Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Monday, October 26, 2009

Deaths during running: Is exercise safe, Part 2

Running and safety continued - some comparisons, and the key point to the debate

One of the best things about this site is that often it is a source of information for me as much as I hope it is for you! And in response to last week's post about the safety of running, we've had some great comments and more information, which warranted a follow up. Also, I felt I should re-emphasize the purpose of that post, which was really a call to the media to change the view they project of running.

And most importantly, it was a message to runners out there to help them 'filter' out what is basically over-hyped reporting about deaths during running, akin to the "shark attack" phenomenon, where sensational reporting skews our beliefs over the relative safety of an activity (surfing in that case, running in ours). Perception does not necessarily equal reality, in other words, and the post was a call to question very hard what the risk of running is. If you do, you'll find that it's not nearly as high as is sometimes portrayed!

And now we have some points of comparison, thanks to you readers!

Host a marathon every weekend

First off, a study last year from the British Medical Journal took a rather creative (Malcolm Gladwell-esque) look at marathon safety. Canadian researchers compared the risk of dying in a motor vehicle accident with that of sudden cardiac death during marathon participation. Because courses are closed to traffic for larger races, it's possible to ask how many motor accident deaths would have been prevented as a result of the race, and compare this to actual deaths (thanks to Bengt, Tony and Amby for raising this one! As an aside, the author Redelmeier seems to be an expert on driving fatalities - another 2008 study published in JAMA is called "Driving fatalities on US presidential election days")

The diagram below summarizes the main finding:

So basically, the study found that the closure of roads for the major marathons prevented an estimated 46 deaths as a result of motor vehicle accidents. However, 26 sudden cardiac deaths were reported during those races (based on newspaper reports, it must be pointed out, and thus potentially a slight under-estimate). The relative risk - 35% lower when marathons happened.

And a final really important point - the authors have controlled for the accidents simply be relocated to other areas, and so marathon closures do not simply shift the site of the accidents. If you want the paper to check out their solid stats, please just let me know, as usual.

The author's conclusion is shown in italics above, but it was that "Organised marathons are not associated with an increase in sudden deaths from a societal perspective, contrary to anecdotal impressions fostered by news media."

Some numbers

Interesting that they picked up on the "impressions fostered by news media", which is basically the point of these last two posts. Interesting also that in 14 million hours, there were 26 deaths, which amounts to approximately 1 per 540,000 hours, a figure which agrees with that which our friend Amby Burfoot put forward in his comment to our last post. For a similar analysis, check out Amby's report on marathon deaths written last year, which is far more comprehensive than I've had time to do here - it's a big read, but if you're up for more details after this article, check it out.

This figure of 1 per ± 500,000 hours is also about the same as a (very) crude calculation would provide based on the information that about 6 deaths occur every year in the USA, and that about 3 million hours of running go into those deaths (this was all covered in the last post, if you want to read the numbers more).

My point, however, not covered by this study or the latest reports, is that marathon running hours are not limited to the hours of participation. There is substantially more time spent running in training, and by those who run but don't participate. If you factor all these people in, do the numbers change? Without quantifying training times, it may remain an unanswerable question.

These people, and their training, are the most important component of the debate, because they are most likely to be dissuaded from running as a result of the negative portrayal of running. Yet they are clearly, based on a large body of research by the likes of Paffenbarger (again, see the previous post), more protected than the sedentary population, and so should be hearing affirmation, not condemnation or warning for their choice to run. And if running in events that may have a risk of 1 death per 500,000 hours is the goal, then it too should be encouraged.

Comparing running with some other activities

To compare running a little better, have a look at the comments to the previous post, where you'll see some stats about how many deaths are caused by other activities. Smoking, for example, claims over 300,000 each year. Flu, 15,000, and car accidents, 20,000. Of course, the problem here is that these numbers don't indicate the 'exposure', or how many people spend how much time doing the activity. Smoking may simply be high because many people smoke, whereas 6 sudden deaths during running may be because hardly anyone runs.

Risk and exposure

So, for another comparison, I received data from a reader who had put together some stats on deaths per million hours of the activity. That piece can be read here, and the original article is here.

Admittedly, it's a little uncertain where this data were sourced, and in this field of epidemiology, that is crucially important. So with a proviso that the data is not "gospel", here is the summary list:

Deaths per million hours:

Skydiving - 128.71
General Flying - 15.58
Motorcycling - 8.80
Scuba Diving - 1.98
Swimming (presumably competitive) - 1.07
Snowmobiling - 0.88
Motoring - 0.47
Water skiing - 0.28
Bicycling - 0.26
Airline Flying 0.15
Hunting 0.08

Running? Depending on which number you believe, the risk during marathons is between 1.8 and 2 deaths per million hours, so it's around the same as scuba diving.

One problem - that doesn't factor in the health benefits, which I emphasized previously. Regular physical activity reduces the risk of cardiac disease and a host of other health problems, and so the risk is moderated by the benefit.

Most important of all - applying this dizzying collection of numbers to yourself

To end off (before we tackle some more enjoyable topics like coaching and science in the coming days), a word on applying this to you. I know I've thrown figures and numbers at you, and your head is probably spinning, so let's try to simplify this.

The reality, at the risk of sounding callous, is that people do die during running. A big race, with 30,000 runners, seems likely to experience such an event every 3 to 4 years. Put differently, between the six major marathons each year, there would be a sudden cardiac death each year.

People who are predisposed to cardiac death are more likely to die while running than while sitting on their couch. This is undeniable. But equally, people who run, including those who run marathons, derive enormous benefit from it - their health gains as a result of running are sufficient to over-ride habits like smoking in terms of overall risk of mortality (Paffenbarger, et al).

You can investigate whether you might be one of those predisposed, higher-risk people, but the problem, as we've discussed at length before, is that medical testing cannot currently identify all the people at risk. Some, certainly, and so medical screening, particularly if you are concerned, would be advisable.

Even in the absence of such 'confirmation', however, you still have a choice. With the risk at one death per 500,000 hours of running, and with the knowledge that running can improve your health, your choice is to remain sedentary and avoid that 0.8 in 100,000 runner chance, or you can run and benefit from the numerous positive adaptations you'll experience. It is a risk-management matrix, where running and remaining sedentary must be weighed against one another, benefits and risks understood, and a choice made. Your ability to manage the risks, by undergoing tests, by training and by adopting a healthy lifestyle, makes this choice far simpler than leaving it to guess-work.

And finally, for the media, physical activity should be encouraged, not 'demonized' with threats of death caused by activities like running. Sensational sells, but when it deters people from running, it becomes a problem. So some perspective, some affirmation and positive reporting would go a long way to fixing what is a growing problem of inactivity and obesity, subtly being driven by the media reporting. By all means, educate and inform people of how to maximize benefits, but let's not give a voice to those who view exercise as radical from the safety of their couches.


P.S. As mentioned, a series on coaching and the application of science starts tomorrow. Join us then!

Further reading:

Amby Burfoot, Editor-in-Chief of Runners World has done a comprehensive piece on safety of marathon running. I've tried to make this article (and the one before) more philosophical and directed at the media coverage, whereas his is full of information and 'hard facts'. But this is a great piece, and if you're up for more reading, this is a great read:

Amby Burfoot:


Anonymous said...

One of the many reasons I come back to this blog again and again is because I find it so stimulating.

I read what you guys, and other posters, say - it causes me pause to think. I then ponder for a while (sometimes for a long while) on what I've read (or think I've read).

Now, that brings me in a roundabout way to my point, which in five parts is this:

First, I have not come across any publication by reputable sports scientists or research exercise physiologists that concludes or suggests even that Homo should not compete in marathons. Am I right - there just aren’t any, or have I missed them?

Second, most of the anti-running polemic is therefore from, for want of a better term, ignoramuses. That is, they are ignorant about the exercise physiology of running; they are ignorant about sports science in general; they are ignorant of the history and current practise of competitive middle to long distance running.

Third, their ignorance doesn’t stop them sounding off, and the more ignorant they are the more loudly they sound off. (Read the latest posts on the NYT Well blog to see what I mean.)

Fourth, now for the rhetorical question: why is this? I know, one can just brush the phenomenon aside, one always finds ignorant gain sayers etc, but I think it’s important. This is because Homo’s perception of his world IS always important. For instance, in Anticipatory regulation of exercise, posted here last week right on this blog, Ross has summarised his convincing thesis that the runner’s perception of his or her race effort is a critical factor affecting the racer’s pace. (I hope I've got that right Ross!)

Fifth, my summing up. I don’t think that the message about the importance of running to get or keep fit is getting through. Depressing, I know. But let’s recognise the truth here. In general, modern Homo chooses not to run; he doesn’t like running; he doesn’t want to run; ergo, in order to justify not running, to himself or to others, he claws desperately at every excuse served up to him (such as the recent sad deaths of three runners in the Detroit Half). And this is DESPITE the convincing research, for example published by Daniel E Lieberman and Bramble, that Homo has evolved to become (with the possible exception of race horses and dogs specially bred by Homo) to become the best adapted distance runner this world has ever known.

Conclusion: we need more research on why people don't run; more research on how, if possible, to motivate them to get them running.

Anonymous said...

Well, excellent text! Could not agree more. I would like to add 2 points that may be interesting to some.

1. One should try to calculate risk of travelling to marathon races. Let's assume that average person travels to race 10M by car. I suppose that actually risk of losing life in travelling to races is much higher than to get a heart attack with a racing number on.

2. The fact that our collegues didn't make it may be connected with the fact of getting close to the finish line when we all try and give our best to squeeze out few extra seconds. Remembering one sprint finish in a 1.5 hour race when I didn't want to give up the podium place, while a glance to the HRM made me aware of getting above 200 bpm, which surely poses a life treathening condition for a 40+ recreational "athlete". My point is that we should be more careful and set our max HR somewhat lower, thinking not about few seconds, but on many still to come.


Bellthorpe said...

Hunting 0.08?

I thought the purpose of hunting was to cause death on a larger scale than that.

fiona said...

To add to Kresimir's point - I would expect that the greatest risk to marathoners arises from their training if they train adjacent to or on roads (as opposed to the lucky ones who can train entirely off-road, in woods or parkland) - this is just based (loosely) on statistics for death rate of pedestrians (which in the stats includes anyone on foot) in the UK.
A more genreal point about the risk versus the health benefits, this echoes the debate around helmet wearing for cyclists where, if making helmet wearing compulsory, decreases the number of cyclists and the amount of cycling undertaken by the population, the overall result is a higher mortality irrespective of the influence of the helmets in mitigating head injuries (open to dbate in itself) due to the loss of fitness health benefits. People get so wound up in 'cycling is dangerous', not helped by the media interest in the deaths (e.g. due to 'bendy-buses' in London) that they cannot conceive of the benefits of the activity.

Meg & Dave said...

This is all awesome discussion. Thanks for the plug guys... not sure my blog is worthy though.

Relative risk is always fun to ponder but it seems to be a serious rabbit hole. I am a whitewater kayaking enthusiast, constantly pondering these issues. People's perception is that whitewater paddling is SUPER dangerous. "Not compared to driving to the river," I say to them. They are usually not convinced. With something like cycling on the shoulder of a road I prefer to not think about it because, like driving, most components of risk are completely out of our control and not really worth getting stressed out about.

Maybe that's what it comes down to, until the activity you are doing poses a risk that is tangible and affectable, trying to weigh the relative risk is just silly talk. Of course running is good for you... as long as you wouldn't otherwise be doing something more dangerous... whatever that might be. Rabbit hole anyone?

In terms of the media, they got caught up with one six year old boy who was obviously not in the balloon to anyone who understands how balloons lift things. Should we really pay any attention to the sensationalism called news that just keeps the sheeple watching commercials?

oh cynicism, will you ever get old?

leagz said...

If the Canadian benefit is calculated, as you state, just "Because courses are closed to traffic" during the race, then this seems to ignore the signifiant risk reduction in the fact that, both while racing and training, the runners are not driving, which a proportion of them would otherwise be doing, and thus relieving them of the risk of dying at the wheel.
Also, the death argument is but a tiny part of the picture, given that exercise promotes fitness and health, which is arguably a much more significant quality of life benefit than reduction of death risk.
Alas though, Colenso is right - you are preaching to the converted on this website/newsletter, and no-one who should be reading it is likely to be. Better to spend effort moving on to another topic. The media won't change its spots.

maryka said...

Great post as usual! Fun stuff too, I'll be reposting this link.

One thing I'm surprised at is that motoring, snowmobiling, and cycling all supposedly have fewer deaths per million hours than running. That seems really counterintuitive.

Ross Tucker and Jonathan Dugas said...

HI everyone

Thanks for the great comments as always!

Just one quick response to Leagz:

The last article was very widely read by the media, and that's kind part of why this follow up was done. It was also read by many of the big marathon doctors and the medical community who do share the media's opinion of how dangerous the sport can be!

So while most of the athletes like yourself agree and are "converted", a good many are not, and amazingly, many of them are doctors or scientists who read this blog. And the media who read it as well! So hopefully ONE person may now look more positively on the sport as a result of this!


Derek said...

How ironic, I found this piece of news just days after you asked for information on deaths in other sports


My sincere condolences go out to the family of Kevin Widemond.

Unknown said...

I have a tip for you. Check out this thread/poll about the safety of running on the roads.

It's something most runners do, and in the poll 75% of runners said they had either been hit by a car or had a close call. When told they were doing something dangerous, most denied that fact or just said they accept the risk.

I think this is so interesting because runners, who are always touting the health benefits of their sport, regularly "play in traffic" and risk serious injury or even death. I know of two runners who were killed by cars in recent years in the Atlanta area alone. And that's just the ones I know about.

So what is the relative safety of training on the roads? I think nobody really knows, but for the good of our sport it would behoove us to find out.

At any rate, it certainly made for an interesting topic rife with self-denial.


maryka said...

Hey Stan, do us all a favour and instead of hijacking the comments on this site, go back to trolling your thread on Training Peaks.


Unknown said...

Well Maryka, I certainly can't match your brilliant insights on this topic.

But in case you missed the point, the risk in our sport from people training on the roads may far outstrip any risk of a sudden cardiac event.


maryka said...


You posted a link to a thread where a guy named Stan (presumably you) wrote:

"The roads were designed for cars, not people. If you get hit it's your own fault."

If that's not trolling, I don't know what is. Again, please take it somewhere else, like back to that thread.


Unknown said...

Every online community has its norms. What I said over there was controversial, yes, but not unusual for that board. It's all in good fun. Furthermore, if I caused some people to rethink the relative safety of the places they train then I've done a good thing.

I assure you that my intent here is to bring this issue to light in the context of the current topic, which is the relative safety of our sport.

I've suggested topics for further discussion on this blog before and one suggestion even resulted in a four-part series on muscle cramps.
http://www.sportsscientists.com/2007/10/fluid-intake-dehydration-and-exercise_26.html As you can see, Ross even says to me, "Thanks ahead for your readership, and keep those ideas coming!"

So I do take issue with your pronouncement that I somehow have no business posting here.

In short, I try to be a good citizen. I'm sorry if you misunderstood my intent.