Originally posted to 13.7 Cosmos & Culture on 7/9/2017
The largest study of its kind — analyzing data from 24,763,389 results between 1996 to 2016 — has found that the average American runner, from 5k runners to marathoners, is getting slower.
The study was led by Jens Jakob Andersen and Ivanka Andreeva Nikolova from RunRepeat.com. Andersen is a former competitive runner and statistician from Copenhagen Business School. Nikolova holds a Ph.D. in Mathematical Analysis. The results include only races with more than 2,000 finishers. The researchers have data from 1902 to 2017 (May), but chose to publish only the trimmed data set from 1996 to 2016. Results can be trusted at a 99 percent confidence level (a statistical measure based on the sample size) and don't include top elite American runners, who are faster than ever.
The graph below summarizes the results for marathons. On the vertical axis, we see the average finish times, while in the horizontal axis the different years. The straight line is a best fit between all points. There is a clear upward trend with the years, which means that finishing times are growing longer on average. The years 2015 and 2016 are the slowest in history.
Although the numbers differ, the upward trend is similar across all four distances. The interesting question, of course, is why. After showing the data, the authors disprove a few possible answers. Here is a quick list:
- The proportion of women participants is increasing. This would mean that since women tend to be slower runners in general, if there are more women running, the average finishing times will tend to increase. Analyzing the data by gender, Andersen and Nikolova show that the rise in the number of female runners has less effect (46 percent) in the slowing pace than the decrease in the speed of male runners (54 percent), concluding that men are becoming slower much faster than the increase in the number of female runners.
- People with inappropriate fitness level just walk the race. Since the average walking pace is roughly 19 and half minutes per mile, Anderson and Nikolova show that across the four distances the proportion of runners finishing with times comparable or slower to a walking time has remained constant through the years. So, an increasing number of walkers doesn't explain the data either.
- Slow runners are getting slower. Here, the idea is that if slower runners are getting slower, they will push up the average finishing time over the years. Anderson and Nikolova studied the average finishing times of the 100th, 1,000th, 2,000th, and 5,000th finishers to see if, indeed, the slower finishers were taking longer to finish the races. They found that the times were slower across the board and with similar trends, be it the 100th or the 2,000th. On average, women slowed down by 9.87 percent over the past 17 years, and men by 9.94 percent.
With these arguments out of the way, the question remains: Why are Americans getting slower?
I exchanged a few emails with Nikolova about this. They believe that the reason is the overall decline in health of Americans, which is reflected in the finishing times of races across the four distances. They looked for correlations using data from the Center for Disease Control and Prevention, in particular, the increase in adult and teenage obesity, diabetes and hypertension, and overall annual medical expenditure.
As the authors write:
"We correlated the average race finish time with each of those parameters and saw clear trends. These correlations proved to be strong, 99 percent statistically significant, consistent, plausible, coherent and replicable in different circumstances.
Nevertheless, these are just correlations. We cannot infer from the national statistics the health condition of the race runners. Is it deteriorating or improving? Also, in no way we argue that these are the only possible explanations and reasons for the observed effect."
So, although there is a correlation between the finishing times in races and the general deterioration of the health of the American population, the authors are careful to state that this may or not be reflected on the statistical results. To be sure, they'd need to know the medical condition of more than 20 million runners — an impossibility.
As I mentioned to Nikolova, I am somewhat skeptical of this correlation, especially for half-marathons and marathons. Being a runner, I find it hard to believe that many more people with obesity are tackling such distances. I surely don't see them at the starting line.
To my mind, the reason why Americans are getting slower on average is a good one: With more people interested in running to get and stay fit, there is an unavoidable drop in the level of training and preparedness. More people run, and the results are slower overall.
To settle this, I encourage Andersen and Nikolova to repeat their study for other countries, where there isn't a marked increase in the level of obesity, diabetes, and hypertension, as in the US. If the finishing times in these other countries are either faster or have remained stable over the years, then they have a strong argument in their hands. Otherwise, if the finishing times are also growing in different countries, then it's all due to having more people running, a very good thing for everyone.