Association Does Not Equal Causation

Here is the claim: “Running 5 minutes a day could add years to your life.”

And here is how it was reported on the Time Magazine Website:

According to a new study published in the Journal of the American College of Cardiology, running 5 minutes per day can reduce an individual’s risk of premature death by about 3 years. Researchers found that people who ran less than an hour per week also saw an increase in lifespan, not just a decrease in risk of premature death. The study took place over the course of 15 years, testing participants ranging in age from 18-100.(http://time.com/3053081/running-daily/)

The basis for the headline is an article published in the August 2014 Journal of the American College of Cardiology (ACC) written by researchers from several universities (including the University of South Carolina) who used a database of information about people who were treated at the Cooper Clinic and the Cooper Institute in Dallas, Texas, since 1970.  The some of the information in the database was collected by asking Cooper patients to complete a questionnaire about their physical activity; for example, what kind of exercise do they do, how often, and for how long. They are also asked health related questions about smoking, alcohol consumption, diet and so forth. 

The research question for the investigators was “what behaviors of the sample population were associated with a reduction in premature death?”

First, they selected a random sample 55,000 records that were taken at least fifteen years ago, including both men and women, ages 18 to 100, with a mean age of 44 years of age.

The investigation proceeded to do a search of death records to identify the subjects who may have died, the age of death to determine whether the death was premature, and the cause of death. 

It was found that 3,413 of the 55,000 subjects had died of all causes with 1,217 deaths attributable to cardiovascular factors.

The behavior that was associated with a reduction in premature death was running. Running was associated with a reduction of 30% in deaths from all causes and 45% reduction in deaths from cardiovascular causes.

The surprise finding was that it didn’t seem to matter whether the running was done for fewer than 51 minutes per week or at a pace of less than 6 miles per hour (10 minute miles). In addition, the runners who had identified themselves as “persistent” runners, had a reduction of 50% fewer premature death attributable to cardiovascular problems.

The statistical analysis of the data identified correlations among the factors. Correlation finds factors that vary together; an increase (or decrease) in one factor is associated with an increase (or decrease) in another. In this case, running was associated with a decrease in premature death.

Correlation is a powerful tool but it comes with a caveat: correlation can identify the strength of an association but it cannot tell anything about why two factors are related. The caution is therefore not to attribute the relationship to cause-and effect. 

So while physicians and health care workers should certainly share the association to encourage the large numbers of Americans who are sedentary to become more active, the actual mechanisms are yet to be fully researched.

Resources:

You can find the abstract and additional information about Leisure-Time Running Reduces All-Cause and Cardiovascular Mortality Risk on the American College of Cardiology Website.

A technical discussion of correlation may be found in Stephen Jay Gould’s The Mismeasure of Man, Revised and Expanded Edition, W.W. Norton & Company. New York. 1996. pp. 269-285.

Advertisements

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