Association Between Conventional Bicycle Helmet Use and Facial Injuries After Bicycle Crashes

Key Points

Question  Do bicycle helmets provide sufficient protection to the face during bicycle crashes, and, if so, how much protection do they offer?

Findings  In this cohort study of 85 187 bicycle crashes from records in the US National Trauma Database, helmets were found to not sufficiently protect the face in bicycle crashes. Helmet protective capacity for the face lessens for the mid face and lower face.

Meaning  Bicyclists are at risk for experiencing facial fractures and soft-tissue injuries during crashes, particularly to the mid and lower face; public education regarding potential facial injury despite helmet use may be warranted.

Importance  Bicycling is an increasingly common activity in the United States that is often associated with fall injuries to the head and face. Although helmets lessen head injury, their role in reducing facial injuries is less clear; therefore, it is important to understand the protective capacity for the face in current helmet design.

Objective  To estimate the conventional bicycle helmet’s association with the rate of facial injury after bicycle crashes.

Design, Setting, and Participants  This retrospective cohort study accessed records from January 1, 2010, to December 31, 2014, from the National Trauma Databank, which collects data from emergency departments in US hospitals. Each record pertained to 1 emergency department admission for a bicycle crash. The National Trauma Databank registry data are collected and recorded by incident, which is equivalent to an injury-related hospital admission. All injuries involving patients aged 18 to 65 years for whom data on helmet use and injury were available were included. Statistical analysis was conducted from July 19 to October 17, 2016.

Exposures  Helmeted and nonhelmeted bicycle crashes.

Main Outcomes and Measures  Head and facial injuries among helmeted and nonhelmeted bicycle crashes.

Results  A total of 85 187 facial injuries met inclusion criteria (patient age 18-65 years, availability of helmet use status, and type of injury). Demographic information on bicycle riders was frequently unavailable. Among all injuries, fractures to the head (11.6% [9854]) and face (11.3% [9589]) occurred at similar rates. Helmets reduced head fractures by 52% (from 14.0% [7623] to 7.3% [2231]) and head soft-tissue injuries by 30% (from 15.0% [8151] to 10.9% [3358]), but had lower rates in protecting against facial injuries. While reducing facial injuries overall, the amount of protection with helmet use varied with facial location of the injury. Reduction in facial fractures was 35% (95% CI, 31%-39%) for upper face, 28% (95% CI, 23%-32%) for mid face, and 21% (95% CI, 15%-26%) for the lower face. Helmets were less protective against facial soft-tissue injuries, with a reduction of 33% (95% CI, 30%-36%) in the upper face, 21% (95% CI, 16%-26%) in the mid face, and 2% (95% CI, 0%-6%) in the lower face.

Conclusions and Relevance  Although bicycle helmets provide some protection against facial injuries after bicycle crashes, the level of protection depends on the proximity of the injury to the helmeted head. The lower face is particularly vulnerable to injury despite helmet use.

Bicycling is an activity commonly practiced throughout the United States, particularly among adults.¹ With bicycles used for recreation, exercise, and as a mode of transportation, the number of bicycle trips made in a single year is in the billions.² As the number of commuter bicyclists has grown because of worsening vehicle traffic and a push for ecofriendly transportation methods, the number of bicycle-related injuries has increased. Quiz Ref IDIn 2014, there were a total of 50 000 reported bicycle-associated injuries and 726 bicycle-associated fatalities in the United States. Injuries and deaths owing to bicycling crashes totaled over $6 billion in related costs in 2012.³

The helmet is the most commonly used protective device associated with bicycling.⁴ Although the current helmet design is effective in reducing head injuries, the face is less protected and remains vulnerable to injury during a bicycle crash. In a European study assessing bicycling injuries in inner city emergency departments, facial injuries occurred 30% of the time, with nonhelmeted bicyclists experiencing significantly more facial injuries than helmeted bicyclists (48% vs 26%).⁵

Facial injury can have severe consequences, particularly involving the loss of facial structural support, which can lead to disfigurement and impairment of vision, mastication, and speech.⁶ Protection from facial injury and its sequelae are an important public health concern. We therefore investigated the frequency with which such injuries occur and the extent to which helmet use might protect against them. We examined the incidence of bicycle-related facial injury, with a focus on facial fractures and soft-tissue injuries, in a large population cohort. Our objective was to estimate the conventional bicycle helmet’s association with the rate of facial injury. We hypothesized that conventional helmet use does not significantly reduce facial injury sustained in bicycle crashes.

This was a 5-year retrospective cohort study using records from the National Trauma Databank (NTDB), a registry that collects records from emergency departments and trauma centers throughout the United States. We accessed the NTDB through the American College of Surgeons on June 20, 2016, following receipt of approval from the University of California San Francisco Institutional Review Board.

Records collected included those from January 1, 2010, through December 31, 2014, that had a mechanism of injury pertaining to pedal bicycling (ie, NTDB e-code 826.1). Injuries occurring among those aged 18 through 65 years in which an injury to the head or neck occurred were included. Injuries in which the status of helmet use was unknown or not recorded were excluded. To determine whether injury severity differed between helmeted and nonhelmeted groups, Injury Severity Score (ISS) data were also collected.

The NTDB registry data are collected and recorded by incident, which is equivalent to an injury-related hospital admission; each incident may, however, comprise numerous injuries. We used the total number of incidents as our denominator and estimated the proportion of all incidents that contained specific types of injuries, either collectively grouped as any fracture of the face or regionally grouped as upper face fractures. Because the database was deidentified, we were unable to link multiple incidents that may have occurred to the same person, and therefore made the general assumption that all injuries were independent of all other injuries. Although this assumption may not be true, because of the relative nature of our comparisons (helmeted vs unhelmeted bicycle crashes), we believed that this approach minimized issues related to the database.

Demographic information on incidents was collected, but these data were frequently unavailable and are included for descriptive purposes only. Injuries were divided into regions of head and face. The face was further subdivided into upper face (frontal bone), mid face (nasal bones and naso-orbital-ethmoid complex, zygoma, and maxilla), and lower face (mandible). Injuries included and assessed were fractures, concussions, soft tissue, intracranial, ocular, and dental. Analysis was performed to compare each injury type separately, with comparison made between helmeted and nonhelmeted bicyclists, and in aggregate (eg, any fracture) or by region (eg, any upper face fracture).

Statistical analysis was conducted from July 19 to October 17, 2016. Frequencies and proportions were used to describe summary statistics for injury type, both in total and stratified by helmet status. Mantel-Haenszel odds ratios (ORs) were calculated and are presented with 95% CIs to compare the odds of sustaining an injury of a given type when helmet use was documented compared with the same injuries occurring in bicycle crashes during which no helmet was worn. The association between helmet use and ISS scores was also calculated and is presented with 95% CIs. Statistical analysis was performed using Stata, version 14 (StataCorp).

A total of 111 646 bicycle crashes fitting our criteria were reported in the NTDB during 5 years from 2010 to 2014. Of these, 85 187 crashes (76.3%) included information regarding type of injury and helmet use and were included in the study analysis. Demographic information of the bicyclists was available for approximately 60% of these crashes. Owing to low demographic reporting in the NTDB, it is difficult to draw conclusions, but it is noted that 5 times more crashes were ascribed to male bicyclists than to female bicyclists.

Quiz Ref IDOverall, fractures to the head (11.6% of all crashes [9854]) occurred at a similar rate to fractures of the face (11.3% of all incidents [9589]). Helmets reduced head and face injuries but provided greater protection against head fractures (52% reduction, from 14.0% to 7.3%; OR, 0.48 [95% CI, 0.46-0.51]) than facial fractures (32% reduction, from 12.6% to 8.8%; OR, 0.67 [95% CI, 0.64-0.70]) (Table 1). When injuries to the face were examined by facial region, this distinction became even more pronounced, with the helmet offering greater protection to regions of the face more proximal to the head (Table 2). Whereas helmets reduced upper face fractures by 35% (95% CI, 31%-39%), protection against mid face fractures fell to a 28% (95% CI, 23%-32%) reduction, while lower-face fractures were reduced by only 21% (95% CI, 15%-26%) (Figure 1).

Quiz Ref IDThis distinction became more apparent for soft-tissue injuries. Soft-tissue injuries not only occurred with more frequency in the face than head (17.5% vs 13.5% of all incidents, respectively), they were also less affected by helmet use. Whereas helmets reduced soft-tissue injuries of the head by 30% (95% CI, 27%-33%), this reduction for facial soft-tissue injuries overall was 23% (from 18.9% to 15.2%; OR, 0.77; 95% CI, 0.74-0.80). As was observed in the case of fractures, reduction in soft-tissue injuries with use of the helmet was dependent on the region of the face involved. Helmets reduced upper face soft-tissue injuries by 33% (95% CI, 30%-36%), but this protection decreased to 21% (95% CI, 16%-26%) for the mid face (Figure 2). Helmets provided no protection for soft-tissue injuries to the lower face. Only a 10% reduction in dental injuries was observed in helmet users (from 3.8% to 3.4% [OR, 0.90; 95% CI, 0.83-0.97]) (Table 2).

Table 3 reports that ISS data were similar between the helmeted and nonhelmeted incidences, with minimal to no association between helmet use and ISS. Differences between the groups did not cross severity classification.

This study analyzes the facial injury patterns from bicycle crashes reported to the NTDB throughout the United States over a 5-year period. Our study results show that, while conventional bicycle helmet use is protective for the head, its protective capacity for the face lessens for the mid face and lower face. Particular to the face, this decrease manifests as soft-tissue injuries, including lacerations, abrasions, and contusions, and skeletal injuries, such as fractures.

Similar studies have shown the lower face to be a region of high impact in bicycle injuries; however, findings in these studies differ. In Germany, a prospective study on head injury mechanisms in helmet-protected motorcyclists showed that the lower face experienced the highest percentage of injuries.⁷ Joseph et al⁸ studied 6267 patients with intracranial hemorrhages from the NTDB and reported a 13% reduction in mandibular fractures with helmet use, but this finding was not statistically significant (OR, 0.87; 95% CI, 0.70-1.05).⁸ In contrast, Stier et al⁹ studied 5350 bicycle crashes and reported a 4-fold risk of mandibular fractures with helmet use but did not observe significantly protective effects to the maxilla, orbit, nasal, or zygomatic bones. Given disparate findings in the literature, the present study aims to provide large cohort data.

Quiz Ref IDA Cochrane Review from 2000 is the most recent Cochrane Review on head and facial protection provided by bicycle helmets, stating, “Helmets reduce bicyclist-related head and facial injuries.”^10[p2] Specifically, the findings are based on a 700-patient review that showed a 65% injury reduction to the upper face and mid face with helmet use and no protection to the lower face.¹¹ In contrast to the Cochrane Review, our data showed the helmet to be less protective to the upper face and mid face, offering a 35% fracture risk reduction for the upper face and a 28% fracture risk reduction for the mid face. Our findings further describe the risk reduction to the lower face fracture by only 21%. The Cochrane Review advocated for development and use of bicycle helmets with chin protection. Since that time, facial protection in bicycle helmets has been developed, but its use and advocacy are limited.¹⁰ Powered by a larger cohort, our study results also continue to support additional facial protection for bicyclists. We hope that recognition of the limited protection of the bicycle helmet to the face may enable some bicycle helmet users to seek further facial protection.

Although head injuries can be devastating, facial injuries also lead to a host of serious complications, including facial deformity, loss of facial sensation, motor dysfunction of chewing and smiling, vision impairment, and financial and psychological stressors. Notably, disfiguring facial injuries cause anxiety and depression. Results showed that anxiety and depression scores were higher in patients with disfiguring injuries compared with those with nondisfiguring injuries.¹² There is also a medical system burden associated with facial injury. The mean length of hospital stay for a mandible fracture is 2.65 days, and the mean hospitalization cost is $35 804.¹³

Since the first commercially successful bicycle helmet, a polystyrene-lined hard shell, was released in 1975, most bicycle helmet designs have not added significant protection of the face.¹⁴ Currently, the Consumer Product Safety Commission determines bicycle helmet standards in the United States. The Code of Federal Regulations reports that the purpose of the published guidelines is to “reduce likelihood of serious injury and death to bicyclists resulting from impacts to the head.” The definition of a bicycle helmet reported in the code is “a device intended to provide protection from head injuries while riding a bicycle.”¹⁵ The regulations and standards for bicycle helmets that took effect on March 10, 1999, consider peripheral vision, helmet stability on the wearer’s head, and impact attenuation tests. Requirements do not include protection from facial injuries. However, as demonstrated in our study, facial injuries occur at rates similar to those of head injury, and the helmet’s protective ability diminishes for the mid and lower face.

A full-face helmet or a helmet with a chin bar may improve head and face protection for bicyclists, and such helmets are commercially available. However, these helmets may be used less frequently owing to potential wearer discomfort, isolation, hearing impairment, decreased peripheral vision, and decreased heat dissipation. To our knowledge, there are no published studies measuring the effectiveness of such facial protective helmet designs. Therefore, whether potential disadvantages of full-face helmets outweigh facial protective benefits remains to be studied, and further studies are needed.

In recent years, attention has been drawn to the long-term physical and neurologic implications of sports-related injury. With recreational and commuter bicycling gaining popularity, it is important to provide public education and awareness regarding this common activity. Educating bicyclists about the importance of the conventional bicycle helmet in providing head protection while describing vulnerability of the face to injury may be an important public safety initiative.

Quiz Ref IDSeveral limitations exist in our study. The NTDB currently does not offer information on types of helmets or how patients wore their helmets. Helmet design and fit are factors to consider because the effectiveness of the helmet is dependent on its ability to cover the designated body parts. For example, helmets that do not sit level on the head can leave other areas of the face exposed to injury in a crash. Helmet design and construction standards are not mandatory; therefore, a number of helmets do not meet these standards and will not offer sufficient injury protection during a crash. As a result, it is possible that lack of specific data on individuals who wore substandard helmets or did not wear their helmet correctly affected our results. Although we were unable to retrieve this information for our population cohort, it remains a topic for future studies.

The NTDB also does not collect data on an individual basis. Rather, data are collected per incident, or crash. Therefore, we were unable to calculate injury rates on a per-person basis but rather estimated injury rates as a percentage of all incidents. This lack of data affected helmeted and unhelmeted incidents equally, therefore making a relative comparison between the groups justifiable. Because multiple injuries were recorded per incident, more serious crashes would have contributed more weight to our estimates than minor crashes. However, it was our intention to highlight the overall number of incidents in which particular kinds of injuries were documented. Because a particular type of injury could be reported only once in an incident and helmet status would be the same for all injuries within an incident, we believed that this approach generated good estimates of helmet association with reduction of injury rates. Furthermore, bicyclist demographic data were not available for all bicycle injuries in the NTDB; therefore, we were unable to accurately compare injury rates between sexes or across age groups.

We used a large population cohort to examine the protective ability of the bicycle helmet. Although helmets provide protection to the head, they do not provide comparable protection to the exposed face. Bicyclists are at risk for facial fractures and soft-tissue injuries, particularly to the mid and lower face. Public education regarding potential facial injury despite helmet use may be warranted.

Accepted for Publication: October 2, 2018.

Corresponding Author: Rahul Seth, MD, Department of Otolaryngology–Head and Neck Surgery, University of California, 2233 Post Street, Third Floor, San Francisco, CA 94115 (rahul.seth@ucsf.edu).

Published Online: December 13, 2018. doi:10.1001/jamaoto.2018.3351

Author Contributions: Ms Benjamin and Dr Seth had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Knott, Seth.

Study concept and design: Benjamin.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Benjamin, Hills, Seth.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Benjamin, Hills, Seth.

Obtained funding: Seth.

Administrative, technical, or material support: Knott, Murr, Seth.

Supervision: Benjamin, Knott, Seth.

Conflict of Interest Disclosures: None reported.

Funding/Support: Funding for this project was provided through a monetary gift from a trauma patient and the University of California, San Francisco (UCSF) Department of Otolaryngology–Head and Neck Surgery. This project was supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through UCSF-Clinical and Translational Science Institute (CTSI) grant UL1 TR000004.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.

Meeting Presentation: The study was presented at the Triological Society; April 29, 2017; San Diego, California.

Additional Contributions: Jennifer Creasman, MSPH (UCSF CTSI), assisted with data extraction. Compensation was received.