Criteria for Selection of Cases

Medical records of all dogs examined at the University of Florida Veterinary Medical Center between January 1, 1996, and December 31, 2005, were reviewed to identify dogs in which primary appendicular OSA had been diagnosed. Dogs were eligible for inclusion in the study if the diagnosis had been confirmed by means of histologic or cytologic examination and information on breed, age, sex, neuter status, and primary tumor location at the time of diagnosis was available. Greyhounds included in the study were confirmed to be retired from racing or race training if an ear tattoo listed in the National Greyhound Association registry was identified or if the owner reported having adopted the dog from a group known to acquire Greyhounds from racetracks and farms that trained racing Greyhounds. For comparison purposes, the reference population consisted of all dogs evaluated at the Veterinary Medical Center for reasons other than OSA during the same period.

Procedures

Information obtained from the medical records of cases included in the study consisted of breed, age at diagnosis, sex, neuter status, and location of the primary lesion. Information on racing and training schedules for retired racing Greyhounds included in the study was not collected, as such information was either unknown or only partially known.

Statistical analysis—For each breed, period prevalence of OSA was calculated as the number of dogs of a particular breed with OSA examined during the study period divided by the total number of dogs of that breed examined during the study period. Unadjusted crude ORs and 95% CIs for risk of OSA in individual breeds were calculated by comparison with risk for mixedbreed dogs. For each of the 3 breeds with the highest odds of developing OSA, sex (male vs female) and age were examined as potential risk factors within breed. For this analysis, age was categorized as tertiles by ranking all dogs with and without OSA for each breed by age, starting at the earliest age at which OSA had been diagnosed, and dividing the population into 3 groups at the 33rd and 67th percentiles. Crude ORs and their 95% CIs were calculated for sex and age, with the category with the lowest number of cases of OSA used as the reference. In addition, a χ² test for linear trends was used to determine whether risk of OSA increased with age. Mean age at the time OSA was diagnosed was compared among the 3 breeds by means of 1-way ANOVA (Holm-Sedak test). For each breed, mean body weight of dogs with OSA was compared with mean body weight of age-matched dogs without OSA by use of a Student t test. Finally, for each breed, univariate χ² and Fisher exact tests were used to determine whether location of the primary lesion differed significantly between forelimbs and hind limbs or between right limbs and left limbs. All statistical analyses were performed with standard software.^c Values of P ≤ 0.05 were considered significant.

Results

Primary appendicular OSA was diagnosed in 179 dogs examined at the University of Florida Veterinary Medical Center between 1996 and 2005. Twenty-seven of the 179 dogs were of mixed breeding, with the remaining 152 dogs representing 30 breeds. Breeds represented by > 1 dog included the Rottweiler (n = 51), Greyhound (21), Labrador Retriever (20), Great Dane (13), Golden Retriever (11), Doberman Pinscher (8), Irish Setter (3), Akita (2), and Siberian Husky (2). All 21 Greyhounds with OSA were identified as retired racing Greyhounds.

Breed period prevalence of OSA, calculated as the number of dogs of a particular breed with OSA during the study period divided by the total number of dogs of that breed examined during the study period, was highest for Greyhounds (21/339 [6.2%]), followed by Rottweilers (51/969 [5.3%]) and Great Danes (13/297 [4.4%]). For all 3 breeds, there were no significant trends in annual prevalence of OSA from 1996 through 2005. For Greyhounds, annual prevalence ranged from 0% in 1999 (0/28) to 10.5% in 2001 (4/38). For Rottweilers, annual prevalence ranged from 0% in 1997 (0/82) to 10.8% in 2003 (11/102). For Great Danes, annual prevalence ranged from 0% in 1996 (0/21) to 11.1% in 1997 (2/18). A total of 7,116 mixed-breed dogs were examined during the study period. Compared with mixed-breed dogs, Greyhounds (OR, 17.3; 95% CI, 9.3 to 32.1), Rottweilers (OR, 14.6; 95% CI, 8.9 to 24.0), and Great Danes (OR, 12.0; 95% CI, 5.8 to 24.5) had a significantly (P < 0.001) increased risk of OSA.

For the 3 breeds with the highest prevalences, body weight did not appear to be a risk factor for OSA. Mean body weight of Greyhounds with OSA (mean ± SD, 31.6 ± 3.1 kg [69.5 ± 6.8 lb]) was not significantly (P = 0.5) different from mean body weight of age-matched Greyhounds without OSA (30.9 ± 5.3 kg [68.0 ± 11.7 lb]), mean body weight of Rottweilers with OSA (46.7 ± 9.9 kg [102.7 ± 21.8 lb]) was not significantly (P = 0.2) different from mean body weight of age-matched Rottweilers without OSA (44.3 ± 9.3 kg [97.5 ± 20.5 lb]), and mean body weight of Great Danes with OSA (69.3 ± 11.2 kg [152.5 ± 24.6 lb]) was not significantly (P = 0.06) different from mean body weight of age-matched Great Danes without OSA (61.6 ± 9.8 kg [135.5 ± 21.6 lb]).

Similarly, sex did not appear to be a risk factor for OSA. Dogs were not stratified on the basis of neuter status for analysis of sex as a risk factor for OSA because all of the Greyhounds and Great Danes with OSA were neutered at the time of diagnosis, as well as most of the Rottweilers (21/26 male Rottweilers and 23/25 female Rottweilers with OSA were neutered). Age at the time of neutering was unknown for all dogs. For all 3 breeds, males were not significantly more likely to have OSA than females (Tables 1–3).

Mean ± SD age at the time of diagnosis was 9.9 ± 2.6 years for the 21 Greyhounds with OSA (range, 6 to 15 years), 8.3 ± 2.1 years for the 51 Rottweilers with OSA (range, 5 to 14 years), and 7.8 ± 2.2 years for the 13 Great Danes with OSA (range, 5 to 12 years). Overall, 12 of the 21 (57%) Greyhounds, 36 of the 51 (71%) Rottweilers, and 7 of the 13 (54%) Great Danes were between 7 and 10 years old at the time OSA was diagnosed. Greyhounds were significantly older at the time OSA was diagnosed than were Rottweilers (P = 0.006) or Great Danes (P = 0.009). Mean age at the time of diagnosis was not significantly (P = 0.8) different between Rottweilers and Great Danes with OSA. In all 3 breeds, the risk of OSA increased significantly with age (Tables 1–3).

In the Greyhounds, tumors were located in the proximal end of the humerus (n = 5), distal end of the radius (7), proximal end of the femur (4), distal end of the femur (2), proximal end of the tibia (1), and distal end of the tibia (2). In the Rottweilers, tumors were located in the proximal end of the humerus (n = 19), proximal end of the radius (1), distal end of the radius (13), proximal end of the femur (1), distal end of the femur (8), proximal end of the tibia (2), and distal end of the tibia (7). In the Great Danes, tumors were located in the proximal end of the humerus (n = 1), distal end of the radius (8), distal end of the femur (3), and distal end of the tibia (1).

Table 1—

Univariate analyses of sex and age as potential risk factors for appendicular OSA in Greyhounds.

View Table

Table 2—

Univariate analyses of sex and age as potential risk factors for appendicular OSA in Rottweilers.

View Table

Table 3—

Univariate analyses of sex and age as potential risk factors for appendicular OSA in Great Danes.

View Table

For the Greyhounds, the proportion of forelimb tumors (12/21) was not significantly (P = 0.4) different from the proportion of hind limb tumors (9/21), and the proportion of right limb tumors (10/21) was not significantly (P = 0.8) different from the proportion of left limb tumors (11/21). For the Great Danes, the proportion of forelimb tumors (9/13) was significantly (P = 0.048) higher than the proportion of hind limb tumors (4/13), but the proportion of right limb tumors (5/13) was not significantly (P = 0.2) different from the proportion of left limb tumors (8/13). In the Rottweilers, the proportion of forelimb tumors (33/51) was significantly (P = 0.003) higher than the proportion of hind limb tumors (18/51), and the proportion of right limb tumors (31/51) was significantly (P = 0.03) higher than the proportion of left limb tumors (20/51).

Discussion

Results of the present study suggested that Greyhounds, Rottweilers, and Great Danes had an increased risk of developing OSA, compared with mixed-breed dogs. The finding that Rottweilers and Great Danes had an increased risk of developing OSA was consistent with findings of previous studies.^4,5,8-10 To our knowledge, however, this is the first time Greyhounds have been identified as having an increased risk of developing OSA.

All of the Greyhounds in the present study were identified as having been retired from race training or racing. In a previous study⁵ of cases reported by various veterinary teaching hospitals to the Veterinary Medical Data Bank from 1980 through 1994, the Greyhound was not identified as a breed at risk for developing OSA. In our opinion, this likely reflected the low number of retired racing Greyhounds in the pet population during the study period. Adoption of retired racing Greyhounds did not gain popularity until the 1990s, and a review of medical records at the University of Florida Veterinary Medical Center revealed that OSA was diagnosed in only 7 Greyhounds between 1985 and 1995, but was diagnosed in 21 Greyhounds between 1996 and 2005.

Similar to findings in previous studies,^5,8-10 the risk for OSA among dogs in the present study increased with age, with most (55/85 [65%]) of the Greyhounds, Rottweilers, and Great Danes in the present study between 7 and 10 years old at the time OSA was diagnosed. Greyhounds were significantly older at the time of diagnosis than were Rottweilers or Great Danes, but the reason for this could not be determined. As was the case in previous studies,^2,3,4,8,9,10 sex (male vs female) was not a risk factor for development of OSA among Greyhounds, Rottweilers, or Great Danes in the present study.

Previous studies^2,3,5,11,12 have found that OSA developed more frequently in the forelimbs of dogs than the hind limbs, which may be related to the fact that the forelimbs support about 60% of the body weight in dogs. In the present study, Rottweilers and Great Danes were more likely to have OSA involving the forelimbs, but there was no difference between forelimbs and hind limbs in the Greyhounds. Consistent with results of previous studies,^2,3,10-12 the most frequent sites of OSA in the Greyhounds, Rottweilers, and Great Danes in the present study were the proximal end of the humerus and distal end of the radius. The proximal end of the femur was also a common site for OSA among Greyhounds in the present study. Previous studies^2,3,9-12 have found that the proximal end of the femur is an uncommon site for OSA in dogs, but these studies did not include Greyhounds.

Fatigue-induced microcracks occur in cortical bone in the limbs of dogs in association with repetitive weight-bearing stresses of daily activities.¹³ These microcracks are repaired by bone remodeling processes characterized by increased cell turnover.^14,15 Thus, it has been speculated that racing Greyhounds may be predisposed to developing OSA because of stresses placed on the limbs during racing, particularly the right limbs, because these sustain the most trauma and fatigue in association with running counterclockwise on oval tracks in the United States. However, there was no difference in the proportion of right limb versus left limb tumors among Greyhounds in the present study. Although retired racing Greyhounds in the present study had a significantly increased risk of developing OSA compared with mixed-breed dogs, we were unable to determine whether racing was a risk factor for development of OSA because there were no Greyhounds that had not trained for racing or raced competitively. Unfortunately, analysis of possible associations between time spent in training or racing and risk for OSA was not possible because details of the training and racing histories of these dogs were unavailable. A previous study¹³ found no association between microcracks in the metaphyses of the humerus and radius and an increase in body size or age, 2 known risk factors for OSA, but the study did not evaluate racing Greyhounds or dogs involved in other athletic pursuits. Further studies are needed to determine whether repetitive exercise associated with athletic competition increases microdamage in bones and contributes to the risk of OSA.

Limitations of the present study include the retrospective design and the potential for institutional bias associated with analysis of a population restricted to a single veterinary medical teaching hospital. Florida has the largest number of Greyhound race tracks and the largest number of chapters of the Greyhound Pets of America, so the number of retired racing Greyhounds evaluated at the University of Florida may not be representative for institutions in other states. The relatively small number of OSA cases limited the statistical analyses to crude analyses without adjustments for potential confounding factors. Despite these limitations, our results suggested that retired racing Greyhounds should be added to the list of dogs at high risk for development of appendicular OSA.