I cycle regularly to my
practice, partly to set a good example to my
patients, partly to keep my middle-age spread at
bay, but mostly because I enjoy it. There is good
evidence for the health benefits of cycling and,
anyway, without my spin, I would never hear the
Three hours of ordinary commuter cycling per
week – 15 minutes each way, perhaps three miles a
day or 30 miles a week – reduced death rates by
nearly half, about 40 per cent. (Andersen LB,
Schnohr P, Schroll M, Hein HO, All-Cause Mortality
Associated With Physical Activity During Leisure
Time, Work, Sports, and Cycling to Work. Arch
Intern Med. 2000; 160:1621-1628).
Even among keen sportsmen in this study,
cycling reduced death rates (pers. comm. Prof
Andersen). There are very few other interventions
that can have such a dramatic effect on health.
My family worried about the danger on the
roads. For careful riders the risk is real, but it
is small. In England and Wales, one cyclist dies
for every 23 million miles cycled. (Passenger
death rates: by mode of transport 1981-1999:
Social Trends 31. www.statistics.gov.uk.)
If I cycle 30 miles most weeks, for 50 years, I
have one chance in 440 of dying as a result. That
is not negligible, but it is many times less than
the health benefits.
I bought a cycle helmet. It seemed obvious that
it was a good idea. I wore it for a while. It was
hot, heavy, and sweaty. I did not enjoy wearing
it. I started looking at the evidence that cycle
helmets actually do much good.
It is easy to find sales literature from helmet
manufacturers (www.bellbikehel mets.com).
It is also easy to find well-meaning
enthusiasts who would like to save cyclists
(www.bhit.org). Genuine evidence is difficult to
extract, even from the professional literature.
Some hopeful figures are often quoted (“85 per
cent” and up). They turn out to have a very poor
scientific base. One early study compared helmeted
children, mostly white, well-off, and cycling in
parks, with unhelmeted children, mostly black,
cycling in busy urban streets.
The latter had more injuries, including head
injuries. No amount of statistical manipulation
can turn that sort of data into sound conclusions
on the effect of helmets. Unfortunately, they have
been widely quoted.
There are other and better case-control studies
in the helmet literature, e.g. Maimaris et. al.
(Maimaris C, Summers CL, Browning C, Palmer CR.
Injury patterns in cyclists attending an accident
and emergency department: a comparison of
helmet-wearers and non-wearers. BMJ 1994;308:
They are based only on cyclists presenting with
injuries. In these studies, those who said they
were wearing helmets were less likely to have
injuries recorded to their heads. (Thompson DC,
Rivara FP, Thompson R. Helmets for preventing head
and facial injuries in bicyclists (Cochrane
Review). In: The Cochrane Library, Issue 3, 2004.
Chichester, UK: John Wiley & Sons, Ltd.)
The odds ratios for protection are in the
neighbourhood of 75 per cent. That sounds good.
These papers often conclude that cycle helmets are
a good thing, which should be encouraged or even
made compulsory. However, there are words of
caution to anyone reading these studies.
They often use the term “serious” head injury
to mean anyone attending hospital, or admitted for
observation overnight. In a sense that is correct;
some of us have seen cyclists very upset, with
scalp grazes and gashes, and a large proportion
would be admitted to rule out intra-cerebral
bleeding. But very few actually suffer any
long-term effects, apart from increased wisdom.
More evidence is required to show that cycle
helmets are useful in the real world. It should be
possible to identify people who suffered serious
injury, or death, that could have been prevented
by helmets. And it should be possible to show,
after a large increase in helmet wearing, that the
proportion of head injuries has declined.
Deaths and design
In London, more than two-thirds of all cyclist
deaths were due to collision with a heavy good
vehicle, 97 per cent with a motor vehicle of some
sort. (Gilbert K, McCarthy M. Deaths of cyclists
in London 1985-92: the hazards of road traffic.
I have no access to post-mortem reports, but
one may doubt that any helmet could save the
majority of cyclists killed by a heavy goods
Cycle helmets are not designed to protect
against most serious impacts. Design standards
involve dropping them from a height of a metre or
so, directly onto a hard surface. The design
intention is that the foam should compress, thus
slowing deceleration of the head. Some of our
patients may produce stories about damaged
helmets, with the implication that the helmet has
avoided a much worse problem.
In fact, it requires very little energy to
break a helmet, even to smash it. Most broken
helmets have simply failed.
Controversy and wiggly lines
Cycle helmets arouse strong feelings and
violent disagreements. In one study of trends in
head injury, in a period when helmet use was
increasing, we read: “The wearing of a cycle
helmet is estimated to prevent 60 per cent of head
injuries” (Cook A. Sheikh A. Trends in serious
head injuries among English cyclists and
pedestrians. Injury Prevention. 9(3): 266-7,
However, another paper concludes that
“increased helmet wearing percentages has had
little association with serious head injuries to
cyclists” (Scuffham PA. Langley JD. Trends in
cycle injury in New Zealand under voluntary helmet
use. Accident Analysis & Prevention.
29(1):1-9, 1997). Why are there such direct
These studies are of trends in injury or death
ratios over time. Such ratios vary considerably
from year to year. In general, we cannot account
in detail for most of the variations.
Any statistician will point out that relatively
rare events, such as cyclist deaths, will
inevitably show considerable random variation from
year to year.
It can be easy to find an expected answer,
merely by selecting the time period of the study,
selecting the age-groups to be analysed, choosing
the variables to be adjusted for, and so on.
At www.cyclehelmets.org there is a re-analysis
of the data from one time-trend study. The
original concluded that a helmet law had been “an
effective road safety intervention”. The
re-analysis, over a longer time period, suggests
strongly that any apparent benefit is an artifact.
Costs of encouraging
Few studies make these explicit. But
enforcement has costs. The most obvious one is the
expense of helmets. As one enthusiastic retailer
put it: “Just think, an accessory that isn’t
particularly cheap and that everyone has to buy”.
That, at least, can be calculated, and if it is a
worthwhile expense, paid for by public funds.
But the main impact on health would come from
an effect on the numbers of people cycling. Even a
small decline in the numbers cycling would do more
harm than preventing all injuries to cyclists.
To repeat, cycling is good for health. It does
far more good than associated injuries do harm.
The work reviewed at www.cyclehelmets.org
strongly suggests that overall, cycling in
Australia declined by about a third after a helmet
law was passed. This would be enough to make the
law extremely bad for health, even if it prevented
all injuries to cyclists.
There are other worrying possibilities,
including the likelihood that the law may increase
risk to the remaining cyclists (Wardlaw M. Three
lessons for a better cycling future. BMJ
It has been suggested that “just one” serious
injury saved would be enough to justify any amount
of cost. No economist would agree. But no health
worker should agree either. The costs are not only
money, but are people dying with the consequences
of insufficient exercise.
With the present political worries about
obesity, politicians too may think it unwise to
criminalise people out for a healthy ride.
Some countries have forced cyclists to wear
helmets or stop cycling. A full analysis of the
natural experiments provided by these laws would
be difficult. It should include head injuries and
other injuries, to cyclists, pedestrians, and
other road users, in relation to the amount of
time they spend on the roads.
It should analyse these over a long
time-period, using as potential explanatory
variables not only changes in helmet use, but also
road safety initiatives. And it should include
In particular, putting people off cycling is a
threat to health far greater than any possible
benefit that helmets could provide.
I have not found any comprehensive analysis of
these factors in the literature. In many cases,
laws appear to have been declared a success
without serious evaluation. But some figures are
A graph (from www.cycle helmets.org) shows some
figures from Western Australia, which made cycle
helmets compulsory in 1991.
It is not clear why the ratio of head injury
among pedestrians showed changes around the time
of a helmet law. Nor is it clear why there is a
downward trend for all groups over most of this
period. But there is certainly no obvious benefit
from the helmet law.
There are few reasonably thorough analyses of
helmet lawsand some of the best attempts may be
found at www.cyclehelmets.org and
www.cycle-helmets.com. But, if the case-control
studies are a good guide to what happens in real
life, the benefits should be obvious.
There are no obvious benefits. It is necessary
to indulge in disputable selection of the data in
order to show any useful effect on any parameter.
On balance, and admitting that the evidence is
often unsatisfactory, cycle helmets do not show
the benefits that some have claimed.
Since I started looking at the evidence, I have
not worn my helmet. If I am worried about my
health, I should get on my bike. If I am worried
about accidents, I would follow the advice in
Cyclecraft (Franklin J, HMSO, 1997). If I am
worried about my patients’ health, I tell them to
consider how they could use a bike in their daily
lives, to listen to the skylarks again, and wear a
helmet if they actually want to.
At best, cycle helmets protect against a few,
mostly minor, injuries. At worst, and especially
when enforced, they contribute to the obesity
The evidence suggests that helmet laws are bad
for health. In a world which bombards politicians
with complicated and worrisome issues, here is one
with a simple resolution – ignore cycle helmets.
There is much that can be done to make cycling
a safer and commoner choice. Helmets are not a
useful part of that process.
Dr Richard Keatinge,