Last week, on a rare evening out sans les enfants, the Cyclist mentioned that she had read a news article earlier in the week on research that showed too much running, more than two and a half hours a week, reduces life expectancy. Shooting from the hip I opined that, even if it were true, approaching 40 miles / 5 hours 20 minutes a week as I am now (the highest levels in my running career to date) I would much prefer to die a little early having lived an active and healthy life rather than eke out a few extra years osmotically sharing my putrefacting humours with the sofa. Since this didn’t entirely allay what turned out to be genuine concern for my wellbeing I assured her I would check out the article and the research behind it.
It was easy enough to find a BBC News article reporting the research and follow the link provided to the Journal of the American College of Cardiology (JACC) where the scientific article was published. A quick review of the publicly available parts of the JACC article immediately revealed the flaw within the conclusions expressed by the researchers as follows:
Light and moderate joggers have lower mortality than sedentary non-joggers, whereas strenuous joggers have a mortality rate not statistically different from that of the sedentary group.
Whilst I don’t pretend to be fully competent in the statistical methods used by the researchers this much is clear from the adjusted figures provided in the Central Illustration in the article: The mortality conclusions are based on statistical comparison with a control sedentary group. The sedentary group consisted of 394 people of whom 120 died over the period of the study. The light joggers group consisted of 570 people of whom 7 died and the moderate group consisted of 252 of whom 8 died. So far, so good. Reasonable sample sizes and hence reasonable statistical conclusions.
The strenuous joggers group consisted of only 36 people of whom 2 died. This is not a decent sample size by any definition and drawing the conclusion that there is no statistical difference from the sedentary group based on this appears mindless at best. (Bear in mind that the mortality data is all cause. The researchers had no access to cause of death and therefore no opportunity to make any assessment of whether activity or inactivity could possibly be a contributing factor.) A more honest conclusion would have read:
Light and moderate joggers have lower mortality than sedentary non-joggers. The sample size of the strenuous joggers group is too small to make meaningful statistical comparison of the mortality rate with the sedentary group possible.
So far, so bad science. I can’t begin to understand the motivations of the researchers concerned if they consciously chose to add the fatuous element to their conclusion. Unless perhaps it was to make it newsworthy? I hope not, I prefer to think that they were just foolish, but if so they certainly succeeded as is evident from the sample below:
- “Fast running is as deadly as sitting on couch, scientists find” – The Telegraph, 2 February 2015
- “Stop that binge jogging! Three times a week is best for you… and too much is as bad as doing nothing” – Daily Mail, 2 February 2015
- “Too much jogging may be as bad for you as not running at all, study suggests” – Independent, 4 February 2015
Whilst I was composing the opening lines of this post, the BBC News Article was amended, changing the beginning of the title – from “Too much jogging …” to “Training very hard …” – but leaving the remainder “… as bad as no exercise at all” and adding the following text to the body of the article:
“This is a small study, particularly when it comes to the people in the most active groups – only 36 were classified as “strenuous” joggers and just two of this group died. So experts caution this makes it harder to detect and be confident of the differences between each group.”
There may have been other significant amendments, but if so I didn’t notice them. (An un-refreshed browser tab caught the change in title, but not the article itself.) Even the additional text barely puts any distance between the BBC author(s) and the conclusions of the researchers and adds little journalistic rigour. Some of the other articles have also been amended since they were first published and some of those amendments appear to have been made to add equally token distance from the conclusions of the researchers.
I really don’t expect journalists to be scientists themselves any more than I expect them to be mathematicians, I am neither myself, but I do expect the application of a basic scientific approach to published conclusions expressed in everyday language just as I would expect a basic mathematical approach to, for example, party political manifesto income tax calculations. It took me literally 5 minutes to go from BBC article to journal article and find the flaw in the conclusion. A journalist has a responsibility to be at least as competent and then, if they choose to report such a non-story at all, to report it in constructively critical terms. Whilst the potential impact on health is not as directly serious in this case as was the original naive reporting and lending of undue credibility to the now infamous MMR vaccine research, it seems little has been learned and mainstream media are apparently to be relied upon only to regurgitate ‘random stories they read on the internet’ without critique.
As is my wont I’ve taken several days to come up with this searing riposte. So many responded much more timeously than I and most of them more elegantly. Of those my favourites are “Strenuous jogging ‘as bad as doing no exercise’ claim” – NHS Choices, 3 February 2015; a carefully even toned and thorough piece. And for a much more meticulous rebuttal, along similar lines to my own, from within the running community – “The (Supposed) Dangers of Running Too Much” – Alex Hutchinson, RunnersWorld.com, 3 February 2015. Linked to in that article, I also found Alex Hutchinson’s previous writing on the same topic, “Will Running Too Much Kill You?”, RunnersWorld.com, 3 April 2014, enlightening.