Pacing and placing in 161-km ultramarathons: Effects of sex and age

Abstract

Ultramarathons are growing in popularity, owing especially to the participation of women and masters athletes. Pacing strategy, which can vary by sex and age, is a critical variable in determining finishing place in races up to the marathon distance. Whether this is true in 161-km ultramarathons is unclear. We tested the hypotheses that pacing is a determining factor in finishing place and that pacing differs by sex and age in 161-km ultramarathons. Publicly available data from 161-km races (n = 6) were analyzed for years 2012-2022 (n = 56). Linear regression was used to analyze the proportion of time in each segment (between timing checkpoints) by place, sex, and age. In general, runners used the same percentage of their total race time in each segment independent of finishing place, sex, or age. The exception was that later finishers often ran proportionally faster at the start compared to earlier finishers. Finishing times increased with age but pacing was unaffected by age or sex. We conclude that slower finishers paced the same as faster finishers following a relatively quick start in these ultramarathons, and pacing was not consistently affected by sex or age. These findings should inform training decisions and racing strategy in 161-km ultramarathons.

Fig 1. Summary of generalized pacing-placing findings.

While all finishers slowed down over the duration of the races, slower finishers started their races fast relative to their overall performance. Finishers paced their races similarly after the first ~30 km of the race. Solid lines depict a flatter course where pace is less variable, and the dashed lines depict a mountainous course where pace fluctuates substantially due to significant elevation changes. These generalized findings were true regardless of sex or age.

Fig 2. Finishing time increases with age.

Although pacing did not differ by age, older finishers had slower average pace than younger finishers. For ease of presentation, all years of each race are pooled in respective figures; open squares and trendlines are males and filled circles and trendlines are females. Significant age-time relations (p < 0.001) were found for all years in HR, UTMB, and WS; 2020 in HURT; 2012-2017, 2019, 2021 in RR; but none of the years in TP. See Table S2 for specific beta and p values. The flat ‘ceiling’ on data in each graph is due to the cut-off times imposed by each event, beyond which participants are no longer allowed to continue or be counted as official finishers. RR = Rocky Raccoon, HR = Hardrock, TP = Thames Path, WS = Western States.