Will A Lot of Running Kill You?


A manuscript was published yesterday in the Journal of the American College of Cardiology that concluded, in a nutshell, that people who participate in strenuous running, in terms of intensity or volume, have the same rate of long-term mortality as sedentary people.   Let’s look at the study further.

This study was performed with retrospective data accumulated by the Copenhagen City Heart Study.  In this study, 1098 healthy “joggers” and 3950 healthy “nonjoggers” were followed.  9% of overall participants were age 50 or older.  Participants were excluded for a history of coronary heart disease, stroke, and cancer.  From 2001-2013, there were 28 deaths among joggers and 128 among nonjoggers (the causes of death were not recorded).  Participants rated their physical activity on a graded scale of one to four.  One: almost entirely sedentary.  Two: light physical activity 2-4 hours per week.  Three: vigorous activity for 2-4 hours per week or light physical activity for more than four hours per week. Four: high vigorous physical activity for more than four hours per week.  Joggers were further subdivided into three subgroups based on “dose” of jogging: slow (5 miles per hour, less than 2.5 hours per week, less than or equal to three times a week), moderate (5-7 miles per hour, between 2.5 and 4 hours per week), and strenuous (greater than 7 miles per hour, more than 4 hours per week, more than three times per week).

The results included the following findings:

  • Participants who ran 1-2.4 hours per week had the lowest rate of mortality.  In fact, their rate of death was 71% lower than for sedentary participants in the study.
  • Participants who ran less than 1 hour per week had a 53% lower rate of mortality than the sedentary group.
  • Participants who ran 2.5 to 4 hours per week, or more than 4 hours per week did not have a lower rate of mortality.
  • The “optimal dose” of frequency of jogging was 2-3 times per week, associated with a 68% lower rate of death compared with the sedentary group.
  • Jogging more than 3 times a week was not associated with a lower rate of mortality compared with the sedentary group.
  • Slow joggers had a 49% lower rate of death than sedentary participants.
  • Strenuous joggers had the same rate of death as sedentary participants.

Overall, the lowest mortality was in light joggers with regard to hours per week, frequency, and pace of jogging.  Moderate joggers had a significantly higher mortality rate compared with light joggers, while strenuous joggers did not have a statistically different mortality rate than sedentary participants.

This is a thought-provoking report.  But it is important to understand the limitations of the study.  For example, the data used is self-reported (by study participants) and is subject, therefore, to error.  This study is retrospective.  The best studies are prospective, in which participants are given an intervention to follow over time (light jogging, moderate jogging, etc.) and then the effects are followed.  In the case of a retrospective study, there are unknown reasons why participants end up in each group and this can affect results.  The study is also weakly powered:   there were only 47 participants who jogged more than 4 hours per week and only 80 who jogged more than three times per week.  What this means is that the number of participants in each group (sedentary, light joggers, etc.) are highly uneven and it is much more difficult to measure the statistical significance of differences between such uneven groups.  The risk of under-powering the different arms of research studies, by the way, is one more reason why a prospective design is superior.  Another weakness is the lack of reporting the cause of death.  What if, among the 28 joggers who died during the course of the study, there were 8 participants in the moderate and strenuous groups who died after being hit by cars (their rate of death, therefore, being higher because they were outdoors running for more hours and more often)?  Would the differences in the groups still be clinically significant?  Would runners, health care providers, coaches, and epidemiologists really attach much concern to such a difference in rate of mortality if it could be explained by motor vehicle accidents?  Finally, I am neither a statistician nor a cardiologist, but I am concerned about those media reports of this study that have used the term “increased risk of death.”  Retrospective data gives information about RATE of death not RISK of death.  This is a very important distinction.

So is this study wrong?

Absolutely not.  I am confident that the authors gave an accurate report of the data they collected and analyzed.  My concern is about over-interpretation, in the media, of retrospective, self-reported, under-powered data, leading to changes in lifestyle.  Please let me be clear: the authors may have identified an important association between intense running and rate of death, but this study has important weaknesses.  If you want to use the results of this study, or other information you receive in the media, to alter your exercise program, please discuss your plans with your health care provider.

Oh no...This form doesn't exist. Head back to the manage forms page and select a different form.

Published February 3, 2015


Schnohr P, O’Keefe JH, Marott JL, et al. Dose of jogging and long-term mortality. J Am Coll Cardiol 2015; 65:411-419



Leave a Reply

Your email address will not be published. Required fields are marked *