339

Work 32 (2009) 339–350 DOI 10.3233/WOR-2009-0832 IOS Press

Backpack load limit recommendation for middle school students based on physiological and psychophysical measurements Denise H. Bauer∗ and Andris Freivalds The Harold and Inge Marcus Department of Industrial and Manufacturing Engineering, The Pennsylvania State University, University Park, PA, USA

Received 10 January 2008 Accepted 24 May 2008

Abstract. The load of student’s backpacks has raised questions over the safety and health of schoolchildren everywhere. The purpose of this study is to use electromyography (EMG), posture evaluation, heart rate, and ratings of perceived exertion and perceptions of pain to find an acceptable backpack load limit for middle school students. Twenty middle school students aged 11 to 14 (10 female and 10 male) volunteered for the study. The subjects completed two tests, standing stationary and walking on a treadmill, where they carried 5% incremental loads from 0% body mass (BM) to 20% BM. The study indicated that the BorgCR10 ratings and trunk flexion angle for the walking trial indicated a possible load limit of 10% BM due to the non-significant difference between 0 and 10% BM and the significant difference between 10 and 15% BM.

Keywords: Child safety, electromyography (EMG), percent body mass

1. Introduction When one thinks of children and school, one usually also pictures some type of bag to carry books and other school materials. The bag of choice for over 40 million students in the United States is the backpack [26], which allows students to carry various objects in an organized manner [11] with both shoulders. However, a recent increase in schoolchildren reporting back pain [24,38,42,45] has health professionals concerned about the long-term affects of heavy backpacks [25, 31,38]. Research has shown a strong relationship between incorrect backpack use and musculoskeletal injuries [18,26], which has caused a debate over the ex∗ Address

for correspondence: Denise H. Bauer, Ph.D., School of Engineering Design, The Pennsylvania State University, 213 Hammond Bldg, University Park, PA 16802, USA. Tel.: +1 785 766 3034; E-mail: bauer.dh@gmail. com.

act reason for the increase in back pain reports [45]. Many think the load of children’s backpacks [26,31,37, 42] combined with how long they must carry them due to schools removing lockers [45,47] is the reason for the incidents. Still others feel students are simply carrying their backpacks incorrectly and using the proper techniques would eliminate the problem [32]. With the introduction of “No Child Left Behind” in 2002, some believe schools are assigning more homework to meet the demands. This increase is likely due to previous research that indicates schools that regularly assign homework have students who are higher achievers [6,13]; as a result, students must transport heavy textbooks between school and home almost every night. Others relate the problem to all the extra supplies, such as athletic equipment, musical instruments, or mp3 players, a student tries to pack into the bag that may add extra weight [16,31]. Still another issue is the mere lack of knowledge regarding the proper method

1051-9815/09/$17.00  2009 – IOS Press and the authors. All rights reserved

340

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

in which to load and carry a backpack [32]. Whether the problem is thought to be related to the belief that textbooks are too large for children to carry or the backpacks are loaded and carried incorrectly, the consensus is that there is a definite safety issue present [18,40] and a need to be proactive in providing recommendations before severe health issues present themselves. Some groups have tried to address the problem by issuing laws or awareness programs. State governments are taking the issue to Congress and attempting to pass laws that will require textbook printers to decrease the weight of books [32]. However, the textbook publishing groups blame the states for increasing the curriculum standards covered in a school year [32]. The American Occupational Therapy Association (AOTA) has created a National School Backpack Awareness Day [33] while the Congress of Chiropractic State Associations (COCSA) has designated a National Backpack Safety Month [10] to encourage students to use a backpack safely by educating them on both the proper techniques used to load a backpack as well as the risks associated with carrying too much weight. Others are introducing school-based backpack health promotion programs [18] to promote behavior change of the students, parents, and administrators. Although these are all valid attempts to solve the problem, the underlying issue of why students’ backpacks are getting heavier and whether or not this results in an increasing incidence of pain is still unresolved. Previous studies have even shown that the current loads of schoolchildren’s backpacks [7,14,15,17,36,50,51] are higher than the current recommended limits. Determining an acceptable limit for the load a child can safely carry is important to reduce injuries to the back, neck, and shoulders as well as posture problems [11,33]. In fact, a study on the load Italian schoolchildren with an average age of 11.6 years carried reported that the average load of 9.3 kg and maximum load of 11.5 kg were equivalent to an average load of 17.6 kg and maximum load of 22.0 kg carried by an 80 kg adult male [36]. These loads are approaching the 23 kg limit allowed by the National Institute for Occupational Safety and Health (NIOSH) to help prevent back injuries in the workplace [28,49]. A few studies present evidence to support backpack load limits for children, but there are discrepancies in the suggested limit. Chow et al. [9], Hong and Brueggemann [21] and Hong et al. [23] suggest a 10% BM load limit based on their respective studies while the results from a study from Hong and Cheung [22] indicate a load limit of 15% body mass (BM). There is also a disagreement

in the appropriate load limit among national organizations. The American Academy of Pediatrics’ (AAP) limit is 10 to 20% BM [2] while AOTA has suggested a backpack load of 15% BM [3]. However, neither organization gives justification for their recommendation. Other authors that have addressed physiological measurements while using a backpack state facts on children and backpacks, but then use adult subjects in their research [12,34,48]. However, studies on adults cannot be transferred to children because the body of a child is constantly changing as they grow and develop, and thus studies on how children’s bodies react when loaded are needed. A particularly important phase is early adolescence (ages 11 to 14). During this time, there is usually a growth spurt and children enter puberty. This is particularly important when carrying a backpack as adolescents in the peak growth period have been found to be at a greater risk for low back pain [30]. This age is also a key time for spinal growth, which is believed to cause the adolescent spine to be less able to withstand stresses that are normal for the adult spine [19,31] as the early and mid-adolescent spine increases in length and volume without substantially adding mass [30]. In addition, it has been shown that if a child reports back pain they are more likely to experience back pain as an adult [30,31,38,42]. A universal backpack load limit for children would help students, parents, and teachers understand the importance of restricting the load of a child’s backpack to ensure the present and future safety of the child. The purpose of this research was to use electromyography (EMG), heart rate, Borg ratings of perceived exertion (Borg-RPE) and perception of pain (Borg-CR10), and trunk flexion angle to find an acceptable backpack load limit. These experiments will help narrow the acceptable load limit by providing additional analysis to the current literature on backpack loads and children. 2. Method 2.1. Participants Middle school students were recruited from the local area to participate in the study. Twenty healthy students aged 11 to 14 (10 female and 10 male) volunteered for the study, which exceeds the minimum number of subjects required (β = 0.90). Eighty percent of the volunteers regularly participate in athletic activities while the other 20% reported occasional athletic participation. The anthropometric data of the subjects is shown in Table 1.

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

341

Table 1 Anthropometric data of subjects Gender Female

Number of Subjects 10

Male

10

Total

20

Statistic Low High Average Std Dev Low High Average Std Dev Low High Average Std Dev

Age (years) 11 14 12.9 1.1 11 14 12.9 0.8 11 14 12.9 1.0

2.2. Backpacks Two backpacks designed for middle school aged students were used in the study: one with standard comfort features and one with additional comfort features. The first backpack is the number one selling backpack in the world from a well-known, popular backpack manufacturer [46]. This backpack contains standard comfort features of straight-cut, padded shoulder straps and a two-thirds padded back (padded vertically with outer edges unpadded). The second backpack is from a different manufacturer that claims the backpack transfers the load from the shoulders to the lower back, reduces back stress by 80%, lightens the effective load of the pack by 50%, and promotes an upright standing position [1]. The additional comfort features of the backpack include air padded s-shaped shoulder straps that can be adjusted at the top and bottom, an air padded lumbar support, side compression straps, and a padded haul handle. 2.3. Instrumentation EMG is widely used in biomechanical studies to examine muscle activity during a task, such as shoulder muscles during abduction, leg and back muscles during load carrying, shoulder and back muscles in assemblyline work, and hand and arm muscles during hand-tool gripping [8]. Triode surface electrodes were placed over the muscle belly in alignment with the muscle fibers to examine the EMG of the trapezius, latissimus dorsi, and erector spinae muscles. Bipolar electrodes (which are commonly used in adult studies) were not used due to the small size of the subjects and thus lack of space to place three separate electrodes. The skin was cleaned with alcohol and electrolyte gel was used to help reduce interference from the skin and improve

Height (cm) 154.6 176.5 162.8 7.5 146.1 168.3 157.9 6.4 146.1 176.5 160.3 7.4

Mass (kg) 40.0 63.2 53.6 6.5 33.8 60.2 47.8 7.4 33.8 63.2 50.7 7.6

BMI (kg/mˆ2) 16.3 23.1 20.2 1.8 14.6 25.3 19.2 2.9 14.6 25.3 19.7 2.5

Resting Heart Rate (beats/min) 50 73 62.9 10.1 55 70 63.0 5.4 50 73 63.0 8.1

the electrical and mechanical contact of the electrodes. The EMG signal was collected at a sampling rate of 2048 samples/sec and the sensitivity scale was set to 1600 microvolts. A MyoScan-Pro sensor was used and allowed a root mean square computation of the raw data before it was passed to the computer, which converted all negative voltages to positive. Any noise or electrical activity from the heart remaining after the sensor filtered the data was removed from any averages by selecting the range and having the program exclude it from any calculations. Averages of the EMG signal over 10 sec where found for each muscle and then normalized to the EMG averages for 0% BM. Heart rate was a physiological measurement of interest as it is an inexpensive way to estimate energy expenditure [27,41]. A Polar Accurex Plus TM heart rate monitor that consisted of a transmitter and watch was used to monitor and measure the subject’s heart rate. The transmitter was placed around the chest with electrolyte gel on the electrodes to, like for the EMG, improve the electrical and mechanical connection. The signal was transmitted to the watch, which continuously displayed the subject’s heart rate during the testing. The monitor served two purposes: to ensure the subject’s heart rate did not exceed their maximum heart rate (preferably staying within 50 to 75% of the maximum heart rate) and to measure the heart rate at intervals during the study. The maximum heart rate was calculated using the equation (220 – age) that has been shown to be suitable for ages 11 and above [39]. The Borg Rating of Perceived Exertion Scale and Borg-CR10 Scale of Perceived Pain were used as psychophysical measurements of the subject’s assessment of the workload. In 1967, Gunnar Borg first developed (later revised in 1985) the Borg-RPE scale, which is the most popular scale for assessing perceived exertion. The scale, intended for use during dynamic ex-

342

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

ertion, ranges from 6 to 20 and is linearly related to the heart rate that is expected for that exertion level where the heart rate equals the rating times 10. Verbal anchors are provided at different intervals of the scale to assist the subject in assessing their workload. The Borg-CR10 Scale relates pain to work intensity, and, although the correlation is not as strong as the BorgRPE scale, Borg-CR10 ratings have been shown to be correlated to heart rate during dynamic activity. The CR10 scale normally ranges from 0 to 10, but the rater is allowed to rate above 10 to express what they feel is maximal pain to avoid ceiling effects. To relate the Borg-CR10 ratings to heart rate, the rating is first transformed to a corresponding Borg-RPE rating that is in turn multiplied by 10 t get the heart rate [5]. It has been found that children ages eight and above are able to distinguish different intensity levels and thus the Borg rating scales are appropriate for use in the study age although the RPE-heart rate relationship may not be as strong as in adults [20]. Borg-RPE and CR10 scales were provided to the subjects before and during testing. The scales were explained to the subjects before any experiments took place so that they could ask questions if needed. During the testing, the subjects were asked to give a Borg-RPE at 1 min intervals. The subjects were asked to give a Borg-CR10 rating for each of the areas where EMG was being collected as well as an overall rating for the whole body after each trial to examine the pain the subjects perceived after carrying each load. The areas were identified by numbers on a figure of a human body so the subjects could more easily rate the perceived pain. The areas were also referred to as shoulder, middle back, lower back, and whole body as opposed to the muscle names. A forward deviation of the trunk from neutral position can lead to discomfort, pain, or injury in the shoulders and lower back [21]. Therefore, trunk flexion angle was measured through lateral view digital photographs to analyze the trunk posture under increasing loads. Markers were placed on the subject to mark the head, shoulder, elbow, wrist, hip, knee, and ankle to aid in identification for measurement. A horizontal line was drawn through the hip, parallel to the treadmill; a vertical line was drawn through the hip, perpendicular to the treadmill. The trunk flexion angle was measured using a protractor from the vertical line through the hip representing neutral trunk posture.

the purpose and procedure of the study. The subject’s weight, height, and age were then recorded, after which they were asked to change into a shirt containing three flaps cut into the shirt (one at each shoulder and one that spanned the middle and lower back). The provided shirt not only provided privacy by allowing the electrodes to be placed on the subject without having to lift their shirt but also permitted easy access to the electrodes during testing. A Polar heart rate monitor was used to measure the subject’s heart rate throughout the experiment for both data collection and ensuring the safety of the subject. Markers were attached to the subject’s left shoulder, elbow, wrist, hip, knee, and ankle for analysis of the trunk flexion angle. The subject completed two tests during the study; in the first, the subject stood in a stationary position while carrying increasing loads while the subject walked on a treadmill while carrying the same loads in the second. For both tests, the subject first had two control trials where they did not carry a backpack (0% BM). Then loads of 10, 15, and 20% BM were tested in increasing order with the order of the two backpacks randomized within each load level. Various books of known weights were placed in each backpack so that the total weight of the backpack plus the books equaled the %BM being tested. Larger books were placed closest to the back and the backpack was positioned with the bottom placed in the curve of the lower back when possible [3]; there were slight variations in the exact position on each subject due to differences in the height of the subjects. The straps were adjusted to hold the backpack close to the body without discomfort [2]. During the standing trials, the subject stood in place for 3 min; during the walking trials, the subject walked 2.41 km/hr on a treadmill for 3 min. For both trials, every 1 min the subject was asked to give a BorgRPE [5] that was recorded along with their heart rate. Lateral view digital photographs of the subject were taken to analyze the trunk flexion angle. Measurements of EMG of the trapezius, latissimus dorsi, and erector spinae muscles were collected throughout the 3 min but only for the standing trials due to signal interference with the treadmill. At the end of each test, the subject was asked to rate the pain they felt in their shoulders, middle back, lower back, and whole body using the Borg-CR10 Scale of Perceived Pain [5]. The subject then rested for 3 min while their heart rate was recorded at 1/2 min and 2 1/2 min to ensure adequate recovery.

2.4. Experimental procedure

2.5. Statistical analysis

When the subject and their parent arrived at the testing site, they received a consent letter that explained

A best subsets analysis was run for each dependent variable to determine the significant variables since

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

343

Table 2 General linear models – standing Dependent Variable

Source of Variance

P-Value

Heart Rate (beats/min)

Age Resting Heart Rate Backpack Mass Mass Resting Heart Rate Backpack Present Load Backpack Mass Resting Heart Rate Backpack Present Load Height Mass Resting Heart Rate Backpack Present Load Age Height Mass Resting Heart Rate Backpack Present Load Age Height Resting Heart Rate Backpack Present Load

< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.003 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.002 0.465 < 0.001 0.025 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.162 0.093 < 0.001 < 0.001 < 0.001

RPE

CR10 Shoulder

CR10 Middle Back

CR10 Lower Back

CR10 Whole Body

many of the variables were strongly correlated. Body mass index and actual mass of the backpack were also included in the analysis as interaction variables. The models were then tested using the general linear test to determine the model that best explained the variance. General linear models were found for EMG of the trapezius, latissimus dorsi and erector spinae muscles, heart rate, Borg-RPE, Borg-CR10 values for the shoulder, middle back, lower back and whole body as well as trunk posture. The variables used to describe the dependent variables were gender, age, height (cm), mass (kg), resting heart rate (beats/min), backpack present (the subjects did not wear a backpack at all for 0% BM), load (%BM), and two interaction terms of backpack mass (kg) and body mass index (BMI in kg/m 2). Although the analysis was not done to predict values for the dependent variables, regression was used because many of the variables were correlated and the model was unbalanced and thus a balanced MANOVA was not possible. Comparisons for gender and load were examined for each dependent variable. All data was analyzed using Minitab 15.0 statistical software with a 5% level of significance.

R-Squared Adjusted for Model 0.7985

0.7612

0.7103

0.6228

0.6027

0.5815

3. Results The results of the general linear models for the standing trials are shown in Table 2. The characteristic and independent variables that accounted for the variance in the dependent variable are listed along with their pvalue. There were no suitable models describing the variance of the EMG data and trunk flexion angle (zero degrees was neutral position and a positive angle indicated the trunk was leaning forward). The resting heart rate was significant in every model while whether or not a backpack was worn as well as load were significant for every variable except heart rate. However, although the factor of whether or not a backpack was worn was indicated by the best subsets analysis in the model for the middle back Borg-CR10 ratings, it does not appear to be significant in the general linear model results (p = 0.465). Table 3 includes the results of the general linear models for the walking trials. Resting heart rate and load were significant in every model, while whether or not a backpack was worn was significant for every variable except heart rate. Since many of the variables were highly correlated, a factor analysis using principal components and

344

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students Table 3 General linear models – walking Dependent Variable

Source of Variance

P-Value

Heart Rate (beats/min)

Gender Mass Resting Heart Rate Load Backpack Mass BMI Gender Age Resting Heart Rate Backpack Present Load Resting Heart Rate Backpack Present Load BMI Resting Heart Rate Backpack Present Load Resting Heart Rate Backpack Present Load Mass Resting Heart Rate Backpack Present Load Resting Heart Rate Backpack Present Load

< 0.001 0.002 < 0.001 < 0.001 < 0.001 0.003 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.004 < 0.001 < 0.001 < 0.001 < 0.001 0.007 < 0.001

RPE

CR10 Shoulder

CR10 Middle Back

CR10 Lower Back

CR10 Whole Body

Trunk Forward Lean (degrees)

equimax rotation was also performed. Three factors were found to describe the data. The first factor, labeled Load Carried, had strong positive influences from the variables backpack present, backpack type, load, and backpack mass. The second factor was labeled Maturity as increasing values of age, height, and weight had the most loading. The last factor was loaded positively by age and negatively by resting heart rate and BMI. This factor was labeled as Fitness. These three factors were then used as the variables in the regression analysis. No meaningful relationships between the three factors and the dependent variables were found as all three factors appeared in every model. The adjusted R-squared values were also lower for the models when using the three factors. Therefore, the original general linear models were used to show which of the measurements influenced the dependent variables. 3.1. Load carried The purpose of the study was to find an acceptable backpack load limit for middle school students; therefore, comparisons between load levels were made for each variable using Dunnett’s comparison with a

R-Squared Adjusted for Model 0.7720

0.7269

0.6689

0.5416

0.5469

0.6566

0.6016

control (0% BM as control) and Tukey-Kramer’s pairwise comparisons (0% BM as a level). There were no significant differences in the standing trials for EMG trapezius, EMG latissimus dorsi, and EMG right erector spinae. In the walking trials, there were no significant differences for heart rate. The only significant changes in the EMG was between loads of 0% and 20% BM and 10% and 20% BM for the left erector spinae muscle and was actually a decrease in activity and not an increase. However, since only one muscle group showed any significant change, a viable conclusion about the increase in load cannot be drawn from the EMG data. For the standing trials, the heart rate showed significant differences between all load levels except between 15 and 20% BM, but the average heart rate for all subjects actually decreased as the load increased. There were no significant differences in heart rate for the walking trials. Comparisons showed that all levels were significantly different for Borg-RPE in both trials, which does not indicate one load limit is better than another is. Dunnett’s comparisons of the Borg-CR10 ratings show the only non-significant differences are between 0% and 10% BM for every area except the shoulder in

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

345

10 0% BM

9

10% BM

15% BM

20% BM

8

Borg-CR10

7 6 5 4 3 2 1 0 Shoulder

Middle Back

Lower Back

Whole Body

Body Area Fig. 1. CR10 average responses for each body area – standing. 10 0% BM

9

10% BM

15% BM

20% BM

8

Borg-CR10

7 6 5 4 3 2 1 0 Shoulder

Middle Back

Lower Back

Whole Body

Body Area Fig. 2. CR10 average responses for each body area – walking.

the standing trial. The Tukey-Kramer results show a similar outcome with an additional non-significant difference between 10% and 15% BM for the middle back while standing. The plot of the data, as seen in Fig. 1, indicates that there may be a large enough increase from 10% to 15% BM for the shoulder and whole body to suggest a load limit of 10% BM. The comparisons for the walking trials show significant differences for all pairs except 0% and 10% BM for all Borg-CR10 ratings, and Fig. 2 shows a similar jump to that in the standing trials between 10 and 15% BM.

In the standing trials for trunk flexion angle, nonsignificant differences were found between 0 and 10% BM, 10 and 15% BM, and 15 and 20% BM. These results do not point to one load limit over another, but it can be seen in Fig. 3 that the trunk flexion angles at 15 and 20% BM were greater than the average trunk flexion angle found during normal gait [39]. From the walking trial results (Fig. 4), comparisons showed that the trunk flexion angles at 0 and 10% BM were the same and significantly different than 15 and 20% BM. Once again, the trunk flexion angles for 15 and 20% BM were greater than the angle during normal gait.

346

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students 16 Trunk flexion angle during normal gait

Trunk Flexion Angle (deg)

14 12 10 8 6 4 2 0 0

10

15

20

Load (% BM) Fig. 3. Average trunk posture (deg) for load (%BM) – standing. 16 Trunk flexion angle during normal gait

Trunk Flexion Angle (deg)

14 12 10 8 6 4 2 0 0

10

15

20

Load (% BM) Fig. 4. Average trunk posture (deg) for load (%BM) – walking.

3.2. Other factors Other variables that showed to be factors in the general linear models are resting heart rate, backpack worn, age, height, and mass. Resting heart rate, height, and mass were not examined because no practical comparisons could be made due to the large number of values since each subject was unique. Comparisons on backpack worn or not indicate wearing no backpack is different than wearing a backpack with any load. When examining the differences between ages, the standing trials indicated no clear differences between the ages.

However, for the walking trials, it appeared that age 11 years might be significantly different than the other ages in their Borg-RPE and Borg-CR10 responses.

4. Disscussion Previous studies on backpacks and children have resulted in different recommendations for a backpack load limit [9,21–23]. Whether the authors suggested values based on where the data simply shows a significant change according to comparison tests or have tried

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

to look for more solid evidence that one load is less safe than another, one fact on which the studies all agree is that there are definite physiological changes that occur as a backpack becomes heavier. However, there is still no consensus on a limit for just how heavy a backpack should be for children. This study was conducted to not only provide more evidence on the vital problem of middle school students carrying backpacks that may cause long-term harm, but to also attempt to support one weight limit over the other. 4.1. Suggestion for load limit Only one muscle (left erector spinae) showed any changes with load. The EMG showed a decrease in muscle activity, which would correspond to the idea that as the load increases, the erector spinae does not have to work to extend the trunk [34]. Instead, the backpack acts as the extensor and the rectus abdominis muscles are activated and work to balance the load and keep the trunk from leaning backward as noted by Devroey et al. [12]. Although the right erector spinae was not found to have any significant differences, the same decreasing pattern in muscle activity was found. The only other muscle to show a trend was the left trapezius, which showed a decrease in the EMG as the load increased, counter-intuitive of what one would believe to happen as a backpack is designed to be carried on the shoulders. A possibility for this occurrence may be that some of the load was transferred to other muscle groups as the load increased causing the total activity of the trapezius to decrease [4]. No other muscles showed any significant changes with an increasing load. This could be explained by the fact that the measurements were taken while standing. Although Devroey et al. [12] found significant differences while standing, the changes were more pronounced while the subjects were walking. This may relate to the idea that one must work to maintain their balance during gait and thus involves more muscle activity. Significant changes in heart rate were observed during the standing trials, but the heart rate decreased as the load increased, opposite of what would be expected. Explanations for this occurrence are that the activity was not dynamic and as the load increased, the subjects became more relaxed. Using heart rate as an estimate for energy expenditure is intended only for dynamic work [35,41], and, as standing is not a dynamic activity, heart rate is not a good measure for determining a load limit in this trial. Another reason for the drop in heart rate is that the loads were tested in increasing

347

order to reduce any effect of fatigue. This also introduced the possibility of an increased heart rate at the beginning of the tests due to anxiety or excitement. As the experiment progressed, the subjects became more relaxed and, as a result, the heart rate decreased as the load level increased. For the walking trials, no significant differences were found for heart rate as the load increased. Although this activity was dynamic, the level of intensity may not have been high enough for many of the subjects to experience any rise in heart rate. The walking speed (2.41 km/hr) was set for the youngest age and kept constant throughout the study. This meant the speed was most likely too slow for the older or taller subjects to affect heart rate. The heart rate did increase from resting heart rate for all subjects, but not enough to show any significant differences between loads. Significant differences were found between all levels of load for the Borg-RPE values. However, the rate of change between each level is approximately the same and no limit can be suggested from the results. A reason no meaningful differences can be found for the standing trials is that Borg’s RPE-Scale is intended for use in dynamic situations just like the heart rate measurements [5]. Although the scale is not proposed for use during static workloads, the ratings were obtained to see how the subjects would rate the load while standing, which is an activity they may perform throughout the day. The results of the walking trials for load may be explained by the speed of the treadmill. Since the activity was not strenuous enough for some of the subjects, the ratings were probably lower than if the subjects had walked at a faster pace. When examining the data for load, a couple of interesting results are seen in the comparisons for the Borg-CR10 ratings. In both Dunnett’s comparison with 0% BM as the control and Tukey-Kramer’s pairwise comparison with 0% BM as a level, there is no significant difference between the loads of 0 and 10% BM (except for one borderline significant comparison for the shoulder while standing). This would indicate that the loads of 0% and 10% BM are similar in terms of discomfort and pain to the subject. Further analysis of the Borg-CR10 ratings shows that the mean values for 15% BM are two to three times higher than the values for 10% BM while the largest change from 15% to 20% BM is about 180%. Although the differences in change may seem small, the average change between 10 and 15% BM is 245% while the average increase between 15 and 20% BM is 176%. This is a disparity of almost 70% and, combined with the non-significant difference

348

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

between no load and 10% BM, leads to the idea that students feel a noticeable change in pain when carrying loads of 15% BM when compared to 10% BM. This disagrees with the findings of Devroey et al. [12] that loads of 0% are significantly different than loads of 10% BM. Non-significant differences between 0 and 10% BM were also found for the trunk flexion angle while walking corresponding to the findings of Li et al. [29]. The difference between 10 and 15% BM was once again examined, which showed an increase of 174%. Compared to a difference of 119% between the 15 and 20% BM loads, the results of the trunk flexion angle during walking support the suggested load limit of 10% BM from the Borg-CR10 rating results. When examining the average trunk flexion angle for both trials, the values for both 15 and 20% BM are higher than the angle experienced in normal gait [44]. This corresponds to the findings of Li et al. [29] and further supports a load limit of 10% BM. 4.2. Other findings The significant difference found between wearing no backpack and wearing a backpack is intuitive, as it would make sense that not wearing a backpack (0% BM) would be different than wearing a loaded backpack. Therefore, no conclusions on load limits can be made from these comparisons except that wearing no backpack is better than wearing a backpack. The walking trials indicate that there is a significant difference between 11 years and all other ages. This may indicate that the age of 11 years should not be grouped with the other ages when determining a load limit for middle school students as they may be physiologically and emotionally immature when compared to the other age groups. The heart rate was also higher for the 11 year olds, which was expected since the normal heart rate is higher in children and decreases with age [47]. 4.3. Recommendations According to the present study, a limit of 10% BM is suggested from the results of the Borg-CR10 ratings and trunk flexion angle. This limit corresponds to other studies and should be considered as the limit for middle school students. However, even with a limit in place, students, parents, and teachers should become more aware of backpack loads and what they can do to help prevent injury. Selecting the right backpack for the age

is the first step in prevention of injury as backpacks that are too big or small can create other problems, such as uneven loads that are awkward to carry or not being able to fit all the books in the bag. Knowing what is carried in the backpack each day will also help ensure that extra items are not needlessly adding to the weight of the bag. Only items required for the day should be included to keep down the mass [3]. 4.4. Suggestions for future research Future research should be conducted to include actual backpack mass instead of %BM to help eliminate confusing percent load limits. Using %BM as a basis for a load limit may be confusing as two students of the same age and school year may be allowed to carry different loads depending on their weight. It may also be possible for a single student to carry more in the second half of the school year as they continue to grow. This idea of actual backpack load was introduced into the present study through an interaction variable and it was shown that the factor was significant in three models (heart rate for both trials and Borg-RPE for standing) and borderline significant in many others. Although there were too many values to make any applicable conclusions, the results lead one to believe that the actual load may be of more importance than the %BM and that there should possibly be a load limit in terms of mass instead of %BM. In fact, the Lativian Ministry of Education presently has load limits for students in terms of actual mass and not %BM (J. Gedrovics, personal communication, August 1, 2007). These limits increase with age and school year, which would take into account both the change in children’s masses and the size of the books needed. If this criterion was adopted for all students, the load limit would be uniform across one grade level and relate more to the actual books required for the school year. This would also take away the idea that an obese student is able to carry more than a physically fit student. To help eliminate the idea that students that have a greater mass can carry more in their backpack, subjects should also be examined in BMI percentile groups. The use of load limits based on %BM indicates that an obese student is able to carry a heavier load when, in fact, they may be at a higher risk for injury due to lack of physical fitness. Almost every study on children and backpacks has focused on the weight of the child instead of the BMI. BMI was entered into the models as an interaction term in this study and showed to be a significant factor in the models of heart rate

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students

and shoulder Borg-CR10 for the walking trials. Just as with the actual backpack weight, the BMI may not have appeared in other models due to the great number of values for the variable. It would be more difficult to get a study population with the same BMI, but the subjects could instead be characterized in percentiles. This would allow a greater range of BMI values by examining groups of under-weight, normal weight, and over-weight and consider how fitness affects the physiological measurements while wearing a backpack. If a load limit must be expressed in a percentage, %BMI may be a better indication of what should be carried to prevent pain and injury. Instead of one load limit, there could be limits based on percentiles that would prevent a child with a high BMI from carrying too much just because their body mass was higher.

[5] [6]

[7]

[8]

[9]

[10]

[11]

5. Conclusions This study was conducted to gather more evidence on the vital problem of children carrying backpacks that may cause long-term harm, and to attempt to support one weight limit over the other. The research indicated a safe load limit for middle school students is 10% BM. This limit is based on the findings that there was no significant difference between 0 and 10% BM in the Borg-CR10 ratings and trunk flexion angle for the walking trial while significant differences were found between 10 and 15% BM. The trunk flexion angle was also higher than that seen during normal gait for loads of 15% and 20% BM. This study only added to the growing number of studies on children and backpacks. There is still much more to examine before a consensus on load limit can be made. Until a universal load limit that everyone can understand is agreed upon, students, parents, teachers, and all others involved should become aware of the problem and things they can do on their own to help prevent injuries.

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

References [21] [1] [2] [3]

[4]

AirPacks, Inc. About Us. Retrieved November 13, 2006, from http://www.airpacks.com//index.html, 2006. Backpack safety, Retrieved December 6, 2005, from http:// www.aap.org/publiced/BR Backpack.htm, 2003. Backpack strategies for parents and students: Pack it light, wear it right, Retrieved December 8, 2005, from http://www. promoteot.org/docs/BackpackStrategies.pdf, 2002. J. Bobet and R.W. Norman, Effects of load placement on back muscle activity in load carriage, European Journal of Applied Physiology and Occupational Physiology 53 (1984), 71–75.

[22]

[23]

[24]

349

G. Borg, Borg’s Perceived Exertion and Pain Scales, Champaign, IL: Human Kinetics, 1998. J.A. Butler, Homework, Retrieved December 16, 2006, from Northwest Regional Educational Laboratory, School Improvement Research Series Web site: http://www.nwrel.org/scpd/ sirs/1/cu1.html, 1987. G. Casey and S. Dockrell, A pilot study of the weight of schoolbags carried by 10 year old children, Physiotherapy Ireland 17(2) (1996), 17–22. D.B. Chaffin, G.B.J. Andersson and B.J. Martin, Occupational Biomechanics, (3rd ed.), New York: John Wiley & Sons, Inc., 1999. D.H.K. Chow, M.L.Y. Kwok, A.C.K. Au-Yang, A.D. Holmes, J.C.Y. Cheng, F.Y.D. Yao and M.S. Wong, The effect of backpack load on the gait of normal adolescent girls, Ergonomics 48(6) (2005), 642–656. COCSA to launch Nat’l Backpack Safety Month in April, Dynamic Chiropractic 20(5) (2002), Retrieved December 17, 2006, from http://www.chiroweb.com/archives/20/05/04. html. J.C. Dale, School backpacks: Preventing injuries, Journal of Pediatric Health Care 18 (2004), 264–266. C. Devroey, I. Jonkers, A. De Becker, G. Lenaerts and A. Spaepen, Evaluation of the effect of backpack load and position during standing and walking using biomechanical, physiological and subjective measures, Ergonomics 50(5) (2007), 728–742. Y. Eddy, Developing homework policies, ERIC Digest, Urbana, IL: ERIC Clearinghouse on Elementary and Early Childhood Education, (ERIC Document Reproduction Service No. ED 256473), 1984. S.N. Forjuoh, School backpack weights: a survey of students in Ghana, Guatemala, and the USA, Injury Control and Safety Promotion 11(4) (2004), 287–289. S.N. Forjuoh, B.L. Lane and J.A. Schuchmann, Percentage of body weight carried by students in their school backpacks, American Journal of Physical Medicine and Rehabilitation 82(4) (2003), 261–266. S.N. Forjuoh, D. Little, J.A. Schuchmann and B.L. Lane, Parental knowledge of school backpack weight and contents, Archives of Disease in Childhood 88 (2006), 18–19. S.A. Goodgold, M. Corcoran, D. Gamache, J. Gillis, J. Guerin and J.Q. Coyle, Backpack use in children, Pediatric Physical Therapy 14(3) (2002), 122–131. S.A. Goodgold and D. Nielsen, Effectiveness of a school-based backpack health promotion program: Backpack intelligence, Work 21 (2003), 113–123. K. Grimmer and M. Williams, Gender-age environmental associates of adolescent low back pain, Applied Ergonomics 31 (2000), 343–360. A. Groslambert and A.D. Mahon, Perceived exertion: Influence of age and cognitive development, Sports Medicine 36(11) (2006), 911–928. Y. Hong and G. Brueggemann, Changes in gait patterns in 10-year-old boys with increasing loads when walking on a treadmill, Gait and Posture 11 (2000), 254–259. Y. Hong and C. Cheung, Gait and posture responses to backpack load during level walking in children, Gait and Posture 17 (2003), 28–33. Y. Hong, J.X. Li, A.S.K. Wong and P.D. Robinson, Effects of load carriage on heart rate, blood pressure and energy expenditure in children, Ergonomics 43(6) (2000), 717–727. S.R. Iyer, An ergonomic study of chronic musculoskeletal

350

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35] [36]

D.H. Bauer and A. Freivalds / Backpack load limit recommendation for middle school students pain in schoolchildren, Indian Journal of Pediatrics 68 (2001), 937–941. K. Jacobs, Are backpacks making our children beasts of burden? Proceedings of the XVI Annual International Occupational Ergonomics and Safety Conference, Toronto: ISOES, 2002. K. Jacobs, R. Lockhart, H.Y. Chiang and M. O’Hara, Book bags for children, in: Ergonomics for Children: Designing Products and Places for Toddler to Teens, R. Lueder and V.J. Berg Rice, eds, Taylor & Francis, New York, 2007, pp. 499– 507. D. Koradecka and J. Bugajska, Physiological Instrumentation, in: The Occupational Ergonomics Handbook, W. Karwowski and W.S. Marras, eds, Boca Raton, FL: CRC Press, 1999, pp. 525–547. S. Legg, J. Whittfield, D. Hedderley and D. Stubbs, Schoolbag weights exceed guidelines for load handling by adults, Retrieved December 17, 2006, from http://www.ergonomics4 schools.com/research/schoolbags.htm, 2003. J.X. Li, Y. Hong and P.D. Robinson, The effect of load carriage on movement kinematics and respiratory parameters in children during walking, European Journal of Applied Physiology 29 (2003), 35–43. R. Lueder and V. Rice, Physical development in children and adolescents and age related risks, in: Ergonomics for Children: Designing Products and Places for Toddler to Teens, R. Lueder and V.J. Berg Rice, eds, New York: Taylor & Francis, 2007, pp. 499–507. W.G. Mackenzie, J.S. Sampath, R.W. Kruse and G.J. SheirNeiss, Backpacks in children, Clinical Orthopaedics and Related Research 409 (2003), 78–84. B. McCarthy and A. Vaishnav, Legislature takes up issue of student loads: Bill filed to limit weight of public schools’ books, The Boston Globe (2003), Retrieved January 24, 2006, from http://www.boston.com/news/local/articles/2003/09/26/ legislature takes up issu of student loads/, 2003. R. Michael, AOTA creates National School Backpack Awareness Day, Ergonomics Today (2002), Retrieved December 17, 2006, from http://www.ergoweb.com/news/detail.cfm?id= 598. R.R.E.E. Motmans, S. Tomlow and D. Vissers, Trunk muscle activity in different modes of carrying schoolbags, Ergonomics 49(2) (2006), 127–138. B.W. Niebel and A. Freivalds, Methods, Standards, and Work Design, (11th ed.), McGraw-Hill, New York, 2003. S. Negrini, R. Carabalona and P. Sibilla, Backpack as a daily load for schoolchildren, The Lancet 354 (1999), 1974.

[37]

[38]

[39]

[40] [41] [42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

T. Puckree, S.P. Silal and J. Lin, School bag carriage and pain in school children, Disability and Rehabilitation 26(1) (2004), 54–59. M.F. Reneman, B.J.J. Poels, J.H.B. Geertzen and P.U. Dijkstra, Back pain and backpacks in children: Biomedical or biopsychosocial model? Disability and Rehabilitation 28(20) (2006), 1293–1297. R.A. Robergs and R. Landwehr, The surprising history of the “HRmax=220-age” equation, Journal of Exercise Physiology 5(2) (2002), 1–10. M.F. Rodts, Backpacks, back pain: Watching what our children carry, Orthopaedic Nursing 23(2) (2004), 99. M.S. Sanders and E.J. McCormick, Human Factors in Engineering and Design, McGraw-Hill, New York, 1993. D. Siambanes, J.W. Martinez, E.W. Butler and T. Haider, Influence of school backpacks on adolescent back pain, Journal of Pediatric Orthopaedics 24(2) (2004), 211–217. D.L. Skaggs, S.D. Early, P. D’Ambra, V.T. Tolo and R.M. Kay, Back pain and backpacks in school children, Journal of Pediatric Orthopaedics 26(3) (2006), 358–363. V.P. Stokes, C. Andersson and H. Forssberg, Rotational and Translational movement features of the pelvis and thorax during adult locomotion, Journal of Biomechanics 22(1) (1989), 43–50. F.C. Trevelyan and S.J. Legg, Back pain in school children – Where to go from here? Applied Ergonomics 37 (2006), 45– 54. VF Outdoor, Inc, Super Break Collection, Retrieved November 13, 2006, from http://www.jansport.com/js home.php, 2007. L.A. Wallis, M. Healy, M.B. Undy and I. Maconochie, Age related reference ranges for respiration rate and heart rate from 4 to 16 years, Archives of Disease in Childhood 90 (2005), 1117–1121. Y.T. Wang, D.D. Pascoe and W. Weimar, Evaluation of book backpack load during walking, Ergonomics 44(9) (2001), 858– 869. T.R. Waters, V. Putz-Anderson and A. Garg, Applications Manual for the Revised NIOSH Lifting Equation (DHHS (NIOSH) Publication No. 94–110), US Government Printing Office, Washington, DC, 1994. L.J. White, H.E. Frasure and P. White, Weight of backpacks carried by school children: students or serpas? Academic Emergency Medicine 7(5) (2000), 487. J.K. Whittfield, S.J. Legg and D.I. Hedderley, The weight and use of schoolbags in New Zealand secondary schools, Ergonomics 44(9) (2001), 819–824.

Copyright of Work is the property of IOS Press and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.