ECSS Paris 2023: OP-AP09
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
Read CV Esther EustacheECSS Paris 2023: OP-AP09
INTRODUCTION: Elite performance is influenced by long-term training and discipline-specific physiological demands (1, 2). Ageing is associated with progressive performance decline due to physiological and neuromuscular alterations (3). Therefore, this study quantified age-related performance decline across sprint, middle-, and long-distance events in Spanish male and female master athletes. METHODS: Official data from the “Spanish Association of Track and Field Statisticians*” were analysed. The all-time top 25 performances were extracted for sprint (100 and 400 m), middle-distance (800 and 1,500 m), and long-distance (5,000 and 10,000 m) events in spanish master athletes, grouped into five-year age categories ranging from 35 to 65 years. Data were analysed separately for male and female athletes, and official race performance time was used as the primary outcome variable. Data are presented as mean ± SD. Percentage change in performance between consecutive age categories was computed to quantify age-related decline in men and women. A two-way ANOVA (age category × sex) was performed. Linear regression examined the association between age and performance separately in men and women. RESULTS: Performance time increased progressively with age in all events (p < 0.05). The magnitude of this increase differed by discipline and sex. In sprint events (100–400 m), men showed a relatively gradual increase across age categories (~2–6% per age category), whereas women demonstrated a more pronounced increase after 50 years, particularly in the 100 m (up to ~16% between 60–65). In middle- and long-distance events (800–10,000 m), performance time increased across all age groups, with larger percentage increases after 50 years. Women generally exhibited greater relative changes than men in the older categories. ANOVA revealed significant effects of age, sex, and age × sex interaction across all events (all p < .001), indicating sex-specific ageing trajectories. Regression analyses confirmed age as a strong predictor of performance (β = .870–.963; p < .001). Slopes were consistently steeper in women than for men (e.g., 100 m: +0.177 vs. +0.086 s·year⁻¹; 10,000 m: +29.19 vs. +20.12 s·year⁻¹). CONCLUSION: Performance decline was evident across all disciplines, with a steeper age-related deterioration in women. Lower historical female participation may partially influence these patterns. Additionally, sex-specific physiological, anthropometric and neuromuscular differences, together with menopause and age-related sarcopenia, may contribute to the amplified decline observed in female master athletes. These findings support sex-specific approaches when analysing ageing in competitive sport. * https://www.aeeaatletismo.es/2025/08/28/ranking-de-espana-master-de-todos-los-tiempos/ 1. Allen & Hopkins, 2015. 2. Haugen et al., 2018. 3. Till & Baker, 2020.
Read CV Sergio Rodríguez-BarberoECSS Paris 2023: OP-AP09
INTRODUCTION: Participation in recreational marathon running continues to increase, yet detailed descriptions of training characteristics across performance levels remain limited. This study examined training volume, training intensity distribution (TID), training load (TL), and training load progression (TLP) in recreational marathon runners during the final 12 weeks preceding a marathon. METHODS: Part 1: General training data (n=182) Twelve weeks of GPS-derived training data from 182 recreational runners were analyzed including, training frequency, running speed, distance, and session duration (age: 44.2 ± 10.0 y; running experience: 10.0 ± 9.8 y; body mass: 72.3 ± 10.9 kg). Runners were classified into five groups based on marathon finishing time, using 30-min intervals ranging from >4:30 h (group 1) to 2:30–3:00 h (group 5). Part 2: TID (n=172) CS was estimated from marathon performance using the linear regression model of Smyth and Muniz-Pumarez (2020), based on the percentage of CS sustained during the marathon.[1] Seven training zones were defined in 10% increments starting at 60-70% of CS (zone 1) up to 120-130% of CS (zone 7), and TID was expressed as percentage of time spent in each zone. Part 3: TLP (n=146) Weekly TL (arbitrary units), was calculated using the TRIMP-method applying zone-specific weighting factors (0.5 for zone 1 up to 12 for zone 7). TLP was then expressed as percentage change relative to the preceding week. RESULTS: Results Part 1: Training frequency and weekly running volume were significantly lower in Groups 1–4 compared with Group 5 (4.0 ± 1.6 vs. 6.6 ± 3.6 sessions·wk⁻¹, p <0.001 ; 51.3 ± 28.7 vs. 82.7 ± 39.4 km·wk⁻¹ ; p <0.001). Mean session duration did not differ between groups (≈70–78 min·session⁻¹, p =0.331). however, Groups 1–3 completed fewer long runs ≥25 km (3.3 ± 2.3 vs. 5.6 ± 2.7, p =0.001) and ≥32 km (0.8 ± 1.1 vs. 2.0 ± 2.1, p =0.002) than Group 5. Part 2:Groups 1–3 spent less training time in zone 2 than Groups 4–5 (≈17–21% vs. ≈27–33% , p <0.001) and more time in zone 4 (≈23–28% vs. ≈16–20%, p <0.001). Part 3: On average TL increased linearly weekly with 7.86 ± 13.62% over all groups, with no differences between groups (p =0.95) . CONCLUSION: The faster marathon runners showed a higher weekly volume (km/week), training frequency, and completed more long runs. Additionally, slower groups spent more time in zones 3 and 4 while faster groups spent a larger proportion in zones 2 and 3. Finally, no differences in training load progression were found between the groups. These findings extend current knowledge by characterizing training patterns observed among faster and slower recreational marathon runners. [1] Smyth, B., & Muniz-Pumares, D. (2020). Calculation of Critical Speed from Raw Training Data in Recreational Marathon Runners. Medicine & Science in Sports & Exercise, 52(12), 2637–2645.
Read CV Rayan VanderpoortenECSS Paris 2023: OP-AP09