The age-related changes and sex difference in master swimming performance

Beat Knechtle, Tiago M. Barbosa, and Pantelis Theo Nikolaidis 1 Medbase St. Gallen Am Vadianplatz, St. Gallen, Switzerland 2 Institute of Primary Care, University of Zurich, Zurich, Switzerland 3 National Institute of Education, Nanyang Technological University, Singapore, Singapore 4 Polytechnic Institute of Bragança, Bragança, Portugal 5 Research Centre in Sports, Health and Human Development, Vila Real, Portugal 6 Exercise Physiology Laboratory, Nikaia, Greece 7 School of Health and Caring Sciences, University of West Attica, Athens, Greece


Introduction
Competitive swimming is a sports discipline which can be performed in different swim strokes, over different distances and in both pools (i.e., indoor or outdoor) and open water (i.e., sea, lake, river). The different swim strokes are butterfly, backstroke, breaststroke, frontcrawl and the combination of the four strokes as individual medley. In addition to these individual events, four swimmers can take part in either a freestyle or medley relay. In swimming competitions, distances for butterfly, backstroke, breaststroke and freestyle are commonly 50, 100 and 200 m whereas, individual medley is held over 200 and 400 m. In freestyle swimming, 400, 800 and 1500 m are additional race distances. In open-water swimming, most often events over 3, 5, 10 and 25 km are held. The 5, 10 and 25 km are held as official FINA (Fédération Internationale de Natation) races for elite swimmers (Baldassarre, Bonifazi, Zamparo, & Piacentini, 2017;Veiga, Rodriguez, González-Frutos, & Navandar, 2019) whereas the 3 km are especially designed for master swimmers (Knechtle, Nikolaidis, Rosemann, & Rüst, 2017a).
A solid body of knowledge has been produced over time about the changes of swimming performance over lifespan. Plenty of research has been carried out on performance impairment by older swimmers (i.e., master swimmers). However, very limited reviews are available consolidating the age-related differences in master swimming performance. In this short review, we summarise recent findings about the age-related performance decline in master swimming. We consider swimming as the single discipline in pool and open-water swimming, but also as one of the three disciplines in triathlon.
For this purpose, the data base PubMed (www.ncbi. nlm.nih.gov/pubmed/) was checked for the term "agerelated performance decline swimming" leading to 42 results. We then extracted all relevant studies for swimming and where swimming was part in a multi-sports discipline event.
2 The age-related performance decline in general in sport It is well-known that performance decreases with increasing age. However, the quantification of the decrease remains to elucidate. For instance, the knowledge of the percentage decrease especially between consecutive age groups would be of great practical value for master athletes. This decrease seems, however, to depend upon the sport discipline. When track and field athletes were analysed, an accelerated decline beyond the age of 70 years in sprint, middle-and long-distance running was found, while in throwing and jumping disciplines the performance continued a linear decline (Ganse, Ganse, Dahl, & Degens, 2018). The age-related decline has very often been investigated in triathlon. When the same individual performs in the three disciplines, a smaller age-related decline in cycling performance than in running and swimming performances was observed .
The distance of an athletic event seems to be crucial for the age-related performance decline. In triathlon, the agerelated performance decline seems to be specific to the race distance (Lepers, Sultana, Bernard, Hausswirth, & Brisswalter, 2010;Stones & Hartin, 2017), the type of triathlon event (e.g., Olympic distance, Ironman) (Lepers & Stapley, 2011) and the discipline (i.e., swimming, cycling and running) (Käch, Rüst, Nikolaidis, Rosemann, & Knechtle, 2018). When the age-related declines in swimming, cycling, and running performances for both Olympic distance triathlon and Ironman triathlon were compared, the magnitude of the declines in cycling and running performance in the Ironman distance was greater than in the Olympic distance triathlon. The age-related decline in triathlon performance is specific to the discipline, with cycling showing lower declines in performance with age than swimming and running. There was a lesser age-related decline in cycling compared with running and swimming after the age of 55 years in the Olympic distance triathlon and after the age of 50 years in the Ironman distance. With advancing age, the performance decline was less pronounced in the Olympic distance triathlon than in the Ironman triathlon in cycling (> 55 years) and running (> 50 years). In contrast, an agerelated decline in swimming performance seemed independent of triathlon distance (Lepers, et al., 2010).
However, in triathlon, there seems to be a difference when cycling and running are performed off-road. When the age-related declines in swimming, cycling, and running performances for road-based and off-road triathlons were analysed, there was a less pronounced age-related decline in cycling performance compared to running and swimming for road-based triathlon independently of age. On the other hand, the age-related decline did not differ between the three locomotion modes (i.e., triathlon legs) for off-road triathlon. With advancing age, the performance decline was less pronounced in road-based than offroad triathlon in swimming (≥ 65 years), cycling (≥ 50 years), running (≥ 60 years), and overall race time (≥ 55 years). These results suggest that the rate of the decline in performance in off-road triathlon is greater than roadbased triathlon, indicating that the type of discipline (i.e., road versus mountain bike cycling and road versus trail running) exerts an important influence on the magnitude of the age-associated changes in triathlon performance (Lepers & Stapley, 2011).
The age-related performance decline in triathlons seems also specific to the disciplines. For Ironman triathletes, the age-related performance decline started in women in age group 25-29 years in swimming and in age group 30-34 years in cycling, running, and overall race time, whereas it started in men in age group 25-29 years in swimming and in age group 35-39 years in cycling, running, and overall race time (Käch, et al., 2018). In Half-Ironman triathlon, the age-related performance decline seems to start in swimming at a higher age compared to the other disciplines. The decline in performance began at a younger cohort age for running (35-39 years) than for swimming (50-55 years) (Stones & Hartin, 2017).
The age-related decline has also been investigated in competitive swimming, cycling and running as single sports events. The age-related decline in swimming, cycling, and running for endurance and sprint events were investigated (i.e., swimming 1500 m and 50 m, cycling 1 h and 200 m, and running 10 km and 100 m). The age-related impairment in 1 h track cycling was less pronounced than in 1500 m swimming and in 10 km running after the age of 60 years. Furthermore, the age-related decline was similar among the three locomotion modes in sprinting events (Lepers, Stapley, & Cattagni, 2018).
Race times of master athletes competing from 1975 to 2013 in the 100 m running, the 400 m running and the 100 m freestyle swimming events were analysed. Master athletes improved their athletic performance significantly and progressively over the years. The improvements in the fastest 100 m running times over time were not significant. However, all the masters' age group records improved significantly over time. The slopes of improvements over the years were progressively greater at older age groups with the greatest progression observed at oldest age groups of 75-79 years old examined. The general trends were similar for 400 m middle-distance running and 100 m freestyle swimming events (Akkari, et al., 2015). This improvement might have an impact on the performance decline with increasing age.
An improvement in performance has been especially shown for master triathletes competing in the Olympic distance (Lepers & Stapley, 2011;Wonerow, et al., 2017) and the Ironman distance (Gallmann, et al., 2014;Lepers, Rüst, et al., 2013) triathlon. For age group triathletes competing in the Olympic distance triathlon at the International Triathlon Union (ITU) World Championships 2009-2014, women and men improved performance in most age groups across all years. Swimming performance improved in athletes in age groups 25-29 years to 55-59 years, but not in younger (i.e., 18-24 years) and older (i.e. > 60-64 years) age groups. Cycling performance improved in athletes in age groups 18-24 years to 70-74 years, but not in age group 75-79 years. In running, athletes in age groups 18-24 years, 30-34 years, 35-39 years and 65-69 years improved, but not in the other age groups. Overall race time was improved by athletes in age groups 18-24 to 65-69 years, but not in age groups 70-74 years and 75-79 years (Wonerow, et al., 2017). When changes in participation and performance trends of older (i.e. > 40 years of age) triathletes competing between 1986 and 2010 in "Ironman Hawaii" were analysed, the relative participation of master triathletes increased while the participation of triathletes younger than 40 years of age decreased. Men older than 44 years old and women older than 40 years old significantly improved their performances in swimming, cycling, running and in overall race time. Sex differences in overall race time decreased in all age groups between the 40-44 and 55-59 years old . Very unique for "Ironman Hawaii" is the fact the age of annual top 10 female and male triathletes in the "Ironman Hawaii" increased over the last three decades while their performances improved (Gallmann, et al., 2014). Similar findings were reported also in an "Ironman Hawaii" qualifying event. In "Ironman Switzerland", triathletes older than 40 years old significantly improved their performance in swimming, cycling, running split times and in overall race time (Stiefel, et al., 2014).
However, also master swimmers competing at the FINA World Championships in pool-swimming in freestyle (Knechtle, Nikolaidis, Konig, Rosemann, & Rüst, 2016), in backstroke (Unterweger, et al., 2016), in butterfly (Knechtle, Nikolaidis, Rosemann, & Rüst, 2017b), in breaststroke , in individual medley  and in 3000 m open-water swimming (Knechtle, et al., 2017a) improved their performances. Although we know that endurance performance decreases with increasing age, the events' distance might be of importance. When the age-related declines in swimming, cycling, running and overall race time for Triple Iron ultratriathlon (i.e., 11.4 km swimming, 540 km cycling and 126.6 km running) and Deca Iron ultra-triathlon (i.e., 38 km swimming, 1800 km cycling and 420 km running) were analysed, the fastest overall race times were achieved between the ages of 25 and 44 years old for both distances. However, Deca Iron ultra-triathletes achieved the same level of performance in swimming and cycling between 25 and 54 years of age, which is higher than for shorter race distances. A potential explanation could be the age of the triathletes in the longer race distance because the mean age of the finishers was significantly higher for Deca Iron ultra-triathletes (41.3 ± 3.1 years) compared to a Triple Iron ultra-triathletes (38.5 ± 3.3 years) (Knechtle, Rüst, Knechtle, Rosemann, & Lepers, 2012).
4 The age-related performance decline in pool-swimming Also in competitive pool-swimming, the age-related performance decline is well-known (Bongard, McDermott, Dallal, & Schaefer, 2007;Fairbrother, 2007aFairbrother, , 2007bLapierre, Baker, & Tanaka, 2018;Tanaka & Seals, 1997) where the decline appears to be associated with the length of the swim distance. Tanaka and Seals (Tanaka & Seals, 1997) analysed the age-related performance decline in 50 to 1500 m freestyle swimming by performing a retrospective analysis of top freestyle performance times from the US Masters Swimming Championships. In both men and women, performance in 1500 m declined linearly from peak levels at the age of 35-40 years old until approximately to the age of 70 years old, whereupon performance declined exponentially thereafter. Performance in 50 m freestyle showed only a moderate decline until the age of 75 and 80 years in both women and men (Tanaka & Seals, 1997).
Potential explanations for the slower swimming speed with increasing age are physiological, biomechanical changes and an increase in energy cost of swimming. The decrease in swimming performance with increasing age is a not only consequence of the physiological decrease in maximal metabolic power but can also be expected to depend on an increase in the energy cost of swimming. It has been shown that older master athletes (60-80 years) swam at a significantly slower pace, with a lower propelling efficiency and a larger projected frontal area than younger master athletes (30-60 years) (Zamparo, Dall'ora, Toneatto, Cortesi, & Gatta, 2012). However, also slowing in swimming is due to biological or behavioural factors affecting physical performance (Fairbrother, 2007b). It is well-known that physiological functional capacity decreased only modestly until the age of 60-70 years, but declined exponentially thereafter (Tanaka & Seals, 2003). A further aspect is training. It has been shown that master athletes spend significantly less time per week training for competition, and their training focus is on endurance, not strength. In contrast, younger swimmers train for endurance, strength, speed, and power, and thus, their performance relies mostly on anaerobic energy transfer systems (Weir, Kerr, Hodges, McKay, & Starkes, 2002). Moreover, a comparison between "young" masters and elite swimmers indicated that elite differed from master swimmers in terms of their larger anaerobic contribution which was in agreement with their different training characteristics where master's training was mainly characterized by a lower volume and intensity than the elite ones (Mejias, et al., 2014).
When the one-hour swimming performance of 4271 presumably healthy men and women, aged 19-91 years, from the 2001-2003 United States Masters Swimming long-distance (1 h) national competition were analyzed, the declines in performance were parallel in men and women at all ages, and the 1-year age-related declines in performance were about twice as great at 40 years old and more than four-times as great at 80 years than at 20 years of age (Bongard, et al., 2007). This finding is similar to the age-related performance decline in master swimmers competing in freestyle, the swimming speed in the age group world records from 25-29 to 100-104 age-groups shows a non-linear decrease for all distances (Fig. 1). Interestingly, only women were able to swim in age group 100-104 years in the longer freestyle distances (i.e., 200 to 1500 m).

The age-related performance decline in open-water swimming
For open-water swimming, the age-related performance decline has mainly been investigated for the swim split in triathlon races of different distances such as Olympic distance and Ironman distance.
It seems that differences do exist between the different distances. The age-related decline in triathlon performance seems to be specific to the discipline, with cycling showing fewer declines in performance with age than swimming and running. When swimming, cycling, running and overall race times of the top 10 males between 20 and 70 years of age were analysed in 5-years-intervals for two consecutive World championships (2006 and 2007) for Olympic and Ironman distances, a less pronounced agerelated decline in cycling performance compared with running and swimming after the age of 55 years for Olympic distance and after the age of 50 years for Ironman distance was found. With advancing age, the performance decline was less pronounced for Olympic than for Ironman triathlon in cycling (> 55 years) and running (> 50 years), respectively. In contrast, an age-related decline in swimming performance seemed independent of triathlon distance (Lepers, et al., 2010).
There also seem to exist differences between the sexes regarding the age-related performance decline for the different distances. In Olympic distance triathlon, the agerelated declines in performance were significantly less pronounced for males compared with females for swimming (> 50 years), for cycling (> 40 years) and for overall race time (> 40 years) (Etter, Knechtle, Rüst, Rosemann, & Lepers, 2013). For Ironman triathletes, the age-related decline in performance started earlier in swimming than in cycling and running. In women, the age-related performance decline started in age group 25-29 years in 6 Sex difference in the age-related performance decline Differences between women and men regarding the agerelated performance decline have been described. For example, in track and field athletes, the steepest declines were observed in javelin throw and 400 m for women and in pole vault and 800 m for men (Ganse, et al., 2018). When the differences in performance between women and men and changes in performance in selected age categories at the ITU Cross World Championships held between 2011 and 2016 were analysed, the best results in cycling and total race time were obtained in age group 30-34 years-old and in swimming and running in age group 40-44 years old in men. Women, however, obtained the best results in running in age group 25-29 years, in cycling in age group 30-34 years, and in both swimming and overall race time in age group 35-39 years (Romero-Ramos, Fernández-Rodríguez, Merino-Marbán, Mayorga-Vega, & Podstawski, 2018). Tanaka & Seals (1997) analysed the age-related performance decline in 50 to 1500 m freestyle swimming by performing a retrospective analysis of top freestyle performance times from the US Masters Swimming Championships and showed that the decrease was linear until the age of ∼70-80 years old whereupon the decline becomes exponential. Moreover, the rate of decline with advancing age appears to be associated with event's duration and sex. The percent of sex difference in performance throughout the age range studied became progressively smaller with increasing distance from 50 m (19 ± 1%) to 1500 m (11 ± 1%). The rate and magnitude of the declines in both 50 and 1500 m swimming performance with age were significantly greater in women than in men. In the women, the percent decline in swimming performance over 50 years old from the age group 19-24 years to age group 69-74 years became progressively greater from the 50 to the 1500 m; whereas in men, no differences were observed in the magnitude of performance decline with age among the five longest distance events (i.e., 100 to 1500 m) (Tanaka & Seals, 1997).
When we consider the sex difference in performance in freestyle swimming from 50 to 1500 m for the master world records (Fig. 2), the sex difference increases in a non-linear fashion-way in both 50 and 100 m after the age of ∼ 90 years; whereas in longer distances (i.e., 200 to 1500 m), sex difference increases rather linearly. There is, however, also a sex difference in swimming when swimming is the first discipline in a triathlon. In Olympic distance triathletes, the age-related declines in performance were significantly less pronounced for males compared with females for swimming (> 50 years), for cycling (> 40 years) Fig. 1. Speed in freestyle swimming by age group (from 25-29 to 100-104 years) and distance (50-1500 m) in men and women. The figure is drawn with data on master world records from http://archives.fina.org/database/main/records.php (access date 1/11/2018). R 2 coefficient of determination for fourth degree polynomial regression ranged from 0.99 to 1.00. The figure was created using GraphPad Prism v. 7.0 (GraphPad Software, San Diego, USA). and for overall race time (> 40 years) (Etter, Knechtle, Rüst, Rosemann, & Lepers, 2013).
Studies investigated the sex difference in performance in master swimmers competing at the FINA World Championships in pool swimming in freestyle (Knechtle, Nikolaidis, Konig et al., 2016), backstroke (Unterweger, et al., 2016), butterfly (Knechtle et al., 2017b), breaststroke , individual medley  and in 3000 m openwater swimming (Knechtle, et al., 2017a). When age groups 25-29 to 90-94 years competing at the FINA World Masters Championships between 1986 and 2014 were analyzed, in older age groups, women were able to achieve a similar performance to men. In freestyle events, women were slower than men in age groups 25-29 to 75-79 years, but not in the age groups 80-84 and 85-89 years. In 50, 100 and 200 m, women reduced the sex difference in age groups 30-34 to 75-79 years. In 400 m, women reduced the gap to men in age groups 40-44, 45-49, and 55-59 years. In 800 m, sex difference became reduced in age groups 55-59 and 70-74 years (Knechtle, Nikolaidis, Konig et al., 2016). In backstroke swimming, women were slower than men in age groups 25-29 to 80-84 years, but not in age groups 85-89 to 95-99 years over all distances. In 50 and 100 m, the sex difference decreased in age groups 40-44, 45-49, 50-54, and 55-59 years. In 200 m, the sex difference decreased in age groups 40-44 and 90-94 years, but increased in age group 25-29 years (Unterweger, et al., 2016). In butterfly swimming, men were faster than women from 25-29 to 85-89 years but not for 90-94 years. Women were able to reduce the sex difference in performance in several age groups. In 50 m butterfly, women reduced the sex difference in age groups 30-34 to 60-64 years. In 100 m butterfly, women decreased the gap to men in age groups 35-39 to 55-59 years. In 200 m, the sex difference was reduced in age groups 30-34 to 40-44 years (Knechtle, et al., 2017b). In breaststroke swimming, men were not faster than women age groups 90-94 to 95-99 years. In 50 m, women reduced the gap to men in age groups 40-44 to 70-74 years. In both 100 and 200 m, women reduced the gap in age groups 50-54 to 60-64 years . In 200 and 400 m individual medley, men were faster than women from 25-29 to 80-84 years, but not in 85-90 and 90-94 years. Women reduced the gap to men in 40-44 and 45-49 years in 200 m, and 40-44 years in 400 m . Finally, in 3000 m openwater swimming, women were slower in age groups 25-29 to 70-74 years. In age groups 75-79 and 85-89 years, however, race times were similar for both women and men (Knechtle, et al., 2017a). Fig. 2. Sex difference in speed in freestyle swimming by age group (from 25-29 to 100-104 years) and distance (50-1500 m) in men and women. The figure is drawn with data on master world records from http://archives.fina.org/database/main/records.php (access date 1/11/2018). R 2 coefficient of determination refers to fourth degree polynomial regression. The sex difference was calculated using the formula "100 Â (speed in menspeed in women)/speed in women". The figure was created using GraphPad Prism v. 7.0 (GraphPad Software, San Diego, USA).
A potential explanation for the improved performance in elderly women is the fact that peak performance ages in women have increased significantly since 1980s. When 118 years from the first Olympics (1896) to the most recent Olympics (2014) were analysed regarding the ages at which peak performance was observed, peak performance ages in women have increased significantly since 1980s and consistently in all the athletic events examined (i.e., track and field, swimming, rowing, and ice skating events). Interestingly, as women's peak performance age increased, they became similar to men's peak ages in many events. In the last 20-30 years, ages at which peak athletic performance is observed have increased in women but not in men (Elmenshawy, Machin, & Tanaka, 2015). In addition, biomechanical differences between women and men would partially explain sex differences in the age-related decline of performance (Ferreira, et al., 2015;Seifert, Barbosa, & Kjendlie, 2011).

Conclusion
In summary, master athletes competing in swimming as single discipline (i.e., pool-swimming and open-water) and in triathlon (i.e., swim split as first discipline) improved their performances across calendar years. The age-related performance decline in swimming seems to be specific to the discipline, the sex and the length of the swimming event.