Do Horse Walkers Cause Stress, and Can we Reduce it by Exercising Horses Together?

Written by Caitlyn Cuthbert

The isolation of equines has been proven to negatively affect their welfare (Mal et al., 1991 and Hartmann et al., 2011), with horses stabled with no contact showing significantly more stress than those in group living situations (Yarnell et al., 2015). Even with substantial research confirming the stress caused by isolation, domesticated horses are subjected to long hours in isolated management (Kay and Hall, 2009 and Bachman et al., 2017).

Horse walkers are seen as a convenient method of exercising a horse. Horses are often put into horse walkers alone, and left alone as they exercise. The horse walker works by manipulating the natural predator-prey mechanism, by having a partition chase the horse to move forward. There is a lack of research to support the use of horse walkers, which may compromise horse welfare. This may be due to the difficulty of experimental design or lack of resources; however, the growing popularity of horse walkers emphasises a need for well-designed research. There is a particular welfare concern for horses owned by recreational riders (Hockenhull and Creighton, 2013 and Hemsworth et al., 2015) who may be using walkers without understanding their potential welfare implications and without a good knowledge of how horses indicate stress and discomfort.

My research, undertaken as part of my final year dissertation project for my BSc (hons) in equine behavioural science at Writtle University in the U.K., aimed to quantify the difference in physiological and psychological stress of a horse whilst isolated on a horse walker, compared to in its companion pair. I wanted to find out how stressed horses were when exercising in a horse walker, and whether they were more or less stressed when they were alone or with another horse from their herd. To do this, I designed an experiment to measure stress in horses exercising in a horse walker by measuring their heart rate, heart rate variability, and behavioural stress indicators.

Measuring stress

Physiological stress parameters like heart rate and heart rate variability are an accurate measurement of stress due to their involuntary nature. Animal subjects are unaware of what is being measured, which guarantees collection of reliable data. The higher the heart rate is, and the less varied it is, the more stressed the horse is (Visser et al., 2002; Janczarek and Kędzierski, 2011; Weinstein et al., 2007). The more varied the heart rate is, the less stressed the horse is (Castaldo et al., 2015).

Horses are expressive animals and show varying degrees of often subtle behavioural indicators dependent upon the level of stress imposed on them (Young et al., 2012). Stress results in behaviours such as a clamped tail, straight lip line, drawn nostrils, ears below neckline, facial tension, high head position, and a guarded body posture.
Aversive behaviours such as rearing, side stepping, backing up and planting are normally seen when the horse is trying to avoid a situation. Severe sudden stress engages the flight response, and explosive behaviours such as bolting are observed. I chose to measure behavioural indicators alongside physiological indicators to get a fuller picture of the horse’s psychological state in the walker. At the end of this article, I’ve included a pictorial ethogram of stress indicators from the horses in my study.

Molenkoning Horse Walker at Writtle University College Equine Training and Development Centre (Source: Author, 2018)

Molenkoning Horse Walker at Writtle University College Equine Training and Development Centre (Source: Author, 2018)


16 horses and ponies (n = 16) from Writtle University College Equine Training Development Centre were used for the trial. This ensured that all horses were stabled and managed in the same manner. The study used 16 subjects (n =16) to ensure statistical accuracy (Ruxton and Colegrave, 2011). The only other study on horse walker stress by Giese et al. (2014) used 12 horses. All horses were habituated to the horse walker for a minimum of three times a week for eight weeks prior to the trial. A varying population of mares, geldings, hot bloods, warm bloods, cold bloods, horses, ponies and ages were used in the trial. Exclusion criteria involved horses that refused to go on the walker or horses that became too stressed on the walker, displaying bolting and rearing behaviours, endangering themselves and the handlers.

A diagram to show the Polar Wrist watch attached to the horse’s head collar to ensure a good connection for the HR and HRV data (left) and show the location of the iPhones whilst fitted onto the bars of the partitions in the horse walker (right). 

A diagram to show the Polar Wrist watch attached to the horse’s head collar to ensure a good connection for the HR and HRV data (left) and show the location of the iPhones whilst fitted onto the bars of the partitions in the horse walker (right).

The horse walker used was a Molenkoning Oval Horse Walker with automated programmable speed settings, six electric partitions mounted on overhead tracks, an internal observation area, roof over horse walking track, mesh screens and rubber matting. The speed was set at 3, which is a slow walk. The walker was programmed to change the rein every five minutes.

The horses wore a PolarÒ EquineWear link transmitter strap with the PolarÒv800 HR monitor. Four Apple iPhones were mounted onto the bars of the horse walker to record the horses continuously

Data collection

The trial was carried out over two days. The horses were grouped into companion pairs based on their turnout groups. One horse went on first isolated, followed by the second horse isolated and then together. This order was swapped on the second day. A wash-out period was not required as the horse walker was not intended as a novel object in this trial.

Each horse was fitted with the Polar v800 heart rate monitor in the stable and a baseline heart rate and heart rate variability was recorded for five minutes. After the baseline measurements were recorded, the first horse from the companion pair was taken to the walker. The horse had a head collar on for the whole trial, but the lead rope was unclipped once on the walker. The horse walked for 10 minutes isolated and then returned to its stable. This was then repeated with the second horse from that companion pair. Then both horses from that companion pair were put on the walker.

The horse order was swapped on the second day of data collection to ensure a fair trial. The whole trial was repeated the next day to take an average and ensure quality and quantity of data to be able to draw accurate conclusions.


Heart rate

To recap, as a rule, the higher a horse’s heart rate is, the more stress it is under. Therefore, we would expect that horses under more stress would have a higher mean heart rate than horses under less stress.

Horses on the walker had an elevated mean heart rate compared to their baseline, whether they were exercising alone or in a pair. When exercised in a pair, however, horses’ mean heart rate was not as high as it was when they were exercising alone. My results showed that the difference between the isolated and paired horses’ mean heart rates was statistically significant (p.033).

The horses’ heart rates in the study all reached 70-100% of their individual calculated maximum heart rate. Five horses exceeded their maximum heart rate whilst isolated. Even with a companion, two of those five horses exceeded their maximum HR, which suggests that there is a cause for concern with use of a walker even if the horse is with a friend.

Heart rate variability

As stated above, existing research suggests that the less varied a horse’s heart rate is, the more stressed it is. We would therefore expect to see a decrease in heart rate variability if the horse is stressed on the horse walker, and a relative increase in heart rate variability if the horse is less stressed when exercised in a pair.

Behaviour score

The most common behaviours seen in all 16 horses were the head raised above the wither, maxillary tension, mandibular tension, drawn nostrils, and a straight lip line. Furthermore, ears rapidly changing direction and orbital tension were observed in 15 out of 16 horses. The difference in the behaviour score of the horses exercised whilst isolated compared to with a companion had the highest significance of all data recorded in the trial (p.0000055). This highlights the importance of measuring behaviour as an indicator of stress, alongside physiological parameters. It also highlights how important an understanding of equine body language is to understanding their psychological state, and therefore of the need for recreational horse owners to be educated on this topic.

Limitations of the study

Heart rate and heart rate variability data was hard to collect due to the heart rate monitors needing a lot of saline solution, especially on those horses who were not clipped around the girth area. The study originally included 20 horses; however, this had to be cut down to 16 horses as the electrodes on the heart rate monitor did not accurately pick up four horses’ heart rates, and therefore they could not be included in the data analysis.

I was only able to access a small sample number (n=16) of horses; this was the maximum sample size available after the exclusion criteria were applied. However, this sample size is actually larger than other published studies on horse walkers: Giese et al. (2014) used 12 horses and Murphy (2008) used 10 horses for their research.

What are horse walkers doing to horses?

Based on the results, I concluded that:

1) Horse walkers do cause horses to display physical and psychological indicators of stress, whether alone or in a pair.

2) There is a significant difference between the stress shown by the horse whilst exercised on the horse walker isolated and whilst in a herd group pair.

I hypothesise that the horses are more stressed when they are isolated because of their gregarious instinct.

What is causing the horses to be stressed on the walker?

There are a few different factors that could be contributing to the stress the horses are experiencing, which horse owners and behavior consultants should be aware of so that they can minimize their effect if they do decide to use a horse walker.

The major reason that horses are stressed on the walker is because of the way the walker operates. Trapping a flight animal inside a partition, with bars in front and behind, and no escape route, mimics the predator-prey chase mechanism, which is likely to cause a stress reaction.

Sensory issues

The way the walker is designed could be causing unpleasant sensory stimuli for the horses, thus causing stress. The model of horse walker used had the partitions moving on an overhead track. There is also a metal roof over the walker, which amplifies noise from rain and wind (Dubout, 1969). It was not raining on the days of the trial. It was not noticeably windy on the days of the trial. The partitions move on overhead tracks, running above the horse’s head in their blind spot (Hanggi and Ingersoll, 2012). A study by Valenchon et al. (2017) found that loud noises contribute to cause stress in horses. The way the trial horses perceived the sound from their environment may have contributed to their stress.

The model of horse walker used in this study has mesh screens to block the weather from coming into the horse walker. The screens also impair the peripheral view of the environment outside of the horse walker. Limiting the visual field of a prey species can cause added stress (Hanggi and Ingersoll, 2012).

Electrified partitions

The model of horse walker used had the ability for the metal bar partitions that move and follow the horses to be electrified, thus administering a shock to horses if they slowed down in the walker. The electricity was not turned on in the trial, but it has been previously used on some of the horses. Giese et al. (2014) found no significant difference in stress between horses on a walker with and without the use of electricity. Additionally, Glauser et al. (2015) found no increased stress response when electric fencing was turned on in paddocks. Therefore, a horse’s previous experience with electricity turned on whilst on the walker was not deemed to influence the results.

How can we educate people to use horse walkers in a more humane way?

Every horse should always be exercised with a companion on the walker. Ideally, the turnout groups should go onto the walker together. I would recommend that any new horse arriving on the yard should be trained to use the horse walker by following a rehabilitation programme. This is particularly important if they have prior experience of using a horse walker. These changes are recommended to improve the horse’s welfare (Waran and Randle, 2017).

Spending a longer time using a different method of initial training (e.g., systematic desensitisation instead of habituation) to the horse walker may have reduced overall stress and improved the horse’s welfare. Using positive methods of training has been proven to leave positive memories of that training with the horse (Sankey et al., 2010). The horses in this study could be slowly rehabilitated to the horse walker, and closely monitored throughout the rehabilitation period to ensure they are not fearful.

In addition to a rehabilitation period and systematic desensitization, I would recommend that the horse walker should not be used during adverse weather conditions to ensure that the noise is kept to a minimum on the walker, thus reducing overall stress. The electricity should be used with caution on horses that will not walk forwards, so long as they have been proven to not be stressed on the walker. All other horses should not need electric partitions.

Areas for further research

Building on this research, it would be interesting to measure whether the horses value a human companion as much as they value an equine companion to cope with a stressful situation. This type of companionship has been recognised in canines (Coppola et al., 2006).

The type of walker could have an effect on the levels of stress of the horses. A horse walker that has an open roof and open sides may be more or less stressful then the type used in this study. A horse walker that has a central pivot point rather than a partition track system could be more or less stressful. The stress would be dependent upon how the horses perceive that environment. It has been found that subordinate horses mimic the reactivity and emotionality of horses higher up in the herd hierarchy (Christensen et al., 2012).

Studying the dynamic of the horse dyad to find out which horse is more dominant could be worthwhile to see whether herd hierarchy (York and Schulte, 2014) has an effect on the horse’s stress level (Giles et al., 2015). This would indicate whether the subordinate horse is feeding off the reactions of the dominant horse, or vice versa.

Furthermore, the type of initial horse walker training (e.g., habituation or systematic desensitisation) may have an impact on stress.


It can be concluded that the horses found the horse walker highly stressful, and the stress may be reduced slightly if they are with a companion. Physiological and psychological displays of stress may be reduced when horses are exercised on a walker with a companion. Exercising and managing horses in a non-isolated manner may benefit the overall welfare of the horse by reducing stress levels and related behavioural tension. Taking the time to train a horse to a frightening stimulus may reduce stress and improve an equine’s welfare.


These images are stills from the video taken using iPhones mounted around the horse walker. I used the recordings to create an ethogram of behaviors indicative of stress in the exercising horses, and to measure the behavior score of each horse.

Behaviors seen: Straight lip line, drawn nostrils, mandibular tension, maxillary tension, orbital tension

Behaviors seen: Straight lip line, drawn nostrils, mandibular tension, maxillary tension, orbital tension

Behaviors seen: Orbital tension, licking, chewing, grinding of teeth

Behaviors seen: Orbital tension, licking, chewing, grinding of teeth










Behaviors seen: Tail clamped under the hind quarters, muscular tension accompanying tail clamping

Behaviors seen: Tail clamped under the hind quarters, muscular tension accompanying tail clamping

Behaviors seen: Defecation, swishing, mandibular tension, maxillary tension, straight lip line, orbital tension, ears changing direction

Behaviors seen: Defecation, swishing, mandibular tension, maxillary tension, straight lip line, orbital tension, ears changing direction









Behaviors seen: Mandibular tension, maxillary tension, head raised high above wither, drawn nostrils

Behaviors seen: Mandibular tension, maxillary tension, head raised high above wither, drawn nostrils

Behaviors seen: Head raised above withers, whites of eyes visible, mandibular tension, maxillary tension, straight lip line, drawn nostrils.

Behaviors seen: Head raised above withers, whites of eyes visible, mandibular tension, maxillary tension, straight lip line, drawn nostrils.











Bachman, A., Thompson, D., Walker, N. and Southerland, C. (2017) Hormonal and heart rate responses to a surprise stimulus and isolation stress in horses. Journal of Equine Veterinary Science. 52, pp.61.

Castaldo, R., Melillo, P., Bracale, U., Caserta, M., Triassi, M. and Pecchia, L. (2015) Acute mental stress assessment via short term HRV analysis in healthy adults: A systematic review with meta- analysis. Biomedical Signal Processing and Control. 18, pp.

Christensen, J., Søndergaard, E., Thodberg, K. and Halekoh, U. (2011) Effects of repeated regrouping on horse behaviour and injuries. Applied Animal Behaviour Science. 133(3-4), pp.199- 206.

Coppola, C., Grandin, T. and Enns, R. (2006) Human interaction and cortisol: Can human contact reduce stress for shelter dogs? Physiology & Behavior. 87(3), pp.537-541.

Dubout, P. (1969) The sound of rain on a steel roof. Journal of Sound and Vibration. 10(1), pp.144- 150.

Giese, C., Gerber, V., Howald, M., Bachmann, I. and Burger, D. (2014) Stressbelastung und Verhalten von Pferden in stromführenden gegenüber nicht stromführenden Führanlagen. Schweizer Archiv für Tierheilkunde. 156(4), pp.163-169.

Giles, S., Nicol, C., Harris, P. and Rands, S. (2015) Dominance rank is associated with body condition in outdoor-living domestic horses (Equus caballus). Applied Animal Behaviour Science. 166, pp.71-79.

Glauser, A., Burger, D., van Dorland, H., Gygax, L., Bachmann, I., Howald, M. and Bruckmaier, R. (2015) No increased stress response in horses on small and electrically fenced paddocks. Applied Animal Behaviour Science. 167, pp.27-34.

Hanggi, E. and Ingersoll, J. (2012) Lateral vision in horses: A behavioral investigation. Behavioural Processes. 91(1), pp.70-76.

Hartmann, E., Christensen, J. and Keeling, L. (2011) Training young horses to social separation: Effect of a companion horse on training efficiency. Equine Veterinary Journal. 43(5), pp.580-584.

Hemsworth, L., Jongman, E. and Coleman, G. (2015) Recreational horse welfare: The relationships between recreational horse owner attributes and recreational horse welfare. Applied Animal Behaviour Science. 165, pp.1-16.

Hockenhull, J. and Creighton, E. (2013) A brief note on the information-seeking behavior of UK leisure horse owners. Journal of Veterinary Behavior: Clinical Applications and Research. 8(2), pp.106-110.

Janczarek, I. and Kędzierski, W. (2011) Emotional Response to Novelty and to Expectation of Novelty in Young Race Horses. Journal of Equine Veterinary Science. 31(9), pp.549-554.

Kay, R. and Hall, C. (2009) The use of a mirror reduces isolation stress in horses being transported by trailer. Applied Animal Behaviour Science. 116(2-4), pp.237-243.

Mal, M., Friend, T., Lay, D., Vogelsang, S. and Jenkins, O. (1991) Physiological responses of mares to short term confinement and social isolation. Journal of Equine Veterinary Science. 11(2), pp.96-102.

Murphy, J. (2008) Innovative Use of an Automated Horse Walker When Breaking in Young Horses. Journal of Applied Animal Welfare Science. 11(3), pp.228-231.

Ruxton, G. and Colegrave, N. (2011) Experimental design for the life sciences. Oxford: Oxford University Press.

Sankey, C., Richard-Yris, M., Leroy, H., Henry, S. and Hausberger, M. (2010) Positive interactions lead to lasting positive memories in horses, Equus caballus. Animal Behaviour. 79(4), pp.869-875.

Valenchon, M., Lévy, F., Moussu, C. and Lansade, L. (2017) Stress affects instrumental learning based on positive or negative reinforcement in interaction with personality in domestic horses. PLoS ONE. 12(5), pp.e0170783.

Visser, E., Vanreenen, C., Vanderwerf, J., Schilder, M., Knaap, J., Barneveld, A. and Blokhuis, H. (2002) Heart rate and heart rate variability during a novel object test and a handling test in young horses. Physiology & Behavior. 76(2), pp.289-296.

Waran, N. and Randle, H. (2017) What we can measure, we can manage: The importance of using robust welfare indicators in Equitation Science. Applied Animal Behaviour Science. 190, pp.74-81.

Weinstein, A., Deuster, P. and Kop, W. (2007) Heart rate variability as a predictor of negative mood symptoms induced by exercise withdrawal. Medicine & Science in Sports & Exercise. 39(4), pp.735-741.

Yarnell, K., Hall, C., Royle, C. and Walker, S. (2015) Domesticated horses differ in their behavioural and physiological responses to isolated and housing. Physiology & Behavior. 143, pp.51-57.

York, C. and Schulte, B. (2014) The relationship of dominance, reproductive state and stress in female horses (Equus caballus). Behavioural Processes. 107, pp.15-21.

Young, T., Creighton, E., Smith, T. and Hosie, C. (2012) A novel scale of behavioural indicators of stress for use with domestic horses. Applied Animal Behaviour Science. 140(1-2), pp.33-43.