The gastrointestinal tract of performance horses: medical, nutritional and surgical considerations Andy E. Durham BSc.BVSc.CertEP.DEIM.DipECEIM.MRCVS., The Liphook Equine Hospital, Forest Mere, Liphook, Hampshire, GU30 7EH, UK. andy@TheLEH.co.uk Nathaniel A. White DVM, MS, DACVS, Marion duPont Scott Equine Medical Center, VMRCVM-Virginia Tech, Leesburg, Virginia 20177, USA. firstname.lastname@example.org
The equine digestive tract evolved anatomically and physiologically in an animal that led a relatively sedentary lifestyle continually browsing for fibre-rich and low-starch food (Houpt 1990). Evolutionary survival pressures in a predated species inevitably led to a high capacity for sudden and high intensity exercise in response to perceived threats. Presumably such extreme exertion was relatively infrequent in the life of the ancestral feral equid as the dietary energy derived from available nutrient sources (grasses, rushes, sedges and perhaps occasional cereals) would probably be inadequate for daily high intensity or prolonged exercise. In a contrasting lifestyle, to facilitate and enhance performance of their intended sport, the almost daily exertional demands of the modern elite domesticated equid require a considerably increased quality and quantity of feed intake and also different feeding patterns and ingestive behaviours in comparison to the evolved physiologic template.
The National Research Council committee on nutrient requirements of horses recently stated that “The goal of feeding management is to efficiently supply dietary ingredients in amounts that will meet the horse’s nutrient needs, while still retaining the horse’s normal feeding behaviour” (National Research Council (NRC) 2007b). If “normal feeding behaviour” is taken to mean that to which horses have become adapted over 54 million years of evolution prior to their recent domestication, then this appears to be unachievable within the inevitable constraints of modern competition horse dietary management with possible adverse consequences on their gastrointestinal (GI) health. Increased exertional activity per se may have additional undesirable consequences on the GI tract.
In accordance with this dietary and behavioural divergence of the modern equine athlete from its feral ancestors, several studies have suggested that competition horses are at increased risk of GI disease (Archer and Proudman 2006; Durham 2007). Certain breeds such as the Thoroughbred and Arabian, developed for athletic pursuits, have been found to have inherently increased risk of GI disease (Cohen et al 1995b, 1999; Cohen and Peloso 1996; Reeves et al 1996; Tinker et al 1997; Traub-Dargatz et al 2001). Eventing, race-training and showing activities have been found to be associated with more than double the risk of colic compared to non-competing horses (Kaneene et al 1997; Tinker et al 1997). However, other large epidemiologic studies have found no association between use of horse and colic risk (Traub-Dargatz et al 2001) and have even suggested that racehorses may have a lower colic risk than less physically active horses (Kaneene et al 1997; Cohen et al 1995b). Dehydration and loss of electrolytes from extended periods of exercise and transport have been associated with GI disturbances leading to colic, although only a few studies have established a significant association (Foreman 1998; McCutcheon 1998; Hillyer 2002). Prevalence studies of gastric ulcers reveal a strong association with exertional activities such as racing in Thoroughbreds and Standardbreds (Hammond et al 1986; Vatistas et al 1994; Rabuffo et al 2002; Dionne et al 2003; Bell et al 2007a), endurance races (Nieto et al 2004) western performance events (Bertone 2000) and showing (McClure et al 1999).
This review considers the interactions between the equine GI tract and performance primarily in terms of dietary and exertional effects on GI health and also evidence linking these effects to performance. Although certain aspects of nutrition are considered, the reader is referred elsewhere within these proceedings (Benoit and Harris) for more specific discussion of nutrition and performance.
Dietary overview in the competition horse Horses participating in different competitive pursuits frequently have markedly different dietary management and therefore subsequently differing health implications of the diets and the dietary interactions with training and competitive exertion. Diet quality may vary from the relatively high forage, low cereal diet of the endurance horse to the low forage, high cereal diet of the flat-racing Thoroughbred with intermediate gradations of dietary quality in other performance disciplines (Bishop 2005; Geor and Harris 2005; Richards et al 2006). Access to grazing varies considerably with generally greater access allowed for the endurance horse than the race horse (Geor and Harris 2005; Bell et al 2007a). Fat-supplemented diets have become widely recommended, especially in endurance disciplines, in order to increase non-starch derived dietary energy with suggested relative benefits to GI health (Geor and Harris 2005). Forage feeding immediately prior to training or racing in Thoroughbreds tends to be limited or absent resulting in exertion with a relatively empty stomach and colon, the former containing primarily gastric fluid and cereal. As forage intake will usually be accompanied by 3 to 4 times its weight in water intake (NRC 2007a), the primary aim of forage restriction in racing Thoroughbreds is to reduce body mass and energy expenditure during exercise (Rice et al 2001; Ellis et al 2002). Horses fed 100% forage vs. hay plus concentrate have more weight and more heat production suggesting inefficiency in horses exercising on forage only diets (Kronfeld et al 1994; Kronfeld 1996). Feeding grain immediately prior to exercise decreases free fatty acid availability and increased glucose disappearance during exercise. Feeding hay with grain or free choice the night before exercise decreased plasma volume and increased lactate and heart rate during exercise on a high-speed treadmill. Feeding only forage did not adversely affect performance but large amounts just prior to exercise could affect performance (Pagan and Harris 1999). In contrast to racing Thoroughbreds, endurance horses are likely to have a GI tract filled with forage-based digesta at the time of competition, a status more similar to the likely evolutionary adaptation of the equine GI tract and providing an effective reservoir of nutrients, electrolytes and water. Fat adaptation further decreases heat production thereby reducing body temperature and needed daily water intake.
Dietary influences on gastric ulceration
Gastric ulceration is a disease of high prevalence in performance horses (Bell et al 2007b) and low-forage, high cereal diets have long been recognised in association with increased prevalence of gastric ulcers in racehorses (Hammond et al 1986). The upper squamous mucosa and lower glandular mucosa of the equine stomach have markedly different properties relevant to the interaction of diet and exercise with gastric disease (Merritt 1999). Consistent with evolutionary development as a “continual browser” or “trickle-feeder”, the equine stomach secretes hydrochloric acid (HCl) almost continually leading to normal prolonged exposure of glandular mucosa to acid (Campbell-Thompson and Merritt 1990; Murray 1997). Unsurprisingly development of a strong antacid defence can be demonstrated in this ventral area of the stomach consisting of prostaglandin-dependent secretion of bicarbonate-rich mucus, good mucosal blood flow (to remove any penetrating protons and maintain potential for a high metabolic rate) and rapid restitution following erosion or ulceration of the epithelium (Murray 1999). Observations of the pH and gross appearance of fibrous ingesta within the forage-fed equine stomach suggest that the gastric squamous mucosa did not evolve in frequent contact with low pH gastric fluid and has not suffered strong evolutionary pressure to develop robust antacid defence unlike its subjacent neighbour (Ethell et al 2000). The stomach content of a horse fed forage/grass ad libitum tends to be a solid but stratified mass of increasing pH from highly acidic ventrally adjacent to the acid-secreting glandular mucosa, to neutral or alkaline pH at the most dorsal aspect of the mass of ingesta exposed to salivary bicarbonate entering the cardia within the squamous mucosal area ([Figure 1] Murray and Grodinsky 1989; Merritt 2003). Forage filling the stomach serves to limit mixing and movement of acidic fluid within the stomach and also incurs a considerably increased secretion of salivary buffer than does a cereal diet (Meyer et al 1985). High-cereal (and low-forage) consumption may promote fluidity, mixing and agitation of gastric contents by reducing intragastric dry matter ([Figure 2] Argenzio 1999) and therefore increase the potential for contact of the poorly defended squamous mucosa with low pH gastric fluid. Additionally high cereal diets may stimulate a more prolonged gastrin secretory response (Smyth et al 1989; Sandin et al 1998) further increasing the corrosive properties of gastric fluid.
The harmful effect of HCl may be compounded by the presence of volatile fatty acids (VFAs) in the gastric fluid. VFAs such as acetic, propionic, butyric, valeric and isovaleric acids arise from fermentation of food (especially starches) within the stomach. Furthermore, high-starch meals tend to reduce gastric emptying rates (Metayer et al 2004) and therefore further promote fermentative production of VFAs (Figure 2). The pKa of most VFAs found in the equine stomach is between 4 and 5 and thus they will be primarily in their non-ionised form in low pH gastric fluid. Diffusion of VFAs across squamous mucosal cell membranes is facilitated by their non-ionised, hydrophobic status and subsequent dissociation of their hydrogen ion within the nearly neutral pH of the cytosol may cause harmful intracellular acidification and cell death (Nadeau et al 2003a,b; Andrews et al 2006). Alfalfa feeding may offer a protective effect by buffering gastric fluid and reducing mucosal uptake of VFAs in comparison to grass hays (Nadeau et al 2000; Lybbert et al 2007).
Consistent with an evolutionary adaptation to continual feed intake, feed deprivation and intermittent feeding programmes have been shown to be an effective means of squamous mucosal ulcerogenesis in experimental studies in horses (Murray and Eichorn 1996). When the equine stomach is continually filled with high-fibre material there are beneficial consequences beyond the buffering and stabilisation effect of the solid mass of fibrous ingesta described above. Additionally, reflux and dispersal of pancreatic and biliary secretions within the stomach is physically limited in comparison to a forage-restricted diet when the stomach is largely devoid of solidly structured matter. As the equine pylorus offers little resistance to retrograde flow, frequent boluses of pancreatic and biliary secretion will enter the stomach and mix freely in an empty or fluid-filled stomach ([Figure 2] Merritt 1999). Although there may be significant buffering of gastric acid by pancreatic bicarbonate (Merritt 1999), the presence of bile acids within gastric fluid may be harmful to the mucosa and this is consistent with the relatively high prevalence of ulcers in the pyloric area in competition horses (Murray et al 2001; Begg and O’Sullivan 2003; Bell et al 2007a). As described above for VFAs, bile acids in their non-ionised form within the low pH gastric fluid may diffuse into mucosal cells leading to cell death and necrosis (Argenzio 1999; Berschneider et al 1999; Murray 1999) although this is not a theory supported by all studies (Widenhouse et al 2002). Bile acids might further disrupt the protective phospholipid barrier covering the squamous mucosal surface (Geor 2000). Dietary oils have been proposed as a means of defending against gastric ulceration. Possible beneficial mechanisms of action might include provision of substrate for prostaglandin synthesis (likely to primarily aid gastric glandular mucosal defence) or possibly by binding free bile acids within the gastric fluid. However, studies have not consistently found dietary oils to reduce gastric ulcers in horses (Cargile et al 2004; Frank et al 2005).
Hypertonic electrolyte solutions and pastes are sometimes administered to horses with the intention of improving performance, hydration and electrolyte status in endurance competitions and this has also been reported as a more surreptitious practice in Thoroughbred and Standardbred racing (Rose and Lloyd 1992; Auer et al 1993; Nyman et al 1996). This treatment may be significantly harmful to the gastric squamous mucosa and has been associated with significant squamous ulcerogenesis (Holbrook et al 2005).
Dietary influences on intestinal digestion and fermentation The horse’s GI tract has evolved to function for a grazing animal that has frequent periods of forage intake interspersed with periods of rest and movement to new food sources. The small intestine serves for energy absorption including sugars, starch, protein and fat. The large intestine or hindgut has adapted for microbial digestion of starch and fibre. Because of the likely inefficient digestion of starch in the horse’s small intestine, when grain is fed rapidly fermentable starch is passed into the large intestine. Rapid fermentation of starch in the large intestine (caecum and large colon) alters volatile fatty acid ratios (increased proprionate and decreased acetate) and increases lactate production (Geor and Harris 2007). A decrease in butyrate transport in the mucosa is altered as well as the increased production of vasoactive substances such as monoamines, endotoxin and exotoxins (Bailey et al. 2002; Bailey et. al. 2003; Geor and Harris 2007). The decreased intake of slow fermenting carbohydrates (cellulose, hemicellulose, ligno-cellulose) decreases the water content in the large colon potentially affecting energy production (decreased acetate), water reserve in the colon and weight. Less fibre with increased starch in the diet decreases water in the colon ingesta and increases gas production (Lopes et al 2004).
Alterations in diet to provide needed energy for performance is associated with an increased risk of colic due to higher levels of cereal or concentrate feeding (Tinker et al 1997, Hudson et al 2001; Cohen et al 2006). Incomplete pre-caecal digestion of starch has long been recognised for its potential adverse effect on large bowel health and function (Goodson et al 1988; Clarke et al 1990; de Fombelle et al 2001; Drogoul et al 2001; Julliand et al 2001; Hussein et al 2004; Lopes et al 2004) and levels as low as 1g starch per kg bodyweight per meal (eg. 1 kg concentrate feed for a Thoroughbred) could overwhelm the limited digestive capacity of the equine small intestine (Potter et al 1992) especially if starch sources of low digestibility are fed (Cuddeford 2000; Hintz 2000; Hussein et al 2004). Richards et al (2006) reported a mean grain intake of 7.3 kg/day (range 3.8-13.2 kg/day) for racing Thoroughbreds in Australia and 27% of horses studied had faecal pH <6.2 that is likely to be associated with an adverse colonic environment for mucosal barrier function and survival of fibre-fermenting bacterial species. The digestibility of fibre, the nutrients for which the equine digestive tract is apparently best adapted, is significantly reduced by cereal feeding and this effect is further exacerbated by exercise (Pagan et al 1998). The adaptation to fat digestibility appears to be a way to avoid high starch intake while maintaining adequate energy intake and glucose metabolism during exercise (Kronfeld et al 1996; Treiber et al 2006). To date there does not appear to be increased risk of colic associated with fat in the form of vegetable oils although some studies have found significantly reduced cellulolytic capacity of the equine hindgut with high fat diets (Jansen et al 2002), and the full effect of fat added to the diet in relation to performance is not yet known (Zeyner et all 2002).
General relationship between exercise and the GI tract Exercise in humans and horses is associated with increases in several hormones that may affect GI secretion, motility, perfusion, absorption and permeability including catecholamines, cholecystokinin, cortisol, dopamine, endorphins, gastrin, glucagon, glucagon-like peptide 1, motilin, neurokinin A, pancreastatin, pancreatic polypeptide, peptide histidine isoleucine, peptide YY, prostaglandins, secretin, somatostatin, and vasoactive intestinal polypeptide (Demers et al 1981; Hall et al 1982; Woie et al 1986; Brouns and Beckers 1993; O’Connor et al 1995; Nagata et al 1999). Exercise may be beneficial or harmful to GI structure and function depending on the nature and intensity of the exercise and the status of the GI tract at the time of exertion. In general mild to moderate aerobic exercise has been found to have significant beneficial effects on human GI function whereas intense or prolonged endurance exercise is frequently harmful (Moses 1990; Brouns and Beckers 1993; Peters et al 2001; Simrén 2002; Bi and Triadafilopoulos 2003). Physical activity in humans has been found to have a marked beneficial effect on the risk of colon cancer (Thune and Furberg 2001) and also on other intestinal complaints such as diverticular disease (Aldoori et al 1995), constipation (de Schryver et al 2005) and inflammatory bowel disease (Klein et al 1998). In marked contrast however, public marathon races have been described as a “preplanned mass casualty incident” with acute GI disease being one of the most common medical complaints amongst human competitors (Sanchez et al 2006). Between 30 and 81% of human marathon runners describe unpleasant GI symptoms during a race (Sullivan and Wong 1992; Lucas and Schroy 1998; Gil et al 1998; Sanchez et al 2006) and this is exacerbated by associated dehydration and hypovolaemia (Rehrer et al 1990; Brouns and Beckers 1993). Between 8 and 85% of human distance runners have positive faecal occult blood after a race (Choi et al 2001; Sanchez et al 2006) and also faecal 1 antitrypsin and lysozyme concentrations are increased following prolonged exercise in humans and dogs indicating mucosal inflammation and damage (Peters et al 2000a; Davis et al 2006). Voluntary exercise in rats is associated with caecal enlargement and beneficial effects on its bacterial and VFA content in comparison to sedentary controls (Matsumoto et al 2008). In contrast, intense exercise appears to be rapidly and significantly harmful to the structure and function of the murine intestine with mucosal necrosis, damage to muscular layers and inhibition of motility all being reported (de Lira et al 2008; Rosa et al 2008).
There are few studies of the effects of exercise on GI disease in horses although the majority appear to suggest a deleterious effect of exercise on GI health. As discussed above several, but not all, studies have found a higher incidence of colic in horses that were actively training and competing. Hillyer et al (2001) found that the months of peak incidence of colic in racing Thoroughbreds (flat and National Hunt) coincided with the timing of peak exertional activity although whether this effect was directly associated with exercise or perhaps confounded by diet or other factors was not examined. In a multivariate analysis where confounding factors were controlled, Cohen et al (1999) found that horses exercised at least once each week had a significantly increased risk of colic when compared with horses living at pasture (OR 1.6; 95% CI 1.2-2.2; P=0.003). Similarly training and competition is well recognised to be associated with development of gastric ulcers in horses (Hammond et al 1986; Vatistas et al 1999a; Hartmann et al 2003; McClure et al 2005; White et al 2007) although again confounding factors make it impossible to directly attribute this effect to exercise and reports of a high prevalence of gastric ulcers in sedentary grazing horses also casts doubt on this putative relationship (le Jeune et al 2006). In contrast, observations consistent with a possible beneficial effect of exercise on GI motility and impaction colic are published. Impaction colic has frequently been noted to follow a reduction in exercise (Love et al 1994; Dabareiner and White 1995; Hillyer et al 2002; Cox et al 2007) although this relationship might be confounded by associated dietary changes or other factors such as pain from musculoskeletal injury (a possible cause of the reduction in exercise) (Little et al 2001) or medication (eg nonsteroidal anti-inflammatory drugs (NSAIDs)) (van Hoogmoed et al 2002). Several possible mechanisms by which exercise could influence GI health and function are discussed below.
Exercise and intestinal barrier function In humans and horses there is a reduction in intestinal blood flow by as much as 80% in response to exercise with even greater exercise-related hypoperfusion of the equine pancreas and stomach having been measured (Manohar 1986; Manohar et al 1995; Clausen 1997). Increased blood viscosity and cellularity associated with exercise in horses may further compromise mucosal perfusion (Geor et al 1994). Non-endocrine mechanisms such as the significantly increased intrabdominal pressure seen at the start of exertional activity (Lorenzo-Figueras and Merritt 2002) might also adversely affect splanchnic perfusion. Reduced GI perfusion could lead to pain and/or compromise of the mucosal barrier (Pals et al 1997). Cortisol has been found to increase GI permeability and the increased cortisol secretion found especially during endurance exercise (Nagata et al 1999) could further exacerbate the effects of reduced perfusion on mucosal integrity and subsequently allow access of endotoxin and other noxious substances into the peripheral circulation (Fink et al 1991). NSAID use at the time of a competitive race is discouraged in human athletes owing to the potential additive deleterious effect of these drugs on intestinal permeability (Smetanka et al 1999; Lambert et al 2007). In humans and horses plasma endotoxin levels increase significantly after intense or prolonged exercise (Cannon and Kluger 1983; Baker et al 1988; Bosenberg et al 1988; Sakurada and Hales 1998; Barton et al 2003) and intestinal permeability has been shown to increase with relatively modest exercise in dogs (Davis et al 2006). Plasma collected from humans after exercise caused pyrexia when transfused into rats, whereas pre-exercise plasma did not, indicating an exercise-induced accumulation of endogenous pyrogens (presumably endotoxin) in the circulation (Cannon and Kluger 1983) and suggesting a role for intestinal-derived endotoxin in the hyperthermic response to intense exercise.
Exercise, intestinal motility and absorption Studies of the effects of exercise on intestinal transit and absorption generally lack consensus. Generally mild intensity exercise appears to improve or have no effect on gastric emptying in humans although gastric emptying is significantly delayed by intense exercise in humans perhaps as a result of increased levels of catecholamines and endorphins (Neufer et al 1989; Rehrer et al 1990; Leiper et al 2005). Some human studies have found orocaecal transit time and colon motility are increased by exercise whereas others have not (Gisolfi 2000; Peters et al 2001). Putative deleterious effects of exercise on intestinal absorption in humans are equivocal and the effects, if present, are probably not marked (Gisolfi 2000; Lang et al 2006). Similarly, several studies have failed to find a marked effect of exercise on GI motility and absorption in horses and donkeys (Orton et al 1985b; Pearson and Merritt 1991; Sosa Léon et al 1997; Pagan et al 1998). Equine studies have shown slightly increased dry matter intake and increased digestibility of the dry matter in association with light exercise consistent with slower intestinal transit (Olsson and Ruudvere 1955; Orton et al 1985a,b). In a study using yearling horses, Orton et al (1985b) found that trotting (12 km/h) for 1 hour/day prolonged the mean retention time of a marker of fluid digesta (51Cr-EDTA) whereas a marker of particulate digesta (ruthenium-phenanthroline) was more rapidly eliminated in the faeces of the exercising horses. Overall digestibility was improved by light exercise and this was explained by the fluid phase containing the more digestible dietary elements. A further study by Pagan et al (1998) examined the effect of more intense exercise on parameters of digestibility and intestinal transit in adult Thoroughbred geldings subject to a gradually increasing treadmill exercise programme culminating in speeds of 4-9 km/hr over 8 km at a 3° incline. In contrast to the study by Orton et al (1985b), exercise resulted in a small but significant decrease in dry matter digestibility, largely because of decreased acid detergent fibre digestibility. However, exercise led to more rapid elimination of a particulate marker of digesta (Ytterbium chloride) in accordance with the previous study.
Electrolytes and Water: Effect on Performance and Colic Dehydration and electrolyte imbalance are considered risks for colic as well as affecting performance (Foreman 1998; Naylor et al 1993). Colic causes dehydration either by lack of intake or by 3rd space sequestration depending on the severity of the obstruction. Horses suffering from dehydration cannot perform optimally. The best examples are endurance horses or three-day event horses which suffer dehydration due to sweating particularly in hot weather (Andrews et al 1994). Continued sweating in the face of dehydration does not affect the amount of sweat produced thereby exacerbating dehydration if there is no fluid replacement. However, acclimation over time in hot environments decreases sweat rate (McCutcheon et al 1999). Though horses start to rehydrate after exercise with normal oral intake, full rehydration is not completed for 24-48 hours (Butudom, 2003). The time required for full recovery without supplementation has not been reported and optimal performance is unlikely until hydration status has returned to normal.
Losses due to sweating during long distance exercise also causes sodium, potassium calcium and chloride depletion (Carlson and Mansmann 1974; Rose et al 1980), though when supplementation occurred all electrolyte concentrations increased with an increased PCV (Hess et al 2005). Hypokalaemia and hypocalcaemia, which has been detected in dehydrated and exhausted horses, is associated with paralytic ileus. The resulting functional obstruction potentially creates more dehydration and electrolyte imbalance.
Synchronous diaphragmatic flutter, an indication of electrolyte imbalance (hypochloraemia, hypokalaemia and hypocalcaemia) during exercise, is associated with ileus or colic (Mansmann et al 1974; White 1998). Hypocalcaemia and hypomagnesaemia have been recognized during colic requiring surgery and strangulating diseases and can be associated with cardiac dysrhythmias and ileus (Dart et al 1992; Garcia-Lopez et al 2001). Intravenous treatment with calcium borogluconate and or magnesium chloride is recommended. Correction is often immediate though magnesium may require oral supplementation for several days to replenish a deficiency.
Exercise, competition and gastric ulceration Diet appears to have a significant influence on development of gastric ulcers in performance horses as discussed above. However, a study by Bell et al (2007a), which found no protective effect of grazing on the development of gastric ulcers in racehorses, emphasised the likelihood of a direct association between other factors such as exercise with ulcerogenesis also. Gastritis is a well-recognised consequence of prolonged exercise in humans and dogs (Simons and Kennedy 2004; Davis et al 2005,2006) and has been mainly attributed to decreased GI perfusion, a key element of glandular mucosal defence (Murray 1992a, 1999). Antacid therapy significantly reduces the incidence and severity of exercise-associated gastritis in humans and dogs suggesting acid-related factors in the aetiology (Choi et al 2006; Williamson et al 2007). Training significantly increases serum gastrin levels in response to feeding in horses (Furr et al 1994) and may be a contributor to exercise induced gastric acidity and ulceration. Gastric ulceration is common in exercising horses although gastric glandular mucosal ulcers (analogous to human and canine gastric lesions) are rare in comparison to squamous mucosal ulcers (Murray et al 1996, 2001) suggesting different aetiology of exercise-associated gastric disease in the horse versus the human and dog. Gastro-oesophageal reflux, a human condition more closely related to gastric squamous mucosal ulceration in horses, is also frequently associated with exercise in humans and has been attributed to the combination of increased intra-abdominal pressure and decreased intra-thoracic pressure at peak inhalation during exercise (Peters et al 2000b; Jozkow et al 2006). Lorenzo-Figueras and Merritt (2002) found the intragastric pressure to increase suddenly and markedly at the walk-trot transition in horses presumably due to external compression following increased tone of abdominal muscles. Furthermore, gastric fluid pH decreased to <4 during trotting and galloping and did not increase until the horse returned to walking. It was suggested that this effect would raise the intragastric fluid level and promote contact of low pH gastric fluid with the upper squamous mucosa although oesophageal reflux was not studied.
The inevitable physical agitation of abdominal viscera during exercise may well have adverse effects on the GI tract and the frequency of most GI symptoms in humans is almost twice as high during running than in less vibrational and concussive exercise such as cycling and swimming (Sullivan 1987; Rehrer and Meijer 1991; Peters et al 2000b). Rapid and forceful diaphragmatic movements during exercise might also contribute to GI disturbance. Equine gaits faster than a walk are all fairly concussive and could theoretically have many adverse consequences such as stretching, compression, bouncing and displacement of various viscera, mesentery and blood vessels and also significant abnormal agitation and disturbance of the normal movement of ingesta contained within. Jumping activities might be even more disruptive. Exercise-associated splashing of low pH gastric fluid may contribute to squamous mucosal ulceration in exercising horses. It is common for racehorses to be exercised when the stomach is relatively free of food content. This practice is partly based on the analogy that human athletes feel uncomfortable if running with a full stomach but it does inevitably promote agitation, movement and splashing of acidic fluid within the stomach during exercise especially if dietary management has promoted fluid gastric content (see above). Although human athletes suffer a high incidence of gastric symptoms during endurance exercise, the contact of low pH fluid with the equine squamous mucosa is likely to be even more poorly defended than that within the human stomach. Thus on the one hand we have a problem with exercising on a full stomach, and on the other we have a problem with exercising on an empty stomach. The premise that horses have probably not evolved to undergo frequent intense or prolonged endurance exercise, indicates that there may not necessarily be a satisfactory solution to this quandary although more consideration might be given to feeding long-fibre pre-exercise to encourage a more physically stable and pH-stratified gastric content (Murray and Grodinsky 1989; Merritt 2003).
Further experiences of the performance horse may contribute to gastric ulceration such as transport and stress (McClure et al 2005) and horses with a more nervous disposition have been found to be predisposed to ulceration (McClure et al 1999). However, theoretically stress and cortisol-mediated prostaglandin suppression would be more likely to affect gastric glandular defence and therefore promote glandular rather than squamous mucosal ulceration (Furr et al 1992; Murray 1999). It may be the case that interruption of dietary patterns, alimentary secretions and even physical vibrational/agitational effects of transport might confound the proposed association between stress and squamous mucosal ulceration in adult horses (McClure et al 2005).
NSAIDs and GI disease Given the frequency of orthopaedic injury and disease in performance horses, NSAIDs are probably the commonest prescribed medication in this group of animals (Kallings 1993). NSAIDs antagonise prostaglandin-dependent mucosal defence and integrity of the GI tract (Wilson 1991) and NSAID therapy has been associated with protein losing enteropathy, gastric ulcers and right dorsal colitis in horses (Karcher et al 1990; Monreal et al 2004; Reed et al 2006).
NSAID-induced gastric ulceration occurs primarily, but not exclusively, in the prostaglandin-dependent gastric glandular mucosa (Monreal et al 2004), a relatively uncommonly ulcerated area in competition horses (Vatistas et al 1999b; Murray et al 2001; Begg and O’Sullivan 2003; Nieto et al 2004). However, NSAIDs might also have a local direct adverse effect on mucosal surface protective phospholipids and cell membranes and therefore potentially contribute to gastric squamous mucosal ulcers in this way (Lichtenberger et al 2006). Whatever the exact mechanism of NSAID-induced gastric ulcers, there appears to be little evidence that NSAID therapy at commonly used doses is a significant cause of gastric ulceration in horses (Hammond et al 1986; Murray et al 1996; Vatistas et al 1999b; Rabuffo et al 2002; Reed et al 2006).
The health and function of the right dorsal colon appears especially NSAID-sensitive and right dorsal colitis is a well recognised, albeit uncommon, consequence of NSAID therapy in horses (Karcher et al 1990; Bueno et al 2000) and it has been suggested that young performance horses are especially predisposed (Cohen 2002). Right dorsal colitis is most commonly associated with excessive NSAID dosing but some cases occur with short term and appropriate dose regimens (Cohen et al 1995a). No difference in the right dorsal colonic mucosa was observed during experimental administration of phenylbutazone for 28 days (McConnico et al 2006), however, right dorsal colonic blood flow was significantly increased suggesting a systemic effect of chronic administration.
It is important to remember that NSAIDs may be a cause of protein losing enteropathy, hypoproteinaemia and weight loss in horses prior to development of obvious clinical signs of gastric or colonic disease (Reed et al 2006).
Relationship between the GI tract, training and performance The putative association between GI disease and poor performance is generally ascribed in terms of anecdote and intuition rather than evidence and does not feature at all in 2 large reviews including 623 poor performance cases (Morris and Seeherman 1991; Martin et al 2000). Traub-Dargatz et al (2001) estimated that 2.2 days of use are lost per colic episode in horses although this is unlikely to have a major impact on training or competition unless occurring at key times.
In addition to the more familiar adaptations to physical training in the cardiopulmonary and musculoskeletal systems, adaptation of the GI tract may also be an important and beneficial component of training. Exercise-associated splanchnic hypoperfusion, loss of barrier function and endotoxaemia as discussed above may be an important cause of fatigue via effects on hyperthermia, hyperlactaemia and cardiac contractility (Cannon and Kluger 1983; Abel 1989; Williamson and Barton 1996; Barton et al 2003). Ovine studies have suggested that improved heat tolerance associated with training results from reduced endotoxin absorption from the better perfused intestine of physically fit versus sedentary animals (Sakurada and Hales 1998). Furthermore, actively competing humans and horses have high levels of anti-endotoxin antibodies suggesting chronic/repeated exposure to endotoxin and, possibly, greater tolerance of exercise-associated endotoxaemia and its putative adverse effects described above (Baker et al 1988; Bosenberg et al 1988); although a consistent correlation between level of physical fitness or performance and endotoxin antibodies has not been shown (Camus et al 1997; Barton et al 2003).
Although several studies have shown reduced gastric emptying in humans in response to intense exercise, trained endurance athletes were found to have faster gastric emptying rates than sedentary controls both at rest and during exercise suggesting gastric adaptation to exercise (Carriό et al 1989). However the suggested mechanism for this adaptation is enhanced gastric distension associated with a greater food intake by highly active human athletes. In an equine context the opposite effect might actually be the case when the volume of a concentrate-rich competition diet is compared with a high-forage, lower energy diet of more sedentary horses.
Gastric ulceration is the most frequently discussed GI disease of competition horses. Studies of gastric ulcers in adult horses have most frequently found signs such as weight loss, poor appetite, poor hair coat, colic and diarrhoea (Murray et al 1989; Murray 1992b; Vatistas et al 1999b; Dionne et al 2003; Stämpfli and Oliver 2006). Several publications describe decreased performance as a consequence of the disease (Murray 1992b; Andrews and Nadeau 1999; Equine Gastric Ulcer Council 1999) despite an absence of good quality evidence existing to justify this widely accepted association. In a study of 202 Thoroughbreds in active race training, Vatistas et al (1999b) compared the prevalence of gastric ulcers against the trainers’ subjective assessments of performance (NB not all horses had previously trained or raced). Those horses performing “at or above expectation” had a significantly lower prevalence of ulcers than those horses performing “below expectations” (78% vs 94%, P=0.01). However, the severity or number of ulcers was not correlated with the degree of under-performance. Mitchell (2001) studied 134 hunter/jumper and dressage horses with a complaint of poor performance and other signs consistent with gastric ulcers and found gastric ulcers in 84 (63%) of them. Following treatment with omeprazole a “noticeable improvement in performance within 5-7 days” was frequently seen and “trainers consistently related that treated horses were more willing to work”. In a study of 80 racing Standardbreds, Jonsson and Egenvall (2006) found a higher prevalence of gastric ulcers in horses performing “worse than expected” (66.7%) than in groups performing “better than expected” (36.4%) or “as expected” (31.1%). Although these data illustrated a significant association between trainers’ expectations and prevalence of gastric ulceration, this only remained significant (P=0.002) in univariable analysis and may have been confounded by other factors. Furthermore, the prevalence of ulcers in horses performing “better than expected” was greater than in those performing “as expected”. Johnson et al (2001) examined 565 horses with trainer complaints thought to be compatible with gastric ulceration and 506 (90%) were indeed confirmed to have gastric ulcers and were then treated with omeprazole. Performance in the 6 months preceding the trial was then compared with performance in the 6 months following treatment of the gastric ulcers. Before the trial 40% of horses finished races in the top 3 whereas this proportion increased to 46% after treatment for ulcers although statistical comparison of these findings was not described. Franklin et al (2008) described 4 poorly performing Thoroughbred racehorses in which no abnormalities other than gastric ulceration were found. Performance judged by Raceform rating and earnings of each horse was improved following treatment with omeprazole.
There are many trainer and practitioner experiences of poorly performing horses found to have gastric ulceration. However, given the high prevalence of gastric ulcers in performance horses it should come as no surprise that ulcers are frequently found in poorly performing subjects. Subsequent improvement in performance following treatment is perhaps more convincing of a “cause and effect” relationship (Mitchell 2001; Johnson et al 2001; Franklin et al 2008), especially in the light of evidence suggesting no performance enhancing effect of omeprazole therapy (McKeever et al 2006), but could still be confounded by many other factors such as time, rest, dietary intake, bodyweight changes, other management changes or placebo effect. Large, controlled studies associating gastric ulcers with objective parameters of performance and subsequent response to specific therapy have not yet been published.
The majority of performance horses found to have gastric ulcers do not show readily detectable signs of pain (colic) although that is not to say that subclinical discomfort is not experienced by affected animals. Low grade abdominal pain is hard to detect and define in horses yet it seems likely that, if present, performance would indeed be compromised. Adverse GI related symptoms such as nausea, fullness/bloating and abdominal cramps have been shown to be quantitatively associated with impaired performance and power in human cyclists (Thorburn et al 2006).
Postoperative nutrition and condition The energy needs of horses after colic or abdominal surgery are not known. Because of the wide variation in the severity of colic, clinicians have assumed that there is no need for an increase in nutritional requirement after simple colic. In most cases horses are only off feed for a short period of time (12-48 hours) if at all and there is no apparent clinical effect in these horses. Hypertriglyceridaemia has been documented in clinically ill horses with GI disease suggesting malnutrition occurs during periods without enteral nutrition and altered metabolism due to disease (Dunkel et al 2003; Durham et al 2004). Sixteen Mcal (67 MJ ; 32 kcal/kg;), the energy requirement for maintenance for an adult horse on pasture, may not be needed for many horses that are stall confined and considered to have basal metabolism. During the immediate postoperative period as much as 15-20% less energy should be expended for digestion (Geor 2007). The basal requirement for the stall confined horse postoperatively is likely sufficient unless they are challenged by a systemic inflammatory response syndrome due to peritonitis or endotoxemia. If the basal requirement is 70% of maintenance, the requirement for the stall rested horse would be approximately 21 kcal/kg per day (Geor 2007). Estimates for the increased energy requirement for horses due to abdominal surgery or shock are based on human information, which suggests an increase of 30% to 100% above maintenance, but this has not been substantiated in the horse. The protein requirement for maintenance is estimated at 1.25 g/kg per day providing 0.9 g/kg per day of available protein. Basal requirements for protein may also be decreased to 0.6-0.8 g/kg per day. Adequate energy from carbohydrates or fat is needed to prevent utilization of protein for energy during nutritional therapy.
Most hay or grass fed to horses is not evaluated for nutrient content, so the daily energy and protein intake frequently is not known for horses during medical care. The use of pellet formed feeds such as Equine Senior® allows calculation of the energy and protein intake. Other total horse feed supplied as pellets are usually made partially from grains thereby increasing soluble carbohydrate concentrations. This may be appropriate for some horses after surgery, but this is dependent on concern for altering the flora with excess starch in the caecum and colon and also perhaps stress induced insulin resistance and glucose intolerance. In some instances this readily available source of carbohydrates may be an advantage. Nevertheless, feedstuffs such as bran are not recommended due to the high amount of carbohydrate often present after milling (White 2006).
Though not currently recognized as the standard of care, horses that are not ingesting at least 50% of their maintenance requirement (15-18 kcal/kg per day) for 48-72 hours should have nutritional support. Supplementation for severe GI disease has been used to increase survival and potentially limit the time required to return equine athletes to competition. Oral liquid diets made for humans have been used in horses for enteral nutrition but contain little fibre and increase the risk of colitis and laminitis (Buechner-Maxwell et al. 2003). Other diets consisting of alfalfa pellets, dextrose, vegetable oils and amino acids maintained horses’ weight but also increased the risk of diarrhoea and laminitis (Naylor et al. 1984). Development of an enteral energy/protein source, which is readily absorbed with minimal risk of disturbing motility or gut flora is needed for horses requiring enteral nutrition during the initial post operative period. Supplementation of glutamine has been suggested to help maintain or restore the intestinal epithelium, reduce the inflammatory response of the epithelium, and prevent bacterial translocation (Zhang et al. 1995; Blikslager et al. 1999; Liboni et al. 2005).
After a period of critical care for GI disease which involves prolonged ileus, shock or infection, horses commonly lose weight and muscle mass. The few studies that have measured the effect of severe disease on nutrition have not provided an objective measure of body condition. In the authors’ experience is it not unusual for horses to lose 25-50 kg after prolonged treatment for colic or colitis. Though both inflammation and healing of the intestine is generally rapid, return to exercise and performance is determined by the horse’s condition or in the case of abdominal surgery by healing of the abdominal wall.
Feeding horses intravenously though shown to be possible has not been proven to be necessary to improve horse survival after abdominal surgery (Lopes and White 2002). In human patients there is no evidence that parenteral nutrition alters mortality of surgical or critical care patients, but there is evidence that complications are decreased particularly in patients that are malnourished (Heyland et al. 1998). A comparison of horses evaluated after abdominal surgery with or without supportive parenteral nutrition, did not show a difference in outcome. Nevertheless there was evidence of significantly decreased triglycerides, total bilirubin, albumin, and urea with significantly higher serum glucose and insulin in horses treated with parenteral nutrition after small intestinal resection and anastomosis (Durham et al. 2004). The treatment did not change time until enteral feeding, time or cost of hospitalization. Parenteral nutrition should be considered immediately, if ileus, shock, or peritonitis is predicted to prevent oral ingestion of feed for more than a few days. Subjectively, it appears to the author that parenteral nutrition has been responsible reducing convalescence time and cost for horses in shock and unable to tolerate enteral nutrition immediately after surgery. Parenteral supplementation may decrease weight loss and lessen the overall time required to rehabilitate horses for competition. Clearly more research in the area of nutrition for horses in need of critical care is required.
Intestinal surgery and performance Equine intestine heals rapidly after enterotomy, or resection and anastomosis. The limiting factor for performance is normally the abdominal incision, which requires 2-4 months to heal and strengthen depending on the horse’s postoperative and nutritional status. Though intestinal inflammation can be present for several weeks after surgery, the healing is normally rapid (Freeman et al 1988). Resection of greater than 60 percent of the small intestine has created nutritional compromise from lack of absorptive capacity with resultant poor body condition. Resection of the large colon can result in altered nutritional status including decreased phosphorus digestion. However, anecdotal observations suggest that after a period of adaptation horses can be successful in numerous types of performance after large colon resection. There appears to be a similar response to bypass of the ileum or the caecum. In both instances most reports success is defined as survival and the ability to compete not included (Craig et al 1987; Ford et al 1990; Gerard et al 1996; Lores et al 2008). In the author’s (NAW) experience the response to these techniques does not decrease a horse’s ability to perform unless the original condition creates chronic inflammation or obstruction.
Colopexy which has been used to prevent large colon displacement or volvulus in horses considered to be at high risk by surgically adhering the right, left or sternal flexure to the abdominal wall (Hance and Embertson 1992). The technique can be completed electively by laparoscopy, but there is hesitancy to complete the surgery in horses used for athletic competition due to the risk of colon rupture at the site of adhesion. Though anecdotal reports suggest horses can compete in events, a risk for fatal rupture does exist. In horses considered to have a high risk of recurrent colon displacement and that will be used for competition, colon resection is considered the best option.
Incisional hernia and performance Abdominal or GI related problems that can limit performance after abdominal surgery include chronic obstruction, decreased absorptive capacity or incisional hernia. Except in cases of intestinal strangulation in the epiploic foramen, risk of death due to the original problem is not increased beyond 2 months after surgery (Proudman et al 2002). Though secondary colic risk is increased after abdominal surgery (Cohen et al 1995b) there is no report that indicates a decreased performance for a particular equine competition after surgery for colic.
The incidence of incisional hernia after ventral midline celiotomy in horses has been reported to range from 5.7-18% (Kobluk et al 1989) Studies have also reported that the risk of incisional hernia increases with relaparotomy, increased pain on recovery, leucopaenia, wound infection and high PCV (Gibson et al 1989; French et al 2002). As many as 20% of horses that develop incisional hernias have multiple smaller hernias along the incision (Gibson et al 1989) and may not require surgical intervention and can often be managed conservatively even for competition.
Ultrasound may be useful in the evaluation of the hernia. Adhesion of intestine to the hernial sac can complicate surgery or may be a risk for intestinal rupture during certain types of competition though this has not been documented. If surgical repair is required adequate time for maturation of the hernia ring and separated abdominal wall is required for adequate tissue strength for sutures. A minimum of three months is considered necessary after the initial surgery or when any infection has been controlled.
If the abdominal wall can be apposed primary closure should be completed. If the abdominal wall cannot be apposed, the hernia should be repaired using a synthetic mesh and fascial overlay using a horizontal or vertical mattress suture (Stick 2006). Though no reports specify results of hernia repair by either method, anecdotal evidence suggests that there is no limitation after satisfactory healing.