The effects of provincial bicycle helmet legislation on helmet use and bicycle ridership in Canada
- 1Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada
- 2Clinical Epidemiology Unit Ottawa Health Research Institute, Ottawa, Ontario, Canada
- 3Departments of Community Health Sciences and Internal Medicine, University of Manitoba, Winnipeg, Manitoba, Canada
- 4Department of Haematology and Medical Oncology, CancerCare Manitoba, Winnipeg, Manitoba, Canada
- Correspondence to Jessica Dennis, McLaughlin Centre for Population Health Risk Assessment, University of Ottawa, 1 Stewart Street, Ottawa, ON K1N 6N5, Canada;
Contributors JD and RZ were responsible for the conception and design of the study, TR provided statistical expertise, and BP assisted in data acquisition and provided analytical guidance. JD drafted the manuscript; all authors participated in the revision process and have approved this submission for publication.
- Accepted 19 January 2010
- Published Online First 29 June 2010
Background Bicycle helmet legislation has been variably implemented in six of 10 Canadian provinces. The objectives of this study were to determine the association between the comprehensiveness of helmet legislation and both helmet use and bicycle ridership.
Methods Analysis of helmet use was based on data from the 2005 Canadian Community Health Survey (CCHS) and included respondents from three Canadian provinces (Saskatchewan, Ontario, and Nova Scotia). Analysis of bicycle use was based on data from the 2000–01, 2003, 2005, and 2007 cycles of the CCHS and included respondents from all provinces. In the time between the 2000–01 and 2007 cycles, two provinces (Prince Edward Island (PEI) and Alberta) implemented helmet legislation.
Results Helmets were reportedly worn by 73.2% (95% CI 69.3% to 77.0%) of respondents in Nova Scotia, where legislation applies to all ages, by 40.6% (95% CI 39.2% to 42.0%) of respondents in Ontario, where legislation applies to those less than 18 years of age, and by 26.9% (95% CI 23.9% to 29.9%) of respondents in Saskatchewan, where no legislation exists. Though legislation applied to youth in both Ontario and Nova Scotia, helmet use was lower among youth in Ontario than among youth in Nova Scotia (46.7% (95% CI 44.1% to 49.4%) vs 77.5% (95% CI 70.9% to 84.1%)). Following the implementation of legislation in PEI and Alberta, recreational and commuting bicycle use remained unchanged among youth and adults.
Conclusions Canadian youth and adults are significantly more likely to wear helmets as the comprehensiveness of helmet legislation increases. Helmet legislation is not associated with changes in ridership.
- head protective devices
- head injuries
- health surveys
- head injury
Bicycling is among the top five physical activities practiced by Canadians1; however, benefits to physical health and to the environment must be considered in light of the risks of injury. Between 1994 and 2004, 44 577 hospitalisations occurred due to cycling incidents in Canada, representing 2% of all hospitalised injuries.2 Twenty-four per cent of these hospitalisations were due to head injuries, of which 61% occurred among those 5–19 years of age.
A meta-analysis of five case–control studies of cyclists seen in emergency departments found that helmets significantly reduced the risk of head, brain, and severe brain injuries by 63–88% among cyclists of all ages.3 Head injuries were defined as injuries to the scalp, skull, or brain, whereas brain injuries were defined as loss of consciousness or some other evidence of brain injury or dysfunction as a result of trauma. The reduction in risk was similar in crashes involving motor vehicles (69%) and in crashes by all other causes (68%). Bicycle helmets also reduce the risk of injuries to the upper and mid face.4 5
The implementation of bicycle helmet legislation has been associated with an increase in the prevalence of helmet use and a reduction in head injuries.6 7 In Canada, four of 10 provinces mandate helmet use for all ages and two mandate use for those less than 18 years of age (table 1). No territory-wide legislation exists in the three Canadian territories. Where legislation only applies to youth less than 18 years of age, compliance may be decreased when compared to all-ages legislation. Although this hypothesis has not been formally tested, a recent study found that 95% of children riding with a helmeted adult wore helmets, compared to 41% of children riding with a non-helmeted adult.8
A recurring argument used by those who oppose bicycle helmet legislation is that legislation might discourage bicycling.9–11 Conflicting results were reported by the two studies that investigated this previously.12 13 The objectives of this study were to examine the association between the comprehensiveness of helmet legislation in Canada and bicycle helmet use in youth and in adults, and to compare bicycle ridership before and after the implementation of helmet legislation.
Analysis for this study was based on data from the 2000–01, 2003, 2005, and 2007 cycles of the Canadian Community Health Survey (CCHS). The CCHS is a cross-sectional survey administered by Statistics Canada that collects information on health status, healthcare utilisation, and health determinants for Canadians 12 years of age and over in all provinces and territories. Individuals living on First Nation's/Aboriginal Reserves, in institutions, and in remote regions, and full-time members of the Canadian Forces are excluded. A complex survey design was used14 and households were the final sampling unit. One interviewee per household was selected based on their age and the household age distribution, except in 18% of households in the 2000–01 survey in which two interviewees were selected. Interviewers were required to obtain verbal permission of a parent or guardian to interview respondents between 12 and 15 years of age. Interviews were conducted either by phone or in person.
The optional module on the use of protective equipment during sporting activities was completed by respondents in Nova Scotia, Ontario, and Saskatchewan in 2005, representing the complete spectrum of helmet legislation types in Canada (table 1). Although this was also the case in the 2007 survey, information on helmet use was not available for all health units in each province and so analysis of helmet use was restricted to the 2005 survey. Information on bicycle ridership was available from all participants in all survey cycles. Sample sizes were 130 827 individuals in 2000–01, 134 072 individuals in 2003, and 132 947 individuals in 2005, with national response rates of 84.7%, 80.7%, and 78.9%, respectively. Beginning in 2007, data collection for the CCHS occurs on an ongoing basis, resulting in changes to the targeted sample size. The 2007 CCHS included 65 946 respondents, with a response rate of 77.6%. Respondents from the three Canadian territories (approximately 2% of respondents) were excluded from all analyses due to differential helmet legislation by municipality. To coincide with the cycling season, analysis of bicycle ridership was restricted to surveys completed between May and October.
Definition of variables
Respondents were asked, ‘(With respect to) leisure time activities… have you done any bicycling in the past 3 months?’ Those who answered in the affirmative were considered recreational bicyclists and were also asked, ‘In the past 3 months, how many times did you bicycle?’ Bicycle commuters were identified by the question, ‘In a typical week in the past 3 months, how many hours did you usually spend bicycling to work or to school or while doing errands?’ Those who responded more than ‘none’ were defined as commuters. Changes to the commuting question in the 2007 survey resulted in incomparable estimates which were therefore excluded from the analysis of commuting bicycle use.
Recreational and commuting bicyclists who participated in the module on the use of protective equipment were asked, ‘When riding a bicycle, how often do you wear a helmet?’ Those who responded ‘always’ or ‘most of the time’ were defined as bicycle helmet users while those who ‘rarely’ or ‘never’ wore a helmet were defined as non-users. The income adequacy ratio was the adjusted ratio of a respondent's total household income to the low income cut-off corresponding to their household and community size and provided, for each respondent, a relative measure of their household income to the household incomes of all other respondents.15 Urban areas were those continuously built-up having a population concentration of 1000 or more and a population density of 400 or more per square kilometre.15 Youth were defined as those less than 18 years of age.
Lifestyle and demographic factors potentially associated with bicycle helmet use were obtained from the published literature 7 16–19 and were investigated using univariable and multivariable logistic regression. Differences in the association between legislation and helmet use by factors associated with helmet use were assessed by two-way interactions between legislation and factors associated with helmet use in the multivariable model. Results of regression analyses were expressed as OR (95% CI). Observations with missing data for cyclist type (n=44) were omitted from the multivariable model. Logistic regression diagnostics available in SAS 9.1 were used to identify observations that influenced the regression results.
In the interval between 2001 and 2007, Alberta and Prince Edward Island (PEI) implemented helmet legislation. Prevalence estimates of recreational and commuting bicycle use in each survey cycle were calculated for respondents in PEI and Alberta and were compared to estimates of bicycle use over the same time period in the other Canadian provinces. Differences in bicycle use between survey cycles were assessed using the Student t test.
All analyses incorporated sampling weights to ensure that the estimates produced were representative of the Canadian population and not just the survey sample. Sampling weights were assigned by Statistics Canada to each person and correspond to the number of persons in the Canadian population that were represented by the respondent. To account for the complex sampling design, variance estimates were calculated using bootstrap weights and a SAS macro provided by Statistics Canada. All analyses were done with SAS 9.1 (SAS Institute).
Of the 55 465 respondents included in the module on the use of protective equipment, 40 326 (72.7%) had not bicycled in the past 3 months, 978 completed interviews by proxy, 1421 were non-respondents, and 215 resided in a territory. The study sample consisted of 12 525 eligible respondents, of which 1742, 10 044, and 739 were from Saskatchewan, Ontario, and Nova Scotia respectively. Youth represented 22.1% of respondents; 60% were male, 83.7% lived in urban areas, and 32.8% were bicycle commuters. Overall, 41.2% (39.9 to 42.5) of participants reported using helmets when riding a bicycle. Helmet use was least common in those aged 16–29 years, males, those with a low education and income, those living in rural areas, and non-commuter cyclists (figure 1).
Use of bicycle helmets was reported by 73.2% (69.3 to 77.0) of respondents in Nova Scotia, where legislation applies to all ages, by 40.6% (39.2 to 42.0) of respondents in Ontario, where legislation applies to those less than 18 years of age, and by 26.9% (23.9 to 29.9) of respondents in Saskatchewan, where no legislation exists. Among adults, the reported prevalence of helmet use was 38.9% (37.3 to 40.6) in Ontario compared to 25.1% (21.9 to 28.4) in Saskatchewan, even though legislation did not apply to adults in either province. Despite the presence of similar legislation applying to youth, the prevalence of helmet use was 77.5% (70.9 to 84.1) among youth in Nova Scotia and 46.7% (44.1 to 49.4) among youth in Ontario (figure 2). In terms of attributable risk percentage, 58% and 66% of helmet non-use among youth in Ontario and Saskatchewan, respectively, is attributable to non-comprehensive helmet legislation and, assuming a causal relationship between legislation and helmet use, could be eliminated if comprehensive legislation was introduced.
In multivariable models adjusted for age, sex, education, income, urban/rural dwelling, and cyclist type, when compared to no bicycle helmet legislation (Saskatchewan), residence in a province with legislation applicable to those less than 18 years of age (Ontario) was associated with an increased OR for bicycle helmet use of 1.8 (1.5 to 2.2). Residence in a province with all-ages legislation (Nova Scotia) was associated with an OR for bicycle helmet use of 9.0 (6.9 to 11.8) (figure 1). Removal of influential observations (n=56) did not significantly change the results, nor did adjusting for the frequency of bicycle use. No differences in the effect of legislation on helmet use by age (<18 vs ≥18) were observed (p=0.95), suggesting that adults and youth were equally likely to report increased helmet use as the comprehensiveness of legislation increased. However, the relationship between legislation and helmet use was significantly different in urban compared to rural settings (p<0.0001). The adjusted ORs for helmet use associated with youth legislation and with all-ages legislation compared to no legislation were 1.5 (1.2 to 1.8) and 7.7 (5.6 to 10.6) in urban areas. Comparable estimates in rural areas were 4.8 (3.2 to 7.1) and 21.9 (13.7 to 35.0).
In each survey cycle, less than 4.3% of responses were excluded from the analysis of recreational bicycle use due to non-response or proxy interview. In 2001, 73.2% (64.4 to 82.0) of youth in PEI used bicycles, compared to 66.4% (56.6 to 76.1) in 2003, the year all-ages legislation was introduced (figure 3A). The mean number of times youth cyclists used their bicycles was 38.6 (28.9 to 48.4) in 2001 and 28.9 (22.6 to 35.2) in 2003 (figure 3B). Although these differences were not statistically significant, they were suggestive of a downward trend that was nonetheless not sustained in subsequent surveys. Among adults in PEI, the prevalence of bicycle use was 17.6% (15.6 to 19.6) in 2001 vs 19.0% (16.0 to 22.0) in 2003 (figure 3C). The mean number of times cycled was 20.0 (17.3 to 22.8) in 2001 and 17.3 (13.4 to 21.2) in 2003 (figure 3D).
In Alberta, the prevalence of bicycle use in youth was highest in 2003, the year immediately after youth legislation was introduced, and then returned to pre-legislation levels (figure 3A). This trend was also evident in Alberta adults, suggesting that the trend was independent of changes in legislation. The mean frequency of bicycle use among youth cyclists was 30.1 (25.8 to 34.4) in 2001 and 28.1 (24.9 to 31.4) in 2003 (figure 3B). From 2005 to 2007, the frequency of bicycle use among youth in Alberta dropped significantly. However a significant decrease was also observed among youth in Manitoba, Saskatchewan, and British Columbia.
With regard to commuter cyclists, less than 4.4% of responses in each survey cycle were excluded due to non-response or proxy interview. In each survey cycle, between 76% and 83% of bicycle commuters also used their bicycles for recreational purposes. In both Alberta and PEI, the change in the reported prevalence of commuting bicycle use following the introduction of legislation was not statistically significant among youth or adults (figure 4).
In this study, we found that self-reported bicycle helmet use was greater in both youth and adults residing in provinces with increasing comprehensiveness of helmet legislation. When bicycle helmet legislation applied only to those less than 18 years of age (Ontario), reported helmet use among youth was less prevalent than if legislation applied to all ages (Nova Scotia), and the majority of helmet non-use in Ontario could be attributed to non-comprehensive legislation. Our results also demonstrate the importance of helmet legislation in rural areas, where helmet use is lower than in urban areas. The implementation of helmet legislation was not associated with significant changes in recreational or commuting bicycle use among adults or youth.
The effect of differential versus comprehensive legislation is remarkable and could be considered when drafting future legislation. In our study, we found significantly reduced self-reported helmet use among Ontario youth where legislation applies to those less than 18 years of age compared to youth in Nova Scotia where legislation applies to all ages. Likewise, we found that adult helmet use was higher in Ontario compared to Saskatchewan (no legislation). Though speculative, reduced helmet use in youth may be due to enforcement difficulties associated with distinguishing teenagers younger than 18 years of age and due to a lack of positive role-modelling by adults. Increased reported helmet use by adults in a province with only youth-specific legislation may reflect greater awareness of bicycle safety.
Mandatory helmet use may discourage bicycle ridership due to acquisition costs or due to notions among youth or adults that helmets are annoying, uncomfortable, ugly, hot, or unfashionable.20 21 A 6-year observational study of cyclists 5–14 years of age in Toronto, Ontario, found that yearly variation in cycling rates was not related to the introduction of provincial legislation applying to those less than 18 years of age.13 In contrast, the number of cyclists less than 16 years of age observed in New South Wales, Australia, decreased by 36% and 44% in the first and second years following implementation of state-wide child helmet legislation.12 In Melbourne, Australia, a similar decrease in the number of all cyclists was observed in the 2 years following the implementation of state-wide all-ages legislation.12 Our analysis builds on these previous studies by including a concurrent comparison group and by expanding the Canadian study population to include adults and youth across Canada. We did not find a significant reduction in bicycle use among youth or adults following the implementation of legislation and even found an increase in bicycle use among Alberta youth in the year immediately following the introduction of legislation.
One important limitation of our study is the potential for ecological confounding. Differences in helmet use cannot be solely attributed to differences in provincial helmet legislation and are likely influenced by variation in the enforcement of the helmet law, incentives for helmet use, educational campaigns, and attitudes towards bicycle use and helmets. It should also be noted that the Canadian provinces differ in climate and terrain, which affects bicycle use. For example, British Columbia has a coastal climate and bicycling is feasible for most months of the year, whereas Saskatchewan has a harsher and colder climate. Nonetheless, our conclusions on the effects of legislation on bicycle ridership are based on changes within the same province and are therefore less likely to be biased by differences in unmeasured confounders between provinces.
Because this study relies on self-reported data, it is also vulnerable to misclassification bias. Helmet use may have been over-reported to a greater extent where legislation was more widespread. Though possible, the published literature suggests that self-reported helmet use overestimates actual helmet use to the same extent under different types of helmet legislation.22 Furthermore, the prevalence of helmet use among youth in Ontario in the present study is similar to that found by other authors using direct observation.19 If over-reporting of helmet use where legislation was more widespread was a strong source of bias, youth in Ontario would have been expected to report helmet use at levels similar to youth in Nova Scotia. Retrospective self-reporting of bicycle use may have resulted in inaccurate estimates of bicycle ridership. However, since identical survey methods were used in each province and survey cycle, comparisons of bicycle ridership remain valid.
Public debate concerning bicycle helmet legislation in Canada is passionate and ongoing. On 9 June 2009, a member of the Legislative Assembly of Manitoba introduced, for a third time, a private member's bill mandating bicycle helmet use for all persons in the province of Manitoba. The member argued that helmets reduced the risk of injuries and that legislation was ‘equally important’ as the education and free helmet programmes that the province has in place.23 An opponent of the bill claimed that legislation was ‘an infringement on our inherent freedoms’ and ‘a misuse of police resources’. The assembly voted against the bill. Similarly, on 16 April 2007, the Saskatoon, Saskatchewan city council voted against a bylaw mandating helmet use for all bicyclists, citing the potential for legislation to discourage bicycle use as justification.24 Data from the present study does not support the potential for legislation to discourage bicycle ridership among youth or adults; however, issues of rights and freedoms or the impact of legislation on police resources could not be addressed.
In conclusion, we found that youth were most likely to wear helmets when residing in a province where helmet legislation applied to all age groups. Furthermore, the presence of youth legislation increased the likelihood of helmet use among adults. Implementation of provincial helmet legislation was not associated with a significant decrease in recreational or commuting bicycle use among adults or youth. These results are timely and applicable to current and proposed public health policies concerning helmet legislation. If the ultimate population health goal is to reduce the risk of cycling head injury in youth and adults, all-ages helmet legislation appears most appropriate and does not significantly impact bicycle ridership.
What is already known on the subject
Bicycle helmets reduce the risk of head, brain, and facial injuries among cyclists of all ages.
Implementation of bicycle helmet legislation is associated with increased helmet use; where legislation only applies to youth less than 18 years of age, compliance may be decreased when compared to all-ages legislation.
Opponents of bicycle helmet legislation claim that legislation reduces ridership.
What this study adds
Self-reported bicycle helmet use was greater in both youth and adults residing in provinces with increasing comprehensiveness of helmet legislation.
Canadian youth were less likely to wear helmets when provincial legislation was differentially applied to those less than 18 years of age.
Implementation of helmet legislation was not associated with changes in recreational or commuting bicycle use among youth or adults.
We would like to thank Jean-Michel Billette at the Carleton, Ottawa, Outaouais local Research Data Centre (COOL RDC) for his statistical guidance and expertise.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.