Partial body-weightsupported. training can improve walking in children with cerebral palsy: a clinical controlled trial

Transcription

1 Partial body-weightsupported treadmill training can improve walking in children with cerebral palsy: a clinical controlled trial Karen J Dodd* PhD; Sarah Foley BApp Sc (Physio), Musculoskeletal Research Centre, School of Physiotherapy, Faculty of Health Sciences, La Trobe University, Victoria, Australia. *Correspondence to first author at School of Physiotherapy, Faculty of Health Sciences, La Trobe University, Victoria 3086, Australia. K.Dodd@latrobe.edu.au This matched-pairs, clinical controlled trial evaluated the effects of a school-based, partial body-weight-supported treadmill training (PBWSTT) programme, conducted twice a week for 6 weeks, on the walking speed and endurance of children with cerebral palsy (CP). Pairs of children (10 males, four females; mean age 8y 10mo [SD 2y 6mo], range 5 14y) matched for sex, age, type of CP (athetoid quadriplegia, n=6; spastic quadriplegia, n=6; spastic diplegia, n=2), and Gross Motor Function Classification System level (10 at Level IV, four at Level III) were allocated to the experimental or control group. Compared with the control group, the seven treadmill-training participants increased their self-selected walking speed over 10 metres (Mann Whitney U=9.00, z= 1.98, p=0.048; mean difference 4.21m/min). A trend was also found for increased distance walked over ground in 10 minutes (t[12]=1.88, p=0.083; mean difference 19.81m). A relatively short PBWSTT programme can improve the walking speed of children with CP and moderate to severe disabilities. Walking endurance might also improve in some children. Treadmill training seems to be a useful gait training option for children with CP, and it seems feasible to conduct such a programme within a school environment. The development of efficient and independent walking is an important therapeutic goal for many children with cerebral palsy (CP). As a result of impairments such as spasticity, muscle weakness, and postural instability, as much as 90% of children with CP have difficulty in walking. 1,2 Reduced walking speed 3,4 and endurance 5,6 are two of the main functional problems, particularly in children with more severe disabilities. 7 Current theories of motor learning suggest that task-specific repetitive practice can improve activities such as walking in people with neurological disorders such as CP. 8,9 Consequently, there has been a growing interest in the effects of treadmill training programmes, particularly with partial body-weight support, for people with neurological disorders. In people without CP, partial body-weight-supported treadmill training (PBWSTT) has been shown to have a range of benefits. For example, a recent review 10 found a trend for people with stroke to increase their over-ground walking speed and walking endurance after participation in a PBW- STT programme. In people with incomplete spinal cord injury, PBWSTT has been shown to increase walking speed 11,12 and endurance, 12 while decreasing oxygen costs, 11 whereas in people with Parkinson s disease, PBWSTT has been found to increase walking speed and stride length. 13 Despite increasing clinical interest, few published articles have reported the effects of PBWSTT on children with CP One article 14 documented the effects of a 12-week programme on 10 children (mean age 11y 6mo, range 6 18y; three had spastic diplegia, four had spastic quadriplegia, and three had spastic quadriplegia and ataxia) using a single-group, pre post research design. In an attempt to control for confounding series effects, a non-intervention baseline phase was incorporated before the programme started and no significant change was detected. After intervention, significant improvements were detected in the Functional Ambulation Category, 22 and the Gross Motor Function Measure (GMFM) 23 dimensions D (Standing) and E (Walking, Running, and Jumping) scores. In another single-group, pre post study, 15 the effects on six children (mean age 10y 2mo [SD 4y 4mo]; spastic diplegia, n=4; spastic hemiplegia, n=1; and athetoid quadriplegia, n=1) of a 12-week PBWSTT programme were investigated. A significant improvement was reported in over-ground walking speed, but no change was detected in the energy expenditure index. A further six relevant articles reported improvements in walking speed 16 19, walking endurance, 17 independent walking function, 16,17,19,20 and GMFM scores. 18,20,21 However, the ability to draw conclusions from these trials is limited because they comprised single-subject case studies without statistical analysis and none established a stable baseline before intervention. Preliminary evidence exists that PBWSTT might be beneficial for children with CP. However, the effects of PBWSTT need to be investigated in a clinically-controlled trial to provide stronger evidence about whether the training can improve walking in such children. Accordingly, our study sought to determine whether a 6- week PBWSTT programme conducted in a school environment could: (1) increase the self-selected over-ground walking speed, and (2) increase the walking endurance of children with CP and moderate to severe disabilities. Method PARTICIPANTS Participants were recruited from two schools for children Developmental Medicine & Child Neurology 2007, 49:

2 with disabilities located in Victoria, Australia. To be included, children had to be aged between 5 and 18 years, be able to understand simple instructions, and have CP with a level of disability categorized as III or IV on the Gross Motor Function Classification System (GMFCS). 24 Volunteers were excluded if they needed physical assistance from another person to walk, if they had undergone lower limb surgery, botulinum toxin injections, or serial casting in the 6 months before the trial started, or if they had a concurrent medical condition, such as uncontrolled epilepsy, that would limit their ability to participate in a physical activity programme. La Trobe University and Education Department Ethics Committees approved this trial, and informed consent was obtained for each participant. APPARATUS A motorized treadmill (Paragon CR; Horizon Fitness, Morwell, Australia), equipped with emergency stop and pause buttons and capable of measuring the amount of time and distance walked in a training session, was used in this trial. The treadmill had a minimum speed of 0.1km/h, with increments of 0.1km/h. During training, partial body-weight support was provided by one of two systems. The first was a commercial system designed for children called the WalkAble (Mobility Research, Tempe, AZ, USA). This system comprised a harness that supported the child around the pelvis and lower trunk. The harness was suspended vertically by clips attached to two horizontal bars positioned directly over the child s head. The bars were attached to a vertical frame that was fitted with lockable castors. The second system was for children who required greater trunk and neck support, and comprised a custommade harness made of heavy-duty cotton corduroy. The harness was suspended vertically by two metal rings fixed at shoulder height, which were attached by clips to the metal crossbar of a manually operated hoist (Elf Hoist; ProMed, Melbourne, Australia). Each experimental group participant was given a trial with both systems and the one that proved more comfortable was used for the duration of the study. PROCEDURE Participants in this matched-pairs, clinical controlled trial were allocated to either the experimental or the control group according to which school they attended. The training programme was implemented in only one of the schools taking part in this trial; eligible children attending this school were recruited first. Each child was then matched for sex, age (within a range of 2y either way), type of CP, and GMFCS level 24 to a child attending the other school. After completing baseline demographic and outcome measures, the experimental group began a 6-week, twiceweekly PBWSTT programme. The programme was conducted at the children s school, within school hours. There was at least 1 day s rest between sessions. One physiotherapist experienced in treating children with CP conducted all training sessions. The children wore their usual footwear and orthoses during training. At each training session, the child was assisted into the apparatus which was placed over the treadmill. The child was then asked to stand as upright as possible and the bodyweight support was reduced until the child began either to flex at the hips or knees or to sit in the harness. The treadmill was then started at the lowest speed (0.1km/h) and increased in 0.1km/h increments to a speed at which the child stepped forward comfortably. If required, the physiotherapist provided assistance to initiate weight shift or the swing phase of the gait cycle. Each session lasted a maximum of 30 minutes. Sessions finished earlier if the child asked to stop, or if the child stopped stepping. A logbook was kept for each child to record the number of sessions attended, the treadmill speed, and the amount of time and distance walked each session. The logbook was also used to record any problems or injuries that might have occurred during training. All participants were instructed to continue their normal daily activities, including any physical activity programmes they usually participated in. Participants also continued their normal physiotherapy programmes, but the children s usual physiotherapists were asked not to increase the amount of gait training or to do any treadmill training. OUTCOME MEASURES Baseline demographic characteristics (age, weight, sex, type of CP, and GMFCS level 24 ) were recorded to measure the success of matching. Walking performance was evaluated immediately before starting the trial and 6 weeks later at the conclusion of the trial, using the self-selected 10-metre walking speed test 25 and the 10-minute walk test. 26 At both measurement sessions, children wore their usual footwear and orthoses, and used their usual gait-assistive devices. A physiotherapist experienced in assessing movement disorders completed all assessments for both groups. Although this assessor was not blind to group allocation, she did not have access to the baseline assessments at the second measurement session. All measurement sessions were completed in a quiet area of the gymnasium within the school that each child attended. Self-selected 10-metre walking speed was used as a measure of walking performance because it has been found to provide an indication of gait function in people with neurological disorders 27 and has high retest reliability (r>0.95) when measured with a stopwatch. 25 Data were collected from the central 10-metre length of a 14-metre linoleum-covered walkway. This allowed a distance of 2 metres before and after the data collection area for acceleration and deceleration of speed. The children were placed in their gait-assistive devices and asked to walk to the assessor who stood at the end of the walkway. The standardized instructions were: I am going to see how long it takes you to walk from here to here [demonstrate walking at comfortable speed]. Don t hurry, just go at your normal speed. I am going to time how long it takes using this watch. Are you ready? Go! The 10-minute walk test 26 was used as the other measure of walking performance because it has been shown to (1) provide a useful measure of walking endurance, (2) be a representative measure of a child s community walking speed, and (3) have high retest reliability (intraclass correlation coefficients 0.91, 95% confidence interval (CI) ) in children with CP. 26 For this test, a linoleum-covered oval track 20 metres in distance was marked out with masking tape. The standardized instructions were: Can you see the line on the floor around the gym? I want you to follow this line [demonstrate walking around the walkway]. Keep walking until I say stop. The assessor walked at the child s side, and verbal encouragement was given if they stopped. After 10 minutes the number of full and partial laps walked was recorded. Because children were expected to be fatigued after the 10-minute walk test, the selfselected 10-metre walk was always completed first. 102 Developmental Medicine & Child Neurology 2007, 49:

3 STATISTICAL ANALYSIS To examine whether baseline differences existed between the groups for demographic or walking performance measures, independent t-tests or, if the assumptions of parametric statistical analysis were not satisfied, Mann Whitney U tests were used. To examine whether the experimental group s performance had significantly improved in comparison with the control group, the change scores for each group (i.e. test 2 minus test 1) were calculated for the self-selected 10-metre walk and 10- minute walk scores. The between-group change scores were then compared by using independent t-tests or Mann Whitney U tests, as appropriate. To provide estimates of systematic change in the units of measure, mean differences with 95% CI were calculated for each group. To provide proportional indices of the strength of the treatment effect, effect sizes and their associated 95% CIs were calculated. Effect sizes were calculated by subtracting the mean change score of the control group from the mean change score of the experimental group and dividing by the standard deviation of the control group. 28 Power analyses were completed on comparisons that approached significance and, for participants who did not return for follow-up testing, an intention-to-treat analysis was conducted with the last observed carry-forward method. 29 A two-tailed level of significance was set at 0.05 for all analyses. Data were analyzed with the SPSS (version 11.5). Results Fourteen children (10 males, four females) took part in this study. No eligible child who attended the school in which the training programme was implemented refused to participate. Individual and group participant details and walking performance outcome measures are summarized in Table I. No significant between-group difference was detected for age or weight, and the children were matched exactly for sex, type of CP, and GMFCS level. Baseline walking performance measures are shown in Tables I and II. No significant betweengroup difference was detected. All 14 children completed baseline assessments. One male in the control group did not attend assessment session 2 because he had unexpected surgery in week 6 of the trial. As Table I shows, six of the seven children in the experimental group appeared to Table I: Individual and group participant characteristics and performance measures at baseline and at 6 weeks Participant Age Sex Weight GMFCS Type 10m walk test (m/min) 10min walk test (m) (y:m) (kg) level of CP Baseline 6wk Baseline 6wk Experimental group 1 8:1 F 18.5 IV AQ :3 M 17.2 IV AQ :3 M 18.1 III SD :0 M 27.3 IV SQ :7 M 17.3 III SQ :2 F 18.1 IV AQ :7 M 29.4 IV AQ Group means 8:5 (2:6) 5M, 2F 20.8 (5.2) Level III: (7.12) (8.93) (24.32) (42.29) (SDs) and totals Level IV: 5 Control group 1 5:5 F 17.3 IV AQ :10 M 16.0 IV AQ :0 M 20.7 III SD :2 M 36.2 IV SQ a a 5 8:8 M 33.8 III SQ :10 F 27.4 IV AQ :3 M 20.3 IV AQ Group means 9:5 (2:9) 5M, 2F 24.5 (8.0) Level III: (4.48) 7.85 (5.70) (38.28) (39.28) (SDs) and totals Level IV: 5 Level IV of Gross Motor Function Classification System (GMFCS) represents greater functional disability. 24 Weight and walking performance outcome values in group summary rows are means (SD). a Missing data and the use of a carry-forward technique for intention-to-treat analysis. AQ, athetoid quadriplegia; SD, spastic diplegia; SQ, spastic quadriplegia. Table II: Group walking performance measures at baseline and 6 weeks with mean change (95% confidence interval [CI]) scores Walking performance Baseline 6wks Mean difference (95% CI) outcome measure Experimental Control Experimental Control Experimental Control (n=7) (n=7) (n=7) (n=7) (n=7) (n=7) Self-selected speed 10m (7.12) 7.75 (4.48) (8.93) 7.85 (5.70) 4.21 ( 0.20 to 8.61) a 0.10 ( 1.96 to 2.16) walk test (m/min) 10-min walk test (m) (24.32) (38.28) (42.29) (39.28) ( 1.92 to 41.53) 3.69 ( to 17.64) Values are mean (SD) except where noted otherwise. a p<0.05. Treadmill Training for Children with CP Karen J Dodd and Sarah Foley 103

4 improve their self-selected walking speed over 10 metres, in comparison with only two in the control group, whereas five of the children in the experimental group improved their walking endurance, in comparison with three in the control group. Children in the experimental group adhered to the programme, attending a mean of 11.4 (SD 0.98; range 10 12) of the 12 prescribed sessions. Five children attended all 12 sessions. The other two children missed two sessions each: one went on vacation and missed two consecutive training sessions, and the other missed one session because he was on vacation, and another session due to illness. The logbooks showed that training intensity increased during the programme. Median treadmill speed was 0.40km/h (interquartile range [IQR] 0.30, range ) at the start of training. This was increased until a median speed of 0.60km/h (IQR 0.40, range ) was reached at the end of the 6-week programme. The amount of time walked within a session also progressed from a median of minutes (IQR 4.73, range ) at the start of training, to a median of minutes (IQR 10.55, range ) at the end. Similarly, the distance walked increased from a median of 90m (IQR 30, range ) at the start of training, to a median of 230m (IQR 160, range ) at the end. Results of PBWSTT on outcomes of group walking performance are shown in Table II. The Shapiro Wilks test 30 revealed that the self-selected 10-metre walk change score data were not normally distributed, so non-parametric analysis was used and a significant between-group difference was detected (Mann Whitney U=9.00, z= 1.98, p=0.048). As Table II shows, the experimental group demonstrated a mean increase of 4.21m/min (68% increase) in speed, and the effect size was large (d=1.84; 95% CI ). Although it did not reach statistical significance, a trend was found (t[12]=1.88, p=0.083) for the experimental group to walk for longer distances over-ground in 10 minutes in comparison with controls. Table II also shows a mean difference of 19.81m (57% increase) in the experimental group, and again the size of the effect was large (d=1.02; 95% CI 0.83 to 1.27). Power analysis suggested that with the effect size maintained and the sample increased to n=8 in each group, there would be an 80% chance of detecting a significant betweengroup difference in the 10-minute walk test. There was no adverse event that led to a missed training session. No muscle soreness, joint pain, or increased fatigue was reported during or after the treadmill training programme, and no injures due to trips or falls were sustained. Discussion School-aged children with moderate to severe functional disabilities benefited from participating in a short PBWSTT programme, with a significant increase in self-selected over-- ground walking speed. In addition, there was a trend for these children to increase their over-ground walking endurance over 10 minutes. These benefits were achieved without any adverse events. Walking speed is an important criterion in deciding whether walking is likely to be a functional method of locomotion for a child with CP, because slow speeds are likely to restrict the child s ability to function in the community and interact with their peers. 26 These changes in walking speed could, therefore, have a meaningful impact on the child s independent mobility, particularly over shorter distances, such as when walking at home or between classes at school. The changes found in walking speed support those reported by other studies The magnitude of change (68%) was also comparable to the large (56% and 72%) increases reported in two of these previous studies. 16,17 Both previous studies also included participants with moderate and severe functional disability, which suggests that children with relatively severe walking disabilities can make appreciable changes in performance after participation in a relatively short PBWSTT programme. Another study that included children with milder disability reported a smaller (10%) increase in speed. 15 A further two previous articles also reported increases in walking speed, but the magnitude of the change was not reported. 18,19 The trend found toward increased walking endurance is also supported by the only other study to include a similar measure; 17 that study, of two males aged 13 years with quadriplegic CP, reported increases during the 6-minute walk test of 18.6m and 14.7m after a 9-week PBWSTT programme. A distinguishing characteristic of participants in the present study was that they had relatively severe walking disabilities. Children aged 8 to 9 years without impairment walk at an average speed of 71.4m/min, and 9- to 10-year-olds walk at 81.6m/min. 31 In comparison, as Table II shows, the children in the present study walked very slowly over even a short distance, and they walked only relatively short distances over 10 minutes. Contemporary clinical practice suggests that taskspecific repetitive practice of walking is required to improve walking performance. 8 Because of impairments such as postural instability and muscle weakness, none of the children in our study could easily practice whole gait cycles in other conventional gait training situations. The PBWSTT programme provided an environment in which these children could repetitively practice a more rhythmical and efficient walking pattern at higher speeds. A feature of the PBWSTT programme in this study was that it was conducted in the children s usual specialist school environment, reducing the travel required by the child and their family and the inconvenience of separate therapy appointments that might interfere with social activities or school attendance. Apart from one study that investigated the effects of a parent-administered training programme, 20 previous studies have evaluated programmes conducted in rehabilitation centres that are relatively expensive in terms of therapy resources. The feasibility of our programme was improved by scheduling two short 30-minute training sessions each week, and running the programme for only 6 weeks. To our knowledge the present study is the first to investigate the effects of PBWSTT for children with CP that includes a matched control group to control more fully for potentially confounding series effects. However, there are some limitations. The main limitation was that group allocation was not randomized, which may have introduced biases that overestimated or underestimated the differences between the PWB- STT programme and the usual care programme. Another limitation was that the assessor was not blind to group allocation. However, to minimize bias, the assessor did not have access to baseline data at measurement session 2, and it was unlikely that scores taken 6 weeks earlier were recalled. This study adds to the accumulating evidence that PBW- STT can be beneficial for increasing the walking speed of children with CP who have moderate to severe disabilities. The results also suggest that PBSWTT might increase the 104 Developmental Medicine & Child Neurology 2007, 49:

5 walking endurance of some children. PBWSTT could be a feasible and useful gait training option for children with a limited capacity to practise walking in conventional therapy situations. Accepted for publication 27th September References 1. Pharoah P, Cooke T, Johnson M, King R, Mutch L. (1998) Epidemiology of cerebral palsy in England and Scotland, Arch Dis Child Fetal Neonatal Ed 79: F21 F Hutton J, Pharoah P. (2002) Effects of cognitive, motor, and sensory disabilities on survival in cerebral palsy. Arch Dis Child 86: Norlin R, Odenbrick P. (1986) Development of gait in spastic children with cerebral palsy. J Pediatr Orthop 6: Abel M, Damiano D. (1996) Strategies for increasing walking speed in diplegic cerebral palsy. J Pediatr Orthop 16: Berg K, Olsson T. (1970) Energy requirements of school children with cerebral palsy as determined from indirect calorimetry. Acta Paediatr Scand Suppl 204: Bottos M, Feliciangeli A, Sciuto L, Gericke C, Vianello A. (2001) Functional status of adults with cerebral palsy and implications for treatment of children. Dev Med Child Neurol 43: Duffy C, Hill A, Cosgrove A, Corry I, Graham H. (1996) Energy consumption in children with spina bifida and cerebral palsy: a comparative study. Dev Med Child Neurol 38: Barbeau H. (2003) Locomotor training in neurorehabilitation: emerging rehabilitation concepts. Neurorehabil Neural Repair 17: Hesse S. (2001) Locomotor therapy in neurorehabilitation. NeuroRehabilitation 16: Moseley A, Stark A, Cameron I, Pollock A. (2005) Treadmill training and body weight support for walking after stroke. Cochrane Database Syst Rev (4): CD Protas E, Holmes S, Qureshy H, Johnson A, Lee D, Sherwood A. (2001) Supported treadmill ambulation training after spinal cord injury: a pilot study. Arch Phys Med Rehabil 82: Hicks A, Adams M, Martin G, Giangregorio L, Latimer A, Phillips S, McCartney N. (2005) Long-term body-weight-supported treadmill training and subsequent follow-up in persons with chronic SCI: effects on functional walking ability and measures of subjective well-being. Spinal Cord 43: Miyai I, Fujimoto Y, Yamamoto H, Ueda Y, Saito T, Nozaki S, Kang J. (2002) Long-term effect of body weight-supported treadmill training in Parkinson s disease: a randomized controlled trial. Arch Phys Med Rehabil 83: Schindl M, Forstner C, Kern H, Hesse S. (2000) Treadmill training with partial body weight support in nonambulatory patients with cerebral palsy. Arch Phys Med Rehabil 81: DeJong S, Stuberg W, Spady K. (2005) Conditioning effects of partial body weight support treadmill training in children with cerebral palsy. Pediatr Phys Ther 17: 78. (Extended abstract) 16. Richards C, Malouin F, Dumas F, Marcoux S, Lepage C, Menier C. (1997) Early and intensive treadmill locomotor training for young children with cerebral palsy: a feasibility study. Pediatr Phys Ther 9: Beard L, Harrow C, Bothner K. (2005) The effect of body weight support treadmill training on gait function in cerebral palsy: two case studies. Pediatr Phys Ther 17: 72. (Extended abstract) 18. Begnoche D, Sanders E, Pitetti K. (2005) Effect of an intensive physical therapy program with partial body weight treadmill training on a 2 year-old child with spastic quadriplegic cerebral palsy. Pediatr Phys Ther 17: 73. (Extended abstract) 19. Sanders E, Begnoche D, Pitetti K. (2005) Effect of an intensive physical therapy program with partial body weight treadmill training on a 9 year-old child with spastic diplegic cerebral palsy. Pediatr Phys Ther 17: 82. (Extended abstract) 20. Dannemiller L, Heriza C, Burtner P, Gutierrez T. (2005) Partial weight bearing treadmill training in the home with young children with cerebral palsy: a study of feasibility and motor outcomes. Pediatr Phys Ther 17: Day J, Fox E, Lowe J, Swales H, Behrman A. (2004) Locomotor training with partial body weight support on a treadmill in a non-ambulatory child with spastic tetraplegic cerebral palsy: a case report. Pediatr Phys Ther 16: Holden M, Gill K, Magliozzi M, Nathan J, Piehl-Baker L. (1984) Clinical gait assessment in the neurologically impaired: reliability and meaningfulness. Phys Ther 64: Russell D, Rosenbaum P, Cadman D, Gowland C, Hardy S, Jarvis S. (1989) The gross motor function measure: a means to evaluate the effects of physical therapy. Dev Med Child Neurol 31: Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. (1997) Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 39: Wade D. (1992) Measurement in Neurological Rehabilitation. Oxford: Oxford University Press. 26. Pirpiris M, Wilkinson A, Rodda J, Nguyen T, Baker R, Nattrass G, Graham HK. (2003) Walking speed in children and young adults with neuromuscular disease: comparison between two assessment methods. J Pediatr Orthop 23: Andriacchi T, Ogle J, Galante J. (1977) Walking speed as a basis for normal and abnormal gait measurements. J Biomech 10: Thomas J, Salazar W, Landers D. (1991) What is missing in p less than 0.05? Effect size. Res Q Exerc Sport 62: Shao J, Zhong B. (2003) Last observation carry-forward and last observation analysis. Stat Med 22: Norusis M. (1993) SPSS for Windows Base System User s Guide. Release 6.0 ed. Chicago: SPSS Inc. 31. Stansfield B, Hillman S, Hazelwood M, Lawson A, Mann A, Loudon I, Robb JE. (2001) Normalized speed, not age, characterizes ground reaction force patterns in 5- to 12-year-old children walking at self-selected speeds. J Pediatr Orthop 21: Treadmill Training for Children with CP Karen J Dodd and Sarah Foley 105