Refining the Para-Dressage Classification System

Results of the FEI study to establish a research-backed system of evaluating riders for para-dressage competition

Reprinted from the November/December 2025 issue of USDF Connection

By Hilary M. Clayton, BVMS, PhD, Dipl. ACVSMR, FRCVS

Para sports are organized sports that have been adapted to facilitate participation by people who have a functional (physical) impairment. (“Para” is short for “parallel,” meaning that the para sport is parallel to its able-bodied counterpart.)

In the equestrian world, dressage and driving are contested as para sports. However, the only equestrian discipline in the Paralympic Games is para-dressage, which has been included since Atlanta 1996. Interestingly, para-dressage is the only Paralympic sport with an artistic component. There are team, individual, and freestyle para-dressage tests. Para-dressage is judged similarly to able-bodied dressage, with scores awarded according to the horse’s performance.

Becoming a Para Athlete

To be deemed eligible to participate in a para sport, an athlete undergoes a process known as classification. For para-dressage, a classifier—a specially trained official who’s often also a medical professional—assesses the person’s functional limitations in terms of mobility, strength, and coordination using clinical assessments and by watching them ride.

Specific minimal impairment criteria must be met in order to qualify as a para-equestrian. Classification is sport-specific because the effect of a particular impairment is different for each sport. To be classified as a para-dressage athlete, the rider must show at least one of the following physical, intellectual, or visual impairments:

• Impaired muscle power that reduces strength
• Permanently impaired passive range of motion in one or more joints (an increase beyond the average range of joint motion, joint instability, or conditions such as arthritis are not considered eligible impairments)
• Limb deficiency with total or partial absence of bones or joints from birth or due to trauma or illness
• Leg-length difference present from birth or as a result of trauma
• Short stature due to abnormal dimensions of bones in the limbs or trunk
• Hypertonia—an abnormal increase in muscle tension and reduced ability of a muscle to stretch resulting from injury, illness, or a health condition
• Ataxia in which muscle movements are uncoordinated as the result of a neurological condition
• Athetosis, characterized by unbalanced, uncontrolled movements and difficulty maintaining a symmetrical posture
• Intellectual impairment originating before 18 years of age and that affects functioning and adaptive behavior in conceptual, social, and practical adaptive skills
• Visual impairment due to deficient structure of the eye, the optic nerve, or the visual cortex of the brain. (Note: Visual impairments are evaluated through a separate process and are not part of this article.)

In Paralympic sports, athletes are classified into sports “grades,” or degrees of impairment severity, with the aim of providing equitable competition by grouping together athletes with similar limitations in their athletic activity resulting from their impairment. In para-dressage, athletes are classified as one of five grades, from Grade I (most severely impaired) to Grade V (least impaired). The grades are determined by level of impairment, not by riding skill. Each grade has its own para-dressage tests, with the grade dictating the gaits and movements performed in competition as well as the size of the competition arena used.

Developing an Evidence-Based Classification System

The classification process is intended to provide a relatively level playing field. However, it is not a simple task to determine equivalence among riders with different types and severities of impairment.

In the 1990s, Dr. Christine Meaden developed a classification system for para-dressage, which has been modified over time. In 2015, the International Paralympic Committee mandated the development of evidence-based systems of classification across all Paralympic sports. The official governing body of the sport of para-dressage, the Fédération Equestre Internationale (FEI), became responsible for developing and implementing a classification system, including a list of eligible impairments that qualify an individual as a para-equestrian, tests to determine whether a rider meets the minimum qualification criteria for impairment, and a process for assigning riders to the appropriate grade. An evidence-based classification system should use objective measurements that quantify the degree to which an athlete’s impairment limits their activities.

The FEI decided to prioritize the development of a strong base of scientific evidence for the classification system. A team of researchers led by Professor Sarah Jane Hobbs of the University of Central Lancashire in England, in collaboration with colleagues from Hartpury University in the UK and me, undertook a multi-year study. Our task was to conduct a comprehensive review of how different types of physical impairment affect performance in dressage. What we learned would be used to inform recommendations for an evidence-based, sport-specific classification system for para-dressage.

Study Methodology

Our research comprised five areas of focus:

Literature reviews. Performance measures—the relationship between impairment and fundamental sport-specific activities—should be the basis of any classification system for para sports. But no studies had extensively investigated or defined the fundamental abilities required to perform successfully in para-dressage.

In year one of our study, we addressed this gap by conducting two reviews of the existing scientific literature. First, we identified objective measurements of horses’ performance in dressage and riders’ functional abilities that may predict elite dressage performance. Second, we sought to identify objective, valid, and reliable clinical tools that could be used to measure eligible impairments in the later stages of the research project.

Note: The results of these reviews and the other elements of the study can be accessed through the QR codes in “Further Reading” below.

Stakeholder interviews. We conducted semi-structured interviews of 30 para-dressage athletes, coaches, judges, and classifiers from various countries, with the goal of gaining insight into key determinants of sport-specific performance and the impacts of impairment on performance in para-dressage. It was very important to us that stakeholders’ views and experiences informed the classification system.

According to the interviewees, the key determinants of performance are:

• The rider must have sufficient muscular coordination and joint mobility to achieve and maintain correct posture in the saddle, and to maintain dynamic postural control in order to follow the horse’s movement.
• The rider should be able to coordinate leg, seat, and hand aids in order to influence the quality and accuracy of the horse’s movements.

These qualities likely sound familiar because they apply to all riders, but they are typically more challenging for para riders. Because the para-dressage judge’s scores are based on the horse’s performance, gait quality can have a large effect. One para-dressage rider emphasized that both impairments and skill level can affect the horse’s performance, for better or for worse. In addition, personality traits and their influence on the horse-rider partnership were considered influential determinants of performance.

Riding test on a horse simulator. We designed a study to investigate the effects of specific impairments on dressage performance. The participants in our study were 21 elite para-dressage riders with a diverse range of impairments and representing all five grades; and 11 able-bodied dressage riders who were actively competing at the FEI levels. The study consisted of two phases: a standardized performance-assessment test on a riding simulator, and a clinical assessment performed later the same day (or, for one rider with a severe impairment, the following day).

We chose to use a horse simulator instead of a live horse because of the highly predictable manner of movement and the lack of stride-to-stride variability. We realized, however, that this predictability does not provide the same demands for postural control that are inherent in the slight unpredictabilities in the movements of a real horse.

All participants wore approved equestrian headgear. Depending on their needs, the para-athletes in the study used compensating aids or adaptive equipment: looped reins (six riders), a seat saver (three riders), no stirrups (three riders), enclosed stirrups (one rider), stirrup irons attached to girth (one rider), a whip (six riders), and a soft or hard hand hold (two riders). All riders used the same dressage saddle with the exception of one para rider, who used their own saddle.

The riding simulator is capable of different gaits and dressage movements. For our study, we positioned it in front of a screen showing a virtual small (20 meters x 40 meters) dressage arena with standard letters, as shown in Figure 1.

We equipped the simulator and each rider with reflective markers and motion sensors to measure three-dimensional movements of the rider’s pelvis, trunk, and head; and the simulator’s body, neck, and head. The only instruction we gave the riders was to ride each movement “for a 10.”

All of the participants performed the short dressage test shown in Figure 2. The test took approximately two minutes to complete and was designed to meet the abilities of all riders as well as our research goals. The final trot was optional, and one rider opted out. Riders were responsible for making transitions and for steering the simulator within the virtual arena on the screen in front of them.

The video recordings and data from the motion sensors were separated into individual strides. For each rider, we analyzed three nonconsecutive strides of walk, three right turns, three left turns, and four halts. When trot was performed, three nonconsecutive strides were extracted. A battery of analytic and statistical tests were used. I’ll outline them here; for the full scientific paper, see the link in “Further Reading.”

We evaluated harmony between rider and simulator based on movement frequencies. Actuators move the simulator’s trunk in a regular and repeated motion pattern in each stride. A difference in coordination between rider and simulator can change the simulator’s motion pattern and was evaluated as a lack of harmony.

Stability was evaluated by comparing accelerations of the simulator trunk and the rider’s head during walk-halt transitions, to determine whether the rider’s head accelerated more or less than the simulator. To assess the consistency of the coordination pattern within individual riders and between different riders, we looked at angular changes between the rider’s trunk and pelvic rotations.

A dressage rider’s trunk should be close to vertical at all times, even with the accelerations, decelerations, and rotations of the horse’s body during each stride. The walk is the least challenging gait in this regard due to the slow speed and lack of suspension phases, but there are relatively large phase differences between the trunk motions of the rider and the horse.

Dynamic symmetry of trunk motion, evaluated from the motion patterns of each rider’s pelvis and trunk in the horizontal plane, was significantly different between the two rider groups, with para riders showing greater asymmetry and variability in their average trunk-motion patterns. Since the para riders in the study were known to be highly skilled, trunk asymmetry was attributed to limitations in their functional capacity. The variability among para riders may reflect individual differences in trunk-control strategies used to overcome their specific impairments. Stability of the trunk affects the demand on the upper arms in controlling the distance from the hands to the bit, which poses an added level of coordination complexity.

Trunk dynamic symmetry also influences the collective score for posture and the consistency and lightness of rein contact, which are judged within behavioral and ridability traits. This suggests that trunk dynamic symmetry is an important measure of performance limitation in para-dressage.

Unmounted performance tests. After the riding test, a battery of tests evaluated participants’ relevant physical impairments. The tests were selected and refined through extensive consultation with accredited and highly experienced FEI classifiers and physiotherapists. The following five clinical-impairment assessment measures were evaluated for their ability to measure eligible impair-ments in para-dressage riders. Joints were selected for evaluation based on their relevance for dressage riding performance.

Revised Modified Ashworth Scale: A reliable, accurate, and compre-hensive method of assessing muscle tone, particularly in specific muscle groups and when evaluated in the riding position.

Function in Sitting Test: Measures static, proactive, and reactive balance as well as sensory integration. In people with balance dysfunction, it can measure functional sitting balance. Although this test had not been applied previously to evaluation of para athletes, it showed sport-specificity for para-dressage and was selected as a potential tool for identifying the impact of impairment on sitting function.

Scale for the Assessment and Rating of Ataxia: A reliable and valid test for detecting impaired balance and coordination in people with ataxia. This test has been used for athlete classification in para-archery and para-cycling. In dressage riders, it was effective in detecting impairments related to sitting function.

Trunk Impairment Scale: Evaluates static and dynamic sitting balance to assess motor impairment of the trunk and sitting function. This test proved useful in detecting asymmetries in para riders, but several able-bodied riders also showed asymmetrical trunk rotation when performing tests of dynamic coordination.

Hand-Held Dynamometry: Measures maximal voluntary muscle contraction using a dynamometer stabilized by the tester. The dynamometer measures the force exerted and assesses the subject’s muscular strength. This test proved useful for assessing strength deficiencies resulting from para-dressage riders’ neuromuscular deficits. Strength of the shoulder and elbow were strongly correlated in most para riders, indicating comparable arm strength; but riders who were unable to use their legs scored zero for some measurements in this test. In able-bodied riders, measurements of trunk and hip strength were more strongly correlated.

Relationships Between Impairment and Performance

Our study sought to investigate the extent to which physical impairments may affect dressage performance, then to apply that information to develop sport-specific impairment-assessment measures for classification purposes. Compared with many other para sports, dressage impairments are particularly challenging to evaluate because of the range of eligible rider impairments; the effect of the horse on rider performance; and the high rate of permanent, performancelimiting injuries in the general riding population.Nevertheless, we found relationships between impairment measures and performance measures in para riders that were not evident in the able-bodied study group. The impairment measures of sitting function (including coordination) and muscle tone predicted up to one-third of the impact of impairment on performance.

The association between sitting-function measures and performance measures was stronger in those riders whose impairments were more severe. Movement restrictions in wheelchair users may reduce strength in core postural and movement muscles. Test 2, which measures function in a sitting position, is useful for functional assessment of the performance demands expected of para-dressage riders. This test includes asking the subject to reach forward and to pick up an object, which requires activation and control of the spinal and abdominal muscles.

Individuals with better trunk control in dynamic conditions are more proficient at performing trunk flexion/extension tasks. Coordinated activation of the trunk flexors and extensors is essential for posture control while riding, with one of the challenges being stabilization of the trunk and pelvis in all directions while still allowing the pelvis to follow the horse’s motion. For the elite para athletes in this study, those who struggled more to pick up an object from the floor were less able to position their trunk over their pelvis when riding the simulator.

Research Recommendations for More Accurate Classification

The goal of this research study was to investigate relationships between functional impairments and dressage performance in para-dressage riders. We were tasked with determining which tests of mobility, strength, and coordination are useful for detecting relevant functional impairments. The results indicated that tests assessing impairment of strength, sitting function, and muscle tone detected deficits that affected athletes’ dynamic trunk symmetry when they rode a dressage test on a horse simulator.

Our research showed that trunk control—evaluated using dynamic posture and symmetry performance measures—is the most promising indicator of the impact of impairment on performance, with clear differences between the para-dressage athletes and the able-bodied riders in our study. We identified three assessment tests that proved useful for predicting the impact of impairment on performance in para-dressage riders but not in able-bodied riders. In the para riders, variability in coordination was independent of muscle strength; in the able-bodied riders, coordination variability was linked to strength of trunk rotation and external hip rotation.

These findings provide the basis for a robust, scientific evidence base that can be used to aid in refining the classification system for para-dressage.

---