Laminitis is a debilitating disease in horses and commonly results in severe pain, lameness, and loss of athletic performance.1 The economic and welfare impacts of laminitis are high because of the associated poor prognosis, severe pain, and frequency of recurrence. Furthermore, the estimated incidence of laminitis ranges from 1.5% to 34%,2 with an estimated lifetime risk of 15%.3 Recognizing the impact of laminitis on horses and horse owners, veterinarians desire an improved understanding of laminitis, as evidenced in a 2009 surveya conducted by the AAEP in which its members identified laminitis as the highest priority for research funding and investigation.
The etiopathogenesis of laminitis is multifactorial and complex, comprised of inflammatory, metabolic, vascular, and traumatic pathways and processes.4,5 The term endocrinopathic laminitis has been used recently to describe laminitis in horses with hyperinsulinemia, PPID, or obesity or that have been administered glucocorticoids6 and has been recognized as the most common cause of laminitis in private equine practice.7,8 In the National Animal Health Monitoring System equine study9 performed in 2000, horse owners in the United States reported that pasture-associated laminitis and laminitis of unknown etiology were the most common forms of laminitis. Previous cohort studies10,11 have identified some risk factors (eg, hyperinsulinemia) for PEAL in ponies, and experimental euglycemic-hyperinsulinemic clamp studies have been shown to induce laminitis in ponies12 and horses.13 However, the importance of these and other risk factors for naturally occurring PEAL has not been characterized among North American horses.
Notable efforts have been made in the past decade to further the profession's understanding of this complex condition; however, much of the research on laminitis has been limited to the study of mechanistic pathways following experimental induction of the condition.12,13 Although valuable, experimental models may not fully replicate the multifactorial interactions underlying naturally occurring laminitis. Thus, observational studies of naturally occurring laminitis are necessary to advance our knowledge and understanding of the condition and associated risk factors as well as to design future investigations for its prevention and control.14 Furthermore, identification of modifiable risk factors for laminitis would improve our ability to reduce the burden of disease in the future. A 2012 systematic review by Wylie et al15 identified inconsistent and conflicting results regarding horses' risks of developing laminitis, further supporting the need for well-designed observational studies to improve the strength of evidence. Thus, the objective of the study reported here was to investigate risk factors for the development of PEAL in horses and ponies in North America.
Materials and Methods
Participant recruitment
North American members of the AAEP were recruited to participate in the study through announcements on the association's website and at its 2011 annual convention in San Antonio, Tex. A listing of 7,100 member veterinarians residing in the United States and Canada was obtained from the AAEP, and the members were contacted by electronic mail and postal services to solicit their participation in the study. Recipients were asked to indicate their willingness to participate through the study website or by direct contact with the study coordinator (MCC). Each respondent received a secure username and password for access to the study website, which contained all study instructions, documents, and electronic data entry forms. Each respondent was also mailed the study instructions (including detailed photographic instructions for obtaining body measurements and BCSs), definitions, questionnaires (Supplementary Appendix S1, available at avmajournals.avma.org/doi/suppl/10.2460/javma.253.4.470), blood-collection tubes, measuring tapes, and prepaid return envelope. Data were collected from January 2012 through December 2015. The study was approved by the Texas A&M University College of Veterinary Medicine and Biomedical Sciences Institutional Animal Care and Use Committee as well as the Clinical Research Review Committee. Client consent forms were obtained for all horses included in the study.
Case selection
The study was designed as a matched, case-control study in which participating veterinarians were asked to identify case-control sets that consisted of 1 horse with PEAL (cases) and 2 control horses without laminitis. Cases were defined as incident cases of PEAL with clinical signs detected ≤ 4 weeks prior to examination and collection of survey data (the timeframe for detection of clinical signs was extended from the initial requirement of ≤ 48 hours after detection). In addition, case horses were required to have evidence of bilateral forelimb lameness of Obel gradeb ≥ 2 and at least 2 of the following findings: sensitivity to hoof testers greatest in the region of the toe at the time of initial examination, a characteristic foundered stance, radiologic evidence of laminar thickening, and postmortem evidence (gross or microscopic) of laminitis. Horses were excluded as PEAL cases if they had any of the following findings: previous history of laminitis or navicular disease; laminitis associated with sepsis, a non–weight-bearing lameness, or excessive grain consumption; other concurrent disease conditions of the foot; radiologic signs of chronic laminitis (eg, extensive remodeling of the third phalanx); or gross evidence of chronic laminitis, such as divergent growth rings (founder rings) in the hoof capsule or if they were an equid other than a horse or pony.
Control selection
For each PEAL case, at least 1 horse from each of 2 control populations, healthy controls and lameness controls, was identified. A healthy control was defined as any healthy horse residing at a different farm than that of the PEAL case and ideally the next horse examined by the veterinarian for a routine wellness examination (eg, vaccination, Coggins testing, health certificate completion, or routine dental examination). A lameness control was defined as any horse with lameness in only 1 forelimb and residing at a different farm than the PEAL case. To qualify as a lameness control, the horse's lameness must have been present for ≤ 4 weeks and must have been graded as ≥ 3 according to the AAEP's 5-point lameness grading scale.16 Horses were excluded from the control groups if they had a history of laminitis or had clinical or diagnostic findings indicative of previous laminitis (eg, divergent hoof rings, dorsal hoof dishing, or preexisting radiologic evidence consistent with laminitis, such as rotation of the third phalanx relative to the hoof wall).
An a priori sample size estimation indicated that approximately 200 PEAL cases and 400 controls were required on the basis of a significance level of 5%, statistical power of 80%, an OR of 2 for PEAL cases relative to controls, and 2 controls/PEAL case.
Data collection
For each horse enrolled in the study, a questionnaire (Supplementary Appendix S1) was completed by the veterinarian and horse owner (or owner's agent) to capture the signalment (age, breed, and sex); activity type and level; housing and stable management; pasture exposure and characteristics; diet and feeding practices; BCS and morphometric measurements; physiological factors (eg, pregnancy and lactation); history of PPID, equine metabolic syndrome, or obesity as suspected or diagnosed by the primary veterinarian; hoof care; recent transportation > 4 hours; and history of corticosteroid treatment. Data were entered via the study website by responding veterinarians, or survey forms were completed and mailed to the study coordinator for data entry. All data were manually checked for apparent errors or discrepancies, which when found were resolved through direct communications between the submitting veterinarian and the study coordinator. The questionnaire was pilot tested prior to implementation, and as an incentive for veterinarian participation in the study, a $50 gift certificate was provided to each veterinarian submitting ≥ 1 case-control sets.
Morphometric measurements
A subjective BCS ranging from 1 (emaciated) to 9 (extremely fat) and based on the Henneke scale17 was determined by the submitting veterinarian for each animal. Objective measurements included height (in hands) at the withers (ie, the highest point of the scapulae) and circumference (in inches) at the girth, waist, and neck, and a girth-to-height ratio was calculated. Girth circumference was measured directly caudal to the olecranon and behind the withers. The maximal waist circumference was obtained by measuring body circumference two-thirds of the distance from the point of the shoulder to the point of the tuberis coxae. Neck circumference measurements were obtained while the neck was in a neutral position; the circumference was obtained at the midpoint of the neck, described as the midpoint from the poll to the withers. Photographs illustrating the measurement procedures were provided to all participating veterinarians.
Subjective morphometric measurements included an assessment of whether general or regional adiposity, alone or in combination, was present (yes vs no) and whether adiposity was present (yes vs no) at specific anatomic sites (tail head, ventrum, flank, thoracic region, and periorbital region). In addition, an assessment was made as to whether the horse had a cresty neck (yes vs no).
Data analysis
Descriptive data were generated for case horses and horses in the 2 control groups. Univariable analysis of all variables was performed to evaluate the strength and direction of association with the development of laminitis. Categorical variables were tested with χ2 or Fisher exact contingency tables, and continuous variables were compared with the Wilcoxon rank sum test. Continuous variables were recategorized into biologically plausible categories or quartiles. Contingency tables containing ≥ 1 cells with no entries were sensibly collapsed,18 and missing data were left as missing.
Data for case horses were compared with data for healthy control horses and again with data for lameness control horses by means of CLR (1:1 match), with the analyses conditional on matched sets contributed by individual veterinarians. All variables associated with PEAL at a value of P < 0.15 in univariable analyses were considered for inclusion in a multivariable model created by means of forward stepwise modeling. Variables selected into the final model were considered factors that best predicted the development of PEAL. All possible bivariable interactions among main effects variables significantly (P < 0.15) associated with PEAL were examined. Correlations among independent variables were assessed by means of the Pearson correlation coefficient, and when the Pearson correlation coefficient was > 0.8 for a pair of variables considered to measure similar exposures, 1 variable of the pair was selected for inclusion on the basis of biological plausibility and magnitude of association. Variables were considered to be confounded if the adjusted OR was different from the crude OR by at least 30%.19 The associations of individual variables with PEAL were expressed as ORs, and 95% CIs for the ORs were calculated with maximum likelihood methods. Commercially available softwarec was used for all analyses, and a value of P < 0.05 was considered significant.
Results
Veterinarian participation
Of the 7,100 AAEP member veterinarians in North America initially solicited, 625 veterinarians (599 from the United States and 26 from Canada) representing all 50 states and 6 Canadian provinces registered to participate in the study. Among the 625 registered veterinarians, 115 (18%) submitted data and 109 (17%) contributed data on horses eligible for inclusion in analysis (Figure 1).
Flow diagram illustrating recruitment of horses with PEAL (case horses) and matched control horses (healthy control horses and lameness control horses) in a study of risk factors for development of PEAL in horses and ponies in North America.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Flow diagram illustrating recruitment of horses with PEAL (case horses) and matched control horses (healthy control horses and lameness control horses) in a study of risk factors for development of PEAL in horses and ponies in North America.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Flow diagram illustrating recruitment of horses with PEAL (case horses) and matched control horses (healthy control horses and lameness control horses) in a study of risk factors for development of PEAL in horses and ponies in North America.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
A total of 199 horses with PEAL that were each matched with at least 1 healthy control (n = 198) or lameness control (153) horse were included in the analysis. These 550 horses were located in 32 states and 3 Canadian provinces (Figure 2), and their corresponding data were submitted to the study over a 4-year period (2012, n = 174 [32%] horses; 2013, 177 [32%]; 2014, 91 [17%]; and 2015, 108 [20%]).
Geographic distribution (32 states and 3 Canadian provinces) of matched PEAL cases (n = 199) and corresponding healthy controls (198) and lameness controls (153) for the study in Figure 1.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Geographic distribution (32 states and 3 Canadian provinces) of matched PEAL cases (n = 199) and corresponding healthy controls (198) and lameness controls (153) for the study in Figure 1.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Geographic distribution (32 states and 3 Canadian provinces) of matched PEAL cases (n = 199) and corresponding healthy controls (198) and lameness controls (153) for the study in Figure 1.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Study population
Of the 550 total horses, 281 (51%) were geldings, 251 (46%) were mares, and 18 (3%) were stallions (Supplementary Table S1, available at avmajournals.avma.org/doi/suppl/10.2460/javma.253.4.470). Age was known and reported for 547 horses, with an age range of 1 to 34 years, a mean of 13.3 years, and a median of 13 years. Neither sex nor age was a risk factor for development of laminitis. Breeds included Quarter Horses, American Paint Horses, and Appaloosas combined (n = 237 [43%]); Thoroughbreds (70 [13%]); draft horses and European warmbloods combined (66 [12%]); Arabians (48 [9%]); gaited horses, including Tennessee Walking Horses and Saddlebreds (36 [7%]); ponies and miniature horses (29 [5%]); and Morgans (17 [3%]). Breed was not reported for 8 (1%) horses and was reported as other for 39 (7%) horses.
Season of onset of clinical signs was recorded for 196 of the 199 horses with PEAL (winter, n = 27 [14%]; spring, 66 [34%]; summer, 65 [33%]; and fall, 38 [19%]; Figure 3). Horses were significantly (P < 0.001) more likely to develop PEAL in the spring or summer, compared with fall or winter.
Seasonal distribution for the onset of clinical signs associated with PEAL as reported for 196 of the 199 (98%) case horses for the study in Figure 1. Horses were significantly (P < 0.001) more likely to develop laminitis in the spring or summer, compared with the fall or winter.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Seasonal distribution for the onset of clinical signs associated with PEAL as reported for 196 of the 199 (98%) case horses for the study in Figure 1. Horses were significantly (P < 0.001) more likely to develop laminitis in the spring or summer, compared with the fall or winter.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
Seasonal distribution for the onset of clinical signs associated with PEAL as reported for 196 of the 199 (98%) case horses for the study in Figure 1. Horses were significantly (P < 0.001) more likely to develop laminitis in the spring or summer, compared with the fall or winter.
Citation: Journal of the American Veterinary Medical Association 253, 4; 10.2460/javma.253.4.470
The severity of lameness for each of the 199 horses with PEAL was assessed with the Obel grading system (grade 2, n = 121 [61%] horses; grade 3, 62 [31%]; and grade 4, 16 [8%]). Underlying causes of lameness were reported for 84 of the 153 (55%) horses in the lameness control group (soft tissue injury, n = 20 [13%]; foot pain, 18 [12%]; hoof abscess, 15 [10%]; laceration, 6 [4%]; arthritis, 5 [3%]; fracture, 5 [3%]; and unknown, 15 [10%]).
CLR modeling
Univariable analyses comparing data for case horses and healthy control horses (Supplementary Table S2, available at avmajournals.avma.org/doi/suppl/10.2460/javma.253.4.470) and comparing data for case horses and lameness control horses (Supplementary Table S3, available at avmajournals.avma.org/doi/suppl/10.2460/javma.253.4.470) identified a number of factors potentially associated with the development of PEAL. The final multivariable CLR model comparing data between case horses and healthy control horses (Table 1) indicated that horses with an overweight body condition (BCS ≥ 7), generalized or regional adiposity (alone or in combination), or a preexisting endocrinopathy, along with horses that did not receive concentrates in their diet, had higher odds of developing PEAL. Horses that had received corticosteroids within the 30 days prior to examination were more likely to develop PEAL (OR, 5.65; 95% CI, 1.32 to 24.27) than were horses that had not received corticosteroid treatment within the 30 days prior to examination. However, corticosteroid administration within that timeframe was uncommon in all groups of horses (case horses, n = 12/198 [6%]; healthy control horses, 4/198 [2%]; and lameness control horses, 1/151 [1%]) for which this variable was reported.
Results of multivariable CLR modeling to identify variables associated with PEAL in horses by comparing data for horses with PEAL to data on matched, healthy control horses.
| Variable | OR | 95% CI for OR | P value |
|---|---|---|---|
| BCS | 0.02 | ||
| < 7 | Referent | – | – |
| ≥ 7 | 3.24 | 1.22–8.58 | |
| Generalized or regional adiposity, alone or in combination | 0.002 | ||
| No | Referent | ||
| Yes | 5.21 | 1.86–14.55 | |
| Concentrates in diet | 0.01 | ||
| No | Referent | ||
| Yes | 0.84 | 0.73–0.96 | |
| Preexisting endocrinopathy | 0.02 | ||
| No | Referent | ||
| Yes | 5.65 | 1.32–24.27 |
The final multivariable CLR model comparing data for case horses and lameness control horses (Table 2) yielded similar results, with PEAL more likely in horses with an overweight body condition (BCS ≥ 7), generalized or regional adiposity (alone or in combination), or a preexisting endocrinopathy. There were no substantial bivariable interactions between any pairs of variables in either model.
Results of multivariable CLR modeling to identify variables associated with PEAL in horses by comparing data on horses with PEAL to data on matched horses with forelimb lameness other than laminitis (lameness controls).
| Variable | OR | 95% CI for OR | P value |
|---|---|---|---|
| BCS | 0.01 | ||
| < 7 | Referent | ||
| ≥ 7 | 4.73 | 1.45–15.37 | |
| Generalized or regional adiposity, alone or in combination | 0.01 | ||
| No | Referent | ||
| Yes | 3.65 | 1.23–10.79 | |
| Preexisting endocrinopathy | 0.04 | ||
| No | Referent | ||
| Yes | 2.71 | 1.7–7.28 |
Discussion
To our knowledge, the present study represents the first reported observational study of veterinarian-diagnosed incident cases of PEAL in North America. The use of incident cases in the present study was crucially important because identified risk factors were more likely to have been causal, whereas studies of prevalent cases identify factors that might be causal or that might be associated with surviving with the disease or recurrence of it. Therefore, the risk factors for PEAL identified in the present study may assist not only in managing and preventing this form of laminitis, but also in guiding future research into its pathogenesis.
During the 4-year period of data collection for the present study, PEAL was diagnosed most often during the spring and summer months. These findings conflicted with data from previous studies20–22 regarding seasonality, such as reports of increased odds of developing laminitis in the fall or decreased odds of developing laminitis during the summer.
Signalment factors, including age, sex, and breed, were evaluated as potential intrinsic risk factors for the development of laminitis in the present study and other studies.8,20,22,23 Although several previous reports8,20,23,24 have suggested a positive association between increasing age and risk of both acute and chronic laminitis, other reports22 have shown no association. The present study was limited to first-time cases of laminitis, and our findings did not associate age with odds of developing acute laminitis. Previous reports associating age with laminitis might indicate that increasing age is a risk factor for recurrent disease or merely an indicator for survival of affected horses (ie, cumulative prevalence increases with age). The authors' rationale for studying incident cases of PEAL was to detect risk factors associated with the development of the condition rather than risk factors associated with survival or longevity. Previous studies15 of the association between sex and laminitis have yielded inconsistent results, and no evidence of an association between sex and incident PEAL was observed in the present study.
For purposes of analysis, similar breeds (eg, American Paint Horses and Quarter Horses) were combined into groups to reduce the number of breed categories in the present study. In univariate analyses, the odds of laminitis were significantly (healthy controls, P < 0.002; lameness controls, P < 0.010) higher for the combined group of ponies and miniature horses, relative to the combined group of American Paint Horses, Quarter Horses, and Appaloosas. However, the precision and power of this estimate were limited because the combined group of ponies and miniature horses represented only 13% of PEAL cases and only 1% of healthy and lameness controls. The odds of PEAL appeared lower for Thoroughbreds as well as the combined group of draft horses and European warmbloods, relative to the odds for the combined group of American Paint Horses, Quarter Horses, and Appaloosas, in the present study. However, breed did not retain significance in either multivariable model. These findings were similar to those of a recent systematic review15 of risk factors for laminitis, which revealed inconsistent results between studies, with most studies finding no breed association.
Undergoing a stabling change within 14 days before the onset of clinical signs was significantly (healthy controls, P < 0.005; lameness controls, P < 0.002) associated with the development of PEAL in univariable analyses of data for case horses and both control populations. However, the importance of this finding is unknown. It was possible that this was an effect of PEAL rather than a cause (eg, horses with laminitis may have been more likely confined to a stall for purposes of disease management), especially given that this observation was also true for the lameness control horses. One might expect these horses to also have been more likely to have had a stabling change to greater stall confinement for management of their lameness. When further evaluated, neither an increased nor a decreased exposure to grass was notably associated with odds of developing PEAL, indicating that the change itself, but not a specific type of change, was associated with increased odds of developing PEAL.
It has long been proposed that access to abundant grass with high nonstructural carbohydrate content is a risk factor for development of laminitis, although clinical evidence to support this is limited. In a survey9 performed by the USDA in 2000, owners reported that 46% of horses developed laminitis as a result of exposure to lush pastures. In another study,22 horses with new access to grass, but no previous access during the 4 weeks prior, had greater odds of laminitis, compared with horses with no access or with prolonged prior access. In the present study, horses exposed to high-quality grass had greater odds of developing PEAL, compared with horses without grass exposure or with limited access to grass (when comparing cases with both control populations). However, this variable was confounded by BCS and was not retained in either of the final multivariable models.
Anecdotally, other dietary features, including feeding high concentrate diets, have been suggested to increase the risk of laminitis in horses. Interestingly, in the present study, the final multivariable model including cases and healthy controls indicated that horses receiving concentrates had lower odds of developing laminitis than did horses not fed concentrates; however, this too was confounded by the effects of BCS. In the authors' opinion, the most biologically plausible explanation for this finding was that owners or caregivers had recognized these horses as being easy-keepers and had eliminated unnecessary grain from their diet. This finding also suggested that interventions other than decreasing the amount of concentrates in the diet might have been needed to control body weight in these horses. Furthermore, specific information regarding the feeding practices of horses with PEAL should be evaluated to help guide intervention trials that could be designed to more accurately inform feeding recommendations.
Body morphometrics have been evaluated in previous studies,22,23 with an increased risk of laminitis being associated with decreasing height, generalized obesity, cresty neck, and a recent increase in body weight. In the present study, body morphometrics were extensively evaluated and consistently found to be associated with the risk of developing PEAL. Further, the association between obesity and insulin dysregulation might contribute to the development of laminitis, although the pathophysiologic mechanisms have not been fully elucidated.25,26 Alternatively, increased body weight might contribute mechanically to the development of laminitis. Results of the present study of incident cases of PEAL supported the causal association of obesity and laminitis because the obesity preceded the onset of laminitis. This strong association indicated that the risk of laminitis might be reduced by controlling obesity or modifying the underlying determinants of obesity. Careful feeding and management practices aimed at reducing body weight and adiposity should be considered. Although this concept is not new,27 results of the present study contributed evidence of the association of body morphometrics with the odds of developing laminitis and should be compelling information for convincing veterinarians and horse owners of the risks that obesity and increased adiposity pose relative to laminitis. In human medicine, causes of obesity not associated with increased caloric intake or decreased energy expenditure have been identified, resulting in novel methods of obesity control and prevention.28,29 Similarly and as noted previously, additional strategies other than dietary management and exercise might be needed in some cases to prevent obesity in horses.
Pituitary pars intermedia dysfunction, insulin dysregulation, and equine metabolic syndrome have been previously implicated as risk factors for development of laminitis,30 and a study7 in a primary-care setting in the United States as well as another study8 in a tertiary-care facility in Europe identified underlying endocrinopathies in most horses with laminitis. Further, a recent prospective, cohort study10 in the United Kingdom identified high basal serum insulin concentrations as a risk factor for development of pasture-associated laminitis. In the present study, horses with PPID, insulin dysregulation, equine metabolic syndrome, or a combination of these conditions had greater odds of developing laminitis; however, because the number of horses reported with any particular endocrine disorder was low, the presence of any endocrinopathy (rather than each type independently) was considered as a variable to gain statistical power for analysis. The low frequency of reported endocrinopathies in this population was surprising, but may have resulted from reporting bias (eg, uncertainty regarding medical history prior to evaluations for laminitis or variability in interpreting the questionnaire and whether further diagnostic testing was warranted before a horse was considered to have an endocrinopathy). More importantly, it was possible that laminitis might have been the first clinical sign of an endocrinopathy recognized by horse owners or veterinarians. These findings indicated that early recognition of endocrinopathies is vital to allow earlier intervention with medical treatment or husbandry management strategies to reduce the likelihood of affected horses developing PEAL.
Corticosteroid administration to horses has been implicated as inducing laminitis31; however, no direct evidence of a causal association has been identified. Although the use of corticosteroids in horses is widespread, the incidence of corticosteroid-induced laminitis was low in an observational study25 investigating risk factors for development of laminitis and in experimental studies32–34 of disease. Although the multivariable model comparing PEAL cases to healthy controls in the present study yielded a 13-fold greater odds of PEAL among horses that had received corticosteroids within the 30 days prior to examination relative to horses that had not, the validity and magnitude of this association must be considered with caution. Prior corticosteroid use did not remain in the final multivariable model comparing PEAL cases to lameness controls, suggesting that the association observed with healthy controls could have been confounded by another variable. More importantly, corticosteroid administration was uncommon in all groups of horses, and the small numbers rendered our estimates of the magnitude of an effect unstable, as reflected in the wide 95% CI. In addition, respondents' recall of corticosteroid administration may have been greater for horses with laminitis than for horses in the control groups, creating a potential for marked recall bias. Nevertheless, the authors believe that the results of the present study indicated that this topic merits further investigation through a well-designed, large-scale, hypothesis-driven observational study.
The results from the final multivariable models were similar for both sets of controls in the present study, with an obese body condition (ie, BCS ≥ 7), the presence of generalized or regional adiposity (alone or in combination), and a previously diagnosed endocrinopathy identified as risk factors for development of PEAL. In addition, the results supported the likelihood of a causal relationship between these variables and PEAL because the study was limited to first-time cases of PEAL. To the extent that these risk factors (eg, obesity, insulin dysregulation, and PPID) are potentially alterable through diet, exercise, medication, or other novel approaches,27 the authors consider the results of this study highly informative for laminitis prevention.
In addition to the variables retained in the final multivariable models, other variables that were notably associated with PEAL in univariable analyses merit consideration because of their association with covariates that were retained in the final multivariable model, indicating they could have been important too. For example, breed was associated with development of PEAL, although confounded by BCS or the presence of generalized or regional adiposity. For this reason, breed might have been an important risk factor for laminitis; however, adiposity or BCS was statistically superior and retained in the final model.
There were limitations to the present study. Because participation was voluntary and limited to veterinarian members of the AAEP, the impact on selection bias cannot be definitively determined. Furthermore, the results of the study were totally dependent on the accuracy of data reported by the participating veterinarians, and it was not feasible to validate reported observations for exposures (eg, diet) or outcomes (eg, laminitis). However, the present study had the advantage of including a relatively large number of veterinarians from diverse regions in North America.
The response rate of veterinarians for participation in the present study was low, which may have introduced additional selection bias. Nine percent of contacted veterinarians enrolled to participate (compared with 71% in a 2009 AAEP studya), and of this 9%, only 18% submitted data. Lack of participation might have been attributed to the length of the questionnaire, requirement of data on 3 horses, busy schedules of practicing veterinarians, requirement to identify incident cases of PEAL, and 4-week timeframe from onset of clinical signs.
Participation in the present study was heavily influenced by contributions from veterinarians in Texas, as 19% of cases were contributed by veterinarians in this state. It was unlikely that this was because PEAL was a greater problem in Texas than in other parts of the country, but rather because of the higher horse population in Texas combined with the potential that Texas veterinarians might have had a loyalty motivation to participate in a study coordinated through Texas A&M University. The effect of these potential response biases is unknown. Furthermore, the present study was limited to horses in North America, and the results should not be extrapolated globally because of regional differences in factors, such as diet, management, and breeds of horses.
Information bias associated with misclassifying cases and controls was another possible limitation of the present study. However, any misclassification could be expected to have been nondifferential and thus would have biased the results toward the null. However, diagnoses of PEAL for the present study were made by equine veterinarians rather than by horse owners, thereby likely reducing misclassification of cases and controls. The decision to exclude horses with laminitis that had only mild lameness (Obel grade 1) may have potentially reduced the ability to detect associations between risk factors and development of PEAL; however, the exclusion was deemed important to the design of the present study because mild laminitis can be challenging to detect and accurately diagnose. Furthermore, it was also probable that not all included cases of PEAL were incident cases because previous mild episodes of laminitis may have been unrecognized or unreported by owners.
Another limitation of the present study was the potential bias when selecting controls. Selection bias is a common bias inherent to case-control studies.35 To account for possible limitations of a single control population, 2 groups of controls were selected. Although this is not uniformly accepted as a superior method, compared with a single control group method, the validity of our findings was strengthened because the present study had similar results between the 2 groups of controls. Selection of cases may have also been biased by virtue of the name, PEAL, the use of which was not intended to include only horses with PEAL, but rather to exclude horses with laminitis attributed to grain overload, supporting-limb laminitis, or laminitis secondary to a septic process. On the basis of the low numbers of horses with previous endocrinopathy diagnoses, the authors suspected that the effect of this bias was limited.
Finally, as with any epidemiological study, the effects of confounding were considered. Cases and controls were matched on the basis of input from responding veterinarians in an effort to control for confounding on this variable. Bias introduced by measured variables were accounted for in the multivariable model; however, the association with PEAL may have been confounded by other factors that were not measured or considered.
In conclusion, the present observational study revealed several important risk factors that might contribute to the development of PEAL. A strategy to reduce the incidence of PEAL could involve further elucidating the determinants, identifying risk factors that are modifiable by medical management or husbandry changes, and educating horse owners and veterinarians about these factors and interventions. In addition to continued investigation of screening tests, treatments, and other interventions that can ameliorate insulin dysregulation and obesity, studies are warranted to identify determinants of obesity and adiposity as well as interventions for modifiable risk factors. It is also plausible that earlier recognition and treatment of endocrinopathies might contribute to reducing the incidence of the devastating disease of laminitis.
Acknowledgments
This manuscript represents a portion of the dissertation submitted by Dr. Coleman to the Texas A&M University College of Veterinary Medicine and Biomedical Sciences Department of Large Animal Clinical Sciences as partial fulfillment of the requirements for a PhD.
Funded by the American Association of Equine Practitioners Foundation.
The authors declare that there were no conflicts of interest.
Presented as an abstract at the AAEP Convention, Orlando, Fla, December 2016.
ABBREVIATIONS
| AAEP | American Association of Equine Practitioners |
| BCS | Body condition score |
| CI | Confidence interval |
| CLR | Conditional logistic regression |
| PEAL | Pasture- and endocrinopathy-associated laminitis |
| PPID | Pituitary pars intermedia dysfunction |
Footnotes
Klein K, AAEP Foundation, Lexington, KY: Personal communication, 2011.
Obel N. Studies on the histopathology of acute laminitis. PhD dissertation, Almqvist & Wiksells Boktryckeri AK, Uppsala, Sweden, 1948.
SAS, version 9.2, SAS Institute Inc, Cary, NC.
References
1. Herthel D, Hood DM. Clinical presentation, diagnosis, and prognosis of chronic laminitis. Vet Clin North Am Equine Pract 1999;15:375–394.
2. Wylie CE, Collins SN, Verheyen KL, et al. Frequency of equine laminitis: a systematic review with quality appraisal of published evidence. Vet J 2011;189:248–256.
3. Moore R. Vision 20/20—conquer laminitis by 2020. J Equine Vet Sci 2010;30:74–76.
4. Eades SC. Overview of current laminitis research. Vet Clin North Am Equine Pract 2010;26:51–63.
5. Heymering HW. 80 causes, predispositions, and pathways of laminitis. Vet Clin North Am Equine Pract 2010;26:13–19.
6. McGowan CM. Endocrinopathic laminitis. Vet Clin North Am Equine Pract 2010;26:233–237.
7. Donaldson MT, Jorgensen AJ, Beech J. Evaluation of suspected pituitary pars intermedia dysfunction in horses with laminitis. J Am Vet Med Assoc 2004;224:1123–1127.
8. Karikoski NP, Horn I, McGowan TW, et al. The prevalence of endocrinopathic laminitis among horses presented for laminitis at a first-opinion/referral equine hospital. Domest Anim Endocrinol 2011;41:111–117.
9. USDA. Lameness and laminitis in US horses. Report N318.0400. Fort Collins, Colo: USDA, APHIS, Veterinary Services, Center for Epidemiology and Animal Health, National Animal Health Monitoring System, 2000.
10. Menzies-Gow NJ, Harris PA, Elliott J. Prospective cohort study evaluating risk factors for the development of pasture-associated laminitis in the United Kingdom. Equine Vet J 2017;49:300–306.
11. Carter RA, Treiber KH, Geor RJ, et al. Prediction of incipient pasture-associated laminitis from hyperinsulinaemia, hyperleptinaemia and generalised and localised obesity in a cohort of ponies. Equine Vet J 2009;41:171–178.
12. Asplin KE, Sillence MN, Pollitt CC, et al. Induction of laminitis by prolonged hyperinsulinaemia in clinically normal ponies. Vet J 2007;174:530–535.
13. de Laat MA, McGowan CM, Sillence MN, et al. Equine laminitis: induced by 48 h hyperinsulinaemia in Standardbred horses. Equine Vet J 2010;42:129–135.
14. Menzies-Gow NJ. Laminitis epidemiology data: still severely lacking. Vet J 2011;189: 242.
15. Wylie CE, Collins SN, Verheyen KL, et al. Risk factors for equine laminitis: a systematic review with quality appraisal of published evidence. Vet J 2012;193:58–66.
16. Swanson TD. Guide for veterinary service and judging of equestrian events. 3rd ed. Golden, Colo: American Association of Equine Practitioners, 1984.
17. Henneke DR, Potter GD, Kreider JL, et al. Relationship between condition score, physical measurements and body fat percentage in mares. Equine Vet J 1983;15:371–372.
18. Hosmer DW, Lemeshow S, Sturdivant RX. Applied logistic regression. 3rd ed. Hoboken, NJ: John Wiley & Sons Inc, 2013.
19. Dohoo I, Martin W, Stryhn H. Veterinary epidemiologic research. 2nd ed. Charlottetown, PEI, Canada: VER Inc, 2010.
20. Polzer J, Slater MR. Age, breed, sex and seasonality as risk factors for equine laminitis. Prev Vet Med 1997;29:179–184.
21. Menzies-Gow NJ, Katz LM, Barker KJ, et al. Epidemiological study of pasture-associated laminitis and concurrent risk factors in the South of England. Vet Rec 2010;167:690–694.
22. Wylie CE, Collins SN, Verheyen KL, et al. Risk factors for equine laminitis: a case-control study conducted in veterinary-registered horses and ponies in Great Britain between 2009 and 2011. Vet J 2013;198:57–69.
23. Alford P, Geller S, Richrdson B, et al. A multicenter, matched case-control study of risk factors for equine laminitis. Prev Vet Med 2001;49:209–222.
24. Slater MR, Hood DM, Carter GK. Descriptive epidemiological study of equine laminitis. Equine Vet J 1995;27:364–367.
25. de Laat MA, McGowan CM, Sillence MN, et al. Hyperinsulinemic laminitis. Vet Clin North Am Equine Pract 2010;26:257–264.
26. Menzies-Gow N. Endocrinological aspects of the pathophysiology of equine laminitis. Equine Vet J 2012;44:735–737.
27. Menzies-Gow NJ. Endocrinopathic laminitis: reducing the risk through diet and exercise. Vet Clin North Am Equine Pract 2010;26:371–378.
28. Stenvinkel P. Obesity—a disease with many aetiologies disguised in the same oversized phenotype: has the overeating theory failed? Nephrol Dial Transplant 2015;30:1656–1664.
29. Walley AJ, Asher JE, Froguel P. The genetic contribution to non-syndromic human obesity. Nat Rev Genet 2009;10:431–442.
30. Johnson PJ, Messer NT, Ganjam VK. Cushing's syndromes, insulin resistance and endocrinopathic laminitis. Equine Vet J 2004;36:194–198.
31. Lose MP. Drug-induced laminitis in a colt. Mod Vet Pract 1980;61:608–610.
32. Tumas DB, Hines MT, Perryman LE, et al. Corticosteroid immunosuppression and monoclonal antibody-mediated CD5+ T lymphocyte depletion in normal and equine infectious anaemia virus-carrier horses. J Gen Virol 1994;75:959–968.
33. French K, Pollitt CC, Pass MA. Pharmacokinetics and metabolic effects of triamcinolone acetonide and their possible relationships to glucocorticoid-induced laminitis in horses. J Vet Pharmacol Ther 2000;23:287–292.
34. Dutton H. The corticosteroid laminitis story: 1. Duty of care. Equine Vet J 2007;39:5–6.
35. Lin IF, Paik MC. Matched case-control data analysis with selection bias. Biometrics 2001;57:1106–1112.
