Making Cycling Safer – What Does the Research tell us?


Cities across the globe have begun to embrace the bicycle, making sustainable transportation easier than ever for their residents. But with more people on bicycles, a new priority will begin to emerge – how can we make cycling as safe as possible? Despite numerous pieces of research on the subject with often-contradictory conclusions, there are areas of overlap in the safe cycling debate.

In Cycling and health – what’s the evidence?, Cavill and Davis say that ‘evidence exists for the role of physical activity of at least moderate intensity in promoting health and preventing disease’, citing ‘many different types of research evidence to back up the assertion that cycling has an impact on health’. A second report involving Cavill and Davis also suggests cycling ‘addresses the trend to obesity and consequent disease’. Robinson’s message is similar in Head injuries and helmet laws in Australia and New Zealand, stating those who cycle less experience ‘reduced health and fitness’. Sustrans go into more detail in Enabling good health for all, listing ‘building and maintaining healthy bones, muscles and joints, and improving mental health and self-esteem’ as benefits of ‘regular physical activity’ like ‘cycling and walking’. In addition to this, social benefits are mentioned, stating ‘people on foot or bikes interact more with their community’. The Bicycle Helmet Research Foundation (BHRF) summarise by saying ‘the bottom line is that people who cycle regularly live longer, on average, than people who do not, with healthier lives and less illness’ in Cycle helmets – an overview.

Cavill and Davis suggest that ‘30 minutes a day of moderate intensity physical activity on five or more days of the week’ is necessary for adults to feel health benefits from physical activities, with ’45-60 minutes’ daily for the prevention of obesity. Robinson’s mention of reduced health for those who cycle less does not suggest the required frequency for improved health. Despite talk of ‘helping people make healthy choices’, Sustrans neglect to place a timeframe on the amount of exercise required to feel their listed benefits. Similarly, the BHRF do not mention the amount of exercise required to ‘live longer, on average’, although this is likely due to their primary focus being cycle helmet functionality.

Not all agree that cyclists are likely to be healthier, however. A piece in the British Medical Journal claims is it ‘unlikely that most leisure cyclists, adults or children, are cycling for 45 minutes six days a week’, suggesting exaggerated health benefits. A personal communication with Doyle Baker – ‘an expert in the epidemiology of health and fitness’ – is cited, who states ‘there is still open debate regarding intensity and type of physical activity required to achieve most favourable health changes’. If Baker is indeed an expert in his field, then this qualitative research is valuable. However, considering the supposed open nature of this debate, one discussion with a relevant professional is not enough to discount numerous other reports with differing conclusions.

In addition to the generally agreed health benefits of cycling, there is support for cycle helmets across studies. In The potential for cycle helmets to prevent injury wearing a helmets is said to result in a ‘reduction of impact force’, with a conclusion stating that ‘there are no dis-benefits to helmet wearing’. The BHRF agree that ‘cycle helmets may produce benefit by reducing and spreading the forces that lead to direct injuries’ but add that ‘cycle helmets might make some injuries worse by converting direct forces into rotational ones’. The European Cyclists’ Federation (ECF) claim that while ‘it is true that helmets can save lives… …their potential should not be overestimated’. Franklin cites numerous authors claiming significant reductions in head injuries resulting from helmet-use. He does, however, acknowledge that these are ‘pro-helmet’ examples of research, which reduces their impartiality, and therefore, their reliability.

The main area of difference is mandatory cycle helmet laws. In Three years of mandatory cycle helmet laws in Spain, Merallo Grande suggests that mandatory helmet laws were introduced to the region in order to ‘improve the rates of cycling accidents’, but with cited data showing an 8.93% increase in accidents in the first year of that change, he asks:

Could it be that the effect of risk compensation, the euphoria of feeling more protected, has led those users to commit more illegal maneuvers and run more risks?

This statement raises a valid question, which Merallo Grande unfortunately chooses not answer, offering no data in support of the statement. He concludes that ‘mandatory helmet law is not reducing the rates of dead and injured’, suggesting the opposite is being achieved:

It could be driving people away from cycling, thus contributing to the health problems brought about by sedentary lifestyles.

Whilst Merallo Grande’s opinion is not unique, one cannot be confident in the quality of his work. Citing an absence of data, he acknowledges his quoted accident statistics ‘cannot be totally trustworthy’, suggesting this ‘hinders the understanding of the phenomenon’. These acknowledgements challenge the reliability of his research.

The ECF also takes an opposing stance to mandatory cycle helmet laws, suggesting such a framework would require ‘constant police enforcement’. They expand on this by predicting:

Either the police would divert their attention from dealing with the serious problems caused by motorised traffic in order to catch cyclists, or the authorities would have to accept the fact that this law would be systematically broken, and therefore useless.

Either of these outcomes could be true, but they are nothing more than predictions and cannot be taken overly seriously. However, a study exploring risk compensation theory in the British Medical Journal states that despite such laws, ‘helmet use has not yet become widespread in the Spanish population of cyclists’, adding more credibility to the ECF’s predictions. A more realistic objection of the ECF’s is that mandatory helmet laws ‘would lead to a significant decrease in cycling’, citing resistance from the British Medical Association in supporting such a motion, and referring to the effect of mandatory cycle helmet laws in Australia.

Cavill and Davis acknowledge the benefits of cycle helmets, but oppose enforced use, suggesting this presents ‘cycling as an especially risky mode of travel’. Merallo Grande’s opinion is mirrored, stating that mandatory helmet laws result in a decline in cycling. This is supported by reference to New Zealand’s legal situation, where cycling figures ‘collapsed by 55% following years of helmet promotion and enforced legislation’.

Rarely explicitly stating their opinion on the subject, the BHRF make their stance clear with figures from New Zealand and two Canadian states where cycle helmet use is enforced that show ‘the head injury rate did not decrease more than for the population at large’. Using data from a large sample – the entire population of these areas – gives these statements credibility. However, the fact that this document is presented as an ‘information sheet’ does make one wonder about its reliability as a piece of research.  The BHRF also express a lack of support for mandatory cycle helmet laws by implying that some areas with such legal provisions do not enforce them. They claim ‘unenforced laws can erode public respect for the rule of law generally, and traffic laws in particular’, although by not stating which areas they refer to or providing any relevant data, the argument fails to be proven.

In their commentary on A case study of the effectiveness of bicycle safety helmets – a second report by the BHRF – the improved safety of cycle helmets is acknowledged, but mandatory requirements are challenged:

Even if helmets are effective, it does not follow that all cyclists should wear them. Racing car drivers wear helmets, but not people driving to work. The difference is the level of risk.

Robinson takes a similar approach, referring to data examining head injury rates from Australia and New Zealand before and after mandatory helmet laws were introduced. Citing large data sets from a range of research pieces, Robinson explores the rate of head injuries alongside helmet use and general head injury trends. She concludes that there is ‘no major change in the percentage of head injuries, over and above the general trends’, instead calling ‘new diagnostic techniques (e.g. CAT scans), changes in admission policies, and safer roads’ the main drivers in head injury reduction. Building her opinions around different large-scale studies with similar conclusions, Robinson presents a convincing and reliable argument.

However, there is support for cycle helmet legislation. Writing for the British Medical Journal in 1992, Illingworth claims ‘the argument for helmets is overwhelming’. Citing the deaths of 296 cyclists in the UK in 1985, 1200 bicycle-related deaths in America annually, and 37 serious injuries reported in New South Wales (no date given), Illingworth attempts to convince that increased helmet use is necessary, but uses poor methods in doing so. Calling bicycle accidents ‘potentially dangerous’ is not a strong argument (after all, life is potentially dangerous), neither is citing a small sample group as evidence. Worse still, using vaguely worded personal communications to endorse the introduction of nationwide legislation changes does not make a convincing argument:

At least 30 children in Sheffield were almost certainly saved from serious injury through wearing helmets.

Illingworth’s methods are too subjective and cannot be considered reliable evidence. Concluding with a quote from Richard Ballantyne, she attempts to solidify her argument but instead reduces it to an overly emotional opinion piece:

Wear a helmet. It’s inconvenient but so is not being able to think or talk because your head has been pounded to jelly.

Cited by Illingworth is a well-cited study by Thompson, Rivara and Thompson which aims to ‘examine the protective effectiveness of bicycle helmets’. Originally published in 1989 and finding ‘that helmets reduce head injuries by 85% and brain injuries by 88%’, this piece has become the cornerstone work for mandatory cycle helmet law supporters, despite the fact that its follow-up report in 1996 was unable to match these figures.

Based on a study of 3390 injured cyclists in Seattle, the report monitored differences between those wearing helmets and those not, sending questionnaires and interviewing those involved in accidents to obtain detailed information about accident circumstances. Study groups have been categorised, and injuries separated by severity. The updated 1996 report quoted 69% reduction in head injuries and 65% reduction in brain injuries – reduced, but still suggesting significant improvements in cyclist safety as a result of cycle helmet use. The authors concluded that encouraging an increase in cycle helmet use ‘by a combination of legislative and educational approaches’  should be ‘widely implemented’.

The BHRF notes in their commentary on this publication that ‘very few [papers] have been based upon primary research into helmet effectiveness’, citing Thompson, Rivara and Thompson as an exception to this. However, the fact that their follow-up report was unable to match the widely-cited figures of their 1989 report leaves questions over the reliability of their methods.

Franklin calls their choice of sample groups unrepresentative of bicycle users in Seattle, also noting that of all research, only Thompson, Rivara and Thompson are ‘unequivocal that helmets work for everyone, everywhere, at any time’. Despite criticism, this has become a widely cited piece of work, largely down to the fact that no paper has matched the quoted figures, despite over twenty years passing since publication.

Addressing these criticisms in the British Medical Journal in 2000, Thompson, Rivara and Thompson refer to reports from the United States, New Zealand and Australia, where:

A consistent year to year trend in which the proportion of head injuries related to trauma from bicycles became lower in each successive year.

Crediting ‘an increased use of helmets occurring as a result of educational and legislative initiatives’, it is reasoned that this confirms the logic of their 1989 study, despite its age.

In Cyclists should wear helmets, Pless and Davis criticise those who object to mandatory cycle helmet laws on the grounds that they reduce the number of cyclists as being too speculative. Suggesting this view is ‘not established by sources’, they reason that it is:

Unlikely that anyone would choose to abandon [cycling] forever rather than wear a helmet or to sustain a blow to the head without a helmet.

A claim which, rather strangely, is not established by any sources. In addition to this dismissal, Pless and Davis argue that any reduction in cyclists as a result of mandatory cycle helmet laws is irrelevant:

Precisely how such a law works matters little if the public health issue is to reduce head injuries.

It is impossible to argue with this logic. If the single aim of mandatory cycle helmet laws is to reduce head injuries, they have been successful. However, the more likely public health issue is public health. Whilst there is debate over how much exercise is required to improve health, it is well-established that exercise is beneficial to health. Considering this fact, and the inevitable health benefits of encouraging cycling, the ECF claims that:

If just 10% of the population were to stop cycling – as is likely to happen if a mandatory helmet law were introduced – then overall health would suffer.

Robinson alone presents data illustrating a larger reduction than this across Australia and New Zealand, where mandatory cycle helmet laws have been present since 1990. If the logic of ECF’s argument were true, increased introduction of mandatory cycle helmet laws would have a negative effect on overall health.

Whether through an increase in cycle helmet use or other methods, a desire for improved cyclist safety is this debate’s uniting point. The main issues are a lack of clarity over what generates improved safety and insufficient reliable data to draw a conclusion. Understanding risk compensation theory has been stunted due to ‘a shortage of reliable data’, and many studies on the benefits of cycle helmets are considered questionable. Merallo Grande successfully concludes that this lack of reliable data ‘hinders the understanding of the phenomenon and the discovery of possible solutions’.

This is an area future studies should consider. The various factors that contribute to cycling accidents need to be addressed, as do potential educational, legal and infrastructural changes that can improve cyclist safety. It can be hoped that future research will bring more satisfactory and practical assistance to cyclists by considering not only the role of cycle helmets, but also the broader changes that can be introduced to improve cyclist safety.

Image courtesy of twicepix on flickr

  • Ericonabike

    Excellent read. One point – could the Boris Bikes represent the best possible safeguard against compulsory helmet laws? Thier business moel depends on casual use, which in turn effectively rules out using a helmet – no casual user would cart one around ‘just in case’ and the logistics of renting one out along with the bikes would be horrendous. Which is why I think criticism of the scheme is misplaced – whatever its shortcomings, the amount of money invrested in it should rule out compulsory helmet use in the UK [apart from Jersey and potentially NI of course…]

  • http://twitter.com/vebah Erik Weber

    A good piece Joe, but you leave out an important facet to the argument against mandatory helmet laws: whether we are concerned about number of crashes and injuries or whether we are worried about the *rate* of them.

    Research has shown in on several occasions that as cycling rates fell in the second half of the 20th century, bicycle crash and injury *rates* increased at the same time. At the same time, some anecdotal evidence has been seen in numerous cities that as biking and walking has seen a renaissance, while the number of crashes and injuries may have increased, the *rate* of injuries per population, or per distance traveled has dropped significantly.

    With that in mind, mandatory helmet laws can be argued to have actually the opposite effect than they are intended to have. While being required to wear a helmet may protect a particular individual from being more severely injured in an crash, the law’s effect of decreasing the number of cyclists actually increases that particular individual’s chances of being in an crash.

    Cities should be doing things that actively decrease the crash rate, not the number of injuries. As a cyclist, I’d much rather have an extremely low chance of being in an crash than a moderate chance of being in a crash but in which I would have a slightly better chance of surviving.

    We wrote about this on TheCityFix recently: http://thecityfix.com/qa-with-mikael-colville-andersen-the-controversy-over-bike-helmets/

    Also, I’d caution against the use of “accident,” which perpetuates an idea that there is no fault in road crashes.

    • http://dave.kinkead.com.au Dave Kinkead

      Spot on Eric.

      The claims by Pless and Davis are especially telling of how many medical professionals fail to understand that risk is a function of exposure.

      “Precisely how such a law works matters little if the public health issue is to reduce head injuries.”

      In arguing that the fact that reductions in cyclist head injuries have been caused by a reduction in cycling is irrelevant, they fail to see that the logical extension of their reasoning is to ban cycling all together. No cycling = no cycling injuries. Yet you don’t need to be a medical research to see that limiting harm by banning an activity doesn’t make the activity any safer.

      The safety improvements achieved by reducing the likelihood of being hit by a car are much greater than can ever be achieved by helmet promotion and legislation.


  • http://twitter.com/vebah Erik Weber

    Of course, then I went and used accident myself! (now edited) Bad habits are hard to break!

  • http://twitter.com/bassjunkieuk Mark Skrzypczyk

    Very well put together article, from my personal experience of around 5 years of London cycle commuting I’ve so far had 1 “serious” incident and 4 or 5 minor incidents. The only times I hit my head are in the serious incident where I hit a metal pole with the side of my head and torso and one of the minor crashes where I wiped out on a sharp corner after hitting what appeared to be an oil patch and banged my chin. My helmet “possibly” saved me from a concussion/fractured skull on the impact with the metal pole but I can’t say this for certain as I had a mass of grazes and cuts on the same side of my neck/chin. As it’s not a full face helmet it made no difference to the minor wipe-out.

    In all the other minor incidents it has been other limbs which came off worse, picking up some nice looking road rash and at least one of the accidents could have been avoided if I hadn’t been riding like a tit 😉

    Considering I ride pretty much 5 days a week, around 20 mile a day come rain or shine I don’t think it’s a bad record. In fact I think I’ve probably bumped my head more often whilst I’ve been off the bike over the same period yet you don’t see people suggesting we wear helmets ALL DAY?

    Cycling in and of itself is a pretty safe activity, the danger comes from those who are in charge of the considerably quicker and heavier motorised vehicles.