Author(s): EUSTACHE, E., RABERIN, A., MILLET, G., Institution: UNIVERSITY OF LAUSANNE, Country: SWITZERLAND, Abstract-ID: 1616

INTRODUCTION:
Ultra-trail races pose unique challenges –long distances, large positive (D+) and negative (D-) elevation gains, and variable environmental conditions that may affect performance and rate of non-finishers (did not finish, DNF) differently in females and males [1,2]. Although psychosocial predictors of dropout in ultra-trail running have been studied [3], sex differences in DNF remain largely unexplored, especially in relation to race characteristics. This study examined whether females are less likely to DNF or tend to drop out later than males. We also consider whether uniform race time barriers which are the same for both sexes, influence DNF rates (%DNF) differently between sexes.
METHODS:
520 time points from 18 ultra-trail races, classified as 100km (100K) or 170km (100M), across 9 UTMB World Series events were analysed. For each time point, we extracted split characteristics (distance, altitude, D+, D-, temperature). Runners were grouped by performance levels, defined by their ranking at the first checkpoint: faster (1st decile, D1), mid (2nd–9th deciles, D2-9), and slower (10th decile, D10). For each combination sex x performance, the median race time, the median split time preceding each time point, and %DNF were calculated. This study excluded D1 whose DNF reasons differ. Data was analysed using permutation tests and a beta-binomial generalized linear mixed model to examine variables associated with the DNF probability at each split. All analyses were performed in R.
RESULTS:
Females were not more likely to DNF than males overall (p=0.95). For D2–9, %DNF did not differ between females and males (100K: 12 vs 15%, p=0.33; 100M: 25 vs 26%, p=0.76). The %DNF differences between 100K and 100M was similar between sexes (p=0.91). However, females were more likely to DNF than males within the D10 group (p=0.03). For 100K, females increased the %DNF between D2-9 (11 vs 15%) and D10 (65 vs 48%) to a larger extent of 21% (p<0.01) than males. The distance at which DNF occurred did not differ significantly between sexes in both 100K (D2-9: p=0.73, D10: p=0.63) and 100M (D2-9: p=0.76, D10: p=0.33).
Turning to the split characteristics, %DNF shows a quadratic relationship with race time (p=0.01) and D- is associated with a higher %DNF (p<0.01). Other race characteristics are not significantly associated with %DNF.
CONCLUSION:
%DNF in ultra-trail are similar between sexes. D- and longer race times increase DNF risk for all runners, highlighting the combined impact of race difficulty. Low-performance (D10) females are more likely to DNF than males, suggesting that uniform time barriers, designed without sex-specific considerations, may penalize slower female runners and may contribute to limit their participation.

1. Raberin A et al. Women at Altitude: Sex-Related Physiological Responses to Exercise in Hypoxia. Sports Med
2. Besson T et al. Sex Differences in Endurance Running. Sports Med
3. Corrion K et al. Psychosocial factors as predictors of dropout in ultra-trailers. PLoS One