Among the matched pairs who had completed two races, mean ± SD running speeds for the 161-km event were 102 ± 13 and 103 ± 12 m·min−1 for the women and men, respectively (Table 1). These speeds were significantly slower (P < 0.0001) than the 50-km speeds but were not different between matched pairs. Furthermore, the percent of speed difference within each matched pair was not different (P = 0.14) between the 50- and the 161-km events.
Among the matched pairs finishing all three races, running speeds for the 80- and the 161-km racers were not different between the women and men (Table 2). Additionally, the percent of speed difference within each matched pair was not different (P = 0.09) among the three events. Speeds for the 50 and the 80 km were also comparable.
As would be anticipated, plots displaying the percent difference in speeds between matched pairs show greater scatter for the 80- and the 161-km races compared with the 50-km race (Fig. 1). When examined statistically, the absolute (not considering whether it was a positive or negative difference) variations in speed between matched pairs were larger (P < 0.0001) for the two longer events than the 50-km event (mean ± SD differences were 1.5 ± 1.1%, 5.3 ± 4.8%, and 7.4 ± 7.3% for the 50-, 80-, and 161-km races, respectively). Yet, there were no significant relationships of the percent difference in running speeds between matched pairs with the 50-km running speed for either the 80- or the 161-km races. Furthermore, there were no significant correlations for the percent difference in running speeds between matched pairs with age (r 2 = 0.0023, P = 0.77; r 2 = 0.00044, P = 0.85) or with the difference in number of WSER finishes between matched pairs (r 2 = 0.016, P = 0.45; r 2 = 0.039, P = 0.067) for the 80- and the 161-km races, respectively.
The relationships of average speeds for the two longer races with the 50-km race are shown in Figure 2. Faster runners in the 50-km race were also faster in the 80- and the 161-km events, and the relationships were similar for the two sexes.
The slopes of the linear regressions of 80-km speed with 50-km speed were around 0.9 for both sexes (Fig. 2). As such, the slopes of the linear regressions of the difference in average running speeds between the two races relative to speed in the 50-km event became small or not different from zero (y = 0.135x − 20, r 2 = 0.13, P = 0.027 for the women; y = 0.089x − 12, r 2 = 0.04, P = 0.22 for the men). This indicates that the slower women had a slightly greater speed at the longer race, and the faster women had a slightly slower speed at the longer race but there was no such effect for the men. When this change in running speed was considered as a proportion of the 50-km speed, the women had mean differences from 2% faster to 5% slower for the 80-km race compared with the 50-km race (Fig. 3).
In contrast to the 80-km race, given that the slopes of the linear regressions of 161-km speed with 50-km speed were between 0.4 and 0.6 (Fig. 2), the difference in average running speeds between the two races as a function of speed in the 50-km event also had a positive slope (y = 0.44x − 17, r 2 = 0.59, P < 0.0001 for the women; y = 0.55x − 35, r 2 = 0.67, P < 0.0001 for the men). In other words, the faster runners of either sex had a greater absolute decrease in speed than the slower runners. When this change in running speed was considered as a proportion of the 50-km speed, the mean decrease in speed at the longer race ranged from 31% to 38% for the women and 27% to 41% for the men (Fig. 3).
This analysis revealed that there were no differences in the 80- and the 161-km running performances between men and women who were matched for 50-km running performance. This finding is in contrast with two previous studies, which concluded that women matched for performance with men at 42.2- or 56-km distances were able to run better than men in a 90-km race (1,11). In fact, in those studies, the men ran approximately 5-9% slower than the women at 90-km.
The present study examined a larger sample of matched pairs than the previous studies, which both had only 10 matched pairs for the comparison at the 90-km distance (1,11). With 86 matched pairs in the present study, there was an adequate sample that a 5% difference between sexes in 161-km speed would have been identified with an 80% probability. This is of the magnitude of difference that would have been predicted from the previous studies but may not necessarily represent the smallest difference of meaningful importance. Nevertheless, examination of Figure 1 (demonstrating no apparent trend for the data points to be more on one side of the x-axis than the other) and Figure 2 (demonstrating virtually identical regressions for the women and men) provides considerable assurance that it is unlikely that this study incorrectly concluded there was no difference between women and men in 80- and the 161-km running performance.
Subject matching in this study required running speed to be within 4% at the 50-km race. This difference in running speed was felt to be adequate to allow for appropriate matching and also yield a suitable number of matched pairs. As it turned out, the actual mean absolute variation in speed between sexes was only 1.5% for the 50-km race. Women were also matched with men who were within 5 yr of the same age. Statistical analysis revealed that the average age of the men was 1 yr older than the women. Given that adult men and women typically compete in the 5- to 10-yr age brackets, an age difference of 1 yr is not thought to have any practical relevance. Likely of more importance is prior running experience, and evaluation of the number of previous finishes at the three races involved in this study revealed that both groups were equally well experienced at these races (Tables 1 and 2).
The present study differs from previous studies in that it examined ultramarathon races that were partially or exclusively run on trails, whereas the earlier studies involved road runs. Trail running is considerably different from road running. To begin with, the average running speeds are slower on trails compared with the roads because the surface is typically more irregular and the terrain is more varied in grade. The 50- and the 161-km ultramarathons examined in this study are particularly rugged and involve considerable amounts of climb and descent relative to the total race distances. As a result of the slower speeds, most runners of a 161-km trail ultramarathon will spend some time running during the night, aided by lights that they must carry. Such conditions do not allow for the same visual acuity as daylight, so running speeds are typically even slower at night. Furthermore, in contrast to the longer event examined in both previous studies (the Comrades Marathon), which is indicated to have 51 aid stations over, the 90-km distance, the 80-km AR, and the 161-km WSER currently have 10 and 24 aid stations, respectively. As such, runners would find it necessary to carry more supplies between aid stations in the longer ultramarathons examined in this study. Given that the overall effect is for slower running on the trails compared with the roads, differences between sexes may tend to be attenuated under such conditions.
An explanation is in order for the findings of similar speeds for the 50- and the 80-km races. This was not an issue of data distortion from selecting out the subset of pairs that had completed all three races. In fact, the ages, the 50-km speeds, the 161-km speeds, and the number of previous race finishes were comparable between the two data sets (Tables 1 and 2). Rather, the explanation for the similar 50- and 80-km speeds relates to the relative difficulty of the courses. Over half of the 80-km AR course is run on asphalt as opposed to the entire WTC 50-km course being run on trails. Furthermore, there is more climb and descent relative to total race distance in the 50-km WTC compared with the 80-km AR.
As would be anticipated, there was considerable variability in speed differences between matched pairs for the 80- and the 161-km races (Fig. 1). In fact, the 80-km speeds of the men ranged from 23% slower to 13% faster than the women, and the 161-km speeds of the men ranged from 25% slower to 46% faster than the women. Such a wide variability results from the multitude of factors that relate to performance in events that are as challenging as these ultramarathons such as the relative tolerance for rugged terrain with long climbs and descents and the ability to manage hydration and nutritional needs. In the 161-km event, there were the additional challenges related to altitude, generally high temperatures, and night running. The performance variability between matched pairs did not appear to be related to running speed or age. However, there was a trend (P = 0.067) toward a relationship between the sex differences in speed for the 161-km event and the difference in number of previous WSER finishes. In other words, there was a tendency for a woman with less WSER finishes than the man with whom they were matched to run slower than the man and for a woman with more WSER finishes than the man to run faster than the man.
Average running speeds for the two longer distances were found to be strongly related to the 50-km speeds. Stated simply, the faster runners at the 50-km distance also ran faster at the longer distances. This seems inherently obvious and also consistent with previous findings related to distances of 10 to 90 km (8). However, it is interesting that the faster runners had a more pronounced reduction in running speed (both in absolute and relative terms) than the slower runners when moving up to the 161-km distance. In fact, the linear regressions showed that the slower runners reduced speed by around 30% whereas the faster runners reduced speed by around 40%. The explanation for this may be that as an event becomes longer and more challenging to the point that slower speeds are required, the differences in speeds between faster and slower runners naturally narrow.
Several factors have been suggested as possible advantages for women over men in running long ultramarathon distances. Among these possible advantages are that some studies have shown women to use fat to a greater extent (and thus better conserve carbohydrate) than equally trained and nourished men (9,13), and that there is some suggestion that women can exercise at higher proportions of maximal oxygen uptake during long ultramarathon events than men(11). It has also been postulated that women may have greater psychological tolerance in endurance exercise than men (7), that women might be more resistant to pain than men (12), that women might have muscles with greater fatigue resistance than men unrelated to metabolic differences (3,5), and that the lighter weight of women may be an advantage (1). Of importance to note, Coast et al. (4) has recognized that the proposed metabolic advantage for women may be masked by regular feeding in endurance events. Furthermore, we have also found that pain perception to a standardized pressure pain stimulus was not different between men and women running in the WSER (6). As such, some of the rationale previously provided as support of why women might run faster than men at long ultramarathon distances may not be supported within the context of the WSER.
A limitation of this study worth noting is that there was no manner to ensure that the results used in the analyses were for maximal efforts. Precautions were taken to exclude runners from analysis who had finished together in both the WTC and the WSER races in a given year because this suggests that one member of the pair was likely not performing maximally. However, we had no other means to ensure that the data involved maximal performance efforts. It should also be noted that there was no information on the subjects about training and injury history leading up to the events, or whether the runners had any other performance-limiting issues during the events. Nonetheless, the relatively tight relationships of the 80- and the 161-km speeds with the 50-km speeds (Fig. 2) suggest that there were few major outliers indicative of serious departures from maximal efforts.
It has been recently pointed out that the best running performances are faster for men than women at distances ranging from 100-m to 200-km (4). The fact that the three races analyzed in this study have always been won by men is consistent with that finding. The present analysis is not intended to dispel such information. Rather, the present study was undertaken to determine whether women and men who perform equally at 50 km will perform differently at longer races. From the present results, it can be concluded that there is no difference between men and women in performance at 80- and 161-km trail ultramarathons when matched for performance in a 50-km trail ultramarathon.
The author wishes to thank Gary Wang for assistance with the provision of tabulated race results.
1. Bam J, Noakes TD, Juritz J, Dennis SC. Could women outrun men in ultramarathon races? Med Sci Sports Exerc
2. Cheuvront SN, Carter R III, DeRuisseau KC, Moffatt RJ. Running performance differences between men and women: an update. Sports Med
3. Clark BC, Manini TM, Thé DJ, Doldo NA, Ploutz-Snyder LL. Gender differences in skeletal muscle fatigability are related to contraction type and EMG spectral compression. J Appl Physiol
4. Coast JR, Blevins JS, Wilson BA. Do gender differences in running performance disappear with distance? Can J Appl Physiol
5. Fulco CS, Rock PB, Muza SR, et al. Slower fatigue and faster recovery of the adductor pollicis muscle in women matched for strength with men. Acta Physiol Scand
6. Hoffman MD, Lee J, Zhao H, Tsodikov A. Pain perception after running a 100-mile ultramarathon. Arch Phys Med Rehabil
7. Koltyn KF, O'Conner PJ, Morgan WP. Perception of effort in female and male competitive swimmers. Int J Sports Med
8. Noakes TD, Myburgh KH, Schall R. Peak treadmill running velocity during the V˙O2max
test predicts running performance. J Sports Sci
9. Ruby BC, Roberg RA. Gender differences in substrate utilization during exercise
. Sports Med
10. Sparling PB, O'Donnell EM, Snow TK. The gender difference in distance running performance has plateaued: an analysis of world rankings from 1980 to 1996. Med Sci Sports Exerc
11. Speechly DP, Taylor SR, Rogers GG. Differences in ultra-endurance exercise
in performance-matched male and female runners. Med Sci Sports Exerc
12. Van Aaken E. Van Aaken Method
. Mountain View (CA): World Publications; 1976. p. 81.
13. Venables MC, Achten J, Jeukendrup AE. Determinants of fat oxidation during exercise
in healthy men and women: a cross-sectional study. J Appl Physiol
14. Whipp BJ, Ward SA. Will women soon outrun men? Nature
Keywords:©2008The American College of Sports Medicine
AEROBIC EXERCISE; EXERCISE; SEX DIFFERENCES; SPORT