Performance trends in ultra-endurance sports disciplines have been investigated in running , cycling , triathlon [3, 4] and more recently also in long-distance swimming [5–9]. Besides investigating general trends of performance, a widely discussed topic was the specific sex difference in ultra-endurance performance [1, 10, 11]. Open-water long-distance swimming is a relatively young sports discipline  compared to running where marathons exist for more than a century . The first open-water long-distance swimmers were single athletes. On August 24, 1875, Captain Matthew Webb of Great Britain became the first man to successfully swim the ‘English Channel’ without assistance . Not until 1986, open-water long-distance swimming was established as a World Cup discipline . Later, open-water long-distance swimming was introduced in the 2008 Olympic Games in Beijing with a 10 km open-water competition which was again held in the 2012 Olympic Games in London . Generally, long-distance swimming is defined as freestyle swimming for distances of longer than 400 m. The International Federation of Natation (FINA) furthermore differentiates between long-distance swimming in competitions of more than 400 m in freestyle and more than 200 m in other swimming disciplines and open-water swimming . On a professional level, FINA lists 5 km, 10 km and 25 km as regular race distances . However, any freestyle open-water swimming competition of more than 400 m would meet the required criteria .
The sex difference in open-water long-distance swimming has been investigated in a few studies [5–7, 9], whereas more studies concentrated on classical pool swimming distances of 50 m to 1,500 m [17–21]. For example, in the 2000 Olympic Games, women were on average ~11% slower than men in different types of strokes and distances ranging from 50 m to 1,500 m . Thibault et al.  reported a mean sex difference in performance of ~8.9% in swimming races in the 2008 Olympic Games in Beijing. Tanaka and Seals  reported a decrease in the sex difference in freestyle swimming performance with increasing race distance ranging from 50 m to 1,500 m whereas Eichenberger et al. [5–7] found larger sex differences in freestyle swimming performance in race distances longer than the 1,500 m freestyle. To date, the largest values for sex differences in swimming were reported for the split times in swimming in a Deca Iron ultra-triathlon where women were ~45% slower in 38 km freestyle swimming compared to men . However, it might be argued that results in split times never reflect the reality of a single discipline, since swimming split times in a triathlon might be influenced by either cycling or running as subsequent disciplines.
Concerning the sex difference in open-water ultra-distance swimming, recent studies investigated the performances in the 26.4 km ‘Lake Zürich Swim’ , the 34 km ‘English Channel Swim’ [5, 8] and the ’12 h Swim of Zürich’ . In both the ‘Lake Zürich Swim’  and the ‘English Channel Swim’  men were faster than women. Depending from which side of the ‘English Channel’ the swim was accomplished the sex difference in swimming time differed from 6.7% for England-to-France to 8.9% for France-to-England . Eichenberger et al. [5, 7] stated that it would be unlikely that women would outperform men in ultra-distance swimming in the future. In the ’12 h Swim of Zürich’, however, women were able to accomplish similar distances as their male counterparts . The comparison of different events is even more complicated as for example ‘the English Channel Swim’ is not a race but has to be completed by each athlete individually with a supporting crew . Furthermore, in the ’12 h Swim of Zürich’  and the ‘Lake Zürich Swim’ , mainly recreational athletes compete.
Single human accomplishments recently called the attention of the long-distance swimming community. In the summer of 2013, Diana Nyad  crossed the sea between Havana, Cuba, and Key West, Florida, as the first person ever without a shark cage. She accomplished the 177 km swim within 53 hours . Another milestone in long-distance swimming history was set in 2013 when Christoph Wandratsch crossed the length of Lake Constance (66.7 km) in 20 h and 41 min . As Diana Nyad covered a considerably longer distance at the nearly the same swimming speed (3.3 km/h) like Christoph Wandratsch (3.2 km/h), it could be argued that the world’s best long-distance swimmers would be women rather than men. However, the achievements of Christoph Wandratsch and Diana Nyad are not directly comparable due to different water temperatures. Therefore, real open-water ultra-distance swim competitions such as the ‘Lake Zürich Swim’  are more suitable to investigate the sex difference in ultra-swimming performance than single achievements. However, the above mentioned competitions [5–7] investigated recreational swimmers and never included the entire world elite and more men than women competed in these events. The reported sex differences in ultra-swimming performance may therefore be biased by the number of finishers and the participation of non-professional athletes. Official events held in 5 km, 10 km and 25 km in the World Cup  would therefore be a better option to bypass these factors.
These recent controversial findings allow interpretations of both an increase in the sex difference in performance with increasing race distance or vice versa. However, to date, mainly recreational swimmers have been investigated in in the 26.4 km ‘Lake Zürich Swim’ , the 34 km ‘English Channel Swim’ [5, 8] and the ’12 h Swim of Zürich’  and there is a lack of data for elite swimmers. Additionally, the single performances of Diana Nyad and Christoph Wandratsch do not allow a generalization for the sex differences in ultra-distance swimming performance.
Since professionalism of athletes might influence the sex difference in performance , we intended to investigate the sex differences in swimming performance in 5 km, 10 km and 25 km open-water ultra-distance swimming races in professional athletes. Additionally, in a high-level swimming race, the fastest women may have the possibility to draft behind the fastest men. This might enable the fastest women to reduce the sex difference. The aims of the present study were therefore to investigate the changes in sex differences across years with increasing race distances from 5 km to 25 km. Based upon existing findings for recreational swimmers, we hypothesized for elite swimmers (i) that men would be faster than women from 5 km to 25 km, and (ii) the sex differences in performance would decrease with increasing race distance.