Over the past decade, Sue Dyson, MA, Vet MB, PhD, and a range of collaborators from various specialties have developed the ridden horse pain ethogram (RHpE).
“I have long believed that training or rideability problems often reflect underlying pain and that a horse’s behavior, rideability, and quality of movement can be immediately transformed by removing their pain by using nerve blocks,” Dyson explains. “I felt it was essential to find a new way of making riders, trainers, and other horse people recognize that abnormal behaviors could in fact reflect pain. It was also important to raise awareness of veterinarians about the importance of ridden exercise during investigation of poor performance and to give them a tool to facilitate recognition of pain in horses which were not overly lame in a conventional sense.”
Behaviors such as mouth-opening and tail-swishing are not normal for pain-free horses, she notes. When a horse displays those behaviors when ridden but is sound on the longe or in hand, he’s often accused of being poorly behaved.
After meeting Jeannine Berger, DVM, Dipl. ACVB, Dipl. ACVW (welfare), of Monticello Veterinary Practice, in Winters, California, and discussing their concerns regarding the relevance of “poor behavior,” they decided to collaborate and explore the association between abnormal behaviors and performance under saddle, ultimately developing the RHpE.
“We knew from the outset that this work would face skepticism,” Dyson relayed in her book Harmonious Horsemanship: Use of the Ridden Horse Ethogram to Optimise Potential, Partnership and Performance. “The equine world is entrenched in tradition. We have grown up accepting that there are grumpy horses and that there are naughty horses. Our long-term goals were to change these misconceptions, to demonstrate that many of the behaviors which have been accepted as normal are actually abnormal and reflect discomfort. We need to reset what is regarded as normal.”
In this article, we’ll explain how the RHpE was developed, how veterinarians can use it, its potential limitations and caveats, and where Dyson would like to see the tool used in the future.
The Birth of the RHpE
One of the first studies Dyson published with key collaborators Berger, Andrea Ellis, DVM, and Jessica Mullard, BSc, BVetMed, MRCVS, (2018) described the assessment of facial expressions in ridden horses. In other species and nonverbal humans (neonates and adults with dementia), facial expressions can be used to determine the presence or absence of pain.
“In that 2018 study, we asserted that pain occurs commonly in athletic horses and that pain is often unrecognized in hand and/or when ridden,” Dyson says. “As a result, many horses that are experiencing pain are instead labeled as ‘problem horses’ and that their poor performance is behavioral rather than pain-related. We compared the facial expressions of non-lame horses and lame horses using a standardized method. The assessor was blinded to the category of horse but based on facial expression alone was able to distinguish between lame and non-lame horses.
“At that time, several research groups were working on ethograms to facilitate recognition of pain in horses at rest, but no ethogram (defined as a catalog of behaviors, each with strict definitions) existed in ridden horses in relation to musculoskeletal pain,” she continues. “We therefore developed an ethogram for whole-horse behavior, including the face, body, and gait of ridden horses.”
In that phase of the ethogram development, Dyson et al. considered all three (facial, body, and gait) behavioral markers. They assessed video recordings of several hundred lame and non-lame horses and identified 117 behaviors, many with time-based definitions, that might be useful for determining pain. Examples of facial markers included ear, eye, and mouth position (for example, mouth open with separation of the teeth for ≥ 10 seconds), and body markers included tail carriage and head position and head movement up and down (but not in rhythm with the trot) or side to side. Gait behaviors included straightness, consistency of rhythm, bucking, rearing, stumbling, and others.
Mullard, an equine veterinarian with training in equine behavior, analyzed video recordings of nine horses, three non-lame and six lame. For each of the 117 behaviors, she answered yes (the behavior was present) or no (the behavior was not present). Each horse was observed for an average of 5 minutes. Mullard assessed each video recording twice over several weeks to test the repeatability of the ethogram’s application.
“Overall, there was excellent agreement between the two assessments,” says Dyson. “Based on those data, the behaviors of the lame and non-lame horses were compared. Behaviors for which there was not good agreement between the two assessments were eliminated, and other behaviors were amalgamated. For example, several different types of bucking were combined.”
This adapted ethogram, which comprised 55 behaviors, was applied to 13 non-lame and 24 lame horses. Again, an evaluator who was blinded to the horses’ soundness analyzed the video recordings.
All horses, lame or non-lame, were capable of showing these facial, body, and gait behaviors; however, horses in pain (i.e., those that were lame) showed more behaviors. Specific behaviors strongly associated with pain included head carriage (e.g., tipping the head on one side), ears back continuously for at least 5 seconds, mouth opening repeatedly with separation of the teeth for ≥ 10 seconds, exposure of the tongue outside the oral cavity repeatedly, incorrect canter, and spontaneous breaking from one gait to another.
Based on those behavior assessments, Dyson’s team solidified a “final” ethogram of 24 behaviors, most of which were at least 10 times more likely to be seen in a horse with musculoskeletal pain compared with a non-lame horse. This final ethogram, the RHpE, was applied to the same 13 non-lame and 24 lame horses. The non-lame horses had an average score of 2/24 with a maximum of 6/24, whereas the lame horses had an average score of 9/24 with a maximum of 14/24. These differences were significant.
“The results suggested that a horse with a RHpE score of ≥ 8/24 during a work period of five to 10 ten minutes is likely to have musculoskeletal pain,” says Dyson.
At this point in the ethogram’s development, the researchers advised observing horses from all perspectives: in front, behind, and from the side, at a trot and canter, and transitions between gaits. Horses should move around the periphery of the ring and in 10-meter-diameter circles to left and right in rising trot.
“A horse should perform its full repertoire of movements, because it may only be when performing biomechanically more demanding movements (for example, half pass) that a horse may experience discomfort,” says Dyson.
Emphasizing the importance of this ethogram for recognizing pain in ridden horses, Dyson says, “Changes in body postures or resistance to responding to a rider’s cues could alert riders of potential pain. Recognizing this pain and investigating it further may significantly improve equine welfare and would avoid implementing harder training practices or punishment-based training that can further aggravate physical pain or make horses unwilling to follow the riders’ cues.
“Horses are prey animals and have evolved to be stoic,” she adds. “Listening to these subtle behavioral changes is important. Moreover, the ethogram provides us with a tool to differentiate training and pain-based problems.”
The authors also emphasized it’s the total RHpE score that’s important; a variety of factors could cause each individual behavior.
Based on the data in the 2018 study, Dyson et al. concluded, “Further research is necessary to verify this new ethogram for assessment of pain in ridden horses.”
The Next Step Toward Recognizing Pain
The team continued their investigation into pain as a common cause of behavioral and gait abnormalities that are often overlooked under saddle. In a follow-up study, they applied the RHpE to 10 horses with bilateral hind-limb lameness before and after resolving lameness via diagnostic analgesia (nerve blocks).
“If the number of behavioral markers was reduced following diagnostic anesthesia, then the value of the ethogram would be better appreciated,” explains Dyson.
Her team noted immediate substantial reductions in RHpE scores after removing pain via diagnostic analgesia, indicating a causal relationship between pain and the behaviors and that the behaviors were not habitual.
Dyson also wished to determine if untrained assessors from a variety of professional backgrounds could apply the RHpE with similar accuracy as a trained assessor. So, in another study, they investigated the behaviors of 21 lame horses presented to the assessors both before and after nerve blocks in a randomized order. The assessors scored the 24 behaviors in a binary fashion (i.e., present or absent) using video recordings. Dyson’s team compared the results of the untrained assessors with those of an assessor experienced in applying the RHpE. The RHpE scores prior to diagnostic analgesia ranged from 3 to 12, with a median of 10. Recall that a score ≥ 8 is highly suggestive of the presence of musculoskeletal pain. After diagnostic analgesia, the number of behaviors decreased to 0-6, with a median of 3.
“This provided further evidence of the relationship between pain and behavior,” Dyson says. “There was also reasonable agreement between the untrained assessors and the trained assessor. Behaviors that the untrained assessors found difficult to determine accurately were identified. For example, an intense stare for at least 5 seconds.”
Dyson did point out that some of the lame horses included in this study had total RHpE scores below the threshold of 8.
“Still, after diagnostic analgesia, no RHpE scores exceeded 6, and in nine of the 21 horses the score was £ 2,” she explains. “These results support the previous finding that scores ≥ 8 likely reflect musculoskeletal pain. Eight remains a good cutoff likely to reflect pain.”
Because the ethogram clearly measures behavior changes after pain has resolved, the researchers suggested it could potentially be used for monitoring the progression or resolution of lameness over time.
An Ethogram for the Masses?
Next, Dyson and colleagues wished to explore whether equine veterinarians trained to the ethogram could effectively apply this tool “in real life.” In that 2020 study published in Equine Veterinary Education, 10 equine veterinarians with varying years of experience were trained to use the RHpE. This training involved having the veterinarians watch a PowerPoint video presentation discussing the behavior of ridden horses with musculoskeletal pain prepared by Dyson. The veterinarians then applied the RHpE to video recordings of several horses and received personal feedback on their performances.
The study involved 20 horse-rider pairs whose riders believed their horses were working comfortably. Each pair performed a standardized dressage-type test of 8.5 minutes duration. An independent assessor determined the presence or absence of lameness in person. The horses were also videotaped. Dyson served as the experienced observer, using the ethogram in both live horses and the video recordings after an interval of several weeks.
In the study, 16 horses were mildly lame. All non-lame horses had a score of < 8, and the majority of lame horses had scores ≥ 8.
“Overall, there was good correlation between the inexperienced and experienced assessors, and there was excellent consistency in overall agreement among all assessors,” says Dyson. “Thus, after training, even inexperienced veterinarians can apply the RHpE, making it a useful clinical tool for determining the presence of musculoskeletal pain in ridden horses performing poorly. There was also excellent agreement between application of the RHpE live compared with assessment of video recordings, indicating that it can reliably be used in the field.”
Applying the RHpE to Three-Day Event Horses
By 2020, the researchers were applying the RHpE to horses competing at 5* three-day events. But unlike previous studies designed to compare RHpE scores between lame and non-lame horses, the goal was to assess the RHpE in fit and healthy high-level performing horses to determine if there was any relationship between the scores and performance. To do this, Dyson and Ellis assessed horses during warmup for the dressage phase of the competition. They then analyzed those scores to identify any correlations with performance, such as failure to complete the cross-country phase. They hypothesized that horses exhibiting ≥ 7 behaviors on the RHpE would be less likely to complete the cross-country phase than horses scoring < 7.
“Note that these were elite athletes and, therefore, a lower threshold was used than the 8/24 to differentiate lame and non-lame horses,” relays Dyson.
Dyson evaluated 137 horses and found that RHpE scores ranged from 0 to 9. Scores were lower in horses without evidence of lameness (mean of 3) compared to horses with gait abnormalities (mean of 5). Only 10 of the 137 horses had scores ≥ 8.
Fifty-nine percent of horses with an RHpE score ≥ 7 failed to complete cross-country, whereas only 33% of horses with an RHpE score < 7 failed to complete.
Other key findings included:
- A positive correlation between RHpE scores and dressage penalties.
- No associations between RHpE score and cross-country time penalties or total showjumping penalties.
- A significant relationship between RHpE score and final horse placings.
“Thus, horses that pass an in-hand inspection, such as eventing horses, may still show gait abnormalities presenting as behavior changes,” she says. “These findings support using the RHpE for identifying horses that could benefit from veterinary diagnostics and treatment to improve both performance and welfare.”
Applying the RHpE to Dressage Horses
In 2021 Dyson and Danica Pollard, BSc(Hons), MSc, PhD, applied the RHpE to elite dressage horses.
“The goal was to use the RHpE in a large cohort of apparently pain-free horses and to use the data to educate owners, trainers, judges, and veterinarians about the power of the RHpE,” says Dyson.
They selected dressage horses competing at an elite level: FEI World Cup Grand Prix competitions. They obtained video recordings of 147 competitors from seven FEI Grand Prix events via YouTube or FEI-TV. All competitors’ videos were standard in that each horse was viewed from similar perspectives for each movement of the 6.5-minute test.
As expected in elite dressage horses, the median RHpE score was low—only 3 out of 24 (range, 0-7). However, the authors pointed out that some horses showed episodic discomfort and suggested identifying the cause of this discomfort might enhance equine welfare and performance. Indeed, the data showed a significant negative correlation between the RHpE scores and the judges’ good marks: As the RHpE score increased, the judge’s scores decreased.
“The most commonly observed behaviors included mouth open with separation of teeth for ≥ 10 seconds and the front of the head behind the vertical for ³ 10° for ³ 10 seconds, intense stare for ≥ 5 seconds, repeated tail swishing, and ears back,” describes Dyson.
The RHpE and Saddle Fit
In a recent prospective study, Dyson and Pollard applied the RHpE to 150 horses undergoing assessment of poor performance before and after nerve blocks and a change of saddle. In the 56 horses (37%) in which saddle fit was considered likely to compromise performance, median RHpE scores decreased from 9 to 2 (out of 24) after the interventions. They noted no correlation between RHpE scores and lameness grade.
Further, a previous study (Dyson et al. 2021) involving 148 horses in normal work revealed an association between RHpE scores and tight tree points of the saddle.
“These studies emphasize the importance of correct saddle fit for optimal performance,” says Dyson.
Limitations and Recommendations for Using the RHpE
“The RHpE was developed as an easy-to-use tool to facilitate identifying horses with the likely presence of musculoskeletal pain,” says Dyson. “This tool can be used not only by veterinarians but also riders, trainers, and paraprofessionals. To date it has been validated using no less than 1,500 horses in a variety of disciplines and both lame and non-lame horses before and after diagnostic analgesia.”
Such a tool is important because a substantial proportion of horses deemed comfortable by riders have pain-related gait abnormalities. This finding implies some riders are unable to recognize musculoskeletal pain, which might compromise their horse’s welfare. Considering the social license of riding and competing horses, all tools should be recruited to ensure pain is recognized and addressed.
Despite the success of the RHpE, Dyson herself recognizes its limitations. As she and colleagues relayed in a recently published commentary (Berger et al. 2022), a number of factors can influence the presence or absence of the 24 behaviors included in the ethogram. Those include rider size and morphology; rider skill; saddle fit for both horse and rider; the presence of bit type or fit; and oral lesions.
To maximize the correct application of the RHpE, Dyson makes several recommendations:
- The horse’s sclera should be examined at rest to determine if one or both is visible.
- Footing should be assessed because if the surface is deep, a toe drag might be observed in otherwise normal horses.
- Horses should be evaluated after warming up.
- The walk, trot, and canter should all be assessed, and transitions, 10- and 20-diameter circles in both directions should be included. In fact, the horse’s full repertoire of movements should be evaluated.
- Remember, it is the total RHpE score that is important.
Berger et al. also made it clear that the RHpE is not synonymous with a pain scale.
“The RHpE scores indicate the absence or presence of musculoskeletal pain, not the degree of discomfort,” adds Dyson.
Future Directions and Goals
“The RHpE can be applied anywhere and by anyone,” Dyson says. “It can be used as a tool to regularly monitor a horse’s well-being and provide reassurance to riders that their horse is OK. When applied sequentially, rising scores can indicate a likely impending problem. The RHpE can also be used at prepurchase examinations.”
In sum, these studies identify the RHpE as a reliable, practical, and powerful tool that can be used in the field or in a research setting. This ethogram allows us the power of refuting claims that a horse is poorly behaved (e.g., by bucking under saddle) when this work suggests horses that buck with an RHpE score of ≥ 8 likely have musculoskeletal pain that explains the behavior.
“It is imperative we recognize what pain-free horses look like,” she concludes.
Related Reading
- Horse Back Pain Rehabilitation
- Is That Horse’s Radiographic Neck Lesion Important?
- Managing Lumbosacroiliac Joint Region Pain in Horses
Stay in the know! Sign up for EquiManagement’s FREE weekly newsletters to get the latest equine research, disease alerts, and vet practice updates delivered straight to your inbox.
