Physical Activity Measurement Using MTI (Actigraph) Among Children With Cerebral Palsy
,
,
Institute of Human Performance, University of Hong Kong, Hong Kong, China
Abstract
Capio CM, Sit CH, Abernethy B. Physical activity measurement using MTI (Actigraph) among children with cerebral palsy.
Objective
To investigate the validity of MTI accelerometer as a physical activity (PA) measurement instrument for children with cerebral palsy (CP).
Design
Participants were classified within Gross Motor Function Classification System I to III and took part in 2 activity sessions: (1) a structured activity protocol with increasing intensities and (2) a free play session. Concurrent measurements of activity counts, heart rate, and observed physical activity were performed.
Setting
Data were collected on normal school days in special schools within the participants' 30-minute break period.
Participants
Convenience sample of children with CP (N=31; 17 girls, 14 boys) age between 6 and 14 years (mean ± SD, 9.71±2.52y).
Interventions
Not applicable.
Main Outcome Measures
MTI measured activity counts, a monitoring device measured heart rate, and the System for Observing Fitness Instruction Time (SOFIT) was used for direct PA observation.
Results
There were strong relationships between MTI and SOFIT (r=.75; R2=.56; P<.001) and heart rate monitor (HRM) and SOFIT (r=.65; R2=.43; P<.001) data in structured activities, but the difference between these 2 correlation coefficients was not significant (P=.46). In free play activities, the association between MTI and SOFIT data (r=.67; R2=.45; P<.001) was significantly stronger (P=.01) than that between heart rate and SOFIT data (r=.14; R2=.02; P<.001) . Bland-Altman plots showed better agreement between observed SOFIT and MTI-predicted SOFIT data than observed SOFIT and HRM-predicted SOFIT data from the linear regression analysis.
Conclusions
The findings suggest that the MTI appears to be a valid instrument for measuring raw activity volume among children with CP and is suitable for use in studies attempting to characterize the PA of this population.
List of Abbreviations:
CP (cerebral palsy), GMFCS (Gross Motor Function Classification System), HRM (heart rate monitor), MVPA (moderate-to-vigorous physical activity), PA (physical activity), SOFIT (System for Observing Fitness Instruction Time)
CEREBRAL PALSY IS caused by an injury to the developing brain and is known to manifest in problems of movement and posture.1 Children with CP form the largest diagnostic group treated in pediatric rehabilitation.2 The hallmark of CP is disorder in the advancement of gross motor function.3 The delayed motor development in children with CP may have an impact on their participation in physical activities. PA is considered one of the most basic human functions4 and is importantly associated with health benefits5 in all children, including those with disabilities.6 Low levels of PA are bigger concerns for children with disabilities because they appear to be at risk for more health problems associated with their disabilities.7 In particular, the biological factors associated with physical disability have been seen to result in delay in fundamental movement skills development,8 which can lead children with CP to choose a more sedentary lifestyle to avoid movement difficulties.9
Children with CP are known to be less physically active and have lower levels of physical fitness relative to their peers without disabilities.10, 11, 12 Without adequate fitness capacity, children with CP may not be able to achieve their maximum potentials in motor activities and consequently have limited engagement in daily activities.10 This concern needs to be addressed because it may contribute to the development of secondary conditions such as chronic pain, fatigue, and osteoporosis.13 In relation to this, having valid and reliable measures of PA among children with CP is critical in order to estimate PA participation, monitor compliance to recommendations, and quantify dose-response relationships between interventions and outcomes.14
There has been a paucity of research addressing PA measurement among children with CP. The assessment of PA in this population is of critical importance to the design and implementation of health, therapy, and physical education programs.15, 16 A review of studies examining PA among children with CP shows that measurement methods tend to be frequently subjective and in the form of questionnaires.10, 11, 12, 17 Objective measurements have been verified in some studies but are generally limited to step measurements18 and measurement of time spent in upright positions.16
Direct observation is accepted as the most appropriate criterion measure of pediatric PA behavior.19 Other objective instrumentation such as heart rate monitoring and accelerometry has been used to assess PA during structured and unstructured settings in the general pediatric population.20 Reviews of PA assessment strategies pointed out that direct observation is useful for PA assessment in a small sample and in a confined space, and when observations are done over short periods.21, 22 It has also been identified to have adequate precision and is therefore considered an appropriate criterion for validity studies of other PA measurement instruments.21, 23, 24 However, compared with other objective measures, direct observation may not be suitable for PA assessment in free living conditions such as home settings because of invasion of privacy, and it is time-intensive and labor-intensive depending on the number and ability of raters.15, 21
In contrast, the use of heart rate monitoring is unobtrusive, is relatively inexpensive, and can be used in small to moderate sample sizes in free living conditions.25 As an indicator of PA, it assumes a linear relationship between heart rate and oxygen consumption, providing a valid and reliable estimate of activity patterns.26 However, heart rate may not necessarily reflect PA behavior because it may be influenced by factors other than body movement.19 Furthermore, heart rate responses tend to lag behind changes in movement patterns, particularly in children and adolescents.23
Accelerometers quantify 1 or more dimensions of movement of the body part to which they are attached.21 Because they are both nonreactive and reuseable, accelerometers have been commonly used as an objective method for PA assessment in children.19 Accelerometers have also been shown to provide a practical, reliable, and valid means of measuring the amount and intensity of PA in free living conditions.18 Previous studies27, 28 have shown valid and reliable use of accelerometers in children with physical disabilities. The ambulatory PA of children with CP has been examined using a device to monitor steps.29 Although the device-based monitoring of step activity has been shown to have excellent accuracy (99.7%) with manual step counts in a sample of children with CP, no evidence of its reproducibility has been reported.30 Currently, the MTI accelerometer has been identified to be the most widely used motion sensor in PA research with the most evidence supporting its feasibility, reliability, and validity in children.20, 30 However, only 1 study has been found that supports the use of MTI to measure PA of children with CP.31 This study demonstrated good test-retest reliability for MTI among children with CP but without assessment of the validity of the instruments for use with this population. Evidence has therefore been established to support the use of MTI among children, but verification of its suitability for use among children with CP is still very limited. Because of their potentially different activity patterns,31 lower fitness levels,14,15 and aberrations in movement associated with spasticity,2 it cannot be assumed that the methodologic properties of MTI established in other pediatric populations will generalize to children with CP.
This current study aimed to address limitations in the existing knowledge by investigating the suitability of the MTI for measuring PA levels among children with CP. Specifically, we collected MTI data concurrently with heart rate and direct observation data for a sample of children with CP in structured and free play conditions. We used tests of associations between the various measures to determine the criterion validity of MTI data while using direct observation as the standard for PA measurement. We hypothesized that if the MTI measurements are valid, the MTI data should manifest a positive linear relationship with the increasing intensity of the structured activities and should be able to account for a larger amount of variance in direct observation than heart rate data. In free play, the positive linear association between MTI and direct observation should be significantly stronger than that between heart rate and direct observation. Agreement of directly observed PA should also be better with a predicted set from regression of MTI data than with a predicted set from regression of heart rate data.
Methods
Participants
A convenience sample of 31 children with CP (17 girls, 14 boys) between ages of 6 and 14 years (mean ± SD, 9.71±2.52y) participated in the study. The sample size was based on a balance between the difficulty in procuring large samples of children with CP and previous suggestions that studies need at least 15 participants for each predictor variable in order to develop reliable regression equations.32 The participants were recruited from 2 special schools for children with physical disabilities. Children with CP in the participating schools, who met the following inclusion criteria were invited to join the study: (1) age between 6 to 14 years, (2) able to walk with or without walking aids, and (3) able to follow 2-step commands. Exclusion criteria included neurologic disease and any other medical conditions that limited participation. Participants included different subtypes of CP and were classified in the GMFCS as levels I to III. Table 1shows the characteristics of the participants. Both children and their parents provided written consent prior to study involvement. Ethical approval was granted by the institutional review board of the university.
| Characteristics | Total N=31 (%) | Girls n=17 (%) | Boys n=14 (%) |
|---|---|---|---|
| Classification of GMFCS | |||
| Level I | 14 (45.2) | 6 (35.3) | 8 (57.1) |
| Level II | 9 (29.0) | 6 (35.3) | 3 (21.4) |
| Level III | 8 (25.8) | 5 (29.4) | 3 (21.4) |
| Subtype of CP | |||
| Spastic diplegia | 13 (41.9) | 10 (58.8) | 3 (21.4) |
| Spastic triplegia | 1 (3.2) | 0 (0.0) | 1 (7.1) |
| Spastic quadriplegia | 6 (19.4) | 2 (11.8) | 4 (28.6) |
| Spastic hemiplegia | 3 (9.7) | 1 (5.9) | 2 (14.3) |
| Dyskinetic | 8 (25.8) | 4 (23.5) | 4 (28.6) |
Physical Activity Assessment
Uniaxial accelerometer (MTI)
The MTI accelerometera was initialized with a time stamp on a 15-second epoch, housed in a pouch, and positioned over the right hip, in line with the participant's midaxilla,18 using an elastic belt attached to the waistband of clothes. The MTI data were downloaded and stored on a computer using a proprietary interface and software. This particular device measures acceleration on the vertical plane and may be used without individual calibration.18 Previous studies in typically developing children have used a cutoff of greater than or equal to 2000 counts a minute to represent MVPA.23
System for Observing Fitness Instruction Time
SOFIT33 was used to record each child's activity levels and the amount of time the child spent in MVPA during the sessions. The child's activity level was scored every 15 seconds by entering 1 of 5 codes: lying down (code 1), sitting (code 2), standing (code 3), walking (code 4), or vigorous (code 5), where codes 4 and 5 represented MVPA. These codes have been validated using heart rate monitoring and accelerometry33, 34, 35 and have previously been used for measurement of PA levels during structured and/or unstructured settings at schools in children with intellectual disabilities7, 36, 37, 38 and children with physical disabilities.7 SOFIT has also been shown to characterize the frequency, intensity, and duration of PA in children with disabilities.15
Heart rate monitor
The Polar HRM,b a self-contained system worn around the chest, was used to detect and store each participant's heart rate every 5 seconds. The HRM data were downloaded via the device software. The 5-second epoch HRM data were screened for any nonphysiologic values (≥215 or ≤45 beats per minute). Using heart rate monitors for the assessment of PA may permit a valid and reliable estimate of activity patterns,34 and such instrumentation has been used among children with CP and shown to provide acceptable group estimates of energy expenditure.39, 40
Procedures
Participants were asked to engage in 2 activity sessions: (1) structured activity with a stepwise increase in intensity and (2) unstructured free play, indicative of the typical intermittent activity patterns in children in free living conditions.41 The sessions were conducted indoors, in an activity room with air-conditioning system to control the ambient environment, and with provisions allowing safe play such as nonslip flooring and rubber mats.
For the structured activity session, participants were divided into groups of 3, with each group engaging in one 12-minute session of movements. The 12-minute session consisted of six 2-minute activities of increasing intensities: (1) sitting, (2) standing, (3) standing with intermittent ball dribbling, (4) walking with intermittent ball dribbling in standing, (5) walking continuously, and (6) jogging continuously.41 Verbal instructions that consisted of 2-step commands and activity demonstrations were given prior to the start of the protocol. Participants were also given verbal and visual cues to switch to the next activity at the end of every 2 minutes.
For the unstructured free play, participants were divided into groups of 5, with each group engaging in one 10-minute session. Verbal instructions were given prior to the start of the session, and the participants were informed that they could engage in any game or play activity according to their preference. Similarly, the free play sessions were held indoors where equipment such as balls was available.
All data were collected on normal school days at the special schools within a 30-minute break period for the participants. All structured and unstructured activity sessions were video-recorded for reliability checks and data analyses. The first author served as the primary data collector, and the second author completed the reliability measures. Both were trained using the standardized SOFIT protocol, which included memorizing coding definitions and conventions, viewing videotaped segments, and surpassing the interobserver agreement of 85% on videotaped assessments before beginning data collection. The second author was responsible for reliability checks (20% of the total data), and the interobserver agreement for child activity levels exceeded 98%.
Data Analysis
The activity counts from the MTI and the heart rate recordings from the Polar monitors were downloaded using their respective software applications, and the videos were downloaded in digital format and were analyzed using the SOFIT system for observed PA levels. Codings were done, using a 15-second cycle which was temporally matched with the 15-second cycle of the data from the MTI. The heart rate recordings were initially recorded using a 5-second cycle, but only the recordings that were temporally matched with the 15-second cycle of the 2 other sets of data were used. In low to moderate levels of exercise, the heart rate reaches a steady state within 1 minute of constant work rate.42 As such, in the structured session, only the recordings for the 3 variables in the last 30 seconds of each 2-minute activity were analyzed. This resulted in 12 data points for each participant in the 12-minute structured activity session. With 31 participants, a total of 372 data points were analyzed. For the free play session, 40 data points for each participant were analyzed, resulting in a total of 1240 data points.
Data were analyzed using SPSS 16.0c and MedCalcd software. Descriptive statistics including means, SDs, frequencies, and percentages were obtained for all variables. The Pearson r coefficient and linear regression analyses were used to examine the relationships between the measured variables, and the difference between 2 independent correlation coefficients was compared using the Fisher z test. Using linear regression equations, sets of predicted SOFIT data based on the MTI (predicted SOFIT=2.5105+.0001157×MTI) and HRM (predicted SOFIT=.1160+.02665×HRM) data were generated for structured activity. A similar procedure was performed for the data in free play (predicted SOFIT=2.0722+.0002760×MTI; predicted SOFIT=2.8883+.004707×HRM). The observed and predicted sets of SOFIT data were then analyzed to determine the time (in minutes) spent in MVPA (codes 4 and 5). Bland-Altman plots were used to verify the linear associations by examining the level of agreement between observed (criterion) and predicted (comparison) time spent in MVPA for both structured and free play activities. This method plots the differences between the criterion and comparison data against their mean, where ±1.96 SD of the differences provides an interval within which 95% of the differences between the 2 sets of measurements are predicted to fall.43 Based on visual judgment, a higher level of agreement is illustrated by a smaller range of the 95% confidence interval of the differences between the measurements.44, 45 Because PA data were skewed, a log transformation was used for subsequent analyses that assumed normality. Significance level was set at P less than .05 for all statistical tests.
Results
Structured Activity
The structured activity session was designed to represent a stepwise increase in intensity of physical activities. Table 2shows the means and SDs for the MTI, HRM, and SOFIT outputs for the 6 protocol activities, with predictably greater activity outputs found in the higher levels of the structured activity protocol.
| Activity | MTI | HRM | SOFIT Activity Code | ||
|---|---|---|---|---|---|
| Counts | Natural Log | Beats per minute | Natural Log | ||
| Sitting | 849±741 | 2.76±.48 | 103.9±13.8 | 2.01±.06 | 2 |
| Standing | 4023±2686 | 3.52±.29 | 116.4±19.8 | 2.06±.07 | 3 |
| Standing with intermittent ball dribbling | 8522±4308 | 3.85±.32 | 122.7±15.0 | 2.08±.05 | 3 |
| Walking with intermittent ball dribbling in standing | 10,471±4277 | 3.97±.24 | 127.8±12.9 | 2.10±.04 | 4 |
| Walking continuously | 12,108±5576 | 4.02±.27 | 136.6±20.4 | 2.13±.06 | 4 |
| Jogging continuously | 15,358±3847 | 4.17±.14 | 154.5±21.2 | 2.19±.06 | 5 |
NOTE. SOFIT activity code: 1, lying; 2, sitting; 3, standing; 4, walking; 5, vigorous.
Results of the Pearson correlations and linear regression analyses indicated that all 3 activity measures increased across the higher levels of the structured activity protocol in a linear manner (fig 1). There were moderate to very strong relationships between protocol activity and MTI (r=.78; R2=.60; P<.001), HRM (r=.66; R2=.44; P<.001), and SOFIT data (r=.97; R2=.94; P<.001).
Fig 1
Linear relation of MTI, HRM, and SOFIT data to the 6 components of the structured activity protocol. Dashed lines represent a 95% confidence interval for the regression line. MTI: r=.78, R2=.60, P<.001; natural log of MTI: r=.75,R2=.57, P<.001. HRM: r=.66, R2=.44, P<.001; natural log of HRM: r=.67, R2=.45, P<.001. SOFIT: r=.97, R2=.94, P<.001; natural log of SOFIT: r=.95, R2=.91, P<.001.
Figure 2 demonstrates the relationship of MTI and HRM to SOFIT data. Using SOFIT as the criterion measure, MTI predicted 56% of the variance in the SOFIT data (P<.001), while HRM predicted 43% of the variance in the SOFIT data (P<.001). A z test comparison revealed that the difference between the association of MTI and SOFIT and that of HRM and SOFIT was not statistically significant (P=.46). Bland-Altman plots (fig 3) showed better agreement between the observed time spent in MVPA and MTI-predicted time spent in MVPA, supporting the results of the linear regression.
Fig 2
Linear relation between MTI and SOFIT data, and between HRM and SOFIT data, during structured activity. Dashed lines represent a 95% confidence interval for the regression line. Between MTI and SOFIT data: r=.75, R2=.56 , P<.001; natural log of MTI and SOFIT: r=.79, R2=.63, P<.001. Between HRM and SOFIT data: r=.65, R2=.43, P<.001; natural log of HRM and SOFIT: r=.66, R2=.43, P<.001.
Fig 3
Bland-Altman plot of the differences plotted against the averages between observed and predicted (MTI and HRM) minutes of MVPA based on SOFIT (criterion measure) during structured activity. Solid line represents the mean difference, and dashed lines represent ±1.96 SD. MTI: Meandiff=.03±.57; 95% confidence interval, –1.08 to 1.14. HRM: Meandiff=.06±.62; 95% confidence interval, –1.17 to 1.28.
Free Play
To represent activities in free living, the participants engaged in unstructured play. They took part in play activities that they themselves chose, and the intensity and activities were varied. This resulted in wide variance in both the activity counts (mean ± SD, 5348±2715; log: mean ± SD, 3.67±.22) and HRM (mean ± SD, 140.2±32.9; log: mean ± SD, 2.14±.10). Observed activities based on SOFIT ranged from codes 2 (sitting) to 5 (vigorous activity), with a median of 3 (standing). MTI data were shown to predict 45% (P<.001) of the variance in the SOFIT data, while HRM data predicted only 2% of the SOFIT variance (P<.001) (fig 4). The correlation coefficients between the MTI and SOFIT data and between the HRM and SOFIT data were significantly different (P=.01). Bland-Altman plots (fig 5) also showed better agreement between the observed and the MTI-predicted data than that between the observed and the HRM-predicted data.
Fig 4
Linear relation between MTI and SOFIT data, and between HRM and SOFIT data, during free play activity. Dashed lines represent a 95% confidence interval for the regression line. Between MTI and SOFIT data: r=.67, R2=.45, P<.001; natural log of MTI and SOFIT: r=.68, R2=.47, P<.001. Between HRM and SOFIT data: r=.14, R2=.02, P<.001; natural log of MTI and HRM: r=.15, R2=.02, P<.001.
Fig 5
Bland-Altman plot of the differences plotted against the averages between observed and predicted (MTI and HRM) minutes of MVPA based on SOFIT (criterion measure) during free play. Solid line represents the mean difference, and dashed lines represent ±1.96 SD, MTI. Meandiff= –2.7±2.48; 95% confidence interval, –7.6 to 2.2. HRM: Meandiff = –1.5±3.27; 95% confidence interval, –7.9 to 4.9.
To examine if participants' level of function as measured by GMFCS would influence the variance in MTI data, further linear regression analysis was performed. GMFCS levels explained 0 or less than 1% of the variance in MTI data during structured (R2=.000; P>.05) and free play (R2=.001; P>.05) activities. No regression results relating to diagnostic subgroups were obtained because of the very low number in each subgroup.
Discussion
The protocol in the structured activity sessions engaged the participants in a series of activities that were designed to have a stepwise increase in intensity. Observed PA levels as measured by SOFIT demonstrated a very strong, significant, and positive linear association with the activity protocol, confirming the evidence that direct observation is an appropriate criterion measure for PA measurement validity studies,19, 21, 23, 24 particularly among children.37 In an earlier calibration study on an accelerometer for children, a direct observation instrument was similarly used as a criterion measure for the accelerometer data in order to capture best the intermittent nature of children's physical activities.41
Heart rate measurement in the structured activity was shown to have a moderate, significant, and positive linear association with the intensity of physical activities. Considering that it takes at least 1 minute for heart rate to reach a steady state during low to moderate exercise levels,42 our measurements were taken during the last 30 seconds of each 2-minute activity. In such a situation, heart rate appears to be a valid measure to estimate PA levels because we are able to control the movement demands on the participants. However, in controlled exercise conditions with a stepwise increase in intensity, the MTI was found to be a better measure of activity than HRM. The validity of MTI as a measure of PA levels in controlled conditions was confirmed by its strong linear association with observed PA as measured by SOFIT (MTI accounted for 56% vs HRM 43% of the variance). Although no statistically significant difference was observed in the strength of associations between the 2 independent correlation coefficients, the results of the Bland-Altman plots for structured activity illustrated better agreement between MTI-SOFIT data than that between HRM-SOFIT data based on the narrower 95% confidence interval of the mean difference between the 2 methods.
Validation protocols such as the structured activity we used may not mimic free living;46 thus, we collected data in unstructured free play. Direct observation of PA levels using SOFIT was again used as the criterion measure. HRM remains to be considered as one of the feasible techniques for assessing energy expenditure and associated patterns of PA in field conditions.25 However, our study showed that in a free play situation, HRM was a poor predictor of directly observed PA. In free play conditions, sudden changes in activity may not necessarily have corresponding changes in heart rates. Furthermore, heart rates remain elevated postactivity, which may mask PA.22 Efficiency of movement also influences heart rate.47 Children with CP are in fact known to have lower movement efficiency, particularly in walking.48Our findings suggest that HRM may not be a valid instrument for studies examining the PA levels of children with CP in free living conditions. Conversely, MTI data were observed to predict 45% of the variance in PA measurements by SOFIT in the free play sessions, confirming that MTI may provide a valid measure of PA among children with CP in free living conditions. The association between the MTI and SOFIT data in free play was found to be significantly greater than that between the HRM and SOFIT data. The differences in accuracy between HRM and MTI are clinically important because they show that the use of the MTI to measure activity improves, in a measurable way, the probability of accurate estimation of PA in this particular group of children. This was further supported by the Bland-Altman plots showing reasonable agreement between the MTI and SOFIT, indicating the methodologic viability of using the MTI for PA studies involving children with CP. Our findings showed similar relationships between MTI and direct observation as have been observed in comparable validation studies in typically developing children.49
It must be noted that the MTI data of our participants have a wide variance, as shown by the high SDs. Furthermore, in the activity protocol components that represent MVPA, the MTI data we recorded were much higher than the established cutoff for children.50 As such, the use of regression equations to predict the cutoff point for MVPA seems inappropriate, and any estimates of energy expenditure derived from the MTI data are also likely to be inaccurate. This implies that the validity of MTI that we have demonstrated so far is limited as a measure of activity volume. Nevertheless, this is a valuable finding in the field of PA studies in children with CP. Activity count outputs of accelerometers have been shown to be useful in their raw form as a measure of activity volume51 and may be relevant outcome measures in studies involving children with CP. Furthermore, associated errors with regression models are eliminated.37
A review of the clinimetric properties of the MTI showed that it is well tolerated, is comfortable to wear, and does not hinder the activities of typically developing children.20 Although this study did not examine these characteristics systematically, the participants did not manifest any indications of intolerance with the device. The MTI devices remained in place over the duration of the activity sessions, and the participants did not attempt to remove them or complain about wearing them.
Study Limitations
Limitations of the study include the use of a relatively small sample size and the recruitment of participants from 2 special schools through convenience sampling—both of which may limit the generalizability of our findings. Although GMFCS levels did not influence the variance in MTI data, the accuracy of the data collected may vary within the CP sample used. Accuracy may be affected for those participants with dyskinetic CP, with their extraneous movements and recurrent changes in tone, and for those participants with hemiplegia who wore the MTI device on their impaired side after a standard placement of the instrument.
Conclusions
The present study concludes that the MTI appears to be a valid measure of activity volume among children with CP in both structured and free play activities. The findings support the potential use of MTI in field-based studies that aim to monitor the habitual PA of children with CP, including studies examining the effects of interventions. A minimum of 7 days of monitoring is required to assess children's habitual PA reliably,38 including weekdays and weekends.39 Similar considerations may have to be noted when using the MTI in examining the dose-response relationships of interventions to promote PA participation in a group of children with CP. Because no single measure of children's PA is without limitations,40 studies involving children with CP may benefit from the use of other complementary instruments to capture the diverse dimensions of PA. Subjective instruments such as the Activities Scale for Kids52 and Children's Activity Participation and Enjoyment53 have been shown to be useful in children with CP.29, 53
Further examinations of the use of the MTI to estimate energy expenditures and establish cutoff points require additional studies using instruments that measure oxygen consumption directly. Considering the variance of activity count readings among the participants in this study, it may be suggested that PA measurement among children with CP with the use of the MTI may necessitate individual calibration. The use of calibration based on individual data may help improve the accuracy of estimates of energy expenditure54 in this group of children.
Suppliers
aModel 7164; MTI Actigraph, 15 W Main St, Pensacola, FL 32502.
bModel S810; Polar Electro Oy, Professorintie 5, FIN-90440 Kempele, Finland.
cSPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
dMedCalc Software, Broekstraat 52, 9030 Mariakerke, Belgium.
Acknowledgments
We thank Tina Chan, PT and Kathlynne Eguia, PT for facilitating the data collection. Special thanks to the participating children with CP and to their parents as well.
References
1Rosenbaum, P., Paneth, N., Leviton, A. et al. A report: the definition and classification of cerebral palsy. Dev Med Child Neurol Suppl. 2007; 109: 8–14
2Odding, E., Roebroeck, M.E., and Stam, H.J. The epidemiology of cerebral palsy: incidence, impairments and risk factors. Disabil Rehabil. 2006; 28: 183–191
3Rosenbaum, P., Walter, S.D., Hanna, S.E. et al. Prognosis for gross motor function in cerebral palsy: creation of motor development curves. JAMA. 2002; 288: 1357–1363
4Cavill, N., Kahlmeier, S., and Racioppi, F. Physical activity and health in Europe: evidence for action. World Health Organisation, Geneva; 2006
5Sallis, J.F., Prochaska, J.J., and Taylor, W.C. A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc. 2000; 32: 963–975
6Murphy, N.A. and Carbone, P.S. Promoting the participation of children with disabilities in sports, recreation, and physical activities. Pediatrics. 2008; 121: 1057–1061
7Sit, C.H., McManus, A., McKenzie, T.L., and Lian, J. Physical activity levels of children in special schools. Prev Med. 2007; 45: 424–431
8Goodway, J.D. and Branta, C.F. Influence of a motor skill intervention on fundamental motor skill development of disadvantaged preschool children. Res Q Exerc Sport. 2003; 74: 36–46
9Wrotniak, B.H., Epstein, L.H., Dorn, J.M., Jones, K.E., and Kondilis, V.A. The relationship between motor proficiency and physical activity in children. Pediatrics. 2006; 118: e1758–e1765
10Maher, C.A., Williams, M.T., Olds, T., and Lane, A.E. Physical and sedentary activity in adolescents with cerebral palsy. Dev Med Child Neurol. 2007; 49: 450–457
11Orlin, M.N., Palisano, R.J., Chiarello, L.A. et al. Participation in home, extracurricular, and community activities among children and young people with cerebral palsy. Dev Med Child Neurol. 2009; 52: 160–166
12Palisano, R.J., Copeland, W.P., and Galuppi, B.E. Performance of physical activities by adolescents with cerebral palsy. Phys Ther. 2007; 87: 77–87
13Fowler, E.G., Kolobe, T.H., Damiano, D.L. et al. Promotion of physical fitness and prevention of secondary conditions for children with cerebral palsy: section on pediatrics research summit proceedings. Phys Ther.2007; 87: 1495–1510
14Salmon, J. Physical activity in young people—assessment and methodological issues. J Sci Med Sport. 2009;12: 513–514
15Kim, S.Y. Assessing physical activity levels of students with disabilities in physical education. J Phys Educ Recreation Dance. 2009; 80: 3–4 (51)
16Pirpiris, M. and Graham, H.K. Uptime in children with cerebral palsy. J Pediatr Orthop. 2004; 24: 521–528
17van Eck, M., Dallmeijer, A.J., Beckerman, H., van den Hoven, P.A.M., Voorman, J.M., and Becher, J.G. Physical activity level and related factors in adolescents with cerebral palsy. Pediatr Exerc Sci. 2008; 20: 95–106
18Reilly, J.J., Penpraze, V., Hislop, J., Davies, G., Grant, S., and Paton, J.Y. Objective measurement of physical activity and sedentary behaviour: review with new data. Arch Dis Child. 2008; 93: 614–619
19Sirard, J.R. and Pate, R.R. Physical activity assessment in children and adolescents. Sports Med. 2001; 31:439–454
20De Vries, S.I., Bakker, I., Hopman-Rock, M., Hirasing, R.A., and van Mechelen, W. Clinimetric review of motion sensors in children and adolescents. J Clin Epidemiol. 2006; 59: 670–680
21Corder, K., Ekelund, U., Steele, R.M., Wareham, N.J., and Brage, S. Assessment of physical activity in youth. J Appl Physiol. 2008; 105: 977–987
22Dollman, J., Okely, A.D., Hardy, L., Timperio, A., Salmon, J., and Hills, A.P. A hitchhiker's guide to assessing young people's physical activity: deciding what method to use. J Sci Med Sport. 2009; 12: 518–525
23Bjornson, K.F. Physical activity monitoring in children and youths. Pediatr Phys Ther. 2005; 17: 37–45
24Kohl, H.W., Fulton, J., and Caspersen, C. Assessment of physical activity among children and adolescents: a review and synthesis. Prev Med. 2000; 31: S54–S76
25Livingstone, M.B., Robson, P.J., and Totton, M. Energy expenditure by heart rate in children: an evaluation of calibration techniques. Med Sci Sports Exerc. 2000; 32: 1513–1519
26Rowlands, A.V. and Eston, R.G. Comparison of arm-to-leg and leg-to-leg (standing) bioelectrical impedance analysis for the estimation of body composition in 8- to 10-year-old children. in: T. Jurimae, A.P. Hills (Eds.) Body composition assessment in children and adolescents. Karger, Basel; 2008
27Bjornson, K.F., Belza, B., Kartin, D., Logsdon, R., McLaughlin, J., and Thompson, E.A. The relationship of physical activity to health status and quality of life in cerebral palsy. Pediatr Phys Ther. 2008; 20: 247–253
28Buffart, L.M., Roebroeck, M.E., Rol, M., Stam, H.J., and van den Berg-Emons, R.J. Triad of physical activity, aerobic fitness and obesity in adolescents and young adults with myelomeningocele. J Rehabil Med. 2008; 40:70–75
29Bjornson, K.F., Belza, B., Kartin, D., Logsdon, R., and McLaughlin, J.F. Ambulatory physical activity performance in youth with cerebral palsy and youth who are developing typically. Phys Ther. 2007; 87: 248–257(discussion 57-60)
30De Vries, S.I., Van Hirtum, H.W., Bakker, I., Hopman-Rock, M., Hirasing, R.A., and Van Mechelen, W. Validity and reproducibility of motion sensors in youth: a systematic update. Med Sci Sports Exerc. 2009; 41: 818–827
31Jun, S., Nobuaki, I., and Kazuhiro, F. Measurement of physical activity using Actigraph method in children with cerebral palsy. Bull Spec Educ. 2000; 24: 1–8
32Stevens, J. Applied multivariate statistics for the social sciences. Lawrence Erlbaum Associates, Mahwah;1999
33McKenzie, T.L., Sallis, J., and Nader, P.R. SOFIT: system for observing fitness instruction time. J Teach Phys Educ. 1991; 11: 195–205
34Heath, E.M., Coleman, K.J., Lensegrav, T.L., and Fallon, J.A. Using momentary time sampling to estimate minutes of physical activity in physical education: validation of scores for the System for Observing Fitness Instruction Time. Res Q Exerc Sport. 2006; 77: 142–146
35McKenzie, T.L., Sallis, J.F., Patterson, T.L. et al. BEACHES: an observational system for assessing children's eating and physical activity behaviors and associated events. J Appl Behav Anal. 1991; 24: 141–151
36Rowe, P.J., van der Mars, H., Schuldheisz, J.M., and Fox, S. Measuring students' physical activity levels: validating SOFIT for use with high-school students. J Teach Phys Educ. 2004; 23: 235–251
37Freedson, P., Pober, D., and Janz, K.F. Calibration of accelerometer output for children. Med Sci Sport Exer.2005; 37: S523–S530
38Trost, S.G., Pate, R.R., Freedson, P.S., Sallis, J.F., and Taylor, W.C. Using objective physical activity measures with youth: how many days of monitoring are needed?. Med Sci Sport Exerc. 2000; 32: 426–431
39Rowlands, A.V. Accelerometer assessment of physical activity in children: an update. Pediatr Exerc Sci.2007; 19: 252–266
40Rowlands, A.V. and Eston, R.G. The measurement and interpretation of children's physical activity. J Sports Sci Med. 2007; 6: 270–276
41Welk, G.J., Eisenmann, J.C., Schaben, J., Trost, S.G., and Dale, D. Calibration of the biotrainer pro activity monitor in children. Pediatr Exerc Sci. 2007; 19: 145–158
42Washington, R.L., Bricker, J.T., Alpert, B.S. et al. Guidelines for exercise testing in the pediatric age group (From the Committee on Atherosclerosis and Hypertension in Children, Council on Cardiovascular Disease in the Young, the American Heart Association) . Circulation. 1994; 90: 2166–2179
43Bland, J.M. and Altman, D.G. Measuring agreement in method comparison studies. Stat Methods Med Res.1999; 8: 135–160
44Schneider, P.L., Crouter, S.E., and Bassett, D.R. Pedometer measures of free-living physical activity: comparison of 13 models. Med Sci Sports Exerc. 2004; 36: 331–335
45Myles, P.S. and Cui, J. Using the Bland-Altman method to measure agreement with repeated measures. Br J Anaesth. 2007; 99: 309–311
46Rothney, M.P., Schaefer, E.V., Neumann, M.M., Choi, L., and Chen, K.Y. Validity of physical activity intensity predictions by ActiGraph, Actical, and RT3 accelerometers. Obesity (Silver Spring). 2008; 16: 1946–1952
47O'Beirne, C., Larkin, D., and Cable, T. Coordination problems and anaerobic performance in children. Adapt Phys Act Q. 1994; 11: 141–149
48Bar-Or, O. and Rowland, T. Pediatric exercise medicine: from physiologic principals to health care applications. Human Kinetics, Champaign; 2004
49Hands, B. and Parker, H. Pedometer-determined physical activity, BMI, and waist girth in 7- to 16-year-old children and adolescents. J Phys Act Health. 2008; 5: S153–S165
50Ekelund, U., Sardinha, L.B., Anderssen, S.A. et al. Associations between objectively assessed physical activity and indicators of body fatness in 9- to 10-y-old European children: a population-based study from 4 distinct regions in Europe (the European Youth Heart Study). Am J Clin Nutr. 2004; 80: 584–590
51McClain, J.J. and Tudor-Locke, C. Objective monitoring of physical activity in children: considerations for instrument selection. J Sci Med Sport. 2009; 12: 526–533
52Young, N.L. The Activities Scale for Kids (ASK). Laurentian University, Ontario; 2007
53King, G., Law, M., and King, S. Children's Assessment of Participation and Enjoyment (CAPE) and Preferences for Activities of Children (PAC). Harcourt Assessment, San Antonio; 2004
54Welk, G.J. Principles of design and analyses for the calibration of accelerometry-based activity monitors.Med Sci Sport Exer. 2005; 37: S501–S511
Nenhum comentário:
Postar um comentário