Authors Lee and Mann argue for law compelling use of cycle helmets by children to prevent road deaths and serious injuries.¹ This observer is surprised that the peer reviewers allowed publication of material lacking evidence either that the actual risks faced by child cyclists justify compulsion, or that the real world results of helmet compulsion in other countries justify compulsion in this country. These shortcomings are typical of papers in the medical literature that attempt to address the issue of cyclist safety. I believe that these chronic shortcomings are primarily the consequence of the failure of the peer review process.

In the first place, it is irrational that consideration of helmet laws for children is restricted only to cycling, or even begins with cycling. Although, tragically, around 30 child cyclists have been killed on public roads annually in recent years, typically 110 child pedestrians are killed annually.² Estimates of death risk per kilometre travelled derived from standard data sources³ do not suggest that child cyclists face greater risks than child pedestrians in most age groups. It is in any case evident that the average child is almost four times more likely to become a serious casualty while walking rather than cycling. The peer reviewers ought to have insisted on a more general discussion of the risks faced by children in transport. This would have placed the injuries to cyclists in context and enabled priority, surely the basis of any systematic approach to public health interventions.

In the second place the evidence for the effectiveness of cycle helmets is split by an interesting contradiction. The authors cite research based on case-control trials reporting that helmeted cyclists were much less prone to serious head injuries than the bareheaded, at least at the time and in the locality of the research work. However, there is also a substantial body of evidence based on population-level studies of head injuries with increasing helmet use. These studies consistently fail to show material benefit for cyclist populations that took up helmet wearing. This was even true in New Zealand, where cyclists responded willingly to helmet promotion, with voluntary use reaching 60% even before the well obeyed law of January 1994 came into force.⁴ The famous helmet laws for the states of Australia brought into effect during 1990–94 drew a similar null result on close analysis.⁵ In the United States, population-level data gathered by the Consumer Product Safety Commission (a US government organisation that strongly promotes helmet use) shows that the risk of head injury per US cyclist increased by 40% during the 1990s, while helmet use increased from under 20% to at least 50% of cyclists.⁶ The omission of such evidence places a serious question mark against the competence of the peer reviewers in this case.

The hiatus between clinical trials and population-level results is of scientific interest and draws the curiosity of inquiring minds. Ignoring the hiatus smothers the existence of a mystery. This is unscientific.

It must be added by-the-by that studies of reported casualties in Britain have revealed a disturbing tendency towards increasing severity of injury with increasing helmet use. This has been observed at the national level⁷ and for London,⁸ where helmet use grew much earlier than the national average. Edinburgh has been identified as having the highest level of helmet use in the country. An ongoing analysis of reported casualties by the author has revealed increasingly severe injuries after 1995, especially for child cyclists. These increases cannot be accounted for by worsening road conditions, since this would have been revealed in pedestrian injury trends. It is not absolutely clear whether the effect is coincidence or consequence, but fair peer review ought to have insisted on some commentary on whether helmet use has influenced reported road casualties. The objective is, after all, to reduce deaths and serious injuries in road crashes.

That helmet use has failed to improve reported road casualties is not surprising. A cycle helmet is designed to meet the event of a simple fall at speeds below 12 mph. Such a mild crash is unlikely to incur serious injury when road riding. Safety campaigners are pressing helmets to an application for which they were not intended. The ethics of this are questionable, a point peer reviewers should have highlighted. The use of helmets is more relevant off-road or “at play”, stunt riding on BMX or MTB type machines. The use of helmets in such situations is perhaps to be encouraged, although parental supervision should come first. On the other hand, these comparatively high risk activities are the consequence of children being denied the freedom to cycle for transport. Riding sensible road bikes on public streets either is, or ought to be, a safe mode of travel for children, not rationally to be distinguished from walking.

In summary, the peer review process has failed to stop incomplete evidence being presented as reliable knowledge. The readership may in consequence be led into two levels of misconception:

• The factoid that child cyclists are more at risk from motor traffic than child pedestrians

• The factoid that cycle helmets can protect children in road traffic accidents. A famous line from Schiller comes to mind as apt to the occasion:

Nur die Fuelle fuehrt zur Klarheit. [Only wholeness leads to clarity]

Let us hope, for the sake of the public understanding of cycling, that in future peer reviewers apply this wisdom. There will be resurgence of children walking and cycling only when the perceived danger from motor traffic in urban areas is addressed. Proposing compulsory use of inappropriate safety equipment evades this simple truth. Public health interventions must focus on the source of the perceived danger, not burden the innocent with the consequences of adult licentiousness.