Dear Editor

Hagel and Rowe reject(1) my criticisms(2) of their study of the impact of a child bicycle helmet law in Alberta, Canada(3). However they appear to have missed the point.

The first issue raised in my letter concerning reduced cycling as a result of Alberta's helmet law, Hagel and Rowe say there could have been confounders which would discount the drawing of a conclusion from the smaller proportion of children observed cycling. Actually I did not suggest the drawing of a conclusion, but rather commented that there was a red flag deserving of attention considering the reported negative impact of legislation on cycling that had occurred elsewhere(4). Similar or other confounders could equally apply to the authors' own conclusion that the proportion of children wearing helmets increased as a result of Alberta's law.

Hagel et al also argued that the increased proportion of children wearing helmets was due to legislation not education or enforcement, yet they admit there were educational programs in the Alberta area studied and there was some police enforcement. The authors supported their view by citing Maryland work(5) which concluded that legislation was more effective than education. In that case however the legislation was enforced and therefore is inconsistent with Hagel et al's conclusion. Other evidence exists which contradicts their conclusion also. A Canadian survey(6) found educational programs actually do account for increases in helmet use. I had cited the Ontario evidence which showed helmet use increases are not sustained by legislation alone anyway in my first letter.

There were other inconsistencies. Police enforcement in Alberta may have been more than the authors pointed out. They say 48 tickets were issued by the Edmonton Police. Were the authors aware that persons under the age of 16 cannot be charged in Canada? The tickets could only have been validly issued to a subset of the age group covered in the original study. It is probable that enforcement among the lower age group took another form, possibly stern police lecturing. In addition, the Edmonton Police Service does not and do not have jurisdiction over all of the areas covered, some of the latter are outside of City boundaries. It must be assumed that the authors did not contact the RCMP who police the remainder of the areas to obtain a count of tickets. These two factors suggest Hagel et al underestimated the degree of police enforcement.

Regarding the second issue concerning their errors in logic, Hagel and Rowe say that it was their intention to illustrate the degree of variation in helmet use depending on the helmet and age characteristics of companions. If that were the case they should have stated so in the report. Instead they and their colleagues chose to say, "considering that children riding with helmeted adults are almost 10 times more likely to be wearing a helmet than children riding with nonhelmeted child companions, policy makers should consider extending current children-only helmet legislation in Alberta and other locations." I stated that the relevant comparison was with non-helmeted adults and showed that children were only slightly more likely to wear a helmet when with adults. That information was available from the first Alberta observations yet the authors chose to use an irrelevant comparison from another jurisdiction, and used it to reinforce a weak case to extend a helmet law to adults.

References

1. Brent Hagel, Brian Rowe, Emergency Medicine, University of Alberta (5 September 2006), Re: Alberta helmet article - logic problem and missing data. Authors reply. http://ip.bmj.com/cgi/eletters/12/4/262 (accessed 16 December 2006)

2. Avery Burdett (5 September 2006), Alberta helmet article - logic problem and missing data http://ip.bmj.com/cgi/eletters/12/4/262 (accessed 16 December 2006)

3. Hagel BE, Rizkallah JW, Lamy A, Belton KL, Jhangri GS, Cherry N, et al. Bicycle helmet prevalence two years after the introduction of mandatory use legislation for under 18 year olds in Alberta, Canada. Inj Prev 2006;12(4):262-265

4. Robinson DL. No clear evidence from countries that have enforced the wearing of helmets. BMJ 2006;332:722-725.

5. Cote´ TR, Sacks JJ, Lambert-Huber DA, et al. Bicycle helmet use among Maryland Children: effect of legislation and education. Pediatrics 1992;89:1216–20.

6. Parkin PC, Spence LJ, Hu X, Kranz KE, Shortt LG and Wesson DE, Evaluation of a promotional strategy to increase bicycle helmet use by children. Pediatrics 1993; Vol 91, 772-777.

Dear Editor

According to good public policy, all laws with potentially detrimental effects (such as reduced cycling and reduced safety in numbers) should be evaluated. Far from being selective, my review examined every jurisdiction with a large increase in helmet wearing (more than 40 percentage points within a year). If helmet laws were beneficial, there should have been an obvious response. Yet there was no clear benefit to offset the expected harm from reduced cycling.[1]

Decisions concerning helmet legislation should, as Hagel and Rowe suggest, be based on the best available evidence. It would be naïve to assume that, even if voluntary wearing were beneficial, the same would be true for helmet laws. Case-control studies cannot determine the effects of reduced cycling, reduced safety in number or risk compensation. Consequently, the only reliable way to determine if helmet laws are beneficial is evaluate their effects on cycle use and head injury rates, in comparison to the cost buying millions of helmets or introducing other road safety measures.

It is surprising and disappointing that Hagel et al. chose not to collect the information needed to evaluate Alberta’s helmet law. Instead, they devoted 3.5 pages of Injury Prevention to the fact that, at 22 sites in Edmonton, helmet use of adults remained stable. In contrast, at these sites, in 2000, 29 children and 17 adolescents wore helmets, 37 children and 81 adolescents did not. In 2004, 13 children and 21 adolescents wore helmets; 0 children and 7 adolescents did not. Although the study was not designed to measure cycle use, this undeniably represents a significant reduction in the proportion of children and adolescents, compared to adults who were not affected by the legislation.

The onus should be on those who advocate laws taking away freedom to choose to prove those laws are beneficial. My systematic review found no evidence of benefit and probable harm from reduced cycling. Hagel et al. provide no evidence to contradict this. Most people would consider it more sensational to argue, as Hagel et al. do, that the helmet law should extended to adults than, with no evidence of benefit and some suggestion of harm, that it should be repealed.

References

1 Robinson DL. No clear evidence from countries that have enforced the wearing of helmets. BMJ 2006;332:722-725.

Dear Editor,

In response to our article, Burdett makes two main criticisms. The first relates to the issue of level of cycling activity in the community post-legislation. The second relates to our interpretation of the evidence for child cyclist helmet wearing when accompanying adults are helmeted compared with non-helmeted children. We consider these points separately.

On the issue of the level of cycling post-legislation, we commented on this in the text. The analysis of this complex issue cannot be solved by the rudimentary analysis provided by Burdett. For example, cycling activity conclusions require a larger overall sample, conducted over many more sampling sites, accounting for weather and road construction differences, and changing demographics within a region. In Alberta, large population influxes have changed the dynamics of both Calgary and Edmonton since the original survey. Despite these issues and the variations in methodology between the 2000 and 2004 surveys, Burdett concludes that “The proportion of cyclists affected by the law dropped from over 25% of those observed in 2000 to approximately 15% of those observed at the same sites in 2004.” Burdett goes on to call this a “disturbing indicator” that should have caught our attention. Clearly, this should not have caught our attention, since the analysis performed by Burdett cannot reasonably reach the conclusion that helmet laws are somehow responsible for decreased cycling activity. A control for time trends in the demographic distribution of the population studied would have been, at the very least, an essential element of any comprehensive assessment. The aim of our report was to examine the prevalence of helmet use and comment on the implications of changes in this outcome. A much more comprehensive and considered assessment of changes in levels of cycling would be required to evaluate the helmet legislation’s influence on cycling. Moreover, drawing the conclusion that this is somehow partly responsible for the rising trend of childhood obesity is, quite frankly, fear mongering.

On the issue of how much more likely children are to be wearing a helmet if accompanied by a helmeted adult vs. a non-helmeted child companion, Burdett argues that the relevant comparison is with non-helmeted adult companions. Our intention was to illustrate the degree of variation in helmet use depending on the helmet and age characteristics of companions. We do, however, agree that the comparison Burdett suggests is informative. Certainly an almost 2.5 fold increase in the likelihood of helmet use when a child is accompanied by a helmeted compared with a non-helmeted adult in the study by Khambalia et al is noteworthy.(1) Referring back to the original report by Nykolyshyn et al,(2) which presents the baseline data for the present study, the results also suggest that a significantly greater proportion of children wear helmets when accompanied by a helmet wearing adult, compared with a non-helmeted adult. We thank Burdett for this comment and the opportunity to point out the additional evidence that suggests if adults wear helmets, accompanying children are more likely to as well.

The issue of reduced cycling raised by Robinson has been addressed in our response to Burdett. In addition, this issue and the other issues she raises about consideration of head injury trends and the evidence on helmet effectiveness have been previously addressed elsewhere.(3-5) Most of the evidence cited by Robinson is not systematically collected and suffers from selection bias. Clearly, Robinson has focused on the fatality end of the injury spectrum and argues that helmets fail to protect riders because of the large kinetic energy of traumatic events, especially those involving motor vehicles. Despite this, well-accepted systematic reviews arrive at different conclusions. We suggest the readers draw their own conclusions. Finally, we take exception to the comment that “the significant reduction in children's cycling relative to adults is enough to suggest that the law should be repealed.” We firmly believe that decisions regarding helmet legislation should be based upon the best available evidence, a balanced assessment of the risks and harms, and avoidance of personal opinion and rhetoric.

Respectfully,

Brent Hagel, Brian Rowe

REFERENCES

1. Khambalia A, Macarthur C, Parkin PC. Peer and adult companion helmet use is associated with bicycle helmet use by children. Pediatrics 2005;116:939-42.

2. Nykolyshyn K, Petruk J, Wiebe N, Cheung M, Belton K, Rowe BH. The use of bicycle helmets in a western Canadian province without legislation. Canadian Journal of Public Health 2003;94:144-148.

3. Cummings P, Rivara FP, Thompson DC, Thompson RS. Misconceptions regarding case-control studies of bicycle helmets and head injury. Accident Analysis & Prevention 2006;38:636-643.

4. Hagel B, Macpherson A, Rivara FP, Pless B. Arguments against helmet legislation are flawed. British Medical Journal 2006;332:725-726.

5. Hagel BE, Pless IB. A critical examination of arguments against bicycle helmet use and legislation. Accident Analysis & Prevention 2006;38:277-278.

Dear Editor,

There is considerable debate about enforced helmet laws. Surveys in Australia counted several thousand cyclists before and after legislation, at the same sites, observation times and time of year. Percent helmet wearing (%HW) increased mainly because non-helmeted cyclists were discouraged from cycling – reductions in numbers counted were 2 to 15 times greater than the increases in numbers wearing helmets.[1] Despite the large increases in %HW, there was no obvious response in percent head injury, suggesting that helmet laws have very little benefit to counteract the drawback of discouraging this healthy exercise and environmentally-friendly transport.[2]

It is therefore surprising that Hagel et al. consider only %HW and try to draw useful conclusions from an outcome that everyone expected – all studies show enforced laws increase %HW.

Despite limitations, Alberta’s data strongly suggest that cycling was discouraged. Child cyclists fell from 10.3% to 4.8% of all cyclists, and teenagers from 15.4 to 10.3%, a very significant reduction in the proportions of children and adolescents (p=0.0015). Relative to numbers of adults counted, the data suggest that about 50% of children and a third of teenagers were discouraged from cycling.

Post-law, remarkably few children were counted – only 13 primary school children and 28 adolescents – at all 22 observation sites. With 3 observers per site, that equates to 0.2 children and 0.4 adolescents per observer per site. Even though pedestrians were also recorded, surely the survey could have been designed to at least count all cyclists? It would be unfortunate if this were the only information on cycle-use before and after Alberta’s helmet law. But if so, the significant reduction in children’s cycling relative to adults is enough to suggest that the law should be repealed.

Indeed, before helmet laws can be recommended, there should be evidence of substantial reductions in head injuries that clearly outweigh losses from discouraging cycling. Unfortunately, the authors provide no evidence of this. Preliminary reports stated that head injuries doubled from 5% to more than 10% of cyclist injuries.[3] Although this may have been due partly to changes in data coding, the increase in %HI was slightly greater for children than adults,[3] the opposite of what would be expected if helmets are beneficial, given that %HW of children increased, but %HW of adults did not.

Two other issues deserve comment. First, the authors state: “Considering that children riding with helmeted adults are almost 10 times more likely to be wearing a helmet than children riding with non-helmeted child companions, policy makers should consider extending current children-only helmet legislation in Alberta and other locations.” In fact, the pre-law survey in Alberta shows that 81% of children riding with non-helmeted adults wore helmets,[4] so in reality forcing adults to wear a helmet is unlikely to make much difference.

Second the authors claim that two systematic reviews found that helmets reduce fatal injuries by 73%. Only one review is cited, and that simply lists a crude odds ratio, based on 47 fatalities, with no adjustment for confounders.[5] Helmet wearers are more likely to obey traffic laws,[6] wear fluorescent clothing and use lights at night[7] and ride in playgrounds or bicycle paths than city streets.[8] All these factors affect the risk of bike/motor vehicle collisions, the main cause of cyclist fatalities. There is no way of knowing whether helmet wearers had fewer deaths relative to the number of minor injuries (but not necessarily population wearing rates) because of these factors, or because of helmets.

Recent research showed that, even with adjustment for confounders, case-control studies can produce incorrect and misleading results. For example, a systematic review of more than 30 studies of hormone replacement therapy (HRT) concluded that it reduced the risk of heart disease by 50%. We now know that HRT does not decrease and probably increases the risk of heart disease.[9] To avoid such problems, researchers should cite only reliable evidence, and discount odds ratios with no attempt to adjust for counfounders.

One reliable source of evidence is a detailed study of serious head injuries to cyclists in Brisbane. Based on the type and severity of injuries, it concluded that helmets would prevent very few fatalities. All deaths were caused by bike/motor vehicle collisions. For 13 of the 14 non-helmeted cyclists who died, there was no indication that a helmet would have made any difference. The authors were very concerned about brain damage from rotational injuries and recommended developing a test to measure sliding impact friction of helmets.[10]

Cyclist deaths were also investigated in Auckland. 16 of 19 non-helmeted cyclists died from multiple injuries, so helmets would not have changed the outcome. Only one cyclist died of head injuries in a bike-only crash, the most likely situation for a helmet to help. That cyclist died despite wearing a helmet. The authors concluded: "This study indicates that the compulsory wearing of suitable safety helmets by cyclists is unlikely to lead to a great reduction in fatal injuries, despite their enthusiastic advocacy."[11]

In the three years after helmets were made compulsory in New South Wales, Australia, 80% of fatally injured cyclists wore helmets, an almost identical proportion to population wearing rates (75% of children, 84% of adults), again suggesting that helmets are ineffective at preventing fatalities.[1]

Comparing Australia-wide fatalities in 1988 (before any helmet law) with 1994 (when all states had enforced laws and about 80% helmet wearing); cyclist, pedestrian and all road user deaths fell by 35%, 36% and 38% respectively; head-injury deaths fell by 30%, 38% and 42%. Thus the reductions for cyclists were less than for other road users. Factoring in the reduction in cycling, cyclists were probably at greater risk with compulsory helmet laws than without them.[12]

In inner London, 58% of cyclist fatalities were caused by collisions with heavy goods vehicles, as were 30% of those in outer London. It is implausible that a polystyrene helmet could be of significant benefit in such circumstances. The well-known tragic case of 4 helmeted cyclists killed by a car travelling at 50 miles/hr demonstrates that cyclists often die in impacts too severe for a helmet to help.[13]

Thus the most reliable evidence suggests that helmets prevent few, if any, fatalities. In Australia, implementation of other measures, such as random breath testing, speed cameras, and fixing up accident blackspots, resulted in large and immediate reductions in fatalities.[1]

In conclusion, the survey data for Alberta show a significant reduction in the proportions of children and teenagers, strongly suggesting the main effect of the law was to discourage cycling. Unless the authors can demonstrate a large response in percent head injury coinciding with the change in %HW that clearly outweighs the lost health and environmental benefits from reduced cycling, perhaps in future attention can be focussed on measures (such as those listed above) that have been shown to reduce injuries, instead of reducing cycling.

Dr Dorothy L Robinson

References

1 Robinson DL. Head injuries and bicycle helmet laws. Accid Anal Prevent 1996;28:463-475.

2 Robinson DL. No clear evidence from countries that have enforced the wearing of helmets. BMJ 2006;332:722-725.

3 Sands D. Helmet law stats called surprise. Calgary Sun 2003 Saturday, July 12, 2003.

4 Alberta Centre for Injury Control & Research. Bicycle Observational Study Summary. 2001.

5 Attewell R, Glase K, McFadden M. Bicycle helmet efficacy: a meta- analysis. Accid Anal Prev 2001;33:345–52.

6 Farris C, Spaite DW, Criss EA, Valenzuela TD, Meislin HW. Observational evaluation of compliance with traffic regulations among helmeted and nonhelmeted bicyclists. Ann Emerg Med 1997;29(5):625-9.

7 McGuire L, Smith N. Cycling safety: injury prevention in Oxford cyclists. Inj Prevent 2000;6(4):285-7.

8 DiGuisseppi CG, Rivara FP, Koepsell TD. Bicycle helmet use by children. Evaluation of a community-wide helmet campaign. JAMA 1989;262:2256-61.

9 Petitti D. Commentary: hormone replacement therapy and coronary heart disease: four lessons. Int J Epidemiol 2004;33(3):461-3.

10 Corner JP, Whitney CW, O'Rourke N, Morgan DE. Motorcycle and bicycle protective helmets: requirements resulting from a post crash study and experimental research. Federal Office of Road Safety, Report CR 55., 1987.

11 Sage M, Cairns F, Koelmeyer T, Smeeton W. Fatal injuries to bicycle riders in Auckland. N Z Med J. 1985;98:1073-4.

12 Curnow WJ. The Cochrane collaboration and bicycle helmets. Acc Anal Prevent 2005;37(3):569-73.

13 Gilbert K, McCarthy M. Deaths of cyclists in London 1985-92: the hazards of road traffic. BMJ 1994;308(6943):1534-7.