Bicycle helmet

A bicycle helmet is designed to attenuate impacts to the head of a cyclist in falls while minimizing side effects such as interference with peripheral vision. There is ongoing scientific research into the degree of protection offered by bicycle helmets in the event of an accident, and on the effects of helmet wearing on cyclist and motor vehicle driver behaviour. There is active debate over what can be concluded from available studies, and on whether the use of helmets by cyclists should be promoted or mandated, either just for children, or for cyclists of all ages. In particular the debate over bicycle helmet laws is intense and occasionally bitter, often based not only on differing interpretations of the scientific and other academic literature, but also on differing assumptions and interests of various parties. The evidence on bicycle helmets is mixed – some studies indicating that helmets may provide a benefit and others finding either no clear benefit or that they have failed to improve safety and that they may even make cycling more dangerous.

A bicycle helmet.

About helmets

History of designs

"Hairnet" helmet

A cycle helmet should generally be light in weight and provide ample ventilation, because cycling can be an intense aerobic activity which significantly raises body temperature, and the head in particular needs to be able to regulate its temperature. The dominant form of helmet up to the 1970s was the leather "hairnet" style. This offered acceptable protection from scrapes and cuts, but only minimal impact protection, and was mainly used by racing cyclists. More widespread use of helmets began in the U.S. in the 1970s. After many decades, when bicycles were regarded only as children's toys, many American adults took up cycling during and after the bike boom of the 1970s. Two of the first modern bicycle helmets were made by MSR, a manufacturer of mountaineering equipment, and Bell Sports, a manufacturer of helmets for auto racing and motorcycles. These helmets were a spin-off from the development of expanded polystyrene foam liners for motorcycling and motorsport helmets, and had hard polycarbonate plastic shells. The bicycle helmet arm of Bell was split off in 1991 as Bell Sports Inc., having completely overtaken the motorcycle and motor sports helmet business.

The first commercially successful purpose-designed bicycle helmet was the Bell Biker, a polystyrene-lined hard shell released in 1975. At the time there was no appropriate standard; the only applicable one, from Snell, would be passed only by a light open-face motorcycle helmet. Over time the design was refined and by 1983 Bell were making the V1-Pro, the first polystyrene helmet intended for racing use. In 1984 Bell produced the Li'l Bell Shell, a no-shell children's helmet. These early helmets had little ventilation.

In 1985, Snell B85 was introduced, the first widely adopted standard for bicycle helmets; this has subsequently been refined into B90 and B95 (see Standards below). At this time helmets were almost all either hard-shell or no-shell (perhaps with a vacuum-formed plastic cover). Ventilation was still minimal due mainly to technical limitations of the foams and shells in use.

A Giro Atmos helmet, showing seamless in-mould microshell construction.

Around 1990 a new construction technique was invented: in-mould microshell. A very thin shell was incorporated during the moulding process. This rapidly became the dominant technology, allowing for larger vents and more complex shapes than hard shells.

Use of hard shells declined rapidly among the general cyclist population during the 1990s, almost disappearing by the end of the decade, but remain popular with BMX riders as well as inline skaters and skateboarders.

The late 1990s and early 2000s saw advances in retention and fitting systems, replacing the old system of varying thickness pads with cradles which adjust quite precisely to the rider's head. This has also resulted in the back of the head being less covered by the helmet; impacts to this region are rare, but it does make a modern bike helmet much less suitable for activities such as unicycling, skateboarding and inline skating, where falling over backward is relatively common. Other helmets will be more suitable for these activities.

Since more advanced helmets began being used in the Tour de France, carbon fiber inserts have started to be used to increase strength and protection of the helmet. The Giro Atmos and Ionos, as well as the Bell Alchera were among the first to use carbon fiber.

Some modern racing bicycle helmets have a long tapering back end for streamlining. This type of helmet is mainly dedicated to time trial racing as they lack significant ventilation, making them uncomfortable for long races.

History of standards

In the United States the Snell Memorial Foundation, an organization initially established to create standards for motorcycle and auto-racing helmets, implemented one of the first standards, since updated. Snell's standard includes testing of random samples. In 1990 the Consumers' Association (UK) market survey showed that around 90 % of helmets on sale were Snell B90 certified. By their 1998 survey the number of Snell certified helmets was around zero.^{[citation needed]} There are two main types of helmet: hard shell and soft/micro shell (no-shell helmets are now rare). Hard shells declined rapidly among the general cyclist population over this period, almost disappearing by the end of the decade, but remained more popular with BMX riders as well as inline skaters and skateboarders.

The American National Standards Institute (ANSI) created a standard called ANSI Z80.4 in 1984. Later, the United States Consumer Product Safety Commission (CPSC) created its own mandatory standard for all bicycle helmets sold in the United States, which took effect in March 1999.

In the European Union (EU) the currently applicable standard is EN 1078:1997.

In the UK the current standard is BS EN 1078:1997, which is identical to the EU standard, and which replaced BS 6863:1989 in 1997.

In Australia and New Zealand, the current legally-required standard is AS/NZS 2063. A 2004 report concluded that the performance requirements of the 1996 version of this standard was slightly less strict than the Snell B95 standard but incorporated a quality assurance requirement, making it arguably safer. The Australian/New Zealand standard was improved in 2008. SAI Global (a profit-making arm of the Standards Australia organisation) publishes the AS/NZS 2063:2008 standard and provides a commercial service to helmet manufacturers and importers to test and certify helmets for compliance with the standard. Compliance is denoted by a "5 ticks" certification trademark sticker on the inside of helmets.

Brian Walker has stated that the CPSC and EN1078 standards are lower than the Snell B95 and B90s standards; Snell helmet standards are externally verified, with each helmet traceable by unique serial number. EN 1078 is also externally validated, but lacks Snell's traceability. The most common standard in the US, CPSC, is self-certified by the manufacturers. It is generally true to say that Snell standards are more exacting than other standards, and Walker asserts that most helmets on sale these days in the UK and Europe will not meet them.

Although the current European standard EN1078 is consider to be weaker than the earlier Snell standards there are no data on which to base an assessment of how this has affected the design goal of mitigating minor injuries. Minor injuries are substantially under-reported, and it is difficult to measure such injuries validly in any large population.^{[citation needed]}

In brief, the primary design goal of a helmet is to decelerate the skull (and by implication, the brain inside it) more gently than would be the case if no helmet were worn. A helmet's ability to manage linear deceleration of the skull could be improved by providing a greater thickness of expanded polystyrene foam and also by making this foam softer, but this would make the entire helmet bulkier, heavier, and hotter to wear.^{[citation needed]} Another hypothetical concern is that a thicker helmet increases the risk of rotational-type brain injuries (see discussion below). Ultimately, every helmet design represents some sort of compromise.

The trend is toward thinner helmets with many large vents. This trend to lower standards has been noted in some of the studies.

Design intentions and standards

Both are intended to reduce acceleration to the head due to impact, as a stiff liner made of expanded polystyrene is crushed against the head.

Standards involve the use of an instrumented headform which is dropped, wearing a helmet, onto various anvils. The speed of impact is designed to simulate the effect of a rider's head falling from approximately usual riding height, without rotational energy and without impact from another vehicle.

Collision energy varies with the square of impact speed. A typical helmet is designed to absorb the energy of a head falling from a bicycle, hence an impact speed of around 12 mph or 20 km/h. This will only reduce the energy of a 30 mph or 50 km/h impact to the equivalent of 27.5 mph or 45 km/h, and even this will be compromised if the helmet fails.^{[citation needed]} As a subsidiary effect, they should also spread point impacts over a wider area of the skull. Hard shell helmets may do this better, but are heavier and less well ventilated. They are more common among stunt riders than road riders or mountain bikers. Additionally, the helmet should reduce superficial injuries to the scalp. Hard shell helmets may also reduce the likelihood of penetrating impacts, although these are very rare.

Criticism of current standards

Some helmet liners may be too stiff to be effective. Standards require the use of headforms more rigid than the human head; these are more capable of crushing foam than is the human head.

Fit and care

correctly worn helmet

It is important that a helmet should fit the cyclist properly – in one study of children and adolescents aged 4 to 18 years, 96 % were found to be incorrectly fitted. Efficacy of incorrectly fitted helmets is reckoned to be much lower; one estimate states that risk is increased almost twofold.

Most manufacturers provide a range of sizes ranging from children's to adult with additional variations from small to medium to large. The correct size is important. Some adjustment can usually be made using different thickness foam pads. Helmets are held on the head with nylon straps, which must be adjusted to fit the individual. This can be difficult to achieve, depending on the design. Most helmets will have multiple adjustment points on the strap to allow both strap and helmet to be correctly positioned. Additionally, some helmets have adjustable cradles which fit the helmet to the occipital region of the skull. These provide no protection, only fit, so helmets with this type of adjustment are unsuitable for roller skating, stunts, skateboarding and unicycling. In general, the more skull coverage a helmet provides, the more effectively it can be fitted to the head and hence the better it will remain on the head in an accident.

Newer helmets for toddlers and children feature flat backs that prevent the helmet from tilting too far forward when worn while riding in a trailer or child seat with a headrest. Pinch-proof buckles and light-weighted helmets are also available for small children. Although these features do not increase protection upon impact, they are convenient.

The Snell Memorial Foundation recommends that any helmet that has sustained a substantial blow should be discarded and replaced, including any helmet involved in a crash in which the head has hit a hard surface or in which a fall has resulted in marks on the shell. Because some helmet materials deteriorate with age, the Snell Memorial Foundation recommends that a helmet be replaced at least every 5 years, or sooner if the manufacturer recommends it.

History of use

Helmets use varies greatly between populations and between groups. Downhill mountain bikers and amateur sportive cyclists normally wear helmets, and helmet use is enforced in professional cycle sport and in a few legal jurisdictions. Utility cyclists and children are much less likely to wear helmets unless compelled.

Required helmet use in cycling sport

Historically, road cycling regulations set by the sport's ruling body, Union Cycliste Internationale (UCI), did not require helmet use, leaving the matter to individual preferences and local traffic laws. The majority of professional cyclists chose not to wear helmets, citing discomfort and claiming that helmet weight would put them in a disadvantage during uphill sections of the race.

The first serious attempt by the UCI to introduce compulsory helmet use in 1991 was met with strong opposition from the riders. An attempt to enforce the rule at the 1991 Paris–Nice race resulted in a riders' strike, forcing the UCI to abandon the idea.

While voluntary helmet use in professional ranks rose somewhat in the 1990s, the turning point in helmet policy was the March 2003 death of Kazakh Andrei Kivilev. The new rules were introduced on 5 May 2003, with the 2003 Giro d'Italia being the first major race affected. The 2003 rules allowed for discarding the helmets during final climbs of at least 5 kilometres in length; subsequent revisions made helmet use mandatory at all times.

No studies have been published yet into whether injuries to racers have reduced as a result, but modern helmets can help to decrease aerodynamic drag by approximately 2% over a rider with no helmet, giving a competitive edge in a bicycle race.

Cycling risk and head injury

In the USA, two-thirds of cyclists admitted to hospital have a head injury. Ninety per cent of cyclist deaths are caused by collisions with motor vehicles. For cyclists admitted to hospital in Western Australia before the helmet law, about 30% of cyclists and 30% of pedestrians had head injuries. Trends and proportions of cyclists admitted to hospital with head injury were similar for all road users.

A 1996 study by Robinson found that based on Australian data, the risk of dying of head injury per million hours was 0.19 for cyclists, 0.34 for pedestrians, and 0.17 for car occupants.

An article published by the Bicycle Helmet Research Foundation (BHRF) reported that, per mile in the United Kingdom, cycling has an overall risk of injury and death similar to walking but higher than driving, and that in France cycling is safer per hour than motoring. An article by Roger Ford in Informed Sources, and a paper by Malcolm Wardlaw, reported that measured per hour, the risk of driving, cycling and walking are similar. A 2010 study by Tin Tin et al., at the School of Population Health at the University of Auckland, found that in New Zealand the average number of serious (AIS>2) injuries per million hours spent travelling in 2003-07 was 6.2 for cyclists, 1.0 for pedestrians, and 0.8 for car/van drivers.

Data published by the UK Department for Transport Statistics shows that in 2011, cyclists comprised 1% of the kilometres travelled; cars/taxis comprised 78% of the kilometres travelled; cyclists comprised 13% of reported road deaths and serious injuries; and car occupants comprised 37% of reported road deaths and serious injuries. These data suggest that the risk of death or serious injury per kilometre travelled in the UK is about 7 times higher for cyclists than for car occupants. The risk for car occupants 2001-2011 in the UK has reduced by about 37% but for cycling it has decreased by 13%.

In March 2013, Malcolm Wardlaw was reported by the Guardian saying that helmet compulsion "should be challenged. It's sending out the wrong message about cycling being dangerous. It's as pernicious as saying smoking is safe."

Science: measuring helmet effectiveness

Desirable effects of helmet use

No randomized controlled trials have been done on the subject. The effects of cycle helmets have been studied mainly through case-control studies and time-trend analyses. Case-control studies show a lower proportion of head injuries for helmet wearers compared to non-wearers, , indicating a substantial protective effect of helmets for head and brain injuries, and for fatality, as do meta-analyses of them. The results of time-trend analyses have been variable, some showing no apparent effects, others showing distinct effects on head injury rates in cyclists, as well as reductions in numbers of non-head injuries (a proxy for the amount of cycling or overall road safety) at the population level.

Time-trend analyses

Time-trend analyses compare changes in helmet use and injury rates in populations over time. Some of these studies have found that, when helmet use increased after the introduction of compulsory helmet laws in some jurisdictions, head injury rates among cyclists did not fall faster than head injury rates for other road users who do not wear helmets, such as pedestrians and motorists. The sudden increase in helmet wearing by cyclists in NSW at the time of the introduction of compulsory helmet laws, corresponded with a sudden drop in the by-month ratio of head injury rates to limb injury rates in cyclists. Criticisms of this study and a response by its authors have been published in the peer-reviewed literature, with additional commentary on web sites and blogs.

Authors do not agree on how data should be selected for analysis, nor on what summary statistics are most relevant. Potential weaknesses of this type of study include: simultaneous changes in the road environment (e.g. drink-drive campaigns); inaccuracy of exposure estimates (numbers cycling, distance cycled etc.), changes in the definitions of the data collected, failure to analyse control groups, failure to analyse long-term trends, and the ecological fallacy.^{[citation needed]}

A 1994 study by Marshall et al. compared hospital admissions for 1 year before and after the law, estimating 228 fewer injuries from reduced exposure (fewer cyclists or safer roads), 22 fewer injuries because of changes in hospital admission policies and 37 fewer injuries (4.3% of total cycling injuries) from increased helmet wearing. Comparing 2 years before and after the law suggested reductions of 149 injuries from reduced exposure, 80 from admission policies and 154 (9.7% of total) from helmet wearing. The authors concluded that the best estimate was the average of the 1 and 2-year comparisons, an 18.4% reduction in head injuries (7.5% of total cycling injuries). Census data for Adelaide for 1991 and 1996 had commuter cycling counts of 7186 and 4494 respectively and indicated a 37% reduction in cycling.

A 1995 study by Carr et al. found that there was a 40% drop in the proportion of serious and severe (AIS 3/4) cyclist head/brain injury admissions after the introduction of the helmet law in Victoria. Carr et al. also found that there had been a 46% drop in the proportion of serious and severe motor-vehicle involved cyclist hospital admissions. The data in the Carr study also showed that pedestrian head injuries had dropped by 20%, and cyclist non-head injuries had dropped by 25%. Carr et al. noted that there was some evidence that that the introduction of the helmets laws had let to decreases in cycling. After taking into account various factors that could have contributed to the reduction in cyclist casualties (including trends in pedestrian head injuries), Carr et al. concluded that "…the major part of this reduction is attributable to the introduction of the helmet wearing law" and that "…this analysis has confirmed the substantial reductions being made in both the number and severity of bicycle injuries since the introduction of the mandatory helmet wearing law." A 2006 review by Robinson that included the Carr et al. study noted that Carr et al. had reported a 40% reduction in head injuries, and stated that "…the authors could not tell whether the main cause was increased helmet wearing or reduced cycling because of the law. Non-head injuries fell by almost as much as head injuries, suggesting the main mechanism was reduced cycling, with perhaps some benefit from reduced speeding and drink-driving".

A 2005 study by Robinson that included a review of the Finch and Carr studies concluded that after adjusting for a 74% decrease in pedestrian deaths and serious head injuries (DHSI) due to general road safety measures, and a 30% reduction in cyclists due to the helmet law, cyclist DHSI should have fallen to 52% of the pre-law level, but only fell to 57% of the pre-law level. An article by Robinson on the Bicycle Helmet Research Foundation (BHRF) web site shows the declining trends in percent head injury in Western Australia, noting that that the analyses of percent head injury by researchers in Victoria found similar trends but "mistakenly concluded helmets were remarkably effective. They didn’t bother to check that the same trend was evident for pedestrians, so had nothing to do with helmets!".

A 1997 study by Scuffham and Langley noted that the New Zealand helmet law applies only to on-road cycling, and that the helmet wearing surveys were of on-road cyclists. Scuffham and Langley also noted that there was a (gradual) 19% reduction in the number of cyclists counted in the 2 years prior to the introduction helmet legislation in New Zealand.

A 1999 study by Povey et al. found that, after adjusting for time-trends (by including cyclist limb injuries in their statistical models), in New Zealand, the helmet law reduced cyclist head injuries by between 24 and 32% in crashes with no motor vehicle involvement, and by 20% in crashes with motor vehicle involvement.

A 2001 study by Robinson re-analysed a subset of the New Zealand data that had been analysed by Povey et al. The original analysis by Povey et al. had included cyclist injuries following crashes with motor vehicles, but Robinson's re-analysis was restricted to examining changes in cyclist injuries with no motor vehicle involvement and changes in on-road helmet wearing rates. Robinson concluded that "Unless voluntary wearing is 15 times more effective in reducing head injuries, it seems likely that the apparent effects (as described by Povey et al., 1999) were an artifact caused by failure to fit time trends in their model."

A 2010 study by Tin Tin et al. found that measured per hour spent cycling, in 1996-99 there were 70% fewer traumatic brain injuries in New Zealand than in 1988-91 (for crashes both with and without motor vehicle involvement), but that injuries to other body parts increased.

An evaluation in 2012 of the New Zealand helmet law by Clarke, reported that "…for the period 1989–1990 to 2006–2009, New Zealand survey data showed that average hours cycled per person reduced by 51%" and concluded that "…the helmet law has failed in aspects of promoting cycling, safety, health, accident compensation, environmental issues and civil liberties" and that "…the introduction of New Zealand's helmet law …[is] estimated to have contributed to 53 premature deaths per year (due to reluctance to cycle and hence people not exercising)." A critique of the Clarke paper has been published.

Robinson's reviews of cyclists and control groups in jurisdictions where helmet use increased by 40% or more following compulsion conclude that enforced helmet laws discourage cycling but produce no obvious response in percentage of head injuries. These studies have been the subject of vigorous debate. A review, by Macpherson and Spinks, which included two original papers (neither of which meet the criteria for inclusion in Robinson's review), concluded that "Bicycle helmet legislation appears to be effective in increasing helmet use and decreasing head injury rates in the populations for which it is implemented. However, there are very few high-quality evaluative studies that measure these outcomes, and none that reported data on an (sic) possible declines in bicycle use." Later work by Macpherson's group admitted that this conclusion had been erroneous and that "Although bicycle-related injuries are generally declining, this decline is not consistent, nor is it clearly associated with helmet laws."

Rodgers analysed annual (1973 to 1987) national time-series estimates of cycling injuries, annual counts of cycling and pedestrian road accident fatalities, estimates of bicycle use and bicycle standards compliance, and estimates of helmet use based on sales of one particular brand of (now obsolete) hard-shell helmet (helmet wearing rates used by Rodgers rose from 0% in 1973 to 5% in 1985). After fitting a linear regression model to these data he concluded that: "There is no evidence that hard shell helmets have reduced the head injury and fatality rates. The most surprising finding is that the bicycle-related fatality rate is positively and significantly correlated with increased helmet use" This was a time-trend analysis with the potential weaknesses mentioned above; the correlation may not be causal. A strong association between voluntary helmet wearing and a lower risk of serious non-head injuries as well as head injuries was reported in a study of cyclists injured in accidents involving motor vehicles. Different analyses of the same data can produce different results. For example, Scuffham analysed data on the increase of voluntary wearing in New Zealand to 1995; he concluded that, after taking into account long-term trends, helmets had no measurable effect. His subsequent re-analysis without accounting for the long-term trends suggested a small benefit.

A 2002 cost-benefit analysis of the New Zealand helmet law by Taylor and Scuffham found that there was no overall cost savings with respect to (initial) hospital admissions averted. Taylor and Scuffham noted that their analysis did not include the savings associated with reduced fatalities, the life-time care costs associated with severe traumatic brain injuries were also not taken into account.

A 1999 cost-benefit of the helmet law in Western Australia by Hendrie et al. and a 1995 study by the (Australian) New South Wales Roads and Traffic Authority are discussed in a related article.

A 2010 study by Dinh et al. of cyclist head injuries seen at an Australian major trauma centre found that unhelmeted cyclists were 5.3 times more likely than unhelmeted cyclists (p<0.01) to sustain a severe head injury (AIS score >2). Dinh et al. noted that 'the lifetime care cost of a traumatic brain injury of this severity was estimated to be about $4.8 million'.

The Victorian hospital admission data in the study by Carr et al. showed that in the 3 years after the helmet law, compared to the 3 years before the helmet law, there was a 20% reduction in the number of pedestrian head injuries, and a 60% reduction in number severe (AIS>2) cyclist head injuries. In the 3 years before the helmet law, there were an average of 73 severe (AIS >2) head/brain hosptial admisssions per year, in the 3 years after the helmet law, there were an average of 29 severe head/brain hospital admissionss per year. A subsequent analysis by Robinson reported the effect of the helmet law on both deaths and serious head injuries (DSHI) and serious injuries not involving the head (OSI), noting that the decline from 72.5 to 41 DSHI was matched by a decline in OSI (from 274 per year to 165) and that the ratio of DSI:OSI showed almost identical changes for pedestrians and cyclists. In 1991, Victoria comprised approximately 1/3 of the population of Australia.

Australian Transport Road Safety Bureau road fatality data shows that between 1992 and 1994, compared to 1987-1989, the number of pedestrian fatalities fell by 32%; and that the number of cyclist fatalities fell by 45%, from an average of 88 to 48 annually. These fatality reductions suggest that the helmet legislation was responsible for preventing 20 deaths a year when it was introduced in Australia. A 2011 report by the (Australian) National Roads and Motorists Association estimates the cost of a road fatality to be $3,180,598. However, when fatality data for 1988 and 1994 were split into deaths for head and other injuries, Curnow noted a fall of 30% in cyclist head injuries, 38% in pedestrian head injuries and 42% in head injuries to all road users and concluded that any reductions in fatalities were due to reductions in cycling.

A 2012 study of post-law data by University of New South Wales researchers found that the population-based hospitalisation rate for cyclist arm injuries had increased by 145 percent in NSW from 1991 to 2010 (a period when helmet wearing rates did not increase). In the same time period head injuries rates only increased by 20 percent. When trends are assessed relative to available estimates of cycling participation from 2001-2010, arm injuries increased by 46% while head injuries remained flat. In terms of absolute numbers of hospitalisations, between 1991 and 2000, the number of arm injuries increased by 100%, while the number of head injuries increased by 40%.

A 2013 study by Dennis et al. examined annual 1994 to 2008 hospital admission counts for head injuries and all injuries in cyclists in ten Canadian provinces, six of which had enacted mandatory cycling helmet laws for either children or for all ages during that time period. They found that between 1994 and 2003, the rate of head injuries among young people decreased by 54.0% (95% confidence interval 48.2% to 59.8%) in provinces with helmet legislation compared with 33.1% (23.3% to 42.9%) in provinces and territories without legislation. Among adults, the rate of head injuries decreased by 26.0% (16.0% to 36.3%) in provinces with legislation but remained constant in provinces and territories without legislation. However, after taking baseline trends into consideration, they were unable to detect an independent effect of helmet legislation on the rate of hospital admissions for cycling related head injuries.

A 2013 study by Yilmaz et al. found that between 2001 and 2009, cyclists admitted a trauma centre in the Netherlands suffered from more serious (AIS>=3) head injuries than cyclists admitted to a trauma centre in Australia (88% vs 62%, p<0.001), and that cyclists admitted to the Australian trauma centre had significantly more serious (AIS>=3) non-head injuries. Yilmaz et al. also found that the cyclist major trauma admissions in the Netherlands trauma centre had a higher mortality rate, associated with a higher percentage of serious head injuries than did the cyclist major trauma admissions in the Australian trauma centre.

Case-control studies

In a case-control study, hospitalised cyclists are divided into those with head injuries (cases) and those without (controls). Many studies of this type have been conducted, and they usually conclude that helmets provide some level of protective effect against head injuries. For example, one recent French study analyzing over 13,000 cyclist casualties during a ten-year period "…confirms the protective effect [of helmets] for head and facial injuries…" and finds that "…the reduction of risk is greater for serious head injuries. The study is inconclusive about the risk for neck injuries." The most widely quoted case-control study, by Thompson, Rivara, and Thompson, reported an 85% reduction in the risk of head injury by using a helmet. These studies have been criticised in the scientific literature, and the authors of the studies have responded.

A 2012 study by Persaud et al. using coronial records of 129 deaths of cyclists in Canada between 2005 and 2010 found that unhelmeted cyclists in fatal crashes were more likely than helmeted cyclists to have sustained a head injury (adjusted odds ratio [OR] 3.1, 95% confidence interval [CI] 1.3-7.3). The published replies detail that 23% of the bicyclists had been under the influence of alcohol or drugs at the time of death, while the status of 30% was unknown. Of the bicyclists, 16% were carrying unsafe loads at the time of death. Some details are mentioned from other reports, comparing wearers to non-wearers, e.g. A 1993 report from Ontario stated "In teenagers, drinking alcohol (OR: 2.8) and smoking (OR: 4.4) were strongly associated with helmet non-use. In the adult group, female gender (OR: 1.26), higher income (OR: 1.43), higher education (OR: 1.68), non-smoking status (OR: 2.0) and abstinence from alcohol (1.27) were associated with helmet use."

Meta-analyses

There are several meta-analyses and reviews which synthesize and evaluate the results of multiple case-control studies. A Cochrane review of case-control studies of bicycle helmets by Thompson et al. found that "helmets provide a 63 to 88% reduction in the risk of head, brain and severe brain injury for all ages of bicyclists. Helmets provide equal levels of protection for crashes involving motor vehicles (69%) and crashes from all other causes (68%). Injuries to the upper and mid facial areas are reduced 65%.". Robinson pointed out that in the Thompson et al. review "…the small numbers and potential problems of confounding… suggest that the conclusions concerning [more serious] brain injury >AIS2 should be treated with caution", and that "…what the authors describe as a 69% reduction in head injuries actually means that in crashes severe enough for 90% of non-helmet wearers to be head injured, so would 73.6% of helmet wearers". Curnow criticised the Thompson et al. review for relying on observational case-control studies and failing to include any randomised-control trials (there are none for bicycle helmets), and for taking "...no account of scientific knowledge of types and mechanisms of brain injury". Cummings et al. subsequently asserted that Curnow's criticisms with respect to case-control studies stemmed from misunderstandings of the epidemiological principles of injury case-control studies. Hagel and Barry Pless also responded to Curnow's criticisms of the Cochrane review, pointing out that "…the crux of his [Curnow's] argument is that in theory helmets should not protect [against] all mechanisms of brain injury and, therefore, all epidemiological research showing [bicycle helmets] are beneficial in a variety of circumstances is invalid". Curnow provided replied to these responses.

A 2001 meta-analysis of sixteen studies by Attewell et al. found that, compared to helmeted cyclists, unhelmeted cyclists were 2.4 times more likely to sustain a brain injury; 2.5 times more likely to sustain a head injury; and 3.7 times more likely to sustain a fatal injury.

A 2012 re-analysis of the 16 studies in the Attewell meta-analysis, by Elvik, found that, compared to helmeted cyclists, unhelmeted cyclists were 2.5 times more likely to sustain a brain injury; 2.3 times more likely to sustain a head injury; and 4.3 times more likely to sustain a fatal injury. When 5 new head-injury studies were added to the model, Elvik found that unhelmeted cyclists were 1.9 times more likely than helmeted cyclists to sustain a head injury. When head, face and neck injuries were combined, Elvik found that unhelmeted cyclists were 1.4 times more likely than helmeted cyclists to sustain an injury to the head, face or neck. Elvik noted that the findings were inconsistent with a Cochrane review published in 2009, and that "the study inclusion criteria applied in the Cochrane review are debatable".

Undesirable effects of helmet use

Health benefits of cycling

Studies from China, Denmark, the Netherlands and the United Kingdom show that regular cyclists live longer because the health effects far outweigh the risk of crashes. A reduction in the number of cyclists is likely to harm the health of the population more than any possible protection from injury. UK figures show that it takes at least 8000 years of average cycling to produce one clinically severe head injury and 22,000 years for one death. De Jong developed a mathematical model to evaluate the health-risk trade-offs of all-age mandatory helmet laws, if they were to be introduced in various North American and Western European countries. He concluded that helmet laws appear to offer net health benefit only in those countries with more dangerous bicycling environments under optimistic assumptions of the efficacy of helmets. Newbold suggested improvements to the de Jong model, and, using published cycling statistics for the United States in his revised model, found that mandatory bicycle helmet laws would seem to have positive net public health benefits there. However, Newbold stressed that there were many parameters to these models which require further research to properly quantitate, and that results should be considered provisional rather than definitive.

Concerns have been raised that enforced mandatory bicycle helmet laws might lead to a reduction in the number of cyclists, with worse health as a result. This suggestion has been criticized. Fewer cyclists might lead to increased risks per cyclist due to the "safety in numbers" effect. According to one source, the probability of an individual cyclist being struck by a motorist declines with the 0.6 power of the number of cyclists on the road. This means that if the number of cyclists on the road doubles, then the average individual cyclist can ride for an additional 50% of the time without increasing the probability of being struck. It is thought that the increased frequency of motorist-cyclist interaction creates more aware motorists.

Bicycle usage related to helmet promotion and compulsion

Most work, and the only studies using concurrent control groups, come from Australia; see Bicycle helmets in Australia for more detail. Observational surveys of helmet wearing and cyclist numbers were carried out in Melbourne, Adelaide and throughout NSW, before and after the introduction of compulsory helmet laws.

A 1993 study by Finch et al. found that in Melbourne, compared to a survey the year before the helmet law, in the year after the helmet law, 29% fewer adult and 46% fewer teenage cyclists were counted. Finch et al. noted that in the post-law survey, more observation sessions coincided with rain than in the pre-law survey. After comparing the 82% of sites that had the same weather classification in both surveys, Finch et al. concluded that there was a 13% reduction in adult cylists, and a 41% reduction in teenage cyclists. Robinson noted that the reduction or 623 teenagers counted (1293 in 1990, 670 post-law in 1991) was much greater than the increase of just 30 teenagers wearing helmets (272 pre-law, 302 post-law.)

A 1994 study by Marshall and White found that surveys taken in South Australia 18 months before and after the introduction of helmet law showed that there was no (statistically significant) reduction the proportion of adults and children who cycled at least once a week. Marshall and White also found that prior to the helmet law, cycling to school comprised about 20% of cycling activity in that age group, and that after the helmet law cycling to school dropped by about 40%, which was accompanied by an increase of about 40% in cycling to/around other venues for this age group.

In 2006 Robinson re-analysed the data from the study by Finch et al., but excluded one site because a bicycle rally had passed through it. Finch et al. had noted that from a statistical point of view, it would not be valid to exclude that site. (Excluding the site would have excluded cyclists who would have been cycling anyway, but through a different site, had it not been for the rally.) After excluding the site, Robinson concluded that cycling in Australia fell by roughly one-third as a result of the helmet laws.

A 1997 study by Scuffham and Langley noted that there was a (gradual) 19% reduction in the number of cyclists counted in the 2 years prior to the introduction helmet legislation in New Zealand. A 1999 study by Povey et al. noted a claim in a 1996 article by Robinson that helmet compulsion reduced cycling, and stated that there was no evidence of this in the New Zealand data.

A study by Robinson found that after the helmet law, there was a reduction of about one-third in commuting by bicycle in New South Wales and Perth. Despite the fact that surveys soon after helmet laws were introduced showed that 51% of schoolchildren in NSW said they had not cycled because of helmet legislation, and the equivalent of 64% of current adult cyclists in Perth also said they would cycle more if not legally required to wear a helmet the conclusion that this was due to helmet laws was criticized. Significant upgrades to the Perth rail system, including electrification of the 3 existing lines in 1992, and the opening of a new line in 1993, have been the subject of a transport mega project case study. The census data compiled by Mees et al. shows that in Perth, between 1991 and 1996, travel to work by train increased by 37%, and travel to work by bus decreased by 41%.

Bicycle hire schemes in Australia have had low usage rates, of the order of one-tenth of schemes in areas without helmet laws, and mandatory helmet legislation was thought to reduce spontaneous use.

Census data on single-mode journeys show that 1.7% of people cycled to work in 1986, compared to to 1.6% in 1991 (when 4 states, comprising 70% of the Australian population, had enforced helmet laws). In 1996 (when all jurisdictions had helmet laws), 2001, and 2006, 1.2% of the workforce cycled to work. A 2007 study by Mees et al. noted that "a dramatic increase in the number of cars driven to work each day in Australia’s capital cities, with a total increase of 1,439,024 cars, or 70.1%, between 1976 and 2006" had been associated with a shift away from more sustainable transport modes. Between 1981 and 2001, travel to work by bicycle dropped from 1.6% to 1.3%, travel by bus dropped from 5.5% to 3.3%, walking to work dropped from 6.4% to 4.7%, and travel as a car passenger dropped from 12% to 6.7%.

Gillham and Rissel used cycling data from two different national surveys of persons aged 9 or over and found a 21% increase in cycle trips between 1985/86 and 2011, compared to a 58% increase in the population aged 9+ years. According to the Australian Bureau of Statistics (ABS) census data, between 1986 and 2011, there was a 42% increase in the Australian population aged 9+ years. Based on a 58% increase, any per capita increases from 1985/86 until the introduction of helmet laws were reversed, with a further 24% reduction in the number of cycle trips per person compared to 1985/86 levels. Davies noted concerns about the comparability of these surveys, and Olivier et al. noted the failure to adjust for the ageing of the Australian population in the three decades between the surveys. When this adjustment was made, per capita cycling rates were found to have increased slightly. ABS Participation in Sport and Physical Recreation reports show that there was a 38% increase in cycling participation rates between 1997/8 and 2009/10. The Australian National Cycling Strategy cites an ABS data series on Environmental Issues that the modal share of cycle commuting rose from 1.1% to 1.5% between 2000 and 2009, but omits to say that the same ABS data series reported a modal share of 1.9% in 1996.

In the UK between 1994 and 1996, in areas where cyclist counts dropped, wearing rates increased and where the number of cyclists increased, helmet wearing rates fell.

Several mechanisms by which cycle helmet promotion or compulsion may deter cycling have been suggested. Helmets and their promotion may reinforce the misconception that bicycling is more dangerous than traveling by passenger car. Referring to the use of "human skull" images in a campaign, the CTC suggests that "this macabre imagery, with its associations of hospitals and death, is likely to reduce cycle use, thereby undermining efforts to realize the health and other benefits of increased cycling". Bicycle helmets can be an additional expense.

A 2010 telephone survey of 600 residents in Sydney, Australia, found that 22.6% of the respondents said that they would cycle more if they did not have to wear a helmet. The survey's authors conclude: "While a hypothetical situation, if only half of the 22.6% of respondents who said they would cycle more if they did not have to wear a helmet did ride more, Sydney targets for increasing cycling would be achieved by repealing mandatory bicycle helmet legislation." This prediction was questioned after being subject to numerical scrutiny.

Risk compensation

It has been hypothesised that the wearing of helmets may make cyclists feel safer and thus take more risks. This hypothetical effect is known as risk compensation or risk homeostasis. Some authors have suggested that risk compensation occurs with other road safety interventions such as seat belts and anti-lock braking systems, but these views are widely disputed by road safety experts.

A Spanish study of traffic accidents between 1990 and 1999 found that helmeted cyclists involved in accidents were less likely to have committed a traffic law violation than unhelmeted cyclists, and that helmeted cyclists were no more likely to have committed a speeding violation in association with the accident than unhelmeted cyclists. The authors concluded that "…although the findings do not support the existence of a strong risk compensation mechanism among helmeted cyclists, this possibility cannot be ruled out."

In one experimental study, adults accustomed to wearing helmets cycled more slowly without a helmet, but no difference in helmeted and unhelmeted cycling speed was found for cyclists who do not usually wear helmets. An experimental study found that that children negotiating an obstacle course on foot went faster and took more risks when wearing safety gear (including helmets). A telephone interview study found that in hypothetical scenarios of their children wearing protective equipment or not, parents' ratings of permissible risk for their children was higher if protective gear was hypothetically worn.

Motorists may also alter their behavior toward helmeted cyclists. One study by Walker in England found that 2500 vehicles passed a helmeted cyclist with measurably less clearance (8.5 cm) than that given to the same cyclist unhelmeted (out of an average total passing distance of 1.2 to 1.3 metres). The significance of these differences has been re-analysed by Olivier.

As described above, in 1988 Rodgers re-analysed data which supposedly showed helmets to be effective; he found data errors and methodological weaknesses, and concluded that in fact the data showed "bicycle-related fatalities are positively and significantly associated with increased helmet use". He mentioned risk compensation as one possible explanation of this association.

Rotational injury

A 1987 study measured rotational accelerations averaging 58,000 rad/s⁻² in a crash test head wearing a hard-shell bicycle helmet in bicycle crash experiments simulating a fall over the handlebars at 45 km/h. The rotational accelerations averaged 30 percent higher than those found in similar tests using a full-face polymer motorcycle helmet. These accelerations were described as "enormous" compared with 4,500 rad/s⁻² for onset of intracranial vein rupture. Comparison tests on unhelmeted crash test heads were not performed.

A 1993 study of bicycle helmet chin-strap forces in simulated oblique road surface impacts measured peak rotational acceleration in crash test dummy heads of 28000 rads/s² for a non-shell helmet (also known as "soft-shell" helmets) in a 34km/h test. However, the investigators reported that "...[hard] shell helmets slid against the asphalt surface and there was only a slight angular movement of the head when the head was pushed upwards." Comparison tests on unhelmeted crash test heads were not performed.

Curnow has suggested that the major causes of permanent intellectual disablement and death after head injury may be torsional forces leading to diffuse axonal injury (DAI), a form of injury which usual helmets cannot mitigate and may make worse. However, Curnow's hypotheses are disputed: a 2012 University of NSW study found that of all reported cycling crashes with motor vehicles in NSW between 2001 and 2009 where helmet and hospitalisation information was present, DAI could have occurred in at most 0.2% of cases (12/6745); seven of these twelve possible cases of DAI were unhelmeted. A 2012 experimental study by McIntosh et al. tested Curnow’s hypothesis that bicycle helmets increase angular acceleration during a crash, and found that they actually reduced both linear and angular acceleration by a considerable margin. In a recent report on motorcycle helmets McIntosh states "Further research and development is required on the oblique test rig to establish its reliability and validity, the latter through comparisons to real world impacts".

Full-face and slip-plane helmet designs

A full-face helmet for downhill mountain bicycling.

A 1991 study by Hodgson, in which bicycle helmets were tested for ease of skidding, found that adding facial protection to a standard bicycle helmet (thus, in effect, converting it into a full-face helmet) brought the benefit of reduced twisting forces on the brain. A full-face helmet design was also found to reduce the cervical-spine injury index. It also assisted in keeping the helmet in place during the crash tests and in protecting the face. It reduced "neck injury by reducing the facial-pavement friction and subsequent twisting, bending, and compression; and brain injury by reducing sudden rotational movements during facial impact, and lowering linear head accelerations by absorbing energy by deformation in the event impact occurs on the guard."

A bicycle helmet with its own synthetic "scalp" has been designed with the aim of mitigating rotational injury. This is one of several slip-plane-type designs that are intended to reduce rotational acceleration on the brain during an oblique impact. Although many bicycle helmets are designed with rounded, smooth shells that slide easily along pavement, adding a slip plane to such an already-good design may be valuable in situations where the helmet impacts obliquely against a high-friction surface. The concept of slip planes is not new, but the implementation in production bicycle helmets is.

Accidental hanging by helmet straps

There are cases of young children playing (on or near bunk beds, trees, clothes lines, play equipment etc.) suffering death or severe brain damage as a result of hanging by the straps of their bicycle helmets. As a result, European Standard EN 1080 was developed, which uses a weak retention system designed to open under load. Such helmets are not intended for use anywhere motor vehicles are present. To avoid serious accidents, parents and carers should take care to ensure that children do not wear bicycle helmets during unsupervised play, or when using climbing equipment. There is also a case of a toddler saved by his helmet from slipping through a hole in a bridge over the Burn of Mosset.

Opinions for and against the compulsion or strong promotion of helmets

Supporters

Significant helmet promotion preceded epidemiological studies evaluating the effectiveness of bicycle helmets in bicycle crashes. Received opinion in some English-speaking countries is that bicycle helmets are useful and that every cyclist should wear one; helmets had become a ‘ "Mom and apple pie" issue’ in the United States by 1991 according to the League of American Bicyclists.

A number of cycling advocacy organizations support helmet use. The League of American Bicyclists "has encouraged the wearing of helmets via its publications and its education program for many years. Since 1991 the League has required participants in League-sponsored events to wear helmets." Cycling Advocates' Network (CAN), a nationwide New Zealand cycling advocacy group, "fully supports the use of helmets when undertaking recreational cycling in difficult terrain or high-speed competitive racing."

Numerous health and medical groups, such as the World Health Organization, Centers for Disease Control and Prevention, and American Academy of Pediatrics recommend wearing helmets.

Some medical groups support helmet laws, as well. These include the British Medical Association, American Medical Association, and American College of Emergency Physicians.

Groups in other sectors also advocate helmet use or laws. For instance, safety groups Safe Kids USA and the National Safety Council urge helmet wearing. Temple University's Public Health Law Research program classifies bicycle helmets laws as an "effective" public health intervention, based on a review of scholarly research.

U.S.-based cycling activist John Forester suggests that helmet wearing could prevent 300 deaths a year in the US out of a total of 1530 preventable deaths, behind Effective Cycling at 500 but ahead of all other interventions.

Opponents

Dorothy Robinson reviewed data from jurisdictions where helmet use increased following legislation, and concluded that helmet use did not demonstrably reduce cyclists' head injuries. Mayer Hillman, a transport and road safety analyst from the UK, does not support the use of helmets, reasoning that they are of very limited value in the event of a collision with a car, that risk compensation negates their protective effect and because he feels their promotion implicitly shifts responsibility of care to the cyclist. He also cautions against placing the recommendations of surgeons above other expert opinion in the debate, comparing it to drawing conclusions on whether it is worthwhile to buy lottery tickets by sampling only a group of prizewinners. The prominent UK-based cycling activist John Franklin is skeptical of the merits of helmets, regarding proactive measures including bike maintenance and riding skills as being more important. Cyclists' representative groups complain that focus on helmets diverts attention from other issues which are much more important for improving bicycle safety, such as road danger reduction, training, roadcraft, and bicycle maintenance.

In 1998 the European Cyclists' Federation adopted a position paper rejecting compulsory helmet laws as being likely to have greater negative rather than positive health effects. The UK's largest cyclists' organisation, the CTC, believes that the "overall health effects of compulsory helmets are negative." The British National Children's Bureau has said "The 2004 BMA statement announcing its decision to support compulsory cycle helmets shows how the uncritical use of accident statistics can lead to poor conclusions." The same report estimated that, at most, universal helmet use would save the lives of three children aged 0 to 15 each year. That figure "assumes universal and correct use of helmets, it assumes that risk compensation does not occur and it assumes that no children die as a result of strangulation or other injuries caused by helmet use. These assumptions are most unlikely to be correct in the real world." The Dutch Fietsersbond summarized existing evidence and concluded that a compulsory helmet law (for utility cyclists) would have a negative impact on population health: "Helmet laws save a few brains, but destroy a lot of hearts". No policy position was provided for other types of cycling, particular mountain biking (MTB) and all forms of on- and off-road cycle sports.

Legislation and culture

See also: Bicycle helmet laws by country

The following countries have mandatory helmet laws, in at least one jurisdiction, for either minors only, or for all riders: Australia, Canada, Czech Republic, Finland, Iceland, New Zealand, Sweden, and the United States. Spain requires helmets on interurban routes. In the U.S. 21 states have state-wide mandatory helmet laws for minors of varying ages, and 37 states have mandatory helmet laws for varying age groups in varying jurisdictions. Nearly 9 in 10 American adults support helmet laws for children. Israel's helmet law was never enforced or obeyed, and the adult element has been revoked; Mexico City has repealed its helmet law.

In 2004, a Bill proposing to make the wearing of bicycle helmets compulsory came before the UK Parliament, and was defeated. Horton observed: "The 2004 Parliamentary Bill was unanimously opposed by the cycling establishment, with every major cycling organisation and magazine rejecting helmet compulsion."

Although the link is not causal, it is observed that the countries with the best cycle safety records (Denmark and the Netherlands) have among the lowest levels of helmet use.^{[citation needed]}. Their bicycle safety record is generally attributed to public awareness and understanding of cyclists, safety in numbers, education, and cycling infrastructure. A study of cycling in major streets of Boston, Paris and Amsterdam illustrates the variation in cycling culture: Boston had far higher rates of helmet-wearing (32% of cyclists, versus 2.4% in Paris and 0.1% in Amsterdam), Amsterdam had far more cyclists (242 passing bicycles per hour, versus 74 in Paris and 55 in Boston). Cycle helmet wearing rates in the Netherlands and Denmark are very low. An Australian journalist writes: "Rarities in Amsterdam seem to be stretch-fabric-clad cyclists and fat cyclists. Helmets are non-existent, and when people asked me where I was from, they would grimace and mutter: "Ah, yes, helmet laws." These had gained international notoriety on a par with our deadly sea animals. Despite the lack of helmets, cycling in the Netherlands is safer than in any other country, and the Dutch have one-third the number of cycling fatalities (per 100,000 people) that Australia has." The UK's CTC say that cycling in the Netherlands and Denmark is perceived as a "normal" activity requiring no special clothing or equipment. Pucher and Buehler state: "The Dutch cycling experts and planners interviewed for this paper adamantly opposed the use of helmets, claiming that helmets discourage cycling by making it less convenient, less comfortable, and less fashionable. They also mention the possibility that helmets would make cycling more dangerous by giving cyclists a false sense of safety and thus encouraging riskier riding behavior."

See also

• Bicycle helmet laws
• Bicycle helmet laws by country
• Bicycle helmets in Australia
• Bicycle helmets in New Zealand
• Bicycle safety

External links

• World Health Organization Helmet Initiative
• Bicycle Helmet Research Foundation
• Bicycle Helmet Safety Institute
• Helmet Development and Standards