The Impact of Nike Vaporflys on Elite Marathon Performance

Jamie Langley (17.12.2019)

Since the introduction of the Nike Vaporfly 4% at the USA Olympic Marathon Trials in February 2016 controversy over the magnitude of performance benefits associated with the shoe has followed. Kara Goucher missed out on a place on the Olympic team to Shalane Flanagan an athlete wearing a prototype of Nikes Vaporfly 4% by 65s (1), thus the question was born what difference do Nikes Vaporflys make to performance?

Nikes Vaporfly differs from the conventional racing flat in three ways: (1) highly compliant and resilient midsole material, (2) an embedded stiff curved carbon fibre plater, and (3) a thicker midsole thickness (31mm) compared to the 23mm heel height of both the Nike Zoom Steak 6 and Adidas Boost (2,3). This trend in the evolution of running trainers looks set to continue with shoe manufactures including; Brooks, Saucony, New Balance, Hoka One One and Asics all releasing new models imbedding a carbon fibre plate within 2019 or soon to be released in 2020 (4)

Since the USA Olympic Marathon Trials Nikes Vaporfly shoes have been at the forefront of endurance running performance with all three of the male medal winners from the 2016 Rio Olympic Marathon wearing Vaporflys. Furthermore, since 2016 male athletes wearing a version of Nikes Vaporflys have broken world records over the 10k, 15k, half marathon and marathon distances (5). On the female side Brigid Kosgei ran an astonishing world record of 2:14:04 for the marathon in a pair of Nikes Next %. Additionally, the improvement in endurance running performances is not an isolated exceptional few athletes distorting statistics, in the period from 1st January 2017 – 13th December 2019, of the top 100 all-time list 45 of the fastest female and 38 of the fastest male marathons have been ran, in contrast between the years 2015-2010 the mean number of performances per year for female in the top 100 = 5 ± 4 and males 8 ± 4 (6). Moreover, the 2019 Valencia Marathon seen an unprecedented number of females break 2:19:00 (4), 2:20:00 (5), 2:21:00 (7) and 2:23 (9) in a single race of which all 9 athletes are sponsored by Nike and wearing a version of either the Vaporfly or next % shoes (7). Interestingly Adidas sponsored athlete Herpasa Negasa felt the performance benefits were great enough to risk his sponsorship by attempting to disguise he was wearing a pair of Nike Vaporflys by painting them and wearing them in the Dubai marathon where he finished 2nd in 2:03:40 a PB by 5 minutes and 34 seconds.

Testing of the Nike Vaporflys in contrast to other leading racing flats (Nikes Zoom Streak and the Adidas Boosts) revealed Nikes Vaporflys to be on average 4% more economical (8). These claims have been substantiated by Barnes & Kilding (2018) in independent laboratory testing (9). Furthermore, a study by Kipp et al (2019) aimed to evaluate the improvements in running economy (RE) whilst wearing the Nike Vaporflys translates into increased running speed (3). Kipp et al (2019) report that the improvements in RE to running velocity is non-linear with running velocity > ~3m/s the predicted percentage of improvement in velocity is less than the degree of percentage improvements in RE. At the previous marathon world record pace of 5.72m/s, a 4% improvement in RE translates to an increase in running velocity of 2.64% which would reduce the marathon world record to 1:59:47, however it is worth noting that since the introduction of the Vaporfly the marathon world record has only fallen by 1.03% (3). Since the publication of improvements in RE of 4% Nike have subsequently released the Nike Next % shoes and a prototype Nike Alpha Fly which Eliud Kipchoge wore in his unofficial marathon time of 1:59:40, both shoes are believed to have a further improvement on RE and subsequently running velocity. This further propagates the question to what magnitude do the Vaporflys increase running velocity within elite athletes when racing?

Aim:

This study aims to investigate if there is a difference in the top 20 elite male and female marathon times and if so by how much from performances pre-Vaporfly (before February 2016) to December 13th 2019. The study looks to evaluate the differences from not only the top 20 world performances but also evaluate the individual variability of athletes ranked within the world’s fastest top 20 times of 2019 who have ran a marathon prior to February 2016.

Method:

Data regarding marathon times was collated from the World Athletics Organisation website and filtered for males, females and dates for individual year performances and all time bests: (https://www.worldathletics.org/records/toplists/road-running/marathon/outdoor/men/senior/2019; https://www.worldathletics.org/records/all-time-toplists/road-running/marathon/outdoor/women/senior?regionType=world&drop=regular&fiftyPercentRule=regular&page=1&bestResultsOnly=false&firstDay=2019-01-01&lastDay=2019-12-19). Data was collated and analysed regarding the athletes marathon time, gender, DOB, country of allegiance, venue, and number of marathons ran and the shoes ran in. To establish what trainers the athlete ran in video footage of races and race images. The top 20 world best times for both males and females from 2019 and 2015 were recorded to analyse the difference in times. The athletes best time ran in 2019 was recorded for analysis along with the athlete’s best time prior to February 2016 for individual analysis. Statistical analysis was conducted comparing variations using both a T-Test and Cohen’s D to calculate any differences between the groups and effect size.  

Results:

Table 1. Comparison of the males’ world’s fastest times for the marathon from 2019 up to 13th of December vs. 2015.

Table 1 signifies there is a statistical significant difference (P < 0.001) between the top 20 male athletes of 2019 (7441 ± 59s) to 2015 (7558 ± 40s) the mean difference in times is -1.55 ± 0.31% translating to an average 117 ± 23s faster in 2019 (Table 1). It is also worth noting in 2019 the top 12 athletes and 16 of the top 20 performances were by athletes wearing a version of the Nike Vaporfly or Next%.  

Table 2. Male individual performance variation of athletes ranked in the top 20 in 2019 who had ran a marathon prior to February 2016.

Table 2 demonstrates the mean improvement for the 14 athletes who had completed a marathon prior to February 2016 -4.02 ± 2.88% a significant reduction of the mean marathon time of -319 ± 243s. This is a significantly greater improvement in mean time in contrast to the -1.55 ± 0.31%, 117 ± 23s when comparing the top 20 fastest times from 2015-2019 (Table 1). Excluding the male athletes not sponsored by Nike (11 total) reported a larger effects size D= -1.52 comparing the differences in performance between their best times prior to 2015 compared with their SB in 2019.

However, when excluding the athletes sponsored by Adidas (3) the magnitude of performance improvements is not any greater for the Nike athletes (14) -4.01 ± 2.98%, -318 ± 253s, this may be due to the majority of the athletes being sponsored by Nike. Furthermore, the SD 2.88% or 243s demonstrates a significant individual variability which is likely to be seen to other factors than just the shoes. The mean (4 ± 3) number of marathons completed prior to February 2016 in comparison to 2019 (11 ± 3) maybe a significant contributing factor to the development of the athlete. Interestingly, the mean fastest marathon time is the same as the number completed by 2019 (11 ± 3), furthermore athletes have been achieving PB’s after their 12th, 14th, 15th and 17th marathons. The mean age for the quickest marathon recorded is 29±4 years, but athletes are achieving PB’s at 37 & 35 years of age respectably (Table 2), this is in line with Hunter et al (2011) (9) study reporting elite male marathoners mean age 28.9 ± 3.4 years and females 29.8 ± 4.2 years. Lehto (2016) reported amateur runners improve up to the age of 34.3 ± 2.6 years (10).

Table 3. Male athletes ranked in the top 20 in 2019 who had completed a marathon prior to February 2016 marathon progression, excluding non-eligible for records, *= Olympics of WC. The athletes’ fastest marathon highlighted in red.

Reviewing the athletes marathon progression 10/14 males have ran their fastest time on their last attempt and 3/4 on their penultimate marathon to date. Are the improvements in marathon times on the athlete’s latter attempts due to the continual progression due to the improvement in shoe technology or as a consequence of improving their craft over the distance? Athlete’s times have gradually improved since 2016, however there is not a sudden reduction in running times post 2016 since the introduction of the Vaporflys as one might suspect.

Figure 1. Marathon progression of the top 5 ranked male athletes in 2019 who had completed a marathon prior to 2016 who are sponsored by Nike.

Figure 1 illustrates the progression of the top 5 ranked marathoners from 2019 from their first marathon until December 2019 who had completed a marathon prior to 2016. From reviewing the individual data of marathon progression there is a trend in improvements that correlate with the release of the Vaporflys. Bekele’s first 3 marathons up to April 2016 range from 7504-7596s however after the introduction of the Vaporfly Bekele’s next marathon in September 2016 is marathon time improved to 7383s improving by 121s. From 2013-2015 Eliud Kipchoge’s fastest marathons range from 2:04:11-2:04:00 demonstrating a plateau in performance, in 2016 Kipchoge’s marathon PB decreased to 2:03:05 in 2016 and further to the now WR time of 2:01:39 in 2018. In 2015 Mule Wasihun ran a time of 2:10:57, this time rapidly progressed by in excess of 5 minutes to 2:05:44 in 2016 and has gradually declined each year (2017: 2:05:39, 2018: 2:04:37, 2019: 2:03:16), however, it is worth noting 2015 was his first marathon. Lemma’s marathon times steadily progress to 2016 to a time of 2:05:16, but plateaus over his next 4 marathons until 2018 where he runs 2:04:08. Herpasa Negasa’s marathon times from May 2012- January 2018 range from 7817s-8055s, after this point his marathon time decreases from 8000s-7420s by January 2019. All of these athletes have improved their PB since 2016 despite periods prior where there PB’s have remained constant.

Figure 2. The fastest male marathon time (s) from 2009-2019 & the mean time of the top 20 athletes.

Figure 2 illustrates the progression of the best time for the marathon and the average time for the top 20 athletes over the last 10 years from 2009-2019. The marathon progression of the fastest marathon time has decreased from 2016-2018 and remained constant from 2018-19 (Table 2). Prior to 2016 the fastest marathon time has gradually decline from 7467s (2009) to 7377s (2014) a reduction in 90s in 5 years, from 2014 (7377s)-2018 (7229s) the fastest time has decreased by 148s in 4 years and remained this low in 2019, showing a decline of 58s more than the previous 5 year progression.  Since the introduction of Nikes Vaporflys in 2016 the fastest marathon time progressed for 2 years in a row and remains low in 2019 just 2s slower than 2018.

The mean time has remained relatively consistent for 8 years between 2009–16 ranging from 7545s (2009) to the lowest 7502 (2012) a difference of just 43s. However, over the last 2 years the mean time has declined by 79s from 7520s (2017) to 7441s (2019), this demonstrates a rapid rate of progression with the mean times declining by -0.04% from 2009-2016, in contrast from 2017-19 the mean times have declines by -0.53%. Figure 2 supports the evidence that the rate of marathon progression has coincided with the introduction of Nike Vaporfly shoes and shows an abnormal acceleration in the rate of progression within the last 2 years demonstrating a leap in performance.

Figure 3. The mean male marathon times (s) of the top 20 performances from 1980-2019.

Figure 3 illustrates how the mean male marathon times (s) has progressed over the last 30 years with performances improving from 7807s in 1990 to 7441s in 2019, throughout this period there have been decreases and plateaus in the mean performance. From 1990-2000 the mean marathon times have decreased by an average of -0.18%, however from the years 2000-05 we see a plateau in performance with the mean increase in performance difference increasing by 0.01% each year, this may coincide with the formation of WADA in 1999 and introduction of out of competition tests in 2004 (12, 13). From 2010-2019 all except 1 of the athletes who is ranked within the top 20 times for that year are of East African descent which may be a contributing factor to the improvements in mean times, 2003 sees the first time that ≥15 of the athletes who comprise the fastest times are of East African descent(5). However, from 2010-17 a plateau in the mean performance is observed with the mean variation year to year fluctuating by -0.07%. Whereas, from 2017-19 the mean marathon time has decreased on average -0.53% each year, this reduction in marathon times coincides with the Vaporfly and due to the majority of these athletes wearing a version of the Vaporfly this suggests the is potential, at least in part, due to the shoe.  

Figure 4. The mean male marathon times (s) of the top 20 performances from 2010-2019.

Figure 4 highlights between 2010- 2017 the mean male marathon times of the top 20 performers are within 0.5% either way of the mean. However, the mean marathon time increases substantially in the years 2018 and 2019 demonstrating the progression of the marathon has excelled in recent years beyond the norm of rate of progression, this indicates a change to the norm has occurred, this improvement in performance correlates with the introduction and use of Nike Vaporfly shoes.

Table 4. Comparison of the female’s world’s fastest times for the marathon from 2019 up to 13th of December vs. 2015.

Table 4 highlights there is a clear statistical significant difference (P = <0.001) in mean times of the top 20 female athletes between 2019 (8358 ± 98s) in comparison to 2015 (8535 ± 87s) is a difference of -2.08 ± 0.5% this equates to a mean reduction of marathon times by 177 ± 42s nearly 3 minutes faster on average.

Conclusion:

From reviewing the statistics it is clear there has been a significant improvement in marathon times ran across both males (Cohen’s D = -1.52) and females’ fastest times and the top 20 athletes mean times, these improvements coincide with the introduction of the Nike Vaporfly shoes post 2016 and significant improvement over the last two years. The question of to what magnitude do Nikes Vaporflys translate to in increased running velocity? A conservative estate in the improvement in running velocity from reviewing the men’s mean top 20 performance across the last 10 years would indicate an improvement in running speed of 0.53% equating to 79s over the marathon, however not all of these runners are sponsored by Nike. On the contrary, I believe the analysis of the individual improvements from prior to February 2016-2019 of -4.02 ± 2.88% (-319 ± 243s) overestimates the performance benefits and does not consider other factors such as training variability and the development of learned marathon craft, this can be seen in the mean number of marathons ran prior to 2016 4 ± 3 to 11 ± 3 in 2019. I believe a more realistic answer lies when reviewing the progression of the top 20 male and female performances in 2019 in comparison to 2015 indicating that running velocity is increased by -1.55 ± 0.31% (117 ± 23s) to -2.08 ± 0.5% this equates to a mean reduction of marathon times by 177 ± 42s nearly 3 minutes faster on average. The difference of performance benefit between males and females may shows female athletes gain a disproportionate advantage over their male counterparts which may be due to the mean female marathon time being greater than males. Rational behind this would indicate the greater the duration of the event the more important economy of locomotion becomes to determining performance outcomes. A crude calculation of the mean difference between the males and females times would indicate that Nike Vaporflys are improving performance by ~135s or 2 minutes 15s thus a 2:04:00 marathon pre Vaporfly = 2:01:45 and a 2:18:00 = 2:15:45. But a more likely outcome is there is intersex differences with females seeing a larger benefit from the Nike Vaporflys, based off the differences between Nike athletes signifies Nike Vaporflys are improving performance by ~93s for males, thus a 2:04:00 marathon pre Vaporfly = 2:02:27 and females see a larger benefit of ~177s, therefore a 2:18:00 = 2:15:03.

Reference List:

1. Flotrack (2016) ‘2016 U.S. Olympic Team Marathon Trials’: https://www.flotrack.org/results/5997269-2016-us-olympic-team-marathon-trials/3572 [13.12.19].

2. Burns, G.T. & Tam, N. (2019). Is it the shoes? A simple proposal for regulating footwear in road running, British Journal Sports Medicine. 0(0) 1-2.

3. Kipp, S., Kram, R., Hoogjamer, W. (2019). Extrapolating metabolic savings in running: Implications or performance predictions. Frontiers in Physiology. 10(79):1-8.

4. Kelly, M. (2019). ‘2019 shoe prototype round-up: A closer look at shoes that aren’t on the market yet, but will be soon (hint: they all have carbon plates)’: https://runningmagazine.ca/sections/gear/the-2019-shoe-prototype-round-up/ [25.10.19].

5. World Athletics. (2019). Marathon Men: https://www.worldathletics.org/records/toplists/road-running/marathon/outdoor/men/senior/2019?regionType=world&drop=regular&fiftyPercentRule=regular&page=1&bestResultsOnly=true [13.12.19].

6. World Athletics. (2019). All-time top lists marathon: https://www.worldathletics.org/records/all-time-toplists/road-running/marathon/outdoor/women/senior?regionType=world&drop=regular&fiftyPercentRule=regular&page=1&bestResultsOnly=false&firstDay=1899-12-31&lastDay=2019-10-13 [13.12.19].

7. Track & Field News. (2019). Best results by placings: https://trackandfieldnews.com/discussion/search.php?s=a2bc27f759be5a39d0fed919e0f96d63&do=process [1.12.19].

8. Hoogkamer, W., Kipp, S., Frank, J.H., Farina, E.M., Luo, G., & Kram, R. (2017). A Comparison of the Energetic Cost of Running in Marathon Racing Shoes, Sports Medicine. 48:1009-19.

9. Barnes, K.R. & Kilding, A.E. (2018). A randomised crossover study investigating the running economy of highly-trained male and female distance runners in marathon racing shoes verses track spikes. Sports Medicine. 1-12.   

10. Hunter, S.K., Stevens, A.A., Magennis, K., Skelton, K.W., & Fauth, M. (2011). Is there a sex difference in age of elite marathon runners? Medicine & Science in Sport & Exercise. 43(4):656-664.  

11. Lehto, N. (2016). Effects of age on marathon finishing time among male amateur runners in Stockholm Marathon 1979-2014. Journal of Sport & Health Science. 5:349-354.

12. World Anti-Doping Agency. (2019). Who we are: https://www.wada-ama.org/en/who-we-are [19.12.19].

13. World Anti-Doping Agency. (2018). WADA welcomes ECHR decision to back Whereabouts rules: https://www.wada-ama.org/en/media/news/2018-01/wada-welcomes-echr-decision-to-back-whereabouts-rules [19.12.19].

Follow My Blog

Get new content delivered directly to your inbox.