Sixteen- to 20-month-old Thoroughbreds are frequently purchased at public auctions from July through September for the purpose of selling them for profit (ie, pinhooking) the following year (from February through April) at what are known as 2-year-old in-training sales.1 The price of yearling Thoroughbreds can vary considerably depending on their pedigree and conformation.2 Elite Thoroughbreds with a high-quality pedigree offered for sale during the first 4 days of a yearling sale at a Thoroughbred sales facilitya in September 2006 had a higher median sales price, compared with Thoroughbreds sold during the subsequent 10 days of the sale; median price of yearling Thoroughbreds sold ranged from $310,000 on first day of the sale to $4,600 on the last day (day 14) of the sale.2 After the yearling sales, trainers have approximately 5 months to prepare horses to train at high speed before the 2-year-old in-training sales. At the 2-year-old in-training sales, Thoroughbreds run 1 or 2 furlongs (1/8 or 1/4 mile, respectively [approx 200 or 400 m, respectively]) at race speeds to assess their athletic performance. After those sales, Thoroughbreds start racing as 2-year-olds.
Knowledge of yearling Thoroughbred sale purchase prices and factors affecting 2-year-old in-training sales prices (eg, lameness during training and speed of horses recorded at those sales) in the United States is limited. Results of a study3 conducted at a training center in Florida indicated lameness was the most common cause of loss of training days for yearling Thoroughbreds in preparation for the 2-year-old in-training sales. Results of another study1 conducted in Florida revealed that the incidence of lameness in horses was highest within 60 days prior to a sale, and joint injury was the most common cause of lameness among yearling Thoroughbreds entered in 2-year-old in-training sales. Typically, Thoroughbreds with a high-quality pedigree, good conformation, and a high yearling sale purchase price are sold for higher 2-year-old in-training sale purchase prices, compared with horses without these characteristics. Speed of a Thoroughbred can also influence the 2-year-old in-training sale purchase price.1
The relationships between training characteristics (eg, total distance exercised during training) and sales prices of Thoroughbreds sold at 2-year-old in-training sales have not been extensively investigated. Results of 1 study1 indicated that the number of furlongs galloped within the 60 days prior to a sale was lower for horses classified as fast at the sale, compared with horses classified as slow at the sale. Results of that study also indicated that horses classified as fast at the sale sold for a higher price (median, $65,000) than did horses classified as slow (median, $20,000). The reasons for this relationship were not identified in that study. Lameness can affect the training schedule of a yearling Thoroughbred via a decrease in number of training days3 or reduction in the total amount of exercise. However, to the authors' knowledge, the effect of yearling sale purchase price on training schedules of yearling Thoroughbreds with lameness is not known. The objectives of the study reported here were to determine the effects of yearling sale purchase price, exercise history, lameness, and athletic performance (ie, speed) classification of Thoroughbreds at 2-year-old in-training sales on sales prices of 2-year-old in-training Thoroughbreds and to compare total exercise distance of horses within 60 days prior to those sales between horses with high versus low purchase prices and between horses with lameness versus horses without lameness during training.
Materials and Methods
Animals—Fifty-one yearling Thoroughbreds purchased at a yearling salea in September 2006 that were trained in Florida for resale at 2-year-old in-training sales (February through April 2007) were included in the study. The median yearling sale purchase price of these horses was $55,000 (range, $7,000 to $725,000). All of the horses were shipped from Lexington, Ky, to a training center in Ocala, Fla, in late September 2006. Horses were vaccinated by an attending veterinarian at the training center against equine herpesvirus types 1 and 4, influenza virus types A1 and A2, West Nile virus, Eastern and Western equine encephalomyelitis viruses, Clostridium tetani, and Streptococcus equi; horses were vaccinated in October 2006, February 2007, and April 2007. That training center was selected because of the large number of horses in training and willingness of the trainer to participate in the study.
Training procedures—All of the horses were trained for riding and underwent similar exercise training programs starting in October 2006. Briefly, all horses trotted to the racetrack, galloped (approx 25 to 30 s/furlong) 1 to 1.5 miles (approx 1.6 to 2.4 km), and walked back to their stalls 6 d/wk; horses were not exercised 1 d/wk. Starting in December 2006, yearlings started to gallop 5 d/wk (Monday through Friday), underwent highspeed exercise (ie, breezing; 1 or 2 furlongs at approx 15 s/furlong) 1 d/wk (Saturday), and did not exercise 1 d/wk (Sunday).
Outcomes of interest—For one of the study objectives (determination of the effects of yearling sale purchase price, exercise history, lameness, and athletic performance [ie, speed] classification of Thoroughbreds at a 2-year-old in-training salea on sales prices of 2-year-old in-training Thoroughbreds), the outcome of interest was 2-year-old in-training sale purchase prices of horses in the study. For the other study objective (comparison of total exercise distance of horses within 60 days prior to sales between horses with high yearling sale purchase prices versus those with low yearling sale purchase prices and between horses with lameness versus those without lameness during training), the outcomes of interest were number of furlongs galloped and number of furlongs breezed by horses within 60 days prior to sales.
Demographic data—Demographic data were recorded for each horse. Demographic data recorded included yearling sale identification number, sex (colt or filly), history of arthroscopic surgery (yes vs no), yearling sale purchase price, date of arrival at the training facility, date on which a horse was first trained to ride, and date on which a horse was first trained on the racetrack at the training facility.
Exercise data—Exercise distance (ie, number of furlongs galloped and number of furlongs breezed during training) of each horse was recorded daily by the trainer during the training period. Only data for the 60 days prior to the 2-year-old in-training sale for each horse were included in the study because the incidence of lameness in yearling Thoroughbreds is highest during that period.1
Identification of lameness—Horses were identified as lame by a rider or the trainer during or immediately after exercise (galloping or breezing) on the racetrack. Horses identified as lame by a rider or the trainer were examined by an attending veterinarian, and a lameness grade (0 to 5 on the basis of the American Association of Equine Practitioners lameness scoring system4) was determined. Horses with a lameness grade ≥ 2 were classified as lame in the present study. Data regarding anatomic locations of lameness and results of radiographic or ultrasonographic examinations were not available for inclusion in the study. The sensitivity and specificity of riders or the trainer to correctly identify lameness in horses were not determined.
Classification of athletic performance—At the 2-year-old in-training sales, trial times of horses for highspeed running sessions (ie, breezes) were measured with an electronic timer operated by a professional racetrack timer operator. On the basis of these times, the athletic performance (ie, speed) of horses was designated as fast or slow in accordance with a modification of a classification scheme used in another study.1 Briefly, horses were designated as fast if their trial time was shorter than the mean trial time for all horses at the sale. Horses were designated as slow if their trial time was equal to or longer than the mean trial time for all horses at the sale.
Statistical analysis—The null hypotheses that median sales price is not different between horses with a low yearling sale purchase price versus those with a high yearling sale purchase price, horses with a low number of furlongs galloped within 60 days prior to sales versus those with a high number of furlongs galloped during that period, horses with lameness versus those without lameness, and horses classified as fast versus those classified as slow were tested via the Wilcoxon rank sum test and multivariable ANOVA. The continuous variables (yearling sale purchase price and total number of furlongs galloped and total number of furlongs breezed within 60 days prior to sales) for each horse were categorized as high or low on the basis of the median values; values that were equal to or higher than the median value were categorized as high, and values that were less than the median value were categorized as low. For univariate analyses, the relationship between the value for the outcome of interest (2-year-old in-training sales price) and yearling sale purchase price, athletic performance classification at the 2-year-old in-training sale, sex (colt vs filly), history of arthroscopic surgery (yes vs no), and exercise (galloping and breezing) distances, lameness status, and number of days lame within 60 days prior to the 2-year-old in-training sale were determined via the Wilcoxon rank sum test. For multivariable analyses, the relationship between 2-year-old in-training sales price (rank-transformed data) and those other variables were determined via ANOVA. Initially, variables determined to be associated (on the basis of a P value ≤ 0.20) with 2-year-old in-training sales price via univariate analyses were evaluated via an ANOVA model; a forward stepwise approach (a 2-sided P value of 0.05 was used to include variables in the model and a 2-sided P value of 0.10 was used to remove variables from the model) was used to identify variables associated with the main outcome of interest (2-year-old in-training sales price). The variables yearling sale purchase price and speed at the 2-year-old in-training sale were included in the final models regardless of the P values for these variables in the univariate analysis because these variables can affect the sales price of horses at the 2-year-old in-training sales. Because exercise can be affected by lameness,3 one model including the number of furlongs galloped within 60 days prior to 2-year-old in-training sales and another model including lameness status within 60 days prior to 2-year-old in-training sales was evaluated.
The null hypotheses that the number of furlongs galloped and the number of furlongs breezed in the 60 days prior to a sale are not different between horses with low yearling sale purchase prices versus those with high yearling sale purchase prices and between horses with lameness within 60 days prior to 2-year-old in-training sales versus those without lameness during that period were tested via the Wilcoxon rank sum test. To determine whether yearling sale purchase price and lameness had a combined effect on exercise distance within 60 days prior to 2-year-old in-training sales, exercise distance during that period was compared (via the Kruskal-Wallis test) among horses with a high purchase price with lameness (n = 13), horses with a high purchase price without lameness (13), horses with a low purchase price with lameness (7), and horses with a low purchase price without lameness (18). For all statistical analyses, values of P ≤ 0.05 were considered significant.
Results
Median 2-year-old in-training sale purchase price of the 51 horses in the study was $37,000 (range, $3,000 to $1,000,000; interquartile range, $20,000 to $150,000). Twenty (39%) horses developed lameness within 60 days prior to 2-year-old in-training sales (median number of days lame, 1 day [interquartile range, 1 to 10 days]).
Results of univariate analysis revealed the median 2-year-old in-training sale purchase price of the horses was significantly associated with yearling sale purchase price, number of furlongs galloped within 60 days prior to 2-year-old in-training sales, and speed at 2-year-old in-training sales (Table 1). Results of multivariable analysis controlled for yearling sale purchase price and speed at 2-year-old in-training sales revealed the median 2-year-old in-training sale purchase price was significantly higher for horses with a low (16 to 190) number of furlongs galloped within 60 days prior to the sale versus horses with a high (191 to 250) number of furlongs galloped during that time (Table 2). Variables included in the model accounted for 63% of the variation in 2-year-old in-training sale purchase price of horses (R2 = 0.63). Lameness was not significantly (P = 0.90) associated with 2-year-old in-training sale purchase price of horses.
Effects of various variables on 2-year-old in-training sale purchase price for 51 horses purchased at a Thoroughbred yearling sale in September 2006 and trained at a facility in Florida for the purpose of selling them for profit.
| Variable | No. of horses | Median 2-year-old sale price ($) | Mean ± SE 2-year-old sale price* | P value† |
|---|---|---|---|---|
| Sex | 0.92 | |||
| Colt | 25 | 40,000 | 26 ± 3 | |
| Filly | 26 | 35,000 | 25 ± 3 | |
| History of arthroscopic surgery | 0.36 | |||
| Yes | 4 | 155,000 | 33 ± 8 | |
| No | 47 | 32,000 | 25 ± 2 | |
| Purchase price at the yearling sale | < 0.01 | |||
| High ($45,000–$725,000) | 26 | 145,000 | 36 ± 2 | |
| Low (< $45,000) | 25 | 22,000 | 16 ± 2 | |
| Lameness‡ | 0.19 | |||
| Yes | 20 | 56,000 | 29 ± 4 | |
| No | 31 | 30,000 | 24 ± 3 | |
| No. of days lame‡ | 0.19 | |||
| High (1–26 days) | 20 | 56,000 | 29 ± 4 | |
| Low (0 days) | 31 | 30,000 | 24 ± 3 | |
| Total distance galloped‡ | < 0.01 | |||
| High (366–480 furlongs) | 26 | 25,000 | 19 ± 3 | |
| Low (50–365 furlongs) | 25 | 85,000 | 33 ± 3 | |
| Total distance breezed‡ | 0.11 | |||
| High (7–22 furlongs) | 26 | 30,000 | 23 ± 3 | |
| Low (0–6 furlongs)§ | 25 | 62,000 | 30 ± 4 | |
| Speed at 2-year-old in-training sales‖ | < 0.01 | |||
| Fast | 9 | 350,000 | 45 ± 2 | |
| Slow | 42 | 30,000 | 22 ± 2 | |
Horses underwent training from October 2006 until 2-year-old in-training sales (February through April 2007) that included galloping (approx 25 to 30 s/furlong; 1 furlong is 1/8 mile [approx 200 m]) and breezing (approx 15 s/furlong).
Rank-transformed data.
The P values were determined via the Wilcoxon rank-sum test.
Within 60 days prior to 2-year-old in-training sales.
Three horses breezed 0 furlongs during the 60 days prior to 2-year-old in-training sales.
Speed was classified by use of trial times recorded for horses at 2-year-old in-training sales during high-speed running (ie, breezing) sessions. Times were measured with an electronic timer operated by a professional racetrack timer operator. Horses were designated as fast if their trial time was shorter than the mean trial time for all horses at the sale. Horses were designated as slow if their trial time was equal to or longer than the mean trial time for all horses at the sale.
Results of multivariable ANOVA for determination of relationships between 2-year-old in-training sale purchase price and various variables for the horses in Table 1.
| Variable | No. of horses | Mean ± SE 2-year-old sales price* | P value | R2 |
|---|---|---|---|---|
| Model 1 | ||||
| Total distance galloped‡ | 0.01 | 0.63 | ||
| High (191–250 furlongs) | 26 | 26.9 ± 2.2 | ||
| Low (16–190 furlongs) | 25 | 34.0 ± 2.3 | ||
| Purchase price at the yearling sale | < 0.01 | |||
| High ($45,000–$725,000) | 26 | 36.8 ± 1.9 | ||
| Low (< $45,000) | 25 | 24.2 ± 2.7 | ||
| Speed at 2-year-old in-training sales‖ | < 0.01 | |||
| Fast | 9 | 37.6 ± 3.4 | ||
| Slow | 42 | 23.4 ± 1.4 | ||
| Model 2 | ||||
| Lameness‡ | 0.90 | 0.58 | ||
| Yes | 20 | 31.3 ± 2.6 | ||
| No | 31 | 29.7 ± 2.1 | ||
| Purchase price at the yearling sale | < 0.01 | |||
| High ($45,000–$725,000) | 26 | 38.1 ± 2.0 | ||
| Low (< $44,000) | 25 | 22.9 ± 2.9 | ||
| Speed at 2-year-old in-training sales‖ | < 0.01 | |||
| Fast | 9 | 37.6 ± 3.6 | ||
| Slow | 42 | 23.5 ± 1.5 | ||
Because exercise can be affected by lameness,3 one model (model 1) including number of furlongs galloped within 60 days prior to 2-year-old in-training sales and another model (model 2) including lameness status within 60 days prior to 2-year-old in-training sales were evaluated.
See Table 1 for remainder of key.
Horses with a high ($45,000 to $725,000) yearling sale purchase price and horses with lameness had a significantly lower number of furlongs galloped within 60 days prior to 2-year-old in-training sales versus horses with a low (< $45,000) yearling sale purchase price and versus horses without lameness, respectively (Tables 3 and 4). The median number of furlongs galloped by horses within 60 days prior to 2-year-old in-training sales was significantly lower for horses with a high yearling sale purchase price and with lameness (250 furlongs) versus horses with a low yearling sale purchase price and with lameness (412 furlongs; Table 5).
Median number of furlongs galloped and breezed within 60 days prior to 2-year-old in-training sales for the 51 horses in Table 1 with high (n = 26 horses) or low (25) yearling sale purchase prices.
| Variable | Yearling sale purchase price | P value | |
|---|---|---|---|
| High ($45,000–$725,000) | Low (< $45,000) | ||
| No. of furlongs galloped | 268 | 410 | 0.03 |
| No. of furlongs breezed | 6 | 7 | 0.65 |
Median number of furlongs galloped and breezed within 60 days prior to 2-year-old in-training sales for the 51 horses in Table 1 with (n = 20) or without (31) lameness.
| Variable | Lameness | P value | |
|---|---|---|---|
| Yes | No | ||
| No. of furlongs galloped | 256 | 409 | 0.04 |
| No. of furlongs breezed | 6 | 7 | 0.62 |
Combined effects of yearling sale purchase price and lameness status during training on median number of furlongs galloped and median number of furlongs breezed within 60 days prior to 2-year-old in-training sales for the 51 horses in Table 1.
| Median yearling sale purchase price | Lame | No. of horses | No. of furlongs galloped | No. of furlongs breezed |
|---|---|---|---|---|
| High | ||||
| $225,000 | Yes | 13 | 250a | 6a |
| $120,000 | No | 13 | 372a,b | 7a |
| Low | ||||
| $20,000 | Yes | 7 | 412b | 4a |
| $25,000 | No | 18 | 408a,b | 8a |
Within each column, values with different superscript letters are significantly (P < 0.05) different.
Discussion
Horses included in the present study were limited to 51 yearling Thoroughbreds at 1 training center in Florida. Therefore, results of the study cannot be extrapolated to reflect data for other training centers in Florida or training centers in other states. Furthermore, the median yearling sale purchase price of horses in the present study was $55,000, compared with $45,000 for all 3,355 Thoroughbred yearlings sold at the September 2006 yearling salea; this finding indicates the quality of horses in the present study may have been different from that for other Thoroughbreds sold at the sale.
For all horses in the present study, lameness identified by a rider or the trainer was confirmed by an attending veterinarian. However, the accuracy of lameness identification by riders or the trainer was not assessed. Therefore, horses may have been misclassified regarding lameness status. Horses with mild signs of lameness may not have been identified as lame (false-negative results). This may have introduced bias, in which case results regarding associations between lameness status and 2-year-old in-training sales price or number of furlongs galloped within 60 days prior to 2-year-old in-training sales would apply only to horses with signs of lameness identifiable by riders or trainers and confirmed by an attending veterinarian at the training center used in this study. Failure of riders and the trainer to detect severe signs of lameness was less likely than failure of riders and the trainer to detect mild signs of lameness in horses. Because yearling sale purchase price and speed of horses at 2-year-old in-training sales can influence 2-year-old in-training sales prices of Thoroughbreds, we considered exercise history and lameness as the most important factors influencing 2-year-old in-training sales prices.
Results of analyses controlled for yearling sale purchase price and speed of horses at 2-year-old in-training sales indicated high 2-year-old in-training sales prices of horses were associated with low numbers of furlongs galloped within 60 days prior to 2-year-old in-training sales. In addition, the number of furlongs galloped within 60 days prior to 2-year-old in-training sales was influenced by yearling sale purchase price and lameness status during training. Results of this study indicated yearling sale purchase price can affect a trainer's decision regarding amount of training for horses, particularly for horses with lameness.
Median sale purchase price of horses sold at 2-year-old in-training sales was higher for horses with a low number of furlongs galloped within 60 days prior to 2-year-old in-training sales than it was for horses with a high number of furlongs galloped during that period. This finding was supported by results of analyses for combined effects of yearling sale purchase price and lameness status on number of furlongs galloped within 60 days prior to 2-year-old in-training sales. For the 13 lame horses with a high yearling sale purchase price (median yearling sale purchase price for these horses, $225,000), the trainer may have reduced the amount of galloping exercise to prevent musculoskeletal injuries that would lower 2-year-old in-training sale purchase prices (to protect the large yearling sale investment). If the trainer's decision had been influenced in this manner, galloping exercise would not have been reduced for the 7 lame horses with a low yearling sale purchase price (median yearling sale purchase price for these horses, $20,000) because that investment was small. Results of another study5 indicated an increase in amount of galloping exercise increases the risk of injury to the dorsal aspects of third metacarpal bones in 2-year-old Thoroughbred racehorses. Results of other studies6–8 including Thoroughbred racehorses indicate horses that are exercised at high speed for long distances within 60 days prior to a sale have a high risk of catastrophic musculoskeletal injury.
Although we detected an association between yearling sale purchase price and distance exercised within 60 days prior to 2-year-old in-training sales, lameness alone was not associated with low 2-year-old in-training sale purchase prices for horses. One explanation for this finding was that horses affected with lameness may have been identified early and improved with treatment, given that the duration of lameness in most affected horses was ≤ 10 days. The anatomic locations and causes of lameness in horses in this study were not analyzed. However, the duration of lameness in affected horses in the present study was similar to that of yearling Thoroughbreds in another study1 conducted in Florida. Joint injury was the primary cause of lameness in horses in that study,1 and those horses were treated via intra-articular injection of sodium hyaluronate or methyl prednisolone acetate and amikacin. In addition, horses with mild acute injury of the dorsal aspects of third metacarpal bones can resume training after 5 to 10 days of rest and treatment with anti-inflammatory and analgesic drugs.9
In the present study, results of analyses controlled for yearling sale purchase price and speed recorded at 2-year-old in-training sales indicated that a high 2-year-old in-training sale purchase price of horses was associated with a low amount of exercise within 60 days prior to those sales. Lameness alone did not influence 2-year-old in-training sale purchase price. However, lameness was associated with reduced amount of galloping exercise within 60 days prior to 2-year-old in-training sales, particularly for horses with a high yearling sale purchase price. This finding suggested that the trainer's tolerance for risk of a musculoskeletal injury in lame horses in the present study was different between horses with low yearling sale purchase prices versus those with high yearling sale purchase prices.
Keeneland Thoroughbred Racing and Sales, Lexington, Ky.
References
1. Preston SA, Trumble TN, Zimmel DN, et al. Lameness, athletic performance, and financial returns in yearling Thoroughbreds bought for the purpose of resale for profit. J Am Vet Med Assoc 2008;232:85–90.
2. Preston SA, Zimmel DN, Chmielewski TL, et al. Prevalence of various presale radiographic findings and association of findings with sales price in Thoroughbred yearlings sold in Kentucky. J Am Vet Med Assoc 2010;236:440–445.
3. Hernandez JA, Hawkins DL. Training failure among yearling horses. Am J Vet Res 2001;62:1418–1422.
4. American Association of Equine Practitioners. Lameness scale. Definition and classification of lameness. In: Guide for veterinary service and judging of equestrian events. 4th ed. Lexington, Ky: American Association of Equine Practitioners, 1991;19.
5. Boston RC, Nunamaker DM. Gait and speed as exercise components or risk factors associated with onset of fatigue injury of the third metacarpal bone in 2-year-old Thoroughbred racehorses. Am J Vet Res 2000;61:602–608.
6. Estberg L, Gardner IA, Stover SM, et al. A case-crossover study of intensive racing and training schedules and risk of catastrophic musculoskeletal injury and lay-up in California Thoroughbred racehorses. Prev Vet Med 1998;33:159–170.
7. Estberg L, Gardner IA, Stover SM, et al. Cumulative racing-speed exercise distance cluster as a risk factor for fatal musculoskeletal injury in Thoroughbred racehorses in California. Prev Vet Med 1995;24:253–263.
8. Estberg L, Stover SM, Gardner IA, et al. High speed exercise history and catastrophic racing fracture in Thoroughbreds. Am J Vet Res 1996;57:1549–1555.
9. Bertone AL. The metacarpus and metatarsus. In: Stashak TS, ed. Adams' lameness in horses. Philadelphia: Lippincott Williams & Wilkins, 2002;800–805.
