- Email: [email protected]
Within- and between-day stability of treadmill walking VO2 in children with hemiplegic cerebral palsy
Gait and Posture 21 (2005) 80–84
Within- and between-day stability of treadmill walking VO2 in children with hemiplegic cerebral palsy Stability of walking VO2 in children with CP Daniel J. Keefer a,∗ , Wayland Tseh b , Jennifer L. Caputo c , Kathy Apperson d , Sheri McGreal d , Don W. Morgan e a
Department of Wellness and Sport Sciences, Millersville University, 113 Pucillo Gymnasium, PO Box 1002, Millersville, PA 17551-0302, USA b Department of Health, Physical Education, and Recreation, University of North Carolina at Wilmington, Wilmington, NC, USA c Department of Health, Physical Education, Recreation, and Safety, Middle Tennessee State University, Murfreesboro, TN, USA d Department of Exercise and Sport Science, The University of North Carolina at Greensboro, Greensboro, NC, USA e Department of Kinesiology, Arizona State University, Temple, AZ, USA
Abstract Within- and between-day stability in locomotor energy use was quantified in 13 children with hemiplegic cerebral palsy (CP). During testing, subjects were familiarized with the laboratory environment (Session 1), performed three 5 min level treadmill walks (Trials 1–3) at 0.67 m∗ s−1 (Session 2), and completed a single 5 min walk (Trial 4) at 0.67 m∗ s−1 (Session 3). In Sessions 2 and 3, heart rate (HR) was assessed and expired air was collected and analyzed to determine VO2 . Data analyses revealed no significant difference (P > 0.05) in either net VO2 (ml kg−1∗ min−1 ) or EEIHR (b∗ m−1 ) across the three trials performed in Session 2 and between average measures of net VO2 and EEIHR quantified in Session 2 and those obtained in Session 3. Mean within-day coefficient of variation (CV) values for net VO2 and EEIHR were 8.6% ± 8.5% and 13.9% ± 7.8%, respectively. Analysis of between-day variability and energy expenditure revealed a between-day CV value of 13.1% for net VO2 and 24.0% for EEIHR . In addition, significant inverse relationships between Gross Motor Function Measure (GMFM) scores and within- (r = −0.61) and between-day (r = −0.58) CV values for net VO2 were detected. Viewed in concert, these data suggest that fairly stable within- and between-day measures of locomotor energy expenditure during level treadmill walking can be achieved in subjects with hemiplegic CP if testing is preceded by a short period of treadmill accommodation. However, children with greater motor dysfunction may require an extended period of treadmill accommodation to reduce trial-to-trial variability in walking energy use. © 2004 Elsevier B.V. All rights reserved. Keywords: Stability; Oxygen consumption; Cerebral palsy
1. Introduction Knowledge of the stability of walking energy expenditure in able-bodied or physically-challenged children is of particular importance to researchers interested in comparing differences in oxygen consumption (VO2 ) between intact groups of subjects varying on a particular factor (e.g. sex, age, disease status, activity level, and motor function) or assessing the efficacy of treatments designed to lower walking energy cost. Previous research has shown that in able-bodied children, stable VO2 data during treadmill walking can be secured within 10 min if data collection is pre∗ Corresponding author. Tel.: +1-717-871-2182; fax: +1-717-871-2393. E-mail address: [email protected] (D.J. Keefer).
0966-6362/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2004.01.004
ceded by a short period of treadmill walking practice [1]. Little is known, however, regarding the ability of children with cerebral palsy (CP) to establish reliable measures of energy expenditure. In an attempt to address this question, Bowen et al. [2] quantified between-day variability in VO2 and the energy expenditure index (EEIHR ), a heart rate-based measure of energy use, in five children with CP and five able-bodied (AB) children who walked while covering various distances. Although variability in VO2 was slightly higher in children with CP (17.5%) compared to AB children (14.3%), variability in EEIHR was nearly identical for both groups (CP = 20.3%; AB = 20.5%). Additionally, Keefer et al. [3] reported a between-day coefficient of variation of 8% for oxygen consumption in four children with CP. With respect to the stability of locomotor energetics in children with CP, few data are available regarding the
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
81
2. Methods
familiarized with the use of a Hans Rudolph 8950 series facemask that was employed to measure energy expenditure in later test sessions. After receiving instruction on how to mount and walk on the treadmill, each subject walked on the treadmill at 0.67 m∗ s−1 for a 5 min period while wearing the facemask. This treadmill speed was chosen because it fell near the middle of a range of speeds used in previous studies of similarly-aged children with CP [5,6].
2.1. Subjects
2.4. Session 2
Thirteen children with spastic hemiplegia, ranging in age from 6 to 15 years old, were recruited to participate in this study. To assist with subject recruitment, several local physical therapists were contacted and asked to provide names of prospective subjects with hemiplegic CP who were able to walk under their own power without the use of assistive devices. Subjects were excluded from testing if they had undergone a surgical procedure influencing functional mobility in the year preceding data collection or were taking medications that could influence energy expenditure or gait function. Table 1 (found in the electronic library of Science Direct) presents information regarding the physical characteristics and gross motor function of each child. At the time of the study, five subjects were undergoing physical and occupational therapy, two were completing only physical therapy, and six were not participating in any type of therapy. Prior to the initial laboratory visit, parents were asked to obtain a written letter from the doctor of their child providing medical clearance for each subject to participate in this investigation.
Within 1 week of completing Session 1, subjects returned to the laboratory for assessment of resting energy expenditure and to accommodate more fully to treadmill walking. Upon arrival at the laboratory, subjects sat in a chair and relaxed for 10 min while resting VO2 was measured. This was accomplished by having each subject breathe into a Hans Rudolph facemask held to the child’s face by a head-cap assembly and connected to a one-way valve directing expired air into a meteorological balloon. A gel sealant, designed specifically for Hans Rudolph facemasks, secured the mask to the face to prevent air seepage. Expired gas samples were then analyzed for CO2 and O2 content using Ametek O2 (S-3A/I) and CO2 (CD-3A) analyzers calibrated previously with primary standard gases. A Rayfield gas meter, calibrated against a Collins 120-1 tissot, was used to measure expired gas volumes. During the final minute of the 10 min resting period, heart rates (HR) were recorded manually using a Polar heart watch and transmitter. During this period, HR registered by the heart watch were recorded and averaged to generate a mean value for resting HR. Following the 10 min rest period, each subject performed three 5 min walks at 0.67 m∗ s−1 . The treadmill speed was adjusted to 0.67 m∗ s−1 during the first 2 min of each walk to ensure valid speed measurements. This was accomplished using a photoelectric cell mounted above the treadmill belt and connected to an electronic timer that counted the elapsed time for eight treadmill belt revolutions. During each walking bout, aerobic demand (VO2 ) and HR were quantified. To establish aerobic demand, gross VO2 was determined from an expired air sample collected during the final 2 min of each 5 min walk using methods described previously. Our decision to collect expired air following 3 min of submaximal walking is supported by published data indicating that children with CP can attain a steady state condition following 2 min of treadmill walking [6]. Resting VO2 was subtracted from the gross walking VO2 of each bout to obtain net VO2 . HR was recorded during two 15 s periods within the final 2 min of each walking bout and averaged to derive a mean walking HR value. Resting HR was subtracted from walking HR and the resultant value was divided by walking speed to yield a measure of EEIHR (expressed in beats per meter). The three walking bouts were separated by a minimum of 5 min of rest and a new trial did not start until subjects indicated that they were ready to continue walking.
ability of children with CP to achieve stable measures during treadmill walking. Hence, the purpose of this study was to quantify within- and between- day stability of VO2 and EEIHR in children with hemiplegic CP during treadmill walking.
2.2. Procedures Each subject completed three treadmill testing sessions over the course of a 2–3-week period. To ensure that circadian variability did not influence energy expenditure, testing occurred generally at the same time of day (±1 h) for each child. In addition, subjects were instructed to refrain from ingesting caffeine and eating at least 2 h before exercising.. During each session, a member of the testing team was situated directly behind each subject to provide an added measure of safety. The following section describes procedures used in each laboratory visit. 2.3. Session 1 On the first visit to the laboratory, parents of each subject were asked to complete an informed consent form and a health history questionnaire detailing the health status of their child. Height was measured with an anthropometer and body mass was determined using a digital electronic scale. Gross motor function was determined using Dimension 5 of the Gross Motor Function Measure (GMFM), which describes the ability of children with CP to perform walking, running, and jumping activities [4]. The subject was then
82
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
2.5. Session 3 Within 1 week of Session 2, subjects returned to the laboratory to complete a final testing session. During this session, VO2 and HR data were obtained while subjects completed a single walking trial on the treadmill at 0.67 m∗ min−1 . Net VO2 and EEIHR were determined using methods described earlier (see Session 2). 2.6. Statistical analysis Within-day stability of net VO2 and EEIHR was quantified by performing two repeated-measures analyses of variance (ANOVA) on metabolic data obtained during the three trials of Session 2. To further characterize within-day variability in energy measurements, a mean group within-day coefficient of variation (CV) value for net VO2 and EEIHR was determined and intraclass correlation coefficients (ICC) were calculated for net VO2 and EEIHR across the three walking trials completed in Session 2. The calculation of ICC values provided a reliability index sensitive to changes in both the order and magnitude of the repeated net VO2 and EEIHR measurements [7]. To determine between-day stability in energy cost, two paired t-tests were performed on average net VO2 and EEIHR values obtained during the three trials of Session 2 and the analogous value from the lone trial of Session 3. To provide an additional measure of stability among baseline energy values, between-day CV values were derived for measures of net VO2 and EEIHR . For each subject, CV values were calculated from the average net VO2 and EEIHR values measured during Trials 1–3 of Session 2 and the net VO2 and EEIHR values obtained during Session 3. These individual CV values were then averaged across subjects to yield a mean group between-day CV value for the entire cohort. Relationships among GMFM scores and variability in walking energy use within and between sessions were determined using Pearson-product moment correlations. Statistical significance was set at P ≤ 0.05.
3. Results Repeated-measures ANOVA yielded no within-day differences in net VO2 and EEIHR . In considering measures of VO2 stability, mean within-day CV values for net VO2 and EEIHR were 8.6% ± 8.5% and 13.9% ± 7.8%, respectively. Corresponding within-day ICC values for net VO2 and EEIHR were 0.78 and 0.85. Mean between-day CV for net VO2 was 13.1% with a range of 0.3%–41.8%. For EEIHR , the average between-day CV value was 24.0%, with values ranging from 0.7%–74.8%. Results from paired t-tests demonstrated no significant differences between the average of the net VO2 and EEIHR values from the three trials of Session 2 and the value from Session 3. Correlational analyses revealed significant relationships between
Table 2 Average net VO2 and EEIHR values for treadmill walking during Session 2
(ml kg−1∗ min−1 )
Net VO2 EEIHR (b∗ m−1 )
Trial 1
Trial 2
Trial 3
6.63 ± 1.70 0.60 ± 0.20
6.55 ± 1.18 0.55 ± 0.20
6.16 ± 1.15 0.53 ± 0.18
Values are mean ± standard deviation.
GMFM scores and within- (r = −0.61, p = 0.03) and between-day (r = −0.58, p = 0.03) CV values of net VO2 . However, no significant associations were observed between GMFM scores and either measure of EEIHR stability.
4. Discussion As shown in Table 2, no significant differences were noted for measures of net VO2 and EEIHR across the three 5 min walking trials conducted in Session 2. Comparison of average metabolic values obtained in Session 2 with the walking trial performed in Session 3 also revealed no overall difference in net VO2 and EEIHR (Table 3). Viewed in concert, these data suggest that subjects were metabolically accommodated to treadmill walking within the same day and across days. The stability of exercise energy expenditure values in our cohort of children with CP was also evaluated by calculating mean within- and between-day CV values as well as within-day ICC values. The between-day CV value of 13.1% for net VO2 obtained in the present investigation was somewhat higher than the CV value of 8% determined in a recent investigation from our laboratory [3] documenting the influence of auditory feedback on walking energy cost in children with CP. In comparing these studies, it should be emphasized that net VO2 was measured using the same facemask and energy expenditure collection procedures. However, the CV value obtained in the auditory feedback study [3] was based on only three children with hemiplegic CP and one child with diplegic CP who were tested during overground walking. Hence, it is possible that with this small number of subjects, an accurate view of the stability of metabolic measures in this limited population may not have emerged. Results from the current project also indicated that within-day CV values for net VO2 and EEIHR (8.6% and 13.9%) were lower compared to respective between-day CV values (13.1% and 24.0%). This finding is not surprising, Table 3 Average net VO2 and EEIHR values from Session 2 compared to Session 3 values
Net VO2 (ml kg−1∗ min−1 ) EEIHR (b∗ m−1 )
Average value from Session 2
Session 3
6.45 ± 1.26 0.56 ± 0.18
5.85 ± 1.71 0.52 ± 0.29
Values are mean ± standard deviation.
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
since energy expenditure values determined within the same day would be expected to be less variable compared to values obtained on different days. Findings from our investigation also revealed an inverse relationship between variability in oxygen use and functional mobility status. This suggests that children with hemiplegic CP who display lower levels of gross motor function may require longer periods of treadmill accommodation to achieve reliable trial-to-trial measures of VO2 stability. Interestingly, this association was not found when the EEIHR was used as a measure of locomotor energy expenditure. The level of within-day stability for net VO2 and EEIHR was further quantified using intraclass correlation coefficients (ICC). The within-day ICC values for net VO2 and EEIHR from the present study were 0.78 and 0.85, respectively. Both of these values are lower than the within-day ICC value for net VO2 of 0.96 obtained on young ablebodied children walking on a treadmill [1]. The lower withinday ICC values reported in the current investigation may be due to a more variable metabolic response in children with CP. As partial support for this hypothesis, Steinwender et al. [8] noted lower within-day repeatability of gait analysis data in children with CP compared to a cohort of age-matched, able-bodied children. In attempting to quantify the between-day stability of walking VO2 values in children with CP during overground walking, Bowen and colleagues [2] expressed variability in oxygen uptake as the absolute value of the difference between gross VO2 measures from two different days divided by the average of these two respective values. This calculation was also performed to determine variability in EEIHR . Using this approach, between-day variability in gross walking VO2 and EEIHR in their study of five children with CP (four with diplegia and one with hemiplegia) was 17.5% and 20.3%, respectively. If between-day variability as defined by Bowen and colleagues [2] had been calculated in the present investigation using average gross walking VO2 and EEIHR values from Sessions 2 and 3, the percent variability in gross VO2 and EEIHR would be 8.8% and 33.9%, respectively. Although speculative, the paradox of why VO2 may be less variable and EEIHR more variable during treadmill walking compared to overground walking may be a function of the methodology employed to assess walking energy expenditure, the topographical distribution of CP, or anxiety experienced by subjects during testing. The greater variability in oxygen consumption in the Bowen et al. study [2], for instance, may have been related to the greater ease associated with varying gait mechanics during overground walking. When speed is not controlled, as would be the case during free-walking conditions, a greater within-subject variation in oxygen consumption across testing bouts would be expected. The Bowen et al. investigation [2] also used four children with diplegic CP and one with hemiplegia compared to the larger number of solely hemiplegic subjects tested in the current study. Related to this point, children with diplegic CP may display a greater propensity to exhibit more variability
83
in VO2 due to the greater involvement of both lower limbs during locomotion. Conversely, HR may have been more variable in the present investigation due to fluctuating levels of nervousness or anxiety experienced by subjects during treadmill walking compared to normal overground locomotion. Concerning the use of the EEIHR , a recent study has placed doubt upon the validity of this measure as a proxy for oxygen consumption [9]. In light of this finding, and given the greater variability associated with this indirect assessment of gait energy cost, clinicians and researchers should avoid using the EEIHR as an index of locomotor energy expenditure. In conclusion, our data suggest that in subjects with hemiplegic CP, fairly stable within- and between-day measures of locomotor energy expenditure during level treadmill walking can be achieved if testing is preceded by a short period of treadmill practice. Children with hemiplegic CP who exhibit lower GMFM scores, however, may need extended exposure to treadmill walking to reduce variability in locomotor energy use to acceptable levels. Given the relative paucity of information in this area, future research is needed to confirm our findings and determine whether similar results would be obtained in children with diplegic CP. In addition, the hypothesis that gross motor function may be related to stability of energy expenditure in children with CP needs to be validated across a broad range of classification levels.
Acknowledgements This study was supported by funding from the National Institute of Child Health and Human Development (HD 30749).
References [1] Tseh W, Caputo JL, Craig IS, Keefer DJ, Martin PE, Morgan DW. Metabolic accommodation of young children to treadmill walking. Gait Posture 2000;12:139–42. [2] Bowen TR, Lennon N, Castagno P, Miller F, Richards J. Variability of energy-consumption measures in children with cerebral palsy. J Pediatr Orthop 1998;18:738–42. [3] Keefer DJ, Tseh W, Caputo JL, Griffith KE, Akins MB, Morgan DW. Effect of a heel-loaded auditory feedback system on gait speed and the aerobic demand of walking in children with cerebral palsy: a preliminary study. Dev Med Child Neurol 1999;41(Suppl 80): 35. [4] Russell DJ, Rosenbaum PL, Cadman DT, Gowland C, Hardy S, Jarvis S. The gross motor function measure: a means to evaluate the effects of physical therapy. Dev Med Child Neurol 1989;31:341– 52. [5] Rose J, Gamble JG, Medeiros J, Burgos A, Haskell WL. Energy cost of walking in normal children and in those with cerebral palsy: comparison of heart rate and oxygen uptake. J Pediatr Orthop 1989;9:276–9. [6] Unnithan VB, Dowling JJ, Frost G, Bar-Or O. Role of cocontraction in the O2 cost of walking in children with cerebral palsy. Med Sci Sports Exerc 1996;28:1498–504.
84
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
[7] Vincent WJ. Statistics in Kinesiology. Champaign: Human Kinetics, 1995. [8] Steinwender G, Saraph V, Scheiber S, Zwick E, Uitz C, Hackl K. Intrasubject repeatability of gait analysis data in normal and spastic children. Clin Biomech 2000;15:134–9.
[9] Keefer DJ, Apperson K, McGreal S, Tseh W, Caputo JL, Morgan DW. Comparison of direct and indirect measures of walking energy expenditure in children with cerebral palsy. Dev Med Child Neurol 2003;45(S 96):19.
Within- and between-day stability of treadmill walking VO2 in children with hemiplegic cerebral palsy Stability of walking VO2 in children with CP Daniel J. Keefer a,∗ , Wayland Tseh b , Jennifer L. Caputo c , Kathy Apperson d , Sheri McGreal d , Don W. Morgan e a
Department of Wellness and Sport Sciences, Millersville University, 113 Pucillo Gymnasium, PO Box 1002, Millersville, PA 17551-0302, USA b Department of Health, Physical Education, and Recreation, University of North Carolina at Wilmington, Wilmington, NC, USA c Department of Health, Physical Education, Recreation, and Safety, Middle Tennessee State University, Murfreesboro, TN, USA d Department of Exercise and Sport Science, The University of North Carolina at Greensboro, Greensboro, NC, USA e Department of Kinesiology, Arizona State University, Temple, AZ, USA
Abstract Within- and between-day stability in locomotor energy use was quantified in 13 children with hemiplegic cerebral palsy (CP). During testing, subjects were familiarized with the laboratory environment (Session 1), performed three 5 min level treadmill walks (Trials 1–3) at 0.67 m∗ s−1 (Session 2), and completed a single 5 min walk (Trial 4) at 0.67 m∗ s−1 (Session 3). In Sessions 2 and 3, heart rate (HR) was assessed and expired air was collected and analyzed to determine VO2 . Data analyses revealed no significant difference (P > 0.05) in either net VO2 (ml kg−1∗ min−1 ) or EEIHR (b∗ m−1 ) across the three trials performed in Session 2 and between average measures of net VO2 and EEIHR quantified in Session 2 and those obtained in Session 3. Mean within-day coefficient of variation (CV) values for net VO2 and EEIHR were 8.6% ± 8.5% and 13.9% ± 7.8%, respectively. Analysis of between-day variability and energy expenditure revealed a between-day CV value of 13.1% for net VO2 and 24.0% for EEIHR . In addition, significant inverse relationships between Gross Motor Function Measure (GMFM) scores and within- (r = −0.61) and between-day (r = −0.58) CV values for net VO2 were detected. Viewed in concert, these data suggest that fairly stable within- and between-day measures of locomotor energy expenditure during level treadmill walking can be achieved in subjects with hemiplegic CP if testing is preceded by a short period of treadmill accommodation. However, children with greater motor dysfunction may require an extended period of treadmill accommodation to reduce trial-to-trial variability in walking energy use. © 2004 Elsevier B.V. All rights reserved. Keywords: Stability; Oxygen consumption; Cerebral palsy
1. Introduction Knowledge of the stability of walking energy expenditure in able-bodied or physically-challenged children is of particular importance to researchers interested in comparing differences in oxygen consumption (VO2 ) between intact groups of subjects varying on a particular factor (e.g. sex, age, disease status, activity level, and motor function) or assessing the efficacy of treatments designed to lower walking energy cost. Previous research has shown that in able-bodied children, stable VO2 data during treadmill walking can be secured within 10 min if data collection is pre∗ Corresponding author. Tel.: +1-717-871-2182; fax: +1-717-871-2393. E-mail address: [email protected] (D.J. Keefer).
0966-6362/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2004.01.004
ceded by a short period of treadmill walking practice [1]. Little is known, however, regarding the ability of children with cerebral palsy (CP) to establish reliable measures of energy expenditure. In an attempt to address this question, Bowen et al. [2] quantified between-day variability in VO2 and the energy expenditure index (EEIHR ), a heart rate-based measure of energy use, in five children with CP and five able-bodied (AB) children who walked while covering various distances. Although variability in VO2 was slightly higher in children with CP (17.5%) compared to AB children (14.3%), variability in EEIHR was nearly identical for both groups (CP = 20.3%; AB = 20.5%). Additionally, Keefer et al. [3] reported a between-day coefficient of variation of 8% for oxygen consumption in four children with CP. With respect to the stability of locomotor energetics in children with CP, few data are available regarding the
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
81
2. Methods
familiarized with the use of a Hans Rudolph 8950 series facemask that was employed to measure energy expenditure in later test sessions. After receiving instruction on how to mount and walk on the treadmill, each subject walked on the treadmill at 0.67 m∗ s−1 for a 5 min period while wearing the facemask. This treadmill speed was chosen because it fell near the middle of a range of speeds used in previous studies of similarly-aged children with CP [5,6].
2.1. Subjects
2.4. Session 2
Thirteen children with spastic hemiplegia, ranging in age from 6 to 15 years old, were recruited to participate in this study. To assist with subject recruitment, several local physical therapists were contacted and asked to provide names of prospective subjects with hemiplegic CP who were able to walk under their own power without the use of assistive devices. Subjects were excluded from testing if they had undergone a surgical procedure influencing functional mobility in the year preceding data collection or were taking medications that could influence energy expenditure or gait function. Table 1 (found in the electronic library of Science Direct) presents information regarding the physical characteristics and gross motor function of each child. At the time of the study, five subjects were undergoing physical and occupational therapy, two were completing only physical therapy, and six were not participating in any type of therapy. Prior to the initial laboratory visit, parents were asked to obtain a written letter from the doctor of their child providing medical clearance for each subject to participate in this investigation.
Within 1 week of completing Session 1, subjects returned to the laboratory for assessment of resting energy expenditure and to accommodate more fully to treadmill walking. Upon arrival at the laboratory, subjects sat in a chair and relaxed for 10 min while resting VO2 was measured. This was accomplished by having each subject breathe into a Hans Rudolph facemask held to the child’s face by a head-cap assembly and connected to a one-way valve directing expired air into a meteorological balloon. A gel sealant, designed specifically for Hans Rudolph facemasks, secured the mask to the face to prevent air seepage. Expired gas samples were then analyzed for CO2 and O2 content using Ametek O2 (S-3A/I) and CO2 (CD-3A) analyzers calibrated previously with primary standard gases. A Rayfield gas meter, calibrated against a Collins 120-1 tissot, was used to measure expired gas volumes. During the final minute of the 10 min resting period, heart rates (HR) were recorded manually using a Polar heart watch and transmitter. During this period, HR registered by the heart watch were recorded and averaged to generate a mean value for resting HR. Following the 10 min rest period, each subject performed three 5 min walks at 0.67 m∗ s−1 . The treadmill speed was adjusted to 0.67 m∗ s−1 during the first 2 min of each walk to ensure valid speed measurements. This was accomplished using a photoelectric cell mounted above the treadmill belt and connected to an electronic timer that counted the elapsed time for eight treadmill belt revolutions. During each walking bout, aerobic demand (VO2 ) and HR were quantified. To establish aerobic demand, gross VO2 was determined from an expired air sample collected during the final 2 min of each 5 min walk using methods described previously. Our decision to collect expired air following 3 min of submaximal walking is supported by published data indicating that children with CP can attain a steady state condition following 2 min of treadmill walking [6]. Resting VO2 was subtracted from the gross walking VO2 of each bout to obtain net VO2 . HR was recorded during two 15 s periods within the final 2 min of each walking bout and averaged to derive a mean walking HR value. Resting HR was subtracted from walking HR and the resultant value was divided by walking speed to yield a measure of EEIHR (expressed in beats per meter). The three walking bouts were separated by a minimum of 5 min of rest and a new trial did not start until subjects indicated that they were ready to continue walking.
ability of children with CP to achieve stable measures during treadmill walking. Hence, the purpose of this study was to quantify within- and between- day stability of VO2 and EEIHR in children with hemiplegic CP during treadmill walking.
2.2. Procedures Each subject completed three treadmill testing sessions over the course of a 2–3-week period. To ensure that circadian variability did not influence energy expenditure, testing occurred generally at the same time of day (±1 h) for each child. In addition, subjects were instructed to refrain from ingesting caffeine and eating at least 2 h before exercising.. During each session, a member of the testing team was situated directly behind each subject to provide an added measure of safety. The following section describes procedures used in each laboratory visit. 2.3. Session 1 On the first visit to the laboratory, parents of each subject were asked to complete an informed consent form and a health history questionnaire detailing the health status of their child. Height was measured with an anthropometer and body mass was determined using a digital electronic scale. Gross motor function was determined using Dimension 5 of the Gross Motor Function Measure (GMFM), which describes the ability of children with CP to perform walking, running, and jumping activities [4]. The subject was then
82
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
2.5. Session 3 Within 1 week of Session 2, subjects returned to the laboratory to complete a final testing session. During this session, VO2 and HR data were obtained while subjects completed a single walking trial on the treadmill at 0.67 m∗ min−1 . Net VO2 and EEIHR were determined using methods described earlier (see Session 2). 2.6. Statistical analysis Within-day stability of net VO2 and EEIHR was quantified by performing two repeated-measures analyses of variance (ANOVA) on metabolic data obtained during the three trials of Session 2. To further characterize within-day variability in energy measurements, a mean group within-day coefficient of variation (CV) value for net VO2 and EEIHR was determined and intraclass correlation coefficients (ICC) were calculated for net VO2 and EEIHR across the three walking trials completed in Session 2. The calculation of ICC values provided a reliability index sensitive to changes in both the order and magnitude of the repeated net VO2 and EEIHR measurements [7]. To determine between-day stability in energy cost, two paired t-tests were performed on average net VO2 and EEIHR values obtained during the three trials of Session 2 and the analogous value from the lone trial of Session 3. To provide an additional measure of stability among baseline energy values, between-day CV values were derived for measures of net VO2 and EEIHR . For each subject, CV values were calculated from the average net VO2 and EEIHR values measured during Trials 1–3 of Session 2 and the net VO2 and EEIHR values obtained during Session 3. These individual CV values were then averaged across subjects to yield a mean group between-day CV value for the entire cohort. Relationships among GMFM scores and variability in walking energy use within and between sessions were determined using Pearson-product moment correlations. Statistical significance was set at P ≤ 0.05.
3. Results Repeated-measures ANOVA yielded no within-day differences in net VO2 and EEIHR . In considering measures of VO2 stability, mean within-day CV values for net VO2 and EEIHR were 8.6% ± 8.5% and 13.9% ± 7.8%, respectively. Corresponding within-day ICC values for net VO2 and EEIHR were 0.78 and 0.85. Mean between-day CV for net VO2 was 13.1% with a range of 0.3%–41.8%. For EEIHR , the average between-day CV value was 24.0%, with values ranging from 0.7%–74.8%. Results from paired t-tests demonstrated no significant differences between the average of the net VO2 and EEIHR values from the three trials of Session 2 and the value from Session 3. Correlational analyses revealed significant relationships between
Table 2 Average net VO2 and EEIHR values for treadmill walking during Session 2
(ml kg−1∗ min−1 )
Net VO2 EEIHR (b∗ m−1 )
Trial 1
Trial 2
Trial 3
6.63 ± 1.70 0.60 ± 0.20
6.55 ± 1.18 0.55 ± 0.20
6.16 ± 1.15 0.53 ± 0.18
Values are mean ± standard deviation.
GMFM scores and within- (r = −0.61, p = 0.03) and between-day (r = −0.58, p = 0.03) CV values of net VO2 . However, no significant associations were observed between GMFM scores and either measure of EEIHR stability.
4. Discussion As shown in Table 2, no significant differences were noted for measures of net VO2 and EEIHR across the three 5 min walking trials conducted in Session 2. Comparison of average metabolic values obtained in Session 2 with the walking trial performed in Session 3 also revealed no overall difference in net VO2 and EEIHR (Table 3). Viewed in concert, these data suggest that subjects were metabolically accommodated to treadmill walking within the same day and across days. The stability of exercise energy expenditure values in our cohort of children with CP was also evaluated by calculating mean within- and between-day CV values as well as within-day ICC values. The between-day CV value of 13.1% for net VO2 obtained in the present investigation was somewhat higher than the CV value of 8% determined in a recent investigation from our laboratory [3] documenting the influence of auditory feedback on walking energy cost in children with CP. In comparing these studies, it should be emphasized that net VO2 was measured using the same facemask and energy expenditure collection procedures. However, the CV value obtained in the auditory feedback study [3] was based on only three children with hemiplegic CP and one child with diplegic CP who were tested during overground walking. Hence, it is possible that with this small number of subjects, an accurate view of the stability of metabolic measures in this limited population may not have emerged. Results from the current project also indicated that within-day CV values for net VO2 and EEIHR (8.6% and 13.9%) were lower compared to respective between-day CV values (13.1% and 24.0%). This finding is not surprising, Table 3 Average net VO2 and EEIHR values from Session 2 compared to Session 3 values
Net VO2 (ml kg−1∗ min−1 ) EEIHR (b∗ m−1 )
Average value from Session 2
Session 3
6.45 ± 1.26 0.56 ± 0.18
5.85 ± 1.71 0.52 ± 0.29
Values are mean ± standard deviation.
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
since energy expenditure values determined within the same day would be expected to be less variable compared to values obtained on different days. Findings from our investigation also revealed an inverse relationship between variability in oxygen use and functional mobility status. This suggests that children with hemiplegic CP who display lower levels of gross motor function may require longer periods of treadmill accommodation to achieve reliable trial-to-trial measures of VO2 stability. Interestingly, this association was not found when the EEIHR was used as a measure of locomotor energy expenditure. The level of within-day stability for net VO2 and EEIHR was further quantified using intraclass correlation coefficients (ICC). The within-day ICC values for net VO2 and EEIHR from the present study were 0.78 and 0.85, respectively. Both of these values are lower than the within-day ICC value for net VO2 of 0.96 obtained on young ablebodied children walking on a treadmill [1]. The lower withinday ICC values reported in the current investigation may be due to a more variable metabolic response in children with CP. As partial support for this hypothesis, Steinwender et al. [8] noted lower within-day repeatability of gait analysis data in children with CP compared to a cohort of age-matched, able-bodied children. In attempting to quantify the between-day stability of walking VO2 values in children with CP during overground walking, Bowen and colleagues [2] expressed variability in oxygen uptake as the absolute value of the difference between gross VO2 measures from two different days divided by the average of these two respective values. This calculation was also performed to determine variability in EEIHR . Using this approach, between-day variability in gross walking VO2 and EEIHR in their study of five children with CP (four with diplegia and one with hemiplegia) was 17.5% and 20.3%, respectively. If between-day variability as defined by Bowen and colleagues [2] had been calculated in the present investigation using average gross walking VO2 and EEIHR values from Sessions 2 and 3, the percent variability in gross VO2 and EEIHR would be 8.8% and 33.9%, respectively. Although speculative, the paradox of why VO2 may be less variable and EEIHR more variable during treadmill walking compared to overground walking may be a function of the methodology employed to assess walking energy expenditure, the topographical distribution of CP, or anxiety experienced by subjects during testing. The greater variability in oxygen consumption in the Bowen et al. study [2], for instance, may have been related to the greater ease associated with varying gait mechanics during overground walking. When speed is not controlled, as would be the case during free-walking conditions, a greater within-subject variation in oxygen consumption across testing bouts would be expected. The Bowen et al. investigation [2] also used four children with diplegic CP and one with hemiplegia compared to the larger number of solely hemiplegic subjects tested in the current study. Related to this point, children with diplegic CP may display a greater propensity to exhibit more variability
83
in VO2 due to the greater involvement of both lower limbs during locomotion. Conversely, HR may have been more variable in the present investigation due to fluctuating levels of nervousness or anxiety experienced by subjects during treadmill walking compared to normal overground locomotion. Concerning the use of the EEIHR , a recent study has placed doubt upon the validity of this measure as a proxy for oxygen consumption [9]. In light of this finding, and given the greater variability associated with this indirect assessment of gait energy cost, clinicians and researchers should avoid using the EEIHR as an index of locomotor energy expenditure. In conclusion, our data suggest that in subjects with hemiplegic CP, fairly stable within- and between-day measures of locomotor energy expenditure during level treadmill walking can be achieved if testing is preceded by a short period of treadmill practice. Children with hemiplegic CP who exhibit lower GMFM scores, however, may need extended exposure to treadmill walking to reduce variability in locomotor energy use to acceptable levels. Given the relative paucity of information in this area, future research is needed to confirm our findings and determine whether similar results would be obtained in children with diplegic CP. In addition, the hypothesis that gross motor function may be related to stability of energy expenditure in children with CP needs to be validated across a broad range of classification levels.
Acknowledgements This study was supported by funding from the National Institute of Child Health and Human Development (HD 30749).
References [1] Tseh W, Caputo JL, Craig IS, Keefer DJ, Martin PE, Morgan DW. Metabolic accommodation of young children to treadmill walking. Gait Posture 2000;12:139–42. [2] Bowen TR, Lennon N, Castagno P, Miller F, Richards J. Variability of energy-consumption measures in children with cerebral palsy. J Pediatr Orthop 1998;18:738–42. [3] Keefer DJ, Tseh W, Caputo JL, Griffith KE, Akins MB, Morgan DW. Effect of a heel-loaded auditory feedback system on gait speed and the aerobic demand of walking in children with cerebral palsy: a preliminary study. Dev Med Child Neurol 1999;41(Suppl 80): 35. [4] Russell DJ, Rosenbaum PL, Cadman DT, Gowland C, Hardy S, Jarvis S. The gross motor function measure: a means to evaluate the effects of physical therapy. Dev Med Child Neurol 1989;31:341– 52. [5] Rose J, Gamble JG, Medeiros J, Burgos A, Haskell WL. Energy cost of walking in normal children and in those with cerebral palsy: comparison of heart rate and oxygen uptake. J Pediatr Orthop 1989;9:276–9. [6] Unnithan VB, Dowling JJ, Frost G, Bar-Or O. Role of cocontraction in the O2 cost of walking in children with cerebral palsy. Med Sci Sports Exerc 1996;28:1498–504.
84
D.J. Keefer et al. / Gait and Posture 21 (2005) 80–84
[7] Vincent WJ. Statistics in Kinesiology. Champaign: Human Kinetics, 1995. [8] Steinwender G, Saraph V, Scheiber S, Zwick E, Uitz C, Hackl K. Intrasubject repeatability of gait analysis data in normal and spastic children. Clin Biomech 2000;15:134–9.
[9] Keefer DJ, Apperson K, McGreal S, Tseh W, Caputo JL, Morgan DW. Comparison of direct and indirect measures of walking energy expenditure in children with cerebral palsy. Dev Med Child Neurol 2003;45(S 96):19.