This blog has moved

This blog has re-located to Chester

Friday 15 April 2011

Rune Elvik’s Bicycle Helmet Efficacy Review

A new review of helmet research has recently been published by Rune Elvik in Accident Analysis and Prevention. The paper can be found here, although it may be behind a paywall.

The way that more mainstream media outlets report science is often interesting. Monday’s Science article about a new drug molecule demonstrating a 15% reduction in tumour mass in a specific type of cancer in rats all too often becomes Tuesday’s Daily Mail article about a new miracle drug to cure cancer. Quite often, if a story doesn’t fit the news narrative*, it is quietly ignored altogether. In the UK, the standard line from the mainstream press is that cycling is an extreme sport, like rock-climbing or sky-diving, and anyone who doesn’t use a helmet is either partially or fully to blame if they are killed by a speeding and/or drunk driver. Unsurprisingly I have not seen this new research mentioned in the mainstream press. picked up on the article, and ran a surprisingly scientific piece on it, although they did fall into the usual media trap of adding “balance” to science reporting, by quoting pro-helmet Professor Alistair Woodward, head of the School of Population Health at the University of Auckland. I am glad they consulted a relevant researcher rather than a non-expert as is surprisingly common in science reporting (look out for interviews with homeopaths being added to “balance” reporting on non-placebo medicine articles**). However, Professor Alistair Woodward was quoted as saying;

“Whether they cause the neck to bend more than otherwise, I suppose it's possible. If there is an effect [on neck injuries], it's much smaller than the protective effect from head injuries.”

Despite the paper, a review article of the body of research on the effectiveness of helmets rather than an individual study, stating that,

“When the risk of injury to head, face or neck is viewed as a whole, bicycle helmets do provide a small protective effect. This effect is evident only in older studies. New studies, summarised by a random-effects model of analysis, indicate no net protective effect.

In doing so, the article put the opinion of one pro-helmet man at the same level as an academic research review. It is hard to base too much on a single study, as there will be limitations in the design of any single study. This is why reviews of the existing literature such as this are a useful tool, they help to even out the limitations and biases of the individual studies. Sadly, because the review doesn’t fit nicely in the news narrative of “helmets are good,” it was felt necessary to include the opinion of one pro-helmet individual in an article about a review of the existing helmet literature.

The study was also picked up on by the New Zealand Herald (who interviewed Professor Alistair Woodward), who in the opening paragraph describe this review of the existing body of literature as “Contentious,” to immediately attempt to discredit the work, presumably because it doesn’t sit well with the existing news narrative. The pro-helmet bias in the NZ Herald piece continues, throughout, and culminates in the following bullet points at the end of the article:


* New research indicates wearing a helmet reduces the risk of head injury in a crash by 43 per cent.” [Neglecting to state the paper’s finding that the increased risk of neck injury brings the benefit of helmet wearing to a net of zero]

“* Previous research found the risk reduction was at least 60 per cent.

* The new findings are disputed.” [Because the NZ Herald went out to ask a pro-helmet individual to dispute the findings, and treating his opinion on the matter as having equal merit to the research to the contrary in the paper]

Bike Biz also ran a piece on the paper, largely based on the NZ Herald piece, focussing heavily on the opinions of Professor Alistair Woodward, and describing Elvik’s review of the existing body of literature as “controversial.”

The review focuses on the effects of helmet use on cyclists in the aftermath of a crash, finding there to be no overall net benefit to helmet wearing cyclists over non-helmet wearing cyclists in the aftermath of a crash. It would be interesting to see some quality research into risk compensation; whether the crash frequency and severity of helmet wearing cyclists is different from that of non-helmet wearing cyclists. I wouldn’t be surprised if the inclusion of such research into a review such as this wouldn’t bring the benefits of helmet wearing down from net zero benefit to a negative benefit to helmet wearers. However, this is just my opinion.

As I have stated previously, I see helmet wearing as a choice and I don’t have anything against anyone who choses to, or not to wear a helmet whilst cycling, walking, cooking or driving a car. It is perfectly reasonable to expect individuals to make decisions based on their own subjective fears, whilst simultaneously I’d prefer governments to make policy decisions based on objective risk. As a politician, Norman Baker’s recent defence of his choice to ride without a helmet is setting a good example rather than a bad one, by making an objective decision to not wear a helmet based on the minimal risk involved in cycling, he is portraying cycling as the safe, normal everyday activity it should be seen as.

* For more on the interesting peculiarities of mainstream news, I recommend Charlie Brooker’s entertaining and informative TV series Newswipe

** For more on the distortion and mis-reporting of science in the mainstream press (and more science-related stuff), I recommend Ben Goldacre’s entertaining Bad Science articles, ironically published in The Guardian


  1. I am extremely grateful for posts like this one and the recent statement from Portlandize. I will admit it: I am not emotionally prepared to tackle this topic on my own blog; I just can't do it. Thank you for writing this, for spreading this information.

  2. Thanks for posting this - your link to the paper doesn't work for me, but I think I've found the paper here...

    I'm also a fan of Ben Goldacre's work, alongside Ian Walker and his very practical approach to investigating driver reaction to helmets,

  3. @Mike

    I had to write something when I saw how badly this was being reported online. For more on this topic, there are some great posts going up at the moment on At War With The Motorist.

  4. Risk of injury to head, face, or neck is not that interesting a statistic, even if it is not entirely clear how Elvik calculated it, given the likliehood of a great deal of correlation between the types of injuries.
    What is interesting is the risk of fatalities. The original study gave a statistic and Elvik reevaluates that too. Odd that you failed to mention it. Perhaps it doesn't match your narrative?-)

  5. Woodward's point that your brain will be better off in a helmet during a collision than not is valid.

  6. @Unknown (Feb 9 2012),

    The difficult part is how you go about reconciling that with both the fact that your neck will not be better off, and that you may be more likely to be involved in a collision due to the effects of risk compensation.

  7. @Anonymous (Jun 30 2011),

    The risk of injury to different areas is an interesting statistic, as despite the potential correlation between the injuries to adjacent areas, it is important to consider that helmets may effectively be shifting a tiny amount of the damage to the head into the neck instead, whilst still offering only a marginal benefit to the head in the first place. As for the fatality figure, the effect is so tiny as to deserve little attention without further work to identify whether the cause is the helmet or one of any number of other factors which could produce a between-group variation of such a small size.


This blog has moved to Chester. All the old posts can be found at Chester Cycling where I invite you to continue the discussion instead

Note: only a member of this blog may post a comment.