Why “Running Too Much May Be as Bad as Sitting Around” is BS

Its not a surprise that people are confused about the best medical practices to improve health.  Whether the recommendations are for diet or exercise it seems that every 3 months there is a contradicting advice from the medical community.  Eat eggs, don’t eat eggs, high carb diet, low carb diet (I still don’t know the right answer to that one).

In part, the flood of seemingly contradictory advice is a consequence of the how the media reports on medical research.  This seems especially true when the medical research contradicts what we had previously or instinctively known.  Its these studies and stories which gain traction through social media megaphones and are promoted by people that would rather not have to follow the existing medical advice.  The advice justifies their unhealthy habits of the non adherents, but any even worse consequence is that flurry in the media can scare those that had adhered to the good behaviors that they were doing into stopping those behaviors.  In was just this past week that several of my friends on Facebook (and in real life) brought the latest sensational headlines to my attention for scrutiny.  “Running Too Much May Be as Bad as Sitting Around” or similar was the headline in numerous papers last week (here, here, here, here). Just reading the title I knew that I was in for a real treat of stretching the data thin, but little did I know how thin of a ledge the authors conclusion rested on.

Running KILLS

Running KILLS – or if thats what some media outlets will have you believe

The problem with that headline is really two fold, one the study they are basing that information on is inherently flawed and two other much better studies show the opposite.  Lets start with the first point.  This particular study, published in the Journal of the American College of Cardiology (JACC), is part of a bigger study called the Copenhagen Healthy Heart study that had been going on for more than 10 years.  Over the 10 years the researchers have tracked more than 5000 individuals living in Copenhagen, noting all of their behaviors changes in risk factors, and eventually their survival. The goal is to find behaviors or traits that associated with lower cardiovascular disease and improved longevity.  Within the entire study the authors identified a group of regular runners (878) and within this group of runners were some that did a little running, some moderate running, and strenuous running.  They tried to compare these three groups with each other, but because they only 17 deaths in the 10 years in the 878 exercisers they could not make any robust conclusions about whether more or less exercise was better for you.  Instead they compared these runners to non-runners that more or less looked similar and found, not surprisingly, that those that did some or a moderate amount of running had a significantly lower risk of death (Figure 2, note that the confidence intervals are huge in the strenuous group, which is because of the insufficient number of subjects).

Good thing a third runner didn't die

Good thing a third runner didn’t die

What was a surprise, and lead to the sensationalize media headlines, was the strenuous group did not have a lower risk of death than those so called couch potatoes.  Sure on first glance, or first headline, this looks bad for those of us that run a lot (myself certainly included), but looking further into the data and its easily explained why they saw no difference.  There were only 40 runners in the run a lot. This is an insanely low number of subjects to make any conclusions on mortality from no matter what the conclusion. Remember that reproducibility starts with number of subjects.  So it only took two of those runners in the strenuous group did in fact die during the 10 year follow up period, for there to be no difference between them and the non-runners.

So not only dose this study have far too few subjects it goes against several studies with much larger numbers of subjects.  One example, also published in the JACC shows that as you increase the amount of running, be by time, distance, or speed there is a similar reduction in mortality (Figure 3).  That particular study had 55,137 adults in it. Yes it has about 40 times more subjects in it.

Mo' running, mo' living.

Mo’ running, mo’ living.

In a second study of Finnish twins with more than 16000 subjects published in the Journal of American Medicine Association found something similar with the group of most strenuous exercisers having the same risk as the occasional exercise group. Both the exercise groups had a lower risk of death.  I could go on, but will instead refer you to an excellent and detailed summary of where the entirety of the research really stands.  I recommend this article by Alex Hutchinson, a staff writer for Runners World (or this in the NYTimes) whom fully admits that there is likely an upper limit of beneficial exercise, just not the one the current study claims.  His article is balanced and takes into consideration all of the pre-existing studies out there, something the authors of the JACC study clearly don’t in their comments to the media.  Instead they put out sensationalized quotes which make their results look much more robust than they actually are.  I am not sure which is crazier, the fact that these scientists are quoted as saying as much or that there are journalists that actually believe them.  This study should have been an easy one for a journalist.  Its not a hard study to detect the flaws and if the media outlets from which we are suppose to trust cannot do so initially, then we the public must take it upon ourselves to do the due diligence concerning inflammatory headlines.  Yes, this means scientific literacy.  Yes, this means a healthy skepticism around reversal of what is currently believed.  Yes, this means that us (scientists) should not and cannot oversell our work merely to generate buzz.

Even with all those changes it will continue to be difficult to get the story right. There were more than 700,000 studies added to the U.S. National Library of Medicine database of scientific studies in 2013.  It’s a staggering amount of data and is overwhelming to keep up even for those of us whose job it is to do so.  Furthermore we can’t allow ourselves to become dogmatic to a point in which new findings are ignored and ostracized for merely being change.  So, we must achieve a balance with responsible scientific communication, responsible reporting, and the relative context provided when necessary an informed and well-educated public is possible.

We devote so much time to this sport of ours that we love, so naturally we are skeptically of data that suggests running might not be beneficial or may even harm us. While the data shows otherwise, we need to remain vigilant and realize running is not a panacea. We ourselves much not oversell.


5 thoughts on “Why “Running Too Much May Be as Bad as Sitting Around” is BS

  1. Observational epidemiology is great for generating hypotheses to be tested in randomized trials but its a stretch to draw any conclusive information from it, especially lifestyle recommendations. The Lee et. al. study you linked is an example of that.

    And one thing that’s still unclear: does running improve health or are healthy people more likely to be runners? Chicken, egg situation. If a sedentary person becomes a runner, would their health parameters improve to a level comparable to long-term runners? That’s what we don’t know. And when people start running, or start any new exercise regimen, they usually improve their diet and other lifestyle factors as well. Which of these is responsible for the health improvements? Or is it the combination of all of them? You’re right that we need to be skeptical about research. But that includes data that appears to support our preconceived notions.


    • Its a fair point. Reductionism can only take you so far in figuring out the “ideal” lifestyle factors. These do not act in isolate and certainly interact with each other. As far as running specifically, it is clear that when you start exercising you become more healthy and when you stop exercising you become less healthy, but not to the same level as people who have always been healthy.
      https://dl.dropboxusercontent.com/u/5187409/Screen%20Shot%202015-02-15%20at%2010.22.23%20PM.png The question is, does it really matter that if its because you make other changes to your lifestyle as well?

      Sure, while I agree you should be skeptical and all population based recommendations will “miss” the personalized levels of recommendations MOST effective for an individual. For some maybe diet (of some sort), for others exercise (of some specific type), for most a combination of the two. That doesn’t mean people should not act. There is plenty more evidence that inactivity causes a plethora of chronic diseases. http://www.ncbi.nlm.nih.gov/pubmed/21836048 The evidence point to even small amounts of exercise in a variety of ways as clinically significant.

      I am not sure if you are skeptical about that science? If so, what level of evidence would convince you otherwise?

      • Data from randomized trials should drive our discussions, not associations from observational data. Too much room for error in making inferences from correlation.

        Questioning if exercise can improve our health (rather than noting that healthy people are more likely to exercise) is hugely important. While it seems like common sense to get the 200 lb diabetic housewife out jogging, research into the metabolic effects of endurance exercise might dampen our enthusiasm. Consider that plenty of runners arrest out in the field and plenty more die of an MI after coming home and showering. Also consider that plenty of people who can’t run a hundred feet (and who have never exercised) live well into their nineties.

        Remember that fitness and health aren’t the same thing. Maximal oxygen uptake is great to measure in athletes in labs but it doesn’t very well predict how long you’re going to live. Improving fitness doesn’t always have corresponding improvements in health markers. There’s so many other confounding factors to consider. A person can run eighty miles a week and still have diabetes, hypertension, rhythm disorders, some subclinical artery disease, arthritis (presumably from running but we need to test that further), etc. and also be overweight.

        There are safer, more efficient ways to exercise and achieve these benefits than long-distance running (or endurance anything). The literature on resistance exercise is growing and is really interesting. And this is coming from a former runner.
        (If this is posted twice, I apologize)

      • I appreciate the post.

        You can’t say you require randomized clinical trials to drive discussions and then cite a bunch of anecdotal evidence. I have never heard of a person developing type 2 diabetes who runs 80 miles a week. Even as an extreme example that one is pretty far out there.

        There is certainly evidence that in specific populations exercise does not improve the primary outcomes (http://www.nejm.org/doi/full/10.1056/NEJMoa1212914) and increasingly evidence about the potential dangers of long duration endurance exercise on cardiac arrythmias (I hope to do a post on this at some further point), however that is not the case. Exercise is not a panacea and some clinical conditions and not going to be reversed by exercise.

        For cardiovascular disease the epidemiology is supporting by the biological data in the form of intervention studies. http://circ.ahajournals.org/content/107/24/3109.full In prevention of Type 2 diabetes the data is even stronger http://www.nejm.org/doi/full/10.1056/NEJMoa012512. Those two diseases alone are major killers of the Western world. Osteoporosis, breast cancer, colon cancer all have extremely strong data in support of exercise as an effective preventive measure.

        I 100% agree that exercise for health does not always improve fitness and am writing a small paper with data that supports this at the moment. Most of the health benefits from exercise are likely to occur from a relatively small amount of exercise, which may not increase VO2max. I also agree that strength is an important part of a healthy lifestyle, but to say that it is safer or MORE effective in preventing heart disease isn’t yet supported by the data and can’t warrant a comment like “there are safer, more efficient….”. Its just simply wrong to say so without any supporting data. If anything resistance and endurance exercises are complimentary rather than redundant types of exercise.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s