Pacing during 24-hour mountain bike races - a modelling approach

We aimed to develop evidence-based guidance on pacing for 24-hour mountain bike races. Data from 36 men and 9 women who competed in at least 3 editions of the World 24-Hour Mountain Bike Championship were analysed. Average racing speed per lap was normalised to each competitor's race average, and lap number to total laps completed. Linear mixed-effects models were fitted to examine the relationship between normalised speed and percentage of total laps, and to determine inter- and intra-individual variability. Pacing adopted by 24-hour mountain bikers was best described by a cubic polynomial. Intercept was 127.4% (95%CI: 124.0-130.9) of the average racing speed. Inter-individual SD was 8.8% (95%CI: 7.0-10.3) for the intercept and 0.2% (95%CI: 0.1-0.2) for the slope. Intra-individual SD was 10.4% (95%CI: 10.2-10.7). This model outperformed simpler alternatives (all P < 0.001). Typically, 24-hour mountain bike races are characterised by a flattened, reversed S-shape pacing profile. Inter- and intra-individual variability in pacing were similar, despite course changes between editions. In the absence of further experimental evidence, a conservative range of 117% to 137% of the average racing speed is recommended for the first lap based on model predictions. Guidance to develop personalised racing strategies for 24-hour events is provided.