Norris, Rosemary A. PT, MS; Wilder, Elaine PT, PhD; Norton, Jennifer PT, MS
Children with developmental disabilities experience postural problems that include difficulties with balance.1–5 Physical therapists use various tests to examine static and dynamic components of balance for children, at different ages, with and without disabilities. Some of these tests are designed specifically to evaluate balance1,5–12 and other tests evaluate more generalized gross motor functions that include balance components.2,4,13–15 Atwater et al1 examined static balance using the one-leg standing balance test in children aged 4 through 9 years without disabilities. They examined each child's ability to maintain balance on a moveable surface by using the tiltboard balance test. Franjoine et al10 modified the Berg Balance Scale16 to assess balance in school age children with mild to moderate motor impairments. Donohoe et al11 reported that the Functional Reach Test (FRT) may be a useful tool for the pediatric population.
The FRT was developed by Duncan et al17 to measure dynamic balance in adults. The FRT defines the maximal distance an individual is able to reach forward beyond arm's length in a standing position without loss of balance, taking a step, or touching the wall.17 In a series of studies with adults, Duncan et al17 established a test–retest reliability of 0.92 [intraclass correlation (ICC) 1,3] and interrater reliability of 0.98 (ICC [1,3]). The FRT was found to be valid for measuring dynamic balance when compared with a center of pressure excursion (Pearson correlation coefficient 0.71) and an electronic functional reach measurement device (Pearson correlation coefficient 0.69). Duncan et al17 reported that as age of adults increased FR scores decreased; height, weight, trunk length, and arm/foot length were all associated (r > 0.80) with the FR distance, the most significant predictors were height and age. The value of the FRT to predict fallers is controversial as some studies found the FRT scores predicted older adults at risk for recurrent falls17–19 while other studies20,21 reported no significant difference in FRT scores of elder fallers and nonfallers.
The FRT has been used to study dynamic balance in children.9,11,12 Donohoe et al11 reported results for 116 children of 5 through 15 years of age without disabilities for a reliability study and a descriptive study. Two experienced pediatric physical therapists collected the data. There were five age groups, in increments of 2 years, and the group sizes varied from 10 per group (13–15 years old) to 36 per group (7–8 years old). The investigators reported that all subjects were able to successfully perform the FRT.11 Donohoe used four trials for the reliability study and two trials for the descriptive study; this differed from Duncan's original protocol that used three trials. Reliability results reported an interrater reliability (ICC [2,1]) of 0.98; the intrarater reliability (ICC [2,1]) ranged from 0.83 to 0.97 and test–retest reliability (ICC [2,1]) ranged from 0.64 to 0.75. Mean and critical reach values (the lowest value at the 95% CI) were reported for each of the five age groups. In the regression analysis, age contributed to FR; as children aged, they reached farther, but none of the anthropometric variables (height, weight, and arm length) significantly explained FR variance. Donohoe et al11 hypothesized that the inability to reach beyond a critical reach value may indicate delay in the development of balance.
Habib and Westcott12 studied the effects of anthropometric factors (height, weight, and base of support) on the FRT scores in 180 children, 4.5 to 13.5 years, from Pakistan who were “typically developing.” Two pediatric physical therapists collected all data. Intrarater percent agreement ranged from 73% to 96%, and intrarater values across all ages (ICC [3,1]) were 0.73. For children 5 to 7 years old, height was the only variable that significantly predicted FRT mean scores (R2 = 0.25). For children aged 8 to 10 years, base of support (defined as length of feet) significantly predicted FRT mean scores (R2 = 0.33). None of the anthropometric factors predicted FRT mean scores in children aged 11 to 13 years. Collectively, 17% of the variance in FRT scores was attributed to age, and another 15% of the variance in FRT scores was attributed to height, weight, and base of support (BOS). According to Habib and Westcott,12 height was a more important predictor of FR than BOS in younger children aged 5 to 7 years.
Bartlett et al9 developed a balance test, the Pediatric Reach Test, for children developing typically, aged 3 to 12.5 years, and for children with a diagnosis of cerebral palsy, aged 2.6 to 14.1 years. Their study evaluated the validity and reliability of testing balance in forward and side reaching directions while sitting and standing. The researchers compared the reaching data with data on the limits of stability collected using a force platform. Although Bartlett et al9 did not compare their data with the FRT, the standing reach forward test in the Pediatric Reach Test is similar to the FRT. A general outcome of their study was that the Pediatric Reach Test is a “simple, valid, and reliable measure that can be used with children.”
Niznik et al6 reported FRT scores from a sample of 32 children with lower extremity spasticity. Subjects included children aged 5 to 18 years, and an age-matched group without disabilities. The purposes of the study were to determine if the FRT was reliable with this population, to determine the optimal number of trials for reliability, and to collect data on mean FR scores for this population. Intrarater reliability was 0.94 and 0.98 on days 1 and 2 of testing, respectively. Test–retest reliability was greater for the group of children with spasticity than the group of children without disabilities. Children with spasticity had a shorter FR distance, by an average of 40%, than children without disabilities. Niznik et al6 determined that children with lower extremity spasticity were able to successfully perform the FRT and suggested one practice trial and one test trial to determine FR.
The FRT is an appropriate, functional, and reliable measure of dynamic balance for children aged 4.5 to 15 years6,11 and for adults.17–19 Variability in the FRT has been reported for children and adults. We examined the variability of the FRT in available studies6,11,17,19,20,22 using the coefficient of variation. The coefficient of variation is a measure of relative variation between groups taking into account the differences in magnitude of the groups' means.23 The coefficient of variation for adults from 20 to 40 years old17,22 ranged from 9.6% to 11.6% and from 14.8% to 82.3% for adults 41 to 104 years old.17,19,20,22 The coefficient of variation for children 5 to 15 years old6 ranged from 20.8% to 61.9%. In children 5 and 6 years old with lower extremity spasticity, the coefficient of variation was 20.8%.6 In the Donohoe et al11 study, the group of 5- and 6-year-old children without disabilities had a coefficient of variation of 21.3%. As reported in the above studies, variability for young children is fairly low. Interrater reliability of the FRT has not been reported; neither have normative values for children younger than 4.5 years of age without disabilities, nor the influence of anthropometric characteristics. The FRT reportedly is sensitive to change in adults receiving rehabilitation.24
The purposes of this study were to describe normative values and variability of the FRT scores and to examine the effect of anthropometric measures, ie, age, height, and weight, on the FRT scores in children 3 to 5 years old without disabilities.
We used a sample of convenience to recruit children from two early childhood screening sites for 3- and 4-year-old children and from 3 day-care facilities for 5-year-old children. Each age category included children from their respective birthday to 1 day before the next birthday, eg, 3-year-old children were eligible from their third birthday until 1 day before their fourth birthday. The early childhood screenings were performed for school placements. Children were recruited by placing signs in the site waiting areas where parents would gather asking for volunteers for the study. After informed consent was granted by a parent or legal guardian, a medical questionnaire was completed. Results from the early childhood screenings were not available. Children were excluded from the study if a parent or legal guardian reported medical, neurological, orthopedic, balance, or visual disorders, or history of treatment for an ear infection within the past 6 weeks, or if they wore splints or orthotic devices. To generalize the results to a more typical population of children from 3 through 5 years of age, potential participants were also excluded if their height or weight fell below the 10th or above the 90th percentile for gender and age.25 The most common exclusions were recent ear infections in 3-year-old children, recent ear infections and height or weight above the 90% in 4-year-old children, and height or weight above the 90% in 5-year-old children. The children who were not excluded were asked for their assent to participate. While 188 children were initially recruited, 67 were excluded and 121 successfully completed the study (Table 1). After agreeing to participate, children had to complete three successful FR trials within six attempts. Of the 121 children who successfully completed the study, 55 were 3 years old, 44 were four 4 years old, and 22 were 5 years old. Sixty-four were boys and 57 were girls. Table 2 presents the height and weight characteristics of the children included in the study. Anthropometric measures increased as a function of age.
Materials used in this investigation included an aluminum meter-stick to measure FR distance, a bubble level to ensure the meter-stick was level, a calibrated scale for measurement of body weight, and an anthropometer to measure standing height. The same equipment was used at all data collection sites. Before measuring FR, the children's shoes and socks were removed. Duncan et al,17 Donohoe et al,11 and Niznik et al6 used foot tracings of individual subjects to increase the likelihood that foot position was similar over multiple trials. Based on the first author's clinical experience, 3- and 4-year-old children were distracted by the paper under their feet and refused to perform the FRT. Therefore, a piece of masking tape was placed on the floor perpendicular to the wall and all children were instructed to align their toes behind the tape.
Each child was asked to stand with the lateral aspect of the right shoulder parallel to the wall. A leveled meter-stick was secured to the wall at the height of the child's acromion. The FRT was demonstrated and described as follows: “Make a fist. Raise your (right) arm this high (shoulder height). Reach forward as far as you can, but don't fall or take a step.” Each child was allowed to perform two practice trials that were not recorded. Demonstrations and verbal instructions were repeated in the same format each time. To measure FR distance an initial measure was taken with the child's arm raised horizontally (approximately 90° of shoulder flexion) using the placement of the third metacarpal along the meter-stick. A second measure was taken after reaching, again using the location of the third metacarpal along the meter-stick (Fig. 1). The parameter measured was the distance the child reached while standing. The distance was measured to the nearest 0.5 cm. A trial was considered successful if the FR was performed without the child stepping, touching the wall, or receiving assistance from the investigator.17 The method of reach was not controlled. The average of the difference between the initial and second measures for each of the three successful FR trials was used as the datum for each child.17 Brief rest periods of approximately 5 to 10 seconds were allowed.
Two researchers, with 28 years of combined experience, established interrater reliability with the first 10 children from the 3- and 5-year-old groups and the first nine children from the 4-year-old group who successfully performed the FRT. The raters viewed each child's reach simultaneously and recorded the initial and final positions. After the FR values in each age group were collected, the ICC values (ICC [2, k]) were calculated and were 0.97, 0.95, and 0.99, respectively, for the 3-, 4-, and 5-year-old children. These ICC values were considered appropriate for continuing data collection.
Interrater reliability, ICC values (ICC [2,k]), were calculated using a random effects model.26 Means and standard deviations for height and weight were calculated for each age group and gender. The mean of the three successful FR trials was analyzed for a given child. The 95% confidence interval (CI)27 of the mean FR for each age group was also calculated.
To determine FR cutoff values by age, we first converted the mean FR distances to z scores to standardize and normalize the data. In a normal distribution 2.27% of the cases will fall more than two standard deviations below the mean. A z score of −2.0 corresponds to two standard deviations below the mean.23 Because few functional reach scores in a normal distribution fell two standard deviations below the mean, we decided to use values equal to two standard deviations as our cutoff point to indicate a short FR for a given age. Finally, since z scores have no unit value, we converted the z score of −2.0 to centimeters. To determine the influence of anthropometric measures on FR distance, a stepwise multiple regression analysis was performed (SPSS, Inc., Chicago, IL). Mean FR was the dependent variable and the predictor variables were age (yr), weight (kg), and height (m). Although studies in adults17 and children12 measured foot length, the study by Habib et al12 did not find foot length to be predictive of FR in the children younger than 8 years old. Donohoe did not measure foot length of the children in her study.11 Our pretesting revealed that trying to measure foot length was too distracting for the younger children, so we did not use this measure. Before performing the regression analysis, the data were analyzed for normality, linearity, and outliers. The residuals were analyzed for homoscedasticity. On the basis of this analysis, the assumptions for the regression analysis were met and transformations of the data were not needed.28
The mean FRT scores and the width of the 95% CI of the mean increased with age (Table 3). With a 95% CI, if samples of the same size are drawn repeatedly from a similar population, we are 95% confident that they include the population mean. The width of the CI is an indication of precision and a smaller CI width indicates that the sample mean is a more precise estimate of the population mean.27 Table 3 shows that the 5-year-old group was the most variable (13.8–17.7 cm), while the 3- and 4-year-old children had similar levels of variability (10.7–12.1 cm and 12.7–14.5 cm, respectively). Cutoff FR distances that indicated a short FR were equal to or less than 6.2 cm for 3-year-old children, equal to or less than 7.6 cm for the 4-year-old group and equal to or less than 6.9 cm for 5-year-old children. Using these criteria, two children who were 3 years old, no child who was 4 years old, and one 5-year-old child had FR distances that were short for their ages. Multiple regression analysis was performed to examine the influence of age, height, and weight on the FRT score. Across the three age groups, the only significant predictor of FRT scores was weight, which accounted for 34% of the variance in FR.
We found that mean FRT scores increased with age, similar to the results reported by Habib and Westcott.12 The 3- and 4-year-old children's FR data add to the existing literature and can be used by therapists when interpreting FRT scores in young children. For example, a 3-year-old child with an FR at or less than 6.15 cm has a short FR. Furthermore, since Donohoe et al11 combined FR data from 5- and 6-year-old children, we believe that our data from the 5-year-old group provide therapists more specific information to interpret FRT scores for children who are 5 years old. The cutoff value for the 5-year-old group (7.0 cm) was a shorter reach when compared with the 4-year-old group (7.7 cm). This is because the variability of the 5-year-old children was greater than the variability observed in the 4-year-old children. The most variable functional reach was in the 5-year-old group of children compared with the 3- and 4-year-old children. A larger sample of 5-year-old children may provide a better estimate of the variability of this age group.
We also examined the influence of anthropometric measures on FRT scores. For the 3- and 4-year-old children, the significant predictor was weight. No anthropometric measures were significant predictors for the 5-year-old children; however, the regression analysis may be limited by the small sample size of the 5-year-old group. In contrast to our findings, Habib et al12 found that height was significantly related to FRT in children 5 to 7 years old. Donohoe et al11 reported that FRT scores increased as a function of age up to 11 to 12 years and then the FRT scores reached a plateau. According to Donohoe et al,11 gender, height, weight, and arm length did not significantly predict FR.
Children may use different strategies to maintain their balance when it is challenged. We noted that children in this study consistently used a hip strategy during the FRT. Nashner29 describes the hip strategy as an automatic postural reaction used to correct anterior–posterior body sway by rotating the body about the hips in the sagittal plane. Children 4 to 6 years old had more variability in their hip strategies than did 3-year old children, but neither age group had adult hip strategy responses for postural control.30 In our study and the study by Donohoe et al,11 children commonly rose onto their toes while flexing forward at the hips in an attempt to increase the reach distance. We also found that 3-year-old children commonly attempted to increase the reach distance by rotating their trunks and upper bodies in the transverse plane.
Donohoe et al11 used two test trials to determine critical values and variability. They recommended the use of two practice trials followed by one measured trial. In contrast, Niznik et al6 found no difference among six test trials and recommended one practice trial followed by one test trial. Our post hoc analysis of the FR distance between trials by age revealed no significant difference, indicating that performance did not change over the three FRT trials.
In general, children understood the instructions and performed the FRT. However, 26% (n = 19) of the 3-year-old group refused to perform the FRT. Refusals were due to difficulty in understanding the task, distractions within the testing environment, and possible fatigue from participation in the screening process. Therefore, the FRT may need to be used with caution for 3-year-old children. Fewer 4-year-old children refused (6.4% or n = 3), and no 5-year-old children refused to perform the FRT.
The FRT is a measure of dynamic postural control in children. Other authors suggest that since the FRT challenges balance in one plane, multiple balance tests give more complete findings. We suggest using the FRT to determine a child's ability to reach forward and to maintain balance. The FRT may be used with other tests to assess dynamic and static balance abilities in children. These other tests may include the Alberta Infant Motor Scale,31 the Posture and Fine Motor Assessment of Infants,32 Peabody Developmental Motor Scales,15 one-leg standing,33 heel-to-toe standing, tandem walking, the Smart Balance Master,33 or the Pediatric Clinical Test of Sensory Integration for Balance.33
Results of this study indicate that the FRT is a feasible, clinical test that examines the dynamic balance of 3- to 5-year-old children. Use of the FRT may identify children with potential balance deficits at an early age. Although the FRT can be used effectively in children 4 to 5 years old, FRT is only one measure of dynamic balance and should be used in conjunction with other tests available to examine balance in children. The FRT should be used with caution with 3-year-old children because 26% of our sample refused to perform the FRT for various reasons. Further research is needed to determine the reliability and validity of the FRT when used to examine the balance abilities of children with disabilities.
1. Atwater S, Crowe T, Deitz J, et al. Interrater and test–retest reliability of two pediatric balance tests. Phys Ther. 1990;70:79–87.
2. Zaino C, Marchese V, Westcott S. Timed up and down stairs test: preliminary reliability and validity of a new measure of functional mobility. Pediatr Phys Ther. 2004;16:90–98.
3. Westcott S, Lowes L, Richardson P. Evaluation of postural stability in children: current theories and assessment tools. Phys Ther. 1997;77:629–645.
4. Palisano R, Hanna S, Rosenbaum P, et al. Validation of a model of gross motor function for children with cerebral palsy. Phys Ther. 2000;80:974–985.
5. Casselbrant M, Furman J, Mandel E, et al. Past history of otitis media and balance in four-year-old children. Laryngoscope. 2000;110:773–778.
6. Niznik T, Turner D, Worrell T. Functional reach as a measurement of balance for children with lower extremity spasticity. Phys Occup Ther Pediatr. 1995;15:1–15.
7. Pellegrino T, Buelow B, Krause M, et al. Test–retest reliability of the pediatric clinical test of sensory interactions for balance and the functional reach test in children with standing balance dysfunction. Pediatr Phys Ther. 1995;7:197.
8. Richardson P, Atwater S, Crowe T, et al. Performance of preschoolers on the pediatric clinical test of sensory interaction for balance. Am J Occup Ther. 1992;46:793–800.
9. Bartlett D, Birmingham T. Validity and reliability of the pediatric reach test. Pediatr Phys Ther. 2003;15:84–92.
10. Franjoine M, Gunther J, Taylor M. Pediatric balance scale: a modified version of the Berg Balance Scale for the school-age child with mild to moderate motor impairments. Pediatr Phys Ther. 2003;15:114–128.
11. Donohoe B, Turner D, Worrell T. The use of functional reach as a measurement of balance in boys and girls without disabilities ages 5 to 15 years. Pediatr Phys Ther. 1994;6:189–193.
12. Habib Z, Westcott S. Assessment of anthropometric factors on balance tests in children. Pediatr Phys Ther. 1998;10:101–109.
13. Williams E, Carroll S, Reddihough D, et al. Investigation of the timed ‘Up and Go' test in children. Dev Med Child Neurol. 2005;47:518–524.
14. Bruininks R, Bruininks D. (BOT-2) Bruininks-Oseretsky test of motor proficiency. 2nd ed. Bloomington, MN: AGS Publishing; 2005.
15. Folio M, Fewell R. Peabody Development Motor Scales Manual. Allen, TX: DLM Teaching Resource; 1983.
16. Berg K, Wood-Dauphinee S, Williams JI, et al. Measuring balance in the elderly: preliminary development of an instrument. Physiother Can. 1989;41:304–311.
17. Duncan P, Weiner D, Chandler J, et al. Functional reach: a new clinical measure of balance. J Gerontol. 1990;45:M192–M197.
18. Westcott D, Duncan P, Chandler J, et al. Functional reach: a marker of physical frailty. JAGS. 1991;40:203–207.
19. Duncan P, Studenski S, Chandler J. Functional reach: predictive validity in a sample of elderly male veterans. J Gerontol. 1992;47:M93–M98.
20. Gill D, Williams K, Williams L, et al. Multidimensional correlates of falls in older women. Int J Aging Hum Dev. 1998;47:35–51.
21. Cho C-Y, Kamen G. Detecting balance deficits in frequent fallers using clinical and quantitative evaluation tools. JAGS. 1998;46:426–430.
22. Hagemann P, Leibowitz M, Blanke D. Age and gender effects on postural control measures. Arch Phys Med Rehabil. 1995;76:961–965.
23. Pourtney L, Watkins M. Foundations of Clinical Research. 2nd ed. Norwalk, CT: Applegate and Lange; 2000.
24. Weiner D, Bongiorni D, Studenski S, et al. Does functional reach improve with rehabilitation? Arch Phys Med Rehabil. 1993;74:796–800.
25. Nelson W. Nelson Textbook of Pediatrics. 15th ed. Philadelphia, PA: WB Saunders Co.; 1996.
26. Shrout P, Fleiss J. Interclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86:420–428.
27. Sim J, Reid N. Statistical inference by confidence intervals: issues of interpretation and utilization. Phys Ther. 1999;79:186–195.
28. Tabachnick B, Eidell L. Using Multivariate Statistics. 3rd ed. New York: Harper Collins; 1996.
29. Nashner L. Sensory, neuromuscular, and biomechanical contributions to human balance. Paper presented at APTA Forum, 1989, Nashville, TN.
30. Shumway-Cook A, Woollacott M. The growth of stability: postural control from a developmental perspective. J Mot Behav. 1985;17:131–147.
31. Piper M, Pinnell L, Darrah J, et al. Construction and validation of the Alberta Infant Motor Scale (AIMS). Can J Public Health. 1992;83: S46–S50.
32. Case-Smith J. A validity study of posture and fine motor assessment of infants. Am J Occup Ther. 1992;46:597–605.
33. Liao HF, Mao PJ. Test-retest reliability of balance tests in children with cerebral palsy. Dev Med Child Neurol. 2001;43:180–186.
© 2008 Lippincott Williams & Wilkins, Inc.