Current exercise-related practices

Study results indicate that respondents’ horses are typically managed conventionally, via a combination of individual stabling and grass pasture turnout. Though all horses had access to turnout over summer months, with most living out full-time, many did receive stabling for at least a portion of each day, and over winter months, most horses were stabled regularly, with a few receiving no turnout. Respondents reported that grass fields, with relatively low environmental complexity, were the prevailing environment for their horses when not stabled. The key implications of these outcomes are two-fold. Importantly, they suggest that respondent horses are representative of the wider UK domestic horse population, as prevailing management strategies are consistent with previous findings (Hockenhull & Creighton Reference Hockenhull and Creighton2014). Additionally, these results imply that a substantial number of domestic horses are kept in obesogenic environments, as grass pastures can facilitate the development of obesity (Giles et al. Reference Giles, Rands, Nicol and Harris2014) and comorbidities including laminitis (Geor Reference Geor2009). As was confirmed in this study, typical turnout areas are simplistic grass fields (Rioja-Lang et al. Reference Rioja-Lang, Connor, Bacon and Dwyer2020). Therefore, while access to these open areas facilitates SDE through free movement, the grass-based composition of these turnout areas may afford horses the opportunity to easily and substantially exceed caloric needs, as grasses are nutrient-rich and obtainable with very little energy expenditure (i.e. need or motivation to move, despite ample opportunity), particularly over the summer growing season (Longland & Byrd Reference Longland and Byrd2006; Wyse et al. Reference Wyse, McNie, Tannahil, Murray and Love2008). This may negate some of the health benefits associated with SDE, including increased energy expenditure, due to overconsumption. Many horses received daily stabling in summer months, which is often presumed to prevent overconsumption of pasture grasses. However, previous research indicates that horses simply increase intake rates during allocated grazing times, nullifying the suggested benefits of such management (Glunk et al. Reference Glunk, Pratt-Phillips and Siciliano2013). Additionally, stabling restricts movement, further limiting opportunities for energy expenditure via SDE (Geor Reference Geor2010). Overall, therefore, the typical management of respondents’ horses may contribute to excess weight gain and subsequent physical health issues, with negative consequences for welfare.

While most respondents’ horses received HLE, associated energy expenditures would be unlikely to neutralise the obesity risk promoted by grass-rich environments. Though a small proportion of horses received only unridden HLE, most were ridden. However, the most popular types of HLE (flatwork, hacking and polework) are low-intensity activities, with lesser energetic demands than activities such as jumping and interval training (Clayton Reference Clayton1991). This finding, particularly when combined with the modest overall levels of HLE received by most respondents’ horses, suggests that workloads are generally submaximal, limiting energy expenditures that could otherwise contribute to weight management. This inference is corroborated by Giles et al. (Reference Giles, Rands, Nicol and Harris2014), who determined that low-intensity HLE was insufficient at mitigating obesity risk in their sample of UK leisure horses. Moreover, horse owners frequently overestimate their horses’ workloads (Hale et al. Reference Hale, Hemmings and Randle2016). It is likely that actual exercise durations, frequencies, and intensities were lower than what was described, which is a limitation of this study. The findings related to HLE should therefore be regarded as a high estimation, further suggesting that HLE is unlikely to fully mitigate obesity risk and may consequentially be inadequate at safeguarding welfare. Collectively, findings regarding respondents’ current practices suggest that prevailing management and exercise regimes likely promote excess body condition in their horses, and that both SDE and HLE in the ways they are currently being provided are inadequate at mitigating obesity risk among horses.

Nearly 40% of respondents reported that, on average, over the past year their horse was overweight or obese. These results echo those of Wyse et al. (Reference Wyse, McNie, Tannahil, Murray and Love2008) who reported that 45% of surveyed riding horses in Scotland were “fat” or “very fat”. Additionally, given the propensity of owners to underestimate their horses’ body condition, and their struggles to recognise obesity (Furtado et al. Reference Furtado, Perkins, Pinchbeck, McGowan, Watkins and Christley2020), excess weight among respondents’ horses is likely more extensive than that reported. Only 12.6% of respondents reported that their horse had been diagnosed, treated, or managed for excess weight in the past year, suggesting most horses who are owner-identified as being overweight are not being managed accordingly, which is a welfare concern. This may indicate that respondents believe excess weight to be an inherent characteristic of their horses, without recognising it as a health problem requiring treatment. This would be consistent with Furtado et al.’s (Reference Furtado, Perkins, Pinchbeck, McGowan, Watkins and Christley2020) finding that owners struggle to differentiate between obesity and the “normal shape” of their horse. Alternatively, this result may suggest that while owners realise their horses are overweight, they feel unable to intervene effectively, potentially indicating a lack of opportunity.

Regression results indicated that breed type was the only significant predictor of a horse being owner-recognised as overweight, with native types being more vulnerable than either sports horses or hot-bloods. HLE parameters (e.g. if a horse was ridden, total amount of HLE) and opportunities for SDE measured via turnout time were not correlated with being overweight. These findings are in accordance with those of Giles et al. (Reference Giles, Rands, Nicol and Harris2014), who also found breed to be the strongest predictor of obesity, such that native horses/ponies were more vulnerable to obesity than other breeds. This further supports the point that the grass-based environments that facilitate SDE likely also promote overconsumption, with subsequent consequences for welfare.

While obesity is a standalone health issue, it can also contribute towards laminitis, the painful and debilitating condition involving inflammation of the hoof lamellae (Bailey & Bamford Reference Bailey and Bamford2013). When obesity and laminitis combine with hyperinsulinaemia, the condition is collectively referred to as Equine Metabolic Syndrome (EMS) (Agne Reference Agne2010). Increased exercise may contribute to prevention or management of laminitis and EMS (Geor Reference Geor2010). This study found that ridden horses were less likely to have been diagnosed, treated, or managed for either laminitis or EMS than unridden horses. This may suggest that HLE aids in their prevention and also reinforces that horses afflicted with these conditions are often unable to be ridden due to lameness (Menzies-Gow et al. Reference Menzies-Gow, Stevens, Barr, Camm, Pfeiffer and Marr2010). Horses with either/both of these conditions were more likely to receive alternative turnout (i.e. not grass fields) than horses without. While one might therefore presume that these alternative systems caused the conditions, the more likely scenario is that these systems are used to promote SDE to manage laminitis and EMS, in accordance with the findings of a report from the University of Liverpool, UK (Furtado et al. Reference Furtado, King and Pinchbeck2021). The premise of such systems is to increase SDE while facilitating species-normal socialisation and foraging behaviours (Furtado et al. Reference Furtado, King and Pinchbeck2021). This is particularly true of ‘Paddock Paradise’ track systems, which are low- or no-grass tracks often built around the perimeter of a field, with resources distributed throughout to encourage movement (Jackson Reference Jackson2006). Owners are typically prompted to implement such systems once serious health conditions, including laminitis and EMS, already affect their horses (Furtado et al. Reference Furtado, King and Pinchbeck2021). This likely explains why overweight horses were not more likely to receive alternative turnout than their fitter conspecifics: owners often wait until obesity triggers further complications before acknowledging the problem (Furtado et al. Reference Furtado, Perkins, Pinchbeck, McGowan, Watkins and Christley2020).

Incidences of laminitis were negatively correlated with total summer HLE, whereas occurrences of EMS were positively correlated with total winter HLE. It is likely that increased HLE over summer helped mitigate laminitis risk in the season when it can easily occur due to high growth rates of grasses containing laminitis-provoking simple sugars (Longland & Byrd Reference Longland and Byrd2006). Additionally, horses suffering from laminitis would presumably not be able to be ridden. The association between EMS and HLE in winter suggests that owners of horses afflicted with EMS are motivated to provide substantial HLE during a time when respondents collectively indicated exercise levels are typically lower than in summer. Interestingly, breed type did not predict either laminitis or EMS among respondents’ horses. This suggests that all horses have the potential to develop these conditions. It is important to raise owner awareness regarding the vulnerability of all horses to these diseases, regardless of breed type, and to promote practical, accessible ways to mitigate risks, while upholding equid welfare.

Though physical health indicators were clearly correlated with exercise-related variables, the impacts of exercise on behaviour were less conclusive. Breed type was a significant predictor of both behavioural welfare indicators: presence of locomotor stereotypies, and aggression towards humans. Natives and sports horses were less likely to display locomotor stereotypies than hot-bloods, and both hot-bloods and sports horses were more likely to show aggression towards humans than natives. This is ostensibly less about differential welfare than it is about temperament as hot-bloods, whose bloodlines are the foundation of sports horses, were selectively bred for heightened reactivity, predisposing them to displays of unwanted behaviours, whereas coldblooded breeds, including natives, are comparatively stoic (McGreevy Reference McGreevy2012).

Controlling for breed effects, horses on livery yards were more likely to display locomotor stereotypies than those kept elsewhere. Locomotor stereotypies are thought to indicate frustration at the inability to access desired resources (e.g. food) due to confinement (Hothersall & Casey Reference Hothersall and Casey2012). Most horses kept at livery yards were on ‘DIY’ livery, meaning owners are responsible for all care, including feeding and turnout from stables to fields. Consequently, horses often have asynchronous routines and see their conspecifics fed and/or turned out without receiving these desirable resources themselves. Cooper et al. (Reference Cooper, McCall, Johnson and Davidson2005) reported incidences of locomotor stereotypies increased when horses viewed neighbouring horses being fed but did not receive food themselves. Unpredictable and asynchronous routines that frustrate motivations to acquire resources therefore likely caused increased locomotor stereotypy presence among horses kept at livery yards.

Increased number of conspecifics in turnout areas was negatively correlated with presence of locomotor stereotypies. Horses are gregarious, so providing appropriate companions is an integral component of upholding welfare (Hartmann et al. Reference Hartmann, Søndergaard and Keeling2012; Ruet et al. Reference Ruet, Lemarchand, Parias, Mach, Moisan, Foury, Briant and Lansade2019). Increased access to, and choice of, conspecifics therefore likely enhanced opportunities for horses to engage in meaningful SDE through socialisation. These results also underscore agency as important contributors towards positive welfare (Littlewood et al. Reference Littlewood, Heslop and Cobb2023). Keeping animals in appropriate groups affords them the freedom to choose with whom to interact, adding a level of social complexity and active engagement that is otherwise inaccessible (Špinka Reference Špinka2019). Such interactions are emphasised in the Five Domains model of welfare both as a standalone component of welfare, and as a crucial contributor to overall mental state (Mellor et al. Reference Mellor, Beausoleil, Littlewood, McLean, McGreevy, Jones and Wilkins2020). Other movement-related variables, including total amount of HLE and environmental complexity, did not predict presence of locomotor stereotypies. Collectively, these results indicate that there is no substitute for socialisation.

Interestingly, ridden horses were more likely to display aggression towards humans than unridden horses. This may simply be because ridden horses are handled more often, creating greater opportunity for aggression. However, riding may involve applications of inappropriate training methods, use of ill-fitting tack, or may exacerbate any musculoskeletal or other pain-related issues (Waran et al. Reference Waran, McGreevy, Casey and Waran2006; Greve & Dyson Reference Greve and Dyson2014, Reference Greve and Dyson2015; Dyson & Van Dijk Reference Dyson and Van Dijk2020). Aversive training can provoke displays of conflict behaviours, including biting and kicking (McLean & Christensen Reference McLean and Christensen2017). A preference test conducted by König von Borstel and Keil (Reference Keil2012) indicated horses prefer to access feed and/or conspecifics than to be ridden. Lee et al. (Reference Lee, Floyd, Erb and Houpt2011) found that most horses choose to remain stabled rather than running on a treadmill but prefer being turned out into a paddock over remaining stabled, particularly if being turned out with another horse. Together, these findings suggest that HLE may not be the average horse’s preferred type of movement, which has important ramifications for welfare.

Overall, behavioural analyses stress the importance of social opportunities for welfare and highlight the effect of breed type on displays of unwanted behaviours. However, they do not provide robust evidence to either support or refute the importance of exercise on welfare. Accurate assessment of equine affective state remains a challenge among welfare scientists (Hall et al. Reference Hall, Randle, Pearson, Preshaw and Waran2018) and horse owners (Rioja-Lang et al. Reference Rioja-Lang, Connor, Bacon and Dwyer2020). This study relied upon simplistic owner-reported presence of certain behaviours and consequently was unlikely to provide a robust psychological welfare assessment. Comparatively, Qualitative Behaviour Assessment (QBA) has been validated for use in horses (Hintze et al. Reference Hintze, Murphy, Bachmann, Wemelsfelder and Würbel2017) and donkeys (Equus asinus; Minero et al. Reference Minero, Dalla Costa, Dai, Murray, Canali and Wemelsfelder2016), and should be considered in future experimental protocols to assess the emotional well-being of horses receiving variable quantities and forms of exercise.

Owner perceptions

Lack of owner knowledge of equine needs is commonly cited as the root of mismanagement, with subsequent impacts on welfare (Horseman et al. Reference Horseman, Buller, Mullan and Whay2016; Rioja-Lang et al. Reference Rioja-Lang, Connor, Bacon and Dwyer2020). This study is among the first to take a comparatively holistic approach to understanding horse owner behaviour, using Michie et al.’s (Reference Michie, van Stralen and West2011) COM-B model of human behaviour. This model describes three main drivers of behaviour: capability, opportunity, and motivation. Capability includes psychological ability (i.e. knowledge) and physical capacity (e.g. skills) to produce a given behaviour. Opportunity refers to the extrinsic physical resources that facilitate behaviours, and social opportunities (e.g. cultural norms, peer pressures) that prompt actions or beliefs. Reflective motivation involves the plans and introspection that shape behaviour while automatic motivation describes subconscious impulses and ingrained responses. Therefore, rather than stemming solely from knowledge, human behaviours, including those relating to equine management, are more accurately the result of complex, interwoven processes that must be identified and addressed to first understand, and then change, current practice.

The study results, when framed using the COM-B model, confirm that knowledge is not the only barrier to adopting new management strategies that could potentially improve domestic horse welfare. Respondents were more likely to be satisfied with their control over turnout conditions (versus preferring more control) as turnout duration increased and as forage became increasingly available over winter, suggesting acknowledgement of species-normal needs for unstructured movement and foraging opportunity (McGreevy Reference McGreevy2012). Respondent satisfaction was also more likely with increasing habitat complexity. This may show awareness of environmental enrichment as a means to improve welfare by facilitating movement and exploration in confinement (Mason et al. Reference Mason, Clubb, Latham and Vickery2007). As the number of unwanted behaviours displayed by a horse increased, that horse’s carer became less likely to be satisfied with their control over turnout conditions, potentially demonstrating knowledge that behavioural issues communicate suboptimal welfare (Normando et al. Reference Normando, Meers, Ellery, Faustini and Ödberg2011). Finally, respondents were more likely to be satisfied with their level of control as the number of horses sharing winter turnouts decreased. Antagonistic encounters tend to occur when forage is limited, as is likely the case in grass fields over winter, but not when sufficient forage availability prevents competition (Benhajali et al. Reference Benhajali, Richard-Yris, Ezzaouia, Charfi and Hausberger2009). Respondents seemingly acknowledge the role of limited resources in provoking competition in a species that otherwise coexists peacefully. Overall, therefore, responses were broadly in line with those that would be expected from knowledgeable horse people, potentially reflecting the high levels of experience across participants, and suggesting that a lack of knowledge is unlikely to be the sole driver of management decisions that can lead to welfare issues in horses.

Results indicate that keeping horses at a livery yard is a substantial barrier to promoting SDE, due to inopportunity. When given a choice of three equine weight-loss strategies (increasing HLE, restricting SDE through stabling and muzzling during turnout, or altering field set-ups to promote SDE), respondents with horses at livery and those with horses on their own premises both believed altered field set-ups that promote SDE were the most welfare-friendly option, as well as their ideal solution. However, those on livery felt this was less realistic than either muzzling/stabling or increasing HLE, whereas those with their own properties thought altering the field set-up was the most viable option. This variance is likely due to a lack of social and physical opportunity at livery yards, whereby control over land management, rules and sociocultural norms act as barriers (Furtado et al. Reference Furtado, Perkins, Pinchbeck, McGowan, Watkins and Christley2020). This is an important result as it highlights the role of livery yards, and yard owners in cultivating opportunity for owners to manage their horses in ways that could be beneficial for welfare.

Interestingly, when asked which solution would be the most effective, chosen strategies mirrored those selected as most realistic within each group of respondents. This suggests horse owners may undervalue practices that they perceive to be unobtainable. This may impact uptake of welfare-enhancing recommendations throughout multiple facets of the equine industry and should therefore be the subject of future research. At present, it is unknown which of the presented weight-control strategies is actually the most effective. Muzzling has previously been shown to substantially reduce forage intake rates in ponies (Longland et al. Reference Longland, Barfoot and Harris2011), though muzzles may need to be worn 24 h per day to avoid compensatory eating (Davis et al. Reference Davis, Iwaniuk, Dennis, Harris and Burk2020). It is also uncertain if horses display the same rebound effect as their smaller conspecifics. The benefits of HLE for physical health are discussed above, as is the evidence that suggests respondents are likely correct in believing increased HLE is most probably not the solution most horses would choose. Alternative management, particularly track systems, are believed to facilitate weight loss (Furtado et al. Reference Furtado, King and Pinchbeck2021) and their theoretical foundation is rooted in upholding welfare through facilitation of movement and species-normal behaviours (Jackson Reference Jackson2006). However, further research is necessary to determine if track systems increase energy expenditure in practice. Hampson et al.’s (Reference Hampson, Morton, Mills, Trotter, Lamb and Pollitt2010) preliminary findings were that racetrack-style turnout did not increase movement. They also reported that complicated spiral enclosures decreased activity compared to typical rectangular paddocks. This may be because horses are adapted to open grasslands (MacFadden Reference MacFadden and MacFadden1992), and therefore potentially lack cognitive adaptations to navigate complex maze-like environments. Practical applications of existing track systems are diverse in terms of land use, surfaces, enrichment, and equid density (Furtado et al. Reference Furtado, King and Pinchbeck2021). They therefore likely vary in success at increasing movement and exploratory behaviours. In comparison with strip-grazing, however, track systems have been shown to promote increased SDE and decreased agonistic behaviours among ponies, thus being more conducive to good welfare (Kirton et al. Reference Kirton, Sandford, Raffan, Hallsworth, Burman and Morgan2024). This is an encouraging finding, though further research is needed to better understand if, and how, track systems should be implemented to promote physical and emotional equine well-being.

Though livery yards appear to be a barrier to providing increased SDE, respondents with horses on livery were more likely to be satisfied with the amount of HLE their horse received than those with horses kept elsewhere. Still, 90% of respondents indicated that barriers limit their provision of HLE. Time, weather, and riding facilities were the primary barriers, suggesting an overall lack of physical opportunity to provide sufficient HLE. As most respondents were leisure owners, they presumably have numerous other demands on their time, including families and careers, which are unlikely to be modifiable in order to increase their horse’s HLE. When identifying the main reason for relocating their horse, “improved living conditions” was chosen over four times more than “improved riding facilities”, presumably reflecting respondent priorities. Collectively, these findings indicate that, although increased HLE is suggested to improve horse health (e.g. Geor Reference Geor2010), this may not be a viable solution for many horse owners due to barriers that are beyond their control. Therefore, finding practical, effective ways to increase SDE is potentially a more useful path to pursue in both research and practical applications.

Animal welfare implications

Collectively, analyses on welfare indicators suggested that increased exercise, particularly via SDE rather than HLE, is worthy of further investigation as a means to improve welfare. Our results provided preliminary indications that increased exercise can lesson risk of health conditions and show the importance of management systems that give horses opportunities to display species-normal behaviours, including socialisation. When the results of this study were contextualised within the COM-B model of human behaviour, horse owners appeared to be most limited in their provision of increased movement by opportunity, rather than knowledge or motivation. It is therefore important to acknowledge that upholding animal welfare is not only the responsibility of the horse’s direct carer, but also other potential change agents, such as livery yard owners, who are responsible for land-management decisions. Future studies directed at livery yard owners, in particular, would be useful to identify their own determinants of behaviour, to build a robust picture of the management factors contributing to horse welfare. Our results justify further research to explore ways of increasing SDE while facilitating species-normal behaviours such as foraging and socialising opportunities, paying particular attention to track systems as a potentially viable alternative management strategy.