This article was downloaded by: [New York University] On: 08 June 2015, At: 03:26 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Traffic Injury Prevention Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/gcpi20

Bicycle Helmet Size, Adjustment, and Stability a

bc

Kim T. Thai , Andrew S. McIntosh

d

& Toh Yen Pang

a

School of Aviation, University of New South Wales, Sydney, Australia

b

ACRISP, Federation University Australia, Ballarat, Australia

c

Monash Injury Research Institute, Monash University, Melbourne, Australia

d

School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, Melbourne, Australia Accepted author version posted online: 20 Jun 2014.Published online: 10 Nov 2014.

Click for updates To cite this article: Kim T. Thai, Andrew S. McIntosh & Toh Yen Pang (2015) Bicycle Helmet Size, Adjustment, and Stability, Traffic Injury Prevention, 16:3, 268-275, DOI: 10.1080/15389588.2014.931948 To link to this article: http://dx.doi.org/10.1080/15389588.2014.931948

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Traffic Injury Prevention (2015) 16, 268–275 C Taylor & Francis Group, LLC Copyright  ISSN: 1538-9588 print / 1538-957X online DOI: 10.1080/15389588.2014.931948

Bicycle Helmet Size, Adjustment, and Stability KIM T. THAI1, ANDREW S. MCINTOSH2,3, and TOH YEN PANG4 1

School of Aviation, University of New South Wales, Sydney, Australia ACRISP, Federation University Australia, Ballarat, Australia 3 Monash Injury Research Institute, Monash University, Melbourne, Australia 4 School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, Melbourne, Australia 2

Downloaded by [New York University] at 03:26 08 June 2015

Received 1 December 2013, Accepted 3 June 2014

Objectives: One of the main requirements of a protective bicycle helmet is to provide and maintain adequate coverage to the head. A poorly fitting or fastened helmet may be displaced during normal use or even ejected during a crash. The aims of the current study were to identify factors that influence the size of helmet worn, identify factors that influence helmet position and adjustment, and examine the effects of helmet size worn and adjustment on helmet stability. Methods: Recreational and commuter cyclists in Sydney were surveyed to determine how helmet size and/or adjustment affected helmet stability in the real world. Anthropometric characteristics of the head were measured and, to assess helmet stability, a test analogous to the requirements of the Australian bicycle helmet standard was undertaken. Results: Two hundred sixty-seven cyclists were recruited across all age groups and 91% wore an AS/NZS 2063–compliant helmet. The main ethnic group was Europeans (71%) followed by Asians (18%). The circumferences of the cyclists’ heads matched well the circumference of the relevant ISO headform for the chosen helmet size, but the head shapes differed with respect to ISO headforms. Age and gender were associated with wearing an incorrectly sized helmet and helmet adjustment. Older males (>55 years) were most likely to wear an incorrectly sized helmet. Adult males in the 35–54 year age group were most likely to wear a correctly adjusted helmet. Using quasistatic helmet stability tests, it was found that the correctness of adjustment, rather than size, head dimensions, or shape, significantly affected helmet stability in all test directions. Conclusions: Bicycle helmets worn by recreational and commuter cyclists are often the wrong size and are often worn and adjusted incorrectly, especially in children and young people. Cyclists need to be encouraged to adjust their helmets correctly. Current headforms used in standards testing may not be representative of cyclists’ head shapes. This may create challenges to helmet suppliers if on one hand they optimize the helmet to meet tests on ISO-related headforms while on the other seeking to offer greater range of sizes. Keywords: bicycle helmets, head anthropometry, stability testing

Introduction The stability of a helmet in a crash is critical to the performance of the helmet because it will ensure optimal coverage of the cranium. Helmet stability is determined by a number of precrash helmet design elements as well as human factors and describes the extent that the helmet will remain in position on the head without translating or rotating. A helmet with inadequate stability may rotate forward and obscure vision or rotate rearwards and expose the forehead to a direct impact. Bicycle helmets are designed to protect cyclists from injuries resulting from impacts to the head during crashes, including falls. Epidemiological research has shown that bicycle helmets

Managing Editor David Viano oversaw the review of this article Address correspondence to Dr. Andrew S. McIntosh, Australian Centre for Research Into Injury in Sport and Its Prevention (ACRISP), Federation University Australia, P.O. Box 668 Ballarat Victoria 3353, Australia. E-mail: as.mcintosh@ bigpond.com

compared to no helmet reduce the risk of head and brain injury in the range 50–88% (Attewell et al. 2001; McIntosh, Curtis, et al. 2013; Thompson et al. 1999). Helmet stability may be one factor that affects helmet performance in crashes. Bicycle helmets are designed for a mass market and are intended to meet the broad needs of the population through a combination of manufacturer-specified discrete sizes (e.g., small, medium, large, and extra-large) that are labeled on the helmet, user-determined comfort/sizing inserts, and an adjustable retention system, including webbing straps and a clasp that fastens under the chin. In addition, many helmets have an adjustable nape strap (also known as an occipital stabilizer) at the rear of the helmet that may be considered part of the retention system. There are fixed design factors that affect fit and stability—for example, the geometry and dimensions of each size in a helmet model relative to the individual’s head, where size is typically based on head circumference—and userdetermined factors; for example, selection of helmet model and size and adjustment of the restraint system. Poor helmet stability can compromise the impact performance of the helmet by exposing the head to direct impacts and resultant

Downloaded by [New York University] at 03:26 08 June 2015

Bicycle Helmet Size, Adjustment, and Stability injury (Ching et al. 1997; Grimard et al. 1995; Rivara et al. 1999; Thompson et al. 1997). At worst, a bicycle rider may not fasten the clasp, which results in the helmet falling off often before the head is struck. Alternatively, the restraint system is not tightly adjusted, the helmet is the incorrect size, and/or the fit is achieved with an overreliance on comfort/sizing inserts. In oblique impact tests (drop height 1 m and striker speed 25 km/h), reductions in angular headform acceleration and impact force were observed with the helmet restraint system adjusted tightly compared to the “2-finger tight” adjustment commonly recommended (McIntosh, Lai, et al. 2013). At present the evidence base on helmet size, adjustment, and stability in a population of cyclists is limited. In most bicycle helmet test standards (AS/NZS 2063 (2008) Standards Australia, CPSC (1998) Consumer Produce Safety Commission, EN 1078 (1997) European Standard, JIS T 8134 (2007) Japanese Standards Association, Snell B95 (1998) Snell Foundation), retention system strength and positional stability are assessed. Impact performance is assessed within a surface defined by specific test lines in helmet standards. The helmet supplier specifies the position of the helmet for the purpose of marking test lines on the helmet. Therefore, the impact performance is assessed with respect to an assumed optimal helmet position on the head. It is widely regarded that helmet stability is more critical in protecting the head than retention system strength alone (Andersson et al. 1993; Chang et al. 2001; Gilchrist et al. 1988; Parkinson and Hike 2003; Rivara et al. 1999; Thompson et al. 1997). In Australia, all bicycle helmets sold must meet AS/NZS 2063 (Standards Australia/Standards New Zealand 2008) wherein the stability of a helmet is assessed in a “static” test using the following method: the helmet without comfort/sizing inserts is fastened to an ISO A or J size headform, testing is performed on the helmet size which is “judged to be an appropriate,” a test band is defined between the basic plane (equivalent to the Frankfurt plane) and a parallel test line at a headform specific height above the basic plane, a 50 N static tensile force is applied separately to the front and rear rims of the helmet, and, the helmet fails if the test band is fully exposed or obscured (AS/NZS 2512.7.1 2006; Standards Australia/Standards New Zealand 2008). Most standards have adopted ISO headforms, whose dimensions are based on those in the draft standard ISO/DIS 6220–1983, Headforms for Use in the Testing of Protective Helmets. Hagel et al. (2010) found a 15% prevalence rate of incorrect helmet use. The authors found that age influenced correct helmet use, with children and adolescents less likely to wear helmets correctly than were adults, but gender did not. Grimard et al. (1995) reported that the helmet had been ejected during a crash in 15% of hospitalized child cyclists who had worn a helmet. Ching et al.’s (1997) observation that the majority of helmeted cyclists with injuries to the forehead also had impact damage to the front rim of their helmets suggests that either the helmet was out of position at the time of the impact or rotated during the impact. Information on the helmet adjustment was not available in either study. Rivara et al. (1999) identified helmet wearing position and fit as factors that contributed to head injury risk in hospitalized child cyclists: cyclists who wore their helmets tilted posteriorly had a 52% greater risk of head injury than those who had it posi-

269 tioned at the correct position, and head-injured cyclists wore helmets that were significantly wider than their heads compared to uninjured cyclists. Mihora et al. (2007) measured the distance between the bottom of the helmet retention strap and the chin of 36 cyclists when a 0.5 kg load was applied. They found that helmets were worn loosely with an average 4.7 cm of slack. Dynamic stability testing in the study (Mihora et al. 2007) using a pendulum device found that with this amount of chin strap slack, a bicycle helmet to ground impact resulted in significant forward rotation of the helmet on the headform and the helmet could easily be rotated off the head after the test. These papers suggest that the position of the helmet on the head (e.g., tilted back), the congruity in shape and size between the helmet and head shape (i.e., fit), and the adjustment of the restraint system are important factors in determining helmet stability and crash performance. The aims of the current study were to (1) identify factors that influence the size of the helmet worn; (2) identify factors that influence in use helmet position and adjustment; and (3) examine the effects of helmet size worn and adjustment on helmet stability.

Methods Participants Participants from the Sydney Metropolitan Area were recruited over a 12-month period in 2008–2009. Helmet use is mandatory for all ages in Australia. The target sample size was 300. Recreational cyclists were approached at 4 popular recreational sites on 11 separate weekend occasions. Commuter cyclists were recruited through advertisement at the University of New South Wales and were surveyed by appointment. The inclusion criteria were cyclists aged over 4 years and wearing a helmet. Participants received a gift voucher as compensation for the time spent participating in the study. The study methods were approved by the University of New South Wales Human Research Ethics Committee. Informed consent was obtained from the participants or their parent/guardian. Measurements and Assessments Participants completed a short structured interview in which data on anthropometry, demographics (gender, age and ethnicity), cycling type (recreational or commuting) and habit (frequency and location), and helmet use were collected. Helmet use, fit, and stability tests were assessed. At the conclusion of the assessment all participants were advised on proper helmet positioning and adjustment. Head Anthropometry Standard techniques were used to measure the anthropometric characteristics of the head (Greulich et al. 1938). A sliding caliper was used to measure head anterior–posterior length, the maximum distance between the glabella and the occiput, and head width (or breadth), the maximum horizontal distance between the temples above the ears (Figure 1). Head

270

Thai et al.

Downloaded by [New York University] at 03:26 08 June 2015

Fig. 1. Anthropometric measurement of the head using a sliding caliper: width (left) and anterior–posterior length (right).

circumference was measured using a flexible measuring tape placed above the brow. A helmet was considered to be the correct size if the wearer’s head circumference fell within the specified size range; for example, a child with a 54-cm head circumference wearing a helmet with a specified range of 53–56 cm was wearing the correct size. Assuming that bicycle helmets are designed to fit the ISO test headforms, comparisons between cyclist head length and width dimensions and the corresponding dimensions of ISO headform sizes A, E, J, M, and O specified in AS/NZS 2512.1 (2009) were made. Dimensions within ±5 mm of headform dimensions were considered to be similar (Meunier et al. 2000). The cephalic index (CIx), the ratio of head width to length expressed as a percentage, was derived. Cyclist head shapes were categorized into 3 groups: dolichocephalic (CIx < 75%), mesocephalic (75% ≤ CIx < 80%), and brachycephalic (CIx ≥ 80%). These categories were based on the definitions proposed by Garson (1887). These groupings permitted a comparison between cyclists’ head shapes and the ISO headforms (A, E, J, M, and O), which were in the mesocephalic range (75.6–80.2%). Helmet Wearing Position and Adjustment The size, age, and standard compliance of each helmet was transcribed from the helmet marking and labeling. The condition of each helmet was recorded as either “new,” “good,” “old/worn,” or “damaged” based on a visual inspection. Deformation of the liner, sliding abrasions to the shell, and/or fractures of the shell were considered signs of prior crash involvement. Each participant was asked to rate his or her helmet in terms of comfort, fit, and usability on a 5-point Likert scale (very poor, poor, average, good, excellent). Helmet position and adjustment were judged from visual and physical assessments, in line with common definitions of correct use (Lee et al. 2009). A helmet was considered correctly adjusted and positioned if it sat squarely on the head with the front rim no more than 2 finger-widths (≈50 mm) above the brow and with no more than 2 finger-widths of slack in the strap present under the chin. Helmet Stability To assess helmet stability a test analogous to the requirements of the Australian Standard (AS/NZS 2512.7.1 2006;

Fig. 2. Static stability test orientations from left to right: photograph from field, rearward, forward, and lateral.

Standards Australia/Standards New Zealand 2008) was undertaken. The interior comfort liner of each helmet was fitted with a hook and loop fastener strap before the participants were asked to fasten the helmet “as they would normally.” The end of the strap was attached to a Mecmesin AFG digital force gauge (Mecmesin Ltd., Sinfold, England) and a tensile force was applied quasistatically over the helmet horizontally in the forward, rearward, and lateral directions from the rear, front and side rims, respectively (Figure 2). The force was applied by hand at as close to a constant rate as possible. The minimum tensile force required to displace the helmet and the angular displacement of the helmet were recorded. Each test was stopped either when the helmet ceased rotating on the wearer’s head or when it became too uncomfortable for the wearer. The position of the helmet was then assessed. The degree of helmet rotation on the wearer’s head was measured using a goniometer. For the forward and rearward tests, the goniometer was placed on the coronal plane above the side of the external ear opening parallel to the side rim of the helmet (Figure 3). For the lateral tests, the goniometer was placed on the midsagittal plane parallel to either the front or rear rim of the helmet. A helmet was considered unstable if the angular displacement exceeded 10◦ in any direction. After the initial 3 tests were complete, the helmet was repositioned and the restraint system adjusted by the tester for optimum fit and stability, as per the judgement of the tester, and the tests were repeated. Data and Statistical Analyses Helmet size and adjustment, as well as stability in the 3 test directions, were treated as dichotomous dependent

Bicycle Helmet Size, Adjustment, and Stability

271

Downloaded by [New York University] at 03:26 08 June 2015

Fig. 3. Measurement of helmet position using goniometer in stability test. The line of action of the tensile force is also presented in the figure.

variables with the categories as defined above. Relationships between the independent variables (behavioral, physical, and helmet-specific factors) and the dependent variables were assessed using binary logistic regression analysis (Hosmer and Lemeshow 2000). The following factors and associations were assessed:

Fig. 4. Age and gender distribution for cyclists surveyed in the study (N = 267).

• Behavioral factors: Associations between helmet size and adjustment by cyclist age, gender, main riding type, riding habit (experience, frequency, carriageway preference), and prior crash involvement were assessed. • Physical factors: Cyclists’ head shapes were tested as potential influences on size wearing and adjustment. Associations between helmet stability and the differences between cyclists’ and headform length and width dimensions and cyclists’ head shapes were assessed. • Helmet specific: The following associations were assessed: Size correctness by helmet age and the user ratings of comfort and fit; helmet adjustment correctness by helmet condition, retention system design, and usability ratings; and helmet stability by size correctness, adjustment correctness, and retention system design were assessed.

A total of 267 cyclists were recruited over the 12-month period, slightly fewer than the target sample size. Eighty-three percent were recruited at the recreational sites and 17.6% were recruited through advertising. The sample included 188 males (70%) and 79 females (30%) between the ages of 4 and 68 years (M = 30.1 years, SD = 17.8; Figure 4). Eighty-one participants (30%) were children under 18 years of age. Cyclists surveyed were of European (Caucasian; 71%), Asian (18%), Middle-Eastern (5%), American (2%), Australian-Aboriginal (1%), and African (1%) ethnicity. Fifteen participants (6%) were of mixed ethnicity. The majority of those surveyed cycled largely for recreation (72%) and cycled infrequently (up to 5 h per week) (78%). The sample included mainly cyclists having at least 3 years of riding practice (57%). Most cyclists (67%) indicated that they were only comfortable cycling in dedicated cycleways and parks or on suburban streets. One hundred and fifty-one cyclists (57%), including 51% of recreational riders and 69% of commuters, had previously been involved in at least one cycling crash or fall. Most cyclists (90%) stated that they always wore a helmet when riding.

In calculating the unadjusted odds ratios for factors with multiple categories (more than 2), individual categories were compared with all other categories combined; for example, for age, children were compared with all people aged 18 and above. Odds ratios were then adjusted for other significant factors using multivariable logistic regression analyses. In order to obtain parsimonious models, selected variables were sequentially excluded if they were found to have nonsignificant contributions in the multivariable model, following a backward stepwise logistic regression method, which tests the probability of the likelihood ratio. The minimum level of contribution for the model α was set at .20 (Hosmer and Lemeshow 2000). Due to the presence of missing values, the models were then reapplied to the full data set. Chi-squared (χ 2) test for independence was used to compare helmet size wearing and adjustment for stable and unstable helmets. Comparisons of tensile forces for stable and unstable helmets were conducted using Student t tests. All statistical analyses were conducted using PASW 18 for Windows software (SPSS Inc., Chicago, IL). The statistical significance α level was set at .05.

Results Characteristics of Cyclists

Cyclist Head and ISO Headform Dimension Comparisons Cyclist head circumferences were comparable to the ISO headforms used in helmet testing (A, E, J, M, O) but differed significantly in shape (Table 1). Head length ranged from 1.8 cm longer to 2.6 cm shorter than matching headform length (mean difference = 0.4 cm, SD = 0.73 cm, P < .001), with less than half (45.5%) of head lengths similar to headform length (within ±5 mm). Head width ranged from 4.6 cm wider to 2.4 cm narrower than headform width (P > .05). Only 40% of heads were similar in width to the matching headforms. These differences were reflected in CIx, which ranged from 21.3% higher to 11.9% lower than headform CIx (mean difference = −1.77%, SD = 5.1%, P < .001). Twelve percent (12.1%) of cyclists had a dolichocephalic head and 40.9% had a brachycephalic head. There were

272

Thai et al.

Table 1. Comparison of cyclists’ head measurements with those of ISO headforms A, E, J, M, O Measure Circumference (cm) Length (cm) ‡ Width (cm) ‡ CIx (%)∗ ∗P

ISO range (mean, SD) 50.0–62.0 (56.6, 4.77) 17.6–21.7 (19.9, 1.68) 13.4–17.4 (15.6, 1.63) 76.1–80.2 (78.3, 0.02)

Table 2. Helmet wearing correctness among children (under 18 years) and adultsa

Sample range (mean, SD) 50.0–62.5 (56.5, 2.61) 15.9–22.0 (19.2, 1.06) 13.3–19.9 (15.3, 0.84) 66.8–98.5 (79.7, 4.73)

< .001.

Helmet worn

Downloaded by [New York University] at 03:26 08 June 2015

Helmets Characteristics Ninety-one percent of the 267 helmets bore a label indicating compliance with AS/NZS 2063 (Standards Australia/Standards New Zealand 2008). Five percent indicated compliance with another helmet standard and the remaining 4% of helmets had no compliance label. Most helmets (60%) were manufactured within the preceding 5 years. Twenty percent of helmets were more than 5 years old and the date of manufacture was unclear in the remainder. On visual inspection, most helmets (78%) were in new or good condition; 20% were old or worn, and 2% were damaged. The majority of cyclists were satisfied with the comfort, fit, and usability aspects of their helmets. In terms of fit, the mean score was 3.72 (SD = 0.86), with 71% of participants rating their helmet fit as “good” or “excellent.” Cyclists generally chose particular helmets for their affordability (23%), comfort and fit (22%), and/or their styling or design (21%). It is important to note that 22% of helmets were not selected by the cyclists but were received as a gift, recommended to them, or came with the bicycle purchase. Most cyclists (55.4%) wore the correct size of helmet according to their head circumference. Eighteen percent of cyclists wore a helmet that was too large and 9% wore on that was too small. Helmet size could not be determined in 17.6% of helmets, for example, because the label was missing. Factors found to be significantly associated with wearing a helmet of incorrect size included age and gender. Specifically, 59.1% of cyclists 55 years or older wore incorrect size helmets and were 3.5 times more likely than all other age groups combined to do so (odds ratio [OR] = 3.462, 95% confidence limit [CL], 1.403–8.544, P < .01). Males were twice as likely as females (OR = 1.953, 95% CL, 1.02–3.738, P < .05) to wear the incorrect size helmet. When adjusted for gender, older cyclists (55 years and over) were 3.3 times (OR = 3.333, 95% CL, 1.340–8.289, P = .01) more likely to wear incorrectly sized helmets than all other age groups. When adjusted for age, gender was not a significant factor influencing size selection.

% of adults (N = 186)

% of all (N = 267)

43.2 40.7 14.8 7.4 1.2

57.3 29.2 8.6 5.4 2.7

53.2 30.3 9.0 5.2 2.2

Correctly Loose/unsecured High on forehead Over other headwear Other aχ 2

significant differences between mean CIx for the various ethnic groups (P < .001) but not for age, gender, or body mass index. People of northeast Asian descent had significantly higher (P < .01) cephalic indices than those of northwestern European ancestry, with a mean difference of 4.9% (SD = 1.14%), and higher than people of North African or Middle Eastern ancestry by an average of 6.2% (SD = 1.67%).

% of children (N = 81)

= 6.61, P > .05.

Helmet Adjustment Only 53.2% of cyclists were deemed to have worn the helmet correctly (Table 2). Age and gender were found to be significantly associated with helmet adjustment. Adults in the 35–54 year age group were most likely to wear a correctly adjusted helmet (OR = 0.481, 95% CL, 0.282 to 0.822, P < .01), and males were 1.8 times more likely than females (OR = 1.779, 95% CL, 1.045 to 3.021, P < .05). The multivariable regression analysis resulted in adjusted odds ratios that were similar to the unadjusted ratios for the 2 factors (Table 3). When adjusted for gender, adult cyclists (35–54 years) were more than twice as likely as children (under 18 years) and young adults (18–34 years) to correctly adjust their helmets. When adjusted for age, males were 1.9 times more likely than females to correctly adjust their helmets.

Helmet Stability Two hundred and sixty-six pedal cycle helmets were subjected to in vivo stability tests in the rearward and forward directions and 265 in the lateral direction. The overall proportion of helmets stable in the forward direction was 80.5%, and in the lateral and rearward directions it was 73.2 and 57.5%, respectively. One hundred and fifteen (43.4%) helmets were found to be stable in all 3 directions. Mean tensile forces of 7.58 N (SD = 3.50), 9.37 N (SD = 4.08), and 8.24 N (SD = 3.79) were measured for the rearward, forward, and lateral directions, respectively (Table 4). The measured tensile forces were significantly higher for Table 3. Logistic regression model for incorrect helmet adjustment among cyclists (N = 266) B

SE

Wald’s χ 2

df

P

exp(B) (95% CL)

0.898

0.323

7.733

1

0.005

2.454 (1.303–4.620)

0.772

0.336

5.273

1

0.022

2.164 (1.120–4.182)

0.412

0.468

0.775

1

0.379

1.510 (0.603–3.781)

0.625 −1.033

0.278 0.252

5.059 16.748

1 1

0.024 0.000

1.869 (1.084–3.223) 0.356 (—)

Factor †

Child (vs. 35–54 years) 18–34 years∗ (vs. 35–54 years) 55+ years (vs. 35–54 years) Female∗ Constant ∗P

< .05. †P < .01.

Bicycle Helmet Size, Adjustment, and Stability

273

Table 4. Mean tensile forces measured in rearward, forward, and lateral directions by degree of helmet rotation

Table 6. Rearward, forward, and lateral stability by helmet adjustment

Mean (SD, N) tensile force (N) Rearward∗

Degree of rotation ≥10◦ Rotation (unstable) .05). Stability was improved in terms of both degree of rotation and force in 48, 39, and 41% of rearward, forward, and lateral postreadjustment tests, respectively. Head dimensions and shape were found not to influence stability. The presence of an adjustable nape strap was significantly associated with helmet stability. Helmets without an adjustable nape strap were 2.3 times more likely to be unstable (in at least one direction; OR = 2.295, 95% CL, 1.022–5.156). In terms of the specific directions, it was found that the inclusion of an adjustable nape strap improved stability in the rearward and lateral directions but not in the forward direction.

Table 5. Helmet stability in the rearward, forward, and lateral directions. Tests were undertaken in 219 out of the 220 cases in which helmet size was known Helmet stability (%) Helmet size worn Incorrect Correct χ2 Any ∗P

> .05.

Rearward (N = 219)

Forward (N = 219)

Lateral (N = 218)

All (N = 218)

58.3 57.8 0.005∗ 58.0

88.9 78.9 3.287∗ 82.2

73.2 71.4 0.078∗ 72.0

45.2 43.5 0.046∗ 44.0

Rearward (N = 266)

Forward (N = 266)

Lateral (N = 265)

All (N = 265)

34.2 74.2 42.267∗ 57.5

70.3 87.7 12.554∗ 80.5

60.4 82.5 16.073∗ 73.2

21.6 59.1 36.868∗ 43.4

< .001.

Discussion A field survey of 267 cyclists found that bicycle helmets worn by recreational and commuter cyclists are often the wrong size and are often positioned and adjusted incorrectly. A novel approach was applied to study a cohort of cyclists and with their helmets, which included measurements of helmet stability that were similar to those applied in standards testing. Stability testing found that bicycle helmets could be significantly displaced on a wearer’s head with a mean static force of approximately 7 N in all 3 directions. This is much lower than the prescribed 50 N static force used in the test standard (Standards Australia/Standards New Zealand 2008). Because 98% of labeled helmets were approved to AS/NZS 2063, this suggests that helmets may not fit users as well as they do ISO headforms in certification and/or compliance tests. In practical terms, helmets can be displaced at relatively low forces. This finding may be attributed to the differences between the size or shape of cyclists’ heads and headforms used in testing or the differences between the levels of adjustment by cyclists and technicians involved in standards tests and interface differences; for example, hair. It was found that 1 in 3 cyclists (for which helmet size was known) was wearing a helmet too small or large according to the manufacturer’s sizing recommendation. Factors that affected correct helmet size selection included age and gender, although gender became nonsignificant when adjusted for age. This may reflect a gender bias in the sample but suggests that bicycle helmet size recommendations may not be suitable for a large proportion of the population, especially older cyclists. Helmet size selection did not, however, affect helmet stability in our tests. This relationship may arise because one head dimension—for example, length—is suitable for the helmet size, whereas the overall head circumference is incorrect for the helmet size. A broader range of head shapes was observed in the cyclist sample than is represented by the ISO headform range. Ethnic differences identified in the results agreed with research that found that East Asian heads are rounder than Western or Caucasian heads (Ball et al. 2010). Although head shape was found not to influence helmet size correctness or stability, it was found to affect helmet use, with 11.1% of cyclists in the brachycephalic range (P > .05) and 18.8% in the dolichocephalic range (P < .05) stating that they did not always wear a helmet while cycling, compared to only 7.3% of cyclists in the mesocephalic range. This might reflect the comfort levels of wearing a helmet that is better fitted to the contours of the head.

Downloaded by [New York University] at 03:26 08 June 2015

274 Almost half of all cyclists surveyed were wearing helmets that were incorrectly positioned or adjusted. Children and young adult cyclists were less likely to adjust their helmets correctly than were mature adults (35 years and older), as were female cyclists. Adjustment was found to be a significant determinant of helmet stability in our tests. Readjustment of poorly adjusted helmets was found to increase helmet stability for the majority of wearers. It was interesting to note that in only 20% of cases where the helmet was readjusted was stability improved in both the rearward and forward directions. This suggested that the adjustment and/or the helmet–head interaction may favor stability in one direction over the other. Results, revealing that the inclusion of a rear adjustment mechanism improved stability in the rearward and lateral directions but not the forward direction, reiterate this point. Findings of this study highlight the importance of providing consumer advice and instruction on proper helmet fit, position, and adjustment. Research by Plumridge et al. (1996) found that adequate advice on helmet fit was offered in less than half of retail outlets. The Australian standard AS/NZS 2063 (Standards Australia/Standards New Zealand 2008) requires that helmets be marked with instructions to the user to the effect “Fasten helmet securely under the jaw.” Packaging is required to include instructions on correct positioning, adjustment, and fastening of the helmet, including diagrams of correct and incorrect positioning. Many cyclists surveyed emphasized the difficulties and inconveniences of adjusting the straps of their helmets. Some cyclists noted that it was difficult to position the straps correctly under the ear by themselves without the aid of another person or a mirror. We recognized that parents were often unaware that children’s helmets were improperly worn or fitted and admitted to not reading the material included with the helmet. This might account for the high proportion of children in the sample with poorly adjusted helmets, despite the fact that the sample included only children who were supervised by a parent or guardian. This is of special note because young people are overrepresented among injured bicyclists (Sikic et al. 2009), so consumer advice and educational programs on proper helmet use should be targeted toward children and their parents. The findings highlight 4 main areas for improvement in the current bicycle helmet standards: • An “appropriate fit” on a headform may be inconsistent with an appropriate fit on all users with the same head circumference. It may be necessary to evaluate helmet fit separately from helmet stability, perhaps by requiring helmets to be positioned on differently shaped headforms. This may result in better fitting helmets for a larger range of users. • The specific protocols employed in the adjustment of helmets in current stability tests may not reflect the real world. This might lead to situations where the helmet meets the stability requirements but displays poorer stability in situ. A larger amount of slack prescribed in the adjustment for stability testing in the standard would better approximate the adjustment on wearers and give a more conservative measure of stability. This is in line with recommendations by Mihora et al. (2007).

Thai et al. • The technician should adjust the retention system optimally for all test directions but not readjust between tests to circumvent the design of retention system geometries that favor stability in one direction over the other. This might also encourage manufacturers to produce retention systems that have minimal potential for misuse. • The addition of a lateral stability requirement should be considered. McIntosh et al. (1998) found that primary impacts to bicycle helmets were most frequently to the anterior–lateral aspects and that there was a greater risk of head injury in this region. This study had several limitations that should be addressed in future research. Firstly, there were disproportionately low numbers of teenage cyclists in the 13–17 year age group surveyed. This was attributed to the nature of the locations at which cyclists were recruited, which may not be popular sites for teenage cyclists. Though this should not affect the sample greatly in terms of head size and shape data, because these are not significantly affected by age there may be significant behavioral factors in this age group that may have been overlooked. Secondly, it was not possible during the survey to measure the internal dimensions of the helmets worn. We have assumed that helmets complying with the Australian standard have been manufactured to fit ISO A, E, J, M, and O headforms used in testing, but this may not be the case. Consequently, the differences in head and headform sizes may not be a useful measure for comparison of stability. Future work should allow direct comparisons of helmet testing in situ and using test headforms. In this study we have defined head shape solely by the cephalic index, which is a simple and useful measure. We expect that other cranial dimensions, including the height of the head, will also affect how a helmet sits on the head and stability. Because we were concerned with how users were wearing helmets in the field, we were limited both by time and the availability of portable measuring tools. Three-dimensional scanning methods such as those used by Meunier et al. (2000) may more accurately determine the roles of fit and accommodation on helmet stability for cyclists. Finally, although the methods applied to measure helmet stability were derived from the static stability test specified in AS/NZS 2063 (Standards Australia/Standards New Zealand 2008), there are some differences. Those differences are required in order to make the tests acceptable to the participants; for example, comfortable and negligible risk, the surroundings—that is, in the field—and the need to perform the tests manually as opposed to using a test rig. A dynamic stability test may replicate real-world impact incidents; however, such a test may not be ethically acceptable due to the dynamic loads involved. Bicycle helmets worn by recreational and commuter cyclists are often the wrong size and are often positioned and adjusted incorrectly, especially in children and young people. This study found that helmet adjustment, rather than size, was a significant determinant of stability in the rearward, forward, and lateral directions. It was also found that current headforms used in standards testing may not be representative of cyclists’ head shapes, resulting in helmets that are inappropriate for many users.

Bicycle Helmet Size, Adjustment, and Stability Acknowledgments This work was conducted at the former School of Risk and Safety Sciences at UNSW. The authors thank the Centennial Park Trust, Canterbury Municipal Council, Sydney Olympic Park Authority and the park rangers, as well as Edgar Schilter and Hong Hua for their assistance in carrying out the surveys. We also acknowledge the advice of Dr. Jake Olivier.

Downloaded by [New York University] at 03:26 08 June 2015

Funding The study was funded through an Australian Research Council (ARC) Linkage Grant LP0669480 Pedal and Motor Cycle Helmet Performance Study. The project partners are the Australian Transport Safety Bureau (ATSB), Roads and Traffic Authority (NSW RTA), Transport Accident Commission (TAC Victoria), NRMA Motoring & Services, NRMA-ACT Road Safety Trust, and DVExperts International. This article does not represent the views of any of these organizations. The chief investigators are Adjunct Professor Andrew McIntosh, Associate Professor Paul McCrory, Dr. George Rechnitzer, and Professor Caroline Finch. Dr. McIntosh was the project leader. The Australian Centre for Research Into Injury in Sport and Its Prevention (ACRISP) is one of the International Research Centres for Prevention of Injury and Protection of Athlete Health supported by the International Olympic Committee (IOC).

References Andersson T, Larsson P, Sandberg U. Chin Strap Forces in Bicycle Helmets. Bor˚as, Sweden: Swedish National Testing and Research Institute; 1993. AS/NZS 2512.1. Definitions and headforms. In: Methods of Testing Protective Helmets. Sydney, New Zealand: Standards Australia/Standards New Zealand; 2009. AS/NZS 2512.7.1. Method 7.1 Determination of stability of protective helmets—static stability. In: Methods of Testing Protective Helmets. Sydney, New Zealand: Standards Australia/Standards New Zealand; 2006. Attewell RG, Glase K, McFadden M. Bicycle helmet efficacy: a metaanalysis. Accid Anal Prev. 2001;33:345–352. Ball R, Shu C, Xi P, Rioux M, Luximon Y, Molenbroek J. A comparison between Chinese and Caucasian head shapes. Appl Ergon. 2010;41(6):832–839. Chang LT, Chang CH, Chang GL. Fit effect of motorcycle helmet—a finite element modeling. JSME Int J Series A. 2001;44:185– 192. Ching RP, Thompson DC, Thompson RS, Thomas DJ, Chilcott WC, Rivara FP. Damage to bicycle helmets involved with crashes. Accid Anal Prev. 1997;29:555–562. Electronic Code of Federal Regulations, Title 16, Chapter II, Subchapter B, Part 1203—Safety Standard for Bicycle Helmets, US Government Printing Office.

275 EN 1078. Helmets for Pedal Cyclists and for Users of Skateboards and Roller Skates. Brussels, Belgium: European Committee for Standardization; 1997. Garson JG. The International Agreement on the Classification and Nomenclature of the Cephalic Index. J R Anthropol Inst Great Brit Ireland. 1887;16:17–20. Gilchrist A, Mills NJ, Khan T. Survey of head, helmet and headform sizes related to motorcycle helmet design. Ergonomics. 1988;31:1395–1412. Greulich WW, Day HG, Lachman SE, Wolfe JB, Shuttleworth FK. A Handbook of Methods for the Study of Adolescent Children. Vol 17. Washington, DC: Society for Research in Child Development; 1938. Grimard G, Nolan, T, Carlin JB. Head injuries in helmeted child bicyclists. Inj Prev. 1995;1:21–25. Hagel BE, Lee RS, Karkhaneh M, Voaklander D, Rowe BH. Factors associated with incorrect bicycle helmet use. Inj Prev. 2010;16(3):178–184. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 2000. ISO DIS 6220-1983. Headforms for Use in the Testing of Protective Helmets. International Standards Organization; 1983. JIS T 8134. Protective Helmets for Bicycle Users. Tokyo, Japan: Japanese Standards Association; 2007. Lee RS, Hagel BE, Karkhaneh M, Rowe BH. A systematic review of correct bicycle helmet use: how varying definitions and study quality influence the results. Inj Prev. 2009;15(2):125–131. McIntosh AS, Curtis K, Rankin T, et al. Associations between helmet use and brain injuries amongst injured pedal- and motor-cyclists: a case series analysis of trauma centre presentations. J Australasian College Road Safety. 2013;24(2):11–20. McIntosh AS, Dowdell B, Svensson N. Pedal cycle helmet effectiveness: a field study of pedal cycle accidents. Accid Anal Prev. 1998;30(2):161–168. McIntosh AS, Lai A, Schilter E. Bicycle helmets: head impact dynamics in helmeted and unhelmeted oblique impact tests. Traffic Inj Prev. 2013;14:501–508. Meunier P, Tack D, Ricci A, Bossi L, Angel H. Helmet accommodation analysis using 3D laser scanning. Appl Ergon. 2000;31:361–369. Mihora D, Jutchinson J, Friedman K, et al. Bicycle helmet retention system testing and evaluation. Int J Crash. 2007;12:211–215. Parkinson GW, Hike KE. Bicycle helmet assessment during well visits reveals severe shortcomings in condition and fit. Pediatrics. 2003;112:320–323. Plumridge E, McCool J, Chetwynd J, Langley JD. Purchasing a cycle helmet: are retailers providing adequate advice? Inj Prev. 1996;2:41–43. Rivara FP, Astley SJ, Clarren SK, Thompson DC, Thompson RS. Fit of bicycle safety helmets and risk of head injuries in children. Inj Prev. 1999;5:1994–1997. Safety standard for bicycle helmets: Final rule. 1998. Sikic M, Mikocka-Walus AA, Gabbe BJ, McDermott FT, Cameron PA. Bicycling injuries and mortality in Victoria, 2001–2006. Med J Aust. 2009;190:353–356. Snell B95. 1995 Standard for Protective Headgear for Use in Bicycling 1998 Revision. North Highlands, CA: Snell Memorial Foundation; 1998. Standards Australia/Standards New Zealand. Bicycle helmets. AS/NZS 2063. 2008. Available at: http://www.saiglobal.com/online/. Thompson DC, Rivara FP, Thompson RS. Helmets for preventing head and facial injuries in bicyclists. Cochrane Database Syst Rev. 1999;(4):CD001855. Thompson DC, Thompson RS, Kaufman A, Rivara FP. Increasing bicycle helmet effectiveness: the importance of proper helmet fit and wearing position. HMO Pract. 1997;11:25–26.