We investigated, firstly, the change in sex difference performance for the overall top ten finishers and the top ten swimmers, the top ten cyclists, and the top ten runners in Ironman Hawaii between 1983 and 2012, and, secondly, the change in the age of these triathletes across time.
Sex difference in performance for top ten overall
The present findings showed for the overall top ten finishers that overall race time and the split times in cycling and running decreased for both women and men, whereas the split time in swimming decreased for men, but not for women. However, the sex difference in performance remained unchanged for the split disciplines but decreased from 15.2% to 11.3% for overall race time. It seems that women were able to reduce the sex difference in overall performance in the Ironman Hawaii during the last years. The overall top ten women improved their overall race time by 15.6%, the overall top ten men only by 12.7%.
These results are different from the findings of Lepers  investigating the time period between 1981 and 2007 in the Ironman Hawaii. During that period, the sex difference in performance remained stable for swimming, increased for cycling, decreased for running, and remained unchanged for overall race time. These different findings are most probably due to the longer time period investigated in the present study. While Lepers investigated the time period 1981–2007 , we investigated the time period 1983–2012. Lepers  showed a steep decrease in overall race times between 1981 and 1988. However, overall race times changed not very much between 2007 and 2012 (Figure 2D).
In 2012, the sex difference in overall performance was at 11.3%. A sex difference of approximately 10–12% in Ironman triathletes seems to be of biological origin. Success in endurance performance is mainly determined by aerobic capacity and muscular strength. Because men possess a larger aerobic capacity and a greater muscular strength compared to women, the gap in endurance performances between women and men is unlikely to narrow naturally . Thibault et al.  reported a dominance of men when investigating the sex differences in different sports disciplines such as swimming, athletics, track cycling, weightlifting, and speed skating. They determined sex as a major variable of athletic performance by comparing the best performances of women and men in different sports. In every discipline, male performances were significantly better than female performances with a sex difference of 10.0 ± 2.9%.
The sex difference in performance of approximately 10–12% between female and male endurance athletes might be explained by physiological differences such as maximum oxygen uptake (VO2max)  and anthropometric characteristics such as skeletal muscle mass and body fat [26, 27]. Endurance-trained male triathletes have a VO2max of 61.3 ml·min−1·kg−1 compared to female triathletes with a VO2max of 52.8 ml·min−1·kg−1 and a corresponding sex difference of 8.5 ml·min−1·kg−1 (13.9%). It has been shown that male ultra-endurance athletes had a higher skeletal muscle mass than female ultra-endurance athletes [26–29]. Regarding ultra-runners, male ultra-runners with approximately 38 kg of skeletal muscle mass  had an approximately 28% higher muscle mass compared to female ultra-runners with approximately 27.4 kg of muscle mass . The higher skeletal muscle mass in men might be an important variable of performance for ultra-endurance athletes and may explain the sex difference. Also for Ironman triathletes, differences in anthropometric characteristics between women and men have been reported. Male Ironman triathletes with approximately 41 kg of skeletal muscle mass had an approximately 32% higher skeletal muscle mass compared to female Ironman triathletes with approximately 28 kg of skeletal muscle mass . The higher skeletal muscle mass may help men to sustain a higher speed for a longer time during an ultra-endurance race compared to women. When women and men were compared regarding their cycling performance, the main factor accounting for sex differences in peak and mean power output during cycling was skeletal muscle mass of the lower extremities . Women with a lower lean leg volume developed a lower peak power compared to men . Apart from skeletal muscle mass, also body fat might be responsible for the sex difference in performance. Body fat has been shown as an important predictor variable for male Ironman triathletes . The lower body fat in men might enhance their endurance performance. Male triathletes with 19.1% body fat have a 29% lower body fat compared to female triathletes with 26.9% body fat .
The sex difference in performance for split disciplines
For the ten fastest performers in each discipline, the cycling and running times improved across years, but not the swim times. The ten fastest cyclists and the ten fastest runners became faster across time, but the sex difference in performance remained unchanged. The sex difference was at 8.0 ± 2.4% in swimming, 12.7 ± 1.8% in cycling, and 15.2 ± 3.0% in running. The low sex difference in swim performance of 8.0 ± 2.4% is explained by the non-significant decrease in swim times of approximately 1 min for women and the non-significant change in swim times for men. Women swam 57 ± 3 min in 1983 and 56 ± 2 min in 2012, while men swam 51 ± 1 min in 1983 and 51 ± 1 min in 2012.
We assume that Ironman triathletes invested more time in the preparation to improve in those split disciplines with more importance for the race. The 3.8-km swim split accounts for only 1.7% of the total distance, whereas the 180-km cycling accounts for 79.6% and the marathon for 18.7% of the total race distance of 226 km. It has been shown for ultra-triathletes that both the cycling and the run splits were associated with overall race performance, but not the swim split .
Difference between the performances of the overall top ten finishers and the performances of the ten fastest swimmers, the ten fastest cyclists, and the ten fastest runners
When the performances between the overall top ten finishers and the performances of the ten fastest swimmers, the ten fastest cyclists, and the ten fastest runners were compared, both the ten fastest male and the ten fastest female swimmers swam faster than the overall ten fastest finishers. For cycling, the overall top ten finishers and the ten fastest cyclists achieved the same split times for both sexes. For running, the overall top ten finishers and the ten fastest marathoners achieved the same split times in women, but not in men. A potential explanation for the finding that the ten fastest male and the ten fastest female swimmers swam faster than the overall ten fastest finishers could be the ‘history’ of the athlete. It has been reported that many triathletes have previously been competitive swimmers . For recreational Ironman triathletes competing in ‘Ironman Lanzarote’, 28% of the athletes had a background as a runner, 14% as a swimmer, and 13% as a cyclist . For elite Ironman triathletes competing in Ironman Hawaii, however, the history of the athletes has not been investigated.
A more likely explanation for the better performance in the top ten swimmers compared to the overall top ten finishers could be the fact that wetsuits are prohibited in Ironman Hawaii. Wearing a wetsuit leads to a significantly lower swimming cadence (−14%), a significantly lower heart rate (−11%), and significantly lower lactate values (−47%) compared to swimming without a wetsuit . Moreover, cycling efficiency was significantly higher (+12.1%) after swimming with a wetsuit compared to swimming without a wetsuit. In an Ironman triathlon where athletes can wear a wetsuit in the swim split, athletes without the background as a swimmer may profit from wearing a wetsuit and achieve faster swim times.
However, in Ironman Hawaii, athletes with a background as a swimmer may swim fast in open water independent of whether they wear a wetsuit or not since they are used as swimmers to swimming fast in contrast to a weak swimmer. It has been shown that wearing a wet suit improves swim performance more in inefficient swimmers with low buoyancy when swimming at low speeds .
The age of the fastest Ironman triathletes
A further major finding was that the age of the overall top ten finishers and the ten fastest swimmers, the ten fastest cyclists, and the ten fastest runners increased across time for both women and men. Similar findings have been reported for athletes competing in Triple Iron ultra-triathlons and Deca Iron ultra-triathlons where the age of the finishers increased across years .
A potential explanation for the top ten athletes becoming older and faster in the same time could be the aspect of experience. In Ironman triathletes, the personal best time in an Olympic distance triathlon is a strong predictor variable for Ironman race time [36, 40–42]. In ultra-triathletes competing in distances longer than an Ironman such as a Triple Iron ultra-triathlon, the personal best time in an Ironman triathlon and in a Triple Iron triathlon were positively and highly significantly related to overall race time . In ultra-triathletes finishing a Deca Iron ultra-triathlon, both the number of finished Triple Iron ultra-triathlons and the personal best time in a Triple Iron ultra-triathlon were related to overall race time . The number of Olympic distance triathlons worldwide is high; however, the number of Ironman triathlons worldwide is low (http://ironman.com). Since Ironman triathlons as a means to gain experience in long-distance triathlons are little available, the number of experienced athletes in distances of longer than an Ironman triathlon will become dramatically reduced. While in 2012, 1,984 athletes started in Ironman Hawaii and 1,886 starters (95%) finished; a total of 1,686 athletes started in a Double Iron ultra-triathlon between 1985 and 2009 . Of these 1,686 starters, 1,333 athletes finished (79.1%).
Apart from this aspect, recent studies showed that the performance in master Ironman triathletes improved [13, 24]. In ‘Ironman Switzerland’, a qualifier for Ironman Hawaii, the number of master triathletes increased, and their performance improved . In Ironman Hawaii, men older than 44 years and women older than 40 years improved both split times and overall race times . Since in both Ironman Switzerland  and Ironman Hawaii , the participation in master triathletes increased, and these master triathletes improved their performance across years; this might explain that also the top ten athletes may become older across years but still seem to be able to improve their performance.
This study is limited since potential predictor variables for an ultra-endurance performance such as body composition , previous experience , training intensity , nutrition [45, 46], motivation [46, 47], and weather conditions [14, 48] were not considered. These variables might have influenced race outcome.