Here you go; helmets became mandatory in 2003 and the difference is very clear – used to be years between the deaths of pro cyclists during a race, now it rare for a year to go by without at least one and maybe more.
Yes, that site was the first thing I found when I saw your claim further upthread. How are you interpreting that to show more deaths since mandatory helmets?
21 deaths between 2003 and present.
1-no helmet
6-heart attacks
1-collision with 18 wheeler
3-vehicle collision
1-blunt force trauma to chest
1-high speed on decent
Interesting how helmets have increased the number of heart attacks occurring during bicycle races too.
Of course what is missing is the denominator. Competitive cycling exploded over that time frame. Amazingly when the denominator is included the rate of serious injuries is unchanged.
Given that bikes have gotten faster and more fragile (carbon fiber), pelotons larger and more crowded, and that the fields in competitive races are twice the size they used to be (with poor visibility of the road ahead due to so many other racers) I am personally shocked that the rate has not increased, even with helmets.
Helmets can only do so much at the speeds these cyclists travel at.
Best to avoid cherry-picking studies and go with a solid meta-analysis. Cochrane does 'em best albeit they are often labelled as having an excessively high bar of evidence. As far as helmet legislation effectiveness goes they conclude:
Data however of limited quality. And no data if such legislation decreased the number of those cycling and thus increased deaths secondary to decreased fitness.
In a typical year 900 deaths occur during cycling. Huge high risk for the typical recreational or commuting cyclist? No. Risk-free also no.
I can work in my job with a separated shoulder or a broken collar bone just fine. I need my brain though. I will continue to wear my helmet, my high viz clothing, and use my defensive riding habits.
Cochrane does cherry-picking best? It is very widely criticized, one reason being they cherry-picked data, using mostly their own previous studies in their analysis.
A couple example with links to many more:
Cochrane review of bicycle helmets: an unsound guide
Cochrane Review: Helmets for preventing head and facial injuries in bicyclists
To take the tangent from this thread further, I think there is not much disagreement here that mandatory helmet laws are a fools errand. As stated, there are millions, perhaps billions of people who ride a bicycle every day all over the globe without a helmet, and are not any more advantaged or disadvantaged regarding safety than anyone else.
However, I think there is a subtle difference that may not have specific data. For example, those millions of people who ride every day without a helmet, perhaps they are using the bike to haul goods or raw materials from/to a small factory, or using a bike to run a quick errand to the corner grocery, or just cruising over to the coffee shop down the street or across a college campus. Let’s say these types of uses are at speeds <10mph.
Then there are your typical North American and European lycra-set, who are using the bike for exercise and fitness, or as transportation to their office job, or are a competitive racer in training, who navigate 1st-world urban and suburban areas while sharing space with other cars and trucks. Let’s say these uses are at speeds >10mph.
My question: the speed of cycling - does that have anything to do with the actual or even perceived benefits of helmets?
FWIW: I cannot ride any bicycle, any distance, for any purpose, without putting on a helmet. I feel weird without it.
Unless you run into a wall, the critical speed is the drop of the head to the ground.
I think this underscores the plausibility of the risk compensation hypothesis, don’t you? Perhaps “feeling weird without it” might mean that you’d ride extremely cautiously and defensively without the perceived protection of the helmet.
Yes, similar to seat belts in cars. I am a lot less cautious when I am belted in, so let’s stop wearing seat belts, right? 
It is of note that a even a site like the one you site that is devoted to coming to the conclusion that helmets cause harms has to start off with the admission that
Yes. They do. Having been doing evidence-based medicine for longer than the phrase has been around their reputation is very solid with the criticisms mainly being that sometimes the bar for acceptable evidence is unreasonably high.
If I lose control making a turn on wet pavement and hit the ground when riding at 18 mph I hit the ground harder than if I fall off my bike at a stop.
More likely “feeling weird without it” means the same as my feeling weird if I was walking around outside in my underwear … it’s just not something I do.
The data on helmets protecting children - who ride at slower speeds - from serious head injuries is great.
At racing speeds I am not sure how much good they do.
To me the biggest variable is the fact that faster speeds, and road cyclists interacting with drivers, are more likely to have accidents.
Gravity is gravity. Unless something gives you an extra acceleration when you wipe out, your forward speed should not affect how fast you go from upright to ground.
True.
Forward speed does however effect how fast your head implants into the side of a tree, a curb, or a car.
Covered that.
If you are imagining a hypothetical circumstance with a perfectly horizontal surface with no friction then it may be true that the only force at play is gravity. The real world is not that extremely hypothetical circumstance.
No it is not true that speed does matter unless is running into a wall. It matters unless you are in an imaginary hypothetical circumstance.
Unfortunately, there’s no source for their information.
Endos are not all that uncommon: drop your front wheel into a deep pothole or fail to miss that possum and you find yourself diving headfirst into whatever is in front of you. In those cases, your forward speed will contribute significantly to the impact of your helmet or head. In the case of losing grip in a turn, that usually means you are leaning into it, which means you are a bit closer to the ground, and the fall vector is mostly across the direction of travel, so speed does not contribute as much to the fall.
The problem is that your head is right there up front. Recumbent bikes are better that way because, as the 'bent riders are wont to say, life is too short to be riding into stuff headfirst.
Not disputing you, just expanding on your work …
Good find. The quote can be summarized as:
“Most bike crashes involve falling off onto flat pavement. In those crashes, and only those crashes, forward speed is almost entirely immaterial.”
Which is completely legit physics. As you and several folks have said.
The key point being the quote explicitly makes no claim plus or minus about the nature of crashes involving something other than flat pavement. No claim at all.
A casual reader might assume they’re dismissing non-flat crashes as not being a risk. That would be a gross misunderstanding. They’re simply saying non-flat crashes are out of scope for what they want to talk about in that article. Which is a very different thing.
I’m not competent to assess the “most crashes are flat” part. For all I know they may be 51% or they may be 95%. Or even only 20% and the quote’s antecedent is flat false. I doubt this latter case, but that’s conjecture, not evidence.
The other thing is that if we’re assessing total risk of the activity we ought to weight the crashes by severity. The harm stemming from 99 blisters plus one beheading is different from the harm from 100 skinned knees. Or even 100 broken legs. Yet all three scenarios are 100 bad events. Saying something like “the accident rate is 1 per 10,000 bike miles ridden” sweeps that info under the rug.
Sadly many academic research papers and consultant’s position pieces are written based on the stats they can find, not the stats that’d be necessary to truly derive useful real world predictions. Understanding the limits of applicability is difficult and it’s usually *not *in the interest of the author(s) to make those limits clear. We need to tease those limits out for ourselves. While avoiding both the reality and the appearance of “cherry-picking”.
There’s little doubt that severity is (roughly) inversely exponentially distributed. IOW, lots of blisters per beheading. But I’d bet there’s enough severity out there that an adult riding on urban or suburban roads receives a very real and statistically obvious benefit from a helmet. But we’ll only know if we actually collected the data needed to run the stats.
Whether a racer on a track or a child on a playground receives a benefit is a very different question. Which will have a very different answer. But we’ll only know if we actually collected the data needed to run the stats.
My overall take, Turble, et al, notwithstanding would be that bike helmets reduce the harm of intermediate severity crashes. Both flat pavement and those involving obstacles, even those as minor as curbs. Considering that curbs are within a couple feet of any non-rural bicyclist nearly 100% of their riding time they are a ubiquitous risk factor easily overlooked.
Yup. It’s an assertion with no support.
Let’s imagine an experiment.
You have two identical golf ball sized spheres of modelling clay.
Clay ball 1 you drop from three feet onto pavement and it hits the ground at about 10 mph.
Clay ball 2 you throw such that it hits the ground with 10 mph force in the vertical axis and 17 mph in the horizontal axis.
Both pretty much stop where they made contact.
Did those clay balls absorb the same amount of force on impact? Will they be deformed to the same degree?
Experiment two.
An egg is dropped vertically such that it it hits a horizontal surface of pavement at a velocity that is just below what would break the shell. Quarter of an inch? Whatever.
Then it is arranged for it to hit the pavement at the same vertical velocity and with horizontal velocity of 17 mph?
Do you believe that it will be at no additional risk of breaking the shell in the latter condition?