Relationship between experience and head kinematics in race riding jockeys

Abstract

Thoroughbred race-riding requires jockeys to attenuate horse movement and maintain postural stability whilst galloping at high speeds. This study aimed to investigate the head movement of jockeys in relation to race-riding experience. Accelerometer and heart rate data were collected from twelve apprentice and two senior jockeys during 85 exercise rides and 82 trial rides. Mean head displacements were determined for each jockey by double integrating the filtered acceleration data. A mixed effect multivariable linear regression model was used to investigate the relationship between jockey experience, physiological variables and head kinematics. The median (IQR) head displacement was higher for exercise riding (0.12 m, 0.09–0.14 m) than trial riding (0.06 m, 0.05–0.09 m). Jockey head displacement decreased with increasing speed of the horse (p < 0.001) and greater jockey experience (p = 0.007). Higher exercise load had a greater effect on head displacement with less experienced jockey’s (p = 0.02). The effect of speed was lower for trial riding than exercise riding (p < 0.001). More experienced jockeys had a greater ability to attenuate horse oscillation than inexperienced jockeys. This ability became more pronounced at higher exercise loads, reflecting a higher level of physical fitness and riding skill level. Measurement of jockey head displacement may provide a simple measure of assessing jockey race-riding ability or fitness.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-98683-9.

Introduction

The Thoroughbred racing industry worldwide faces increasing external scrutiny in relation to the health and safety of industry participants, including jockeys. Jockeys are integral to the sport with jockey’s typically piloting one of 10–15 horses in a race with speeds reaching over 60 km/hr¹. Control of both their individual riding performance as well as the performance of the horse they are riding requires physical strength, stamina and the ability to make quick decisions during a race^2,3. Therefore, ensuring jockeys are adequately prepared, and performing to their best potential is integral to maintaining high health and safety standards for the main industry participants.

Most apprentice jockeys in New Zealand exercise horses daily as part of their race preparation, generally at a steady-state canter (8–9 m/s)⁴. They also participate in trial (“mock” races) and races, both performed at race speeds of 14–16 m/s⁵. Race riding requires greater physical activity and fitness than trial or exercise riding^4–6 reflecting the importance of race-specific physiological fitness. However, some aspects of race riding require specific cognitive skills (race tactics, reaction and coordination) in addition to the physiological component of the sport. The combination of both psychological and physiological fitness may be encompassed by the term “race-specific fitness”, used to describe athletes in elite sport^7–9.

Recent studies have established that measures of race- specific fitness are difficult to quantify using conventional fitness tests^10,11. A jockey-specific test designed to test the stability and balance of the jockey’s riding position on a static sawhorse indicated that experienced jockeys were able to maintain the race riding position longer than less experienced jockeys¹⁰, providing a crude measure of jockey skill. In other equestrian disciplines (dressage), comparisons between riders of high and low experience levels have also indicated that on-horse posture and stability provide the most important indicators of rider performance^12–14. These findings indicate that on-horse posture maintenance is an important learnt skill. Therefore, measuring the head movement of riders could therefore provide an indication of their on-horse postural stability.

A galloping horse oscillates in a regular sinusoidal manner, particularly in the vertical plane¹⁵. The jockey absorbs these oscillations in their legs and body to result in little observable movement of their upper body or head^5,16. Jockey head displacement is lower during the higher speeds of race riding (~ 0.05–0.06 m per stride), compared to ~ 0.11 m per stride in exercise riding (Fig. 1)^4,5,17. This effect can be attributed to race riding jockeys attenuating approximately half the magnitude of the horses’ oscillations during a race compared to 1/3 during exercise riding^4,5. Additionally, horses moving at greater speeds, as they would in race riding compared to exercise riding, have lower vertical oscillations^15,18, contributing to the lower head displacement of race riding jockeys. The ability to dissociate their movement from the horse’s movement requires a high level of skill and physical fitness^10,19. This finding is reflected in observations in other competitive sports such as downhill mountain biking, with muscular activation in the legs and torso acting as an active spring, attenuating the oscillations of the bike²⁰.

Fig. 1.

Mean magnitudes of displacement (cm) of the jockey relative to the horse (%) at trackwork (A), trials (B) and races (C) in the vertical (y), medial/lateral (z) and fore/aft (x) planes. Figure adapted from^4,5.

In many competitive sports, reduced displacement of the head is associated with greater dynamic postural control, cognitive ability and reaction speed^21,22. Intervention studies have identified that the primary role of muscle activity in both apprentices and jockeys was to achieve stability of the torso, but more importantly, stability of the head whilst riding (Fig. 1)^4,5. Therefore, measurement of jockey head movement during racing may prove to be indicative of the effectiveness of the jockey (both physical and mental fitness to compete).

Falling from a horse during the high-speed close contact conditions of a race poses not only an injury risk to the jockey, but may be associated with a lack of this race-specific fitness. Incidence rates for jockey falls during flat racing range from 1.2 to 4.4 per 1,000 rides worldwide, highlighting jurisdictional differences in fall risk^23–29. However, across most jurisdictions, experienced jockeys have a lower incidence of race day falls than inexperienced jockeys^23,30, suggesting that jockey skill and race day injury are highly inter-dependent. In addition, those jockeys with longer careers and more rides per race day also have a reduced incidence of race day falls⁷. Falling from a horse during training is also a potential source of injury to jockeys, being the leading cause for injury to both jockeys and exercise riders in Ireland, Australia and the United Kingdom^31–33.

Fatigue has also been suggested as a possible contributor to fall risk in jockeys, with fall risk increasing with the accumulation of rides on a race day^23,34. The potential effect of fatigue has also been observed in race riding jockeys where a decrease in abdominal activity between the first and last race of the day indicated lower synchronous muscle control in the later races⁵. This effect may result in a change in head kinematics in fatiguing jockeys. Synchronous movement between horse and rider requires a coordinated activation pattern of the rider’s core muscles, resulting in less work done by the horse to carry the rider, possibly leading to a competitive advantage in race riding and lower fall risk^14,16,35. Therefore, early identification of a fatiguing jockey may help to reduce fall risk.

At present there is a paucity of data on how jockey experience and fitness levels influence head kinematics and riding efficiency. Existing evidence suggests that the ability to maintain an optimal riding position, which both minimises work done by the horse and provides a measure of jockey race-specific fitness may be represented by a reduction in head displacement. To investigate the potential of utilising head movement as a measure of “race-specific fitness”, the aim of this study was to investigate the head movement of jockeys in relation to race riding experience and physical exertion. It was hypothesized that more experienced jockeys would have a lower head displacement than less experienced jockeys.

Methods

Participants

A power analysis estimated that 14 participants were required to detect a large effect size (Cohen’s d = 0.8) at a 5% significance level with 80% power. Twelve apprentice and two senior jockeys, holding a current and valid licence with New Zealand Thoroughbred Racing (NZTR); the governing body for Thoroughbred racing in New Zealand, were recruited through NZTR. Apprentice jockeys held a current apprentice licence for 0.5–4 years and had all completed the required 25 trial rides before attaining their apprenticeship. Both senior jockeys had ridden professionally for 14 years. Written informed consent was obtained prior to commencement of data collection and only jockeys over the age of 16 years were considered for the study. Ethical approval for this study was provided by the Massey University Human Ethics Committee (SOA 20/27) and Massey University Animal Ethics Committee (MUAEC Protocol 20/48). All methods were performed in accordance with the relevant guidelines and regulations.

Data collection

Data collection and analysis followed the methods described in Legg, et al.⁴. Anthropometric data were assessed prior to (within 1 month) commencement of field data collection. Stature was assessed to the nearest centimetre using a portable stadiometer (Seca 213, Hamburg, Germany). Body mass was measured in minimal light clothing (shorts and shirt) using a portable digital weighing scales (Tanita InnerScan, Body Composition Monitor, BC-532). Body Mass Index (BMI) was calculated as the ratio of the weight to the square of height in metres (kg/m²).

Field data were collected from jockeys riding during exercise riding, jump outs (unofficial race training) and trials (official mock races) in their normal working schedules. Jump outs were grouped with trials and hereafter are labelled as trials, with the gallop portion identified at speeds ≥ 13.9 m/s. Data were collected from apprentice jockeys on two separate occasions, approximately 2 months apart, for each ride type. Only exercise rides where horses worked at a canter (> 6.9 m/s and < 13.9 m/s) were included in the analysis. All horses ridden were in normal race training and entered in the trial according to their trainer’s instruction.

Speed, distance travelled, and heart rate (HR) of jockey were determined using Polar HR monitors via Bluetooth to their respective watches (Polar V800 sports watch, Kempele, Finland) each containing a global positioning system (GPS) unit at a sampling rate of 1 Hz. Linear accelerations and displacements of the jockey were determined via wireless, tri-axial accelerometers attached to the back of the jockey’s helmet, with a reported accuracy 0.0012 m s² √Hz^− 1 (Emerald, APDM, OR, USA) and sampling rate 128 Hz. All devices were synchronised to universal time and recorded continuously from the jockeys first ride to the completion of the last ride. Data from each device were downloaded after the completion of the day’s competition.

Data analysis

Accelerometer data were filtered with a Butterworth low pass filter between 0.1 and 1.1 Hz to remove acceleration peaks. The magnitude of acceleration vector was calculated as the sum of squares of the three movement planes. Acceleration data were then integrated twice to quantify displacement following published methods³⁶. Three 200 m sectionals (“Start”, “Middle” and “End”) were extracted from the corresponding portion of each canter or gallop section of the ride. Mean and peak linear accelerations and mean displacements were calculated for each 200 m sectional. Vertical displacement described the movement of the jockey’s head in the vertical plane, whereas magnitude of displacement described the sum of all movement of the jockeys head in the three movement planes (vertical, medial/lateral and fore/aft).

The time at which the jockey mounted and dismounted for each ride was manually recorded and subsequently matched to GPS data. Riding time began when the jockey mounted the horse and finished when the jockey dismounted at the conclusion of the ride. Non-riding time was the time spent between dismounting after each ride and mounting the next horse and varied according to the individual jockeys’ racing schedule. Horses ridden per hour was calculated as the total number of horses ridden by a jockey during the data collection day divided by the total riding time from the first horse to the last horse. Training impulse (TRIMP or exercise load) scores were calculated by summing the accumulated time (min) spent in five different HR zones of each 200 m sectional according to Edwards³⁷. A higher TRIMP score indicated higher physiological demand by the jockey. Percentage of HR_max was determined using the average HR for each 200 m sectional divided by the HR_max estimated as 220—jockey age.

Jockey data was matched to the race day records of every race start from 2004 to 2024, provided by NZTR. Days experience for each jockey was calculated as the number of days between the jockey’s first recorded career race ride and the date of data collection for this study. The number of race rides for each jockey was calculated as the number of race rides ridden prior to the day of data collection. The number of placings (placing was defined as passing the finish post in 1, 2 or 3rd place) per 100 race rides and the amount of stakes money won by each jockey were summed prior to the day of data collection.

All data were examined visually for errors and unrealistic values using histograms and scatterplots and then summarised using descriptive statistics. Data were presented as medians and interquartile ranges (IQR) unless otherwise stated. Normality of data distribution was tested using the Shapiro Wilk’s test. Differences between groups were determined using Kruskal-Wallis and Wilcoxon rank sum tests for non-parametric data.

The head displacement of jockeys was investigated using a mixed-linear regression model. The interaction between jockey and ride type (exercise or trial riding) was included in the model as a random effect. Univariable analysis was used within the mixed-linear model to identify variables associated with the mean magnitude of jockey head displacement at p < 0.2; these variables were included in a multivariable model. The multivariable model was built using a combination of forwards and backwards selection procedure whereby variables that improved the model fit based on a chi squared likelihood ratio test (p < 0.05) were retained in the model. Fixed effects included; jockey variables - gender, age of jockey (yrs), days experience, number of races ridden, placings per 100 race rides, log of stakes winnings ($), TRIMP and percentage of HR_max; and ride variables - ride type, ride sectional, speed (m/s), stride length (m), distance of ride (m), total time of ride (mins), non-riding time (mins), horse number, total number of horses ridden on day of measurement, horse contact (a qualitative measure of how hard the horse was to ride on a scale of 1–10), Borg’s rating of perceived exertion 10-point scale (Supplementary Table 1), and number of horses ridden per hour.

Collinearity between fixed effect variables were assessed using descriptive statistics to visualise relationships, with correlations (|ρ|>0.8) considered significant³⁸. Where collinearity was present between two variables, the variable that provided the best model fit was retained in the model. Models were checked against the null hypothesis of equidispersion (p < 0.05) and residual plots examined for goodness of fit.

All statistical analyses were conducted using RStudio (version 3.5.1, 2018; R Foundation for Statistical Computing, Vienna, Austria) with the level of significance set at p < 0.05.

Results

Anthropometric data for all jockeys (12 apprentice and 2 senior) riding on 34 separate occasions (n = 17 exercise days, n = 17 trial days) are presented in Table 1. A total of 85 exercise rides and 82 trial rides were recorded (Table 2). Horses ridden included male and female Thoroughbred racehorses, ranging from 2 to 10 years old. The two senior jockeys (one male and one female) were of similar age, weight and riding experience.

Table 1.

Anthropometric data (median and IQR) for apprentice (n = 12) and senior (n = 2) jockeys.

Variable	Male		Female		All jockeys		p-value
n	8		6		14
Age (yrs)	23	(21–28)	22	(21–26)	22	(20–28)	0.8
Height (cm)	164	(161–167)	160	(156–163)	162	(158–166)	0.07
Body Mass (kg)	52.8	(50.8–53.6)	51.6	(50.5–53.0)	52.6	(50.5–53.5)	0.7
BMI (kg/m2)	19.6	(19.4–19.6)	20.4	(19.5–20.8)	19.6	(19.4–20.4)	0.4
Experience race riding (years)	1.0	(0.1–3.1)	2.2	(2.0–6.9)	2.0	(0.1–3.6)	0.07
Number of race rides ridden	159	(7–462)	453	(187–841)	261	(15–560)	0.01

Table 2.

Ride metrics (median and IQR) for jockeys (n = 14) riding in 17 exercise days and 17 trial days.

Variable	Male		Female		All jockeys		p-value
Trackwork Ride Metrics
Number of jockeys (n)	5		5		10
Number of days recorded (n)	9		8		17
Total number of exercise rides (n)	45		40		85
Number of horses ridden each day (n)	6	(6–7)	6	(4–6)	6	(6–7)	0.05
Individual ride time (mins)	19.0	(16.2–23.7)	15.5	(13.0–18.8)	17.3	(14.0–22.8)	0.003
Median time between rides (mins)	9	(7–17)	10	(6–13)	10	(7–14)	0.3
Number of horses ridden per hour (horses/hr)	2.0	(2.0–2.6)	2.8	(2.1–3.1)	2.3	(2.0–2.8)	0.2
Canter time (s)	352	(305–406)	298	(242–330)	323	(288–376)	< 0.001
Canter distance (m)	2978	(2583–3574)	2551	(2239–2923)	2802	(2342–3074)	0.002
Canter speed (m/s)	8.5	(8.0–8.9)	8.8	(8.4–9.5)	8.7	(8.2–9.3)	0.04
Stride frequency (strides/s)	1.95	(1.89–2.00)	2.00	(1.91–2.05)	1.96	(1.90–2.04)	0.005
Magnitude linear acceleration of jockey head (g)	2.5	(2.3–3.0)	2.3	(1.95–2.6)	2.4	(2.05–2.9)	< 0.001
Peak linear acceleration of jockey head (g)	3.5	(2.7–3.5)	3.0	(2.7–3.5)	3.2	(2.8–3.8)	< 0.001
Vertical linear acceleration of jockeys head (g)	2.9	(2.5–3.3)	2.6	(2.3–2.9)	2.7	(2.4–3.1)	< 0.001
Magnitude of displacement of jockeys head (m)	0.12	(0.10–0.14)	0.10	(0.08–0.13)	0.12	(0.09–0.14)	< 0.001
Vertical displacement of jockeys head (m)	0.13	(0.11–0.15)	0.11	(0.09–0.14)	0.12	(0.10–0.15)	< 0.001
Jockey HR at canter (% of HRmax)	68	(62–76)	62	(59–64)	63	(60–70)	< 0.001
Jockey TRIMP score	0.97	(0.77–1.25)	0.73	(0.41–0.83)	0.83	(0.57–1.07)	< 0.001
Trial Ride Metrics
Number of jockeys (n)	6		4		10
Number of days recorded (n)	11		6		17
Total number of trial rides (n)	52		30		82
Number of horses ridden each day (n)	5	(3–7)	5	(5–7)	5	(3–7)	0.6
Individual ride time (mins)	9.8	(8.7–11.4)	9.4	(8.0–10.4)	9.6	(8.5–11.1)	0.2
Median time between rides (mins)	5.7	(2.2–17.5)	14.3	(2.9–21.2)	8.5	(2.2–19.0)	0.3
Number of horses ridden per hour (horses/hr)	2.5	(2.2–3.2)	2.5	(2.2–2.7)	2.5	(2.2–3.1)	0.7
Trial time (s)	64	(56–75)	65	(60–77)	64	(56–76)	0.4
Trial distance (m)	1072	(930–1256)	1103	(965–1252)	1086	(942–1252)	0.6
Trial speed (m/s)	16.7	(16.2–17.1)	16.4	(16.0–16.7)	16.6	(16.1–16.9)	0.03
Stride frequency (strides/s)	2.33	(2.25–2.40)	2.33	(2.25–2.42)	2.33	(2.25–2.42)	0.3
Magnitude linear acceleration of jockey head (g)	1.7	(1.4–2.1)	1.3	(1.2–1.6)	1.5	(1.3–2.0)	< 0.001
Peak linear acceleration of jockey head (g)	2.8	(2.3–3.5)	2.0	(1.7–2.5)	2.5	(2.0–3.1)	< 0.001
Vertical linear acceleration of jockeys head (g)	2.4	(1.8–3.1)	1.9	(1.5–2.1)	2.1	(1.6–2.9)	< 0.001
Magnitude of displacement of jockeys head (m)	0.06	(0.05–0.07)	0.04	(0.03–0.05)	0.05	(0.04–0.07)	< 0.001
Vertical displacement of jockeys head (m)	0.07	(0.05–0.96)	0.05	(0.04–0.06)	0.06	(0.05–0.09)	< 0.001
Jockey HR during trial (% of HRmax)	87	(79–92)	84	(77–87)	85	(78–91)	0.3
Jockey TRIMP score	0.93	(0.73–1.08)	0.82	(0.69–0.98)	0.87	(0.7–1.03)	0.04

*Canter was defined as speeds > 6.9 m/s and < 13.9 m/s.

Female jockeys in this study had more race experience than male jockeys (p < 0.001). Female jockeys had lower linear accelerations and displacements of the head and exercise load (TRIMP) than male jockeys (p < 0.001) although workload pattern each day (i.e., number of horses ridden each day, number of horses ridden per hour, ride times) were not different (p > 0.05) (Table 2).

A mixed effect linear regression model which contained the interaction between jockey and ride type (trial or exercise riding) as a random effect was fitted to the data to account for the inherent variation between individual jockeys and ride types. Head displacement was the dependant variable in the model, with ten descriptors which affect jockey head displacement included (p ≤ 0.2) after univariable analysis; jockey variables - age of jockey, days experience, number of races ridden, placings per 100 rides, stakes money won, TRIMP, percentage of HR_max and ride variables – ride type, speed and stride length (Supplementary Table 2). Stride length and number of races were removed from the model due to collinearity with their respective variables; speed and days experience (R² > 0.9). The final multivariable model with random effects performed better than an intercept- only baseline model (χ² = 32, p < 0.001) and retained four fixed effects with two interactions between the fixed effects (Table 3). Jockey head displacement decreased with increasing speed and greater jockey experience (Fig. 2). Jockey head displacement was lower in trials, with speed affecting head displacement less compared to exercise riding. Jockey head displacement increased with higher TRIMP, with the effect of experience on head displacement more pronounced at higher TRIMP levels (Table 3; Fig. 2). The interaction term indicates that 20% of the variance in head displacement is attributable to differences between jockeys within each ride type (Table 3).

Table 3.

Mixed effect multivariable linear regression modelling the association of jockey head displacement (m) with predictor variables; speed, days experience and TRIMP with the random effect of jockey interacting with ride type (exercise or trial riding).

Fixed effects	Estimates	CI		S.E		d.f	t-value	p
(Intercept)	0.18	0.16–0.20		9.7e-3		35	18.9	< 0.001
Days experience	-3.1e-6	-7.3e-6–1.2e-6		2.2e-6		17	-1.4	0.16
Speed	-8.3e-3	-9.6e-3 – -6.9e-3		6.5e-4		35	-12.7	< 0.001
TRIMP	0.012	0.003–0.021		4.5e-3		323	2.7	0.007
Random Effects
σ2 (residual variance)	5.7e-4
τ00 (Jockey: Ride Type) Between subjects variance	1.5e-4
ICC	0.21
N (Jockey)	14
N (Ride Type)	2
Observations	413
Marginal R2	0.66
Conditional R2	0.73

Fig. 2.

Predicted values (marginal effects) of mean head displacement (m) of jockeys (n = 14) in a mixed effect multivariable linear regression model with random effect of the interaction between jockey and ride type (trial or exercise riding) and fixed effects of ride type (trackwork or trials), race riding experience (number of race rides at time of measurement), speed (m/s) and TRIMP (exercise load) score. Shaded area denotes 95% CI level.

Discussion

This study demonstrates that the head displacement of jockeys is related to the speed of the horse, the jockey’s exercise load and their race riding experience. The strong association of jockey head displacement with horse speed was expected, due to the synchronisation of horse and jockey movement and the lower horse displacement (and thus jockey displacement) associated with higher speeds of gallop^5,15,18. There is less variation, and higher speeds within a trial ride compared to an exercise ride, which lessens the effect of speed on head displacement. Jockeys with greater race riding experience tended to have lower head displacements than jockeys with less experience, indicating that experienced jockeys are more skilled at attenuating horse oscillation with increased postural control. The tendency for greater head displacement with increasing TRIMP scores indicates that a higher exercise load is associated with a reduced ability of the jockey to attenuate horse oscillation, with this effect exacerbated in less experienced jockeys. The combination of these latter two measures indicates that measurement of jockey head displacement may provide a simple means with which to measure a jockey’s race riding ability or “race-specific” fitness.

In many competitive sports, reduced displacement of the head is associated with greater cognitive ability and reaction speed and may also result in long-term health benefits due to fewer acceleration and deceleration events of the head^21,39. As jockeys ride horses in the canter or gallop daily for approximately 30–150 min, the long-term effect of these chronic low-level accelerations may require further investigation. Mean linear head accelerations of jockeys in this study were similar to those observed in exercise, race and simulated race riding in the United Kingdom and United States^17,40, falling within the range of accelerations experienced in everyday living and below the threshold of 10 g for sports head impact data⁴¹. Peak linear head accelerations, both in track work and trials in the current study were lower than those recorded both in exercise riding (including gallop, 5.8 g)⁴⁰ and race riding (6.2 g) jockeys¹⁷. This may be due to the current study assessing only 200 m sectionals of the steady-state canter or gallop portion of the ride, rather than the entire ride, which may include events such as transitions between paces or pulling the horse up after the race, that can potentially cause greater head perturbation.

Motor skill in sport is inextricably linked with physical fitness, with decreases in both proprioceptive and cognitive ability observed after physical fatigue^42–44. Fatigue (both central and peripheral) reduces movement velocity and muscle activity, which can impact the postural stability of the jockey⁴⁵. Changes in movement patterns have also been linked to fatigue in lifting patterns of weightlifting athletes⁴⁶. Exercise load (TRIMP) measures the (internal) physiological demand of an activity and is related to an individual’s physical fitness. A jockey with a lower level of ‘race-specific’ fitness would experience a greater internal exercise load for the same activity (external workload) than a jockey with a higher ‘race-specific’ fitness level, hence could be more susceptible to fatigue⁴⁷. The association of increased TRIMP (exercise load) with greater head displacement may indicate a fatigue-induced reduction in postural stability caused by the demands of exercise, which is more evident depending on jockey experience. Jockeys could be expected to undergo a higher exercise load or more head displacement towards the end of ride (as they push to the finish), however, there is a lack of association between ride sectional (first, middle or end of the ride) and head displacement. Collectively, findings suggest that the increased head displacement in less experienced jockeys with a higher exercise load, may be attributed in part to the physical fitness of the jockey or accumulation of fatigue. More detailed assessment of fatigue on the postural stability of jockeys, incorporating measurements of controlled exercise and time-to-exhaustion may be a useful avenue for future research.

The physiological attributes of jockeys in various conventional fitness testing batteries have been previously described^10,11,19,48. Despite a consensus that jockeys require high levels of anaerobic and aerobic fitness, high lower body and core strength, and good balance, conventional tests tend to show little differentiation in these factors between jockey experience levels. However, distinct differences in muscle recruitment between apprentices and experienced jockeys, and between trackwork, trial and race riding have been demonstrated^4,5. Experienced jockeys, whilst having greater race day fitness than less experienced jockeys^5,10, also appear to have a more effective technique in moderating the displacement of the head during race riding. These findings are consistent with the relationship between head displacement and experience observed in this study, and indicate that the muscle pattern to attenuate head displacement is a learnt behaviour which can be conditioned through specific exercise. This has been documented in other sports such as sailing, windsurfing and motorcross^49,50.

Female jockeys had lower head displacement than male jockeys in both exercise and trial riding. However, they also tended to ride exercise work at slightly faster speeds than their male counterparts, which may have partly accounted for this effect. The female jockeys in this study had a higher overall level of race riding experience than male jockeys, which would likely correspond to a greater degree of postural control on the horse¹³. Additionally, they experienced a lower physiological exercise load for a similar workload as their male counterparts. Anecdotally, females who pursue a career as a jockey tended to come from riding backgrounds more commonly than the male jockeys in this study, who ‘learnt on the job’. This likely resulted in male jockeys having less postural control at the beginning of their careers than female jockeys, potentially increasing the time taken to achieve a similar level of race riding skill, exacerbating the difference between genders and experience levels in the current study. In contrast, male riders have been reported to have a biomechanical advantage when riding in traditional horse sports⁵¹. The act of ‘decoupling’ themselves from the horse, instead of sitting on their back (as in traditional horse riding) may result in different biomechanical challenges for male and female jockeys. These gender related biomechanical differences may provide different physiological challenges for male and female jockeys, and may warrant further investigation.

Currently, prospective jockeys are assessed qualitatively for riding competency by NZTR professionals before licencing and during their apprenticeship. The relationship between jockey experience and TRIMP (exercise load) in the attenuation of head displacement indicates that measurement of head displacement of jockeys in training rides could be used as a quantitative measure of both jockey fitness, training milestones and their ability to ride safely in races. The addition of a quantitative benchmark for jockeys to achieve before licensing could be a useful guideline to aid NZTR professionals in their qualitative riding assessments.

Limitations

There were a relatively low number of participants in this study, however data from multiple exercise and trial rides from each participant were recorded, which provides strength in the repeatability of the measurements and reduces variability caused by differences between participants. Three of the apprentices had not yet ridden in a race ride at the time of first measurement, but had race ridden by the second measurement day. Additionally, most of the participants provided data from two separate time points, between which race experience was gained, giving insight into the observable changes in apprentice head displacement at the start of their race riding careers. As the risk of falling from a horse during a race falls steadily during the apprenticeship⁵², this time period could be expected to be the time of greatest adaptation to the physiological demands of racing.

A limitation of the study was the lack of data obtained for more experienced or established jockeys, with only two senior jockeys included in the study. However, it could be expected that jockeys in an established career may be unlikely to greatly change their riding fitness (and thus head displacement). The age range of jockeys included was relatively small, which potentially reduced the impact of age within the model. Age was moderately colinear with riding experience (R² = 0.5), however was not significant in either the univariable or multivariable model, indicating that experience was more important than age in determining head displacement. Therefore, the inclusion of these more experienced jockeys may provide a benchmark range of reasonable head displacements in race riding, which warrant inclusion in the model.

The placement of the accelerometer on the jockeys’ helmet may have caused some inaccuracies in the measurement of head displacement by disregarding the amount of motion from the helmet/head interaction. Ideally the accelerometer would have been attached to the jockeys’ head itself, however, for practical reasons this was not feasible. The effects of this were considered to be negligible as the mean linear head accelerations observed were similar to those obtained when attached to skin of a jockeys head in the United Kingdom⁴⁰.

No data was included for race rides within this model. Race riding jockeys experience a higher exercise load, but lower head oscillation than jockeys riding in trials due to a slightly lower crouched riding position⁵. Therefore, whilst trials provide valuable insights, they may not fully capture the conditions and pressures of a race ride. Future research should aim to include raceday data to enhance the validity of the findings.

Conclusion

This study has identified differences in head movement of jockeys in exercise and trial rides according to horse speed, jockey race riding experience and exercise load. Head displacement decreases with increasing speed of the horse and increasing race riding experience of the jockey. A greater physiological load on the jockey exacerbates the effect of experience on head displacement. These results indicate that jockey head displacement may be used as a guideline to quantify jockey on-horse stability and to assess the skill and physical preparation of jockeys to achieve apprenticeship milestones and compete safely in races.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization and methodology, K.L., C.R., D.C. and E.G.; formal analysis, K.L., Y.C.; writing—original draft preparation, K.L.; writing— review and editing, all authors; supervision, C.R., D.C. and E.G.; funding acquisition, C.R. All authors have read and agreed to the published version of the manuscript.

Data availability

The datasets used and analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.