I almost forgot to post my reaction to the “Bicycle Helmets Put You at Risk” article from last weekend:

For years, cyclists who ride on city streets have cherished an unusual superstition: if they wear a helmet, they are more likely to get hit by a car. “I belong to an e-mail list for cyclists, and they complain about this all the time,� says Ian Walker, a psychologist at the University of Bath who rides his bike to work every day. But could this actually be true?

Walker decided to find out — putting his own neck on the line. He rigged his bicycle with an ultrasonic sensor that could detect how close each car was that passed him. Then he hit the roads, alternately riding with a helmet and without for two months, until he had been passed by 2,500 cars. Examining the data, he found that when he wore his helmet, motorists passed by 8.5 centimeters (3.35 inches) closer than when his head was bare. He had increased his risk of an accident by donning safety gear.

Several issues jumped right out at me:

1. Risk is related to the countermeasure to vulnerabilites and assets, not just threats (e.g. likelihood of being hit by a car). Your head is, in fact, more protected from a proper helmet than without. Having survived many crashes myself, I can attest to this fact personally. So whether or not the likelihood of an incident decreases, your brain is far more likely to survive an actual incident in a helmet.
2. “Risk of an accident” is misleading. What percentage of all bicycle accidents are related to being hit by a car? I found some really old data that suggests only 17.5% of all bicycle accidents involve an automobile. Of course it depends, right? If you ride desolate dirt mountain trails, would you remove your helmet to reduce the chance of being hit by an automobile? So the study should clearly exclude the 80% or whatever number of bicyclists are not at risk from autos, even if a threat reduction were possible.
3. If the theory is that “helmets change the behavior of drivers”, then why should the answer be to remove helmets (significantly increasing risk related to vulnerability — see #1) instead of mandating helmets for everyone and reducing the chance of special treatment from drivers? If everybody wears them, then after a settling period helmets could not be accused of changing behavior.
4. Did the study control for other factors like time of day and road, type of clothing and bike, and/or gestures from the bicyclist to the motorist? The study could just as well prove that wearing hot pink socks or a “one less car” jersey makes drivers more likely to side-swipe you than putting on protective gear.

I know I’ve completely destroyed (foam cracked all the way through or seriously dented) at least two, maybe three, bicycle helmets but I have yet to make contact with a car. Perhaps that colors my perspective. Anyway, might be worth a read to see if/how the more complete risk picture is addressed.