postural control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1195
3. Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1198
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1198
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1198

1. Introduction interfere with the activities of daily life (Brogren et al.,

2001; Van der Heide et al., 2004). However, little is
Dysfunctional postural control is one of the key known about the specific nature and development of the
problems in children with cerebral palsy (CP), which postural problems of children with CP. Such knowledge is
needed for the development of successful therapeutic
Corresponding author. Tel.: +31 50 3614247; fax: +31 50 3636905. interventions and to evaluate efficacy of therapeutic
E-mail address: [email protected] (M. Hadders-Algra). interventions.

0149-7634/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neubiorev.2007.04.008
 ARTICLE IN PRESS
1192 V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200

The current paper starts with a review of the develop- caudal-to-cranial sequence or in a reverse order, Hadders-
ment of postural control in typically developing children Algra, 2005).
and in children with CP. The review zooms in on EMG-
studies on muscle coordination during postural adjust- 2.1. Developmental changes in the first level of control
ments. Postural adjustments in children are assessed mainly
with the help of two paradigms: (1) studies using the Hedberg et al. (2004, 2005) suggested that the basic level
external perturbation paradigm in which postural adjust- of control of postural adjustments might have an innate
ments are recorded during a sudden movement of the origin. They evaluated postural muscle activity in response
support surface of the sitting or standing child and (2) to sudden perturbations in sitting position in infants from
studies assessing internally triggered postural adjustments 1 month onwards. The results revealed that at the age of
during reaching movements or locomotor activity. 1 month, direction-specific activity was present in 85% of
In the second part of the paper we review current the trials requesting activity of the dorsal postural muscles
knowledge on the effect of intervention on muscle and in 72% of the trials requiring activity of the ventral
coordination during postural development in children with muscles (Hedberg et al., 2004). Between 2 and 4–5 months
typical and atypical motor development. the rate of direction-specific adjustments remained 70–85%
to increase to a virtually consistent presence in infants aged
2. Basic principals in the organization of postural at least 7–8 months (Harbourne et al., 1993; Hadders-Algra
adjustments et al., 1996; Hedberg et al., 2005).
The early development of internally trigged postural
Postural control is organized to guarantee the main- adjustments has been addressed in EMG-studies during
tenance of equilibrium and to keep the projection of reaching in supine and sitting conditions. ‘Pre-reaching’
the centre of mass (COM) within the stability limits of movements in infants aged 1–3 months are accompanied by
the body (Massion, 1994). It involves the resistance to postural activity which in general lacks direction specificity
forces of gravity and mechanical support during movement (Van der Fits et al., 1999a). At 3–4 months ‘pre-reaching’
(Massion, 1998). activity is replaced by reaching movements towards an
The current paper focuses on the balance maintaining object and from 4 to 5 months reaching movements
aspect of postural control. For human beings keeping increasingly more often result in grasping of an object
balance in vertical position is a motor control problem of (Touwen, 1976). At the age of 4–6 months about 50% of
reputed complexity: a multi-joint body has to be kept the reaches towards an object in supine and sitting position
upright on a relatively small area of support with the help are accompanied by direction-specific activity in the dorsal
of many muscles. In terms of motor control, this means postural muscles (De Graaf-Peters et al., 2007). This means
that the nervous system is faced with a problem of that the presence of direction-specific postural activity is
redundancy in degrees of freedom. Bernstein (1935) not a pre-requisite for the generation of reaching move-
suggested that the adult nervous system solves this problem ments, but it affects the ‘success’ of reaching movements.
by creating motor synergies. This means amongst others Reaching movements accompanied by direction specific
that supraspinal control centres do not need to specify each adjustments end more often in successful touching or
single muscle contraction, but may use pre-structured grasping of a toy than reaching movements accompanied
neural commands, i.e. the repertoire of synergies embedded by postural adjustments lacking direction specificity. In the
in the spinal cord and brain stem. posturally more safe supine condition the success of
Forssberg and Hirschfeld (1994) suggested that in the reaching is independent of the presence of direction-specific
neural control of postural synergies two levels can be activity (De Graaf-Peters et al., 2007).
distinguished. The first level of control is involved in the Van der Heide et al. (2003) showed that children from
generation of basic direction-specific adjustments. Direc- the age of 2 years onwards consistently use direction-
tion specificity means that perturbations inducing a specific adjustments while reaching in sitting position. This
forward sway of the body, such as reaching movements, means that a consistent recruitment of direction-specific
are accompanied by postural activity in the muscles on the adjustments during reaching while sitting emerges at some
dorsal side of the body, whereas perturbations inducing a age between 6 months and 2 years.
backward body sway are accompanied by activity in the Until recently it was unclear whether direction-specific
ventral muscles (Forssberg and Hirschfeld, 1994). Func- activity was present during stance in young infants who
tional activity at the second level of control means were not yet able to stand independently. Sveistrup and
involvement in the fine-tuning of the basic postural pattern Woollacott (1996) had studied postural adjustments with
on the basis of multi-sensorial afferent input from the external perturbation paradigm in infants in the pull-
somatosensory, visual, and vestibular systems. The mod- to-stand phase during standing with support. The results
ulation of postural adjustments can be achieved in various indicated some, inconsistently present, direction-specific
ways, for instance, by changing the number of direction- activity. It remained unclear, however, whether the absence
specific muscles recruited, by modifying the order in which of consistent direction-specific activity could be attributed
the direction-specific muscles are recruited (e.g., in a to the infants’ developmental level of neural control or to
 ARTICLE IN PRESS
V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200 1193

the presence of support, as it is well known that the highest degree of glucose metabolism, which might serve as
presence of support alters postural activity (Cordo and an indicator of functional activity, is found in the primary
Nashner, 1982). Recently, Hedberg et al. (2006) managed sensory and motor cortex, cingulate cortex, thalamus,
to study some infants at a similar developmental stage aged brain stem, cerebellar vermis, and hippocampal region. At
8–10 months by means of external perturbations during 2–3 months of age glucose utilization increases in the
stance without support. The data showed that direction- parietal, temporal, and primary visual cortices, basal
specific activity could be found consistently at least at one ganglia and cerebellar hemispheres.
of the levels recorded, i.e. lower leg, upper leg and neck. After the age of 3–4 months the complete pattern more
But consistent direction specificity at all levels recorded was often occurs, but the muscle recruitment patterns still
first found at the age of 14 months, when all infants were remain rather variable. The increase of the occurrence of
able to stand independently. Previously it had already been the complete pattern has been reported for multiple
demonstrated that young children who are well able to conditions. For instance, at 6 months of age the occurrence
stand independently show consistent direction-specific of the complete pattern is already 50% during backward
postural activity during external perturbations in stance translations in sitting, 30% during forward translations
(Forssberg and Nashner, 1982; Woollacott et al., 1987; and about 50% during reaching movements in supine and
Sveistrup and Woollacott, 1996). sitting position (Hadders-Algra et al., 1996; De Graaf-
To summarize, already at young age, infants show Peters et al., 2007). After 6 months the use of the complete
direction-specific adjustments during lying supine and pattern further increases. At 9–10 months of age it is used
sitting. The extent to which direction-specific adjust- in 75% of perturbation trials causing a backward sway of
ments are found depends on the postural demands of the the body and in 100% of perturbation trials inducing
task. From the developmental age that infants can stand forward body sway (Hadders-Algra et al., 1996). A simi-
independently consistent direction-specific adjustments are larly high prevalence of the complete pattern during
found. reaching in sitting position is first observed at 15 months
of age (Hadders-Algra, 2005). When the child becomes
2.2. Developmental changes in the second level of control older the dominance of the complete pattern disappears. It
disappears during reaching while sitting between the age of
Direction-specific postural muscle activity at young age 18 months and 2 years and during external perturbations in
is characterized by variation (Hadders-Algra, 2005). sitting between 212 and 3 years (Hadders-Algra et al., 1998;
Variation is found in muscles which are recruited, in the Van der Heide et al., 2003).
temporal ordering of muscle activity, in antagonist In stance, the development of muscle recruitment
recruitment, and in the degree to which the postural patterns only has been evaluated in studies using the
muscles are contracted. However, within these variations, external perturbation paradigm. Infants at the verge of
developmental trends can be distinguished. independent stance show substantial variation in which
Prior to the emergence of reaching, the development of direction-specific muscles are activated. The variation
postural muscle activity only has been studied by means of includes the complete pattern, which in the study of
external perturbations in sitting infants. The external Hedberg et al. (2006) meant the activation of direction-
perturbation experiments revealed that infants at 1 month specific lower- and upper leg muscles and neck muscles.
of age use a variable repertoire of direction-specific A preference for the use of the complete pattern emerges at
adjustments. This means that the direction-specific muscles 2 years of age and lasts at least until the age of 10 years
are activated in any possible combination, including the so- (Forssberg and Nashner, 1982; Sveistrup and Woollacott,
called complete pattern during which all direction-specific 1996; Woollacott et al., 1998; Sundermier et al., 2001). It is
muscles are activated in concert (Hedberg et al., 2004). At even likely that the complete pattern remains the preference
1 month of age infants respond during 10–20% of the pattern into adulthood (Keshner et al., 1988; Aruin and
perturbations with the complete pattern. But with increas- Latash, 1995).
ing age the complete pattern is less often found, and The data indicate that the differences in ages at which
around 3 months, this pattern is virtually absent. Prechtl the complete pattern is used preferentially in the various
(1984) already indicated that the age of 3 months may be conditions, depend on differences in the nature of the
considered as a moment of major neurodevelopmental postural task: an external perturbation challenges balance
transition. The finding that postural control until the age of control more than a self-generated reaching movement and
3 months is not related to achievements in spontaneous during standing and walking balance control is challenged
motor behaviour in supine position whereas such a relation more than during sitting.
is present after 3 months, supports the notion of a Also the development of the recruitment order of the
transition (Hedberg et al., 2005; de Graaf-Peters et al., muscles involved in a postural adjustment is characterized
2006). In addition, the positron emission tomography by variation (Hadders-Algra et al., 1996; De Graaf-Peters
studies of Chugani (1998) indicated that around the age of et al., 2007). Infants initially develop a slight preference for
3 months substantial changes occur in the distribution of a top–down recruitment during which the neck muscle is
metabolic activity in the brain. In newborn infants the recruited first (Hadders-Algra et al., 1996; Van der Fits
 ARTICLE IN PRESS
1194 V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200

et al., 1999b; De Graaf-Peters et al., 2007). During which balance is threatened considerably, e.g. during rapid
reaching, the preference for top–down recruitment first backward displacements: adult subjects show in this
becomes stronger between 4 and 6 months (De Graaf- situation a co-activation of the lower leg muscles (Nashner
Peters et al., 2007). But infants of 8–10 months who sit and Cordo, 1981; Oddsson, 1989).
independently have a slight preference for a bottom–up Amplitude modulation of EMG activity is considered to
recruitment (Hadders-Algra et al., 1996; Van der Fits et al., be one of the most subtle forms of fine-tuning of postural
1999b). From pre-school age onwards recruitment order control (Van der Heide et al., 2003). The ability to
during external perturbations and reaching in sitting is modulate postural adjustments in this subtle way emerges
characterized by variation (Hadders-Algra et al., 1998; at 9–10 months. From that age infants are able to adapt
Brogren et al., 1998; Van der Heide et al., 2003). During postural muscle activity with respect to the velocity of a
reaching a gradual preference within the variation develops reaching arm movement and to the amount of forward or
for top–down recruitment. But it is only after puberty that backward rotation of the pelvis in sitting position
the top–down recruitment becomes the dominant recruit- (Hadders-Algra et al., 1996; Van der Fits et al., 1999a–c).
ment order during reaching (Van der Heide et al., 2003). Which muscle’s activity is modulated strongly depends on
During the initial phases of standing, postural adjust- the nature of the task. For instance, sitting children aged
ments are also characterized by variation in recruitment 3–7 years in particularly modulate the amplitude of the
order. However, the perturbation experiments showed that upper leg muscles during external perturbations inducing a
a preference for a bottom–up strategy emerges already backward body sway (Brogren et al., 2001). During
between 10 and 12 months (Hedberg et al., 2006). The reaching while sitting pre-school and school-aged children
preference for bottom–up recruitment order persists into do not show a clear preference to modulate the degree of
adult life, in particular during external perturbations contraction of a specific direction-specific muscle. It is only
(Nashner, 1976; Nashner et al., 1979; Forssberg and after puberty that a preference to modulate the amplitude
Nashner, 1982; Sveistrup and Woollacott, 1996; Woollacott of the more cranially located muscles emerges (Van der
et al., 1998; Sundermier et al., 2001). In this respect it is Heide et al., 2003).
interesting to note that adults during self-produced arm Perturbation experiments during standing revealed that
movements show a considerable variation in recruitment from the age of independent stance children are able to
order (Cordo and Nashner, 1982; Aruin and Latash, 1995). modulate EMG amplitude with respect to the size of the
The data above indicate that the capacity to fine-tune perturbation (Roncesvalles et al., 2004). Like in the sitting
postural activity to task constraints in terms of recruitment position, the way in which the degree of contraction of the
order emerges between 4 and 6 months of age. In addition direction-specific muscles is modulated strongly depends
they underline that recruitment order during childhood is on the nature of the postural task and the age of the child.
characterized by variation. Whether or not a child adopts a Berger et al. (1992, 1995) showed that in the age period
preference for a specific form of recruitment depends in between 2 and 11 years the degree to which lower leg
particular on the child’s age and the nature of the postural muscle EMG amplitude is modulated during external
challenge. perturbations decreases with increasing age. They also
Remarkably, the development of postural adjustments in reported that the activity of the direction-specific tibialis
sitting position is characterized by the absence of anterior was modulated to a larger extent than that of the
antagonistic co-activity (Hadders-Algra et al., 1996; Van direction-specific gastrocnemius muscle.
der Fits et al., 1999a; Hedberg et al., 2005). Only between 6 Summarizing, the development of the second level of
and 24 months some antagonistic activity is observed in control is complex and is characterized by variation. The
specific situations: during external perturbations inducing a developmental timing of the various aspects of the fine-
backward sway of the body (Hadders-Algra et al., 1996, tuning of postural control depends on the difficulty of the
1998) and in the neck muscles during reaching (Van der postural task. The fine-tuning of postural control runs a
Heide et al., 2003). protracted course: it first reaches an adult level of control
During the development of standing the situation is after adolescence.
different. Perturbation experiments indicated that during
the earliest phases of stance development some antagonistic 2.3. Postural control in children with CP
activity may be observed (Sveistrup and Woollacott, 1996;
Hedberg et al., 2006). Children aged 11225 years often Nowadays CP is described as ‘a group of disorders of the
exhibit antagonistic co-activation in the leg muscles when development of movement and posture, causing limitation
balance is disturbed by an external force (Forssberg and in activity, that are attributed to non-progressive dis-
Nashner, 1982; Berger et al., 1992, 1995). Beyond the age turbances that occurred in the developing foetal or infant
of about 5 years the antagonistic muscles no longer are co- brain. The motor disorders of CP are often accompanied
activated with the agonist, but they are recruited in a reci- by disturbances of sensation, cognition, communication,
procal manner (Forssberg and Nashner, 1982; Sundermier perception, and/or behaviour, and/or by a seizure disorder’
et al., 2001). Antagonistic co-activation may be found (Bax et al., 2005). The definition underscores the notion
beyond the age of 5 years, but only in situations during that there is large variation amongst children diagnosed
 ARTICLE IN PRESS
V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200 1195

with CP. Children with CP invariably show deficits in top–down recruitment strategy might reflect that head
development of postural control. The exact nature of the stabilization in space is a major goal of postural control
deficits is only known to a limited extent. Most studies on (Pozzo et al., 1990).
postural muscle activity in children with CP assessed Children with CP show a high amount of antagonistic
relatively small groups of children, dealt with school-age co-activation during perturbation experiments in sitting
only, and only included children with mild to moderate and standing position (Brogren et al., 1998, 2001;
forms of the disorder. Current knowledge is outlined Woollacott et al., 1998, 2005). In sitting position co-
below. activation is especially high during perturbations inducing
Only one study addressed the development of postural a backward body sway (Brogren et al., 1998, 2001). During
control of children with CP during infancy (Hadders-Algra perturbations inducing a forward sway of the body little
et al., 1999a). The study consisted of a longitudinal antagonistic co-activation is found. This could be related to
assessment of postural control during reaching between 4 a higher stability in this situation induced by the
and 18 months in five infants with a spastic hemiplegia and configuration of the sitting body and to more experience
two infants with severe bilateral spastic CP. The infants as forward body sway is frequently used in daily activities
with spastic hemiplegia showed direction-specific adjust- such as reaching (Brogren et al., 1998, 2001). Antagonistic
ments from 15 months onwards. Unlike typically develop- co-activation is rarely found in children with CP during
ing infants they did not develop the ability to modulate reaching in a sitting position (Van der Heide et al., 2004).
EMG-amplitude to the velocity of the reaching arm or to The major postural dysfunction of children with CP is
initial pelvis position in the age period till 18 months. the substantially reduced capacity to modulate the degree
Postural development of one of the infants with bilateral of postural muscle contraction to the specifics of the
spastic CP resembled that of the infants with hemiplegia, situation (Brogren et al., 2001; Van der Heide et al., 2004;
but it proceeded at a slower pace. Postural control of the Woollacott et al., 2005). Children with CP have for
other infant with bilateral spastic CP, who also showed instance difficulties in using information stemming from
signs of dyskinesia and who was not able to sit at the age of the initial body configuration to adapt postural activity
4 years, was severely disorganised: she lacked direction- during reaching while sitting (Van der Heide et al., 2004).
specific postural adjustments and was not able to adjust Children with spastic hemiplegia are able to use the
postural activity to task-specific conditions. information of the body configuration to some extent to
Also beyond infancy children with CP in general show modulate postural activity during reaching; children with
direction-specific postural activity, both during sitting and bilateral spastic CP lack this capacity entirely (Van der
standing (Nashner et al., 1983; Burtner et al., 1998; Heide et al., 2004). Children with CP are able to use the
Woollacott et al., 1998; Brogren et al., 1998, 2001; Van information originating from the reaching arm, including
der Heide et al., 2004). However, mild to moderate the velocity of the arm to adjust the degree of contraction
problems to recruit direction-specific activity may occur of the direction-specific postural muscles (Van der Heide
in particular in the leg muscles (Van der Heide et al., 2004; et al., 2004).
Woollacott et al., 2005). Only children with severe CP not The above data indicate that children with severe forms
able to sit independently show a total lack of direction- of CP, i.e. children who are not able to sit independently by
specific postural adjustments (Hadders-Algra et al., the age of 112 years, are hampered by serious dysfunction of
1999a, b). Pre-school and school-age children with CP the basic level of postural control. In children with less
always show dysfunctions in the fine tuning of the postural severe forms of CP the first level of control is basically
adjustment, i.e. an invariable recruitment order, an intact. At the second level of control, multiple forms of
excessive degree of antagonistic co-activation during disorganization and/or adaptation are found: a dominance
external perturbations, and a reduced capacity to modulate of cranio-caudal recruitment, an increased degree of
postural adjustments (Van der Heide et al., 2004, Brogren antagonistic co-activation and a reduced or absent capacity
Carlberg and Hadders-Algra, 2005). to adapt the degree of muscle contraction to the specifics of
A strong preference for a top–down recruitment of the the situation. The extent to which these problems are
postural muscles in children with CP is found, not only present depends on the postural challenge of the situation.
during perturbation experiments in sitting and standing
position, but also during reaching in a sitting position 2.4. The effect of intervention on development of postural
(Nashner et al., 1983; Brogren et al., 1998; Woollacott control
et al., 1998; Van der Heide et al., 2004). A clear cranio-
caudal recruitment involves an early recruitment of the Paediatric physiotherapeutic intervention in children
neck extensor muscle. The fact that this strategy more often with or at high risk for developmental motor disorders
occurs in children with mild to moderate forms of CP than often aims at improving postural control. However, only
in children with severe CP, might indicate that the limited information is available on the effect of interven-
preference for cranio-caudal recruitment reflects the child’s tion on postural muscle activity.
strategy to cope with deficient postural control (Latash and Two studies addressed the effect of training on the
Anson, 1996; Van der Heide et al., 2004). In addition, the development of postural control in typically developing
 ARTICLE IN PRESS
1196 V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200

infants. Hadders-Algra et al. (1996) studied the effects of adjustments, an increase of the use of the complete pattern,
balance training on postural control in sitting during and of distal to proximal muscle recruitment. This means
infancy. At the age of 5–6 months 20 infants were that balance training in typically developing infants may
randomly assigned to a group which obtained 3 months result in an acceleration of postural development.
of daily balance training or a control group without One study addressed the effect of balance training in
balance training. Balance training was carried out at home children with CP (Shumway-Cook et al., 2003; Woollacott
by family members and consisted of playful presentation of et al., 2005). Woollacott and co-workers trained balance in
toys in the sideward and semi-backward direction in the four children with bilateral spastic CP and two children
border zone of sitting without falling. The balance play was with spastic hemiplegia aged 7–13 years by means of
carried out three times a day during 5 min. Postural control exposure to 100 balance perturbations/day for a period of 5
was assessed by means of external perturbations in sitting days. The training resulted in improvements in the
position at baseline, at the age of 7–8 months and once direction specificity of the adjustments, a faster recruitment
again at 9–10 months. The data revealed a significant effect of the postural muscles, a more frequent use of a
of training on the organization of postural activity at the bottom–up recruitment and an improved ability to
second level of control: the trained infants showed a clearly modulate the degree of muscle contraction (Woollacott
accelerated development of the ability to select the et al., 2005). The improvements in postural muscle activity
complete pattern out of their repertoire and of the ability were associated with a reduction of postural sway (Shum-
to modulate the degree of ventral muscle contraction to the way-Cook et al., 2003).
velocity of the moving platform and to body configuration Recently we evaluated the effect of two forms of early
at perturbation onset. The other study was carried out by intervention on postural development in 20 infants at high
Sveistrup and Woollacott (1997). They studied the effect of risk for developmental motor disorders such as CP. Infants
exposure to 100 balance perturbations/day during three with definitely abnormal general movements at the
consecutive days in 15 infants in the pull-to-stand phase corrected age of 3 months were randomly assigned to
aged 9–11 months. The study indicated that training either traditional paediatric physiotherapeutic intervention
resulted in an increase of the rate of direction-specific (n ¼ 11) or a novel intervention program called COPCA

Fig. 1. Effect of two types of intervention in infants at high risk for developmental motor disorders on postural control. Typical examples of postural
activity during reaching in sitting position at 6 months in an infant who had received traditional paediatric physiotherapy (TPP; left panel), an infant who
had received COPCA intervention (middle) and a typically developing (TD) infant (right panel). Duration of each trial is 2500 ms. DE ¼ Deltoid,
PM ¼ Pectoralis Major, BB ¼ Biceps brachii, TB ¼ Triceps brachii, NF ¼ neck flexor, NE ¼ neck extensor, RA ¼ rectus abdominis, TE ¼ thoracal
extensor, LE ¼ lumbar extensor, RF ¼ rectus femoris, HAM ¼ hamstrings. Dotted vertical lines denote the onset of the reaching movement as indicated
by simultaneously recorded kinematics. Horizontal lines delineate the presence (onset and offset) of significant EMG bursts as defined by the computer
algorithm. During the reaching movement of the TD infant (right) postural activity in neck and trunk is direction specific: NE starts prior to NF; TE and
LE are recruited, the antagonistic RA not. Postural activity in the infants who received COPCA and TPP intervention is also direction specific. However
note the synchronous recruitment of the postural muscles shortly after the onset of reaching in the infant of the TPP group.
 ARTICLE IN PRESS
V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200 1197

(n ¼ 9). COPCA (Coping and caring for infants with months postural control in the group of infants who had
neurological dysfunction—a family centred programme; received COPCA intervention was in some aspects
see Blauw-Hospers et al., 2007, this issue) is based on the significantly better than that of the group who had received
motor developmental principles of the Neuronal Group traditional paediatric physiotherapy (Figs. 1–4). The
Selection Theory (NGST; Edelman, 1989; Hadders-Algra, infants who had received COPCA intervention showed
2000a, b) and on new insights in the field of education and more direction-specific adjustments in sitting (Fig. 2), more
family care (Dale, 1996; Rosenbaum et al., 1998). Tradi- often recruited the complete pattern in supine and sitting
tional paediatric physiotherapy in the Netherlands usually position (Fig. 3) and showed significantly less often a
consists of NeuroDevelopmental Treatment (NDT). Inter- synchronous onset of postural muscle activity (Figs. 1 and
vention was provided for a period of 3 months. Postural 4). Postural performance at 6 months of the high-risk
activity was assessed at 4 and 6 months during reaching infants who had received COPCA intervention had
while lying supine and while sitting with support. At the improved to such an extent that it closely resembled that
age of 4 months both groups of at risk infants showed less of typically developing peers.
direction-specific activity in supine and sitting than age- The two studies on the effect of intervention on postural
matched typically developing infants. At the age of 6 development in children with or at risk for a developmental

Fig. 3. Effect of two types of intervention in infants at high risk for

developmental motor disorders on individual developmental trajectories
of preference patterns between 4 and 6 months, including reference data of
Fig. 2. Effect of two types of intervention in infants at high risk for age-matched typically developing (TD) infants (n ¼ 12). TPP ¼ group
developmental motor disorders on frequency of direction-specific adjust- which received traditional paediatric physiotherapy (n ¼ 11), COP-
ments at 4 and 6 months in supine and sitting position, including reference CA ¼ group which received COPCA intervention (n ¼ 9). The preference
values of age-matched typically developing (TD) infants (n ¼ 12). The pattern is defined as the pattern present in at least 50% of the trials. Each
boxes represent median values. TPP ¼ group which received traditional line represents the development of one infant. N ¼ Neck extensor,
paediatric physiotherapy (n ¼ 11), COPCA ¼ group which received T ¼ Thoracal extensor, L ¼ Lumbar extensor. X indicates participation
COPCA intervention (n ¼ 9). Mann Whitney: **pp0.01, *po0.05. Not of a direction-specific muscle in a pattern. Three Xs represent the complete
indicated with a symbol: significant increase in direction specificity in pattern. Differences in the preference of the complete pattern at 6 months
sitting COPCA infants between 4 and 6 months (Wilcoxon p ¼ 0.04). between the various groups: Mann Whitney: **pp0.01, *po0.05.
 ARTICLE IN PRESS
1198 V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200

3. Concluding remarks

During typical development the basic level of postural

control is functionally active from early infancy onwards.
This means that already at early post-natal age the infant
possesses a repertoire of direction-specific postural adjust-
ments. Whether or not direction-specific adjustments are
used depends on the child’s age and the nature of the
postural task. After the age of major neurodevelopmental
transition at 3 months the capacity to adapt postural
activity to environmental constraints emerges. Yet, it takes
at least till adolescence before an adult type of postural
adaptation has been achieved.
Children with CP in general have the ability to generate
direction-specific adjustments, but they show a delayed
development in the capacity to recruit direction-specific
adjustments in tasks with a mild postural challenge, such as
reaching while sitting. Children with CP virtually always
have difficulties in the fine-tuning of postural activity.
The few studies available on the effect of intervention on
postural development indicate the following. Intervention
by means of balance training accelerates the development
of postural control in typically developing infants. There is
also some evidence that balance training with active trial
and error experience may improve postural control in
children with or at high risk for a developmental motor
disorder. Intervention by means of NDT does not result in
improved postural control.

Acknowledgements

We kindly acknowledge the contribution of our colla-

borators Kune Blaauw PT, Lily van Doormaal MD, Ellen
Janssen PT, Peter Logtenberg PT and Wilma Renkema PT
Fig. 4. Effect of two types of intervention in infants at high risk for MSc to the study on the effect of intervention in infants at
developmental motor disorders on postural control on the frequency of high risk for motor developmental disorders. VBdG-P was
synchronous postural muscle activity at 4 and 6 months in supine and
sitting position, including reference data of age-matched typically
financially supported by the Johanna KinderFonds,
developing (TD) infants (n ¼ 12). TPP ¼ group which received traditional Stichting Fonds de Gavere, the Cornelia Stichting and
paediatric physiotherapy (n ¼ 11), COPCA ¼ group which received the Graduate School for Behavioural and Cognitive
COPCA intervention (n ¼ 9). Synchronous postural muscle activity was Neurosciences (BCN).
defined as present when at least two of the five neck and trunk muscles had
been recruited in the narrow time window ending at 80 ms after the onset
of the prime mover, i.e. the arm muscle recruited first. The boxes represent References
median values. Between group differences: Mann Whitney: **pp0.01,
*po0.05. Within group differences: Wilcoxon: ]po0.01. Differences Aruin, A.S., Latash, M.L., 1995. The role of motor action in anticipatory
between TD group and intervention groups at 4 months not indicated with postural adjustments studied with self-induced and externally triggered
symbols: Mann Whitney: TPP-supine: p ¼ 0.001, TPP-sitting: p ¼ 0.04, perturbations. Experimental Brain Research 106, 291–300.
COPCA-supine: p ¼ 0.002, COPCA-sitting: p ¼ 0.001. Bax, M., Goldstein, M., Rosenbaum, P., Leviton, A., Paneth, N., Dan, B.,
Jacobsson, B., Damiano, D., 2005. Executive committee for the
definition of cerebral palsy. Proposed definition and classification of
cerebral palsy. Developmental Medicine and Child Neurology 47,
571–576.
motor disorder indicate that intervention which requires Berger, W., Discher, M., Trippel, M., Ibrahim, I.K., Dietz, V., 1992.
active trial and self-produced and variation error experi- Developmental aspects of stance regulation, compensation and
ence is able to improve postural control. The data also adaptation. Experimental Brain Research 90, 610–619.
suggest that the more traditional approach of intervention, Berger, W., Trippel, M., Assaiante, C., Zijlstra, W., Dietz, V., 1995.
Developmental aspects of equilibrium control during stance: a
such as NDT, which involves a substantial amount of kinematic and EMG study. Gait and Posture 3, 149–155.
handling and provision of postural support, is not effective Bernstein, N., 1935. The problem of the interrelation of coordination and
in improving postural development. localisation. Archives of Biological Sciences 38, 15–59.Reprinted in
 ARTICLE IN PRESS
V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200 1199

Bernstein, N., 1967. The Coordination and Regulation of Movements. Harbourne, R.T., Giuliani, C., MacNeela, J., 1993. A kinematic and
Pergamon Press, Oxford. electromyographic analysis of the development of sitting posture in
Blauw-Hospers, C.H., de Graaf-Peters, V.B., Dirks, T., Bos, A.F., infants. Developmental Psychobiology 26, 51–64.
Hadders-Algra, M., 2007. Does early intervention in infants at high Hedberg, Å., Forssberg, H., Hadders-Algra, M., 2004. Early development
risk for a developmental motor disorder improve motor and cognitive of postural adjustments in sitting position: evidence for the innate
development? Neuroscience and Biobehavioral Reviews, this issue, origin of direction specificity. Experimental Brain Research 157, 10–17.
doi:10.1016/j.neubiorev.2007.04.010. Hedberg, Å., Brogren-Carlberg, E., Forssberg, H., Hadders-Algra, M.,
Brogren Carlberg, E., Hadders-Algra, M., 2005. Postural dysfunction in 2005. Development of postural adjustments in sitting position during
children with cerebral palsy: some implications for management. the first half year of life. Developmental Medicine and Child
Neural Plasticity 12, 149–158. Neurology 47, 312–320.
Brogren, E., Hadders-Algra, M., Forssberg, H., 1998. Postural control in Hedberg, Å., Schmitz, C., Forssberg, H., Hadders-Algra, M., 2006. Early
sitting children with cerebral palsy. Neuroscience and Biobehavioral development of postural adjustments in standing with and without
Reviews 22, 591–596. support. Experimental Brain Research, November 9; epub. ahead of
Brogren, E., Forssberg, H., Hadders-Algra, M., 2001. Influence of two print.
different sitting positions on postural adjustments in children with Keshner, E.A., Woollacott, M.H., Debu, B., 1988. Neck, trunk and limb
spastic diplegia. Developmental Medicine and Child Neurology 43, muscle responses during postural perturbations in humans. Experi-
534–546. mental Brain Research 71, 455–466.
Burtner, P.A., Qualls, C., Woollacott, M.H., 1998. Muscle activation Latash, M.L., Anson, J.G., 1996. What are ‘‘normal movements’’ in
characteristics of stance balance control in children with spastic atypical populations? Behavioral and Brain Sciences 19, 55–68.
cerebral palsy. Gait Posture 8, 163–174. Massion, J., 1994. Postural control system. Current Opinion in
Chugani, H.T., 1998. A critical period of brain development: studies of Neurobiology 4, 877–887.
cerebral glucose utilization with PET. Preventive Medicine 27, Massion, J., 1998. Postural control systems in developmental perspective.
184–188. Neuroscience and Biobehavioral Reviews 22, 465–472.
Cordo, P.J., Nashner, L.M., 1982. Properties of postural adjustments Nashner, L.M., 1976. Adopting reflexes controlling the human posture.
associated with rapid arm movements. Journal of Neurophysiology 47, Experimental Brain Research 26, 59–72.
287–302. Nashner, L.M., Cordo, P.J., 1981. Relation of postural responses and
Dale, N., 1996. Working with Families of Children with Special Needs. reaction-time voluntary movements in human leg muscles. Experi-
Routledge, London. mental Brain Research 43, 395–405.
De Graaf-Peters, V.B., De Groot-Hornstra, A.H., Dirks, T., Hadders- Nashner, L.M., Woollacott, M., Tuma, G., 1979. Organization of rapid
Algra, M., 2006. Specific postural support promotes variation in motor response to postural and locomotor-like perturbations of standing
behaviour of infants with minor neurological dysfunction. Develop- man. Experimental Brain Research 36, 463–476.
mental Medicine and Child Neurology 48, 966–972. Nashner, L.M., Shumway-Cook, A., Marin, O., 1983. Stance posture
De Graaf-Peters, V.B., Bakker, J., Van Eykern, L.A., Otten, E., control in select groups of children with cerebral palsy: deficits in
Hadders-Algra, M., 2007. Postural adjustments and reaching in sensory organization and muscular coordination. Experimental Brain
4- and 6-months-old infants: an EMG and kinematical study. Research 49, 393–409.
Experimental Brain Research, in press, doi:10.1007/S00221-007- Oddsson, L., 1989. Motor patterns of a fast voluntary postural task in
0964-6. man: trunk extension in standing. Acta Physiologica Scandinavica 136,
Edelman, G.M., 1989. Neural Darwinism. The Theory of Neuronal 47–58.
Group Selection. Oxford University Press, Oxford. Pozzo, T., Berthoz, A., Lefort, L., 1990. Head stabilisation during various
Forssberg, H., Hirschfeld, H., 1994. Postural adjustments in sitting locomotor tasks in humans. I. Normal subjects. Experimental Brain
humans following external perturbations: muscle activity and kine- Research 82, 97–106.
matics. Experimental Brain Research 97, 515–527. Prechtl, H.F.R., 1984. Continuity of neural functions from prenatal to
Forssberg, H., Nashner, L.M., 1982. Ontogenetic development of postural postnatal life. In: Clinical and Developmental Medicine, vol. 94,
control in man: adaptation to altered support and visual conditions Blackwell Scientific Publications, Oxford.
during stance. Journal of Neurosciences 2, 545–552. Roncesvalles, M.N., Woollacott, M.H., Brown, N., Jensen, J.L., 2004. An
Hadders-Algra, M., 2000a. The neuronal group selection theory: a emerging postural response: is control of the hip possible in the newly
framework to explain variation in normal motor development. walking child? Journal of Motor Behavior 36, 147–159.
Developmental Medicine and Child Neurology 42, 566–572. Rosenbaum, P., King, S., Law, M., King, G., Evans, J., 1998. Family
Hadders-Algra, M., 2000b. The neuronal group selection theory: centered service: a conceptual framework and research review. Physical
promising principles for understanding and treating developmental Occupational and Therapeutic Pediatrics 18, 1–20.
motor disorders. Developmental Medicine and Child Neurology 42, Shumway-Cook, A., Hutchinson, S., Kartin, D., Price, R., Woollacott,
707–715. M., 2003. Effect of balance training on recovery of stability in children
Hadders-Algra, M., 2005. Development of postural control during the first with cerebral palsy. Developmental Medicine and Child Neurology 45,
18 months of life. Neural Plasticity 12, 99–108. 591–602.
Hadders-Algra, M., Brogren, E., Forssberg, H., 1996. Training affects the Sundermier, L., Woollacott, M., Roncesvalles, N., Jensen, J., 2001. The
development of postural adjustments in sitting infants. Journal of development of balance control in children: comparisons of EMG and
Physiology 493, 289–298. kinetic variables and chronological and developmental groupings.
Hadders-Algra, M., Brogren, E., Forssberg, H., 1998. Postural Experimental Brain Research 136, 340–350.
adjustments during sitting at pre-school age: presence of a transient Sveistrup, H., Woollacott, M., 1996. A longitudinal study of the
toddling phase. Developmental Medicine and Child Neurology 40, development of postural control in infants. Journal of Motor Behavior
436–447. 28, 58–70.
Hadders-Algra, M., Van der Fits, I.B.M., Stremmelaar, E.F., Touwen, Sveistrup, H., Woollacott, M., 1997. Practice modifies the developing
B.C.L., 1999a. Development of postural adjustments during reaching automatic postural response? Experimental Brain Research 114,
in infants with cerebral palsy. Developmental Medicine and Child 33–43.
Neurology 41, 766–776. Touwen, B., 1976. Neurological Development in Infancy. Clinics in Deve-
Hadders-Algra, M., Brogren, E., Katz-Salamon, M., Forssberg, H., lopmental Medicine No. 58. Heinemann Medical Publications, London.
1999b. Periventricular leukomalacia and preterm birth have a Van der Fits, I.B.M., Klip, A.W.J., van Eykern, L.A., Hadders-Algra, M.,
detrimental effect on postural adjustments. Brain 122, 727–740. 1999a. Postural adjustments during spontaneous and goal directed arm
 ARTICLE IN PRESS
1200 V.B. de Graaf-Peters et al. / Neuroscience and Biobehavioral Reviews 31 (2007) 1191–1200

movement in the first half year of life. Behavioral Brain Research 106, reaching in preterm children with cerebral palsy. Developmental
75–90. Medicine and Child Neurology 46, 253–266.
Van der Fits, I.B.M., Otten, E., Klip, A.W.J., Van Eykern, L.A., Hadders- Woollacott, M., Debû, B., Mowatt, M., 1987. Neuromuscular control of
Algra, M., 1999b. The development of postural adjustments during posture in the infant and child: is vision dominant? Journal of Motor
reaching in 6–18-month-old infants: evidence for two transitions. Behavior 19, 167–186.
Experimental Brain Research 126, 517–528. Woollacott, M., Burtner, P., Jensen, J., Jasiewicz, J., Roncesvalles, N.,
Van der Fits, I.B.M., Flikweert, E.R., Stremmelaar, E.F., Martijn, A., Sveistrup, H., 1998. Development of postural responses during
Hadders-Algra, M., 1999c. Development of postural adjustments standing in healthy children and in children with spastic diplegia.
during reaching in preterm infants. Pediatric Research 46, 1–7. Neuroscience and Biobehavioral Reviews 22, 583–589.
Van der Heide, J.C., Otten, E., van Eykern, L.A., Hadders-Algra, M., Woollacott, M., Shumway-Cook, A., Hutchinson, S., Ciol, M., Price, R.,
2003. Development of postural adjustments during reaching in sitting Kartin, K., 2005. Effect of balance training on muscle activity
children. Experimental Brain Research 151, 32–45. used in recovery of stability in children with cerebral palsy: a
Van der Heide, J.C., Begeer, C., Fock, A., Otten, B., Stremmelaar, E.F., pilot study. Developmental Medicine and Child Neurology 47,
van Eykern, L.A., Hadders-Algra, M., 2004. Postural control during 455–461.