Research in Developmental Disabilities 35 (2014) 1087–1097
Contents lists available at ScienceDirect
Research in Developmental Disabilities
Physical fitness in children with Developmental Coordination Disorder: Measurement matters Gillian D. Ferguson a,b,*, Wendy F.M. Aertssen c, Eugene A.A. Rameckers c,d, Jennifer Jelsma a, Bouwien C.M. Smits-Engelsman b a
University of Cape Town, Faculty of Health Sciences, Department of Health and Rehabilitation Sciences, Suite F45: Old Main Building, Groote Schuur Hospital, Main Road, Observatory 7925, Cape Town 8000, South Africa Katholieke Universiteit Leuven, Faculty of Kinesiology and Rehabilitation Sciences, Department of Kinesiology, Movement Control and Neuroplasticity Research Group, Tervuursevest 101, Postbox 1501, B-3001 Heverlee, Belgium c Avans + University of Professionals, Department of Physiotherapy, Heerbaan 14-40, Postbox 2087, 4800 CB Breda, The Netherlands d Maastricht University Medical Centre, Department of Rehabilitation Medicine & Adelante Center of Expertise in Rehabilitation & Audiology, P.O. Box 616, 6200 MD Maastricht, The Netherlands b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 December 2013 Received in revised form 29 January 2014 Accepted 29 January 2014 Available online 26 February 2014
Children with Developmental Coordination Disorder (DCD) experience considerable difficulties coordinating and controlling their body movements during functional motor tasks. Thus, it is not surprising that children with DCD do not perform well on tests of physical fitness. The aim of this study was to determine whether deficits in motor coordination influence the ability of children with DCD to perform adequately on physical fitness tests. A case–control study design was used to compare the performance of children with DCD (n = 70, 36 boys, mean age = 8y 1mo) and Typically Developing (TD) children (n = 70, 35 boys, mean age = 7y 9mo) on measures of isometric strength (hand-held dynamometry), functional strength, i.e. explosive power and muscular endurance (Functional Strength Measurement), aerobic capacity (20 m Shuttle Run Test) and anaerobic muscle capacity, i.e. muscle power (Muscle Power Sprint Test). Results show that children with DCD were able to generate similar isometric forces compared to TD children in isometric break tests, but were significantly weaker in three-point grip strength. Performance on functional strength items requiring more isolated explosive movement of the upper extremities, showed no significant difference between groups while items requiring muscle endurance (repetitions in 30 s) and items requiring whole body explosive movement were all significantly different. Aerobic capacity was lower for children with DCD whereas anaerobic performance during the sprint test was not. Our findings suggest that poor physical fitness performance in children with DCD may be partly due to poor timing and coordination of repetitive movements. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Physical fitness Strength Anaerobic muscle capacity Muscle power Aerobic capacity Developmental Coordination Disorder South Africa
1. Introduction The American College of Sports Medicine (ACSM) defines physical fitness as a set of measurable health and skill-related attributes that include body composition, cardiorespiratory fitness (CRF), muscular fitness, flexibility, and neuromotor * Corresponding author at: University of Cape Town, Department of Health and Rehabilitation Sciences, Suite F45: Old Main Building, Groote Schuur Hospital, Main Road, Observatory 7925, Cape Town 8000, South Africa. Tel.: +27 21 406 6045; mobile: +27 82 9743924. E-mail addresses: [email protected] (G.D. Ferguson), [email protected] (Wendy F.M. Aertssen), [email protected] (Eugene A.A. Rameckers), [email protected] (J. Jelsma), [email protected] (Bouwien C.M. Smits-Engelsman). http://dx.doi.org/10.1016/j.ridd.2014.01.031 0891-4222/ß 2014 Elsevier Ltd. All rights reserved.
1088
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
fitness (Garber et al., 2011). In the last decade, physical fitness in children with Developmental Coordination Disorder (DCD) has gained recognition as an important factor influencing performance in daily activities and as a mediator of health and wellbeing (Wahi et al., 2011). Children with DCD are reported to have reduced levels of physical fitness (Nascimento et al., 2013; Rivilis et al., 2011; van der Hoek et al., 2012) and are considered to be at increased risk for cardiovascular problems later in life (Cairney, Hay, Veldhuizen, & Faught, 2011). Studies examining body composition report that children with DCD have higher body mass indices (BMI) (Rivilis et al., 2011), higher body fat percentage (Cairney, Hay, Faught, & Hawes, 2005) and increased waist circumference (Cairney, Hay, Veldhuizen, Missiuna, et al., 2010; Wahi et al., 2011) compared to their Typically Developing (TD) peers. Regular participation in moderate to vigorous physical activity has therefore been recommended to reduce the risk of children developing cardiovascular conditions later in life (Lipnowski & Leblanc, 2012). However, participation in physical activity is often hampered by the limited motor performance capacity of children with DCD (Fong et al., 2011; Haga, 2009). Decreased CRF in DCD has been reported in several studies in which aerobic capacity was measured using field-based running tests such as the Le´ger 20 m shuttle run test (20 mSRT) (Rivilis et al., 2011) or less frequently, in laboratory tests using cycle ergometry (Cairney, Hay, Veldhuizen, & Faught, 2010) and treadmill protocols (Chia, Reid, Licari, & Guelfi, 2013). Laboratory-based measures of CRF (i.e. volume of oxygen consumed at maximal physical exertion/VO2max) are considered to be the gold standard for assessing aerobic capacity, whereas field-based measures have been criticized for the confounding factors associated with measuring maximal effort in the absence of objective indicators of exertion (Cairney, Hay, Veldhuizen, & Faught, 2010). In DCD specifically, the main factors associated with poor performance in field-based tests of CRF are thought to be related to lowered perceived self-efficacy (Cairney, Hay, Wade, Faught, & Flouris, 2006), low motivation and reduced levels of physical activity (Cairney, Hay, Faught, Wade, et al., 2005). Despite this, various authors agree that field-based running testing using standardized protocols, such as the Le´ger 20mSRT, are valid and reliable means to assess aerobic capacity in children with and without DCD (Cairney, Hay, Veldhuizen, & Faught, 2010). In contrast to the endurance running tests used to measure aerobic capacity, tests of anaerobic capacity include maximal running speed tests (e.g. 10 m 5 m, 20m and 50m sprint tests). Importantly, Verschuren, Takken, Ketelaar, Gorter, and Helders (2007) highlight that the agility requirements within the 10 m 5 m sprint test may confound the interpretation of anaerobic capacity in children with poor coordination (Verschuren et al., 2007). Another important attribute of physical fitness is flexibility. The sit and reach test is the most commonly reported flexibility measure used among children with DCD (Rivilis et al., 2011). Results show that children with DCD have a heterogeneous flexibility profile, with some studies showing poorer flexibility (Cantell, Crawford, & Tish Doyle-Baker, 2008; Hands, Larkin, Parker, Straker, & Perry, 2009) and others reporting no difference in flexibility compared to TD children (Schott, Alof, Hultsch, & Meermann, 2007; Tsiotra, Nevill, Lane, & Koutedakis, 2009). Concerning muscular fitness, three elements are typically evaluated: muscle strength, power and endurance. Findings from studies using either isometric or isokinetic dynamometry, which are considered the most robust forms of measuring muscular strength report that muscle strength is decreased in most muscle groups in DCD (Raynor, 2001; van der Hoek et al., 2012). Muscle power and muscular endurance tests on the other hand, are commonly used to make inferences about anaerobic muscle capacity. Tests of explosive muscle power examine parameters such as distance covered (e.g. throwing a heavy ball or performing a standing-long-jump) whereas tests of anaerobic muscle endurance measure the maximal number of repetitions within a specific time constraint (e.g. number of sit- or push-ups executed in 30 s). While the extent to which motor coordination deficits influence performance on these tests is acknowledged (Raynor, 2001; Rivilis et al., 2011) few studies have examined the relationship between muscle fitness and task constraints in DCD. The term neuromotor fitness is a collective noun, introduced by Garber et al. (2011) to describe motor skills such as balance, coordination, agility, and proprioceptive ability. Neuromotor fitness is a skill-related component of physical fitness and considered to be important in injury prevention. Neuromotor skills are by definition, functional skills and the motor tasks used to evaluate neuromotor fitness include running, walking on a line or hopping. Outcome measures designed to evaluate motor performance in children with DCD include standardized measures such as the Movement Assessment Battery for Children 2nd Edition (MABC-2) (Henderson, Sugden, Barnett, & Smits Engelsman, 2010), the Bruininks–Oseretsky Test of Motor Proficiency – 2nd edition (BOT-2) (Bruininks & Bruininks, 2005) and the McCarron Assessment of Neuromuscular Development (MAND) (McCarron, 1997). On examination of the items in these tests, it is evident that aspects of neuromotor fitness (i.e. balance, agility, coordination) are evaluated in each test. Importantly, adequate performance in neuromotor fitness tests is influenced by the ability to mitigate the variable influence of external forces and environmental constraints affecting movement quality. Evidently, motor proficiency plays an important role since carefully graded and well-timed muscle contractions lead to more economical and efficient ways of moving. Poor balance and agility in children with DCD (Chia, Licari, Guelfi, & Reid, 2012) may therefore explain their less favorable performance on neuromotor fitness measures. In DCD, compensatory strategies for motor control deficits are likely to influence physical fitness outcomes. One of the strategies frequently used in early stages of skill learning is co-activation of muscles, which leads to increased stability but can also potentially hamper force production (Raynor, 2001). Since DCD is a motor skill-learning deficit by definition, it is likely that the fine-tuning (grading) of multi-joint movements used in agility or dynamic power tests will be harder to optimize. Freezing joints may then be used as a temporary solution for controlling degrees of freedom or as a kind of mechanical filter to suppress the effects of force variability (Smits-Engelsman & Wilson, 2013).
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
1089
The assessment of particular components of physical fitness rest on the premise that the limiting factor in test performance is fatigue following a certain time period or number of repetitions, rather than the level of difficulty of the test (Cairney et al., 2006). Importantly, this may be different for children with coordination problems. For instance, the control of velocity and force changes that occur during an explosive throwing action and the implicit knowledge of how to use stored energy during a standing-long-jump (by exploiting the effect of the stretch-shortening cycle) will give an experienced and coordinated child an advantage in throwing and jumping distance even with the same generated muscle strength. In the current study, we suggest that, as both co-ordination and fitness elements are assessed concurrently. Isolated forms of isometric or isokinetic muscle contractions, using dynamometry, could improve the validity of measuring muscular fitness in children with poor co-ordination. Hand held dynamometry (HHD) for isometric strength attempts to isolate muscle action by controlling for joint movement. Although physical (e.g. BMI, ligament, tendon factors) and other aspects (e.g. motivation, attention, cognitive level) influence HHD outcomes, validity reports from studies conducted among children are generally positive with intra-class correlation coefficient values ranging from 0.73 to 0.99 (Beenakker, van der Hoeven, Fock, & Maurits, 2001; Brussock, Haley, Munsat, & Bernhardt, 1992). Inclusion of isometric measures is thus useful in understanding the contribution of muscle strength to function. However, in real life, actions seldom involve a pure form of isolated muscle action. Moreover, evidence suggests that the relationship between isometric strength and the ability to perform a functional task in which strength is required is not linear (Mattar & Sobreira, 2008). Thus, an evaluation of the functional impact of DCD would need to incorporate both complex and simple muscle actions. The Functional Strength Measure (FSM) is newly developed instrument, which was designed specifically to measure strength within a standardized functional task (Smits-Engelsman & Verhoef-Aertssen, 2012). The FSM measures two aspects of muscular fitness i.e. muscular endurance and explosive strength, within eight functional tasks. Tasks related to muscular endurance include lifting a heavy box repeatedly, repetitive sitting to standing, running up and down stairs, and performing repetitive lateral step ups. Tasks related to explosive muscle power include a standing long jump and performing a chest pass, over- and under-hand throwing using a heavy beanbag. In the current study, the aim was to determine whether poor performance on physical fitness tests is due to muscle weakness and poor CRF or deficits in using this strength in a functional context. To determine whether coordination is an important factor in physical fitness measures, we used various outcome measures, each containing tasks with different coordination requirements. Since most of the tests used require a certain amount of agility, we hypothesized large differences (effect sizes) between TD and DCD groups in tests that require more coordination (e.g. standing long jump; running up and down stairs) and small differences in simpler tasks such as single joint strength measures. The specific objectives were to compare TD and children with DCD in terms of aerobic and anaerobic muscle capacity, isometric strength, muscle endurance and explosive muscle power using a combination of functional and isolated measures. 2. Materials and methods 2.1. Research design and setting A cross-sectional, case–control study design was conducted in three mainstream primary schools situated in a lowincome area in Cape Town, South Africa. The implementation of physical education at these schools was limited by resource constraints and the accessibility of safe play-areas and opportunities to participate in sports and other physical activities were limited for all children. Since only one child participated in sports, information regarding participation in physical activity was not formally assessed. Ethical approval and permission was granted by the University of Cape Town Human Research Ethics Committee and the Western Cape Education Department. All parents and their children provided written informed consent and assent to participate in this study and to publication of the results. 2.2. Participants Children between the ages of six and ten years old, in grades 1–4, whose parents’ had given consent (N = 148) were recruited using convenience sampling. Exclusion criteria were: (1) failing a grade more than once and (2) a diagnosis of cerebral palsy or other significant neurological or medical condition as reported by a parent. Three children were excluded based on these criteria. Children were eligible for inclusion in the DCD group if their motor function in daily life was considered to be problematic according to their teacher and/or parent (Criterion B of the DSM IV diagnostic criteria) and they scored 5th percentile on the MABC-2 (Criterion A of the DSM IV diagnostic criteria). Inclusion criteria for the TD control group were if their motor function in daily life was considered to be within normal range for age and gender according to their teacher and/or parent and they scored above the 16th percentile on the MABC-2. Seventy children were identified as having DCD (boys = 36, girls = 34) and 75 were classified as TD (girls = 39, boys = 36) according to our criteria. A ratio of 1:1 was used to randomly select one TD child, matched for age and gender to every child identified with DCD. Since there were more girls in the TD group, we matched one girl of the same age to a boy with DCD and the remaining four TD girls were excluded. The final effective sample thus consisted of 140 children (boys = 71, girls = 69, mean age: 8.01, SD = 1.35).
1090
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
2.3. Outcome measures 2.3.1. Anthropometric measures Standardized anthropometric measurements were taken. Standing height (in centimeters) was measured without shoes, heels together using a tape measure fixed to a wall. Weight was measured in kilograms (0.1 kg accuracy), using a calibrated scale and BMI was calculated using the formula (weight/height2). Waist circumference was measured in centimeters, using a tape measure positioned around the abdomen in line with the top of the iliac crests. 2.3.2. Stand and reach test The Stand and reach test item, taken from the Beighton Scale of Joint Hypermobility (Beighton, Solomon, & Soskolne, 1973) was used to assess flexibility of the lower back and hamstring muscles. Following a brief warm up period, children were asked to stand with feet together, and then bend forward, with straight arms and knees and place both hands flat on the floor. 2.3.3. The Functional Strength Measure (FSM) The FSM (Smits-Engelsman & Verhoef-Aertssen, 2012) was used to assess maximal explosive power in one movement (four items: standing-long-jump, overarm throwing, underarm throwing, and chest pass), and muscle power where weight (body or object) was moved at maximal speed to generate a maximum number of repetitions within 30 s (four items: sit-tostand, lateral step-up, lifting a box and stair climbing). Each item consists of an instruction and demonstration phase followed by a practice phase, where participants are encouraged to execute the task. Feedback on how to correct the movement where necessary is given to those who demonstrate difficulty during practice phase. Once testers are satisfied that children know what to do (backward demonstration), formal trials may begin. Three trials were conducted for each task and the results from the best trial were scored. After testing, the highest score achieved on items in which left and right were tested was designated at the preferred limb for that particular item. Test reliability is moderate to high (ICC ranging from 0.73 to 0.91) (Smits-Engelsman & Verhoef-Aertssen, 2012). Concurrent validity of the FSM-items and handheld dynamometer is reported to range between 0.46 and 0.69 for the lower limb and between 0.52 and 0.74 for upper limb items. This suggests that while constructs related to strength are being tapped into during the test, it also suggests the presence of additional factors unrelated to one maximal isometric muscle contraction. Divergent validity assessed against the MABC-2 items yielded correlation values between 0.27 and 0.50. This suggests that items of the FSM are weakly correlated with coordination ability. 2.3.4. Hand-held dynamometer (HHD) The MicroFET-2 (Hogan Health Industries Inc., USA) and the Lafayette Manual Muscle Testing System (Model 01163, Lafayette Instrument Company, USA) were used to assess isometric muscle strength. The protocol for positioning and testing children outlined by Beenakker et al. (2001) was adopted. ‘‘Break’’ tests, where the examiner gradually overcomes the muscle strength generated by the participant, were used to evaluate elbow flexors, elbow extensors and knee extensors. The ‘‘make’’ test, defined as exerting maximal strength against the HHD, was used to asses three-point grip strength. Practice trials were given between each new test item using verbal instruction, demonstration and practice. Participants were corrected during practice and research assistants made sure that participants understood the requirements of the test. Each muscle group was tested three times and peak forces, measured in Newton, were recorded for each trial. Research assistants alternated between testing left and right sides of various muscle groups. The best scores of the three trials were used for analysis. After testing, the highest score achieved on items in which left and right were tested was designated at the preferred limb for that particular item. 2.3.5. Muscle Power Sprint Test (MPST) The MPST is an intermittent sprint test consisting of six, timed 15 m sprints (Verschuren et al., 2007). It was designed to measure muscle power with good reliability (Douma-van Riet et al., 2012). The time to complete each run is measured in milliseconds. Power output is calculated using the equations specified by Verschuren et al. (2007). Peak power is defined as the highest power that is generated of all six sprints and the mean power is the average power output over the six sprints. Greater mean power indicates the ability to maintain power output over time and is representative of anaerobic running performance. The 15 m distance required for the MPST was marked by a painted line on the playground and cones were placed 1 m beyond the line where children were encouraged to stop. Children were tested in groups of three to six and were instructed to run as fast as possible and cross the line. Explanations and practice trials were given followed by a brief rest period before the actual trial. A 30-s rest was allowed between trial runs. 2.3.6. 20 m Shuttle Run Test (20mSRT) The 20mSRT was used to measure aerobic fitness (Le´ger, Mercier, Gadoury, & Lambert, 1988). A practice trial was given and a short rest period included before the start of the actual trial. Participants were verbally encouraged to keep running for as long as possible. For some children, it is difficult to run at the audio signal that determines the running speed. Therefore, one tester ensured that children understood the principle of the Shuttle Run Test by running with them throughout the test.
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
1091
Since it is understood from the literature that BMI may be a confounding factor in performance of children with DCD (Tsiotra et al., 2009) we used the Matsuzaka formulae (Matsuzaka et al., 2004) to calculate VO2max in this study. This formula takes into consideration the impact of gender and body mass index (BMI) on VO2max and may yield more valid estimates of CRF (Barnett, Chan, & BruceIain, 1993; Matsuzaka et al., 2004; Ruiz et al., 2008). The Le´ger formula was also calculated for comparison. 2.4. Procedure Recommendations outlined in the Diagnostic and Statistical Manual of Mental Disorders IV (American Psychiatric Association, 2000) were used to identify children with DCD. Initially, teacher and parent questionnaires were administered to determine whether the child’s motor coordination ability affected their performance in everyday life (Criterion B of the DSM IV diagnostic criteria). The questionnaire, developed by the authors, consisted of a description of common functional problems that children with DCD present with and one closed- and one open-ended question asking parents and teachers to give their opinion regarding whether they thought the child had a motor coordination problem and why they thought so. Next, the Movement Assessment Battery for Children 2nd edition (MABC-2) (Henderson, Sugden, & Barnett, 2007) was used to identify children whose motor performance scores were at or below the 5th percentile (Criterion A of the DSM IV diagnostic criteria). All children participated in a standardized warm up phase before testing which including marching across the playground while swinging arms. This was followed by a brief rest period in which the individual tests were explained. Each test was conducted a different day over a period of one week. Running tests (MPST and 20mSRT) were performed on a tarred playground and other tests (FSM and HHD) were conducted in a quiet room on the school premises. All tests were conducted by physiotherapists and physiotherapy students who received training on the instruments used and were unaware of the child’s group affiliation. 2.5. Data analysis A power calculation was conducted based on the means and standard deviations of results obtained in previous studies using the 20mSRT (Cairney, Hay, Faught, Flouris, & Klentrou, 2007) and sit-to-stand scores of the FSM (Smits-Engelsman & Verhoef-Aertssen, 2012). Using an online sample size calculator we determined that sample sizes between 32 and 43 children per group would be sufficient to detect a difference at p = 0.001 level with 80% power (http://www.stat.ubc.ca). All data were analyzed using SPSS 20.0 (IBM, 2011). Pearson’s Chi-square test was used to compare differences in gender and preferred hand between groups. Shapiro–Wilks tests showed that fitness data were not normally distributed, thus raw scores were transformed using natural logs. Independent t-tests (two-tailed) were used to compare outcomes between groups. The post hoc ANCOVA procedure was used to test if the between-group effect remained significant if BMI was used as a covariate. Alpha was set at 0.05. Estimates of effect size (Cohen’s d) were calculated for group comparison. This measures the magnitude of the difference between the mean scores of groups, divided by a pooled SD. The magnitude of the mean effect size estimates (d) were interpreted according to the conventions of Cohen: 0.30 (small), 0.50 (moderate), 0.80 (large) and >1.00 (very large effect size) (Fern & Monroe, 1996). 3. Results Groups (TD n = 70, DCD n = 70) were comparable in terms of age, gender, handedness and height. However, weight, BMI, and waist circumference were significantly higher for the DCD-group. Four children were overweight (>85th percentile) in the TD and nine in the DCD-group. The differences between the two groups of children are presented in Table 1. Table 2 shows the results of the group comparison on the physical fitness tests.
Table 1 Characteristics of participants with Developmental Coordination Disorder (DCD) and age-matched Typically Developing (TD) children. Variable
TD (N = 70)
DCD (N = 70)
Statistics
MABC-2 standard score Gender (n)
11.2 (2.1) Boys = 35 Girls = 35 92.9 7y 9mo (1y 4mo) 1.3 (0.1) 27.4 (5.5) 57.0 (9.6) 16.4 (1.8)
3.7 (1.4) Boys = 36 Girls = 34 91.4 8y 1mo (1y 3mo) 1.3 (0.1) 30.2 (9.3) 60.4 (8.7) 17.6 (3.2)
t = 25.3, p < 0.001 Chi = 0.03; p = 0.87
Preferred hand (% right handed) Age (mean (SD) in years, months) Height (mean (SD) in m) Weight (mean (SD) in kg) Waist circumference (mean (SD) in cm) BMI (mean (SD) in kg/m2)
Chi = 0.10; p = 0.75 t = 0.8; p = 0.45 t = 0.5; p = 0.62 t = 2.2; p = 0.03 t = 2.2; p = 0.03 t = 2.8; p = 0.05
MABC-2: Movement Assessment Battery for Children 2nd Edition; SD: standard deviation; m: meters; kg: kilogram; cm: centimeters; BMI: body mass index; TD: Typically Developing; DCD: Developmental Coordination Disorder.
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
1092
Table 2 Descriptive results [mean (SD)] of aerobic capacity, muscle strength and muscle power of the Developmental Coordination Disorder (DCD) and agematched comparison group (TD). Variable
TD (n = 70) Mean (SD)
DCD (n = 70) Mean (SD)
t
p
Functional strength measure Overarm throwing [cm] Standing-long-jump [cm] Underarm throwing [cm] Chest pass [cm] Sit-to-stand [repetitions/30 s] Lifting box [repetitions/30 s] Stair climbing [steps/30 s] Lateral step-up (preferred leg) [repetitions/30 s] Lateral step-up (non-preferred leg) [repetitions/30 s]
204.9 112.2 284.2 170.4 26.9 19.7 73.6 35.1 32.3
(55.1) (22.4) (68.2) (32.4) (4.4) (5.6) (11.1) (5.7) (4.9)
201.0 (63.3) 99.19 (27.31) 262.8 (81.2) 161.6 (44.2) 20.0 (4.8) 16.2 (5.9) 61.0 (11.3) 27.8 (5.6) 25.5 (5.8)
0.61 3.39 1.99 1.74 8.87 3.94 6.66 7.30 7.14
0.54 0.001 0.048 0.08
Contents lists available at ScienceDirect
Research in Developmental Disabilities
Physical fitness in children with Developmental Coordination Disorder: Measurement matters Gillian D. Ferguson a,b,*, Wendy F.M. Aertssen c, Eugene A.A. Rameckers c,d, Jennifer Jelsma a, Bouwien C.M. Smits-Engelsman b a
University of Cape Town, Faculty of Health Sciences, Department of Health and Rehabilitation Sciences, Suite F45: Old Main Building, Groote Schuur Hospital, Main Road, Observatory 7925, Cape Town 8000, South Africa Katholieke Universiteit Leuven, Faculty of Kinesiology and Rehabilitation Sciences, Department of Kinesiology, Movement Control and Neuroplasticity Research Group, Tervuursevest 101, Postbox 1501, B-3001 Heverlee, Belgium c Avans + University of Professionals, Department of Physiotherapy, Heerbaan 14-40, Postbox 2087, 4800 CB Breda, The Netherlands d Maastricht University Medical Centre, Department of Rehabilitation Medicine & Adelante Center of Expertise in Rehabilitation & Audiology, P.O. Box 616, 6200 MD Maastricht, The Netherlands b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 December 2013 Received in revised form 29 January 2014 Accepted 29 January 2014 Available online 26 February 2014
Children with Developmental Coordination Disorder (DCD) experience considerable difficulties coordinating and controlling their body movements during functional motor tasks. Thus, it is not surprising that children with DCD do not perform well on tests of physical fitness. The aim of this study was to determine whether deficits in motor coordination influence the ability of children with DCD to perform adequately on physical fitness tests. A case–control study design was used to compare the performance of children with DCD (n = 70, 36 boys, mean age = 8y 1mo) and Typically Developing (TD) children (n = 70, 35 boys, mean age = 7y 9mo) on measures of isometric strength (hand-held dynamometry), functional strength, i.e. explosive power and muscular endurance (Functional Strength Measurement), aerobic capacity (20 m Shuttle Run Test) and anaerobic muscle capacity, i.e. muscle power (Muscle Power Sprint Test). Results show that children with DCD were able to generate similar isometric forces compared to TD children in isometric break tests, but were significantly weaker in three-point grip strength. Performance on functional strength items requiring more isolated explosive movement of the upper extremities, showed no significant difference between groups while items requiring muscle endurance (repetitions in 30 s) and items requiring whole body explosive movement were all significantly different. Aerobic capacity was lower for children with DCD whereas anaerobic performance during the sprint test was not. Our findings suggest that poor physical fitness performance in children with DCD may be partly due to poor timing and coordination of repetitive movements. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Physical fitness Strength Anaerobic muscle capacity Muscle power Aerobic capacity Developmental Coordination Disorder South Africa
1. Introduction The American College of Sports Medicine (ACSM) defines physical fitness as a set of measurable health and skill-related attributes that include body composition, cardiorespiratory fitness (CRF), muscular fitness, flexibility, and neuromotor * Corresponding author at: University of Cape Town, Department of Health and Rehabilitation Sciences, Suite F45: Old Main Building, Groote Schuur Hospital, Main Road, Observatory 7925, Cape Town 8000, South Africa. Tel.: +27 21 406 6045; mobile: +27 82 9743924. E-mail addresses: [email protected] (G.D. Ferguson), [email protected] (Wendy F.M. Aertssen), [email protected] (Eugene A.A. Rameckers), [email protected] (J. Jelsma), [email protected] (Bouwien C.M. Smits-Engelsman). http://dx.doi.org/10.1016/j.ridd.2014.01.031 0891-4222/ß 2014 Elsevier Ltd. All rights reserved.
1088
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
fitness (Garber et al., 2011). In the last decade, physical fitness in children with Developmental Coordination Disorder (DCD) has gained recognition as an important factor influencing performance in daily activities and as a mediator of health and wellbeing (Wahi et al., 2011). Children with DCD are reported to have reduced levels of physical fitness (Nascimento et al., 2013; Rivilis et al., 2011; van der Hoek et al., 2012) and are considered to be at increased risk for cardiovascular problems later in life (Cairney, Hay, Veldhuizen, & Faught, 2011). Studies examining body composition report that children with DCD have higher body mass indices (BMI) (Rivilis et al., 2011), higher body fat percentage (Cairney, Hay, Faught, & Hawes, 2005) and increased waist circumference (Cairney, Hay, Veldhuizen, Missiuna, et al., 2010; Wahi et al., 2011) compared to their Typically Developing (TD) peers. Regular participation in moderate to vigorous physical activity has therefore been recommended to reduce the risk of children developing cardiovascular conditions later in life (Lipnowski & Leblanc, 2012). However, participation in physical activity is often hampered by the limited motor performance capacity of children with DCD (Fong et al., 2011; Haga, 2009). Decreased CRF in DCD has been reported in several studies in which aerobic capacity was measured using field-based running tests such as the Le´ger 20 m shuttle run test (20 mSRT) (Rivilis et al., 2011) or less frequently, in laboratory tests using cycle ergometry (Cairney, Hay, Veldhuizen, & Faught, 2010) and treadmill protocols (Chia, Reid, Licari, & Guelfi, 2013). Laboratory-based measures of CRF (i.e. volume of oxygen consumed at maximal physical exertion/VO2max) are considered to be the gold standard for assessing aerobic capacity, whereas field-based measures have been criticized for the confounding factors associated with measuring maximal effort in the absence of objective indicators of exertion (Cairney, Hay, Veldhuizen, & Faught, 2010). In DCD specifically, the main factors associated with poor performance in field-based tests of CRF are thought to be related to lowered perceived self-efficacy (Cairney, Hay, Wade, Faught, & Flouris, 2006), low motivation and reduced levels of physical activity (Cairney, Hay, Faught, Wade, et al., 2005). Despite this, various authors agree that field-based running testing using standardized protocols, such as the Le´ger 20mSRT, are valid and reliable means to assess aerobic capacity in children with and without DCD (Cairney, Hay, Veldhuizen, & Faught, 2010). In contrast to the endurance running tests used to measure aerobic capacity, tests of anaerobic capacity include maximal running speed tests (e.g. 10 m 5 m, 20m and 50m sprint tests). Importantly, Verschuren, Takken, Ketelaar, Gorter, and Helders (2007) highlight that the agility requirements within the 10 m 5 m sprint test may confound the interpretation of anaerobic capacity in children with poor coordination (Verschuren et al., 2007). Another important attribute of physical fitness is flexibility. The sit and reach test is the most commonly reported flexibility measure used among children with DCD (Rivilis et al., 2011). Results show that children with DCD have a heterogeneous flexibility profile, with some studies showing poorer flexibility (Cantell, Crawford, & Tish Doyle-Baker, 2008; Hands, Larkin, Parker, Straker, & Perry, 2009) and others reporting no difference in flexibility compared to TD children (Schott, Alof, Hultsch, & Meermann, 2007; Tsiotra, Nevill, Lane, & Koutedakis, 2009). Concerning muscular fitness, three elements are typically evaluated: muscle strength, power and endurance. Findings from studies using either isometric or isokinetic dynamometry, which are considered the most robust forms of measuring muscular strength report that muscle strength is decreased in most muscle groups in DCD (Raynor, 2001; van der Hoek et al., 2012). Muscle power and muscular endurance tests on the other hand, are commonly used to make inferences about anaerobic muscle capacity. Tests of explosive muscle power examine parameters such as distance covered (e.g. throwing a heavy ball or performing a standing-long-jump) whereas tests of anaerobic muscle endurance measure the maximal number of repetitions within a specific time constraint (e.g. number of sit- or push-ups executed in 30 s). While the extent to which motor coordination deficits influence performance on these tests is acknowledged (Raynor, 2001; Rivilis et al., 2011) few studies have examined the relationship between muscle fitness and task constraints in DCD. The term neuromotor fitness is a collective noun, introduced by Garber et al. (2011) to describe motor skills such as balance, coordination, agility, and proprioceptive ability. Neuromotor fitness is a skill-related component of physical fitness and considered to be important in injury prevention. Neuromotor skills are by definition, functional skills and the motor tasks used to evaluate neuromotor fitness include running, walking on a line or hopping. Outcome measures designed to evaluate motor performance in children with DCD include standardized measures such as the Movement Assessment Battery for Children 2nd Edition (MABC-2) (Henderson, Sugden, Barnett, & Smits Engelsman, 2010), the Bruininks–Oseretsky Test of Motor Proficiency – 2nd edition (BOT-2) (Bruininks & Bruininks, 2005) and the McCarron Assessment of Neuromuscular Development (MAND) (McCarron, 1997). On examination of the items in these tests, it is evident that aspects of neuromotor fitness (i.e. balance, agility, coordination) are evaluated in each test. Importantly, adequate performance in neuromotor fitness tests is influenced by the ability to mitigate the variable influence of external forces and environmental constraints affecting movement quality. Evidently, motor proficiency plays an important role since carefully graded and well-timed muscle contractions lead to more economical and efficient ways of moving. Poor balance and agility in children with DCD (Chia, Licari, Guelfi, & Reid, 2012) may therefore explain their less favorable performance on neuromotor fitness measures. In DCD, compensatory strategies for motor control deficits are likely to influence physical fitness outcomes. One of the strategies frequently used in early stages of skill learning is co-activation of muscles, which leads to increased stability but can also potentially hamper force production (Raynor, 2001). Since DCD is a motor skill-learning deficit by definition, it is likely that the fine-tuning (grading) of multi-joint movements used in agility or dynamic power tests will be harder to optimize. Freezing joints may then be used as a temporary solution for controlling degrees of freedom or as a kind of mechanical filter to suppress the effects of force variability (Smits-Engelsman & Wilson, 2013).
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
1089
The assessment of particular components of physical fitness rest on the premise that the limiting factor in test performance is fatigue following a certain time period or number of repetitions, rather than the level of difficulty of the test (Cairney et al., 2006). Importantly, this may be different for children with coordination problems. For instance, the control of velocity and force changes that occur during an explosive throwing action and the implicit knowledge of how to use stored energy during a standing-long-jump (by exploiting the effect of the stretch-shortening cycle) will give an experienced and coordinated child an advantage in throwing and jumping distance even with the same generated muscle strength. In the current study, we suggest that, as both co-ordination and fitness elements are assessed concurrently. Isolated forms of isometric or isokinetic muscle contractions, using dynamometry, could improve the validity of measuring muscular fitness in children with poor co-ordination. Hand held dynamometry (HHD) for isometric strength attempts to isolate muscle action by controlling for joint movement. Although physical (e.g. BMI, ligament, tendon factors) and other aspects (e.g. motivation, attention, cognitive level) influence HHD outcomes, validity reports from studies conducted among children are generally positive with intra-class correlation coefficient values ranging from 0.73 to 0.99 (Beenakker, van der Hoeven, Fock, & Maurits, 2001; Brussock, Haley, Munsat, & Bernhardt, 1992). Inclusion of isometric measures is thus useful in understanding the contribution of muscle strength to function. However, in real life, actions seldom involve a pure form of isolated muscle action. Moreover, evidence suggests that the relationship between isometric strength and the ability to perform a functional task in which strength is required is not linear (Mattar & Sobreira, 2008). Thus, an evaluation of the functional impact of DCD would need to incorporate both complex and simple muscle actions. The Functional Strength Measure (FSM) is newly developed instrument, which was designed specifically to measure strength within a standardized functional task (Smits-Engelsman & Verhoef-Aertssen, 2012). The FSM measures two aspects of muscular fitness i.e. muscular endurance and explosive strength, within eight functional tasks. Tasks related to muscular endurance include lifting a heavy box repeatedly, repetitive sitting to standing, running up and down stairs, and performing repetitive lateral step ups. Tasks related to explosive muscle power include a standing long jump and performing a chest pass, over- and under-hand throwing using a heavy beanbag. In the current study, the aim was to determine whether poor performance on physical fitness tests is due to muscle weakness and poor CRF or deficits in using this strength in a functional context. To determine whether coordination is an important factor in physical fitness measures, we used various outcome measures, each containing tasks with different coordination requirements. Since most of the tests used require a certain amount of agility, we hypothesized large differences (effect sizes) between TD and DCD groups in tests that require more coordination (e.g. standing long jump; running up and down stairs) and small differences in simpler tasks such as single joint strength measures. The specific objectives were to compare TD and children with DCD in terms of aerobic and anaerobic muscle capacity, isometric strength, muscle endurance and explosive muscle power using a combination of functional and isolated measures. 2. Materials and methods 2.1. Research design and setting A cross-sectional, case–control study design was conducted in three mainstream primary schools situated in a lowincome area in Cape Town, South Africa. The implementation of physical education at these schools was limited by resource constraints and the accessibility of safe play-areas and opportunities to participate in sports and other physical activities were limited for all children. Since only one child participated in sports, information regarding participation in physical activity was not formally assessed. Ethical approval and permission was granted by the University of Cape Town Human Research Ethics Committee and the Western Cape Education Department. All parents and their children provided written informed consent and assent to participate in this study and to publication of the results. 2.2. Participants Children between the ages of six and ten years old, in grades 1–4, whose parents’ had given consent (N = 148) were recruited using convenience sampling. Exclusion criteria were: (1) failing a grade more than once and (2) a diagnosis of cerebral palsy or other significant neurological or medical condition as reported by a parent. Three children were excluded based on these criteria. Children were eligible for inclusion in the DCD group if their motor function in daily life was considered to be problematic according to their teacher and/or parent (Criterion B of the DSM IV diagnostic criteria) and they scored 5th percentile on the MABC-2 (Criterion A of the DSM IV diagnostic criteria). Inclusion criteria for the TD control group were if their motor function in daily life was considered to be within normal range for age and gender according to their teacher and/or parent and they scored above the 16th percentile on the MABC-2. Seventy children were identified as having DCD (boys = 36, girls = 34) and 75 were classified as TD (girls = 39, boys = 36) according to our criteria. A ratio of 1:1 was used to randomly select one TD child, matched for age and gender to every child identified with DCD. Since there were more girls in the TD group, we matched one girl of the same age to a boy with DCD and the remaining four TD girls were excluded. The final effective sample thus consisted of 140 children (boys = 71, girls = 69, mean age: 8.01, SD = 1.35).
1090
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
2.3. Outcome measures 2.3.1. Anthropometric measures Standardized anthropometric measurements were taken. Standing height (in centimeters) was measured without shoes, heels together using a tape measure fixed to a wall. Weight was measured in kilograms (0.1 kg accuracy), using a calibrated scale and BMI was calculated using the formula (weight/height2). Waist circumference was measured in centimeters, using a tape measure positioned around the abdomen in line with the top of the iliac crests. 2.3.2. Stand and reach test The Stand and reach test item, taken from the Beighton Scale of Joint Hypermobility (Beighton, Solomon, & Soskolne, 1973) was used to assess flexibility of the lower back and hamstring muscles. Following a brief warm up period, children were asked to stand with feet together, and then bend forward, with straight arms and knees and place both hands flat on the floor. 2.3.3. The Functional Strength Measure (FSM) The FSM (Smits-Engelsman & Verhoef-Aertssen, 2012) was used to assess maximal explosive power in one movement (four items: standing-long-jump, overarm throwing, underarm throwing, and chest pass), and muscle power where weight (body or object) was moved at maximal speed to generate a maximum number of repetitions within 30 s (four items: sit-tostand, lateral step-up, lifting a box and stair climbing). Each item consists of an instruction and demonstration phase followed by a practice phase, where participants are encouraged to execute the task. Feedback on how to correct the movement where necessary is given to those who demonstrate difficulty during practice phase. Once testers are satisfied that children know what to do (backward demonstration), formal trials may begin. Three trials were conducted for each task and the results from the best trial were scored. After testing, the highest score achieved on items in which left and right were tested was designated at the preferred limb for that particular item. Test reliability is moderate to high (ICC ranging from 0.73 to 0.91) (Smits-Engelsman & Verhoef-Aertssen, 2012). Concurrent validity of the FSM-items and handheld dynamometer is reported to range between 0.46 and 0.69 for the lower limb and between 0.52 and 0.74 for upper limb items. This suggests that while constructs related to strength are being tapped into during the test, it also suggests the presence of additional factors unrelated to one maximal isometric muscle contraction. Divergent validity assessed against the MABC-2 items yielded correlation values between 0.27 and 0.50. This suggests that items of the FSM are weakly correlated with coordination ability. 2.3.4. Hand-held dynamometer (HHD) The MicroFET-2 (Hogan Health Industries Inc., USA) and the Lafayette Manual Muscle Testing System (Model 01163, Lafayette Instrument Company, USA) were used to assess isometric muscle strength. The protocol for positioning and testing children outlined by Beenakker et al. (2001) was adopted. ‘‘Break’’ tests, where the examiner gradually overcomes the muscle strength generated by the participant, were used to evaluate elbow flexors, elbow extensors and knee extensors. The ‘‘make’’ test, defined as exerting maximal strength against the HHD, was used to asses three-point grip strength. Practice trials were given between each new test item using verbal instruction, demonstration and practice. Participants were corrected during practice and research assistants made sure that participants understood the requirements of the test. Each muscle group was tested three times and peak forces, measured in Newton, were recorded for each trial. Research assistants alternated between testing left and right sides of various muscle groups. The best scores of the three trials were used for analysis. After testing, the highest score achieved on items in which left and right were tested was designated at the preferred limb for that particular item. 2.3.5. Muscle Power Sprint Test (MPST) The MPST is an intermittent sprint test consisting of six, timed 15 m sprints (Verschuren et al., 2007). It was designed to measure muscle power with good reliability (Douma-van Riet et al., 2012). The time to complete each run is measured in milliseconds. Power output is calculated using the equations specified by Verschuren et al. (2007). Peak power is defined as the highest power that is generated of all six sprints and the mean power is the average power output over the six sprints. Greater mean power indicates the ability to maintain power output over time and is representative of anaerobic running performance. The 15 m distance required for the MPST was marked by a painted line on the playground and cones were placed 1 m beyond the line where children were encouraged to stop. Children were tested in groups of three to six and were instructed to run as fast as possible and cross the line. Explanations and practice trials were given followed by a brief rest period before the actual trial. A 30-s rest was allowed between trial runs. 2.3.6. 20 m Shuttle Run Test (20mSRT) The 20mSRT was used to measure aerobic fitness (Le´ger, Mercier, Gadoury, & Lambert, 1988). A practice trial was given and a short rest period included before the start of the actual trial. Participants were verbally encouraged to keep running for as long as possible. For some children, it is difficult to run at the audio signal that determines the running speed. Therefore, one tester ensured that children understood the principle of the Shuttle Run Test by running with them throughout the test.
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
1091
Since it is understood from the literature that BMI may be a confounding factor in performance of children with DCD (Tsiotra et al., 2009) we used the Matsuzaka formulae (Matsuzaka et al., 2004) to calculate VO2max in this study. This formula takes into consideration the impact of gender and body mass index (BMI) on VO2max and may yield more valid estimates of CRF (Barnett, Chan, & BruceIain, 1993; Matsuzaka et al., 2004; Ruiz et al., 2008). The Le´ger formula was also calculated for comparison. 2.4. Procedure Recommendations outlined in the Diagnostic and Statistical Manual of Mental Disorders IV (American Psychiatric Association, 2000) were used to identify children with DCD. Initially, teacher and parent questionnaires were administered to determine whether the child’s motor coordination ability affected their performance in everyday life (Criterion B of the DSM IV diagnostic criteria). The questionnaire, developed by the authors, consisted of a description of common functional problems that children with DCD present with and one closed- and one open-ended question asking parents and teachers to give their opinion regarding whether they thought the child had a motor coordination problem and why they thought so. Next, the Movement Assessment Battery for Children 2nd edition (MABC-2) (Henderson, Sugden, & Barnett, 2007) was used to identify children whose motor performance scores were at or below the 5th percentile (Criterion A of the DSM IV diagnostic criteria). All children participated in a standardized warm up phase before testing which including marching across the playground while swinging arms. This was followed by a brief rest period in which the individual tests were explained. Each test was conducted a different day over a period of one week. Running tests (MPST and 20mSRT) were performed on a tarred playground and other tests (FSM and HHD) were conducted in a quiet room on the school premises. All tests were conducted by physiotherapists and physiotherapy students who received training on the instruments used and were unaware of the child’s group affiliation. 2.5. Data analysis A power calculation was conducted based on the means and standard deviations of results obtained in previous studies using the 20mSRT (Cairney, Hay, Faught, Flouris, & Klentrou, 2007) and sit-to-stand scores of the FSM (Smits-Engelsman & Verhoef-Aertssen, 2012). Using an online sample size calculator we determined that sample sizes between 32 and 43 children per group would be sufficient to detect a difference at p = 0.001 level with 80% power (http://www.stat.ubc.ca). All data were analyzed using SPSS 20.0 (IBM, 2011). Pearson’s Chi-square test was used to compare differences in gender and preferred hand between groups. Shapiro–Wilks tests showed that fitness data were not normally distributed, thus raw scores were transformed using natural logs. Independent t-tests (two-tailed) were used to compare outcomes between groups. The post hoc ANCOVA procedure was used to test if the between-group effect remained significant if BMI was used as a covariate. Alpha was set at 0.05. Estimates of effect size (Cohen’s d) were calculated for group comparison. This measures the magnitude of the difference between the mean scores of groups, divided by a pooled SD. The magnitude of the mean effect size estimates (d) were interpreted according to the conventions of Cohen: 0.30 (small), 0.50 (moderate), 0.80 (large) and >1.00 (very large effect size) (Fern & Monroe, 1996). 3. Results Groups (TD n = 70, DCD n = 70) were comparable in terms of age, gender, handedness and height. However, weight, BMI, and waist circumference were significantly higher for the DCD-group. Four children were overweight (>85th percentile) in the TD and nine in the DCD-group. The differences between the two groups of children are presented in Table 1. Table 2 shows the results of the group comparison on the physical fitness tests.
Table 1 Characteristics of participants with Developmental Coordination Disorder (DCD) and age-matched Typically Developing (TD) children. Variable
TD (N = 70)
DCD (N = 70)
Statistics
MABC-2 standard score Gender (n)
11.2 (2.1) Boys = 35 Girls = 35 92.9 7y 9mo (1y 4mo) 1.3 (0.1) 27.4 (5.5) 57.0 (9.6) 16.4 (1.8)
3.7 (1.4) Boys = 36 Girls = 34 91.4 8y 1mo (1y 3mo) 1.3 (0.1) 30.2 (9.3) 60.4 (8.7) 17.6 (3.2)
t = 25.3, p < 0.001 Chi = 0.03; p = 0.87
Preferred hand (% right handed) Age (mean (SD) in years, months) Height (mean (SD) in m) Weight (mean (SD) in kg) Waist circumference (mean (SD) in cm) BMI (mean (SD) in kg/m2)
Chi = 0.10; p = 0.75 t = 0.8; p = 0.45 t = 0.5; p = 0.62 t = 2.2; p = 0.03 t = 2.2; p = 0.03 t = 2.8; p = 0.05
MABC-2: Movement Assessment Battery for Children 2nd Edition; SD: standard deviation; m: meters; kg: kilogram; cm: centimeters; BMI: body mass index; TD: Typically Developing; DCD: Developmental Coordination Disorder.
G.D. Ferguson et al. / Research in Developmental Disabilities 35 (2014) 1087–1097
1092
Table 2 Descriptive results [mean (SD)] of aerobic capacity, muscle strength and muscle power of the Developmental Coordination Disorder (DCD) and agematched comparison group (TD). Variable
TD (n = 70) Mean (SD)
DCD (n = 70) Mean (SD)
t
p
Functional strength measure Overarm throwing [cm] Standing-long-jump [cm] Underarm throwing [cm] Chest pass [cm] Sit-to-stand [repetitions/30 s] Lifting box [repetitions/30 s] Stair climbing [steps/30 s] Lateral step-up (preferred leg) [repetitions/30 s] Lateral step-up (non-preferred leg) [repetitions/30 s]
204.9 112.2 284.2 170.4 26.9 19.7 73.6 35.1 32.3
(55.1) (22.4) (68.2) (32.4) (4.4) (5.6) (11.1) (5.7) (4.9)
201.0 (63.3) 99.19 (27.31) 262.8 (81.2) 161.6 (44.2) 20.0 (4.8) 16.2 (5.9) 61.0 (11.3) 27.8 (5.6) 25.5 (5.8)
0.61 3.39 1.99 1.74 8.87 3.94 6.66 7.30 7.14
0.54 0.001 0.048 0.08
