Abstract
Objective:
To quantify the effect of handhold size (diameter) on the maximum breakaway strength between a hand and handhold for children.
Background:
Falls from playground equipment are a major cause of childhood injury and death. It is unclear if recommendations for handholds on playground equipment are too broad.
Methods:
Breakaway strength was defined as the maximum quasistatic force that can be exerted on a grasped object before the object is forcibly pulled from the grasp. Hand anthropometry, grip, and breakaway strengths were measured for 397 children between the ages of 6 and 11 years. Three cylindrical handhold diameters were tested.
Results:
Breakaway strength was significantly affected by handhold size, gender, and hand dominance. Significant covariate predictors for breakaway strength included grip strength, age, and hand breadth. Breakaway strength was reduced for the largest diameter (3.81 cm) for children of all ages.
Conclusion:
Handhold design factors significantly affect the breakaway strength of children.
Application:
The results can be used as a basis for design recommendations for hand rungs used by children to reliably support their bodyweight.
Introduction
Falls are the third leading cause of death, the number one reason for trips to the emergency room, and the leading cause of nonfatal unintentional injuries among hospitalized children age 1 to 19 years in the United States (Agency for Healthcare Research and Quality, 2011; Centers for Disease Control and Prevention, 2005; National Center for Injury Prevention and Control, 2002, 2005). It is estimated that 63,000 of those injured will require hospitalization and that approximately 261 will die of their injuries. Falls are also responsible for more open wounds, fractures, and brain injuries than any other cause of injury. Between 2004 and 2007, 1,015 children under 18 years were hospitalized due to a fall at the University of Michigan CS Mott Children’s Hospital. Of these, 242 (24%) suffered a traumatic brain injury and 678 (67%) had a fracture. A fall from playground equipment was the leading mechanism accounting for over 50% of these hospitalizations in school-age children.
Each year in the United States, playground-related injuries are responsible for 15 deaths and more than 200,000 emergency department visits for children ages 14 and younger (National Recreation and Park Association, 2005; Tinsworth & McDonald, 2001). Of these injuries, 45% are severe—fractures, internal injuries, concussions, dislocations, and amputations. A recent study that analyzed 2,691 playground-related injuries and deaths from 2001 to 2008 reported specifically to the Consumer Product Safety Commission (CPSC) indicated that falls accounted for 44% of all incidents and that 20% of all incidents were associated with climbers, monkey bars, jungle gyms, and bars (O’Brein, 2009). Though the majority of falls were due to unspecified causes, 2.4% are attributed specifically to loss of hand grip. An earlier study of data from the CPSC’s National Electronic Injury Surveillance System (NEISS) reported that “the most frequently reported cause of falls, accounting for 40 percent of all fall related injuries on public equipment, was the child losing his or her grip (primarily on climbing bars or swing chains)” (Tinsworth & McDonald, 2001, p. 1). These observations suggest that the design of handholds, rungs, rails, or grab-bars on playground structures may be important determinants in preventing falls and injury.
Public playground safety standards are set by each state; however, most endorse the voluntary standards in the CPSC book: Public Playground Safety Voluntary Standards, now in its eighth edition (Consumer Product Safety Commission, 2011). According to those standards, the one physical capacity on which handhold design recommendations are based is grip strength. However, recent studies of adults have shown that grip strength is not a reliable predictor of the capacity to hold onto rungs or rails and support the body with the hands (Young, Woolley, Armstrong, & Ashton-Miller, 2009, 2012). This is because the amount of force that can be exerted on a grasped object before it slips free or is pulled from the grasp of the hand (“breakaway strength”) depends not only on a person’s grip strength, but also on many other factors, including the size and shape of the handhold that is grasped, the orientation of that handhold (horizontal, inclined, or vertical), and the coefficient of friction between the fingers and the handhold. These studies indicate that the basis for the design of handholds on playground equipment may be systematically flawed. Furthermore, anthropometric factors such as hand size and bodyweight, which vary greatly in school-age children as they grow, may influence design recommendations for specific age groups.
The overall aim of this study was to quantify children’s capacity to hang onto a handhold and to determine handhold and biomechanical factors that influence this capacity. To address this aim an experiment was designed to measure breakaway strength for school-age children on cylindrical handholds of different sizes. The experiment tested the hypotheses that there exists a handle diameter that significantly increases the capacity for children to hang on and that hand grip strength does not accurately predict this capacity.
Methods
Participants
Study participants consisted of students from two elementary schools. Prior to participation, research staff visited all classrooms in each elementary school and described the study to the students. A note describing the study including pictures of the equipment to be used and a parental consent form was sent home to the students’ parents. Students whose parents provided consent were eligible to participate in the study. Before a student participated in the study, research staff once again described the study to the student and the students provided assent or consent to participate (verbal assent for students in kindergarten through Grade 3 and written consent for students in Grades 4 to 6). Students saw the study equipment/apparatus in person before providing assent or consent and could withdraw at any point during data collection. After participation, each student was free to select a small gift (total value approximately $1). After the study was completed the school principal was given a $20 educational gift card per student who participated to be divided equally between the classes. The University of Michigan Institutional Review Board approved the study recruitment procedures and protocol.
A total of 397 students from two public elementary schools (212 male and 185 female) ranging in age from 5 to 11 years participated in the study. One child completed all aspects of the study except the breakaway force measurements (Station 3). She became overwhelmed and did not want to complete this part of the experiment. Table 1 contains demographic information on study participants. Of the participants, 91% were right-handed and 6% were obese (World Health Organization, 2011). Mean (±SD) height, weight, and body mass index for the 397 participants were 1.28 ± 0.11 m, 28.0 ± 8.12 kg, and 16.31 ± 2.59 kg/m2. The range (minimum to maximum) for subject height, weight, and body mass were 1.06 to 1.59 m, 16.7 to 84.8 kg, and 13 to 33 kg/m2.
Subject Demographics (count of participants)
Breakaway Force Measurement Apparatus
Two custom apparatus, adapted from previously approved apparatus for testing breakaway strength in adults (Young et al., 2009, 2012), were built to simulate children hanging onto a playground monkey bar (Figure 1). An instrumented handle unit was suspended above participants by a cable-pulley system fixed to a rigid, aluminum structure. The handle unit was mounted on a vertical cable that was mechanically raised and lowered (2.54 cm per second) using a linear actuator controlled by the experimenter. The handle unit also contained a video camera that recorded the posture of the hand during each trial. The United States Consumer Product Safety Commission playground handbook recommends that rungs or handrails for children between the ages of 5 and 12 years have diameters ranging from 2.4 cm to 3.9 cm. We therefore used three different cylindrical handle diameters: small, medium, and large having diameters of 2.5 cm, 3.2 cm, and 3.81 cm, respectively. The handles were made of an aluminum alloy with a smooth surface and were easily interchangeable. The floor of the test apparatus was padded and had two attachment points for straps that were used to secure the standing participant via a padded climbing belt (Figure 1a and 1b). Securing the participant ensured that as the grasped handle was raised the participants’ maximum voluntary strength would be attained without them being lifted off the floor (Figure 1c and 1d). Subjects applied force to resist movement of the handle upward during breakaway trials until the handle was pulled free of the grasp of the hand or the subject let go. This resulted in the fingers opening and a lengthening contraction of the finger flexor muscles (Figure 2).
Breakaway strength apparatus. (a) The device consists of a sturdy metal structure supporting a handle that can be raised and lowered by a linear actuator and pulleys. (b) The subject is secured to the floor of the device. (c) Subject at the beginning of a trial. (d) Subject near the end of a trial (hand starting to come off the handle).
Sequence of hand positions from the start of a trial (at left) to the time when the hand beings to slip off of the handle (at right).
Resultant vertical force exerted by the participant as they held onto the handle was measured using a single-axis load cell (Interface® SM-50), an amplifier, and a 12-bit data acquisition interface (National Instruments USB-6008). LabView™ software was used to record the forces that were exerted at 100 Hz as well as the video at a frame-rate of 15 Hz. The peak force measured at any time during the entire breakaway sequence was recorded and defined as the breakaway strength for that trial.
Study Protocol
This study was conducted in two elementary schools over the course of several regular school days between March and May of 2010. The experiment was set up in three consecutive stations in the same area. Students were paired with a classmate at each station. Study team members escorted participants to and from the data collection room. Before entering the data collection area, participants washed and dried their hands.
Station 1
Demographic information (age, grade, and ethnicity), anthropometry (height and weight), and grip strength data were collected at Station 1. The dominant hand was determined by asking participants to write their name or draw a circle and observing which hand the child used to write with. Grip strength was measured three times for each hand using a Jamar-type grip dynamometer in position 2 at 49 mm (Molenaar et al., 2008). Before starting each trial, the dynamometer was sanitized and participants were provided a piece of paper towel and asked to dry their hands by rubbing the paper towel between their hands. Subjects were asked to raise their hand above their head with their arm fully extended. Study team members then positioned the dynamometer in the subjects’ hand so that grip strength measurements would be recorded in the same posture as breakaway strength measurements. Participants were then asked to “squeeze the dynamometer as hard as they can and to keep squeezing” until they were asked to stop. Two minutes of rest were provided between each grip strength trial.
Station 2: Left and right hand anthropometry (length and breadth) measurements
Hand length was measured from the tip of the middle finger to the wrist crease and the hand breadth was measured as the maximum breadth across the knuckles using calipers (Garrett, 1971). In addition, digital pictures of the palmar and dorsal side of each hand were taken by a digital camera mounted above a table. The hand was positioned against a grid located on the table top before the pictures were taken.
Station 3: Breakaway strength measurement
This test was described to the subjects in the following manner, “At this station we want to know how long you can hold onto this handle. The handle will move up when we start it. We want you to wear a belt to hold you down.” Subjects were fitted with a mountain climbing belt by an experimenter and were secured to the floor of the test device. The vertical handle location was positioned so that subjects could comfortably reach it with their arm mostly extended. Subjects were randomly assigned to test one of the three handle sizes and to randomly start with either their right or left hand. Next, subjects were told, “Now, I want you to reach up and hold the handle as long as you can. There is a camera that will take a video of your hand. Are you ready? Okay, reach up and hold the handle!” Subjects were also instructed to “Keep holding. . . . Hold as long as you can.” Once the trial started, the handle arm was mechanically raised until the subject let go or the hand slipped off the handle (Figure 2). Each subject completed six total breakaway strength trials (three repetitions of only one handle size on each hand). Two minutes of rest was provided between each trial. Prior to starting each trial, subjects dried their hands by rubbing a paper towel between their hands. After completing all the trials, an experimenter removed the belt from the subject and the subject selected a small gift.
Design and Statistical Analysis
The study investigates the effects of different sizes of handle size (small, medium, and large; diameter of 2.54 cm, 3.175 cm, and 3.81 cm, respectively) on several dependent variables, including breakaway strength, grip strength, breakaway/grip, and breakaway/bodyweight. Due to time constraints, it was not possible to have each participant perform breakaway strength trials for all three handle sizes on both hands. Instead, each participant was assigned a single handle size for testing when they arrived at Station 1. The handle size was assigned randomly and balanced among participants (Table 2). Dependent variables were measured for each child both for dominant and nondominant hand.
Count and Anthropometry of Participants Randomly Assigned to Each Handle Size
One subject withdrew from the study before breakaway strength was measured at Station 3.
Measured for the dominant hand.
The grip strength was determined from the grip dynamometer trials and breakaway strength from the breakaway apparatus trials. The maximum peak force from the three repetitions of the measurements was used as the outcome metric for each dependent variable. Two additional outcome measures were calculated: breakaway strength normalized by bodyweight (“breakaway/bodyweight”) and breakaway strength normalized by grip strength (“breakaway/grip”).
The analysis was performed separately for each dependent variable. To accommodate repeated measures, namely, for dominant and nondominant hand, we employed linear mixed effects models (Laird & Ware, 1982). When developing final models, we considered several covariates. The handle size (small, medium, large) as the main covariate of interest was used in all models. To adjust its effect for other covariates, several other variables were considered, such as within-subject fixed effect of hand (dominant vs. nondominant), several subject-specific covariates (gender, age, height, weight, hand length, hand breath, and grip strength), and interaction terms (age by gender, age by handle size, handle size by hand, handle size by hand length, handle size by hand breadth, and handle size by grip strength). To take into account the between-subject variation in the model, we associated random intercept with each subject. The interaction terms turned out to be nonsignificant, therefore they were omitted from the final models. Post hoc Bonferroni correction was performed for significant main effects. Analyses were performed using the linear mixed models module in PASW software (version 18.0, SPSS Inc.).
Results
Initial analysis revealed that no interaction terms were significant (p > .05), so they were removed from the final statistical model, which is presented in Table 3. Breakaway strength was significantly affected by handle size (p < .001), gender (p = .006), and hand (p = .007). Significant covariates for breakaway strength were grip strength (p < .001), age (p = .01), and hand breadth (p = .023). Breakaway strength and grip strength increased with age, as did many anthropometric variables (i.e., height, weight, hand length, hand breadth). The Pearson correlation between breakaway strength and covariates was 0.703, 0.634, and 0.590 for grip strength, hand breadth, and age, respectively.
Statistical RMANOVA Results for Fixed Effects on Breakaway Strength
Note: Num DF = numerator’s degrees of freedom; Den DF = denominator’s degrees of freedom. Bolded numbers represent significant interactions.
Post hoc tests showed that breakaway strength was significantly reduced for the large handle size compared to the medium and small sizes (p < .001), which were not significantly different (p = .269). Breakaway strength for the small and medium handle size were 42 N and 27 N greater, on average, than for the large handle size, respectively (Table 4). Breakaway strength was significantly greater for the dominant hand than for the nondominant hand (p = .007); though the average difference between hands was generally small, at 7 N, 9 N, and 7 N for the small, medium, and large handle size, respectively (Table 4). Gender had a significant effect on breakaway strength, with males being stronger than females (p = .006). The gender effect was particularly noticeable for the large handle size, where breakaway strength was 44 N greater for males than females, on average (Table 4).
Mean (SD) Results for Dependent Variables By Handle Size, Gender, and Hand
Mean breakaway strength for the large handle tended to be lower than the medium and small handle for all ages (Figure 3), although the Handle × Age interaction did not reach significance. Grip strength exhibited similar trends as breakaway strength, with the male gender (F = 21.01, p < .001) and the dominant hand (F = 63.20, p < .001) exhibiting greater strength. Unlike breakaway strength, grip strength was not different between handle size groups (F = 1.04, p = .353), which was expected due to randomization of subjects. Both breakaway and grip strength increased linearly with age (Figure 4).
Breakaway strength for each handle size and grip strength versus age. Gender and hands are pooled. Error bars indicate one standard deviation.
Breakaway and grip strength by age and gender. Handle sizes and hands are pooled. Error bars indicate one standard deviation.
Mean normalized breakaway force is also presented in Table 4. Normalizing breakaway strength by bodyweight provides a metric that may help predict if a child can support their own bodyweight while hanging onto a playground handhold with one hand. The breakaway/bodyweight ratio was affected by gender (F = 14.47, p < .001), hand (F = 14.41, p < .001), and handle size (F = 12.22, p < .001). Breakaway/bodyweight was, on average, 3% greater for the dominant hand than the nondominant hand, 10% greater for males than females, and 12% to 15% smaller for the large size handle than the medium or small sizes (p < .001), which were not significantly different (p = .958). Breakaway/bodyweight and grip/bodyweight ratios were generally constant with age (Figure 5).
Breakaway and grip strength normalized by bodyweight versus age by gender. Handle sizes and hands are pooled. Error bars indicate one standard deviation. A value greater than 1 (plotted as a line) indicates that applied force was greater than subject’s bodyweight.
The breakaway/grip ratio was affected by handle size (F = 8.35, p < .001) and hand (F = 8.73, p = .003), but not gender (F = 1.05, p < .306). Breakaway/grip was, on average, 6% greater for the nondominant hand than the dominant hand and was 24% to 15% smaller for the large size handle than the small or medium sizes (p < .034), which were not significantly different (p = .401). Overall, breakaway strength was much greater than grip strength for all handle sizes (191% to 232%, on average).
Discussion
The aim of this research was to quantify children’s capacity to exert force on and hang on to handholds of different diameters. This study provides the first and only breakaway strength data for children and shows that handhold and individual factors influence this breakaway strength. Furthermore, standard measures of hand strength (grip strength) significantly underestimate the capacity for children to apply a hand force to a grasped handhold.
Breakaway strength increased significantly with individual factors of grip strength and its covariates (age, gender, and hand breadth). It is not surprising that as children get older, their grip strength increases, their hands get bigger, and their ability to hold on to a handhold improves. Grip strength has been quantified for children in many studies (Ferreira et al., 2011; Hägar-Ross & Rösblad, 2002; Molenaar et al., 2010; Rauch et al. 2002; Wind, Takken, Helders, & Engelbert, 2010), which also show that grip strength increases with age, hand size (length and breadth), male gender, dominant hand, and other anthropometric factors such as height and weight. Grip strength for subjects in this study was slightly higher than reported previously, which may be due to arm posture during measurement (overhead in this study) or reporting the maximum of three trials rather than an average.
In addition to individual strength and covariates, breakaway strength was significantly affected by the diameter of the handhold being grasped, with larger handholds (3.81 cm) affording less capacity than medium (3.2 cm) or small diameters (2.5 cm).
The similarity in results for handle size and adult (Young et al., 2012) and child breakaway strength is interesting because of the large difference in grip strength and hand size between children and adults. Since the size of the hand determines the posture of the hand on the handhold, we would expect that optimal handhold diameter would be smaller for younger children than older children or adults. However, there was no significant interaction between age and handhold size. Furthermore, the two smaller handle diameters were similar in this study and Young et al. (2012). A possible explanation for both studies showing that 2.2 cm to 2.5 cm handholds are optimal is that the optimal handle size for younger children (or those with smaller hands) is smaller than the 2.5 cm diameter tested. While it is apparent that larger handhold sizes reduce the capacity to hang on for all ages, it is unclear if age influences the optimal handhold diameter for horizontal cylinder-type handholds.
Gender and hand dominance affect breakaway strength similarly in children and adults, with males and dominant hands being stronger than females and nondominant hands (Rajulu & Klute, 1993; Young et al., 2009, 2012). However, the sex difference in breakaway strength between male and female children is much smaller than for young adults (Young et al., 2012): Breakaway strength on 3.2 cm diameter handholds was 92 N and 388 N less for 11-year-old children than young adults for females and males, respectively. This difference is likely due to disproportionate strength changes during puberty and adolescence between genders (Ferriera et al., 2011; Hägar-Ross & Rösblad, 2002; Rauch et al., 2002; Wind et al., 2010).
The three cylindrical handhold diameters tested in this study cover the range of national CPSC recommendations for playground rung/rail size. These broad standards suggest diameters or maximum cross-section between 0.95” (2.4 cm) to 1.55” (3.9 cm) for children ages 2 to 15 years. These results clearly show that breakaway strength on the 3.81 cm handle was decreased for all ages and suggest that the recommendation should be revised.
While both weight and breakaway strength increase with age, the ratio of breakaway strength to bodyweight provides an estimate of the ability for the child to support their own bodyweight with one hand in a potential fall situation. Values greater than 1.00 suggest that a child can, at least briefly, hang on to playground equipment that is similar to the tested handle diameter. Overall, the average values of this ratio suggest that roughly half the females could not support their bodyweight on the largest handle size (Table 4). When stratified by age (Figure 5), average 5-year-olds of both sexes and females of age 8 and 11 years could not support their own bodyweight with one hand. Therefore, approximately half the very young children and half the females may be at greater risk of playground falls due to loss of hand/handhold coupling. Breakaway/bodyweight ratio is similar for children and adults in males, but greater for female children than female adults (Young et al., 2012). The breakaway/grip strength ratio results can be interpreted as estimates of a “safety factor” formed by considering the ratio of demand over capacity.
While this study provides the first exploration of children’s capacity to hang on, this study has several limitations. First, this was primarily a static assessment of breakaway strength in that the child’s whole body was not moving while holding on, as they would be when swinging on bars or rings, which could definitely affect breakaway strength. Hence, our estimates of breakaway strength may be conservative because, while the finger flexor muscles would have had to act in the lengthening region of their force-velocity relationships, it is possible that they might develop a little more force at the more rapid lengthening velocities associated with a fall arrest, at least until the contractile force saturates. Second, the goal of the breakaway strength measurement proved to be confusing for some children (particularly the youngest ages). These children prematurely let go of the handle as it was raised and did not exert their full capacity to hang on. In many cases, this was noticed immediately by study personnel at the breakaway station (who benefited from real-time feedback of applied forces), and they were able to reiterate and explain the goal of the measurement to the child for subsequent repetitions. Due to this complication, the maximum reading for the three trials was chosen rather than an average of all three repetitions. Third, children were generally very excited to participate, motivation to hang on was not the same as if a child were about to fall, and some were shy or timid and may not have exerted their maximum voluntary capacity. Fourth, although obesity was of interest and of particular importance in the context of climbing, it was not able to be examined because of a paucity of obese children in this sample. Fifth, the study would have been strengthened had the effect of the three grip diameters been examined in every child, but we decided to examine one diameter per child so as to minimize bias due to fatigue. Sixth, friction between the hand and handhold might affect the results (e.g., Young et al., 2009), so the difference between the test handholds being made of aluminum and playground handholds being typically manufactured from galvanized or painted steel should be considered. However, unless the paint has particles in it to purposely increase friction, there should not be significant differences in practical terms.
This study only examined a single handhold factor: diameter. While this factor clearly impacted breakaway strength, other handhold factors, such as orientation and available surface friction, have been shown to have a large influence on hand/handhold coupling in adults (e.g., Young et al., 2009, 2012). Playground equipment may be particularly influenced by these factors since playgrounds are often outdoors and exposed to contaminants (dirt, mud) and environmental factors (rain, snow, ice).
Conclusions
The aim of this study was to quantify the capacity for children to hang on to handholds and determine handhold and biomechanical properties that affect this capacity. An apparatus was therefore built to measure the breakaway force developed while cylindrical handrungs of three different diameters were slowly pulled from each child’s one-handed grasp. The results show that breakaway strength was influenced by grip strength, age, gender, hand dominance, and hand breadth and it was significantly reduced for large (3.81 cm) handhold diameters. While grip strength was positively correlated with breakaway strength, it significantly underpredicted breakaway force and does not account for the handhold design factors that influence breakaway strength. Further studies are needed to improve understanding of predictors for climbing-related falls and design factors that reduce fall risk.
Key Points
Falls from playground equipment are a major cause of childhood injury.
This study provides the first and only breakaway strength data for children and shows that handhold and individual factors influence this breakaway strength.
Grip strength as measured by a grip dynamometer is related to, but may significantly underestimate, breakaway strength in children.
Children’s breakaway strength is reduced for larger diameter (3.81 cm) handholds for all ages between 6 and 11 years.
Consumer Product Safety Commission recommendations for playground rung/rail size in children age 2 to 15 years (between 0.95” [2.4 cm] to 1.55” [3.9 cm] for children ages 2–15 years) appear to be too broad.
Footnotes
Acknowledgements
This study was supported by a grant from the University of Michigan Sports Injury Prevention Center.
Peter F. Ehrlich is an associate professor of surgery and director of the pediatric trauma and injury program at the University of Michigan. He received an MD and master’s degree from the University of Toronto in 1989 and 1994.
Justin G. Young is an assistant professor in the Department of Industrial and Manufacturing Engineering at Kettering University. He completed his PhD in industrial and operations engineering at the University of Michigan in 2010.
Sheryl Ulin, PhD, CPE, is the director of Continuing Education, University of Michigan Center for Occupational Health and Safety Engineering. She received her PhD in engineering from the University of Michigan.
Charles Woolley is an ergonomics research engineer in the Center for Ergonomics at the University of Michigan, Ann Arbor, where he received his MS in bioengineering in 1980.
Thomas J. Armstrong is a professor in the Department of Industrial and Operations Engineering at the University of Michigan, Ann Arbor, where he received his PhD in ergonomics in 1976.
Andrzej Galecki, MD, PhD, is a professor at the University of Michigan, Institute of Gerontology and Department of Biostatistics. He received his Physician Diploma, Medicine at the Medical Academy of Warsaw in 1981 and his PhD in epidemiology at the Institute of Mother and Child Care in Warsaw (Poland), in 1987.
Shu Chen is a statistical research associate at University of Michigan, Institute of Gerontology and Department of Biostatistics.
James A. Ashton-Miller is a research professor in the Department of Mechanical Engineering at the University of Michigan. He received his PhD in biomechanics from the University of Oslo, Norway, in 1982.