Effect of balance training on muscle activity ... - Wiley Online Library

Effect of balance training on muscle activity used in recovery of stability in children with cerebral palsy: a pilot study Marjorie Woollacott* PhD, Human Physiology, University of Oregon, Eugene, OR; Anne Shumway-Cook PhD PT; Susan Hutchinson MS PT, Division of Physical Therapy; Marcia Ciol PhD; Robert Price MSME; Deborah Kartin PhD PT, Division of Physical Therapy, Department of Rehabilitation Medicine, School of Medicine, University of Washington, Seattle, WA, USA. *Correspondence to first author at Department of Human Physiology, University of Oregon, Eugene, OR 97403, USA. E-mail: [email protected]

This study explored possible neural mechanisms that contribute to improvements in balance control produced by reactive balance training in children with cerebral palsy (CP). Six children with CP (four males, two females; mean age 9y 4mo), two with spastic hemiplegia (Gross Motor Function Classification System [GMFCS] level I) and four with spastic diplegia (GMFCS level II,) were given 5 days of intensive training in reactive balance control (100 perturbations per day on a moveable force platform). Surface electromyography was used to characterize changes in neuromuscular responses pretraining, immediately posttraining, and 1 month posttraining. Training in reactive balance control resulted in improvements in directional specificity of responses (a basic level of response organization) and other spatial/temporal characteristics including: (1) faster activation of muscle contraction after training, allowing children to recover stability faster; (2) emergence of a distal–proximal muscle sequence; and (3) improved ability to modulate the amplitude of muscle activity (increased amplitude of agonist and decreased amplitude of antagonist, reducing coactivation). Each child with spastic hemiplegia or diplegia showed a different combination of factors that contributed to improved performance; the level of change in neural factors depended on the severity of involvement of the child: hemiplegia vs diplegia, and level of involvement within each diagnostic category.

See end of paper for list of abbreviations.

Research on skill development in children with cerebral palsy (CP) has documented constraints in the performance of voluntary skills, including manipulatory and mobility skills. These limitations are accompanied by underlying constraints in postural function (Nashner et al. 1983; Brogren et al. 1996, 1998; Burtner et al. 1998; Hadders-Algra et al. 1999). Constraints in locomotor and postural skills include delays in the onset of independent standing and walking, poor levels of performance of these tasks, or an inability to stand and walk, depending on the severity of the disorder (Crothers and Paine 1988). Impaired postural control in children with CP results from multiple factors. Studies of children with both spastic hemiplegia and diplegia using moving platform posturography (Nashner et al. 1983; Brogren et al. 1996, 1998; Burtner et al. 1998) have shown disruptions in the spatial and temporal aspects of responses of postural muscles during the recovery of stability after an unexpected perturbation. Such disruptions include: (1) poor directional specificity, with antagonists activated before agonists (considered a basic level of postural control); (2) additional problems with modulation in the temporal and spatial domain, including delayed onset of muscle activity, poor sequencing within a muscle synergy (reduction in distal–proximal muscle activation), high levels of agonist–antagonist muscle coactivation (Nashner et al. 1983; Brogren, et al. 1996, 1998; Burtner et al. 1998), reduced contraction amplitudes, and the inability to modulate amplitudes in response to changing perturbation characteristics (Roncesvalles et al. 2002). Clinical research (Seeger et al. 1984, McCleneghan et al. 1992) has also shown that children with CP show improvement in manipulation skills when given postural support during sitting or standing. In addition, although in seated children a crouched or rounded trunk posture gives better responses than an erect posture, in standing children it has been shown that a crouched posture contributes to abnormal balance and gait (Crenna 1998, Woollacott et al. 1998, Brogren et al. 2001). Studies on standing balance suggest that constraints on stability contribute to limitations in the performance of manipulation and mobility skills. Because a relation between constraints in balance control and functional limitations has been established, increased efficacy of postural control should be a goal of intervention programs (Pape et al. 1993, Hur 1995, Burtner 1998, Park et al. 2001). However, there is limited evidence to show the efficacy of training programs on improving stability or of neural changes contributing to posttraining improvements. Sveistrup and Woollacott (1997) demonstrated that three days of intensive balance training on a moveable platform (100 perturbations per day) improved the ability of typically developing children, who were in the process of learning independent stance skills, to recover stability after external balance threats. The children showed improvements in the organization and level of involvement of postural muscles in balance recovery. Shumway-Cook et al. (2003) reported that training reactive balance control in children with CP resulted in less center of pressure (COP) displacement and a shorter time taken for balance recovery after a balance threat, with improvements maintained one month posttraining. The purpose of this study was to explore neural mechanisms that could explain the training-induced improvements

Developmental Medicine & Child Neurology 2005, 47: 455–461 455

in balance control in children with CP. Postulated changes included: (1) improved directional specificity (considered the most basic level of balance performance); (2) additional improvements in spatial or temporal characteristics of muscle responses, or both, including faster responses, reducing agonist–antagonist coactivation, improved distal–proximal response sequencing, and improved modulation of contraction amplitudes, including reduced amplitudes due to faster response onsets. Method Six children with CP ( four males, two females; mean age 9y 4mo; four diagnosed with spastic diplegia; two with right spastic hemiplegia) were recruited from local schools and children’s therapy programs in Seattle, WA, USA. Inclusion criteria were: (1) the ability to stand independently for 30 or more seconds; (2) the ability to walk independently with or without an assistive device; (3) to have no uncorrected vision or hearing impairments; and (4) the ability to understand experimental procedures and give informed assent. Once informed consent from parents and informed assent from the children had been obtained, the Gross Motor Function Measure (GMFM; Russell et al. 1993) was performed on each child with CP. Information from the GMFM was used to classify children using the Gross Motor Function Classification System (GMFCS; Palisano et al. 1997). The children with diplegia were rated as level II and the children with hemiplegia were rated as level I on the GMFCS. Table I summarizes participants’ clinical characteristics. Approval for this study was obtained from the University of Washington and University of Oregon Human Subjects Review Committees. A moveable forceplate system (NeuroCom International, Inc, Clackamas, OR, USA) was used to test and train reactive balance control (Shumway-Cook et al. 2003). A single research assistant and a licensed physical therapist both tested and trained the participants. Surface electromyogram (EMG) data were collected from the Noraxon Myosystem 1200 (Scottsdale, AZ, USA; sampling rate 1000Hz) using a Qualisys motion

capture system (Columbiaville, MI, USA) with analog capture capability. Pairs of bipolar surface electrodes (Medicotest N50-K; Medicotest, Olstkke, Denmark) were placed over the gastrocnemius muscle (G) and the tibialis anterior (TA) bilaterally, and over the hamstrings and quadriceps muscles of the more involved leg. As electrodes are repositioned in each recording session, there is controversy about the best way to measure and compare EMG amplitudes under these conditions. Some authors advocate normalizing EMG by noting EMG responses as a percentage of maximal contraction amplitudes. However, this does not take into account changes in actual muscle output from one collection session to another. Alternatively, we use a method in which response amplitude is reported as a ratio of the background level of activity before either the stimulus or muscle response. We analyzed EMG responses by using custom software, with onset of phasic muscle responses defined as the time between onset of platform movement and onset of an EMG response. A single-subject replicated-multiple-baseline experimental design was used; a detailed description of the intervention study may be found in Shumway-Cook et al. (2003). We recorded muscle activity used to recover stability after five forward and five backward platform perturbations pretraining, immediately posttraining, and 1 month posttraining. During all sessions, children stood barefoot on the platform, secured to an overhead harness, and were guarded by an assistant for safety. Instructions were to try to remain standing in one place. Training consisted of 5 days of platform perturbation sessions (one hundred 3 to 6cm amplitude forward and backward perturbations at 12 to 24cm/s velocity). Children watched a videotape during the training and were given rest breaks every 20 to 25 perturbations. Each child’s data were analyzed in detail to determine differences in individual strategies associated with training. Formal statistical hypothesis testing relying on aggregated group data was not deemed appropriate because of the wide range of strategies found among the children (Thelen et al. 1993).

Table I: Clinical characteristics of participants Patient no.

Sex Gestational Age age (wks) (y:m)

Diagnosis GMFCS level

SD1

M

28 12:11

Spastic diplegia

SD2

M

26

7:5

Spastic diplegia

SD3

F

25

9:8

Spastic diplegia

SD4

M

28

7:8

Spastic diplegia

SH1

F

41

7:10

Right hemiplegia

SH2

M

39

10:4

Left hemiplegia

Prior medical procedures

Current therapy

Plantarflexors, Serial casting hamstrings bilaterally II Hamstring tightness Baclofen; botox; (left more than right) serial casting II Plantarflexors, Bilateral shunts hamstrings bilaterally and revisions II Left plantarflexors Botox; serial casting I Right plantarflexors Right tendon transfer; heel cord lengthening; botox I Generalized mild Serial casting tightness

Private PT: 1h/week

II

ROM limitations

Private PT: 1h/week; school OT: 30min twice per week; AFOs Private PT, OT, SLP: 30min/week each; school OT: 45min/week School OT: 45min/week; AFOs School OT: 45 min/week; AFOs

No therapy; AFOs

GMFCS, Gross Motor Function Classification System; ROM, range of movement; Botox, Botulinum toxin; PT, physical therapy; OT, occupational therapy; AFOs, ankle–foot orthoses; SLP, Speech and language pathologist.

456

Developmental Medicine & Child Neurology 2005, 47: 455–461

Results Children with spastic hemiplegia Pretraining, the organization of muscle activity in response to platform perturbations in the hemiplegic leg was characterized by disruption of both temporal and spatial parameters. Figure 1a shows a pretraining example of muscle activity in response to a forward sway perturbation and immediately posttraining (Fig. 1b) in a child with hemiplegia (SH2). In the children with hemiplegia, contraction onset in the ankle muscles of the hemiparetic leg was slow compared with the non-hemiparetic leg. Figure 2a,b compares the onset time (pre-, immediately post-, and one-month posttraining) for G, in response to forward sway, and TA, which is normally the first muscle responding to backward sway perturbations. Posttraining, child SH1 substantially reduced the onset latency for TA muscle onset in the leg with hemiplegia (pretraining: 106ms ±2; posttraining: 92ms ±2 ), but did not change the onset latency for G (pretraining: 70ms ±4; posttraining: 78ms ±3). Child SH2 showed reductions in response onset for G in the hemiplegic leg (pretraining: 105ms ±2; posttraining: 95ms ±13), whereas TA showed no improvement. Pretraining, delayed onset of ankle muscle activity caused disruptions to the sequencing of activity in postural muscles (normally activated in a distal–proximal sequence). Proximal muscles contracted in advance of the distal muscles in both children with hemiparesis during backward sway perturbations, and during forward sway in SH2. Figure 3 shows the percentage of trials in which the children with spastic hemiplegia used a distal–proximal response sequence pretraining, and immediately and one month posttraining in response

to forward sway perturbations. Pretraining, SH2 did not activate muscles in the hemiparetic leg in a distal–proximal sequence. Immediately posttraining, there was a distal–proximal sequence in 40% of the trials, and at one month posttraining in 60%. SH1 was already recruiting muscles in a distal–proximal sequence in all trials pretraining. This improved sequencing in SH2 was associated with faster distal and slower proximal muscle activation. Thus, for forward sway, onset of G activity in SH2 was 105ms pretraining and 95ms posttraining, whereas hamstring activity shifted from 98ms pretraining to 114ms posttraining. Changes in the onset of contraction resulted in a shift to distal–proximal sequencing. Improvements in the timing of postural activity were present at one month, including faster contraction onsets (90ms 1 month posttraining compared with 95 to 97ms immediately posttraining). In child SH2 pretraining, in addition to poor sequencing of agonist muscles, there was a loss of directional specificity in some trials (e.g. in response to forward sway, TA was activated 19ms before G). After training, directional specificity improved, with TA activated 5ms before G in only one of five trials. At one month, the antagonist TA muscle was not activated, eliminating directional specificity problems. Though participant SH1 showed appropriate muscle sequencing in response to forward sway perturbations pretraining, in backward sway perturbations a distal–proximal pattern was found in only 20% of the trials. Posttraining this pattern was found in 100% of the trials, and again resulted from faster distal (TA: 106ms ±21 pretraining; 92ms ±2 posttraining) and slower proximal (quadriceps: 108ms ±24 pretraining; 146ms ±6 posttraining) muscle activation.

a

b

TEMPORAL ORGANIZATION OF POSTURAL RESPONSES

R Gastroc

L Gastroc

R TA

L TA 250µV

250µV

R Ham

L Ham

R Quad

L Quad

L Gastroc

R Gastroc

L TA

R TA

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Time (s)

2.0

2.2

2.4

2.6

2.8

3.0

0.4

0.6

0.8

1.0 1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0

Time (s)

Figure 1: Example of muscle activity in response to a forward sway perturbation (a) before training and (b) immediately after training in one child with spastic hemiplegia (SH2). Onset of platform movement occurs at vertical line. R Gastroc, right gastrocnemius muscle (hemiparetic leg); RTA, right tibialis anterior muscle (hemiparetic leg); Quad, quadriceps muscle; L Gastroc, left (non-hemiparetic) gastrocnemius; LTA, left (non-hemiparetic) tibialis anterior.

Effect of Balance Training on Muscle Activity in Children with Cerebral Palsy Marjorie Woollacott et al.

457

Children with spastic diplegia Pretraining, postural muscle activity in these children was characterized by disruptions in both spatial and temporal aspects of responses. Contraction onset of the distal muscles was delayed

Table IIa: Amplitude asymmetry modulation in gastrocnemius muscle in hemi- versus non-hemiparetic leg, pretraining versus posttraining

SH1 SH2

Pretraining G hemi G non-hemi

Posttraining G hemi G non-hemi

6.8 (1) 9.2 (6)

2.9 (1) 2.1 (1)

42.2 (26) 12.5 (5)

2.6 (1) 7.8 (2)

G hemi, gastrocnemius muscle in hemiparetic leg; G non-hemi, gastrocnemius in non-hemiparetic leg.

Table IIb: Electromyograph amplitude modulation in children with spastic diplegia G SD1 SD2 SD3 SD4

22.2 (7) 5.3 (2) 5.4 (2) 5.6 (1)

Pretraining H 7.3 (2) 0.4 (3) 4.2 (1) 6.5 (3)

P

4.3 (2) 0.8 (2) 13.5 (9) 1.5 (1)

G

Posttraining H

P

12.4 (7) 3.8 (1) 2.0 (1) 2.7 (1) 1.7 (0.5) 2.1 (0.5) 3.4 (1) 1.5 (1) 0.9 (0.1) 3.3 (2) 2.8 (1) 1.3 (0.3)

(G: 115ms ±17; TA: 109ms ±6), allowing proximal muscles to contract in advance of distal muscles. Most children showed a loss of directional specificity, with antagonist muscles contracting either in advance of, or coincident with, agonist muscles. Immediately after training all children with diplegia showed substantial improvements in spatial or temporal characteristics of muscle activity, or both, which affected their directional specificity. Although only two children (SD1 and SD2) modulated the contraction onset of G (first muscle responding to forward sway) after training, all four reduced the onset latency of TA. Figure 2c,d compares the onset latencies for G (Fig. 2c) in response to forward sway, and TA (Fig. 2d) in response to backward sway perturbations in these children. Pretraining, all children showed a proximal–distal activation pattern. Posttraining, two children (SD1 and SD3) demonstrated a distal–proximal sequence (see Fig. 3c,d) associated with earlier onset of activity in distal muscle in combination with later onset in proximal muscle. This improvement did not remain at one month posttraining. For backward sway pretraining, two children showed a distal–proximal sequence whereas two showed distal–proximal coactivation. After training, a distal–proximal sequence was present in all four children; however, this improvement did not remain at one month. MODULATION OF CONTRACTION AMPLITUDES

c 150

Pretest Posttest One month

100

50

0

SH1 Forward sway

200


100

0

SH2 Backward sway

SD1

SD2 SD3 SD4 Gastrocnemius response

150


100

50

0

SH1

SH2

Tibialis anterior response

Onset latency (msec)

d

b Onset latency (msec)

Onset latency (msec)

a Onset latency (msec)

G, gastrocnemius; H, hamstrings; P, paraspinals.

Children with spastic hemiplegia Pretraining, muscle responses showed significantly greater amplitude in the non-hemiparetic leg compared with the hemiparetic leg for both children. Table II compares G amplitude for

150


100

50

0

SD1

SD2

SD3

SD4

Tibialis anterior response

Figure 2: Effect of training on onset of muscle latency to (a,c) forward (gastrocnemius) and (b,d) backward (tibialis anterior) sway perturbations in two children with spastic hemiplegia (SH; a,b) and four children with spastic diplegia (SD; c,d).

458


and SD3). Modulation or response amplitudes were not maintained for child SD2 (participant SD4 was not tested at one month; results are not shown in Table).

forward sway in the hemi- versus non-hemiparetic leg for SH1 and SH2. After training, SH2 reduced the amplitude of activity in the hemiplegic leg, but most strikingly in the non-hemiplegic limb, thus reducing the asymmetry between muscles. Child SH1 had less amplitude asymmetry pretraining. Thus, although training resulted in a decrease in amplitude of both muscles, it did not significantly impact muscle amplitude asymmetry.

Discussion Previous research has examined the effects of training in reactive balance control on balance performance measures in children with CP. It has found that the time required to recover stability after a balance threat and the mean area of COP movement per second during balance recovery were significantly reduced (Shumway-Cook et al. 2003). To determine underlying changes to the nervous system that contributed to those performance improvements, this study compared neuromuscular response characteristics during balance recovery for children with CP during pretraining, posttraining, and at a 1

Children with spastic diplegia The amplitude of both G and TA was reduced from pre- to posttraining for all children (Table IIb). The reduction in response amplitude was seen for all muscles within the synergy. Reduction in response amplitudes was maintained at one month for two of the children with spastic diplegia (SD1

Table III: Summary of strategies used to modify postural control in response to training in children with CP Patient no. Shorten onset of distal agonist

SD 1 SD 2 SD 3 SD 4 SH 1 SH 2

Temporal modulation strategy Lengthen Lengthen proximal onset of muscle antagonist agonist

FS

BS

FS

BS

FS

BS

+

+ + + + +

+

+

+

+ +

N/A +

+ +

N/A +

+ +

N/A +

Amplitude modulation strategy Increase Decrease Decrease amplitude amplitude amplitude of distal of of all muscles agonist antagonist equally

Shorten onset of all muscles equally FS

BS

+

+

FS

BS

FS

BS

FS

BS

+ + + +

+

+

+ + +

+ + +

+

Temporal modulation strategies: (1) shorten agonist and lengthen proximal synergist – emergence of a distal to proximal sequence; (2) shorten agonist and lengthen antagonist – increase directional specificity, decrease in coactivation; (3) shorten onset of all muscles – faster response but not necessarily better organized. Amplitude modulation strategies: (1) increase amplitude of agonist and decrease antagonist amplitude – reduce coactivation; (2) decrease amplitude of all muscles equally – could be compensatory to improved timing. FS, forward sway; BS, backward sway; SD, spastic diplegia; SH, spastic hemiplegia; +, technique used; N/A, not applicable.

Forward sway – spastic hemiplegia

a

b

Forward sway – spastic diplegia 100

100

Pretest

Posttest One month 50

0

Trials for distal to proximal sequence (%)

Trials for distal to proximal sequence (%)

Pretest

Posttest One month 50

0 SH1

SH2

Children with spastic hemiplegia

SD1

SD2

SD3

SD4

Children with spastic diplegia

Figure 3: Effects of training on sequencing of muscle activity. Comparison of number of trials in which children with (a) spastic hemiplegia (SH1 and SH2) and (b) diplegia (SD1–4) used a distal–proximal response sequence in response to a forward sway perturbation pretraining, immediately posttraining, and at 1-month follow-up.


459

month follow-up test session. Two hypotheses about neuromuscular contributions to improvements in reactive postural performance were identified, related to two levels of balance control. We hypothesized that improvements could be due to some combination of (1) changes in directional specificity (level 1), and (2) changes in the spatial or temporal domains, or both, including faster activation of muscle contraction, reducing agonist–antagonist coactivation, and improvements in distal–proximal muscle response sequencing (level 2). Results suggest that changes in both of these neural factors contributed to improvements in reactive balance performance with training, with each child showing a different combination of factors that contribute to their improved performance. In addition, the level of change in neural factors underlying balance control depended on the severity of involvement of the child (hemiplegia vs diplegia, and level of involvement within each diagnostic category). Table III summarizes the strategies used by the children with CP to modify postural control posttraining. We have aggregated these individual strategies into two main categories: temporal (including directional specificity) and amplitude modulation. Table III also summarizes the combination of strategies used by each child to modify postural responses to forward sway and backward sway perturbations. Note that there was improved directional specificity (shortening the onset of agonist or lengthening the onset of antagonist, or both) in children from both groups. This was most evident for backward sway perturbations, possibly because sway in this direction causes most risk for loss of balance. In general, the children with hemiplegia modified both temporal and amplitude characteristics of postural responses. In contrast, only one of the four children with diplegia (SD4) was able to modulate both. Two children with spastic diplegia modulated the relative timing of muscle activity and reduced response amplitude across all muscles. The remaining child (SD1) reduced onset and amplitude of all muscles. This strategy did not target specific muscles but improved the overall efficiency of balance recovery (reduction in COP area and time to stabilization; see Shumway-Cook et al. 2003). For the children with hemiplegia, we hypothesize that the ability to modulate amplitude effectively in hemi- and nonhemiplegic legs is in part dependent on the ability to improve the temporal organization and, thus, the contribution of the hemiplegic side to the overall response. This would decrease the reliance on the non-hemiplegic side. In fact, this was associated with a reduction in the amplitude of the response on the non-hemiplegic side. There are two possible explanations for the reduction in response amplitude of agonist muscles with training. First, improved sequencing of muscle activity and reduced onset latency increased the efficiency of the response, thus allowing a decrease in magnitude of muscle activity. For example, if an ankle muscle response were activated earlier in response to a perturbation, the center of mass would not have moved as far away from the center of stability before muscle activation, and it would take less force to return the center of mass to baseline (Nashner and McCollum 1985). Second, the response magnitude could be due to adaptation from repeated experience on the platform, and not the result of improved sequencing and timing. For example, Woollacott et al. (1988) showed that adults given repeated exposure to

460


platform perturbations showed reductions in response amplitudes. We hypothesize that it was due to both. It is interesting that the children with diplegia tended to show a non-specific change in amplitude factors rather than specific changes in individual muscles. One hypothesis to explain this lack of specificity is that this category of CP shows greater deficit in motor cortex function, which is known to control specificity of individual muscle responses (Ghez 1991). These children also did not retain their changes in muscle sequencing beyond the first posttest measure, possibly due to the short duration of the training. Modulation of muscle activity posttraining was also associated with improvements in the organization and timing of the COP in all six children, as shown in an earlier study (Shumway-Cook et al. 2003). Total sway area decreased, as did the time needed to recover a stable position. Thus, changes in temporal and spatial aspects of muscle responses to loss of balance were associated with improved efficiency, as reflected in COP measurements. The relation between changes in muscle activity and clinical measures (Shumway-Cook et al. 2003) were more variable, possibly because training was limited to one week. Three children improved on dimension D of the GMFM, one child had no change, and one had a slight decrease in score. Improvements in the GMFM may be either an effect of practice (the second test was given within 2 weeks of the first), or the result of improved neuromuscular organization related to postural control. Because of the small sample size, further studies are needed to determine the functional consequences of improvement in neuromuscular activity on gross motor function, as measured by the GMFM. In addition, because testing and training were performed by the same individual, tester bias may have been a factor in determining the clinical results. In summary, our results suggest that reactive balance training in children with CP can result in changes in specific neural factors contributing to balance control. Each of the children with spastic hemi- or diplegia showed a different combination of factors contributing to improved performance. This provides evidence that reactive balance training may be beneficial for postural response organization in children with CP. The presence of variability among children with the same diagnosis underscores the importance of individual analyses when examining the effects of training on neuromuscular mechanisms related to postural control. DOI: 10.1017/S0012162205000885 Accepted for publication 16th August 2004. Acknowledgments This research was supported by NIH grant 2R01 NS038714, Dynamic Balance in Children with Cerebral Palsy, to M Woollacott.

References Brogren E, Hadders-Algra M, Forssberg H. (1996) Postural control in children with spastic diplegia: muscle activity during perturbations in sitting. Dev Med Child Neurol 38: 379–388. Brogren E, Hadders-Algra M, Forssberg H. (1998) Postural control in sitting children with cerebral palsy. Neurosci Biobehav Rev 22: 591–596.

Brogren E, Forssberg H, Hadders-Algra M. (2001) Influence of two different sitting positions on postural adjustments in children with spastic diplegia. Dev Med Child Neurol 43: 534–546. Burtner PA, Qualls C, Woollacott MH. (1998) Muscle activation characteristics of stance balance control in children with spastic cerebral palsy. Gait Posture 8: 163–174. Butler PB. (1998) A preliminary report on the effectiveness of trunk targeting in achieving independent sitting balance in children with cerebral palsy. Clin Rehabil 12: 281–293. Crenna P. (1998) Spasticity and ‘spastic’ gait in children with cerebral palsy. Neurosci Biobehav Rev 22: 571–578. Crothers B, Paine PS. (1988) The Natural History of Cerebral Palsy. Classics in Developmental Medicine No. 2. London: Mac Keith Press. Ghez C. (1991) Voluntary movement. In: Kandel E, Schwartz JH, Jessell TM, editors. Principles of Neuroscience. 3rd edn. NY: Elsevier. p 609–625. Hadders-Algra M, van der Fits IB, Stremmelaar EF, Touwen BC. (1999) Development of postural adjustments during reaching in infants with CP. Dev Med Child Neurol 11: 766–776. Hur JJ. (1995) Review of research on therapeutic interventions for children with cerebral palsy. Acta Neurol Scand 91: 423–432. McClenaghan BA, Thombs L, Milner M. (1992) Effects of seat surface inclination on postural stability and function in the upper extremities of children with cerebral palsy. Dev Med Child Neurol 34: 40–48. Nashner LM, McCollum G. (1985) The organization of human postural movements: a formal basis and experimental synthesis. Behav Brain Sci 8: 135–172. Nashner LM, Shumway-Cook A, Marin O. (1983) Stance postural control in select groups of children with cerebral palsy: deficits in sensory organization and muscular organization. Exp Brain Res 49: 393–409. Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. (1997) Development and reliability of a system to classify gross motor function of children with cerebral palsy. Dev Med Child Neurol 39: 214–223. Pape KE, Kirsch SE, Galil A, Boulton JE, White MA, Chipman M. (1993) Neuromuscular approach to the motor deficits of cerebral palsy: a pilot study. J Pediatr Orthop 13: 628–633. Park ES, Park CI, Lee HJ, Cho YS. (2001) The effect of electrical stimulation on the trunk control in young children with spastic diplegic cerebral palsy. J Korean Med Sci 16: 347–350.

Roncesvalles N, Woollacott M, Burtner P. (2002) Neural factors underlying reduced postural adaptability in children with cerebral palsy. Neuroreport 13: 2407–2410. Russell DJ, Rosenbaum PL, Gowland C. (1993). Manual for the Gross Motor Function Measure. 2nd edn. Hamilton, Ontario: Institute for Applied Health Sciences, McMaster University, CanChild Centre for Childhood Disability Research. Seeger BR, Caudrey DJ, O-Mara NA. (1984) Hand function in cerebral palsy: the effect of hip-flexion angle. Dev Med Child Neurol 26: 601–606. Shumway-Cook A, Hutchinson S, Kartin D, Price R, Woollacott M. (2003) The effect of balance training on recovery of stability in children with cerebral palsy. Dev Med Child Neurol 45: 591–602. Sveistrup H, Woollacott M. (1997) Practice modifies the developing automatic postural response? Exp Brain Res 114: 33–43. Thelen E, Corbetta D, Kamm K, Spencer JP, Schneider K, Zernicke RF. (1993) The transition to reaching: mapping intention and intrinsic dynamics. Child Dev 6: 1058–1098. Woollacott MH, Burtner P, Jensen J, Jasiewicz J, Roncesvalles R, Sveistrup H. (1998) Development of postural responses during standing in healthy children and children with spastic diplegia. Neurosci Biobehav Rev 22: 583–589. Woollacott M, von Hofsten C, Rosblad B. (1988) Relation between response onset and body segmental movement during postural perturbations in humans. Exp Brain Res. 72: 593–604.

List of abbreviations COP EMG G GMFCS GMFM SD SH TA

Centre of pressure Electromyogram Gastrocnemius Gross Motor Function Classification Scale Gross Motor Function Measure Spastic diplegia Spastic hemiplegia Tibialis anterior

Published by

Mac Keith Press

for the

nter national hild eur ology ssociation

Tuberous Sclerosis Complex : from basic science to clinical phenotypes Edited by Paolo Curatolo ISBN 1 898 683 39 5 US$75.00 £55.00 pp315, 2003 Tuberous sclerosis complex is a genetically determined, variably expressed, multisystem disorder that may affect any human organ with well-circumscribed benign, non-invasive lesions. This book provides a review of the historical background and current criteria for diagnosis, and detailed descriptions of the clinical manifestations of this protean disease are given. Signiﬁcant advances in the ﬁelds of neuropathology, molecular genetics, and neurobiology are progressing at a fast rate. This books seeks to encapsulate in a single text, knowledge about modern research, ideas and practice on tuberous sclerosis.

International Child Neurology Association. From the International Review of Child Neurology Series. See inside cover for purchase details.


461