INTRODUCTION
Functional mobility is a component of school participation for children, inside and outside the classroom. Throughout the school day, children are required to move within the school building and the classroom to write on the blackboard, deliver assignments or class supplies to a specified location, retrieve items from lockers or cubbies, and move to different workstations within the classroom or between rooms. First-grade students have been observed to move among activities and locations 15 to 20 times per day,1 whereas students in kindergarten can spend 25% of their day moving within the classroom.2
Physical therapy is one of many related services that may be required by children with disabilities to fully participate in educational settings3 and to complete school-related functional activities, including moving within the classroom. If limitations are noted, a physical therapist (PT) uses the appropriate tests, including established criterion- or norm-referenced measures, to determine the functional limitations and physical impairments that are affecting school participation.4
The Timed Up and Go test (TUG)5 is a norm-referenced measure that has established reliability for quickly assessing functional ambulatory mobility and dynamic balance in adults.6 It measures the amount of time it takes for a person to stand from sitting on a chair/bench, walk a distance, and return to the chair. In adults, the TUG can be performed with various types of chairs; armrests, tilted backs, or soft seats have little influence on test results.7
There are 4 studies of the TUG in children8–11; however, methodological differences across the studies make comparisons challenging. Sample sizes of children with typical development range from 1769 to 503,10 but the sample diversity is limited. Habib et al8 studied Pakistani children, and Williams et al9 studied Australian children, but did not mention ethnic diversity. Of the 503 subjects tested in Pennsylvania by Marchese et al,10 400 (79.52%) were white and the other 103 subjects represented other ethnicities: 32 African American (6.36%), 43 Asian (8.5%), 26 Hispanic (5.17%), 2 others (0.40%). Nicolini-Panisson and Donadio11 studied 459 children with typical development in Brazil, of whom 74% were white. Although different cultures are represented across the 4 studies, whether the results reflect the diversity of children in large American urban settings is not clear.
Procedures for administering the TUG varied among the 4 studies. Williams et al9 instructed participants to “walk only,” touch a target on the wall, and reported the average time for 3 trials. Habib et al8 instructed participants to go as fast as they could without running, did not use a target on the wall, and reported the average time for 2 trials. Marchese et al10 instructed participants to go as fast as they could without running, did not use a target on the wall, and used only 1 trial. Nicolini-Panisson and Donadio11 instructed participants to “walk as fast as possible,” used a target on the wall, and recorded 3 trials per subject. Trials 1 and 2 were performed on the same day, whereas trial 3 was conducted 1 week later. Nicolini-Panisson and Donadio11 tested participants with their shoes off; the other investigators did not report whether shoes were on or off.8–10 The average time to perform the TUG ranged from 3.78 seconds10 to 5.61 seconds,11 and younger participants were slower than older ones when age groups are reported (Table 1). Because faster walking speed seems to reduce gait variability,12 comparison of results across the TUG studies is unreliable if procedural differences (ie, the directions to walk fast vs a natural pace) are not taken into account.13
;)
TABLE 1: Comparison of Average TUG Speed in Seconds
The ages included in previous studies range from 3 to 21 year; however, Marchese et al10 report a single average time for all participants, whereas Williams et al9 and Nicolini-Panisson and Donadio11 report results for groups of children by age. In all the studies that tested children 10 to 13 years of age, participants were instructed to go as fast as they could. None of the investigators reported data on 10- to 13-year-old children with typical development walking at normal indoor or classroom paces.
Other variables such as culture and parenting emphases,8 the testing location (rural vs urban settings), level of opportunity for physical activity, body mass index (BMI), and testing procedures can have an effect on the performance of gross motor tests.14 Previous investigators of the TUG did not indicate whether data were collected in suburban or urban settings, and thus may not reflect the performance of children in urban communities. Nicolini-Panisson and Donadio11 proposed a regression model in which age and weight together are the strongest predictors of TUG values, and other variables, such as height, lower limb length, race, physical activity, sex, and BMI centile, did not improve its predictive ability. Finally, the subjects' BMI categories by age in years were not reported for any of the studies. Since 1980, a nationwide obesity epidemic has been reported by the Center for Disease Control (CDC),15,16 but it is not known whether there is a relationship between BMI and the speed at which children perform the TUG.
The simplicity of the TUG, including the minimal equipment and space it requires, makes it an attractive tool for school-based therapists, but current age-based reference data are needed for children with diverse backgrounds living in urban settings. This study seeks to expand the reference data for the TUG, by using a larger, more diverse sample of children living in an urban setting, and by exploring the effect of sex and BMI category by age on natural speed, appropriate for classroom settings.
METHODS
Subjects
This cross-sectional observation study was part of a larger study of 5 gross motor function assessments (TUG, Timed Up and Down Stairs, Timed Floor to Stand–Natural, 30-Second Walk, Shuttle Run); only the TUG results are reported here. A sample of convenience of school-aged volunteers, ranging from 5 years to 17 years of age, was recruited from 20 public elementary and middle schools from all 5 New York City (NYC) boroughs. Parent/guardian consent and student assent were received from all volunteers. Inclusion criteria for participation included age of 5 to 17 years, no orthopedic surgeries or injuries within the past 6 months, no history of neurological disorders, and no individualized education program. Students were tested during school hours. This study had approval from the NYC Department of Education (DOE) Institutional Review Board.
Measurement Reliability
Before the start of data collection, interrater reliability was established for the TUG using 22 children with typical development, tested simultaneously by 5 NYC DOE PTs, each with a minimum of 10 years of experience. Each had a minimum of 10 hours of training for administering the 5 standardized tests to assure uniformity and proficiency. Laminated cue cards for each test were printed to assure consistency of verbal commands during the testing process. Intertester reliability was established for the TUG using the average of 2 trials per clinician and a 2-way mixed, absolute, single measure intraclass correlation coefficient. The resulting intraclass correlation coefficient (3,1) = 0.988 indicates an excellent level of interrater reliability when the procedures are followed.17 This level is comparable or higher than those previously reported.8–10
Data Collection
On the day of testing, a member of the research team brought students with signed parent consent forms to the testing area in groups of 10 to 25. The testing process was explained and each subject was given a student assent form to sign if they wanted to continue. A random identification number was assigned to each student's data collection forms, where date of birth, age, sex, grade, height, weight, and testing results were recorded. Height was measured using a Charder HM200P PortStad Stadiometer according to CDC methods (Charder Electronic Co, Ltd., Taichung City, Taiwan).18 Weight was measured using a Lifesource Precision Health Scale ProFIT/IntelliSCALE (UC-321; A&D Medical, San Jose, CA). Outer garments and heavy items in pockets were removed. BMI category and percentile were calculated using an online BMI calculator (http://nccd.cdc.gov/dnpabmi/). Each student was assigned a BMI category on the basis of the CDC definitions. “Underweight” was defined as a BMI less than the fifth percentile, “healthy” weight was from the fifth to the 85th percentile, “overweight” was from the 85th to less than the 95th percentile, and “obese” was equal to or greater than the 95th percentile.15 Shoes were worn for all body measures and TUG trials, as this most accurately replicates conditions in the school setting. In a preliminary analysis of 519 students' BMI levels with shoes on versus off, a statistical but nonfunctional difference of 0.31 BMI points was found between the conditions (t (518) = −22.95, P ≤ .000). After heights and weights were recorded, the 10 to 25 students were divided into 3 smaller groups and assigned a starting station for the TUG, the Timed Floor to Stand, or the Timed Up and Down Stairs. After each group completed 2 trials of a test, the students rotated to the next station.
The original instructions from Podsiadlo and Richardson5 were used with the exception of using 2 trials rather than 3. Two pieces of tape (1″ × 24″) were placed on the floor 3 m (9 ft 10 inches) apart, determined by a measuring wheel (Measure Master by Rolatape, RT204; Watseka, IL). At 1 end was a Kaye Adjustable Bench, model S3A or S4A (Kaye Products Inc, Hillsborough, NC), positioned so each student was sitting in approximately 90° of hip and knee flexion. Time was measured with a handheld stopwatch (Super Sport Digital Stop Watch by Sweda, Item#: JW91 [Sweda Company LLC, La Puente, CA]).
Before the start of testing, an initial demonstration and explanation of the task was given. As each subject was tested, the rest of the students would stand or sit in a single row at least 3 ft behind the starting line to limit distractions. The subject was asked to sit on the adjustable bench with hands in his/her lap, hips and knees flexed to approximately 90°, feet flat on the ground, and toes touching the tape. The original verbal instructions from Podsiadlo and Richardson5 were used because the instructions from Williams et al9 were lengthy and less likely to keep the attention of children. The instructions were read from a preprinted cue card: “When I say go, stand up, walk to the line, turn around, walk back to the starting line, and sit back down on the bench. Walk, don't run. 1, 2, 3, GO.” The testers were positioned to allow full view of the starting and end lines. Subjects were timed from the cue “go” until they returned to sitting on the bench. They were permitted to use any method they chose to come to stand or return to sitting (eg, with or without use of hands to push off).
The times for 2 successful trials per subject were recorded. The test was repeated if the student ran, tripped, fell, or did not pass the second line with both feet. It was acceptable to have 1 foot step beyond the line and the second foot swing through, as long as the subject's full body passed the line. After a successful first trial, the subject went to the end of the student line to await the second trial.
Data Analysis
Demographics and normative values for performance of the TUG were determined by age. Means and standard deviations of TUG results were determined for each age. Pearson correlation was used to determine the relationship between TUG speed and age in years, and TUG speed and BMI. Paired t tests compared the 2 trial times for each subject, and an independent t test was used to compare time between males and females. A 1-way analysis of variance (ANOVA) was used to compare time among BMI categories within ages by year; linear and curved regressions were used to explore BMI trends. Statistical analyses were performed using SPSS Version 22.
RESULTS
Sample Description
Consent forms were distributed to the parents of approximately 18 231 students; 1653 students from 20 schools received parental consent to participate, of whom 1481 completed the TUG: 635 males and 846 females, ages 5 to 13 years. Another 162 students were either absent on the day of testing, did not have proper footwear, refused to participate, could not follow the instructions, or time to complete the testing on that day was not sufficient. Sample sizes vary among age groups because of the number of responses received from parents. The sample is ethnically diverse with 438 (29.60%) white, 358 (24.2%) Asian/Pacific Islanders, 292 (19.7%) Latino/Hispanics, 242 (16.3%) African Americans, 21 (1.40%) Native Americans/Alaskans, and 130 (8.80%) reporting multiple ethnicities/other. Of the 1481 participants, parents reported that 4 had heart conditions, 4 had diabetes, 147 had asthma, 20 had multiple conditions, and 85 had other nonspecific conditions. The rest of the subjects had no reported medical conditions.
TUG Speed by Age and Sex
Each subject performed 2 trials of the TUG. A paired t test between trial 1 and 2 indicates that trial 2 was faster than trial 1 by only 0.27 seconds (t (1480) = 15.689, P ≤ .000). Nicolini-Panisson and Donadio11 established the clinically important difference at approximately 2 seconds, far longer than the difference in this study, so the 2 trials were averaged for the remaining analyses. The average speed for males and females combined within years of age ranged from a low of 6.20 seconds to a high of 7.12 seconds. There is an overall trend toward decreasing times from ages 5 to 9 years for males, and 5 to 10 years for females, and then trending upwards (increasing times) again (Table 2). A Pearson correlation of age with the average TUG time for the entire sample is weak but significant (r (1479) = −0.133, P < .000, 2-tailed), suggesting that as age increases, TUG time decreases slightly (Figure 1A). A regression analysis for the best-fit curve yielded a significant model of age as a predictor of TUG time (TUG = 10.47-0.905(age) = 0.49(age)2, t (1478) = −9.495, P < .000). This curve (Figure 1B) describes the decrease in time from ages 5 through 9 years, and the increase that occurs from ages 9 through 13 years.
Fig. 1: Graphs of average Timed Up and Go (TUG) times by age. (A) Average TUG time by age, best-fit line. (B) Average TUG time by age, best-fit curve.
TABLE 2: TUG Averages and t Test Significance for Sex Within Age (n = 1481)
Table 2 presents the average times by age in years for males and females. Although 8-, 9-, and 11-year-old males are about 0.5 seconds faster than females of the same age, this 0.5 seconds does not exceed the 2-second clinically important difference.11 TUG times across the age groups ranged from 5.85 ± 0.88 to 7.24 ± 0.97 seconds for males and 6.34 ± 0.94 to 7.17 ± 1.12 seconds for females. An ANOVA revealed that age and sex are each significant factors accounting for differences in TUG speed; however, there was no interaction effect. Post hoc analyses illustrated that age differences cluster with the 5- to 7-year-old groups differing from 8- to 11-year-old groups, and they are different from 12- to 13-year-old groups (Table 3).
TABLE 3: ANOVA and Post Hoc Tests of Between-Subjects Effects
TUG Speed and BMI
On the basis of the CDC BMI definitions, the sample was classified as 3.87% underweight, 63.83% healthy, 16.38% overweight, and 15.90% obese (Table 4). The average TUG speed was only weakly, but positively correlated with BMI for the entire sample (r = 0.065, n = 1471, P < .012, 2-tailed). Within each year of age, BMI is positively correlated with TUG speed only for 8-, 9-, 11-, and 12-year-old groups (Table 4). No significant differences were found in time to complete the TUG on the basis of BMI categories within ages, except for 11-year-old children, but that did not hold up under post hoc analysis (Table 5).
TABLE 4: Correlation of BMI and TUG
TABLE 5: ANOVA of TUG Speed Among BMI Categories Within Age (n = 1471), BMI Category
TUG Speed and Medical Conditions
Of the 1481 participants, asthma was the most frequently identified medical condition (n = 147 or 9%). A secondary analysis using independent t tests compared this subset with those who had no reported medical conditions; no significant differences were found within age groups. Other medical conditions were not tested because the samples were very small or, in the case of those with multiple conditions, different combinations of conditions were reported.
DISCUSSION
The purpose of this study was to expand the reference values of the TUG for a diverse urban sample of children with typical development and to determine whether differences existed on the basis of sex or BMI. By testing 1481 subjects using the same instructions and testing procedures, the data from this study reflect age-based speeds for an ethnically diverse sample of 5- to 13-year-old urban school children walking at a natural pace.
The average TUG time for all subjects in this study was 6.60 ± 1.1 seconds, which is slower and more variable than previously reported times of 3.78 ± 0.610 to 5.61 ± 1.0611 seconds (Table 1). TUG times decreased from ages 5 through 9 years, and then began to increase from age 9 through 13 years; this uptick in times was unexpected. Differences between 8-, 9-, and 11-year-old males and females were statistically, but not clinically meaningful. The differences between the average time in this study and prior literature, and the unexpected uptick in TUG times after age 9 may be attributed to several factors including differences in procedures, a higher incidence of physical inactivity among urban school children,19 and the potential influence of BMI.
In 3 prior studies, students were instructed to “go as fast as possible,”8,10,11 whereas in this study students were instructed to “walk, don't run,” to mimic students moving within the classroom. “Walk, don't run” can be interpreted by the participants as not walking fast, which may account for the longer average time to complete this test and the greater variability as reflected by a larger standard deviation (Table 1). When walking speed slows, variability of gait parameters increases.12 Having students consciously slow themselves may thus contribute to wider variations in walking speed than might be seen when they “go as fast as possible.”
To increase ease of use in a variety of environments, the investigators chose the line on the floor as a target, similar to Podsiadlo and Richardson,5 rather than a star on the wall as used by Williams et al.9 The difficulty in following instructions without using a wall target, as suggested by Williams et al,9 was not explicitly evident in this study. Of the 5- to 8-year-old groups, 90% were able to complete the task on the first attempt for trial 1, and 96% on the first attempt for trial 2. Of the 9- to 13-year-old groups, 92.5% completed the task on the first attempt for trial 1, and 96.4% on the first attempt for trial 2. Thus, for children with typical development, ages 5 to 13 years, a target may not be necessary, although whether the lack of the target affected the TUG speeds is not clear.
Children living in urban areas have been shown to be less active than their rural counterparts19 and may have less access to exercise opportunities.20,21 In addition, evidence exists that between the years of 9 and 15 years, the amount of physical activity successively decreases from 3 hours per day to 1 hour per weekday and 35 minutes per weekend day.22 The participants in this study live in a large metropolitan community comprised of many urban districts where access to playgrounds, recreation centers, and organized sports can be limited by location, safety, and economics.
Over one third of American children ages 2 to 19 years are considered overweight or obese.23 Children display altered functional movement as a consequence of excess weight.24 In this study, beginning at age 9 years, the percentage of children who are overweight or obese ranges by ages from 35% to 56% (Table 5). This is a higher percentage of children in the overweight and obese BMI categories than is reported for the US population.23
Secondary analysis of the most prevalent health condition (asthma) did not seem to affect TUG performance within age groups. Thus, the instruction to walk at a normal pace, less physical activity associated with life in urban settings, trends toward less physical activity in general after age 9 years and a higher incidence of children in this sample that were overweight or obese, may have interacted to account for the increased overall average time, larger standard deviations and the uptick in TUG times described for this sample.
Utility of the Test
The TUG is an inexpensive and expedient test that can be used to measure functional mobility and the transition between sitting and standing in the classroom. Physical therapists can use the TUG, in conjunction with classroom observations, to assist in decision making when evaluating students for physical therapy or when assessing progress throughout the year. By testing with shoes on, using shorter instructions, and removing the need for a target on the wall, the procedures for using the TUG have been simplified, thus further decreasing the amount of time needed to set up and conduct the test. Moreover, PTs may also choose to test students with just a single trial because the difference between 2 trials was found to be clinically insignificant. This is consistent with prior comparisons of trials within the same day, as well as when separated by 1 week.11 The instructions of “walk, don't run” more accurately reflect typical daily movement in the classroom and other indoor environments. Within age groups in this study, a single standard deviation is approximately 1 second. Students who are 2 standard deviations from the mean would be approximately 2 seconds slower than their peers, and that may be a clinically important difference in some settings,11 but would need to be evaluated for its relevance on the basis of the age of the child and the social demands and expectations in the school environment. Therefore, the results of the TUG for an individual student should not be used in isolation to determine the need for or focus of interventions, but could be a part of a comprehensive assessment that includes observation in the natural setting and other standardized tests and measures.
Difficulties replicating this test may include finding the right size chair so that subjects' hips and knees are at about 90° angles, and ensuring that students follow instructions so that both feet pass the second the line. It should be noted that students performed the test in an open testing area, which does not simulate the obstacles in classroom settings, or natural tasks, such as carrying objects when moving between class locations.
It is suggested that future studies address children with typical development in both younger (3- to 4-years-old) and older (14- to 18-years-old) age groups to establish reference standards while walking at a normal pace, as well as establishing reference standards for individuals with developmental delays.
CONCLUSIONS
The TUG is a functional assessment that parallels a student's transition of getting up from a seat and moving around a classroom. It is a reliable test that requires minimal equipment, time, and space. This study establishes the reference ranges for an ethnically diverse cohort of 1481 children with typical development 5 to 13 years old, from a large, urban school district. The times are slightly longer than reported for prior studies and an unexpected progressive increase in times was found from age 9 to 13 years. These differences are attributable to the instructions to “Walk, don't run,” documented trends toward inactivity in urban youth, and a higher percentages of the sample categorized as overweight or obese, beginning at age 9 years. No significant differences were found between sexes and the broader BMI categories within age groups, and positive correlations between BMI scores and TUG speeds were found for selected ages. The ranges of TUG times may be helpful for comparing individual children to age-matched norms. School-based physical therapists can use the TUG as part of their assessment to determine whether students are functioning slower than age-matched peers when moving between their seats and other classroom locations.
ACKNOWLEDGMENTS
The authors thank the PTs who assisted us with organizing the children at the testing sites: Liann Arnold-Liebman, PT, DPT, Caren Goldberg, PT MS, Deborah Salwen, PT, Sujeeta Sippy, PT, DPT, Hea Jung Fico, PT, DPT, and Michelle Frohlich, PT DPT, the schools that participated, including their administrators and PTs, and the children who volunteered for testing.
REFERENCES
1. Geffers A. Increasing time-on-task behavior through the implementation of classroom transition strategies. Vol MS in Education. Oshkosh, Wisconsin: University of Wisconsin Oshkosh; 2007:1–40.
2. Carta J, Renauer M, Schiefelbusch J, Terry B. Effective Instructional Strategies to Facilitate: Classroom Transitions, Group Instruction, Independent Performance and self-assessment. 4th ed. Teacher's Manual for Project SLIDE: Skills for Learning Independence in Developmentally Appropriate Environments. The University of Kansas; 1998:114.
3. US Department of Education. Building the Legacy: IDEA 2004, Sec. 300.34 Related services.
http://www.ed.gov/explore/view/p/%2Croot%2Cregs%2C300%2CA%2C300%2E34%2C. Accessed December 15, 2014.
4. McEwen IR. IDEA: Providing Physical Therapy Services Under Parts B & C of the Individuals With Disabilities Education Act. Alexandria, VA: Section on Pediatrics, American Physical Therapy Association; 2009.
5. Podsiadlo D, Richardson S. The timed “Up & Go”: A test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39:142–148.
6. Shumway-Cook A, Brauer S. Predicting the probability for falls in community-dwelling older adults using the Timed Up & Go test. Phys Ther. 2000;80(9):896–903.
7. Eekhof JAH, DeBock GH, Schaapveld K, Springer MP. Functional mobility assessment at home: Timed Up and Go test using three different chairs. Can Fam Physician. 2001;47:1205–1207.
8. Habib Z, Westcott S, Valvano J. Assessment of balance abilities in Pakistani children: a cultural perspective. Pediatr Phys Ther. 1999;11(2):73–82.
9. Williams EN, Carroll SG, Reddihough DS, Phillips BA, Galea MP. Investigation of the timed “Up & Go” test in children. Dev Med
Child Neurol. 2005;47(8):518–524.
10. Marchese VG, Oriel KN, Fry JA, et al. Development of reference values for the Functional Mobility Assessment. Pediatr Phys Ther. 2012;24(3):224–230.
11. Nicolini-Panisson RD, Donadio MVF. Normative values for the Timed “Up and Go” test in children and adolescents and validation for individuals with Down syndrome. Dev Med
Child Neurol. 2014;56(5):490–497.
12. Beauchet O, Annweiler C, Lecordroch Y, et al. Walking speed-related changes in stride time variability: effects of decreased speed. J Neuroeng Rehabil. 2009;6(32):1–6.
13. Bergmann JH, Alexiou C, Smith ICH. Procedural differences directly affect Timed Up and Go times. J Am Geriatr Soc. 2009;57(11):2168–2169.
14. Klepper SE, Muir N. Reference values on the 6-Minute Walk Test for children living in the United States. Pediatr Phys Ther. 2011;23:32–40.
15. Center for Disease Control and Prevention. Healthy Weight: Assessing Your Weight: BMI: About BMI for Children and Teens | DNPAO | CDC.pdf. July 11, 2014.
http://www.cdc.gov/healthyweight/assessing/bmi/childrens_bmi/about_childrens_bmi.html, 2014.
16. Thorpe LE, List DG, Marx T, May L, Helgerson SD, Frieden TR. Childhood obesity in New York City elementary school students. Am J Public Health. 2004;94(9):1496–1500.
17. Hallgren KA. Computing inter-rater reliability for observational data: an overview and tutorial. Tutor Quant Methods Psychol. 2012;8(1):23–34.
18. Center for Disease Control and Prevention. National Health and Nutrition Examination Survey (NHANES): Anthropometry Procedures Manual. In: Statistics NCfH, ed. Hyattsville, MD; 2009.
19. Liu J, Bennett KJ, Harun N, Probst JC. Urban-rural differences in overweight status and physical inactivity among US children aged 10-17 years. J Rural Health. 2008;24(4):407–415.
20. Kottyan G, Kottyan L, Edwards NM, Unaka NI. Assessment of active play, inactivity and perceived barriers in an inner city neighborhood. J Community Health. 2014;39(3):538–544.
21. Joens-Matre RR, Welk GJ, Calabro MA, Russell DW, Nicklay E, Hensley LD. Rural-urban differences in physical activity, physical fitness, and overweight prevalence of children. J Rural Health. 2008;24(1):49–54.
22. Nader PR, Bradley RH, Houts RM, McRitchie SL, O'Brien M. Moderate-to-vigorous physical activity from ages 9 to 15 Years. JAMA. 2008;300(3):295–305.
23. Fryar CD, Carroll MD, Ogden CL. Prevalence of Overweight and Obesity Among Children and Adolescents: United States, 1963-1965 Through 2011-2012. Centers for Disease Control and Prevention, National Center for Health Statistics; 2014.
24. Duncan MJ, Stanley M. Functional movement is negatively associated with weight status and positively associated with physical activity in British primary school children. J Obes. 2012;2012:5.
