More helmet use = more cycling deaths

MichaelBluejay · 2011-06-09 23:43:07

Many of you are familiar with the fact that as helmet use surged in the 1990's, head injuries among cyclists actually skyrocketed. On a simple basis head injuries went up by 10%, but when you factor in that cycling decreased by 27% during the same period, head injuries went up a whopping 51% on a per-cyclist basis. (NYT)

One reader recently complained to me that, he said, the reason that head injuries went up is that all those new cyclists with head injuries would have simply died had they not been wearing helmets. Those who would have died otherwise simply because head-injured, instead, making the number of head injuries increase (or so the argument goes).

As usual with these types of complaints, he didn't bother to look up the numbers. I did, and I found that his theory is dead wrong. Head injuries didn't go up because would-be fatalities became head-injuries instead. Biking fatalities shot up too, right along with injuries.

Cyclist deaths went from 843 to 728 from 1991 to 2001. (Traffic Safety Facts 2001, PDF, p. 17) However, as per the NYT article above, the number of people biking went down 27% during this period, so we would expect the 843 deaths in 1991 to shrink to 615 deaths in 2001 even without increased helmet use. But as helmet use surged, deaths didn't drop to 615, they actually went to 728. That's effectively an 18% increase in the number of cycling deaths as helmet use surged.

And for those who complain that we can't look at specific years because of year-to-year variance, let's compare the five-year-average number of deaths from 1987-1991 and 1997-2001. That gives us 879 to 750 deaths, or an effective increase in biking deaths of 17% as helmet use surged.

Of course the question becomes, why did head injuries and fatalities go up as helmet use went up? We don't know, but the take-away is, for whatever reason, helmets aren't the panacea for safety that their opponents claim. Anyway, the most common theories for the increase in injuries and deaths are:

(1) The more cyclists, the safer it is to cycle. So when cycling use went down, danger went up.

(2) Helmet use seem to increase the odds of getting hit, according to multiple studies. Drivers apparently give a wider berth to unhelmeted cyclists because they appear more vulnerable.

For more on the studies regarding helmet efficacy, I recommend the Bicycle Helmet Research Foundation.

bizikletari · 2011-06-10 11:32:07

A hypothesis:
During that same period SUVs gained popularity.
SUVs scared people out of the streets, thus decreasing the bikers numbers.
When SUVs got the newly helmeted bikers, they either killed them or they increse the amount of head injuries.

Maybe.

john the blasphemer · 2011-06-10 12:11:14

When I got my drivers license back in 1981 I had to pass the written portion of the test. Then, I had to drive around with a scary looking DPS officer. The guy was in uniform complete with gun and aviator glasses. I never saw his eyes. I passed on the first go because my mother had made me practice with her in the car for weeks ahead of time.

In 2005 my daughter got her drivers license. She went in, took the 20 question multiple choice test and walked out with a license. I actually did spend time teaching my kids to drive, but I know that a lot of parents don't bother. Drivers ed is no longer done at HS. Parents can elect to teach their own kids and just sign off saying they actually made the effort.

I'm not sure if this is how it is done in other states. I do wonder if the lack of meaningful drivers ed and practical testing has made a difference. I guess looking at changes in overall motor vehicle accident rates might shed some light.

Last edited by john the blasphemer (2011-06-10 12:11:47)

jma6610 · 2011-06-12 07:31:12

These studies clearly don't seem to be accounting for the number of miles ridden yearly by the cyclists. Although I don't know, I strongly suspect that even if the number of cyclists has decreased, the miles ridden has gone up and thus the exposure to accidents will go up. Bikes have become much more high-tech over the years and this has made them both objectively easier to ride for longer distances and more comfortable, and also more appealing to those who simply like to ride something that they are proud of and feels good to them. I also suspect that none of these studies differentiates between the strong recreational (or even professional) cyclist who is riding many hundreds or thousands of miles on more heavily travelled roads each year and the teenage cyclist who is riding around the block twice a week. I suspect that more of the latter are turning towards skateboards and other similar hobbies but that this trend doesn't much affect the adult cyclist who is riding more miles now than before. I rarely hear about a child or teenage rider being hit by a car, but often hear about the more serious adult cyclist who is riding on more heavily travelled roads being hit. I've also heard that more serious road riding has increased in the US since Armstrong become popular (many people I know seem to get motivated during the TdF each year). The studies also do not seem to be accounting for the number of miles driven by the motor vehicles each year. More miles by either cars or bikes will create more accidents if all else is equal.

Another thing to consider is that making comparisons across years by using the percentage change in accidents (or whatever) can be somewhat misleading). E.g., if there is a 1 in 1 million chance of something happening and that increases by 80%, this means that the new rate is only 1.8 in 1 million. ...still a very small number.

What the message should be in these studies is that cycling can be dangerous in some ways and that danger might be increasing somewhat in recent years (although that danger is offset by increased health that will make one live longer and healthier). So you need to protect yourself by wearing a helmet.

dougmc · 2011-06-12 13:11:34

MichaelBluejay wrote:

Many of you are familiar with the fact that as helmet use surged in the 1990's, head injuries among cyclists actually skyrocketed. On a simple basis head injuries went up by 10%, but when you factor in that cycling decreased by 27% during the same period, head injuries went up a whopping 51% on a per-cyclist basis. (NYT)

Such statistics are hard to come by and often inaccurate with large systemic errors. If you're looking at "head injuries" across many different years, the criteria of what qualifies as a "bicycle related head injury" are likely to change. Certain hospitals may not initially track injuries very well, and then start. Some may initially decide that something that merely needs stitches doesn't count -- and then it does, etc. So many variables.

And in general, only hospitals track injuries, possibly only emergency rooms. So if you don't go to the doctor -- no injury recorded. Or if you go to your family doctor -- probably no injury recorded. So things like how many people are insured and the idea that "even minor head injuries should be checked out" play a big role in this.

Which is why so many studies look at deaths rather than injuries -- even if "safety" is more than "not dying", deaths are consistently recorded, injuries are not. Deaths even have investigations done into what caused them, though they're not always very thorough. And if somebody is hit by a car and dies, it's often not recorded in the head injuries caused the death, or even if they could have.

One reader recently complained to me that, he said, the reason that head injuries went up is that all those new cyclists with head injuries would have simply died had they not been wearing helmets. Those who would have died otherwise simply because head-injured, instead, making the number of head injuries increase (or so the argument goes).

Yes, that argument would be stupid, considering how few die and how many times more injuries there are -- with or without bicycle helmets.

And for those who complain that we can't look at specific years because of year-to-year variance, let's compare the five-year-average number of deaths from 1987-1991 and 1997-2001. That gives us 879 to 750 deaths, or an effective increase in biking deaths of 17% as helmet use surged.

Nationwide, there's enough cyclist deaths to make meaningful comparisons from year to year. And ideally you'd compare adjacent years to reduce changes in other factors -- how effectively things are recorded, number of cyclists on the road, etc. Ideally, you have some watershed event happen between your two time periods -- usually a mandatory helmet law. But you have to consider that this law affects more than helmet use -- it will probably reduce the number of cyclists on the road by some amount, hospitals may start tracking head injuries and bicycling deaths more carefully, researchers may start trying harder to quantify how much cycling is done and what percentage of riders wear helmets (which will change the results compared to those taken last year), etc. Such a law affects a lot of things, and they all will muck with the statistics.

And usually mandatory helmet laws only cover a city or perhaps a state. But if you just look at that one area, there's not enough deaths to make accurate statistical conclusions for just one year which makes using this data to come to a good conclusion difficult.

(1) The more cyclists, the safer it is to cycle. So when cycling use went down, danger went up.

This is logical. However, ultimately it's an argument against mandatory helmet laws, not against helmet use.

(2) Helmet use seem to increase the odds of getting hit, according to multiple studies. Drivers apparently give a wider berth to unhelmeted cyclists because they appear more vulnerable.

As far as I know, there was only one study (this link seems to give the most information on it), and while he did find that motorists passed slightly closer when one wore a helmet, whether or not this lead to an increase in the chance of being hit was not determined (though he was hit twice, wearing a helmet both times -- but that's not a large enough sample size to come to a conclusion) and the reasons for this were only guessed at. (I haven't read the study, just articles about it.) He also found that looking female seemed to have a significantly larger effect on passing distance -- so perhaps we need to make helmets that have long hair to make one look like a girl?

For more on the studies regarding helmet efficacy, I recommend the Bicycle Helmet Research Foundation.

It is indeed the best site on the issue, but it has a strong anti-helmet bias. Like other sites, they cherrypick their data and studies and dismiss others out of hand if they mention them at all. I'm not really aware of any site that gives a balanced view of the issue -- but cyclehelmets.org does come the closest to that.

I still think the only way we're going to really determine how effective bicycle helmets really are is to test them in real world crashes. Take 1000 cyclists and come up with 500 real world typical accident scenarios (do actual crashes where people were seriously injured) and recreate each one, once with a helmet and once without. Smash cyclists into cars, into trees, under cars, into each other, etc. Carefully measure and quantify the resulting injuries. Unfortunately, volunteers for such a study might be hard to find. (Though we might be able to get some good data from crash test dummies with the right sensors in the right places -- especially their heads. People don't quite flop around like rag dolls in collisions, but that might be a close enough approximation.)

Either way, such a test would be expensive to do properly (and unethical if really done properly!) with nobody to really benefit, so nobody is going to pay for it, and so it's not going to happen.

m1ek · 2011-06-15 12:31:41

dougmc wrote:

(1) The more cyclists, the safer it is to cycle. So when cycling use went down, danger went up.
This is logical. However, ultimately it's an argument against mandatory helmet laws, not against helmet use.

A fair number of the anti-helmetters, including yours truly, believe that the incessant braying about helmets makes fewer people ride as if the behavior is portrayed as "so unsafe you have to wear a helmet or you're a Darwin award candidate", some non-trivial number of people will conclude it's too unsafe, period.

dougmc · 2011-06-16 23:24:22

m1ek wrote:

dougmc wrote:
(1) The more cyclists, the safer it is to cycle. So when cycling use went down, danger went up.
This is logical. However, ultimately it's an argument against mandatory helmet laws, not against helmet use.
A fair number of the anti-helmetters, including yours truly, believe that the incessant braying about helmets makes fewer people ride as if the behavior is portrayed as "so unsafe you have to wear a helmet or you're a Darwin award candidate", some non-trivial number of people will conclude it's too unsafe, period.

OK, but then that's an argument against "incessant braying about helmets", not helmets themselves. A cyclist quietly riding his bike with a helmet on his head doesn't make cycling more dangerous or turn people off of cycling.

(I guess it's *possible* that somebody might see him and go "hmm, he has a helmet. cycling must be too dangerous -- I'm not going to do it!" but that seems incredibly unlikely. I would imagine that his own cycling would have a larger positive effect than any negative effect his simply wearing a helmet might have.)

(And just so it's clear, the context for my statement was that the question "why did head injuries and fatalities go up as helmet use went up?" -- I was just pointing out that increased helmet use didn't cause cycling use to go down -- that was the mandatory helmet law which caused both.)

m1ek · 2011-06-21 11:53:54

dougmc wrote:

(I guess it's *possible* that somebody might see him and go "hmm, he has a helmet. cycling must be too dangerous -- I'm not going to do it!" but that seems incredibly unlikely.

You must not have as much experience with the suburban mindset as you presumably think you do. Helmet = dangerous. It's not 'unlikely'; it's quite likely; and it at least fits the observed facts (people think cycling is a lot more dangerous today than they thought it was 30 years ago, for instance).

Bicycle Austin forum

#1 2011-06-09 23:43:07

More helmet use = more cycling deaths

#2 2011-06-10 11:32:07

Re: More helmet use = more cycling deaths

#3 2011-06-10 12:11:14

Re: More helmet use = more cycling deaths

#4 2011-06-12 07:31:12

Re: More helmet use = more cycling deaths

#5 2011-06-12 13:11:34

Re: More helmet use = more cycling deaths

#6 2011-06-15 12:31:41

Re: More helmet use = more cycling deaths

#7 2011-06-16 23:24:22

Re: More helmet use = more cycling deaths

#8 2011-06-21 11:53:54

Re: More helmet use = more cycling deaths

Board footer