Average stride length and stride rate of Thoroughbreds and Quarter Horses during racing

Abstract

The main factors influencing speed in athletes are stride length (SL) and stride rate (SR). However, conflict remains whether SL or SR is the key determinant of higher speeds. Quarter Horses (QH) generally reach higher speeds in their races than do Thoroughbreds (TB). However, the influence of SL and SR on this greater speed is unclear. Therefore, the main objective of this study was to compare SL and SR in QH and TB raced in short (sprint) and long (classic) distances. We hypothesized that QH have a higher SR in comparison to TB, and SR decreases as distance increases. Two race distances were analyzed for each breed: QH races of 100.6 and 402.3 m, and TB races of 1,207.0 and 2,011.7 m. Data from 20 horses were obtained, consisting of five horses from each race distance (10 QH and 10 TB). Five individuals watched recordings of each race three times counting the number of strides taken by each winning horse. The SR was calculated using the average number of strides over a given race duration, and SL was determined by calculating the total number of strides over the distance covered. Speed was calculated by dividing the distance by the time of the winning horse. The PROC Mixed Procedure was used to identify statistical differences between breeds, and between distances within the same breed. Results showed that although the SL of the TB was longer in comparison with the QH (P < 0.001), the average SR in QH was higher than in TB (2.88 vs. 2.34 + 0.03 strides/s; P < 0.001). Furthermore, QH classic distance demonstrated a faster speed than TB at either distance (P < 0.001). In conclusion, QH achieve a higher SR in comparison to TB (between 14% and 20% more than TB), confirming the importance of SR in achieving high racing speeds.

INTRODUCTION

Whether it is a human sprinter, a marathoner, or a racehorse, speed is critical to performance. Many sports enthusiasts think that a long stride length (SL) is the most important determinant of speed (Magness, 2011; Stokes, 2013; Henry, 2021). However, velocity is dictated by both SL and stride rate (SR) (Clayton, 2016). Daniels (2014) identified that elite human athletes maximize their performance by reaching an average step rate of 180 steps/min (90 strides/min) or more in racing distances of 800 m or more. Although SL and SR often have an inverse relationship (Hunter et al., 2004), human subjects reach maximum speeds and performance using different strategies—some being more reliant on increasing SR while others favor increasing SL (Salo et al., 2011). In horses, it is currently unknown whether SL or SR is most influential in racing speed. Determining the importance of SL or SR on racing speed could have implications on horse selection.

Similarly, horse enthusiasts often believe that SL is the most important factor influencing speed (Henry, 2021). For instance, the outstanding racehorses Man O’ War, Secretariat, and Justify have SL reported to be 8.5, 7.6, and 7.5 m, respectively (von Hippel, 2019). These stride lengths are often quoted, leading to the potential misconception that speed in horses is most strongly influenced by SL. Thoroughbreds (TB) are the world’s most widely distributed racehorse and are well-known for their speed (Bower et al., 2012). Triple crown winners have achieved peak speeds of 61.2 km/h (Denny, 2008). However, this is not the fastest horse breed. Quarter Horses (QH), developed in the United States, are the fastest horse to run a quarter of a mile (403.3 m) (Nielsen, 2014), reaching top speeds of 87.5 to 92.6 km/h (Pratt, 1991; Nielsen, 2014). QH and TB typically race different distances. For instance, QH most often run distances between 91 and 796 m, and TB usually run between 1,006 and 3,219 m (5–16 furlongs) (Hodgson, 2014).

Previous studies investigating the effect of SL and SR on racing speed have conflicting results. An early study showed that SL was responsible for the increase in speed, up to 8.3 m/s (Dušek et al., 1970). However, at higher speeds (11.7 m/s), increasing SR resulted in faster speeds (Dušek et al., 1970). Another study using three TB at six different speeds reported that, as the speed increases, both SL and SR increased nearly linearly. Yet, the fastest TB from the study had the shortest SL and highest SR at maximum speed, indicating that, at higher speeds, SL may decrease as SR increases (Yamanobe et al., 1992). In another study using nine TB, it was observed that the SR increased linearly without signs of a plateau (Witte et al., 2006).

The objective of the current study was two-fold. The first objective was to evaluate the average SL and SR in both QH and TB breeds within their breed-specific races. The second objective was to determine if SL and SR are influenced by the distance raced. Therefore, it was hypothesized that QH have a greater average SR than TB, and that SR will decrease as running distance increases. Finally, it was hypothesized that SL would decrease as the SR increases.

MATERIALS AND METHODS

To examine differences between breeds and between short and long race distances, “sprint” races and “classic” races were analyzed for both QH and TB from 2008 through 2012. The QH sprint race analyzed was the Texas Twister Stakes, 100.6 m (110 yards), held at Sam Houston Race Park (www.shrp.com), and the QH classic race was the Champion of Champions, 402.3 m (440 yards), held at Los Alamitos Race Course (www.losalamitos.com). For the TB races, the Breeders’ Cup Sprint, 1,207.0 m (6 furlongs), and Breeders’ Cup Classic, 2,011.7 m (10 furlongs), were analyzed. The Breeders’ Cup races were held at Santa Anita Park (www.santaanita.com) in 2008, 2009, and 2012, and at Churchill Downs (www.churchilldowns.com) in 2010 and 2011 (Breeders’ Cup. Races, 2020). This resulted in a total of 20 horses being analyzed (5 QH sprint, 5 QH classic, 5 TB sprint, 5 TB classic). These races were chosen as they represented top-tier racehorses racing at distances classified as short and classic distances within their respective breeds, and for which video replays were publicly available for review. Five individuals watched recordings of each race three times (in slow motion when necessary) and counted the strides of the winning horse. If, during periods of a race, the winning horse was not in camera view, the horse in the lead was chosen as a surrogate horse for which to count strides until the winning horse was back in the camera view. All individuals were provided the guidelines as to how to count the strides and were afforded the opportunity to practice before proceeding to record data. From the 15 viewings of each race (five individuals viewed each race three times), the average number of strides for the winning horse was calculated.

For the QH races, the stride count began when the starting gates opened, signifying the official start to the race. For TB races, the stride count began after the horses left the starting gates, had completed the “run-up” (passed the “flagman” or “tripped the beam”), and the race officially started. The average SL was determined by dividing the length of the race by the average number of strides taken during the race. The SR was computed by dividing the number of strides by the race time in seconds. Furthermore, the average speed and standard error of each type of race were calculated using the distance of each race and then divided by time for each winning horse.

Data were analyzed with SAS (version 9.4, Cary, NC, USA). Using the average number of strides of each winning horse, the PROC Mixed Procedure was used to evaluate differences in SR and SL between QH and TB, along with differences between distances within each breed. Besides providing the means and the standard error of the mean (SEM) for the number of strides, the coefficient of variation (CV) was included to demonstrate reliability of the methodology proposed. The SEM was calculated using the standard deviation divided by the square root of the sample size. The CV was determined by dividing the standard deviation and the mean and then multiplying by 100 in order to determine a percentage of variation. Figures were made using GraphPad Prism software (version 9.1.0, San Diego, CA). Differences were considered significant at P ≤ 0.05.

RESULTS

Figure 1 shows the total number of stride counts (15 per race) of the viewers for QH (Figure 1A) and TB races (Figure 1B), respectively. In Table 1 are presented the average values per year and race with their respective SEM and CV. All CV and SEM values were lower than 9.4% and 6.4%, respectively, suggesting reliability in the number of strides counting method proposed in the present study. Also, Table 1 presents the comparison between the average number of strides, SEM, and CV for each breed and type of race. Differences were detected between the different race lengths (P < 0.001).

Figure 1.

Graph bar with scatter of the average and SEM of the number of strides per year of Quarter Horses (A) and Thoroughbreds (B). Each data point corresponds to the number of strides per viewer. In total, 15 views per distance per year (5 viewers, repeating the counting three times).

Table 1.

Average number of strides (Str) rounded to the nearest stride with corresponding SEM and CV values from the 15 total views for Quarter Horses and Thoroughbreds for short (“Sprint”) and long (“Classic”) distances for each breed and year, including final average and overall P value

Year	QH (100.6 m)			QH (402.3 m)			TB (1,207 m)			TB (2,011.7 m)
	Str	SEM	CV	Str	SEM	CV	Str	SEM	CV	Str	SEM	CV
2008	21	0.4	7.2%	62	1.1	6.6%	171	2.8	6.3%	272	6.4	9.1%
2009	21	0.4	6.7%	61	0.7	4.1%	173	3.0	6.6%	277	5.0	7.0%
2010	20	0.2	3.9%	59	1.3	8.7%	164	3.1	7.2%	271	3.8	5.5%
2011	19	0.3	5.6%	57	1.3	9.0%	160	2.4	5.8%	266	6.1	8.9%
2012	20	0.4	7.3%	59	0.5	3.5%	168	4.1	9.4%	271	5.3	7.6%
Average	20a	0.4	4.1%	60b	0.9	3.3%	167c	2.4	3.1%	271d	1.7	1.4%
P-value	< 0.001

Values lacking common superscripts differ (P < 0.001).

SEM corresponds to the standard error of the mean.

CV corresponds to the coefficient of variation expressed in percentage.

Overall, QH averaged a half of stride more per second than did TB (2.88 vs. 2.34 ± 0.03 strides/s; P < 0.001; Table 2). Furthermore, SR decreased as race distance increased, regardless of the breed (2.96 [100.6 m] vs. 2.81 [402.5 m] vs. 2.45 [1,207 m] vs. 2.23 [2,011.7 m] ± 0.04 strides/s; P < 0.001; Table 2). Average values for SL are presented in Table 3. A greater SL was observed in TB compared to QH at all distances (5.9 vs. 7.4 ± 0.11 m P < 0.001). Furthermore, QH had the shortest SL in the 100.6-m race (5 m), with it increasing in the 402.3-m race (6.8 m). TB stride length did not differ significantly between distances (7.3 vs. 7.5 m; P = 0.49).

Table 2.

Average stride rate (strides/sec) for Quarter Horses and Thoroughbreds for short (“Sprint”) and long (“Classic”) distances for each breed

	Breed				SEM	P-value
	Quarter Horse (n = 10)		Thoroughbred (n = 10)
Stride rate	2.88a		2.34b		0.03	< 0.001
	Race length (meters)
	100.6 (n = 5)	402.3 (n = 5)	1,207.0 (n = 5)	2,011.7 (n = 5)
Stride rate	2.96a	2.81b	2.45c	2.23d	0.04	< 0.001

Values lacking common superscripts differ (P < 0.001).

SEM corresponds to the standard error of the mean.

Table 3.

Average stride length (meters) for Quarter Horses and Thoroughbreds for short (“Sprint”) and long (“Classic”) distances for each breed

	Breed				SEM	P-value
	Quarter Horse (n = 10)		Thoroughbred (n = 10)
Stride length	5.9a		7.4b		0.11	< 0.001
	Race length (meters)
	100.6 (n = 5)	402.3 (n = 5)	1,207.0 (n = 5)	2,011.7 (n = 5)
Stride length	5.0a	6.8b	7.3c	7.5c	0.04	< 0.001

Values lacking common superscripts differ (P < 0.001).

SEM corresponds to the standard error of the mean.

The average speeds for each breed and each race are presented in Table 4. Notably, when averaging the overall speed for both races per breed, the speed value for QH and TB was not different (16.9 vs. 17.1 ± 0.03 m/s; P = 0.75). However, the probable origin in this lack of differences in speed between breeds is how the official races are recorded, where QH races involve a standing start for QH and a running start for TB (Nielsen, 2014). This difference in the way QH races are timed and how TB races are timed thus results in the overall average speed of QH including when they are standing still as opposed to TB when the average speed only includes when the horses are running. Regardless, in opposition to the average speed per breed, the average speed for each distance varied within breeds. QH racing the classic distances had the fastest average speed (402 m:19 m/s; P < 0.001). On the other hand, TB showed a significantly faster speed in sprint (1,207 m: 17.7 m/s) versus classic distances (2,011.7 m: 16.5 m/s; P < 0.001).

Table 4.

Average speed (m/s), time and SEM for Quarter Horses (QH) and Thoroughbreds (TB) calculated from when the race officially started for each breed; from a standing start when the starting gates open for QH and from a running start when TB completed the “run-up”

	Breed				SEM	P-value
	Quarter Horse (n = 10)		Thoroughbred (n = 10)
Speed	16.9		17.1		0.03	0.75
	Race length (meters)
	100.6 (n = 5)	402.3 (n = 5)	1,207.0 (n = 5)	2,011.7 (n = 5)
Speed	14.8a	19.0b	17.7c	16.5d	0.04	< 0.001
Time	6.8a	21.2b	68.9c	121.7d	0.20	< 0.001

Values lacking common superscripts differ (P < 0.001).

SEM corresponds to the standard error of the mean.

DISCUSSION

Results from the study support the hypothesis that the average SR was greatest in the racing QH and SR decreases as race distance increases (Table 2). In terms of SL, the study data suggest QH have a shorter SL than TB. In addition, as hypothesized, the shortest races with the highest SR have the shortest average SL. Similarly, as also hypothesized, as the distance of the race increased, the SL also increased, but only up to the point at which the TB race distances were reached as no differences were seen between TB racing at the sprint or classic distances (Table 3). It is important to note that many strides were taken at the beginning of QH races as the horses accelerated. As higher speeds were reached, the stride frequency tended to decrease as SL increased as has previously been described by Nielsen et al. (2006). In that study, the speed was lower in the beginning of QH races than in the middle and final parts of QH races—contributing to the explanation as to why the average speed of QH was slower at shorter distances (100.6 m) than at the longer distances (402.3 m). This phenomenon is also observed in human athletes where it takes around 10 m to reach a SR plateau and about 30 m of running distance to reach a SL plateau (Nagahara et al., 2014).

Although it is recognized that QH are faster than TB (Pratt, 1991; Nielsen et al., 2006), the results obtained from the calculations used in this study to determine average speed (distance divided by time) appear to lack support for this—providing temptation to conclude the QH and TB race at similar speeds. Likewise, if one compares the world speed records at the same distance (402 m), they appear somewhat similar. The current QH record is held by ‘First Moonflash’, with 20.27 s (Equibase, 2020a), and the TB record is held by ‘Winning Brew’, with 20.57 s (Equibase, 2020b). In reality, while accurate, calculating the speed using the time of the race and distance fails to take into consideration that QH races are timed from when the starting gates begin to open and the horse is standing still while TB races are timed after horses have already started running (Nielsen, 2014) and have traversed the “run-up” which is highly variable (Thoroughbred Idea Foundation, 2020). Although TB tend to be relatively constant in their speed throughout a race and the peak speed reached often is somewhat similar to their average speed, this is not true for QH (Nielsen et al., 2006). The average speed for a QH takes into account the period in which they are standing still and have not yet begun running. Pratt (1991) has estimated it can be 0.6 s before a QH has taken a step away from the starting gates. Using that estimate, roughly 9% of the race time for the QH sprint races is spent on the very first step away from the starting gate. Although constituting only 3% of the race time for the QH classic races, it still represents a period during the race when the speed of the horse is at or near 0 m/s. Even after that, QH are accelerating during the initial portion of the race, whereas TB had that period of acceleration during the “run-up” before the official timing of the race began. Although QH racing at the classic distance demonstrated the fastest speed, higher peak speeds could have been achieved during the sprint races. However, peak speed measurements were not within the scope of this study.

Although a portion of the general public tends to believe that a long stride is correlated strongly with greater speed (Magness, 2011; Stokes, 2013; Henry, 2021), this study indicates that performance is dependent on both SL and SR. That stated, a difference in stride rate within breeds at different distances is noteworthy. At the classic distances for each breed, the average QH speed over the entire race was 2.5 m/s faster than the TB speed (even with the QH timed from a standing start). With QH racing at higher speeds than TB, it is clear that the SR plays a greater role in reaching peak speeds in short distances. These findings go in agreement with the study by Hay (2002) that stated, at maximum speeds, the SL remained constant or decreased slightly, contrary to the SR that was increased with the increase in speed. In fact, the fastest horse analyzed showed the highest SR but the lowest SL, showing the important role of both factors and not only SL. Moreover, the findings of another study showed that both variables (SR and SL) increased linearly, although SL showed a tendency to decrease (Yamanobe et al., 1992). Of note, with the average SR being 2.96 ± 0.04 strides/s during the sprint QH races, some of the horses had over 3 strides per second—a truly amazing physical feat.

In the case of human athletes, they reach maximum speeds and performance using different strategies. Elite human sprinters have been shown to possess individual preference, whether it is a higher SR or SL (Salo et al., 2011). Mixed results are found in the study of Hunter et al. (2004). They concluded that SR may be a significant factor influencing speed in short distances. They examined 28 human athletes performing repeated sprint trials. Results indicate the best individual results were obtained when they had a higher SR. However, the SL is significantly correlated to speed when this factor was analyzed as a group, and not the SR. Also, a negative correlation was found between SR and SL (r = −0.70). Acknowledging that there are distance differences between and within the breeds, the differences in SR and SL average between breeds can be an adjustment in racing strategy according to race distance. In other words, if trained and raced like a QH, a TB running for distances of 402 m or less may have a similar recorded time as a QH racing the same distance. For example, QH are often not ridden every day, usually being galloped only a few days per week at most (Nielsen, 2014). This is in contrast with the TB training in which it is common practice to gallop the horses on most days and for longer distances during each ride, developing more endurance ability required for longer races in comparison with the typical distances in QH (Hodgson, 2014). Interestingly, in the study of Ferrari et al. (2009), the SR was increased after 6 mo of training mature TB.

It is reported that more than 75% of fiber muscles are type-II in TB (Kawai et al., 2009). However, QH have a greater proportion of muscle fibers type-IIx when compared with TB which may explain their increased speed (Stull and Albert, 1980; Nielsen, 2014; Valberg, 2014). Moreover, having this larger IIx fiber proportion in QH could provide a higher muscle glycolytic power and a higher maximum speed when raced short distances. Although differences in muscle fiber composition are likely to be heavily influenced by genetics, differences in training techniques may also have an effect. It is well accepted that type I and type IIa fibers have a smaller size and larger oxidative capacity in comparison with a larger glycolytic and cross-sectional area of IIx (van Wessel et al., 2010). For this reason, when endurance training is incorporated, it will produce a gradual decrease in muscle fiber size. For instance, trained Standardbreds showed smaller gluteal fiber cross-sectional areas than untrained horses (Essen-Gustavsson and Lindholm, 1985). Similarly, Yamano et al. (2002) compared changes in fiber types of trained versus untrained 2-yr-old TB. They found that there is an increase proportionally in fiber type-IIa/IIx. These changes—an increased proportion of fiber type-IIa and smaller fiber size—may be an advantage for a more efficient diffusion of oxygen through the muscle fiber and faster waste product removal, such as CO₂ and lactic acid, during endurance exercise (Rivero et al., 2007; Valberg, 2014). For this reason, it is not surprising that QH present an apparent larger muscle mass in comparison to TB (Nielsen, 2014), due to the inherent differences in size of these muscle fibers (van Wessel et al., 2010).

The current study had some limitations. In the QH races, the distance horses ran was likely very similar to the official distance of the race as QH typically run in a straight line. In contrast, the TB races were run on an oval. Often winning horses traveled greater distances than the official race distance unless they happened to be on the rail for the entire race, thus decreasing the precision in the calculation of the average SL. In truth, the average TB SL is likely slightly greater than what is reported in this study due to that variation. This does not negate the breed differences in SL, but it is acknowledged that the difference between breeds could be even greater.

Another limitation was the difficulty of counting the short steps of QH at the beginning of the races. As mentioned previously, QH races begin as the starting gates open (Nielsen et al., 2006) and acceleration is dramatic. There is a short period in a QH race when the horse is standing still and has not left the gate. During that rapid acceleration, QH take several short strides (Pratt, 1991). However, by having 15 views of each race, the challenge associated with counting those first few strides should have been ameliorated. By contrast, TB races begin when the first horse crosses in front of the flagman or electronic beam—a short distance in front of the starting gates (Nielsen, 2014). Although the TB races officially start while the horses are already running and, hence, are taking full strides (an advantage to being able to count strides), there is some degree of uncertainty in terms of being to determine exactly when the horse crossed that line. With repeated viewing of each race by five individuals, this uncertainty was minimized and a meter or so difference in starting point would have only a minor impact on the number of strides taken over the longer distance of the TB races. This is reflected in Table 1, where reduced CV values (≤ 9.4%) are found for the number of strides for both breeds. This difference in the start and how it is timed (running versus standing start) makes comparing SR and SL during different segments of the race challenging, especially at the start of the race. It was determined that only average differences in SR and SL over the entire race could be performed accurately, as opposed to, for instance, comparing SR and SL in the last 100 m of the race (which would represent the entire race for the QH sprint races). As a result of these differences in the type of racing, comparing TB and QH over the same distance could lead to inaccurate conclusions. Therefore, it is also acknowledged that differences between breeds are confounded with distance.

Besides the novelty of reporting the amazingly high SR seen in racing QH (especially at the shorter distance), this study illustrates other considerations as it relates to other physiological systems within the racing QH. First, the study raises some potential questions regarding how an increasing SR may affect the respiratory system. A locomotor-respiratory coupling system has been described in horses cantering and galloping (Franklin et al., 2012; Lekeux et al., 2014). If this coupling system remains true at high speeds, the average respiratory rate may reach 134 to 147 breaths/min in the TB and 169 to 178 breaths/min in the QH with some QH individuals in the sprint races likely taking over 3 breaths/s. This respiratory rate is between 14% and 20% higher in the QH than the 148 breaths/min previously reported for TB racehorses (Hörnicke et al., 1987; Lekeux et al., 2014). Although clarifying the deeper mechanisms and effects on respiratory system are beyond the scope of this study, further studies are needed to determine the potential impact on this system due to the high SR, especially for QH.

Beyond the respiratory system, the average SR findings also have possible implications for the dynamics of the equine lower limb and hoof. During each stride, the hoof momentarily comes to a halt during the stance phase of the stride (other than minor rotations or sliding of the hoof on the ground). For racing QH taking three strides per second, this suggests that three times during each second the hoof experiences rapid deceleration as the hoof comes to a stop and then experiences rapid acceleration as the hoof leaves the ground during the swing phase (Setterbo et al. 2009). For horses previously reported to reach speeds of around 89 km/h (Pratt, 1991; Nielsen et al., 2006), if the hoof is temporarily not moving during part of the stride, the hoof would have to accelerate at a very high speed (especially when one recognizes the flight pattern of the hoof is curved in the vertical plane as it leaves the ground and then descends onto the ground again, thus traveling even further than the body of the horse). Thus, the hooves of racing QH may reach double the speed of the horse or greater at some point in the swing phase. While not the point of this project, with enhanced technology being developed, it would be interesting to determine peak speeds the hooves of racing QH achieve and determine the ground reaction forces associated with such rapid acceleration and deceleration, as being an inherent part of locomotion of horses (Clayton and Hobbs, 2019).

In conclusion, despite some limitations in methodology, differences between breeds and within breeds support that a higher average SR contributes to the higher speeds previously reported for QH. Therefore, the analysis of an equine athlete must consider both SR and SL as determinants of potential performance in speed competitions. Future work could explore how the increased respiration rates affect the integrity of the respiratory system in animals with high SR, especially in the short QH races.

Conflict of interest statement

None declared.