Heart rate and behavioural responses of crib-biting horses to two acute stressors

The heart rate and behaviour of 14 adult saddle horses, eight crib-biters and six normal controls, were investigated. Initially, the relationship between crib-biting and heart rate was investigated while the horses were undisturbed. The horses were tested when restrained with a lip twitch, and assessed when they were exposed suddenly to the rapid inflation of a balloon. The heart rate of the crib-biters during crib-biting was lower than during other behaviours. The crib-biters had a higher overall mean heart rate (P<0.05) suggesting that they may have had a higher basal sympathetic activity. After the application of the twitch, all the horses had a transient increase in heart rate which returned to basal values more rapidly in the crib-biters. The crib-biters were less reactive to the lip twitch, five of the six investigated remaining calm, and after the release of the twitch, they spent more time nibbling (P<0.05) than the control horses. The crib-biters reacted more strongly to the inflation of the balloon (three of the six reacted), and after it had been inflated they spent more time walking in the box.     

Evaluation of Continuous Infusion of Lidocaine on Gastrointestinal Tract Function in Normal Horses

Objective— To determine the effect of continuous infusion of lidocaine on fecal transit time in normal horses.
Study Design— Experimental randomized cross-over study.
Animals— Healthy horses (n=6).
Methods— Barium-filled microspheres were administered to horses by nasogastric intubation and feces were collected every 2 hours for 4 days. A bolus of 2% lidocaine (1.3 mg/kg) was administered randomly, followed by a continuous infusion of lidocaine (0.05 mg/kg/min) for 3 days or an equivalent volume of saline. The washout period was 10 days. Variables assessed included defecation frequency, weight of feces produced, intestinal transit time (number of microspheres observed on radiographs), fecal moisture content, borborygmus score, heart and respiratory rate, and signs of lidocaine toxicity (e.g., ataxia, CNS depression).
Results— During the first 24 hours of lidocaine administration, mean (±SD) fecal output (10.8±6.9 kg) was decreased compared with controls (15±4.9 kg). Mean (±SEM) time for passing 50% of the barium-filled microspheres was shorter in controls (42±1.13 hours) compared with the lidocaine group (50±1.32 hours).
Conclusions— Continuous infusion of lidocaine increases the transit time of feces in normal horses.
Clinical Relevance— Clinicians need to be aware of the effects of using a continuous infusion of lidocaine on the transit time of feces in normal horses, with a potential for exacerbating those effects when combined with drugs that decrease motility and in horses with medical colic (e.g., impaction) or where a diagnosis has not been made.     

Tegaserod (HTF 919) stimulates gut motility in normal horses

REASONS FOR PERFORMING STUDY: It has been shown that the selective 5-HT4 receptor agonist tegaserod induces an increase in frequency and amplitude of contractions in isolated muscle preparations of equine ileum and pelvic flexure. OBJECTIVES: To investigate the effects of tegaserod on gut motility and transit of spheres in normal horses. METHODS: Six mature Freiberger horses were kept under standardised conditions. Effects of tegaserod (0.02 mg/kg bwt i.v. b.i.d. for 2 days) or vehicle on intestinal transit of barium-filled spheres, defaecation and gut sounds were studied in a cross-over design. Spheres were given via stomach tube prior to the first dosing of tegaserod or vehicle. Faeces were collected every 3 h and spheres eliminated were identified radiologically in the faeces. RESULTS: Tegaserod significantly accelerated the gastrointestinal (GI) transit time of spheres and increased the frequency of defaecation and scores of gut sounds compared to vehicle. The compound was well tolerated; no effects on behaviour, body temperature, heart rate, respiratory rate and clinical laboratory data were observed. CONCLUSIONS: Tegaserod efficaciously stimulated motility and accelerate GI transit in healthy horses. POTENTIAL RELEVANCE: Tegaserod may offer therapeutic potential in horses suffering from impaction or paralytic ileus.     

Enhanced plasma availability of moxidectin in fasted horses

The influence of fasting prior to treatment on plasma availability and kinetic disposition of moxidectin was studied in horses. Eight adult crossbred saddle horses were allocated to two experimental groups of four horses. One group was fasted for 24 hours before treatment and the other group received their usual feed. Both groups were treated with an oral gel formulation of moxidectin at the manufacturers recommended therapeutic dose of 0.4 mg/kg. Blood samples were collected by jugular puncture at different times between 0.5 and 50 days. After plasma extraction and derivatization, samples were analyzed by HPLC with fluorescence detection. Computerized pharmacokinetic analyses were done. Fasting induced marked modifications in the pharmacokinetic behavior of moxidectin in the horse. An extended absorption process, and significantly higher area under the curve (AUC) values were obtained in the fasted animals compared to those fed their usual rations. It is suggested that fasting decreases intestinal transit time, and may have prolonged time for absorption of moxidectin. Since, biliary excretion and intestinal secretion are major routes of moxidectin elimination, fasting may have reduced bile flow and intestinal secretion. Fasting before treatment may be a useful tool for improved anthelmintic treatment in horses.     

Effect of phenytoin on the clinical signs and in vitro muscle twitch characteristics in horses with chronic intermittent rhabdomyolysis and myotonia

In vitro twitch characteristics of the semimembranosus muscle were evaluated in 9 clinically normal horses, in 15 horses with chronic intermittent rhabdomyolysis (CIR) and in 2 horses with myotonia. Effects of phenytoin on in vitro muscle twitch and clinical signs of CIR and myotonia were evaluated in these same horses. Times to 90% relaxation were prolonged in the horses with CIR (mean +/- SEM, 186 +/- 5.9 ms) and in 2 horses with myotonia (197 and 177 ms) compared with those in clinically normal horses (mean +/- SEM, 146 +/- 2.1 ms). Horses with CIR also had significantly (P less than 0.05) longer times to 50% relaxation, compared with clinically normal horses. In the group of horses with CIR, Standardbreds had significantly (P less than 0.05) longer times to 90% and 50% relaxation, compared with Thoroughbreds. Times to 100% peak tension did not differ among the groups. Administration of phenytoin directly into a muscle preparation bath solution had no effect on muscle twitch properties. After the initial muscle biopsy, phenytoin was administered orally for 7 to 10 days to 4 horses with CIR, 2 myotonic horses, and 2 clinically normal horses before repeat biopsy from the same site in the contralateral semimembranosus muscle. Times to 90% relaxation decreased from 197 and 177 ms to 144 and 126 ms, respectively, in the 2 myotonic horses, from a mean of 192 (+/- 9) ms to 170 (+/- 9) ms in the 4 horses with CIR and remained unchanged (154 and 140 ms before vs 155 and 139 ms after treatment) in the 2 clinically normal horses.(ABSTRACT TRUNCATED AT 250 WORDS)     

Evaluation of the effects of the opioid agonist morphine on gastrointestinal tract function in horses

OBJECTIVE: To evaluate the effects of morphine administration for 6 days on gastrointestinal tract function in healthy adult horses. ANIMALS: 5 horses. PROCEDURES: Horses were randomly allocated into 2 groups in a crossover study. Horses in the treatment group received morphine sulfate at a dosage of 0.5 mg/kg, IV, every 12 hours for 6 days. Horses in the control group received saline (0.9% NaCl) solution at a dosage of 10 mL, IV, every 12 hours for 6 days. Variables assessed included defecation frequency, weight of feces produced, intestinal transit time (evaluated by use of barium-filled spheres and radiographic detection in feces), fecal moisture content, borborygmus score, and signs of CNS excitement and colic. RESULTS: Administration of morphine resulted in gastrointestinal tract dysfunction for 6 hours after each injection. During those 6 hours, mean +/- SD defecation frequency decreased from 3.1 +/- 1 bowel movements in control horses to 0.9 +/- 0.5 bowel movements in treated horses, weight of feces decreased from 4.1 +/- 0.7 kg to 1.1 +/- 0.7 kg, fecal moisture content decreased from 76 +/- 2.7% to 73.5 +/- 2.9%, and borborygmus score decreased from 13.2 +/- 2.9 to 6.3 +/- 3.9. Mean gastrointestinal transit time was also increased, compared with transit times in control horses. CONCLUSIONS AND CLINICAL RELEVANCE: Morphine administered at 0.5 mg/kg twice daily decreased propulsive motility and moisture content in the gastrointestinal tract lumen. These effects may predispose treated horses to development of ileus and constipation.     

Pharmacokinetics of ketoprofen in healthy horses and horses with acute synovitis

The pharmacokinetlc properties of a single intravenous dose of ketoprofen (2.2 mg/kg) in plasma and synovial fluid were compared in four healthy animals and four horses with experimentally induced acute synovitis. Synovitis was induced by the injection of a 1% solution of sterile carrageenan into the left intercarpal joint Ketoprofen was administered at the same time as carrageenan infection. The plasma disposition followed a biexponential equation or a two-compartment model in most horses. The plasma harmonic mean half-life in healthy horses (0.88 h) was longer than in horses with synovitis (0.5 5 h). Synovial fluid concentrations of ketoprofen in healthy horses approximated those in plasma by 3 h post-dose. In horses with synovitis, synovial fluid concentrations approximated plasma concentrations by 1 h. Synovial fluid concentrations of ketoprofen in horses with synovitis were 6.5 times higher than those in healthy horses at 1 h. The area under the synovial fluid concentration curve for horses with synovitis was greater than in healthy horses. These data suggest that the inflamed joint serves as a site of sequestration for ketoprofen. Furthermore, these results indicate that plasma pharmacokinetics may be altered by inflammation in a peripheral compartment such as the joint     

Intradermal testing of horses with chronic obstructive pulmonary disease and recurrent urticaria.

Six horses with chronic obstructive pulmonary disease (COPD) and 8 horses with recurrent urticaria were skin tested with 67 extracts from 58 allergens, including pollens, epidermals, cultivated farm plants, dusts, molds, and insects. Reactions were evaluated 3 times over a 24-hour period immediately after the injections. Results were compared with those obtained from 11 clinically normal horses. All horses had positive skin test reactions. Significant differences was evident between horses with COPD and clinically normal horses for only 3.0% of the possible extract reactions, and between horses with urticaria and clinically normal horses for only 4.5% of the possible extract reactions. Horses with COPD or urticaria had greater total percentage of allergen extract reactions than did clinically normal horses. Positive reactions were observed at all 3 evaluation periods, and late-onset reactions were not always preceded by positive reaction at earlier periods. All horses with COPD or urticaria had at least 1 skin test reaction that exceeded the mean +/- 2 SD, as calculated for each of the 67 extracts for the group of clinically normal horses.     

Comparison of hemodynamic, clinicopathologic, and gastrointestinal motility effects and recovery characteristics of anesthesia with isoflurane and halothane in horses undergoing arthroscopic surgery

OBJECTIVE: To compare hemodynamic, clinicopathologic, and gastrointestinal motility effects and recovery characteristics of halothane and isoflurane in horses undergoing arthroscopic surgery. ANIMALS: 8 healthy adult horses. PROCEDURE: Anesthesia was maintained with isoflurane or halothane (crossover study). At 6 intervals during anesthesia and surgery, cardiopulmonary variables and related derived values were recorded. Recovery from anesthesia was assessed; gastrointestinal tract motility was subjectively monitored for 72 hours after anesthesia. Horses were administered chromium, and fecal chromium concentration was used to assess intestinal transit time. Venous blood samples were collected for clinicopathologic analyses before and 2, 24, and 48 hours after anesthesia. RESULTS: Compared with halothane-anesthetized horses, cardiac index, oxygen delivery, and heart rate were higher and systemic vascular resistance was lower in isoflurane-anesthetized horses. Mean arterial blood pressure and the dobutamine dose required to maintain blood pressure were similar for both treatments. Duration and quality of recovery from anesthesia did not differ between treatments, although the recovery periods were somewhat shorter with isoflurane. After isoflurane anesthesia, gastrointestinal motility normalized earlier and intestinal transit time of chromium was shorter than that detected after halothane anesthesia. Compared with isoflurane, halothane was associated with increases in serum aspartate transaminase and glutamate dehydrogenase activities, but there were no other important differences in clinicopathologic variables between treatments. CONCLUSIONS AND CLINICAL RELEVANCE: Compared with halothane, isoflurane appears to be associated with better hemodynamic stability during anesthesia, less hepatic and muscle damage, and more rapid return of normal intestinal motility after anesthesia in horses undergoing arthroscopic procedures.     

Comparative efficiency of moxidectin gel or ivermectin paste for cyathostome control in young horses

Thirty-six young horses were allocated to three similar groups. Horses in Group 1 were treated with moxidectin gel on Days 0, 90, and 180, Group 2 horses received ivermectin paste on Days 0, 60, 120, and 180, and horses in Group 3 were untreated controls. All horses were maintained on a common pasture for the first 180 days. Immediately after the final scheduled deworming, each group was moved to a separate, clean pasture where it remained until Day 360. At monthly intervals, fecal egg counts, body weights, body condition scores, and pasture larval counts were measured. The cumulative costs of both deworming regimens were calculated. Young horses treated three times at 90-day intervals with moxidectin gel had significantly lower monthly fecal egg counts than untreated controls from Days 30 through 300. Horses given ivermectin paste four times at 60-day intervals had significantly lower egg counts than controls 30 days after each treatment and 60 days after the third dose. Average daily gains of treated horses were significantly greater than controls from Days 120 through 360 (moxidectin) and from Days 210 through 360 (ivermectin). Quarterly moxidectin treatments reduced egg counts more effectively and cost less than ivermectin given bimonthly.     

Clinical relevance of monocyte procoagulant activity in horses with colic

Endotoxin-activated monocytes express a thromboplastin-like procoagulant activity on the cell surface that may serve as a focal point for formation of microvascular thrombi. Because coagulopathy is a common sequela to endotoxemia in the equine species, we investigated the ability of monocytes, isolated from horses with colic, to express procoagulant activity. On the day of admission, and on the third and fifth day of hospitalization, monocytes were isolated from 30 adult horses with colic. A coagulation profile, including prothrombin time, activated partial thromboplastin time, thrombin time, and plasma fibrinogen and serum fibrin degradation products concentrations, was determined at each sample collection. The concentration of endotoxin in the plasma was quantitated at the time of admission. Ten clinically normal adult horses served as controls. The procoagulant activity of monocytes isolated from horses with colic was significantly (P less than 0.05) greater than that of the monocytes isolated from clinically normal horses. On the first and third day of hospitalization, the mean prothrombin time was significantly (P less than 0.05) longer in horses with colic, compared with clinically normal horses, and was the most common abnormality in the coagulation profile on the day of admission (25/30; 83%). Mean fibrin degradation products concentration was significantly (P less than 0.05) greater in horses with colic on the day of admission and was the second most common abnormality in the coagulation profile on day 1 (23/30; 77%). In horses with colic, the mean prothrombin and activated partial thromboplastin times were significantly (P less than 0.05) longer in horses that did not survive, compared with horses that survived.(ABSTRACT TRUNCATED AT 250 WORDS)     

Treatment of horses with chronic diarrhea: immunologic status.

All chronically diarrheal horses given (orally) 2 series of treatments with normal horse serum recovered in 2 to 4 weeks. However, mild diarrhea sometimes persisted several months in the group of horses with severe diarrhea. Weight gains were approximately 35% in horses with severe diarrhea and approximately 10% in horses with mild diarrhea. Serum specimens from 12 diarrheal and 20 normal horses were examined for immunoglobulins by single radial immunodiffusion technique. Concentration of immunoglobulin A in serum of diarrheal horses was approximately 50% lower than that in serum of normal horses. By contrast, there was more immunoglobulin G in serum of diarrheal horses than in serum of normal horses. Phytohemagglutinin (PHA-M) responsiveness of blood lymphocytes showed transient suppression during the stage of severe diarrhea. The regaining of PHA-M responsiveness of lymphocytes was observed simultaneously with the recovery process. However, the responsiveness of lymphocytes in recovered horses remained markedly lower than that in normal horses. Allergic reactions in diarrheal and normal horses were studied by observing dermal response to injections of saline extracts from some of the horse feeds. A delayed hypersensitivity reaction to streptokinase-streptodornase and PHA-M was also studied. Allergic reactions to these extracts were not induced in either diarrheal or normal horses; however, inflammatory response to the extracts was approximately 50% greater in normal than in diarrheal horses. Response to intradermal injection of either streptokinase-streptodornase or PHA-M was significantly greater in normal horses than in diarrheal horses.     

Large intestinal capacity, retention times, and turnover rates of particulate ingesta associated with extensive large-colon resection in horses

Fecal excretion of a particulate marker, ytterbium (Yb), was evaluated in 9 horses before surgery and 3 weeks, 3 months, and 6 months (4 trials) after sham-operation (group 1; n = 3) or extensive large colon resection (group 2; n = 6). Fecal excretion curves of total Yb excretion, loge Yb excretion, % Yb excretion, loge % Yb excretion, and cumulative % Yb excretion were evaluated, and kinetic analysis was performed on the loge Yb excretion curves to detect mixing pools and to calculate the fractional rate of particulate passage, turnover rate, and pool size. Calculations were performed to determined transit time, mean overall retention time, adjusted mean retention time, peak time, and disappearance time. Values were statistically analyzed to determine differences between groups and among trials (P less than 0.05). Group-2 horses had significantly shorter transit, peak, and mean overall retention times, compared with preoperative values and with values for group-1 horses. Two mixing pools were identified: a slower emptying pool of 5.7% hour-1 (k1) and a faster emptying pool of 12.3% hour-1 (k2). The rate of passage from the first pool (k1) was not altered by colon resection, and was interpreted as being most influenced by the cecum. In further support of this interpretation, the capacity of the k1 pool approximated the capacity of the cecum (17 L). The capacity of the k1 pool significantly expanded by 6 months in the resected horses. the rate of passage from the second pool (k2) significantly increased initially after colon resection (3 weeks and 3 months), but returned to preoperative values by 6 months.(ABSTRACT TRUNCATED AT 250 WORDS)     

Prekallikrein deficiency in a family of miniature horses

Two sibling miniature horses, a male and a female, had a normal 1-stage prothrombin time and a prolonged activated-partial thromboplastin time (APTT). The addition of as little as 5% of a normal equine plasma pool to the plasma samples of both horses shortened their prolonged APTT to within normal limits. Coagulation factor analysis revealed deficiencies in factor XII (12 and 13 U/dl, control population 77 to 128 U/dl), when determined with a feline factor XII-deficient plasma substrate, but normal concentrations (119 and 96 U/dl) when a human factor XII-deficient plasma substrate was used. Deficiencies of another factor, prekallikrein, were detected with a human prekallikrein-deficient plasma substrate (16 and 6 U/dl, control population 70 to 173 U/dl). Other intrinsic coagulation factors were present in normal concentrations. The APTT was measured with plasma from the 2 horses after various incubation periods (1 to 15 minutes) with a contact activator before the addition of Ca ions. With incubation times of greater than or equal to 10 minutes, the APTT of both horses were essentially the same as that of the normal equine plasma pool. Several family members of the 2 prekallikrein-deficient miniature horses appeared to be heterozygous carriers of the prekallikrein deficiency.     

A subjective descriptive study of the warm‐up and turn to a fence,approach, take‐off,suspension, landing and move‐off in 10 showjumpers

There is limited knowledge about causes of musculoskeletal injury in showjumpers. The objectives were to describe features of the turn, approach and jump in a group of experienced showjumpers believed by their riders to be sound, to relate these findings to clinical findings, and to identify features that may predispose to injury. Ten experienced showjumpers in normal competition training jumped an upright and a parallel fence four times off the left and right reins respectively, after a rider‐defined period of warm‐up. Real‐time and high‐speed motion capture was undertaken. Detailed subjective assessment of the gait was performed during the warm‐up, on the turn and approach to the jump and all phases of the jump. Six horses had thoracolumbar pain or epaxial muscle tension. Six horses exhibited a poor‐quality canter. The mean duration of warm‐up was 7 min (range 5–10 min). All horses had lean of the trunk and hindlimbs >45° on the turn. The inside hindlimb was placed in front of the outside hindlimb in 75% of the turns. Sideways oscillations of the hocks during stance were seen in all horses on the turn and on the straight approach in five horses. The inside hindlimb had greater magnitude of oscillation than the outside hindlimb on the turns. Repeated asynchronous push‐off with the hindlimbs at take‐off was seen in three horses. The hindlimbs drifted to the left or right during the ascent‐phase of suspension in four horses, reflecting asymmetrical push‐off. Only two horses landed consistently with the correct forelimb leading relative to the direction in which the horse had to turn after the fence. Four horses landed seven (n = 2) or eight (n = 2) times each with the left (n = 2) and right (n = 2) forelimbs respectively. Repetitive overload through asymmetrical use of the left and right canter leads, inadequate warm‐up, and limb instability could potentially predispose to injury.     

The Effect of Exercise on Equine Gastric Ulcer Syndrome in the Thoroughbred and Standardbred Athlete

Thoroughbred and Standardbred horses in race training were selected at random and examined for gastric ulcers. The mucosa of the stomach was examined by endoscopy and was scored for ulcers on a scale of 0 (no ulcers, normal mucosa) to 3 (severe ulceration with at least five deep lesions). A total of 79 horses were examined; 37 of these horses had one to four follow-up examinations, and 42 horses had only an initial gastroscopic examination. On the first examination, 33 of 79 (42%) horses had normal mucosa (ulcer score = 0), and 22 of 79 (28%) had ulcers with an ulcer score of 2 or greater. Seven horses (9%) had severe gastric ulceration (ulcer score = 3). The average ulcer score for all 79 horses on initial examination was 1.1 and 1.2 for the 37 horses with follow-up gastroscopic examination. Males (including male castrates) had slightly higher ulcer scores (1.2) than females (1.0), and 2-year-old horses had lower ulcer scores (0.7) than horses older than 3 years (1.4), but these differences did not reach statistical significance. Each of these average scores was significantly (P < .05) greater than zero. For the 37 horses with follow-up examinations, the males had higher scores (1.6) than females (0.7), and 2-year-old horses had higher ulcer scores (1.0) than 3-year-old horses (0.7) and lower ulcer scores than horses older than 3 (1.8), which were statistically significant. Follow-up examinations revealed ulcer scores ranging from 0 to 3; 23 of 37 (63%) horses had an ulcer score of 2 or greater, and 4 of 37 (12%) horses had severe ulcer disease. Average ulcer scores were significantly higher at subsequent examinations than at the initial examination. Risk of gastric ulceration increases with the length of time horses are in race training.     

RADIOGRAPHY AND COMPUTED TOMOGRAPHY IN THE DIAGNOSIS OF NONNEOPLASTIC EQUINE MANDIBULAR DISEASE

We compared the information gained from computed tomography (CT) vs. radiography in horses with nonneoplastic disease of the mandible. We hypothesized that CT would provide additional diagnostic information. Medical records, radiographs, and CT images of horses with nonneoplastic mandibular disease evaluated between 1994 and 2008 were reviewed. Nineteen horses were identified; 11 had a tooth root abscess and related disease, four had a fracture of the teeth and/or mandible, and four had a nonneoplastic mass. Both CT images and radiographs allowed identification of diseased teeth that appeared clinically normal otherwise. CT allowed identification of teeth that were clinically affected but appeared normal radiographically. Parameters such as tooth pulp involvement, lamina dura destruction, presence of bone fragments, lingual and buccal mandibular bone periosteal reaction, and cortical bone destruction were more conspicuous with CT. Performing radiography and CT in horses with nonneoplastic mandibular disease provides a more complete evaluation than either technique alone. CT contributes additional information that could otherwise be overlooked with radiographs alone in horses with a mandibular fracture. CT provides ancillary information to radiographs in horses with dental infection or a nonneoplastic mass of the mandible.     

Use of BAER to identify loss of auditory function in older horses

Objective To use the brainstem auditory evoked response (BAER) to test the hypothesis that auditory function could be worse in older horses than in younger horses. Procedure BAER waveforms in response to click stimuli were measured in five younger horses (5–8 years) and four older horses (17–22 years). Results Compared with the younger horses, the older horses showed significantly (P < 0.02) worse BAER thresholds and significantly (P < 0.02) worse BAER wave V amplitudes to the 90 decibels above normal hearing level stimulus. These results were consistent with partial deafness in the older horse group. Conclusion BAER assessment can be used to identify partial deafness in older horses. Such horses should be managed appropriately, with particular care taken in noisy environments where hearing loss could put the horse and/or its owner at risk of harm.     

Prevalence and risk factors in serous arthritis in the carpal joint of Norwegian coldblooded trotters

Of 508 four-year-old coldblooded trotters examined for carpitis, 135 (27%) had the condition. Thirty horses had unilateral carpitis on the left frontlimb, 52 on the right frontlimb and 53 had bilateral carpitis. Of these horses, 105 were stallions, 135 were geldings and 269 were mares. The prevalence of carpitis in these three groups was 25%, 25% and 28%, respectively.Information was obtained on the training regimes in four periods of the life of 378 horses. The odds ratios for the training variables were of the same order in all four periods. In logistic regression analyses on the training variables for the first period for the variable percentage of carpitis within each progeny group (sire index), the odds ratio corresponding to a 10% increase in sire index was 1.7 (1.4, 2.1) (95% confidence interval of the odds ratio). The odds ratios for leaving the horses out in a paddock, walking in a jogcart and fast training, or not in each case, were 0.4 (0.2, 0.9), 2.5 (1.2, 5.3) and 2.0 (0.9, 4.4), respectively. This implies a decreased risk of developing carpitis for horses left out in a paddock as a two-year-old, and a 2.5 and 2.0 times increased risk of carpitis occuring in horses exposed to walking in a jogcart and fast training, respectively.Information on the feeding regimes during the first, second and third year collected by a questionnaire revealed no association with carpitis beyond that which could be ascribed to chance. The rations offered were within normal ranges except for calcium, which was below recommended levels.