Triathlon is a multi-sports discipline involving swimming, cycling and running and can be held from the Olympic distance (i.e. 1.5 km swimming, 40 km cycling and 10 km running) (Bentley et al., 2002) to the Ironman distance (i.e. 3.8 km swimming, 180 km cycling and 42.195 km running) such as the ‘Ironman Hawaii’ (Lepers, 2008; Lepers et al., 2013). Recent studies showed that triathlon performances changed over the last decades for both the Olympic distance (Etter et al., 2013) and the Ironman distance (Lepers, 2008; Rüst et al., 2012a 2012b).

Etter et al. (2013) showed for short-distance triathlon at national level during the 2000–2010 period that the overall top five women improved overall race time by ~0.8 min per annum, while overall race time remained stable in men. During this period, swimming and running performances remained stable for both women and men while cycling performance decreased significantly by ~0.8 min per annum in women and by ~0.5 min per annum in men, respectively. Similarly, in a long-distance triathlon at national level such as ‘Ironman Switzerland’ as a qualifier Ironman for the Ironman World Championship ‘Ironman Hawaii’, Rüst et al. (2012a) observed that women improved between 1995 and 2011 in all three split disciplines and overall race times whereas men improved only in the cycling split and in overall race time. In ‘Ironman Hawaii’, the world best elite men improved in the three split times and overall race time, whereas women improved only in cycling, running and overall race time during the last 30 years (Rüst et al., 2012b). An increase in running speed for both ITU (International Triathlon Union) male and female junior elite triathletes has been reported since the introduction of sprint distance events (i.e. 750 m swimming, 20 km cycling and 5 km running) at the World Championships 2002 to 2011 (Landers et al., 2013; Vleck et al., 2008). To date, the changes in performance of the world best elite triathletes in short distance triathlon have not been investigated.

The sex difference in triathlon performance has changed during the last decades. For example, the sex difference in overall race time in ‘Ironman Hawaii’ decreased significantly during the last 25 years to stabilize at ~11.3% (Lepers, 2008). During the same period, the sex difference in performance remained quite stable for swimming (~12.5%) and cycling (~12.5%) but it decreased for running from ~13.5% to ~7.3% (Rüst et al., 2012b). At the long-distance duathlon World Championship ‘Powerman Zofingen’ from 2002 to 2011, the sex differences in performance were ~16%, ~17%, ~15%, and ~16% for the 10-km running split, the 150-km cycling split, the 30-km running split and overall race time, respectively (Rüst et al., 2013).

The sex difference in triathlon performance depends upon different variables such as the three disciplines, the distances (i.e. short-distance versus long-distance), race tactics (Landers et al., 2008), training (Etxebarria et al., 2013), race experience (Gilinsky et al., 2013), age (Knechtle et al., 2012), anthropometric characteristics (Knechtle et al., 2010a 2010b), and the level of the triathletes (i.e. elite versus non-elite) (Lepers et al., 2013). For example, in short-distance triathletes at national level, the sex difference appeared greater for running (~17%) compared to swimming (~15%) and cycling (~13%) (Etter et al., 2013). In contrast for elite long-distance triathletes, the sex difference in performance tended to be lower for running and swimming compared to cycling (Lepers, 2008).

Anthropometric characteristics seem to be important predictors for race time in short distance triathlon at world class level (Landers et al., 2000). Potential reasons in the differences in performance between female and male triathletes are the lower maximum oxygen uptake in women (~52.8 ml · kg^-1 · min^-1) compared to men (~61.3 ml · kg^-1 · min^-1) (Knechtle et al., 2004), the lower muscle mass in women (~28 kg) compared to men (~41 kg) (Knechtle et al., 2010a), and the higher percent body fat in women (~23.6%) compared to men (~13.7%) (Knechtle et al., 2010a). If we consider the three individual sports, there is a greater difference between the male and female world records for running (e.g. 10 km running difference ~ 12.1%, 21.1. km running difference ~12.8%) than swimming (e.g. 1,500 m swimming difference ~7.4%) and cycling (e.g. cycle hour record difference ~7.9%). The difference between running and the two other disciplines could be explained in part by the biological gender difference in relative body fatness which is higher in women (Landers et al., 1999). Indeed, greater body fat may represent a limit in weight-bearing activities such as running.

In addition, at international level, the differences in performance for elite triathletes between short- and long-distance triathlon might be explained by the possibility of drafting in the cycling split. In international long-distance triathlon, drafting is prohibited in contrast to short-distance triathlons of the ITU World Triathlon Series. Drafting in swimming and cycling may result in a better tactical approach to increase the overall performance in elite Olympic distance triathlons (Bentley et al., 2008). Pacing strategies are observed by elite athletes who are swimming or cycling in a sheltered position inducing several changes of pace (Hausswirth and Brisswalter, 2008). Drafting may alter the sex difference in cycling and in the subsequent running performance. Fast runners seemed to benefit most from drafting during cycling (Hausswirth et al., 1999). For the run split in a short-distance triathlon, an appropriate pacing appeared to play a key role in high-level triathlon performance (Le Meur et al., 2009). Le Meur et al. (2009) showed that both female and male elite triathletes developed specific pacing strategies in running. The men’s running speed decreased significantly over the whole distance whereas women slowed down in the up- and down-hill sections.

Elite short-distance triathletes intending to compete in the Olympic Games need to undergo a qualification in the four years before the Olympic Games. They have to compete in the ITU World Triathlon Series in order to obtain points to qualify for the 55 start places in the Olympic Games. To date, the changes in performance of world class triathletes during a 4-year period before the Olympic Games have not been analyzed. The first aim of the study was therefore to analyze the changes in performance for both elite men and women and the corresponding sex difference in performance in the ITU World Triathlon Series between 2009 and 2012 including the Olympic Games 2012 in London. A second aim was to investigate the sex difference in performance for overall race time and for split times in these athletes.

All procedures used in the study met the ethical standards of the Swiss Academy of Medical Sciences and were approved by the Institutional Review Board of Kanton St. Gallen, Switzerland, with a waiver of the requirement for informed consent of the participants given the fact that the study involved the analysis of publicly available data.

Between 2009 and 2012, 58 different women and 55 different men recorded a top ten result in a total of 27 World Triathlon Series event and the 2012 Olympic Games. Eight races were held in 2009, seven races in 2010, eight races in 2011 and four races in 2012 before the 2012 Olympic Games. Table 1 reports the number of finishes during the studied period.

Figure 1 shows the changes in overall race times and split times across all 28 races. For women and men, the split times in swimming and running remained unchanged. The cycling split times increased for women from 63.5 ± 3.0 min to 67.2 ± 2.3 min and for men from 57.4 ± 2.6 min to 60.4 ± 1.1 min (Table 2) also when corrected for multiple finishes (Table 3). Overall race times increased for women from 119.5 ± 3.4 min to 123.3 ± 2.9 min and for men from 106.9 ± 3.6 min to 110.6 ± 2.2 min (Table 2) also when corrected for multiple finishes (Table 3). The sex difference in performance remained unchanged for swimming and cycling but decreased for running from 14.9 ± 2.7% to 13.2 ± 2.6% and for overall race time from 11.9 ± 1.2% to 11.4 ± 1.4% (Table 2) also when corrected for multiple finishes (Table 4). The sex difference in running (14.3 ± 2.4%) was significantly (P < 0.001) greater compared to the sex difference in swimming (9.1 ± 5.1%) and cycling (9.5 ± 2.7%) (Figure 2).

The aim of this study was (i) to analyze the changes in performance and sex difference for both elite women and men in the ITU World Triathlon Series between 2009 and 2012 including the Olympic Games 2012 in London and (ii) to investigate the sex difference in performance for overall race time and split times in these athletes. The present findings showed an increase in overall race times and cycling split times between 2009 and 2012 for the top ten finishers in the ITU World Triathlon Series including the 2012 Olympic Games.

However, due to several limitations, e.g. different races each year, different environmental conditions, and potential differences in the course lengths, these changes in cycling and total performance across the period 2009–2012 may be not relevant. Additionally, in most cases, the female and male events were not only held at different times of the day but also on different days. Besides these limitations, the most interesting findings were (i) a decrease in the sex difference for running and overall race time with a stabilization in swimming and cycling and (ii) a greater sex difference in running compared to cycling and swimming performances.

During the 2009–2012 period, the world’s best female short-distance triathletes reduced the gap with male athletes in running and total triathlon performance. The sex difference in performance was greater for running (~14%) compared to swimming (~9%) and cycling (~10%). The influence of cycling drafting on running performance may differ between elite female and male triathletes. Future studies are required to clarify why the sex difference in running is greater compared to swimming and cycling in international short distance triathlon races with drafting.

The authors declare that they have no competing interests.

CAR and BK drafted the manuscript, CAR performed the statistical analyses, MS collected all data, TR and RL participated in the design of the study and helped in interpretation of the results and to draft the manuscript. All authors read and approved the final manuscript.