BICYCLE HELMET
RESEARCH
FOUNDATION
cyclehelmets.org
Home page
Main topics
News Headlines
Frequently asked Questions
For Policy Makers
Research evidence
Misleading claims
Helmet laws
Analysis
Search Engine
Australia
Canada
New Zealand
UK
USA
Other countries
Full index
Links
BHRF
Policy statement
Curran ALM
A briefing for the Board of Science, British Medical Association, March 2008 (revised February 2010)
The briefing relies crucially upon the conclusions of studies which have been shown to have unacceptable flaws, and passes lightly (if at all) over more sceptical, but scientifically stronger, analyses. It misrepresents statistical data by using it out of context, misunderstands basic concepts, and ignores discussion of reduced cycling which previous BMA publications have acknowledged. The use of emotive, and one-sided, anecdotal quotations is out of place in a scientific piece of work.
No reference is made to evidence that cycle helmets can increase risk of the most serious types of head injury through rotational brain injury. There is no mention of research findings that helmets do not work as designed in real crashes. The most comprehensive review of the outcomes of helmet legislation, published in the BMA's own British Medical Journal, is mentioned but without any consideration of its findings, that helmets do not reduce the proportion of head injury among cyclists. These and other serious omissions are unacceptable.
This briefing reflects poorly on the ability of the Board of Science of the BMA to look objectively at a complex and controversial subject. It would seem that the influence of individuals, predisposed to helmet laws, may have exceeded the influence of objective evidence and the need for academic rigour.
The comments below follow the BMA's section headings
In fact, there is significant direct research showing benefit. The most substantive study is by the Copenhagen Centre for Prospective Population Studies (Andersen, Schnohr, Schroll and Hein, 2000). This cohort study followed 30,640 Danish individuals aged 20-93. It concluded that "even after adjusting for other risk factors, including leisure time physical activity, those who did not cycle to work experienced a 39% higher mortality rate than those who did". A study of Chinese women found a 20% to 50% lower risk of mortality (Matthews et al, 2007).
Other research in Holland and Finland has found that cycling to work even short distances a few times a week had significant potential to improve health-related physical fitness in previously sedentary adults. Research in Britain found that amongst a group of mostly overweight individuals that took up cycling to work, body fat was significantly reduced, whilst mood and tolerance to stress increased. Other studies show coronary heart disease developed significantly later in those who cycled as adults (Cavill and Davis, 2007). The life-extending benefits of regular exercise, such as cycling, have been suggested to exceed the risks of cycling by a factor of 20 (Hillman, 1997; BMA, 1992).
.
However, the situation is more complicated than this and there are circumstances worthy of investigation.
Cycle use in Britain grew steadily after the first oil crisis in 1973, reaching a peak in the mid 1980s, but then suffered a severe fall of 33% by 1993. The years 1989 to 1993 saw particularly sharp decline. These were the years during which cycle helmet use was for the first time promoted. After 1993, annual cycle mileage increased very slowly to 2004, but then suffered a set back of 10-15%. It was in 2004 that a bill was presented to Parliament proposing to make helmets compulsory for children. Although a law was not passed, significant media interest was created about the use of cycle helmets. Since 2004, cycle use has again increased.
Promotion and publicity about cycle helmets unavoidably portray cycling as an exceptionally hazardous activity, creating a climate of public opionion less favourable to cycling. The events above may have been, at least in part, a reason for the falls in cycle use. (see also Does it discourage cyclists, below)
Great Britain cycle use data: DfT, census, helmet use data: Bryan-Brown and Taylor, 1997.
.
This unsubstantiated statement is misleading. Serious casualties when cycling on the public highway are in fact quite rare. The National Travel Survey (DfT, 2006) reports there are 2.5-3.0 million regular cyclists in the UK, while HES and STATS19 both report about 2,300 serious injuries in traffic collisions (DfT, 2006b). The rate is only about 1 serious injury per 1,100-1,300 cyclists per year, a low level of risk.
Moreover, the BMA has failed to highlight a critical aspect of safety for cyclists: popularity. There is good evidence that if cycle use increases there is no increase in deaths; indeed, deaths may continue to fall. This 'safety in numbers' effect has been verified across countries (BHRF, 1186). On the other hand, if there is a fall in cycling, risk per cyclist will likely increase. The falling trend in cycle use in Britain noted above is thus far more ominous for the welfare of continuing cyclists than the author appears to appreciate.
.
The vast majority of these are minor injuries to children and relate to falls off-road, not transport injuries. Department of Health HES data show that between 1995 and 2003, of all children admitted to hospital due to injury (all causes), only 8% followed cycling accidents. This places cycling injuries in perspective - cycling is a minor cause of serious injury, even for children.
In the 2006 edition of the Department for Transport's standard reference, Road Casualties, Great Britain 2006 (RCGB 2006), there is a chapter that discusses police-reported injuries and records of hospital admissions (DfT, 2006b). The police-reported injuries are recorded in a database called STATS19, while hospital admissions are recorded in a database of the Department of Health known as the Hospital Episode Statistics (HES).
The analysis in RCGB 2006 reports as a primary conclusion that cyclists suffer 17% of all serious injuries in road traffic accidents. This would certainly appear to support the view that cycling is comparatively hazardous, since cycling accounts for only 0.6% of distance travelled and 2% of time travelled by the UK population. The BMA has repeated this finding in good faith.
The figure of 17% is wrong. The Department for Transport (DfT) has acknowledged this following correspondence (Wardlaw, 2008) with a member of the BHRF Board. The error arose due to a false comparison between cyclist casualties and those of other road users. The DfT now agrees that the police-reported (STATS19) figure of 8% is the correct one to use. That is, cyclists account for 8% of all hospital admissions following traffic collisions.
This is of course still disproportianately high relative to pedestrians and car occupants. The problem is that there is a preponderance of teenage boys and young men among the cyclists using Britain's roads, and few of them have benefitted from more than cursory training. These two groups suffer high injury rates, whatever form of transport they are using, due to their high level of risk-taking. This difference in the age distribution and gender balance of cyclists loads the casualty data, making cycling look disproportionately dangerous. Yet a late 1980's study by the Transport and Road Research Laboratory showed that 17-21 year old males themselves face similar risks as drivers or as cyclists (Morgan 1988).
Despite this, as already noted, the absolute risk remains low at one hospital admission per 1,100 to 1,300 cyclists per year, and one death per 20,000 cyclists per year. These are low risks. A further point is that cyclists spend on average much less time as in-patients than other classes of road user. Data from Odense, Denmark (AAG, 2000) reports cyclists are held as in-patients less than half as long as car occupants, and only one sixth the duration of pedestrians. Thus total hospital days for cyclists will be much less than 8% of all hospital days due to road traffic accidents (we would suggest 3% as a working estimate, pending further investigation).
.
It is misleading to lump deaths and injuries together as this emotionally increases the perception of risk when cycling. In recent years there have been 115-150 cyclist deaths annually in Great Britain. In relation to the number of active cyclists, deaths are very rare, corresponding to a rate of about 1 in 20,000 regular cyclists per year. The percentage of road deaths that are cyclists is around 4%, a proportion that has remained fairly constant for the last 40 years.
|
One of the most disturbing aspects of rising helmet use has been a corresponding increase in deaths. Cyclist deaths sharply increased in the years when helmets first became popular in the UK, despite steady declining trends for pedestrians. Cyclist and pedestrian casualties ubiquitously show near perfect correlation over time. A sudden break with correlation, in the very years when helmets became popular, is thus highly disturbing (Wardlaw, 2000). Analysis of US cyclist fatalities up to the late 1980s also found a positive correlation of deaths with helmet use (Rodgers, 1988). . This sweeping statement is not correct. Private cars are involved in approx. 50% of cyclist deaths, which is disproportionately low in relation to the amount of car traffic. Commercial vehicles, and in particular HGVs, are the greatest hazard to cyclists, especially in the urban setting. Typically victims suffer wide ranging crushing injuries against which a cyclehelmet can offer no protection. |
|
.
However, it is the quality and independence of research that determines how reliable it is, not its publisher. Much helmet research lacks the independence and rigour of best scientific endeavour. The two Cochrane reviews of helmets are no exception and they have attracted much criticism for their poor methodology and bias (BHRF, 1069; BHRF, 1243; Curnow, 2005; BHRF, 1181) . Some of this criticism is published by the Cochrane Collaboration as commentaries to their review. In both cases the reviews have been carried out by people well-known for their predisposition to helmet use and their activities in campaigning for helmet laws. In both cases the reviews have been dominated by the authors' own work. Both have neglected to consider the results of research unfavourable to helmet effectiveness while ignoring criticisms of the research that has been included. The scientific element of the controversy relates not to compulsion, but to the question of whether helmets have any useful effects at all.
.
However, mention should also be made of rotational impacts which may be the most common cause of brain injuries sustained by road crash victims that result in death or chronic intellectual disablement (Curnow, 2003). It should be noted that cycle helmets are not designed to mitigate rotational impacts and have not been shown to be effective in doing so. They may well make rotational injuries worse. (see also below)
.
This is not, however, necessarily the reality. The senior engineer of Bell Sports, the leading helmet manufacturer (Sundahl, 1998), and the Australian Federal Transport Bureau (Corner, Whitney, O'Rourke and Morgan, 1987), have commented that the foam in helmets may be too stiff so that it does not crush as designed.
.
EN1078 is a weaker standard than its predecessor, BS6863, a result of helmet manufacturers seeking to make helmets more acceptable for wearing in terms of comfort and style. Nevertheless, many helmets on sale today fail to meet the EN1078 specification. Helmets that meet the more demanding Snell standards are very difficult to find in the UK (Walker, 2005). Helmet research carried out more than 10 years ago (e.g. BHRF, 1069) took place when standards were higher and is not reliable with regard to current helmets.
.
The DfT review has been extensively criticised for examining only one type of evidence and not including any papers sceptical of helmet effectiveness (Franklin, 2003; BHRF, 1245). Its detailed analysis (published separately from the main report) includes serious criticism of all the papers examined, rendering them unreliable evidence. The DfT has recognised the need to update the research basis in commissioning a new research programme starting in 2008.
.
The studies cited in Table 1 are vulnerable to a common weakness; they make no allowance for differences in riding behaviour and attitudes to risk between cyclists wearing or not wearing helmets, although other research shows this to exist. Detailed analyses of most of these papers have been made by the Bicycle Helmet Research Foundation (BHRF, 1047) and the results of the papers cannot be regarded as reliable, particularly when contradicted by most other types of evidence.
The Cook & Sheikh study misquoted helmet wearing statistics from the Transport Research Laboratory upon which its conclusions rely, and also included mathematical errors. More rigorous examination of its data shows that the fall in head injuries over the period examined was not related to changes in helmet wearing (BHRF, 1099).
The BMA briefing is itself selective in the studies it cites, not including any that are sceptical of helmet effectiveness. Most relevant to the UK are two studies of all hospital admission in England from 1990 to 2002 (Hewson, 2005b) and road casualty injuries from 1995 to 2002 (Hewson, 2005) that found no association between helmet use and head injuries. A third analysis on a similar data set came to the same conclusion (Franklin and Chapman, 2005), as has a study for the Scottish Executive (Scottish Exec, 2005).
The briefing also does not cite important research by the Transport Research Laboratory (St Clair and Chinn, 2007) that confirms earlier Australian research (Corner, Whitney, O'Rourke and Morgan, 1987) (itself building on other research) that cycle helmets can increase harm by increasing the likelihood of rotational injury (which leads to the most serious types of brain injury). Further analysis by the Bicycle Helmet Research Foundation has expanded on this analysis (BHRF, 1182; BHRF, 1183) while current commercial developments have been triggered by professional concerns about the harm caused by current helmets (Morgan, 2007; Phillips helmets).
Other important research sceptical of helmet benefit comes from reviews of helmet laws. These reviews are the best way of analysing the effects of helmets, because they cover fairly short periods when the proportion of helmet use rises rapidly. This minimizes the other variables that otherwise confound analysis. An analysis of all enforced laws (Robinson, 2006), based on the largest data set of any helmet research, found no benefit from increased helmet use, with evidence of an increase in risk of head injury for those who continued to cycle after laws.
There are many other analyses of trends in casualties over time that have found no evidence of helmet benefit. Any comprehensive review of the evidence would need to include at least some of these.
.
The Spanish paper (Lardelli-Claret et al, 2003) looked at cyclist behaviour in voluntary wearers and non wearers of cycle helmets from 1990 to 1999. Spain's helmet law, which is limited in its scope, started in 2004. The BMA briefing might lead the reader into thinking this paper considered risk compensation after Spain enacted legislation. The authors acknowledge in their conclusions that the paper does not adequately test for risk compensation. This study is not evidence against a risk compensation effect for helmets.
Those who voluntarily wear cycle helmets are more likely to be safety conscious than those who choose not to. They are less likely to break the rules of the road , are more likely to come from higher socio economic groups, to wear high visibility clothing and use lights at night (McGuire and Smith 2000), ride with other helmeted children in parks, playgrounds, rather than on city streets and in the USA be of white race (DiGuiseppi, Rivara, Koepsell and Polissar, 1989). A risk compensation effect could be drowned out by these factors.
In a controlled trial, children running an obstacle course in a gym were more reckless and finished the course quicker when wearing a helmet and wrist guards than without (Morrongiello, Walpole and Lasenby, 2007). The differences were quite substantial, with increases in risk taking of up to 60%, particularly among sensation-seeking children.
In the UK, a study (Halliday, White, Finch and Ward, 1996) found that some teenage boys and young men would ride faster and less carefully when wearing a helmet. However, the decision to ride less carefully is often sub-conscious rather than deliberate, such as riding on busier roads because a person feels safer. Monitoring for the DfT has consistently found much higher levels of helmet wearing on busy roads (for example: in 2004 the proportion of cyclists wearing helmets was 28.2% and 9.4% on major and minor roads respectively - Inwood, Whitley and Sexton, 2005). It has also been found that motorists take less care when passing helmeted cyclists (Walker, 2007).
.
It is surprising to see the study of Lee et al included by the BMA considering its many known and serious problems (BHRF, 1142). It claimed a reduction in head injuries that would not be plausible even if helmets were 100% effective; the paper is based largely on opinion and conjecture; it did not consider the effect on levels of cycling; helmet wearing was entirely self reported; the control and intervention towns were too dissimilar; the results conflict with routine road casualty figures for the area concerned; and no attempt was made to put the risk of head injury into perspective.The use of such poor research does not inspire confidence in the BMA briefing.
.
An increase in percentage helmet wearing does not necessarily imply a similar increase (if any at all) in the number of cyclists wearing helmets. That is because a high proportion of people react to helmet laws by ceasing to cycle. For example, in Victoria (Australia) the helmet law increased percentage helmet wearing by teenagers from 21% to 45%. However, the study data showed that only 30 more teenagers wore helmets compared with 623 fewer cycling (Robinson, 1996).
After some years of official encouragement of helmet wearing with wearing rates at 40%, one might think those that wish to don helmets would have done so. The other 60% will be quite aware of them and many will be making an active choice not to wear one. A law presents them with the choice of having to wear one or stop cycling. In Australia, New Zealand and Nova Scotia a large proportion stopped cycling (Robinson, 2006). Surveys in Australia have found the helmet law is the largest reason cited for not cycling more (Heathcote and Maisey, 1994; Blacktown).
. This research (Karkhaneh, Kalenga, Hagel and Rowe, 2006) was concerned only with changes in percentage helmet wearing; it did not address the question of whether helmets work or whether they have harmful effects. However, for most of the before/after studies the numbers size post-law were smaller (and in some cases much smaller) than the numbers pre-law, suggesting that cycle use may have fallen significantly.
.
Table 2 lists five studies. Three conclude legislation worked, two conclude it did not. No attempt at qualitatively appraising them is made, however some have received much criticism.
The 2002 Canadian study by Macpherson et al has important failings that make its conclusions untenable (Robinson, 2003b; BHRF, 1106). Most of the fall in head injuries in provinces introducing helmet laws occurred before the law was enacted, and trends in % of head injuries among both cyclists and pedestrians continued without apparent effect from helmet laws. In fact the helmet laws made no difference to the ongoing slight decline in head injury rates among cyclists.
The California results from Lee et al omits important facts from their data likely to have a bearing on results and used an inappropriate control group, introducing several confounding factors (BHRF, 1151). The authors conclusions seem based more on speculation than the data they present. They cite only research favouring helmet wearing, failing to mention studies contradicting this.
Robinson's 2006 BMJ paper (Robinson, 2006) has the largest and most statistically robust dataset of any study. It looked at 10,479 head injuries treated in hospital across all jurisdictions in the world that saw a 40% point increase in helmet wearing over a few months. Nova Scotia, New Zealand and the Australian states of NSW, South Australia, Victoria and Western Australia were included. Ontario, cited in the past by the Board of Science as showing that cycling levels are not affected by helmet laws was not included, because the law was not enforced and helmet wearing increased only a little before falling back to pre-law levels. Case controlled studies predict that a large increase in helmet wearing should translate into fewer head injuries among cyclists, but Robinson was unable to detect any clear trends. She did find unambiguous evidence of reduced levels of cycling but no apparent effect on percentage head injuries for a large increase in helmet wearing. The BMA briefing offers nothing to explain this important finding.
The Macpherson and Spinks 2007 Cochrane review is mentioned earlier.
A scientific briefing might reasonably have commented more on why the published research is so contradictory.
.
Data in the study (Hagel et al, 2006) suggests that cycling by children and teenagers declined significantly post-law compared with adults (59% decline for children, 41% for teenagers). This alone would explain the percentage increase in helmet wearing. In the three years prior to the Alberta law, the proportion of head injuries was relatively constant at just over 5%. In the six months following the law, 9 health regions reported an increase in the proportion of head injuries to above 10% for children and just under 10% for all age groups (BHRF, 1055).
.
This study covered the same area (East York, Ontario) as the much-criticised MacPherson et al study in Table 2 above (Macpherson et al, 2002), and two of the authors are the same. Most of the increase in helmet wearing took place pre-law and by 1999 (4 years after legislation) helmet use had fallen back to pre-law levels. Despite no net change in helmet use and no change in cycling levels, % head injuries for children fell from 40.6% to 21.2% (Macpherson et al, 2002; CIHI, 2003).
.
The studies only show an increase in percentage helmet wearing. The number of cyclists generally declines; in at least one case where the numbers are available, there was no increase of helmeted cyclists. Instead, unhelmeted cyclists were removed from the road.
.
The review presented evidence only of an increase in percentage helmet wearing and of a decline in the absolute number of head injuries. It also confused ongoing trends with the effects of laws. . Without good evidence of the effect of laws on cycle use, it is not possible to draw any conclusions about the effect of the laws on either cycling or injuries. The Cochrane review did not find any evidence about cycle use because it did not search for any. In fact there is much evidence of a strong connection between enforced helmet laws and substantial falls in cycle use (BHRF, 1020), but the authors of the review chose to make an unsupported assertion that no such evidence exists.
The 2007 Cochrane review has been the subject of much criticism as well as concerns about ethics and morality (BHRF, 1181). The principal author is a known campaigner for cycle helmet laws, clearly not neutral in her enthusiasm for them. One of the papers reviewed is her own, without reference to the many criticisms that have made it unreliable evidence. On the other hand, the most comprehensive work on helmet laws is omitted even from the literature search (Robinson, 2006).
The BMA briefing makes no reference to criticism of the review and it gives the impression of greater certainty about its conclusions than can be justified. Even the authors themselves acknowledge that “There is a paucity of high quality evaluative studies assessing the effect of helmet legislation on bicycle related head injuries: only three were identified for this review.” This requires conclusions from the review to be heavily qualified, and even then to acknowledge that the data applies only to children.
.
The BMA's reference 31 appears to be a misquotation. Serious injury due to transport accidents, Australia, 2003–04, Oct 2007. Australian Institute for Health and Welfare does not include Western Australian data, only that for Australia as a whole. It could not be the source of the author's statistics about head and upper limb injury.
Percentage head injuries for Western Australia were as follows (Meuleners, Gavin and Cercarelli, 2003):
|
1988
|
1989
|
1990
|
1991
|
1992
|
1993
|
1994
|
1995
|
1996
|
1997
|
1998
|
|
31.8%
|
24.8%
|
25.5%
|
24.2%
|
20.2%
|
22.6%
|
19.6%
|
24.2%
|
20.6%
|
18.7%
|
20.6%
|
The helmet law was enforced from 1992. Although there was a significant fall in risk of head injury with time, control groups showed very similar trends. This is consistent with Robinson's findings (Robinson, 2006) that a large increase in helmet wearing did not decrease the risk of head injury among cyclists.
Official sources (Meuleners, Gavin and Cercarelli, 2003) also show that cyclists comprised 12% of road casualties in 1991 (pre-law), rising to 16% in 2000. Of total hospital admissions, cyclists rose from 17% in 1991 to 25.9% in 2000. There is no evidence of helmet law benefit here either.
.
The fall in cyclist casualties in New Zealand has occurred against a background of large reductions for all road users between1989 and 2006. Between 1990 and 2004 all road user deaths fell from 21.4 to 10.4 per 100,000 (IATSS, 2006). The data presented by the BMA briefing is misleading unless these large underlying trends are also mentioned.
.
See previous comments about the qualitative and bias issues of the Cochrane Review, which did not seek out evidence on cycle use. It makes an unsupported assertion.
The BMA has previously accepted there have been large decreases in cycling in New Zealand and Australia coinciding with helmet laws (BMA, 1999). Other places with enforced helmet laws have seen large falls in levels of cycling too: Nova Scotia - 40% to 60% (Chipman, 2002), British Columbia - 28% (Foss and Beirness, 2000), Alberta - 41% to 59% (BHRF, 1176). It is striking how little effort there has been to audit the effect on cycling levels.
There is also evidence that helmet promotion, without legislation, leads to falls in cycle use by making the activity appear more hazardous than is the case. In the UK, those local authorities that had campaigns focused on the promotion of helmet wearing, were found to be “strongly linked to a decrease in the number of cyclists observed” (Bryan-Brown and Taylor, 1997).
In Denmark, helmet promotion was found to be one of the measures to have contributed towards a much less positive attitude towards cycling by parents and children, resulting in a 30% fall in the number of children cycling to school from 1993 to 2000 (Jensen and Hummer, 2002).
.
This statement is not substantiated by the reference given. Sheikh et al's belief is that reductions in cycling are temporary, but they cite no material to back this up in their paper. This reference amounts to another unsupported opinion. If the author knows of evidence suggesting cycling levels recover post law he should cite it so it can be appraised.
In Western Australia, cycling numbers returned to pre-law levels after seven years in absolute numbers but were still considerably depressed in 2006 relative to population. Bridge surveys in the capital city of Perth show just 6.5 percent more cyclists in 2006 than 1992, despite population growth around 23 percent over the 14 years and average petrol prices increasing from AUD 0.68 to AUD 1.13 per litre (WA, 2). The recovery that did take place was mainly due to an increase in leisure cycling; cycling for everyday purposes and by children had not recovered. There was also the loss of more than a decade's cycling growth (levels were increasing pre-law). Moreover, cyclist hospital admissions rose to record levels in the late 1990s and in 2000 were approximately 30% higher than the pre-law average (~ 260 more admissions) (CHC, 1).
.
This argument is unfounded speculation and it contradicts the strong and reliable evidence of reduced cycling after helmet laws. Would the enforced wearing of pedestrian helmets be likely to make walking more popular? Or would the enforced wearing of bullet proof vests in South Central Los Angeles increase public confidence it is any safer to go there?
.
The BMA fails to recognise the large natural falls in road fatalities that occur due to general road safety measures, rather than specific interventions like cycle networks. For instance, between the early 1970s and currently, cyclist deaths in Britain fell from 400 annually to about 140 currently, a decline of 65%. The level of cycle use in the early 1970s was similar to today, according to the National Travel Survey and on-road traffic counting by the Department for Transport. This large fall in deaths occurred without any particular intervention, apart from cycle helmet promotion, which as previously noted, is linked if anything to increases in deaths, not decreases.
Amongst non-cyclists, or cyclists who lack proficiency, there is a view that adding segregated cycle networks will make cycling safer. This view is not confirmed by real-world experience (Franklin, paths). Most studies of segregation conclude that, at best, cycle paths improve safety between junctions, only to increase risk at junctions due to increased complexity and confusion over priority. Cycle routes must be continuous and provide long stretches without junctions across side-roads, otherwise they will hinder riders and increase the risk of collision. All expert cyclists have concluded that skilful riding, sharing road space on the public highway, is the safest type of cycling per unit distance travelled, and the only practical way to cover significant distances (Franklin, 2007).
It has already been emphasised that the most effective cycle-specific intervention to improve safety is to increase the number of cyclists.
Denmark, the Netherlands and Germany have the best safety records and highest amount of cycling in Europe. None of these countries has high helmet wearing rates (Pucher and Beuhler, 2008). In the Netherlands, the government has rejected the promotion of cycle helmets, stating it is more likely to put people off cycling than make cycling safer (Pucher and Beuhler, 2008), though there is promotion in Germany and Denmark. It is striking that countries that have been successful in getting people cycling more do not have high helmet wearing rates and countries that have enforced helmet laws continue to have low participation rates. A 2001 report by the SWOV (Dutch Road Safety Institute) for the European Council of Transport Ministers states (Wittink, 2001):
"From the point of view of restrictiveness, even the official promotion of helmets may have negative consequences for bicycle use, and that to prevent helmets having a negative effect on the use of bicycles, the best approach is to leave the promotion of helmet wear to manufacturers and shopkeepers."
A recent review of cities that have succeeded in promoting more and safer cycling makes clear that helmets have not been a helpful factor (Pucher and Beuhler, 2007).
.
The BMA fails to make its case that there is any justification even to promote the use of cycle helmets, let alone make them compulsory. Some data are presented that superficially suggest cyclists bear a high risk of injury, and several unsubstantiated statements are made to that effect. Better understanding of STATS19 and HES injury records reveal that in fact cyclist injuries are a small element of road casualties and of all-causes injuries in any age group. Risk assessment of cycling in a number of countries, including Britain, shows that cycling, walking and driving all incur similar risks per hour of use when fair comparison is made. The risk in cycling is known to be strongly affected by the number of cyclists, with a fall in cycle use being highly likely to increase the risk of death per cyclist. The large falls in cycle use recorded in countries with helmet promotion programmes thus ought to be far greater cause for concern to those with a true commitment to improving the safety of cyclists. There is no reasonable case to focus on cyclists alone to be the target of helmet laws. For example, pedestrians face higher risks per mile travelled than cyclists.
Evidence about cycle helmets is contradictory and confusing. It amounts to observational studies which as Bandolier has commented are "great for raising questions, but not for answering them" (Bandolier) . The Centre for Evidence Based Medicine would place the case control studies used in helmet research in its class 3 category, or possibly class 4 (CEBM). Doctors have been fooled before by case-control studies (Petitti, 2004).
The use of anecdotes, or 'case studies' in this document is absurd. What part does this play in a scientific debate? Anecdotes are a favourite tactic of complementary therapists and advertising agencies. Their use is risible, not to mention one-sided.
Doctors, even A&E consultants, have no more wisdom about this subject than anyone else. Britain's A&E departments increasingly deal with the acute medical problems of a sedentary population. Along with, happily rare, cyclist trauma they see thousands of heart attacks, strokes and other diseases related to physical inactivity. More cycling, helmeted or not, might save them a lot of work.
Accident Analysis Group: Accident victims treated at the A&E Department, Odense University, Denmark 1998-2000. .
Andersen, Schnohr, Schroll and Hein, 2000
Andersen LB, Schnohr P, Schroll M, Hein HO, 2000. All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work. Arch Intern Med 2000 Jun 12;160(11):1621-8.
Bandolier Journal. .
How helmet promotion and laws affect cycle use. .
Commentaries on published evidence. .
Head injuries up after Alberta law?. .
Cochrane Review: Helmets for preventing head and facial injuries in bicyclists. .
Trends in serious head injuries among English cyclists and pedestrians. .
Bicycle safety helmet legislation and bicycle-related non-fatal injuries in California - Commentary. .
Assessment of current bicycle helmets for the potential to cause rotational injury - Commentary. .
Helmets for preventing head and facial injuries in bicycles - critique. .
Robinson DL, . Comments on Road Safety Research Report 30. 2003. .
Blacktown Bikeplan Study, Final Report. Blacktown City Council, Sydney.
Cycling towards health and safety. British Medical Association ISBN 0-19-286151-4.1992.
Cycle Helmets. BMA Board of Science and Education ISBN 0-7279-1430-8.1999.
Bryan-Brown K, Taylor S, 1997. Cycle helmet wearing in 1996. Transport Research Laboratory Report 286.
Cavill N, Davis A, 2007. Cycling and Health: what's the evidence?. Cycling England . 
Centre for Evidence-Based Medicine. .
Cyclist injury data before and after helmet law in Western Australia. .
Chipman ML, 2002. Hats off (or not?) to helmet legislation. Canadian Medical Association Journal 2002 Mar 5;166(5):602.
Ontario Trauma Registry 2003 Report: Injury Hospitalizations. Canadian Institute for Health Information, 2003.
Corner, Whitney, O'Rourke and Morgan, 1987
Corner JP, Whitney CW, O'Rourke N, Morgan De, 1987. Motorcycle and bicycle protective helmets: requirements resulting from a post crash study and experimental research. Federal Office of Road Safety Report CR55.
Curnow WJ, 2003. The efficacy of bicycle helmets against brain injury. Accident Analysis and Prevention 2003,35:287-292.
Curnow WJ, 2005. The Cochrane Collaboration and bicycle helmets. Accident Analysis & Prevention 2005;37(3):569-573.
Transport Statistics Bulletin: National Travel Survey 2006. Department for Transport, August 2007 .
The Use of Hospital Data in Road Accidents: Road Casualties Great Britain, Annual Report 2006. . Department for Transport.
Department for Transport traffic census data. .
DiGuiseppi, Rivara, Koepsell and Polissar, 1989
DiGuiseppi CG, Rivara FP, Koepsell TD, Polissar L, 1989. Bicycle helmet use by children. Evaluation of a community-wide helmet campaign. J Amer Med Assoc 262:2256-61.
Foss RD, Beirness DJ, 2000. Bicycle helmet use in British Columbia. UNC Highway Safety Research Center; Traffic Injury Research Foundation .
Franklin J, Chapman G, 2005. Quantifying the risk of head injury to child cyclists in England: an analysis of hospital admissions data. BHRF .
Franklin J, 2003. Bicycle helmet effectiveness - a broader perspective. .
Franklin, J, . Cyclecraft. The Stationery Office UK, 2007. ISBN 978-0-11-703740-3.
Cycle path safety: a summary of research. .
Hagel BE, Rizkallah JW, Lamy A, Belton KL, Jhangri GS, Cherry N, Rowe BH, 2006. Bicycle helmet prevalence two years after the introduction of mandatory use legislation for under 18s in Alberta, Canada. Injury Prevention 2006;12:262-265.
Halliday, White, Finch and Ward, 1996
Halliday ME, White C, Finch H, Ward K, 1996. Attitudes to cycle helmets - a qualitative study. Transport Research Laboratory Report 154.
Heathcote B, Maisey G, 1994. Bicycle use and attitudes to the helmet wearing law. Traffic Board of Western Australia May 1994.
Hewson PJ, 2005. Cycle helmets and road casualties in the UK. Traffic Injury Prevention 2005;6(2):127-134. 
Hewson PJ, 2005. Investigating population level trends in head injuries amongst child cyclists in the UK. Accident Analysis & Prevention 2005;37(5):807-815.
Hillman M, 1997. Health benefits of cycling greatly outweigh loss of life years from deaths. BMJ BMJ 1997;314:69.
IATSS research Vol. 30 No 1, 2006. .
Inwood, Whitley and Sexton, 2005
Inwood C, Whitley C, Sexton B, 2005. Cycle helmet wearing in 2004. Transport Research Laboratory Report 644.
Jensen SU, Hummer CH, 2002. Sikre skoleveje: En undersogelse af borns trafiksikkerhed og transportvaner. Danmarks Transport Forskning Rapport 3, 2002.
Karkhaneh, Kalenga, Hagel and Rowe, 2006
Karkhaneh M, Kalenga JC, Hagel BE, Rowe BH, 2006. Effectiveness of bicycle helmet legislation to increase helmet use: a systematic review. Injury Prevention 2006;12:76-82.
Lardelli-Claret P, Luna-del-Castillo J, Jiménez-Moleon JJ, Garcia-Martin M, Bueono-Cavanillas A, 2003. Risk compensation theory and voluntary helmet use by cyclists in Spain. Injury Prevention 2003;9:128-132.
Macpherson AK, To TM, Macarthur C, Chipman ML, Wright JG, Parkin PC, 2002. Impact of Mandatory Helmet Legislation on Bicycle-Related Head Injuries in Children: A Population-Based Study. Pediatrics 2002; 110(5):e60.
Matthews CE, Jurj AL, Shu Xo, Li HL, Yang G, Li Q, Cao YT, Zheng W, 2007. Influence of exercise, walking, cycling and overall nonexercise physical activity on mortality in Chinese women. American Journal of Epidemiology 2007 165(12):1343-1350.
McGuire L, Smith N, 2000. Cycling safety: injury prevention in Oxford cyclists. Injury Prevention 2000;6:285-287.
Meuleners, Gavin and Cercarelli, 2003
Meuleners LB, Gavin AL, Cercarelli LR, 2003. Bicycle crashes and injuries in Western Australia 1987 - 2000. University of Western Australia RR131.
Morgan J, 1988. Risk in cycling. TRRL WP/RS/75.
Morgan D, . Cone head helmets. Click the 'Inspiration' tab.
Morrongiello, Walpole and Lasenby, 2007
Morrongiello BA, Walpole B, Lasenby J, 2007. Understanding children's injury-risk behavior: Wearing safety gear can lead to increased risk taking. Accident Analysis & Prevention 2007 May;39(3):618-23.
Petitti D, 2004. Hormone replacement therapy and coronary heart disease: four lessons. Int Journal of Epidemiology 2004;33:461-463.
Pucher J, Beuhler R, 2007. At the frontiers of cycling:policy innovations in the Netherlands, Denmark and Germany. World Transport Policy and Practice 2007;13(3):8-57.
Pucher J, Buehler R, 2008. Making cycling irresistible: lessons from The Netherlands, Denmark and Germany. Transport Reviews 2008(4):495-528.
Robinson DL, 1996. Head injuries and bicycle helmet laws. Accident Analysis & Prevention 1996 Jul;28(4):463-75.
Robinson DL, 2003. Confusing trends with the effect of helmet laws. Pediatrics P3R 7 Jul 2003.
Robinson DL, 2006. Do enforced bicycle helmet laws improve public health?. BMJ 2006;332:722-725.
Rodgers GB, 1988. Reducing bicycle accidents: a re-evaluation of the impacts of the CPSC bicycle standard and helmet use. Journal of Products Liability 1988 ,11:307-317.
Extent and severity of cycle accident casualties. Scottish Executive Social Research .2005.
StClair VJM, Chinn BP, 2007. Assessment of current bicycle helmets for the potential to cause rotational injury. Transport Research Laboratory PPR213.
Letter to the U. S. Consumer Product Safety Commission, c/o Scott Heh, Project Manager, Directorate for Engineering Sciences, Washington, D. C, 20207. Jim G Sundahl, Senior Engineer, Bell Sports, 19 January 1998.
Bike numbers in Western Australia: government surveys. .
Walker B, 2005. Heads up - the science of helmets. Cycle Jun/Jul 2005.
Walker I, 2007. Drivers overtaking bicyclists: Objective data on the effects of riding position, helmet use, vehicle type and apparent gender. Accident Analysis & Prevention 2007 Mar;39(2):417-25.
Wardlaw MJ, 2000. Three lessons for a better cycling future. BMJ 2000-12-23/30 n321 p1582-5.
Wardlaw MJ, . Correspondence with Mr Jim Fitzpatrick MP, Minister for Road Safety and reply from Barbara Noble of the Department for Transport. June 2008.
Wittink, R, 2001. Promotion of mobility and safety of vulnerable road users. SWOV Report D-2001-03.