Activities related to Thoroughbred horseracing (flat racing and breeding) in Latin America are important industries, and concern regarding the welfare of Thoroughbreds continues to grow. To the best of the authors’ knowledge, there are no published studies that identify risk factors for fatal injury or MSI in horses while racing in Latin America.
International studies have identified important risk factors for fatal injury and MSI worldwide. These include horse-level risk factors, such as age,1–10 sex,2,6,7,9–13 and body weight,12 and race-level risk factors, such as race distance,4,8–12,14 field size (ie, number of horses in the race),14,15 and racing surface and track.5,9,10,13 However, there are several clear differences between racing jurisdictions around the world, suggesting that, although some risk factors will be common to many, some variables will be unique or at least more important in different parts of the world than others. For example, in Argentina, from where data for the present study was provided, there are important differences in medication regulations, track surfaces, and training regimens, compared with many other parts of the world. It is therefore important to use local data and knowledge to develop region- or even track-specific models to identify risk factors and mitigate fatal and nonfatal injury in those particular settings or jurisdictions.
The prerace administration of phenylbutazone and furosemide to racehorses is authorized in specific horse races throughout the Americas. Phenylbutazone is a common NSAID frequently used as an analgesic and anti-inflammatory drug in horses with musculoskeletal conditions. Furosemide is a potent loop diuretic administered to horses before races to reduce or prevent exercise-induced pulmonary hemorrhage. Because of the resultant reduction in plasma volume following furosemide administration, it is believed that blood pressure in the lungs is also reduced and the risk of exercise-induced pulmonary hemorrhage is lowered.
In Argentina, the regulations surrounding prerace administration of phenylbutazone and furosemide have been modified. In 2009, prerace administration of phenylbutazone was forbidden for all horses < 3.5 years of age and in the higher level of races (eg, Group races or Listed races). In 2010, the age restriction was raised to horses < 4 years of age, and all higher-level Black-type races were included in the ruling. The purpose of the study reported here was to examine the association between prerace administration of phenylbutazone and the risk of MSI and fatal injury in Thoroughbred racehorses that raced between 2006 and 2015 at 2 of the 4 official racetracks in Argentina.
Materials and Methods
Study design and data acquisition
A retrospective cohort study design was used to conduct first univariable, then multivariable logistic regression analyses of 2 particular outcomes affecting racehorses during racing, MSI and fatal injury. Instances of MSI were defined as injuries that resulted in euthanasia of horses, retirement from racing, or a period ≥ 6 months away from racing and included fractures, joint luxations, and tendon and ligament injuries. Fatal injuries were defined as events that resulted in euthanasia or death of horses, such as fatal MSI or sudden death. All MSIs and fatal injuries were confirmed as such by racetrack veterinarians working at the racetracks from which data were received.
Data were retrieved from 2 of the 4 Organización Latinoamericana de Fomento del Sangre Pura de Carrera racetracks in Argentina (ie, racetracks A and B). Race start data were recorded in a database at each racetrack, and MSI and fatal injury reports were provided by official veterinarians from each racetrack.
For each race start made by a horse after prerace administration of furosemide, phenylbutazone, or both, a declaration must be made prior to the race to notify the racing authorities. Once this declaration is made, it becomes public information and is recorded in the racing system database.
Over the period of this study at the 2 racetracks, 32% of all race starts (89,443/283,193 starts) were made without a trainer's declaration that phenylbutazone or furosemide had recently been administered to the horse. For the remaining race starts, trainers declared horses to be racing after prerace phenylbutazone administration (with or without furosemide). Data on horses not racing after prerace phenylbutazone administration were obtained from both races in which medications were and were not permitted. Records relating to the exact treatment dates of either medication were not available. However, a declaration most often meant that medications were administered on the day of the race, the day or days before the race, or both.
The accuracy of veterinarians’ reports was checked by confirming injury classifications and validating information contained in reports against information contained in the relevant racetrack database. The studies were conducted with the outcome of interest being measured at the level of a race start (a start referring to a horse starting a race).
The total study population was determined by the number of years for which data were available at the 2 racetracks involved in the study. The analysis of risk factors for MSI involved 561 case starts and 282,632 control starts (359 case starts and 180,965 control starts at racetrack A, and 202 case starts and 101,667 control starts at racetrack B). A case start was defined as a start in a race after which the horse sustained a fatal or nonfatal MSI. Control starts were defined as any start in a race that did not end in a fatal or nonfatal MSI. Starts made by horses that sustained a fatal or nonfatal non-MSI during a race were excluded from the control population.
The analysis of risk factors for fatal injury involved 186 case starts and 283,007 control starts (127 case starts and 181,197 control starts at racetrack A, and 59 case starts and 101,810 control starts at racetrack B). A case start was defined as a start in a race after which the horse died or was euthanized. A control start was defined as any race start that did not end in the death or euthanasia of the horse.
Statistical analysis
Each variable was screened independently for each outcome. Continuous variables were also categorized to facilitate assessment of any relationship with the outcomes of interest. The Akaike information criterion was calculated for continuous variables and their categorized versions to determine the most appropriate form of the variables that would most improve the fit of the final model. Unless otherwise indicated by a substantially improved Akaike information criterion for categorical forms of continuous variables, a linear relationship between each continuous variable and the risk of MSI or fatal injury was assumed. Variables were selected for each multivariable model on the basis of a univariable value of P < 0.25.
Multivariable logistic regression models were developed to identify multiple risk factors for MSI and fatal injury. All variables were included in an automated stepwise logistic regression selection process to adjust for potential confounding that resulted in the development of 2 multivariable logistic regression models. Variables were retained in final multivariable models if their associated P value was < 0.05. Once final multivariable models were identified, all other variables, regardless of the P value identified at the time of the univariable logistic regression analysis, were added back into each model one at a time to examine any effect of confounding. If the addition of the potentially confounding variable resulted in a change of significance of one of the risk factors in the final multivariable model, or its inclusion changed the size of associated coefficients by > 30%, confounding was regarded as being present.16 Biologically plausible interaction terms were examined and included if they improved the overall fit of the models.
Results
The incidence of MSI in horses for which prerace phenylbutazone administration was or was not declared by the trainer to the racing authorities is shown (Figure 1). With the exception of 2009, the incidence of MSI was consistently lower in horses racing without prerace phenylbutazone administration, compared with horses racing after prerace phenylbutazone administration. During the study period, the incidence of MSI in horses racing with and without prerace phenylbutazone administration was 2.36/1,000 starts and 1.16/1,000 starts, respectively. The incidence of fatal injury in horses for which prerace phenylbutazone administration was or was not declared is shown (Figure 2). In all years, apart from 2006, the incidence of fatal injury was consistently lower in horses racing without prerace phenylbutazone administration, compared with horses racing after prerace phenylbutazone administration. Over the study period, the incidence of fatal injury in horses racing with and without prerace phenylbutazone administration was 0.75/1,000 starts and 0.46/1,000 starts, respectively.
Number (point estimate [95% CI]) of MSIs/1,000 starts in Thoroughbred racehorses that did (circles) or did not (squares) undergo prerace phenylbutazone administration at 2 racetracks in Argentina between 2006 and 2015.
Citation: Journal of the American Veterinary Medical Association 257, 6; 10.2460/javma.257.6.642
Number (point estimate [95% CI]) of MSIs/1,000 starts in Thoroughbred racehorses that did (circles) or did not (squares) undergo prerace phenylbutazone administration at 2 racetracks in Argentina between 2006 and 2015.
Citation: Journal of the American Veterinary Medical Association 257, 6; 10.2460/javma.257.6.642
Number (point estimate [95% CI]) of MSIs/1,000 starts in Thoroughbred racehorses that did (circles) or did not (squares) undergo prerace phenylbutazone administration at 2 racetracks in Argentina between 2006 and 2015.
Citation: Journal of the American Veterinary Medical Association 257, 6; 10.2460/javma.257.6.642
Number (point estimate [95% CI]) of fatal injuries/1,000 starts in Thoroughbred racehorses that did (circles) or did not (squares) undergo prerace phenylbutazone administration at 2 racetracks in Argentina between 2006 and 2015.
Citation: Journal of the American Veterinary Medical Association 257, 6; 10.2460/javma.257.6.642
Number (point estimate [95% CI]) of fatal injuries/1,000 starts in Thoroughbred racehorses that did (circles) or did not (squares) undergo prerace phenylbutazone administration at 2 racetracks in Argentina between 2006 and 2015.
Citation: Journal of the American Veterinary Medical Association 257, 6; 10.2460/javma.257.6.642
Number (point estimate [95% CI]) of fatal injuries/1,000 starts in Thoroughbred racehorses that did (circles) or did not (squares) undergo prerace phenylbutazone administration at 2 racetracks in Argentina between 2006 and 2015.
Citation: Journal of the American Veterinary Medical Association 257, 6; 10.2460/javma.257.6.642
Univariable analysis
Full details of the univariable analysis for MSI and fatal injury are available (Supplementary Appendix S1, available at: avmajournals.avma.org/doi/suppl/10.2460/javma.257.6.642). A total of 8 categorical variables were assessed in both analyses of which the following were considered for inclusion in the 2 final multivariable models: sex (included in both MSI and fatal injury models), declared authorized medication (both models), existence of previous racing injury in the database (MSI model), racing in the summer season (both models), racing surface (both models), track configuration (MSI model), and time of day (MSI model). Fifteen continuous variables were analyzed in both univariable analyses of which the following were considered for inclusion in the final multivariable models: race distance (both models), age of horse (both models), horse body weight (both models), total career starts (MSI model), field size (both models), cumulative distance raced prior to the current start (MSI model), number of layups (defined as a period ≥ 60 days) and total days of layups (MSI model), speed of winning horse in the race (MSI model), number of starts in the period 30 to 60 days prior to the current start (MSI model), and jockey race-day weight (MSI model).
The variable relating to the use of declared authorized medication was initially examined as a 4-category variable: no medication, phenylbutazone alone, furosemide alone, and both phenylbutazone and furosemide. Initial univariable analysis showed that, compared with no medication, prerace administration of furosemide on its own was not associated with the risk of either MSI or fatal injury, whereas prerace administration of phenylbutazone alone and of both phenylbutazone and furosemide were associated with an increased risk of both MSI and fatal injury. Additionally, the size of the increased risk (ie, the univariable OR) was similar for horses that had received phenylbutazone alone or both phenylbutazone and furosemide. Because the objective was to achieve a final multivariable model that included as few significant variables as possible, it was decided to collapse this variable into a binary form (ie, furosemide administration only or no drug administration vs phenylbutazone administration only or both phenylbutazone and furosemide administration).
Multivariable analysis for MSI
Horses for which prerace phenylbutazone administration was declared before a race had greater odds (OR, 1.45 [95% CI, 1.03 to 2.04]; P = 0.03) of MSI than horses racing without prerace phenylbutazone administration. Five-year-old horses (OR, 1.48 [95% CI, 1.19 to 1.83]; P < 0.001) and horses ≥ 6 years old (OR, 1.4 [95% CI, 1.06 to 1.86]; P = 0.02) had greater odds of sustaining MSI than 4-year-old horses. Odds of MSI increased (OR, 1.07 [95% CI, 1.04 to 1.09]; P < 0.001) for every extra 10 kg (22 lb) of horse body weight. Female horses had reduced odds (OR, 0.77 [95% CI, 0.64 to 0.94]; P = 0.008) of MSI, compared with male horses (stallions and geldings). Horses racing on turf had lower odds (OR, 0.63 [95% CI, 0.50 to 0.79]; P < 0.001) of MSI than horses racing on dirt. Horses racing in longer races had greater odds (OR, 1.06 [95% CI, 1.0 to 1.11]; P = 0.002) of MSI for each extra 100 m raced. Horses racing in summer had reduced odds (OR, 0.72 [95% CI, 0.56 to 0.92]; P = 0.009) of MSI, compared with horses racing in autumn. Horses that had more layups had reduced odds (OR, 0.87 [95% CI, 0.79 to 0.95]; P = 0.002) of MSI for each additional layup, but the odds of MSI increased (OR, 1.01 [95% CI, 1.0 to 1.02]; P = 0.05) as the layup length increased for each extra 15 days.
Multivariable analysis for fatal injury
Horses racing with declared prerace phenylbutazone administration had greater odds (OR, 1.59 [95% CI, 1.1 to 2.27]; P = 0.015) of sustaining a fatal injury than horses racing without prerace phenylbutazone administration. Odds of sustaining a fatal injury increased (OR, 1.07 [95% CI, 1. 2 to 1.13]; P = 0.012) for every extra 10 kg of horse body weight, and horses racing with a larger number of horses had greater odds (OR, 1.07 [95% CI, 1.2 to 1.13]; P = 0.012) of fatal injury for each extra horse in the race. Horses racing in longer races had greater odds (OR, 1.1 [95% CI, 1.03 to 1.18]; P = 0.006) of a fatal injury for each extra 100 m raced.
Discussion
In the present study, horses having recently received phenylbutazone medication were identified as being at significantly increased risk of MSI and fatal injury during racing; however, this finding does not imply a direct causal relationship between prerace administration of phenylbutazone and the risk of MSI or fatal injury. The finding does concur with that of Dirikolu et al,17 who demonstrated that plasma concentrations of phenylbutazone were higher in fatally and nonfatally injured racehorses than uninjured racehorses from the same race. However, those authors also recognized their inability to account for other potential risk factors and suggested that further work was required to confirm an association between NSAID usage and fatal and nonfatal injury.
Two studies18,19 did investigate the use of prerace administration of phenylbutazone and the risk of catastrophic MSI in Thoroughbreds and Quarter Horses. However, in both investigations, particularly that specific to Quarter Horses,19 the prevalence of prerace administration of phenylbutazone made identifying an association with the risk of injury difficult. In that study,19 all but 1 of 40 Quarter Horses that sustained a catastrophic MSI were recorded as having received prerace administration of phenylbutazone.19
In the present study, the association between MSI or fatal injury and recent phenylbutazone administration was apparent from the 2 plots (Figures 1 and 2) showing the risk of MSI or fatal injury in horses racing with or without prerace phenylbutazone administration. These data suggested that the risk of MSI in horses racing with prerace phenylbutazone administration was more than double that in horses racing without prerace phenylbutazone administration (2.36 MSIs/1,000 starts vs 1.16 MSIs/1,000 starts, respectively). Results of multivariable analysis revealed an increased risk for MSI (OR, 1.45 [95% CI, 1.03 to 2.04]) and an increased risk for fatal injury (OR, 1.59 [95% CI, 1.1 to 2.27]) in horses for which prerace phenylbutazone administration was declared, compared with horses that raced without prerace phenylbutazone administration. Important confounders in the present study that likely also influenced the risk of MSI or fatal injury included the fact that prerace medication administration was not allowed in younger racehorses and in horses entered in top-quality races. This effectively means that older and lower-quality horses were more likely to compete in races that permitted medication, and both of these factors have been reported as risk factors for fatal and nonfatal injuries in racehorses.1,2,4,6,7
The increased risk of MSI and fatal injury in horses competing after prerace administration of phenylbutazone is most likely attributed to the analgesic effect of the NSAID, enabling horses to continue to race in the presence of subclinical injury and pain. Phenylbutazone administration is used to allow horses to continue training and racing, potentially inadvertently resulting in further damage to their musculoskeletal structures and increasing the risk of fatal injury. This is consistent with many postmortem findings, which indicate the presence of preexisting lesions,20–24 and other epidemiological studies2,3,9,10 that show that previous injury is associated with an increased risk of subsequent fatal and nonfatal injury.
In the present study, associations were found between prerace administration of phenylbutazone and risk of MSI and fatal injury; however, these findings did not imply that prerace administration of phenylbutazone directly increased the risk of MSI or fatal injury. The use of NSAIDs has been studied in human athletes, in terms of its pain-masking effect and the risk of recurrence or even exacerbation of some injuries.25–28 Speed and Wolfarth28 examined the dilemma in humans between healing soft tissue injuries and returning to full activity, and the detrimental effect that pain masking with NSAIDs may have on the healing process. As for human athletes,28 the acronym REST (resume exercise below soreness threshold) should be followed in the management of soft tissue injuries in racehorses. It is the authors’ opinion that when dealing with racehorses, the precautionary principle of REST should be applied on their behalf, especially given that racehorses are less able to communicate their soreness threshold. In the present study, the associations found between prerace administration of phenylbutazone and risk of MSI and fatal injury in racehorses call into question the continued use of analgesic medications immediately prior to racing.
In the present study, a number of risk factors, in addition to prerace administration of phenylbutazone, were found to be significantly associated with MSI and fatal injury. Findings in the present study largely agreed with those of previous studies. Older horses are believed to be at greater risk of fatal and nonfatal injury as a result of the gradual accumulation of subclinical injury or, more specifically, microdamage of bony structures.29 Male horses are commonly found to be at greater risk of injury, compared with female horses.7,11,13 In most studies, this is hypothesized to be related to the potentially greater body weight of sexually intact males placing greater strain on musculoskeletal tissues.12 In the present study, horse body weight was included as well as sex in the model for MSI. As with other studies,5,9,10 horses racing on dirt, compared with turf, were at increased risk of MSI in the present study. The reasons for the increased risk of injury on dirt surfaces have not been conclusively determined, but it is thought that the greater amount of slipping and potentially increased concussive forces on dirt, compared with turf, may be responsible.14 Race distance is believed to be a risk factor for injury primarily because of a greater amount of time at risk,14 although it is also possible that fatigue does contribute to an increased risk towards the end of some longer races. The size of the field (ie, number of horses in the race) has also been identified as a risk factor in some studies,14,15 and it is hypothesized that larger fields may increase the competitive nature of the race. Season of the year has rarely been identified as a risk factor, primarily because its impact is largely driven by surface condition, which has been repeatedly associated with increased risk of injury in turf racing.3,4,14 Nevertheless, both surface conditions and season, as independent risk factors, have been identified in some studies.2,3
In the present study, an increase in the number of layups decreased the risk of MSI, but the number of layup days that accumulated over the career of a racehorse slightly increased the risk of MSI. It is plausible that the latter association reflects horses that have had substantial periods of time out of racing (most likely attributable to injury, which itself increases the risk of subsequent MSI) and the former association reflects horses more frequently rested by their trainer as part of a typical training regimen.
One limitation of the present study was the lack of information related to exercise protocols, medical histories, and medication records of racehorses while in training. Because of the retrospective analysis of existing records, it was not possible to collect further details about each racehorse, including the primary reason for prerace administration of phenylbutazone.
To the authors’ knowledge, the present report represented the first large-scale study to clearly demonstrate and quantify the association between prerace administration of phenylbutazone and fatal and nonfatal injury in racehorses. It is hoped that these results will be useful in the further development of equine welfare–oriented medication policies that seek to minimize risk to Thoroughbred racehorses during racing.
Acknowledgments
This manuscript represents a portion of a thesis submitted by Dr. Zambruno to the University of Glasgow School of Veterinary Medicine as partial fulfillment of the requirements for a Master of Veterinary Medicine degree.
Funding for Teresita Zambruno's research was provided by the Organización Latinoamericana de Fomento del Sangre Pura de Carrera.
ABBREVIATIONS
| MSI | Musculoskeletal injury |
References
1. Bailey CJ, Reid SWJ, Hodgson D, et al. Risk factors associated with musculoskeletal injuries in Australian Thoroughbred racehorses. Prev Vet Med 1997;32:47–55.
2. Perkins NR, Reid SWJ, Morris RS. Risk factors for injury to the superficial digital flexor tendon and suspensory apparatus in Thoroughbred racehorses in New Zealand. N Z Vet J 2005;53:184–192.
3. Reardon RJM, Boden LA, Mellor DJ, et al. Risk factors for superficial digital flexor tendinopathy in Thoroughbred racehorses in hurdle starts in the UK (2001-2009). Equine Vet J 2012;44:564–569.
4. Henley WE, Rogers K, Harkins L, et al. A comparison of survival models for assessing risk of racehorse fatality. Prev Vet Med 2006;74:3–20.
5. Mohammed HO, Hill T, Lowe J. Risk factors associated with injuries in Thoroughbred horses. Equine Vet J 1991;23:445–448.
6. Estberg L, Stover SM, Gardner IA, et al. Relationship between race start characteristics and risk of catastrophic injury in Thoroughbreds: 78 cases (1992). J Am Vet Med Assoc 1998;212:544–549.
7. Estberg L, Stover SM, Gardner IA, et al. Fatal musculoskeletal injuries incurred during racing and training in Thoroughbreds. J Am Vet Med Assoc 1996;208:92–96.
8. Lyle CH, Blissitt KJ, Kennedy RN, et al. Risk factors for race-associated sudden death in Thoroughbred racehorses in the UK (2000-2007). Equine Vet J 2012;44:459–465.
9. Georgopoulos SP, Parkin TDH. Risk factors associated with fatal injuries in Thoroughbred racehorses competing in flat racing in the United States and Canada. J Am Vet Med Assoc 2016;249:931–939.
10. Georgopoulos SP, Parkin TDH. Risk factors for equine fractures in Thoroughbred flat racing in North America. Prev Vet Med 2016;139:99–104.
11. Boden LA, Anderson G, Charles J, et al. Risk factors for Thoroughbred racehorse fatality in flat starts in Victoria, Australia (1989-2004). Equine Vet J 2007;39:430–437.
12. Takahashi T, Kasashima Y, Ueno Y. Association between race history and risk of superficial digital flexor tendon injury in Thoroughbred racehorses. J Am Vet Med Assoc 2004;225:90–93.
13. Hernandez J, Hawkins DL, Scollay MC. Race-start characteristics and risk of catastrophic musculoskeletal injury in Thoroughbred racehorses. J Am Vet Med Assoc 2001;218:83–86.
14. Parkin TDH, Clegg PD, French NP, et al. Race- and course-level risk factors for fatal distal limb fracture in racing Thoroughbreds. Equine Vet J 2004;36:521–526.
15. Parkin TDH, Clegg PD, French NP, et al. Risk factors for fatal lateral condylar fracture of the third metacarpus/metatarsus in UK racing. Equine Vet J 2005;37:192–199.
16. Dohoo I, Martin W, Stryhn H. Cohort studies. In: Veterinary epidemiologic research. 3rd ed. Charlottetown, PE, Canada: AVC Inc, 2003;168–177.
17. Dirikolu L, Woods WE, Boyles J, et al. Nonsteroidal anti-inflammatory agents and musculoskeletal injuries in Thoroughbred racehorses in Kentucky. J Vet Pharmacol Ther 2009;32:271–279.
18. Beisser AL, McClure S, Rezabek G, et al. Frequency of and risk factors associated with catastrophic musculoskeletal injuries in Quarter Horses at two Midwestern racetracks: 67 cases (2000–2011). J Am Vet Med Assoc 2014;245:1160–1168.
19. Beisser AL, McClure S, Wang C, et al. Evaluation of catastrophic musculoskeletal injuries in Thoroughbreds and Quarter Horses at three Midwestern racetracks. J Am Vet Med Assoc 2011;239:1236–1241.
20. Stover SM. Nomenclature, classification, and documentation of catastrophic fractures and associated preexisting injuries in racehorses. J Vet Diagn Invest 2017;29:396–404.
21. Diab SS, Stover SM, Carvallo F, et al. Diagnostic approach to catastrophic musculoskeletal injuries in racehorses. J Vet Diagn Invest 2017;29:405–413.
22. Janes JG, Kennedy LA, Garret KS, et al. Common lesions of the distal end of the third metacarpal/metatarsal bone in racehorse catastrophic breakdown injuries. J Vet Diagn Invest 2017;29:431–436.
23. Gray SN, Spriet M, Garcia TC, et al. Preexisting lesions associated with complete diaphyseal fractures of the third metacarpal bone in 12 Thoroughbed racehorses. J Vet Diagn Invest 2017;29:437–441.
24. DeLay J. Postmortem findings in Ontario racehorses, 2003–2015. J Vet Diagn Invest 2017;29:457–464.
25. Warden SJ. Prophylactic misuse and recommended use of nonsteroidal anti-inflammatory drugs by athletes. Br J Sports Med 2009;43:548–549.
26. Warden SJ. Prophylactic use of NSAIDs by athletes: a risk/benefit assessment. Phys Sportsmed 2010;38:132–138.
27. Mueller-Wohlfahrt HW, Haensel L, Mithoefer K, et al. Terminology and classification of muscle injuries in sport: the Munich consensus statement. Br J Sports Med 2013;47:342–350.
28. Speed C, Wolfarth B. Challenges of pain masking in the management of soft tissue disorders: optimizing patient outcomes with a multi-targeted approach. Curr Med Res Opin 2014;30:953–959.
29. Riggs CM. Aetiopathogenesis of parasagittal fractures of the distal condyles of the third metacarpal and third metatarsal bones—review of the literature. Equine Vet J 1999;31:116–120.
