Pacing Profiles of Middle-Distance Running World Records in Men and Women

the current study were to compare the pacing patterns of all-time 800 m, 1500 m and mile running world records (WRs) and to determine whether differences exist between sexes, and if 800 m and 1500 m WRs were broken during championship or meet races. Overall and lap times for men and women's 800 m, 1500 m, and mile WRs from World Athletics were collected when available and subsequently compared. A fast initial 200 m segment and a decrease in speed throughout was found during 800 m WRs. Accordingly, the rst 200 m and 400 m were faster than the last 200 m and 400 m, respectively (p< 0.001, 0.77 ES 1.86). The rst 400 m and 409 m for 1500 m and mile WRs, respectively, were faster than the second lap (p< 0.001, 0.74 ES 1.46). The third

in speed throughout was found during 800 m WRs. Accordingly, the rst 200 m and 400 m were faster than the last 200 m and 400 m, respectively (p< 0.001, 0.77 ES 1.86). The rst 400 m and 409 m for 1500 m and mile WRs, respectively, were faster than the second lap (p< 0.001, 0.74 ES 1.46). The third 400 m lap was slower than the last 300 m lap and 400 m lap for 1500 m and mile WRs, respectively (p< 0.001, 0.48 ES 1.09). No relevant sex-based differences in pacing strategy were found in any event. However, the rst 409 m lap was faster than the last 400 m lap for men but not for women during mile WRs. Women achieved a greater % of WRs than men during championships (80% vs. 45.83% in the 800 m, and 63.63% vs. 31.58% in the 1500 m, respectively). In conclusion, positive, reverse J-shaped and U-shaped pacing pro les were used to break 800 m, men's mile and 1500 m, and women's mile WRs, respectively. WRs are more prone to be broken during championships by women than men. Keywords:pacing; middle-distance running; world record; athletics 1. Introduction Pacing, described as the work or effort distribution over a race, has been extensively studied over the last 40 years in endurance sports [1]. There is consensus that pacing is a prerequisite to achieve successful endurance performance, and that depends on several internal (i.e., muscle fatigue [2], and psychophysiological variables [3]) and external (i.e., tactics [4], and ambient conditions, such as wind resistance [5]) factors [6]. Ideally, an even effort distribution would be the most optimal pacing strategy from an energetic point of view [7], but the accumulated evidence demonstrates that different pacing pro les can be observed mostly depending on the distance/duration of the race [8]. The analysis of pacing strategies of track running world records (WRs) is an excellent paradigm of study because they were used to achieve the most optimal and outstanding performances in history. These strategies may be considered by runners exhibiting lower performance and help them to improve it

les can be observed mostly depending on the distance/duration of the race [8]. The analysis of pacing strategies of track running world records (WRs) is an excellent paradigm of study because they were used to achieve the most optimal and outstanding performances in history. These strategies may be considered by runners exhibiting lower performance and help them to improve it through a learning process. Interestingly, while cross-sectional data suggest a reduced variability of velocity with WRs improvements over different distances [7,9], a more recent analysis of WRs performed by the same athletes suggest that these individuals' running performances can be improved without changes in Int. J. Environ. Res. Public Health2021,18, 12589.

Int. J. Environ. Res. Public Health2021,18, 12589 2 of 9 their own pacing strategies [10]. Pacing strategies in men during 800 m [11], 1500 m [12] and mile [13] running WRs were previously studied. However, the analysis of men's 800 m WRs is not updated in the current literature, with the latest three WRs achieved by Kenyan runner David Lekuta Rudisha. Therefore, considering the new WRs and the potential impact of exceptional WR holders on performance and pacing, such as Rudisha for 800 m, it is required an updated analysis of pacing strategies of middle-distance running races in which the contribution of anerobic capacity is more relevant [14] than in other endurance running races. In addition, to date, no previous study has analyzed the pacing strategies during women's 800 m and 1500 m WRs. Sex in uence on pacing behavior is also another topic of interest in pacing research, as it would be expected that internal factors associated to sex differences may be relevant. A previous study compared both men's and women's pacing strategies from marathon WRs [15] and found that women tended to follow a less uniform pace, while men typically adopted a more even pace with a fast end spurt at the nal stages of the race. Regarding middle-distance running WRs, Foster et al. [7] observed that the pacing pattern of men's mile WRs is characterized by a progressive reduction in the within-lap variation of pace, while in women, the pattern of lap times has almost not changed over time, likely secondary to a lack of performance depth in the women's elds. However, this previous analysis of mile WRs in women was conducted up to 1996. In addition, a sex-based comparison between 800 m and 1500 m WRs is warranted to better understand whether physiological differences between sexes in uence pacing patterns over middle-distance runs. The use of pacemakers strongly assist for the achievement of the fastest possible nish- ing performance by means of a reduction in the cognitive load associated to a continuous decision-making process [3,16], and also allowing WR aspirants to take advantage of draft- ing

is warranted to better understand whether physiological differences between sexes in uence pacing patterns over middle-distance runs. The use of pacemakers strongly assist for the achievement of the fastest possible nish- ing performance by means of a reduction in the cognitive load associated to a continuous decision-making process [3,16], and also allowing WR aspirants to take advantage of draft- ing [3,5]. However, pacemakers are typically used during meets (i.e., non-championship races in which the main goal is to achieve the fastest nishing performance) rather than championship races [8]. In addition, setting a WR is a very different goal to winning a gold medal during major championships. Therefore, it is expected that middle-distance running WRs were achieved during meets rather than championship races. However, WRs may also be achieved during championship races, at which world-class athletes typically peak. Nonetheless, to date, no previous study analyzed the type of race (i.e., meets vs. championship races) in which WRs were broken. Therefore, the aims of the current study were (1) to describe and compare the pacing pro les of all-time middle-distance running WRs, (2) to verify whether differences exist between men and women, and (3) to determine whether 800 m and 1500 m WRs were broken during either championship or meet races. 2. Materials and Methods 2.1. Pacing Data Overall and split times recorded during 800 m (21 men and 10 women), 1500 m (37 men and 10 women) and mile (32 men and 9 women) world records (WRs) from the World Athletics (WA, formerly International Amateur Athletic Federation (IAAF)) era until 2014 were collected from the Hymans and Metrahazi [17] database when available. WRs rati cation by WA and 2 other WRs broken from 2015 to 2020 were extracted from the WA website (www.worldathletics.org 2.2. Design and Methodology The present study followed an observational approach. In the men's 800 m event, 24 WRs were rati ed by WA from 1912 to 2012. However, WRs with split times in yards or without split times were excluded. Finally, 400 m lap times were available for 13 races, while 200 m lap times were available

website (www.worldathletics.org 2.2. Design and Methodology The present study followed an observational approach. In the men's 800 m event, 24 WRs were rati ed by WA from 1912 to 2012. However, WRs with split times in yards or without split times were excluded. Finally, 400 m lap times were available for 13 races, while 200 m lap times were available for the remaining 8 races. They represented 87.5% of all WRs. In women's 800 m events, 29 WRs from 1922 to 1983 were rati ed by WA. However, split times were only available for 10 WRs, which represented 34.5% of all WRs. Finally, 200 m and 400 m lap times were available in 2 and 8 WRs, respectively.

Int. J. Environ. Res. Public Health2021,18, 12589 3 of 9 In the men's 1500 m event, 38 WRs were rati ed by WA from 1912 to 1999. Lap times (3 400 m lap times and the last 300 m) were available for 37 men's WRs from 1917 to 1998. They represented 97.4% of all WRs. In the women's 1500 m event, 14 WRs were rati ed by the WA from 1967 to 2015. Only split times for 11 WRs were available. They represented 78.6% of all WRs. In the men's mile event, 32 WRs were rati ed by WA from 1913 to 1999. Lap (a rst lap of 409 m and 3 400 m laps) times were available in all these WR. In the women's mile event, 14 WRs from 1967 to 2019 were rati ed by WA. Split times were available in 10 WRs, which represented 71.4% of all WRs. Each lap time was expressed as a percentage of the average race speed (%RS) for further comparisons. Categorization of the different pacing strategies was conducted according to the statistical differences found between lap times. For example, a U-shaped pacing strategy during either 1500 m or mile events is considered if the rst and second laps are covered at signi cantly faster paces than those during the second and third laps, without signi cant differences between paces during the rst and last laps. Rather, a reverse J-shaped pacing strategy is considered if the rst and last laps are covered at signi cantly faster paces than those during the second and third laps, and the rst lap is completed at a signi cantly faster pace than that during the last lap. 2.3. Statistical Analysis All data are presented as mean and standard deviation (mean SD). Data were checked for normality of distribution, equality of variances, and assumption of sphericity. When the sphericity assumption was violated, the Greenhouse–Geisser correction was employed. A 2-factor analysis of variance (ANOVA) with repeated measures with `race average speed at each lap' as the between laps' factor and sex as the between subjects' factor was conducted to

standard deviation (mean SD). Data were checked for normality of distribution, equality of variances, and assumption of sphericity. When the sphericity assumption was violated, the Greenhouse–Geisser correction was employed. A 2-factor analysis of variance (ANOVA) with repeated measures with `race average speed at each lap' as the between laps' factor and sex as the between subjects' factor was conducted to determine the differences between %RS at each lap and between sexes. A Bonferroni post hoc correction was used in all pairwise comparisons. Effect sizes (ES) were calculated using partial eta-squared ( p2) for the repeated measured ANOVA test, and Cohen's d [18] for the Bonferroni post hoc test. The p2 was considered to be small (0.01), moderate (0.01–0.06) or large (>0.15) [19]. The Cohen's d was considered to be small (0.21–0.50), moderate (0.51–0.80) or large (>0.80) [18]. Statistical signi cance was set atp< 0.05. All analyses were performed with the JASP software (version 0.13.1 for Mac OS, JASP Team, Amsterdam, the Netherlands). Figures were performed with the Graph Pad Prism software (version 8.0 for Mac) (San Diego, CA, USA). 3. Results The repeated measures ANOVA revealed a non-signi cant difference in %RS between sexes for all middle-distance running events (Table). However, there were signi cant differences in %RS within all middle-distance events in both sexes. Table 1.Repeated measures ANOVA outcomes of lap and lap and sex interaction. Repeated Measures ANOVA Between Laps Lap and Sex Interaction Event Number of Laps df Residual p F df ES p F ES 800 m 2 19 <0.00131.55 1 0.6230.978 0.00 0.000 4 21 <0.00110.029 3 0.5610.483 0.483 0.047 1500 m 4 138 <0.00131.40 3 0.3970.358 1.08 0.014 Mile 4 95.80 <0.00139.54 2.39 0.4920.486 0.771 0.010 F—variation between sample means/variation within the sample; df—degrees of freedom; ES—effect size ( p 2 ). In the 800 m event (Figure), the rst 400 m lap was covered at a signi cantly faster speed than the second 400 m lap (p< 0.001, ES = 0.99 and 0.77 for men and women,

of freedom; ES—effect size ( p 2 ). In the 800 m event (Figure), the rst 400 m lap was covered at a signi cantly faster speed than the second 400 m lap (p< 0.001, ES = 0.99 and 0.77 for men and women,

Int. J. Environ. Res. Public Health2021,18, 12589 4 of 9 respectively). The rst 200 m lap was also covered at a signi cantly faster speed than the fourth 200 m lap for men (p< 0.001, ES = 1.86).Int. J. Environ. Res. Public Health 2021, 18, 12589 4 of 9 1500 m 4 138 <0.001 31.40 3 0.397 0.358 1.08 0.014 Mile 4 95.80 <0.001 39.54 2.39 0.492 0.486 0.771 0.010 F—variation between sample means/variation within the sample; df—degrees of freedom; ES — effect size (ηp 2 ). In the 800 m event (Figure 1), the first 400 m lap was covered at a significantly faster speed than the second 400 m lap (p < 0.001, ES = 0.99 and 0.77 for men and women, respec- tively). The first 200 m lap was also covered at a significantly faster speed than the fourth 200 m lap for men (p < 0.001, ES = 1.86). Figure 1. Mean and standard deviation of (A) average race speed of two 400 m laps, (B) four 200 m laps during male and female 800 m world record performances, and (C) world records set by Rud- isha and Kratochvilova with two 400 m laps. ** p < 0.01; *** p < 0.001. In the 1500 m event (Figure 2), the first lap was covered at a significantly faster speed than the second lap (p < 0.001, ES = 0.91 and 0.78 for men and women, respectively). The second lap was covered at a significantly slower speed than the last 300 m lap (p < 0.001 and p = 0.05, ES = 1.02 and 0.65 for men and women, respectively). The third lap was cov- ered at a significantly slower speed than the last 300 m split ( < 0.001, ES = 0.86 and 0.48 for men and women, respectively). There were no differences between %RS at the first lap and the last 300 m lap for men or women. Figure 2. Mean and standard deviation of (A) average race speed of three 400 m laps and last 300 m during men and women’s 1500 m world

( < 0.001, ES = 0.86 and 0.48 for men and women, respectively). There were no differences between %RS at the first lap and the last 300 m lap for men or women. Figure 2. Mean and standard deviation of (A) average race speed of three 400 m laps and last 300 m during men and women’s 1500 m world record performances (A), and (B) world records set by El Guerrouj and Dibaba. * p < 0.05; *** p < 0.001. Figure 1. Mean and standard deviation of (A) average race speed of two 400 m laps, (B) four 200 m laps during male and female 800 m world record performances, and (C) world records set by Rudisha and Kratochvilova with two 400 m laps. **p< 0.01; ***p< 0.001. In the 1500 m event (Figure), the rst lap was covered at a signi cantly faster speed than the second lap (p< 0.001, ES = 0.91 and 0.78 for men and women, respectively). The second lap was covered at a signi cantly slower speed than the last 300 m lap (p< 0.001 andp= 0.05, ES = 1.02 and 0.65 for men and women, respectively). The third lap was covered at a signi cantly slower speed than the last 300 m split (p< 0.001, ES = 0.86 and 0.48 for men and women, respectively). There were no differences between %RS at the rst lap and the last 300 m lap for men or women.Int. J. Environ. Res. Public Health 2021, 18, 12589 4 of 9 1500 m 4 138 <0.001 31.40 3 0.397 0.358 1.08 0.014 Mile 4 95.80 <0.001 39.54 2.39 0.492 0.486 0.771 0.010 F—variation between sample means/variation within the sample; df—degrees of freedom; ES — effect size (ηp 2 ). In the 800 m event (Figure 1), the first 400 m lap was covered at a significantly faster speed than the second 400 m lap (p < 0.001, ES = 0.99 and 0.77 for men and women, respec- tively). The first 200 m lap was also covered at a significantly faster speed than the fourth 200 m lap for

). In the 800 m event (Figure 1), the first 400 m lap was covered at a significantly faster speed than the second 400 m lap (p < 0.001, ES = 0.99 and 0.77 for men and women, respec- tively). The first 200 m lap was also covered at a significantly faster speed than the fourth 200 m lap for men (p < 0.001, ES = 1.86). Figure 1. Mean and standard deviation of (A) average race speed of two 400 m laps, (B) four 200 m laps during male and female 800 m world record performances, and (C) world records set by Rud- isha and Kratochvilova with two 400 m laps. ** p < 0.01; *** p < 0.001. In the 1500 m event (Figure 2), the first lap was covered at a significantly faster speed than the second lap (p < 0.001, ES = 0.91 and 0.78 for men and women, respectively). The second lap was covered at a significantly slower speed than the last 300 m lap (p < 0.001 and p = 0.05, ES = 1.02 and 0.65 for men and women, respectively). The third lap was cov- ered at a significantly slower speed than the last 300 m split ( < 0.001, ES = 0.86 and 0.48 for men and women, respectively). There were no differences between %RS at the first lap and the last 300 m lap for men or women. Figure 2. Mean and standard deviation of (A) average race speed of three 400 m laps and last 300 m during men and women’s 1500 m world record performances (A), and (B) world records set by El Guerrouj and Dibaba. * p < 0.05; *** p < 0.001. Figure 2. Mean and standard deviation of (A) average race speed of three 400 m laps and last 300 m during men and women's 1500 m world record performances (A), and (B) world records set by El Guerrouj and Dibaba. *p< 0.05; ***p< 0.001. In the mile event, the rst 409 m lap was covered at a signi cantly faster speed than the second lap

This research analyzes pacing strategies in middle-distance running world records.