Associations between fitness and mobility capacity in school-aged children with cerebral palsy: a longitudinal analysis
Abstract
Aim
The aim of this study was to determine the longitudinal associations among fitness components and between fitness and mobility capacity in children with cerebral palsy (CP).
Method
Forty-six children (26 males, 20 females; mean age 9y 7mo [SD 1y 8mo]) with a bilateral (n=24) or a unilateral spastic CP (n=22) participated in aerobic and anaerobic fitness measurements on a cycle ergometer and isometric muscle strength tests (Gross Motor Function Classification System [GMFCS] level I [n=26], level II [n=12], level III [n=8]). Mobility capacity was assessed with the gross motor function measure (GMFM) and a walking capacity test. Associations over longitudinal measurements (three or four measurements over 1y) were determined since longitudinal data allow a more accurate estimation. The associations were determined using a mixed model with fixed effects (mobility capacity as dependent variables and fitness components as independent variables) and a random intercept.
Results
In children with bilateral CP, changes in aerobic fitness were associated with changes in anaerobic fitness (p<0.001), and changes in aerobic fitness showed an association with changes in muscle strength (p<0.05). Anaerobic fitness was not associated with muscle strength. No associations between fitness components were found in unilateral CP. Anaerobic fitness and muscle strength were significant determinants for GMFM and walking capacity in bilateral but not in unilateral CP.
Interpretation
The longitudinal associations between aerobic and anaerobic fitness and mobility indicate that increasing either aerobic or anaerobic fitness is associated with improvements in mobility in children with bilateral CP. While increasing anaerobic fitness might be beneficial for mobility capacity in children with bilateral CP, this is less likely for children with unilateral CP.
What this paper adds
- There are no associations between changes in fitness components and mobility capacity in children with unilateral CP.
- In children with bilateral CP, better anaerobic fitness relates to better aerobic fitness and higher mobility capacity.
- Different interventions to improve mobility capacity are required for children with bilateral or unilateral CP.
This article is commented on by Narayanan on pages 597–598 of this issue.
Abbreviations
-
- GMAE
-
- Gross Motor Ability Estimator
-
- ICC
-
- Intraclass correlation coefficient
-
- VO2peak
-
- Peak aerobic capacity
Emerging evidence shows that fitness is an important health indicator in all individuals.1 Adequate fitness might be even more important for people with a physical disability, who often have lower levels of fitness and higher risks of developing secondary health conditions including obesity, diabetes, and cardiovascular disease.2 It has also been suggested that adequate fitness in children with a physical disability is associated with a better capacity to execute daily activities.3 The most common physical disability in childhood is cerebral palsy (CP), a group of motor disorders characterized by activity limitations caused by motor impairments.4 Improving fitness might be an important issue to consider when aiming to lower the impact of the disorder on activities of daily living, and to improve general health in persons with CP.
There is accumulating evidence that children with CP have decreased aerobic fitness, anaerobic fitness, and muscle strength.5 Decreases in muscle strength and anaerobic fitness might be expected in light of the impaired muscle activation in CP, whereas aerobic fitness seems to be more closely associated with physical inactivity.5 While muscle strength training has received increasing attention over the years,6 interventions aimed at improving aerobic fitness have been undertaken less often.7 In particular, objective assessments that measure aerobic fitness directly, such as peak oxygen uptake, the anaerobic threshold, oxygen pulse, and ventilatory factors have been undertaken less frequently following interventions in children with CP.8, 9 More recently, it was suggested that more high-intensity short duration components be included in fitness training.10 This type of training better suits the activity patterns of children and is promising with regard to improving aerobic and anaerobic fitness.10 However, a recent multi-component intervention including fitness training with strength and anaerobic exercises, combined with a lifestyle intervention, showed no mean group effects on physical fitness.11 These randomized controlled trial results have been published in a previous article.11 In both groups a large interindividual variation in changes in physical fitness was found, indicating that some participants improved, while others deteriorated despite their group allocation. The present article presents a secondary analysis of the fitness and mobility capacity data of both the intervention and control group. A secondary analysis of these data allows investigation of associations between changes in physical fitness which might have been averaged out in the group mean values, and also allows subgroup differentiation in these associations.12
A better understanding of the factors limiting maximal exercise provides guidance in setting up training programmes aimed at improving fitness in children with CP. A recent cross-sectional study suggested that aerobic and anaerobic fitness are limited by reduced muscle strength.13 This was indicated by the stronger correlations between these fitness components in adults with CP than in controls.13 Because of the differences in coordination and muscle function between unilateral and bilateral CP, a different association among fitness components might also be expected in these two groups. The strong cross-sectional correlations in the previously mentioned study may be influenced by the large interindividual variability of fitness levels in participants with CP. This cross-sectional association does not consider the association between changes in fitness components. Therefore, longitudinal associations might provide better insights into whether a change in one fitness component is associated with a change in the other.14
Associations of muscle strength and anaerobic fitness with mobility capacity were shown in cross-sectional studies.15 However, strength training studies showed that improvements in muscle strength did not necessarily lead to increased mobility capacity.16, 17 Moreover, evidence for better mobility capacity as a result of improved anaerobic and aerobic fitness is even more sparse.18 The above-mentioned differing associations in unilateral and bilateral CP may also apply to associations between fitness and mobility capacity. Stronger associations among fitness components and between fitness and mobility capacity are expected for children with a bilateral CP than for children with a unilateral CP, because children with a bilateral CP are more limited by muscle strength and coordination. The aim of this study was to investigate the longitudinal associations between changes in fitness components aerobic fitness, anaerobic fitness, and muscle strength, and between fitness and mobility capacity in children with CP.
Method
Participants
This study is a secondary analysis of a 6-month single-blinded, randomized controlled trial, with a 6-month follow-up, which was aimed at assessing the effects of a physical activity stimulation programme (LEARN 2 MOVE 7–12y study).11 The intervention programme included physical fitness training, counselling, and home-based physiotherapy while the control group continued their regular physical therapy. Children with spastic CP were recruited in special schools for children with disabilities and paediatric physiotherapy practices in the Netherlands, between September 2009 and February 2012. The intervention programme included 49 children with CP. A total of 46 children completed the study. Children were included when the following criteria were met: 7 to 13 years old; Gross Motor Function Classification System (GMFCS) level I, II, or III; no history of botulinum toxin injections and/or serial casting in the past 3 months or surgery in the past 6 months; no contraindications for maximal exercise; and less active than the international physical activity norm – no regular participation in sports or (physiotherapeutic) fitness programme, and experience of problems related to daily life mobility or sports.11 The institutional ethics committee of the VU University Medical Center approved this study and all parents and participants above 12 years of age signed an informed consent agreement.
Study design
This study included measurements at baseline, 4, 6, and 12 months of both the intervention and the control group. Within each child the four measurements were performed at approximately the same time of the day. A measurement session took place at the outpatient unit of a university medical centre and included determination of anthropometry, physical fitness, and mobility capacity. The children were instructed not to eat or drink (except for water) 1.5 hours before the measurement session. All measurements, except for the walking test, were performed in a laboratory with a temperature between 19°C and 22°C and a relative humidity of 41 to 50%. Fitness testing consisted of a maximal aerobic exercise test, an anaerobic Wingate test, and isometric strength testing, after which mobility capacity was determined by measuring gross motor function and walking capacity.
Outcome measures and materials
Anthropometry
Measures of body height (m) and weight (kg) were determined using an electronic scale (DGI 250D, KERN DE Version 3.3 10/2004; Kern & Sohn GmbH, Balingen-Frommern, Germany). Body mass index (BMI [kg/m2]) was calculated. Skinfold was measured at the supra-iliac and the sub-scapular site using a Holtain skinfold caliper (ProCare BV, Groningen, the Netherlands) with an accuracy of 0.2mm, providing a summed score of skinfold thickness (mm).
Physical fitness
Physical fitness comprised aerobic fitness, anaerobic fitness, and isometric muscle strength. Aerobic fitness was determined during a maximal aerobic exercise test performed on a cycle ergometer (Corival V2; Lode B.V., Groningen, the Netherlands). The protocol was based on the McMaster protocol and consisted of a resting period enabling habituation to the mask used for gas analysis, 3 minutes of warming-up, 5 minutes of submaximal cycling, and a maximal phase with the load increased each minute. The load increments were adapted to match the abilities of the child.5 Respiratory gas was collected breath-by-breath with a gas analysis system Quark CPET version 9.1b (Cosmed, S.r.l., Rome, Italy). Corresponding software was calibrated with a known mixture (ambient air and a calibration gas of 16% O2 and 5% CO2) before testing. Heart rate (HR) was measured with a heart rate monitor (Cosmed, S.r.l., Rome, Italy). Maximal aerobic parameters were calculated over the 30s with the highest sustained load. Achievement of maximal aerobic exercise was checked using objective criteria: HR >180 or respiratory exchange ratio >1.00, and if participant exhaustion was present.5 Maximal aerobic parameters were peak aerobic capacity (VO2peak), peak O2pulse (VO2/HR[mL/beat]) and peak ventilation (VEpeak [L/min]). The anaerobic threshold was determined with the V-slope method by two independent raters. The test–retest reliability of this protocol for VO2peak was shown to be excellent in children with CP, with an intraclass correlation coefficient (ICC) of 0.94 and a standard error of measurement (SEM) of 2.06mL/kg/min.19
Anaerobic fitness was measured using the 20 seconds Wingate anaerobic cycling test, which is a sprint test against a constant breaking torque. The test was performed for 20s, providing the delivered mean power over 20s (P20mean [W/kg]), which was used for analyses. Reliability assessment of P20mean using this protocol showed high ICCs of 0.96 to 0.99 and SEMs of 0.148 to 0.270W/kg.20 Isometric muscle strength of the knee extensors and hip abductors of the most-affected leg was determined via hand-held dynamometry. The ‘make test’ was performed with the child pushing against the dynamometer (MicroFet; Biometrics, Almere, the Netherlands) for 3s with maximal force, while the leg was fixed by the assessor according to standardized procedure.21 Strength (N) was multiplied by the moment arm resulting in the torque (Nm), and the mean of three measurements was calculated. Intersession reliability was found to be good with ICCs >0.82 and SEMs of 0.55N/kg (knee extensors) and 1.21N/kg (hip abductors).21 Measurements were performed by the same assessor in each measurement session.
Mobility capacity
Gross motor function was determined with the Gross Motor Function Measure item sets (GMFM-IS).22 The Gross Motor Ability Estimator (GMAE) was used to calculate the GMAE score. Walking ability was assessed using the 1-minute walk test. Children were asked to walk as fast as possible, without running, on an oval walkway for 1 minute. Covered distance (m) was registered. This test was reliable in children with CP with an ICC of 0.94 and limits of agreement of 13.1m.
Statistics
Distribution of the data was checked using visual inspection of the data (mean values, SDs, and ranges) and using inspection of the histogram and normal P–P plot of the residuals of the mixed model. Since the data were normally distributed, a parametric test was applied. Longitudinal associations among fitness components and between fitness components and mobility capacity were determined using a mixed model with fixed effects (mobility capacity as dependent variables and fitness components as independent variables) and a random intercept. This linear regression analysis corrects for the dependency of observations within one individual by estimating a random intercept and slope (if this shows a better fitting model).14 The regression coefficient derived from this analysis combines the between-participant and within-participant association and represents the association between the independent and dependent variable.14 Effect modification was investigated to determine whether the associations were different for children with a different GMFCS level, for unilateral or bilateral involvement and group allocation: intervention or control group. To assess determinants of mobility capacity a multiple random coefficient regression model was applied, using a forward selection procedure. Fitness variables that were significantly associated with mobility capacity (p<0.05) in the univariate analysis were included as determinants in the multiple regression model. Variables with the strongest association were added to the model first and included if the level of significance was p<0.05. Data on walking capacity was corrected for height if the regression coefficient changed >10%. Analyses were carried out using IBM spss Statistics, version 20 (SPSS Inc., Chicago, IL, USA).
Results
Participants
Characteristics of the participants are listed in Table 1.
| All children (n=46) | Bilateral CP (n=24) | Unilateral CP (n=22) | |
|---|---|---|---|
| Male/female (%) | 26/20 (57/43) | 12/12 (50/50) | 14/8 (64/36) |
| Mean age, y (SD) | 9y 7mo (1y 8mo) | 9y 6mo (1y 4mo) | 9y 8mo (2y) |
| Height, cm (SD) | 136.8 (12.4) | 133.3 (8.6) | 140.7 (14.7) |
| Weight, kg/m2 (SD) | 34.8 (11.1) | 31.7 (7.9) | 38.2 (13.1) |
| BMI, kg/m (SD) | 18.2 (3.3) | 17.7 (3.3) | 18.7 (3.1) |
| Skinfold, mm (SD) | 26.3 (10.3) | 26.0 (10.9) | 26.7 (9.7) |
| GMFCS level (I/II/III) | 26/12/8 | 8/8/8 | 18/4/0 |
- CP, cerebral palsy; BMI, body mass index; GMFCS, Gross Motor Function Classification System.
Longitudinal associations
Table 2 shows the indicators of fitness and mobility capacity. Significant interactions showed that associations among fitness components and between fitness components and mobility capacity were different for children with unilateral involvement than for those with a bilateral involvement. There was no interaction effect for GMFCS level and group allocation, indicating that the association between fitness and mobility were similar for different GMFCS levels and for the intervention and control group. Therefore, indicators and associations were presented for the whole group, and separately by unilateral or bilateral involvement (Tables 2, 3, and 4).
| Mean (SD) | Mean (SD) difference | |||||||
|---|---|---|---|---|---|---|---|---|
| Baseline | n | T4-T0 | n | T6-T0 | n | T12-T0 | n | |
| Mobility capacity | ||||||||
| GMAE score | ||||||||
| Overall | 78.8 (13.8) | 46 | na | 0.4 (4.7) | 46 | 1.6 (3.8) | 44 | |
| Bilateral | 72.5 (15.6) | 24 | na | 1.1 (5.0) | 24 | 2.2 (4.0) | 23 | |
| Unilateral | 85.6 (7.1) | 22 | na | −0.4 (4.4) | 22 | 0.9 (3.6) | 21 | |
| Walking capacity, m | ||||||||
| Overall | 89.0 (20.3) | 46 | 3.2 (8.0) | 44 | 4.7 (10.0) | 45 | 3.5 (10.2) | 42 |
| Bilateral | 77.5 (19.3) | 24 | 3.5 (6.8) | 24 | 4.2 (10.5) | 23 | 2.7 (11.8) | 23 |
| Unilateral | 101.6 (12.4) | 22 | 2.8 (9.3) | 20 | 5.2 (9.6) | 22 | 4.5 (8.1) | 19 |
| Fitness | ||||||||
| Aerobic fitness | ||||||||
| VO2peak, mL/kg/min | ||||||||
| Overall | 31.4 (6.2) | 38 | 2.0 (4.7) | 32 | 1.7 (4.3) | 33 | 2.2 (4.4) | 34 |
| Bilateral | 29.0 (6.3) | 18 | 3.1 (5.3) | 15 | 1.5 (4.0) | 13 | 2.8 (4.7) | 16 |
| Unilateral | 33.5 (5.4) | 20 | 1.0 (3.9) | 17 | 1.8 (4.6) | 20 | 1.6 (4.2) | 18 |
| Anaerobic threshold, mL/kg/min | ||||||||
| Overall | 16.8 (4.7) | 41 | 2.9 (4.6) | 38 | 1.6 (4.8) | 39 | 0.9 (5.0) | 39 |
| Bilateral | 15.5 (5.2) | 20 | 3.7 (5.3) | 20 | 1.6 (5.6) | 18 | 1.9 (5.8) | 19 |
| Unilateral | 18.1 (3.9) | 21 | 2.0 (3.5) | 18 | 1.7 (4.0) | 21 | −0.1 (4.1) | 20 |
| O2 pulse, mL/beat | ||||||||
| Overall | 5.9 (2.3) | 37 | 0.6 (0.6) | 31 | 0.8 (0.7) | 32 | 1.1 (0.8) | 33 |
| Bilateral | 5.0 (1.4) | 18 | 0.7 (0.7) | 15 | 0.6 (0.5) | 13 | 0.9 (0.7) | 16 |
| Unilateral | 6.8 (2.6) | 19 | 0.6 (0.6) | 16 | 0.9 (0.8) | 19 | 1.2 (0.8) | 17 |
| VE/VO2 | ||||||||
| Overall | 42.2 (7.9) | 38 | −1.5 (6.7) | 32 | −1.9 (4.1) | 33 | −3.9 (7.2) | 34 |
| Bilateral | 44.4 (9.0) | 18 | −2.8 (7.6) | 15 | −1.8 (4.3) | 13 | −5.4 (8.4) | 16 |
| Unilateral | 40.3 (6.3) | 20 | −0.3 (5.8) | 17 | −2.0 (4.0) | 20 | −2.5 (5.9) | 18 |
| Anaerobic fitness | ||||||||
| P20mean, W/kg | ||||||||
| Overall | 3.5 (1.5) | 44 | −0.01 (0.43) | 42 | −0.08 (0.64) | 43 | 0.02 (0.59) | 42 |
| Bilateral | 3.0 (1.6) | 23 | −0.03 (0.43) | 23 | −0.16 (0.50) | 23 | 0.14 (0.46) | 22 |
| Unilateral | 4.1 (1.3) | 21 | 0.01 (0.44) | 19 | 0.01 (0.77) | 20 | −0.12 (0.70) | 20 |
| Muscle strength | ||||||||
| Knee ext, Nm/kga | ||||||||
| Overall | 1.17 (0.31) | 46 | 0.04 (0.23) | 44 | 0.03 (0.26) | 46 | 0.01 (0.25) | 44 |
| Bilateral | 1.13 (0.34) | 24 | 0.04 (0.24) | 24 | 0.00 (0.23) | 24 | −0.05 (0.28) | 23 |
| Unilateral | 1.21 (0.28) | 22 | 0.04 (0.22) | 20 | 0.06 (0.30) | 22 | 0.06 (0.20) | 21 |
| Hip abd, Nm/kga | ||||||||
| Overall | 0.84 (0.27) | 46 | −0.01 (0.19) | 44 | 0.05 (0.19) | 46 | −0.02 (0.17) | 44 |
| Bilateral | 0.76 (0.31) | 24 | 0.03 (0.15) | 24 | 0.06 (0.13) | 24 | −0.01 (0.15) | 23 |
| Unilateral | 0.92 (0.21) | 22 | −0.06 (0.23) | 20 | 0.04 (0.25) | 22 | −0.04 (0.19) | 21 |
- a Of the non-dominant leg. GMAE, Gross Motor Ability Estimator; na, not assessed; knee ext, isometric knee extensor muscle strength; hip abd, isometric hip abductor muscle strength; P20mean, mean anaerobic power; VO2peak, peak aerobic capacity.
| Anaerobic fitness | Aerobic fitness | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| VO2peak, mL/kg/min | Anaerobic threshold, mL/kg/min | O2 pulse, mL/beat | VE/VO2 | |||||||||
| B (SE) | 95% CI | p | B (SE) | 95% CI | p | B (SE) | 95% CI | p | B (SE) | 95% CI | p | |
| P20mean, W/kg | ||||||||||||
| Overall | 1.79 (0.48) | 0.85 to 2.73 | <0.001 | 0.93 (0.34) | 0.26 to 1.60 | 0.008 | 0.22 (0.14) | −0.05 to 0.49 | 0.112 | −0.15 (0.56) | −1.27 to 0.97 | 0.791 |
| Bilateral | 3.31 (0.67) a | 1.98 to 4.64 | <0.001 | 1.10 (0.47) | 0.16 to 2.05 | 0.023 | 0.43 (0.21 | 0.00 to 0.85 | 0.049 | 0.25 (0.83) | −1.40 to 1.90 | 0.766 |
| Unilateral | 0.52 (0.65) | −0.77 to 1.81 | 0.429 | 0.49 (0.57) | −0.63 to 1.61 | 0.388 | 0.05 (0.18) | −0.30 to 0.41 | 0.760 | −0.07 (0.84) | −0.17 to 1.59 | 0.933 |
| Muscle strength | ||||||||||||
| Knee ext, Nm/kgb | ||||||||||||
| Overall | 1.40 (1.52) | −1.60 to 4.41 | 0.357 | 2.52 (1.26) | 0.02 to 5.02 | 0.048 | −0.01 (0.36) | −0.72 to 0.71 | 0.987 | 3.97 (1.84) | 0.32 to 7.62 | 0.033 |
| Bilateral | 3.00 (2.07) | −1.09 to 7.10 | 0.149 | 2.56 (1.67) | −0.74 to 5.86 | 0.127 | −0.37 (0.50) | −1.36 to 0.62 | 0.456 | 4.93 (2.46) | 0.07 to 9.79 | 0.047 |
| Unilateral | −0.11 (2.13) | −4.34 to 4.11 | 0.958 | 1.93 (1.93) | −0.19 to 5.74 | 0.320 | 0.35 (0.49) | −0.62 to 1.32 | 0.478 | 3.28 (2.67) | −1.99 to 8.55 | 0.221 |
| Hip abd, Nm/kgb | ||||||||||||
| Overall | 3.96 (1.95) | 0.11 to 7.81 | 0.044 | 2.30 (1.68) | −1.02 to 5.62 | 0.173 | 0.15 (0.46) | −0.76 to 1.07 | 0.741 | 5.63 (2.41) | 0.87 to 10.39 | 0.021 |
| Bilateral | 5.38 (3.08) | −0.71 to 11.48 | 0.083 | 3.03 (2.31) | −1.55 to 7.61 | 0.192 | −0.19 (0.79) | −1.74 to 1.37 | 0.814 | 7.97 (3.52) | 1.00 to 14.75 | 0.025 |
| Unilateral | 3.11 (2.56) | −1.95 to 8.17 | 0.226 | 0.71 (2.52) | −4.27 to 5.69 | 0.778 | 0.34 (0.58) | −0.80 to 1.48 | 0.557 | 4.34 (3.31) | −2.20 to 10.89 | 0.192 |
| Anaerobic fitness | |||
| P20mean, W/kg | |||
| Muscle strength | B (SE) | 95% CI | p |
| Knee ext, Nm/kgb | |||
| Overall | 0.38 (0.19) | 0.00 to 0.76 | 0.051 |
| Bilateral | 0.37 (0.26) | −0.14 to 0.88 | 0.150 |
| Unilateral | 0.44 (0.28) | −0.11 to 0.99 | 0.117 |
| Hip abd, Nm/kgb | |||
| Overall | 0.19 (0.24) | −0.29 to 0.67 | 0.431 |
| Bilateral | 0.52 (0.38) | −0.24 to 1.28 | 0.179 |
| Unilateral | 0.07 (0.33) | −0.58 to 0.72 | 0.839 |
- aThe magnitude of this regression coefficient is twofold: a difference of 1W/kg P20mean is associated with an increase of 3.31mL/kg/min in VO2peak: (1) between participants; and (2) within participants. bOf the non-dominant leg. Significant results are marked in bold text. B, unstandardized regression coefficient; SE, standard error; CI, confidence interval; knee ext, isometric knee extensor muscle strength; hip abd, isometric hip abductor muscle strength; VO2peak, peak aerobic capacity.
| GMAE | Walking capacity | |||||
|---|---|---|---|---|---|---|
| B (SE) | 95% CI | p | B (SE) | 95% CI | p | |
| Aerobic fitness | ||||||
| VO2peak, mL/kg/min | ||||||
| Overall | 0.063 (0.124) | −0.184 to 0.309 | 0.613 | 0.32 (0.20) a | −0.08 to 0.71 | 0.112 |
| Bilateral | 0.295b (0.143) | 0.011 to 0.579 | 0.042 | 0.52 (0.22) a | 0.09 to 0.96 | 0.019 |
| Unilateral | −0.208 (0.167) | −0.541 to 0.125 | 0.217 | 0.23 (0.27)a | −0.30 to 0.76 | 0.395 |
| Anaerobic threshold, mL/kg/min | ||||||
| Overall | 0.023 (0.109) | −0.193 to 0.239 | 0.832 | 0.19 (0.19) | −0.17 to 0.56 | 0.299 |
| Bilateral | 0.099 (0.143) | −0.186 to 0.384 | 0.491 | 0.34 (0.24) | −0.14 to 0.81 | 0.164 |
| Unilateral | −0.106 (0.160) | −0.424 to 0.213 | 0.512 | −0.04 (0.27) | −0.56 to 0.49 | 0.895 |
| Anaerobic fitness | ||||||
| P20mean, W/kg | ||||||
| Overall | 2.75 (0.67) | 1.43 to 4.07 | <0.001 | 5.4 (1.1) a | 3.2 to 7.5 | <0.001 |
| Bilateral | 5.14 (0.90) | 3.34 to 6.94 | <0.001 | 7.9 (1.3) a | 5.4 to 10.5 | <0.001 |
| Unilateral | 1.54 (0.91) | −0.26 to 3.34 | 0.093 | 2.5 (1.5)a | −0.46 to 5.4 | 0.097 |
| Muscle strength | ||||||
| Knee ext, Nm/kgc | ||||||
| Overall | 3.25 (1.86) | −0.43 to 6.94 | 0.083 | 3.5 (3.3)a | −3.1 to 10.1 | 0.294 |
| Bilateral | 7.88 (2.49) | 2.93 to 12.82 | 0.002 | 8.4 (4.3)a | −0.1 to 17.0 | 0.053 |
| Unilateral | −0.97 (2.61) | −6.14 to 4.21 | 0.712 | 1.3 (4.9)a | −8.4 to 11.1 | 0.788 |
| Hip abd, Nm/kgc | ||||||
| Overall | 1.01 (2.47) | −3.89 to 5.92 | 0.683 | 12.4 (4.2) a | 4.1 to 20.7 | 0.003 |
| Bilateral | 7.65 (4.01) | −0.29 to 15.58 | 0.059 | 27.7 (5.8) a | 16.1 to 39.2 | <0.001 |
| Unilateral | −0.77 (3.35) | −7.44 to 5.90 | 0.820 | 5.1 (5.9)a | −6.5 to 16.7 | 0.386 |
- aOf the non-dominant leg. bThe magnitude of this regression coefficient is twofold: a difference of 1mL/kg/min VO2peak is associated with an increase of 0.295 GMAE score: (1) between participants; and (2) within participants. cCorrected for height. Significant results are marked in bold text. GMAE, Gross Motor Ability Estimator; B, unstandardized regression coefficient; SE, standard error; CI, confidence interval; knee ext, isometric knee extensor muscle strength; hip abd, isometric hip abductor muscle strength; VO2peak, peak aerobic capacity.
Associations between fitness components
Table 3 shows the associations among the fitness components: aerobic fitness, anaerobic fitness, and muscle strength. In children with bilateral CP, aerobic fitness parameters (VO2peak [p<0.001] and anaerobic threshold [p=0.023] and O2pulse [p=0.049]) were associated with anaerobic fitness, while VE/VO2 was the only aerobic parameter that showed an association with muscle strength (knee extensors [p=0.047] and hip abductors [p=0.025]). Anaerobic fitness was not associated with muscle strength. In children with unilateral CP, no associations among fitness components were found.
Associations with mobility capacity
Fitness components and mobility capacity were not associated in children with a unilateral CP (Table 4). The variables that were significantly associated with mobility capacity in children with a bilateral CP were used to determine which variables were to be included in the multivariate models. Figure 1 presents the association between VO2peak and the fixed predicted GMAE values to illustrate the clinical significance of the size of the regression coefficient. The multivariate models showed that P20mean and knee extensor strength (regression coefficient [SE] 3.7 [1.1], p=0.001 and 9.1 [3.0], p=0.004 respectively) were the most important determinants of gross motor function in bilateral CP. For walking capacity, P20mean and hip abductors (5.9 [1.5], p<0.001 and 15.2 [6.2], p=0.016 respectively) were the most important determinants in bilateral CP. In children with a unilateral CP, no fitness components were determinants for mobility capacity.
Discussion
This was the first study to investigate the longitudinal associations among fitness components aerobic fitness, anaerobic fitness, and muscle strength; and between fitness and mobility capacity in children with CP. The results showed that in children with bilateral CP, changes in aerobic fitness were strongly associated with changes in anaerobic fitness, and changes in aerobic fitness were associated with changes in muscle strength. In children with a unilateral CP, no longitudinal associations between fitness components were found. Changes in anaerobic fitness, and, to a lesser extent, changes in muscle strength and aerobic fitness, determined mobility capacity in children with a bilateral CP, while none of the fitness components were associated with mobility capacity in children with a unilateral CP.
Anaerobic fitness was positively and strongly associated with aerobic fitness in children with a bilateral CP. It appears that low anaerobic capacity might limit aerobic fitness, while improving capacity to achieve short-duration high-intensity exercise might contribute to a higher peak oxygen uptake. This strong association was also found in a cross-sectional study undertaken in adults with CP, with a weaker association found in adults without CP.13 The associations between changes in aerobic and anaerobic fitness concurs with the finding that aerobic fitness also improved through anaerobic training in children with CP,18 suggesting that anaerobic training also includes aerobic training components. In our study, the longitudinal component of the regression coefficient, along with the contribution of a cross-sectional share, is indicative of the actual association in changes between parameters. In more detail, when exercise is performed with a gradual increase in intensity, as in the maximal aerobic exercise test, the oxygen debt in the energy requirement that occurs with every increase in load requires an anaerobic contribution to achieve a new steady state.23 Therefore, for children with physical activity patterns that are characterized by short, intermittent activities, anaerobic training programmes might also have the potential for improving aerobic fitness, which might eventually be beneficial to general health.1, 2
Muscle strength appears to be a less important limiting factor for aerobic and anaerobic fitness, as indicated by weaker statistical associations. This corresponds to the weak associations reported in adults with CP for isometric knee muscle strength of the more impaired leg and aerobic and anaerobic fitness,13 although the associations for muscle strength of the less impaired leg were stronger.13 However, these previous conclusions were based on a combined sample of adults with unilateral or bilateral CP. As indicated in our study, results may differ for persons with CP with a different anatomical involvement. In addition, our study included prepubertal children for whom muscle characteristics differ from adults because of hormonal factors.24 Our results show that muscle strength relates to VE/VO2 in children with bilateral CP. This might be indicative of better ventilation efficiency caused by changes in muscle volume, which could be beneficial in terms of muscle strength and oxidative capacity.25
The lack of associations among fitness components in children with a unilateral CP in this study cannot be compared to previous studies as these associations have not previously been evaluated for these subgroups separately. However, in adults with CP, a stronger correlation among fitness components was shown compared to those without CP.13 This finding was attributed to the limiting effect of reduced muscle strength in CP on maximal aerobic capacity. The stronger correlations found for the less impaired leg might indicate that the muscle strength of the less impaired leg is more important for cycling performance. This might explain the lack of associations between muscle strength of the more impaired leg in this study of children with a unilateral CP, with the least affected leg being more strongly associated with aerobic and anaerobic performance on a bicycle.
The magnitude of the longitudinal association between anaerobic fitness and mobility is that 1W/kg improvement in anaerobic capacity is associated with an improvement of 5.14 points on the GMAE score. A previous study showed an increase of 25% anaerobic capacity after fitness training, which would correspond to an increase of a 3.86 point GMAE score for children with a bilateral CP.18 This indicates that improving anaerobic fitness might lead to a better mobility capacity in children with a bilateral CP. This confirms the suggestion of Bar-Or et al.3 that peak mechanical power (i.e. anaerobic fitness) is a more important determinant of the ability to execute activities than aerobic fitness. The findings are in agreement with the positive correlations between running sprint power and gross motor function found in a cross-sectional study including children with unilateral or bilateral CP15 and with the increase in gross motor function following combined aerobic and anaerobic fitness training.18
Regarding muscle strength, knee extensors were associated with a better gross motor function, confirming previous findings.25 The stronger association between hip abductors and walking ability confirms earlier findings on this association.26 In conclusion, improving gross motor function through fitness training focused on increasing anaerobic power, and, to a lesser extent, through increasing muscle strength, seems to have potential in children with a bilateral CP.
Current findings suggest that increasing anaerobic power might not contribute to improved gross motor function and walking capacity in children with a unilateral CP, as indicated by the lack of an association in this group. The different associations in children with unilateral and bilateral CP might explain the fact that previous strength training studies did not find that gross motor function and walking capacity improved with increased knee extension and hip abduction strength, since no distinction was made between children with bilateral or unilateral CP in these studies.16, 17 A previous study found better gross motor function and walking capacity in children with a unilateral CP than in children with a bilateral CP within a given GMFCS level and suggested these components be evaluated separately for unilateral and bilateral CP.27 While improvements in mobility capacity might be expected following anaerobic fitness training in children with a bilateral CP, this is not expected for children with a unilateral CP. A different intervention is required for improving mobility capacity in children with either a unilateral or a bilateral CP.
Limitations
A limitation of this study was that isometric muscle strength of the dominant leg was not measured, while the dominant leg might determine aerobic and anaerobic performance. Another limitation was that only two muscle groups were included in the measurements. Nevertheless, the significant associations between strength and mobility capacity in children with bilateral involvement indicate that strength may be a limiting factor for mobility in this population. Further research is required to determine whether these associations also hold for other muscle groups.
Conclusion
The strong longitudinal associations between aerobic and anaerobic fitness indicate that increasing either aerobic or anaerobic fitness is associated with improvements in the other fitness component in children with bilateral CP only. While increasing anaerobic fitness might be beneficial for mobility capacity in children with bilateral CP, this is less likely for children with unilateral CP. Different interventions are required for improving mobility capacity in children with either a unilateral or a bilateral CP.
Acknowledgements
This study was supported by a grant from The Netherlands Organization for Health Research and Development (ZonMw) and the Phelps Foundation for Spastics. They were not involved in the design of the study, data collection, data analysis, manuscript preparation, and publication decisions. The authors would like to acknowledge Prof. Dr J Twisk for his advice on statistical analysis, Kim van Hutten for her help with data collection, and Joyce van Tunen for her help with data analysis. The authors have stated that they had no interests that might be perceived as posing a conflict or a bias.
