Speed, Gender, And Age: An Ultrarunning Comparison

[Writer’s Note: The data and statistics used in this article were obtained from the 2015 February/January edition of UltraRunning Magazine and from my personal blog posts in 2013 and 2014.]

The sport of ultrarunning is growing. There were roughly 80,000 ultrarunning finishes around the country in 2014, which marked a 15% increase in finishes as compared to 2013. (One runner can account for more than one finish, so the exact number of finishers, rather than finishes, is not specified by UltraRunning Magazine.) Folks are getting faster, too, as evidenced by significant course records set last year at the 50-mile distance: notably those set on both the men’s and women’s sides at the Ice Age Trail 50 Mile and the Lake Sonoma 50 Mile. At the 100-mile distance, UltraRunning Magazine’s Male Ultrarunner of the Year, Rob Krar, posted the second-fastest times ever on three different courses, all with competitive fields: Western States 100, Leadville Trail 100 Mile, and Run Rabbit Run 100 Mile.

Despite the increase in speed, one might be surprised by the age of participants at ultramarathons. Non-ultra friends of mine who have spectated at races have proclaimed, “Why, these people aren’t very young!” And they are correct! Nearly half of ultramarathon finishers in 2014 were 40 years of age or older. Yet runners have ventured into the sport at a young age, too: runners in college and high school and even younger still. Interestingly, both groups—the young and (relatively) old—have seen success: Jared Hazen, at 18 years of age, finished third at the fast and competitive Rocky Raccoon 100 Mile in 2014 and, not even out of high school at the ripe age of 17, Andrew Miller won the 2014 Orcas Island 50k. He also won in 2015 at the age of 18. Yet, at 49 years young, Bev Anderson-Abbs won the Jed Smith 50 Mile in the second-fastest 50-mile time in the country last year, 6:14:46, and Traci Falbo, at the age of 43, set a then American record for the fastest trail 100 mile. (That record has since been bested.) These, of course, are the outliers, but they raise questions about age, gender, and performance: How does age affect performance in ultras? What are the ages of the runners who are turning in the fastest times and the most impressive performances? Are the ages of top female performers generally greater than the ages of top male performers, as the previous examples might suggest?

In this article, we explore the links between age, gender, and performance in ultramarathons. Various data on men and women with top times and performances at the 50k, 50-mile and 100-mile distances are presented in sets and then analyzed; perceived trends are noted; tentative conclusions are drawn; and future projections are given. To put it more succinctly: this article is an attempt to look at age-related performances in ultras from several different angles.

However, it should be noted that this article comes with some qualifications. First, the data are based on what races people choose to run in a given year. For example, Anna Frost might have failed to post a top overall finishing time at the 50-mile distance in 2014, but she chose not to run a 50-mile race in 2014, and the data sets presented are thus skewed: they don’t represent the average ages of the runners who could post top times, they represent the average ages of runners who chose to run a certain distance and who did run a top time. Similarly, Michael Wardian, at the ages of 39 and then 40 last year, chose to run several fast 50k courses, and so the data for top overall times at the 50k distance are notably skewed upward. Second, and relatedly, the data sets below concerning “top overall times”—rather than the data concerning the ages of runners who posted top graded performances or top finishes in the “most competitive races,”—are skewed toward times from the most runnable/fastest courses. Thus, Sage Canaday may have posted several top graded performances at the 50k distance in 2014, even though he failed to post a top overall time at the distance. But, we have done our best with the available data.

Data Sets and Analysis

Overall Times: Average ages of runners who ran the 15-fastest-finishing times at the respective distances

Men’s Age Women’s Age Men’s vs. Women’s
50k 34.7 30.5 + 4.2 years
50 mile 31.5 31.8 – 0.3 years
100 mile 33.3 36.2 – 2.9 years
  • The average male age for top overall times is lesser than the average female age at the 50-mile and 100-mile distances but greater at the 50k distance.
  • The 50k data for men is notably skewed upward compared to previous years as can be seen in overall times-comparison data below.

Graded Performances: Average ages of runners who ran the top-15 graded performances at the respective distances

Graded Performances are, according to UltraRunning Magazine, “rankings determined by using the Comparative Difficulty Ratios developed by Gary Wang at RealEndurance.com… using this approach [UltraRunning Magazine] can make a relative assessment for courses of widely varying difficulty.” See UltraRunning Magazine for more details.

Men’s Age Women’s Age Men’s vs. Women’s
50k 28.5 31.5 – 3.0 years
50 mile 28.7 31.5 – 2.8 years
100 mile 33.5 35.5 – 2.0 years
  • The average male age for those with top graded performances is uniformly and notably lesser than the average female age.
  • This has been the case in data from previous years, too, as can be seen the graded performance comparison data below.

Overall Times Comparison: 2014 Overall Times compared to the same data from 2012 and 2013

2012 Men’s Age 2013 Men’s Age 2014 Men’s Age 2012 Women’s Age 2013 Women’s Age 2014 Women’s Age
50k 34.1 31.3 34.7 34.9 32.8 30.5
50 mile 32.7 29.7 31.5 35.7 35.2 31.8
100 mile 37.3 37.6 33.3 36.9 38.6 36.2
  • At each distance on the women’s side, there is a nearly uniform decline over the last three years in the average age.
  • At the 50k distance on the men’s side, there was a significant age drop from 2012 to 2013.
  • At the 50-mile distance on the men’s side, there was a slight overall decline in age and at the 100-mile distance there was a significant decline in age since 2012.
  • For those with top overall fastest times, the average male age is lesser than the average female age in six of the nine comparisons by year and distance.

Graded Performances Comparison: 2014 Graded Performances compared to the same data from 2012 and 2013

2012 Men’s Age 2013 Men’s Age 2014 Men’s Age 2012 Women’s Age 2013 Women’s Age 2014 Women’s Age
50k 30.5 30.5 28.5 34.1 33.1 31.5
50 mile 30.6 27.4 28.7 32.7 33.3 31.5
100 mile 34.5 36.9 33.5 36.6 36.2 35.5
  • At each distance on the women’s side, there is a nearly uniform decline in the average age over the last three years.
  • On the men’s side, there is also a general decline in the average age at each distance since 2012, although the declines are not as uniform as they are for the women.
  • The average male age is lesser than the average female age almost uniformly—this time in eight of the nine comparisons.

Most Competitive Races: Average ages of runners who ran the 10 fastest times, i.e. had “top finishes,” at the three most competitive races at the respective distances

The three “most competitive races” of 2014, listed in no particular order, were individually picked by Eric Senseman, Bryon Powell, and Meghan Hicks, and then chosen by consensus.

Men’s Age Women’s Age Men’s vs. Women’s
50k 31.4 32.6 – 1.2 years
50 mile 31.2 32.4 – 1.2 years
100 mile 34.9 36.5 – 1.6 years
  • In the three most competitive races in the country last year, the average male’s age is lesser than the average female age at each distance.

The breakdown for most competitive races by gender and distance, with the average ages in parentheses, were as follows:

  • Men 100 mile: Western States (33.7), Run Rabbit Run (34.3), Leadville Trail (36.6)
  • Men 50 mile: Lake Sonoma (29.5), The North Face Endurance Challenge 50 Mile Championships (31.6), Sean O’Brien (32.4)
  • Men 50k: Run the Rut (30.0), Speedgoat (32.9), Chuckanut (31.2)
  • Women 100 mile: Western States (37.0), Wasatch (37.4), Rocky Raccoon (35.2)
  • Women 50 mile: Lake Sonoma (34.4), The North Face Endurance Challenge 50 Mile Championships (33.4), Ice Age (29.4)
  • Women 50k: Run the Rut (29.5), Speedgoat (34.2), Chuckanut (34.1)

Overall Times (OT) versus Most Competitive Races (MCR): 2014 Overall Times compared to 2014 Most Competitive Races

Men’s OT Men’s MCR Men’s OT vs. MCR Women’s OT Women’s MCR Women’s OT vs. MCR
50k 34.7 31.4 + 3.3 yrs 30.5 32.6 – 2.1 yrs
50 mile 31.5 31.2 + 0.3 yrs 31.8 32.4 – 0.6 yrs
100 mile 33.3 34.9 – 1.6 yrs 36.2 36.5 – 0.3 yrs
  • At the 50k distance, the ages of men with top overall times was considerably greater than men with a top finish at one of the most competitive races while the average age of women with a top overall time was notably lesser than women with a top finish at one of the most competitive races.
  • At the 50-mile distance, the ages of men with top overall times were slightly greater than the ages of men with top finishes at the most competitive races; the ages of women with top overall times were slightly lesser than women with top finishes at the most competitive races.
  • At the 100-mile distance, the ages of both men and women with top overall times were lesser than their counterparts with top finishes at the most competitive races, although the average male age was notably lesser.

Graded Performances (GP) versus Most Competitive Races (MCR): 2014 Graded Performances compared to 2014 Most Competitive Races

Men’s GP Men’s MCR Men’s GP vs. MCR Women’s GP Women’s MCR Women’s GP vs. MCR
50k 28.5 31.4 – 2.9 yrs 31.5 32.6 – 1.1 yrs
50 mile 28.7 31.2 – 2.5 yrs 31.5 32.4 – 0.9 yrs
100 mile 33.5 34.9 – 1.4 yrs 35.5 36.5 – 1.0 yrs
  • At each distance, the average age of men and women with top graded performances was lesser than the average age of men and women with top-10 finishes at the most competitive races.
  • The decline in age was more significant for men than for women.

Trends, Conclusions, and Projections

To do anything more than speculate about trends, conclusions, and projections would seem more than a bit hasty. After all, it would be very difficult, if not altogether impossible, to identify clear and significant trends with just three years of data (or less in the case of top finishes at the most competitive races), to conclude what causes improved performances and times based on average ages alone, or to make projections about an unknown future. Many other factors beyond age and gender surely have an impacting influence on performance. So, as it is said, take what follows with a grain of salt—or several, and perhaps a whole handful.

Trends
Trend 1: Fast ultrarunners may be getting younger, regardless of gender.
The ages of top performers—both male and female—have lowered over the last several years. In fact, the average ages of males and females who posted top overall times and top graded performances at the various distances have lowered considerably since 2012.

Men:

  • At the 50k distance, the average age of top overall times for males has, contrary to the trend, increased by 7% while the average age of males with top graded performances has lowered by 6.6%.
  • At the 50-mile distance, the average age of top overall times for males has lowered by 6% while the average age of males with top graded performances has lowered by 6.2%.
  • At the 100-mile distance, the average age of top overall times for males has lowered by 7% while the average age of males with top graded performances has lowered by 2.9%.

Women:

  • At the 50k distance, the average age of top overall times for females has lowered by 6% while the average age of females with top graded performances has lowered by 7.6%.
  • At the 50-mile distance, the average age of top overall times for females has lowered by 9% while the average age of females with top graded performances has lowered by 3.7%
  • At the 100-mile distance, the average age of top overall times for females has lowered by 9% while the average age of females with top graded performances has lowered by 3.0%.

This trend might be caused by the fact that more young runners competing in 2014 than in 2012; it might be that younger runners haven’t grown in quantity, but have improved in quality as compared to their relatively older competitors. The trend might be caused by both of those things or by something else altogether. Whatever the cause, the trend seems to present itself clearly: fast folks are getting younger!

The single case contrary to the overall trend is also explainable, as was previously mentioned: Michael Wardian. He ran three of the top-15 overall times at the 50k distance in 2014 and, at the ages of 39 and 40 last year, significantly skewed the data. Notice that the ages of men with top graded performances lowered significantly between 2012 and 2014, which suggests that, indeed, and consistent with the overall trend, an increasing number of younger runners are performing better.

Trend 2: The best male runners are younger than the best female runners.
Quickly think of several top female ultrarunners and then bring to mind several top male ultrarunners. Then check their ages. You’ll likely find that, on average, the women you bring to mind are older than the men.

To make the point more concrete, let us turn to the data presented here. The average age of male runners with top performances and times is notably and nearly uniformly lower than the age of female runners with top performances and times and this has proven true each of the last three years.

  • At the 50k distance, males with top overall times were, contrary to the trend, 8% older than top overall female times; males with top graded performances were 9.5% younger than top overall female times; and males with top finishes in the most competitive 50k races were 3.7% younger than their female counterparts.
  • At the 50-mile distance, males with top overall times were 0% younger than top overall female times; males with top graded performances were 8.9% younger than top overall female times; and males with top finishes in the most competitive 50 miler were 3.7% younger than their female counterparts.
  • At the 100-mile distance, males with top overall times were 0% younger than top overall female times; males with top graded performances were 5.6% younger than top overall female times; and males with top finishes in the most competitive 100 milers were 4.4% younger than their female counterparts.

This trend, which has a single data point against the tide, and which again seems accountable to Mike Wardian, suggests many possibilities—or may be caused by many differing things. It’s possible that last year there were more young men competing in ultras than young women. This possibility seems fairly likely since males accounted for more than two-thirds of the ultra finishes last year. Secondly, and relatedly, it might be that there are a larger percent of young men competing in ultras relative to the total number of men than there are young women competing in ultras relative to the total number of women. This would, consistent with the first possibility, result in a much greater number of young men competing in ultras than young women. Thirdly, it might be that females perform at a high level for a longer time, or until a later age, than males.

Trend 3: Top finishers in the most competitive races are relatively older.
The effects of this trend are somewhat less severe than in the first two trends. In general, the average ages for men and women with top overall times and top graded performances are younger than the average ages for men and women with top finishes at the most competitive races in 2014. (See the data above comparing overall times and graded performances to most competitive races.) Keep in mind that only one year of data from the so-called “most competitive races” was collected so it’s difficult to determine a genuine, rather than spurious, trend. Nevertheless, it’s a curious fact that this trend presents itself in the single year of data and its presence suggests many possibilities: perhaps, for example, the so-called “most competitive races” weren’t all that competitive after all, or there were high dropout rates in those races, and so less talented runners found their way into the top 10. Rather, it might be that older runners possess more of the tactical skills needed to race well in competitive fields that younger runners haven’t yet developed. Or these data points may simply be irregular to actuality: the older folks just got lucky last year! I leave it to the more ingenuous readers to consider further possibilities here.

Conclusions
What can be rightly concluded from all this analysis? Are these trends merely coincidence or are there genuine links between age, gender, and performance?

Conclusion 1: Age is at least correlated with performance.
Age and performance seem to ebb and flow: when we are very young, our bodies aren’t yet developed enough to perform at our best and once we age past a certain point our bodies begin to lose the capacities they once had—due to wear and tear, or overuse, or whatever. There is a period in the middle when we can perform very well and an even smaller window when our bodies perform their best. Two people might reach peak performance in ultras at very different ages simply because one started running very early and the other didn’t begin running until later in life: there is a certain “shelf life” to most people’s performance longevity. Of course, there are outliers to that fact, too, and some perform at a high level for much longer than others. The general conclusion remains: age has an affect—indirect or direct—on performance. In the case of ultrarunning—a sport that demands much of its pursuers—youth does seem advantageous, since younger runners possess characteristics that prove beneficial for speed and endurance.

Conclusion 2: Gender causes peak performance at different ages.
It seems within the bounds of reason to conclude that males and females peak physically, at least for the purpose of ultrarunning, at different ages. The data is almost perfectly clear and nearly uniform in supporting this fact: each of the last three years, at each of the distances, average male ages are generally lesser than average female ages. Exactly how, why, and where those physical peaks differ is a further conclusion that cannot be rightfully reached in this article. There are simply too many factors that influence this gender-based difference!

Projections
Can anything be justifiably said about the future? Will these trends hold into the future? Will average ages continue to decline or have they reached a plateau? What will the average ages be next year, 10 years from now, and so on?

Projection 1: The first two trends will continue…
If we assume that the sport of ultrarunning will continue to grow in popularity, and thus that more and more people will choose to run ultra-distance races, then it would seem to follow that competition will continue to become more stout, that course records will continue to be reset, and that finishing times in general will continue to improve. If all of that holds true, then the trends noted and analyzed here would seemingly continue to hold true as well. Think about it: the sport of ultrarunning has grown vastly in the last three years and has, in part as a result of the growth, become much more competitive. At the same time, we see the average ages of top runners trending downward since 2012. The evidence seems to side with more youngsters prospering!

Projection 2: …but the trends won’t continue forever!
Not so fast, there is a small caveat: the ages of top ultrarunners will not continue to decline ad infinitum. They couldn’t! They must plateau at some point. I’m suggesting that we have not yet seen that plateau. When will the ages plateau? I can’t say with any degree of certainty when that will be; it only seems to me that—based on the fact that the sport continues to grow and draw fast, young talent—it hasn’t happened yet. But what do you think?

It is also worth noting that if females begin to account for closer to half of the ultrarunning finishers instead of less than half—males accounted for 68.2% of ultra finishes in 2014 and races on the men’s side were, unsurprisingly, more competitive—it seems quite possible that the average age of top female performers might become closer to those of top male performers, although, given what is known about peak male and female performance in endurance sports, men’s ages are likely to always be higher.

No matter which trends die, continue, or emerge, the sport of ultrarunning is still a relatively young one and it will be very interesting to keep an eye on what is happening in the sport as it grows. Tune in next year to see what trends live on!

Call for Comments (from Meghan)

  • What do you make of the data, trends, and projections here? Do you see any trends, or a lack thereof, that haven’t been mentioned?
  • Anecdotally, according to your personal experiences at ultramarathons, do your observations match with the data and trends presented here?
  • Where do you think high-speed and competitive ultrarunning is going? Will ages of top performers drop? Will they level off? Will a difference in age of top performance between men and women always be present?

There are 7 comments

  1. Andy

    I would agree in general with the observation that the declining age of top ultra runners is due to more youngsters entering the sport. As for most of the data and conclusions — viewed from my perspective as a behavioral scientist — it's all meaningless and and can't be intepreted without looking at the qualititative and individual aspects of the numbers. The whole concept of "average age" of top finishers is not statistically sound. And, "Gender causes peak performance at different ages"? I hope that's intended as a joke! Still, it is an entertaining effort to make sense of trends. Hell, I'm in my 50s and still getting faster, thanks to only being in my 6th season of this crazy sport. And, of course, clean living :)

  2. idonotrunfar

    Also science person here.

    When doing any sort of basic statistics, the average means absolutely nothing, it is helpful to have the standard deviation and median at the very least to know what the average represents. (also "3" is a bit limited as a population)

    I understand it is for simplification purposes but a billionaire entering a bar makes "on average" everybody in this bar a millionaire. :-)

    Also correlation is not causation.

    Since the 1950s, both the atmospheric CO2 level and obesity levels have increased sharply.
    But it doesn't mean atmospheric CO2 causes obesity. :-)

    1. @goodsenseruns

      Thanks for reading. You are very right that additional information, like standard deviation, would make the statistics presented more meaningful. Alas, over simplification for mass readership prevailed! And, of course, correlation does not equal causation–I try to make that point and other cautionary points clear in the brief introduction to the last section, but perhaps I failed to do that clearly. The trends and conclusions are merely speculation; no causal claims could possibly be made with any certainty here. Thanks for the feedback!

  3. sharmanian

    What I'd love to know, but it would require detailed physiological testing over years for many top level athletes, is what the optimum age, on average, is for different distances. Or it may be more related to the number of years since starting running rather than the absolute age.

  4. jesseluna

    One factor that wasn't mentioned was the "progression" from 50k to 50 milers to 100 milers. It seems like the 50k is turning into the new 5k. As people get into ultras the 50k seems like "speed work" or "a nice training run" and thus more experienced folks are moving to the higher distances because they are perceived as more difficult and more prestigious and they present opportunities to quality for races that are hard to get into like Western States. The age of these speedy folks moving up a way from 50k's is decreasing.

  5. Oluf_RaddGnar

    The most startling and newsworthy conclusion that could be drawn from this data seems to be overlooked: the best ultra-runners are of non-african heritage. Some "genetic wall" must exist between miles 26.2 and 31! (Or.. perhaps this data is too limited to draw any conclusions from…)

Post Your Thoughts