Effects of transportation on gastric pH and gastric ulceration in mares

Abstract Background Transportation has been suggested as a risk factor for gastric ulceration in horses, but limited evidence supports this assumption. Animals Twenty‐six Standardbred, Thoroughbred, and Warmblood mares from a university teaching herd. Methods Twelve mares were confined for 12 hours, overnight, in reproductive stocks with indwelling nasogastric tubes (NGTs) to assess pH of gastric fluid (GF). Gastric ulceration was assessed endoscopically before and after confinement. Subsequently, 26 horses were transported for 12 hours, overnight, in 2 consignments. During transportation, GF was aspirated from indwelling NGT placed in the same 12 mares used in the confinement study, and gastric ulceration was assessed endoscopically before and after transportation in all horses. Results The median pH of GF in confined horses was 1.70‐2.49 at each sampling point, and there was no apparent effect on gastric squamous ulcer scores. The median pH of GF from the same 12 horses at corresponding sampling times during transportation was 6.82‐7.22. Transportation was associated with increased gastric squamous ulcer scores, particularly in horses fasted for gastroscopy and NGT placement immediately before departure. Gastric emptying appeared delayed after transportation in horses fed before departure. Conclusions and Clinical Importance Transportation is associated with increased gastric squamous ulceration and with increased pH of GF. These findings may be a consequence of impaired gastric emptying and reflux of alkaline small intestinal content, with factors such as duodenal bile salts and short‐chain fatty acids mediating mucosal injury.


| INTRODUCTION
Gastric ulceration is the most common disease condition of the equine stomach 1 and is associated with colic, decreased appetite, failure to thrive, and poor performance. 2,3 Equine gastric ulcer syndrome has been recommended as an encompassing term for all erosive and ulcerative conditions of the equine stomach. 4 Importantly, however, given likely differences in the pathogenesis, clinical presentation and management of diseases of the squamous and glandular gastric mucosa, 5,6 the terms equine squamous gastric disease (ESGD) and equine glandular gastric disease (EGGD) are now preferred. 4 The pathogenesis of ESGD is likely related predominantly to increased exposure to highly acidic gastric contents, 5 although bile acids and volatile fatty acids also have been implicated in the induction of disease. [7][8][9] Possible causes of EGGD are less well characterized although, in common with ESGD, it is hypothesized that erosion and ulceration occur as a result of an imbalance between aggressive and protective factors such that mucosal integrity is compromised. In horses, as in people with peptic ulcer disease, factors that decrease prostaglandin production 6,10 and possibly altered gastric microbiota 11,12 may contribute to the development of disease.
Horses travel frequently for performance, breeding, and other purposes. Transportation has been suggested as a risk factor for gastric ulceration 13 and a recent survey of transport-associated health problems indicated that gastrointestinal (GI) problems occur frequently in association with transportation of horses. 14 Limited evidence currently is available regarding the effects of transportation on gastric ulceration or gastric pH.
Transportation is associated with factors that may contribute to gastric ulceration such as physiologic stress responses (increased heart rate and serum cortisol concentrations), [15][16][17][18][19][20][21] changes in feeding practices and water consumption, 22,23 and changes in GI microbiota. 12,24 Our study was conducted to determine the impact of 12 hours of transportation without food or water on gastric ulcer scores and gastric fluid (GF) pH in horses. A secondary objective was to determine the effect of pretransport feeding practices on these outcomes. We hypothesized that transportation would be associated with ulceration of the squamous mucosa associated with acidic GF and that these effects would be more severe in horses fasted before transportation.

| Animals
Twenty-six Standardbred (n = 14), Thoroughbred (n = 10), and Warmblood (n = 2) mares were included in the study. Mean age was 9.9 years (range, 4-20 years), and mean body weight was 518.8 kg (range, 416-658 kg). All horses were Charles Sturt University teaching or research horses and had been resident on site for ≥4 weeks. Prior transport history was unknown for each horse, although all had been transported on at least 1 prior occasion.

| Experimental design
The study was conducted in 2 parts. Part 1 was conducted as a preliminary observational study to assess the effect of overnight confinement (1800-0600 hours), without feeding, on gastric pH and gastric ulcer scores in 12 mares. Part 2 was conducted as an interventional study to determine the effect of overnight (1800-0600 hours) transportation on gastric pH and gastric ulcer scores in 26 mares. Feed management before transportation consisted of feeding <60 minutes before departure (group 1, n = 7), feeding 6 hours before departure (group 2, n = 7) and fasting for 12 hours for gastroscopy and nasogastric tube (NGT) placement immediately before departure for the 12 mares used in study part 1. Study design is presented in Table 1, with further detail provided in Data S1 ( Figure S1).

| Part 1
Mares (n = 12) were confined in reproductive stocks (150 × 71 cm; height of front gate, 122 cm) as 2 groups, each of 6 horses, on consecutive nights. Horses were fed alfalfa hay (1%-1.5% body weight) between 0600 and 0700 hours on the morning of confinement. Water was withheld from 1200 hours. Each horse underwent veterinary clinical examination and venous blood was collected for hematology, serum biochemistry, and blood gas analysis at 1400 hours, 4 hours before confinement (T0). Intestinal borborygmi were graded subjectively by one investigator (B.P.) based on auscultation of 4 abdominal quadrants (upper left, lower left, upper right, lower right) and recorded as 0 (no intestinal sounds auscultated in 60 seconds), 1 (decreased activity), 2 (normal activity, 2 or 3 discrete rumbling or gurgling noises in 30s), or 3 (increased activity) for each quadrant. These results were summed to give a GI activity score, as previously described. 19 Horses were sedated (200 mg xylazine and 10 mg acetylpromazine, or 10 mg detomidine and 5 mg butorphanol IV) between 1600 and 1800 hours for gastroscopy and placement of an indwelling NGT (Veterinary

| Gastroscopy and GF pH determination
Endoscopic evaluation of the gastric mucosa was performed using a 3-m endoscope (9 mm outer diameter, Olympus Medical Systems Corporation, Tokyo, Japan), as previously described, 25 after insufflation of the stomach with air to permit complete visualization of the glandular and squamous gastric mucosa. Each horse's stomach was assessed using a validated equine scoring system 26 with separate scores for the greater curvature, lesser curvature and fundus summed to give a squamous score, and separate scores similarly summed for fundic and pyloric glandular mucosa, as previously described. 25 Gastroscopy examinations were video recorded and stored using a unique, randomly generated, 4-digit identification number. Gastroscopy findings were scored in real  (Table 2).
Gastric fluid was aspirated via the endoscope at the completion of mucosal scoring. Indwelling NGT were then placed by grasping a small piece of tape at the distal end of the NGT using endoscopic biopsy forceps and passing the 3-m endoscope and NGT into the stomach, as previously described. 25 GF (1-10 mL) was aspirated every 2 hours via the NGT. Fluid was placed immediately into 50-mL polypropylene containers, and pH was determined within 15 minutes of T A B L E 2 Feed retention score for subjective assessment of gastric emptying based on the amount of ingesta retained in the gastric lumen

| Hematology and serum biochemistry
Routine hematology parameters were determined using an automated cell counter analyzer (Cell Dyn 3700; Abbott, Chicago, IL

| Statistical analyses
Power analysis before the study and based on previous work conducted by our group suggested that 6 horses (the minimum subset of horses in the experimental design described above) afforded satisfactory power (>0.8, α = .05) to discriminate minimum differences in mean results for CK of 16.5 IU/L, cortisol of 0.5 mmol/L, and gastric pH of 2.2, assuming SD of 9.9 IU/L, 0.3 mmol/L, and 1.3 respectively.
All data were evaluated initially using summary statistics. Differences between hematological and serum biochemistry parameters were compared by 2-way repeated measures analysis of variance (ANOVA) after testing for normality using the D'Agostino and Pearson test.
Where possible, nonparametric results were log-transformed before analysis, or nonparametric tests were used.  Figure S1).
Mean rectal temperature still was increased at T2, 8 hours after return, for transported horses. Body weight and respiratory rate showed no changes attributable to transportation or confinement.
Abdominal auscultation identified decreased GI sounds at each examination after transportation (Figure 1), but no changes were observed in confined horses. Both transportation and confinement were associated with transient increases in plasma cortisol concentrations ( Figure 2). Transportation was associated with increased white cell and neutrophil counts, and these changes were not observed in horses in association with confinement ( Figure S2). A significant difference (P = .04) was evident for red cell count for transported horses in comparison with results during confinement, but pairwise comparisons were not significant. Transportation, but not confinement, was associated with increased plasma activity of CK, and plasma F I G U R E 3 pH results for gastric fluid (GF) aspirated every 2 hours via indwelling nasogastric tubes from 12 horses during confinement (left) and transportation (right). Results are shown as median (horizontal bar) and range (whiskers), with all available data shown. Transportation, but not confinement, was associated with significantly increased pH of GF, with results significantly greater than obtained at departure shown (**P < .01; *P < .05) F I G U R E 4 Summed squamous ulcer scores before (T0), immediately after (T1), and 60 hours after (T3) overnight confinement or transportation. Median squamous score increased after transportation (blue), but not after confinement (white), with significant differences observed between T0 and T1 for transported horses, and at T1 when transported horses were compared with confined horses. A minor effect on glandular ulcer score was observed in confined horses (white), but transportation had no effect on glandular ulcer scores (blue). Results are shown as median (horizontal bar), mean (cross), quartiles (box), and range (whiskers), with all data points shown. Significant differences are shown concentrations of protein, albumin, sodium, ionized calcium, lactate, and glucose ( Figures S3-S5). Plasma potassium concentrations varied with both transportation and confinement ( Figure S4).

| Effect of confinement on gastric ulceration and pH
Median pH of aspirated GF was 2.43 (range, 1.31-6.65) before confinement. Results obtained from samples aspirated during confinement ranged from 1.27 to 6.80, with median values between 1.70 and 2.49 at each sampling point. No significant effect of time was observed on median pH of aspirated GF during confinement (P = .39, Figure 3).

| Effect of transportation on gastric ulceration and pH
Attempts were made to aspirate GF through the endoscope before Median pH of samples aspirated before departure from horses with indwelling NGT was 4.00 (range, 1.48-6.88; n = 12). Median pH of GF samples aspirated during transportation was between 6.82 and F I G U R E 5 Effect of time of feeding before departure on feed retention scores at T1 (immediately after 12-hour transportation). Median scores were increased in horses fasted for 1 or 6 hours before departure, in comparison with horses fasted for 12 hours. Results are shown as median (horizontal bar), mean (cross), quartiles (box) and range (whiskers), with all data points shown. Significant differences are shown F I G U R E 6 Representative images showing gastric ulceration in a transported horse at T0 (before departure), T1 (immediately on return from transportation) and T3 (60 hours after transportation). Ulceration was attributed to splashing of liquid content against the gastric mucosa, and was particularly marked in horses fasted before transportation for placement of indwelling nasogastric tubes. All images show the greater curvature of the stomach. At T0 there is hyperkeratosis only (grade 1). At T1 there is extensive ulceration extending from the margo plicatus, with bleeding evident in some affected areas (grade 4 of greater curvature and fundus). By T3 the affected area appears less inflamed and ulcers are no longer bleeding; however, extensive ulceration is still apparent, and remains deep in some affected areas (grade 4) 7.22, significantly higher than before transportation (P < .001,  Figure S6).

| DISCUSSION
We evaluated the effect of transportation on gastric ulceration and pH of GF in horses. In part 1, a preliminary study was conducted in which horses were subjected to confinement, sampling and gastroscopy procedures that mimicked those experienced during transportation. Transported horses had well-characterized clinical, hematological, and blood biochemistry changes reported after transportation in previous studies, 16,19,20,[28][29][30][31] and these changes were not observed in horses confined for a similar duration. Plasma cortisol concentrations increased in transported horses, as has been reported previously. 19,21,32,33 However, study design may have prevented discrimination of effects of transportation on cortisol because confinement was associated with a similar increase in plasma cortisol concentration and these results might reflect expected diurnal variations. 34,35 Indeed, observed changes were remarkably consistent with diurnal changes in plasma cortisol concentration reported previously. 35 Our findings suggest that transportation of 12 hours without feeding may contribute to the development of ESGD because travel induced or exacerbated ulceration of the squamous mucosa in the majority of horses and was associated with a significant increase in F I G U R E 7 Gastric ulcer scores in horses according to time of feeding before transportation. Horses fasted for 12 hours before placement of nasogastric tubes had higher summed squamous scores (blue) than horses that were fed at 1200 hours (6 hours before departure) or at 1700 hours (immediately before departure), but observed differences were not significant (P = .06 Because the fundic squamous mucosa was discolored and had residual feed particles adherent after transportation in many horses, it was assumed that contact with GF contributed to mucosal damage.
Lesions extending into the squamous mucosa from the margo plicatus are uncommon in spontaneous disease. 38 These observations contradict a recent review emphazing the role of HCl as the dominant erosive agent in the development of ESGD 5 and suggest that less acidic factors, such as short-chain fatty acids 8,9 and duodenal bile salts, 7 are important contributors to mucosal damage in some circumstances.
Reflux of alkaline small intestinal content is observed commonly during gastroscopic examination, and samples obtained from NGT in our study were typically yellow, consistent with bile content. These observations are important because gastric ulcer prophylaxis before transportation using proton pump inhibitors such as omeprazole 39 will not be effective if gastric acid is not mediating mucosal injury.  and metabolic responses to diet and exercise. 44 Together these findings suggest that transportation may predispose to GI disease by affecting GI motility. Further study is required to better characterize this effect, and to determine optimal feeding strategies before and during transportation.
Although treatment with omeprazole is unlikely to have prevented development of gastric ulceration in our study, several horses required omeprazole treatment to ensure resolution of ESGD, with lesions of the lesser curvature seemingly more refractive to healing, as has been reported previously. 45 Lesions of the greater curvature and fundus, areas that are unlikely to experience ongoing contact with acidic gastric content after transportation, appeared to heal rapidly when horses were kept on pasture between T1 and T3 in our study. No effect of transportation, feeding or confinement was observed on glandular ulcer scores in our study.
Limitations of our study relate to the need to manipulate feeding practices before departure in order to place NGT for sampling of gastric content during transportation. For this reason, and because of limited access to some compartments of the transport vehicle during travel, aspiration of GF during the journey was possible for only 12 horses.
However, this number was approximately twice the required number based on power analysis, and a clear effect of transportation, relative to confinement, was observed in horses with NGT. The study design also allowed assessment of changes in horses after different feeding practices and without the presence of the NGT. Although the macroscopic appearance of aspirated GF was consistent with refluxed small intestinal content, and changes were not observed in horses confined with indwelling NGT, we cannot definitively exclude the possibility that the NGT influenced GF pH, for example by stimulating excessive salivation.
Further studies therefore are required to determine the generalizability of these findings, and to assess the effects of other management strategies, such as provision of water during transportation. Our findings suggest that the pH of GF might rapidly decrease on cessation of transportation, and future studies should control for delayed aspiration of GF at the end of transportation. It was not possible to randomize the order of treatment (confinement or transportation) in horses fitted with indwelling NGT, nor to randomize feeding before transportation (all group 1 horses were allocated to trip 1, and all group 2 horses to trip 2).
Subjective grading of gastric ulcer scores, gastric feed retention and abdominal auscultation are inherent limitations of our study. The validity of endoscopic findings was protected by evaluation of deidentified video recordings, but assessors were not blinded during assessment of GI motility when completing physical examination of horses. Feed retention limited visualization of fundic glandular mucosa for some horses, although squamous and pyloric mucosa was satisfactorily visualized. The study was not specifically designed to evaluate the effect of transportation on GI motility, and effects of confinement on gastric emptying were indirectly assessed, without variation of feeding practices before confinement. Further studies are warranted therefore to better characterize possible effects of transportation on GI motility.
To our knowledge, our study is the first to unequivocally identify gastric ulceration associated with transportation of fasted horses, and therefore to support transportation as a risk factor for the development of ESGD. To our knowledge, rapid development (within 12 hours) of squamous ulceration has not been reported previously. The severity of changes, particularly in horses fasted before travel, suggests that horses should have access to feed until the time of departure, and potentially during transport, to limit contact between the squamous mucosa and gastric secretions or small intestinal reflux. Unexpectedly, gastric ulceration was associated with increased pH of GF during transportation, consequently the administration of proton pump inhibitors or histamine receptor antagonists to horses before transportation cannot be supported based on our findings. Although spontaneous recovery occurred in a number of horses, 9 (of 26) horses were treated with omeprazole after completion of the study, primarily to ensure healing of ulcerative lesions of the squamous mucosa at the lesser curvature.
Study findings suggested that GI motility was decreased during transportation, but further studies are warranted to better characterize possible effects of transportation on GI motility, to ascertain whether our findings in mares are generalizable also to male horses and to determine whether feeding before departure is protective against development of gastric ulceration.