Materials and Methods

Case selection—Requests to submit information regarding dogs with CCLD that had an excessive TPA and for which pretreatment radiographs were available for review were sent to a convenience sample of 134 veterinary surgeons in North America known to at least 1 of the authors. Submitted cases were selected for inclusion in the present study if the dog weighed ≥ 18.1 kg (40 lb) at the time CCLD was diagnosed and TPA in 1 or both hind limbs was ≥ 35°. For determination of TPA, angles were measured by 3 individuals (RHP, ELE, and FMD) on pretreatment radiographs as described⁵ with the aid of a radiographic analysis program,^b and mean angles were calculated. The mean TPA was used for all statistical analyses.

Control selection—Control dogs were identified by searching the medical records of the Colorado State University Veterinary Medical Center to identify dogs with CCLD that weighed ≥ 18.1 kg at the time CCLD was diagnosed and in which TPA was ≤ 30°. For control dogs, the TPA recorded in the medical record was used; all angles had been measured by or under the direct supervision of a board-certified veterinary surgeon. Dogs were excluded from the control group if TPA in either limb was > 30°. Control dogs were randomly selected from the candidate list until the number of control dogs equaled the number of case dogs.

Data collection—For case dogs, preoperative radiographic views of the stifle joints submitted by the contributing veterinary surgeons were examined by 4 individuals (RDP, RHP, ELE, and FMD) for any radiographic abnormalities, such as cortical, intramedullary, and physeal changes (ie, premature closure) and other deformities and fractures, that could potentially have caused development of excessive TPA. Abnormalities were included only if at least 2 evaluators were in agreement with regard to the abnormality.

A standardized telephone questionnaire was administered by a single individual (RSS) to owners of case and control dogs. The questionnaire focused on host and environmental factors that could potentially have been associated with development of excessive TPA and on factors related to CCLD in general.

Specific host factors that were studied included breed, personality, and overall activity level of the dog, particularly during the first year of life. Information was also collected on age at the time of neutering, history of any physical trauma, and other medical conditions. Information regarding signalment (date of birth, sex, breed, and body weight), age at the time of neutering, whether lameness was unilateral or bilateral, and age at the onset of hind limb lameness was confirmed by review of medical records when possible.

Environmental factors that were studied included the dog's feeding regimen prior to 6 months of age; body weight prior to the onset of hind limb lameness; number and type of other pets in the household; and time spent each day on controlled leash walks, off-leash activity, and unobserved activity. Owners were also asked whether the dog regularly engaged in any strenuous activities, such as hiking, hunting, working, or agility, before 1 year of age. Finally, owners were asked to grade the dog's overall activity level during the first year of life and the amount of agitation in response to strangers or environmental changes on a scale from 1 (very little) to 6 (very much).

Statistical analysis—Age, body weight, and time spent each day on controlled leash walks, off-leash activity, and unobserved activity were recorded as continuous variables; all other factors were recorded as categoric variables. The t test was used to compare continuous variables between case and control groups, and the χ² test was used to test for associations between categoric variables and group (case vs control). For dichotomous variables, the measure of effect was reported as an odds ratio and 95% confidence interval. When possible and necessary, categoric variables were collapsed for statistical analysis. For all analyses, a value of P ≤ 0.05 was considered significant.

Age at the time of neutering was categorized as ≥ 6 months or < 6 months. Only data confirmed by review of medical records were used for analysis. Whether lameness was unilateral or bilateral prior to treatment was determined on the basis of owner responses to the questionnaire; if the owner was not available, lameness was assumed to be bilateral if surgery was performed on both hind limbs. Owner-reported factors that could potentially have contributed to CCLD were categorized as high activity level, trauma, or other.

Results

Fifty-eight case dogs (ie, dogs with unilateral or bilateral CCLD and TPA ≥ 35° in 1 or both hind limbs) and 58 control dogs (ie, dogs with unilateral or bilateral CCLD and TPA ≤ 30°) were included in the study. Records for case dogs were submitted by 27 veterinary surgeons.

For the case dogs, 9 owners could not be located, 4 owners declined to participate in the study, and 2 owners were not contacted at the request of the attending veterinarian. Thus, questionnaires were completed by owners of 43 of the case dogs. Owners of control dogs were contacted until 43 had also completed the questionnaire.

Thirty-eight case dogs had bilateral stifle radiographs, and 20 had unilateral radiographs. Thirty-nine control dogs had unilateral, and 19 had bilateral stifle radiographs. Mean ± SD TPA was 41.7° ± 4.9° (range, 35° to 59°) for case dogs and 24.7° ± 3.6° (range, 11° to 30°) for control dogs. Of the 38 case dogs with bilateral stifle radiographs, 27 (71%) had TPA ≥ 35° in both hind limbs, 7 (18%) had contralateral TPA between 30° and 35°, and 4 (11%) had contralateral TPA ≤ 30°. All 19 (100%) control dogs with bilateral stifle radiographs had contralateral TPA ≤ 30° as required for control group inclusion.

Nineteen of the 58 dogs in the case group were of mixed breeding. The remaining dogs consisted of 12 Labrador Retrievers, 6 Rottweilers, 5 Golden Retrievers, 2 Australian Shepherds, 2 Great Danes, 2 Great Pyrenees, 2 Newfoundlands, 2 pitbull-type dogs, and 6 dogs representing 6 other breeds. Similarly, 17 of the 58 control dogs were of mixed breeding, with the remaining dogs consisting of 17 Labrador Retrievers, 3 Golden Retrievers, 3 Siberian Huskies, 2 Rottweilers, 2 Newfoundlands, 2 pitbull-type dogs, and 12 dogs representing 12 other breeds. Breed distribution for case dogs was not significantly (P = 0.44) different from breed distribution for control dogs.

Mean ± SD body weight was 37.7 ± 10.8 kg (83.1 ± 23.8 lb) for dogs in the case group and 36.3 ± 9.3 kg (80 ± 20.5 lb) for dogs in the control group; body weight was not significantly (P = 0.47) different between groups. Case dogs consisted of 31 (53%) spayed females, 20 (34%) castrated males, and 1 (2%) sexually intact female; sex and neuter status were not available for the remaining 6 (10%) dogs. Control dogs consisted of 29 (50%) spayed females, 27 (46%) castrated males, 1 (2%) sexually intact female, and 1 (2%) sexually intact male. Case and control groups did not differ significantly with regard to sex distribution (P = 0.06), proportion of neutered dogs (P = 0.62), or proportion of male dogs (P = 0.30).

None of the stifle joints in case dogs had radiographic evidence of overt abnormalities, such as a previous tibial fracture, that might have caused the excessive TPA. Abnormalities that were observed included widening of the physis of the proximal tibial apophysis (n = 3), soft tissue swelling cranial to the patellar tendon (2), subjectively long fibula (2), irregular shape of lateral tibial condyle (2), calcification of the patellar tendon (1), previous femoral fracture (1), and periosteal bone reaction involving the tibial crest (1).

Information on age at the time of neutering was available for 38 case dogs and 42 control dogs. Twenty of the 38 (53%) case dogs and 11 of the 42 (26%) control dogs were < 6 months old at the time of neutering. Case dogs were 3 times (95% confidence interval, 1.2 to 8.0) as likely to have been neutered before 6 months of age as were control dogs. Seventeen of 22 (77%) case dogs with bilateral CCLD in which TPA was ≥ 35° in both limbs and 3 of 15 (20%) control dogs with bilateral CCLD in which TPA was ≤ 30° in both limbs were neutered before 6 months of age. Case dogs with TPA ≥ 35° in both limbs were 13.6 times (95% confidence interval, 2.72 to 68.1) as likely to have been neutered before 6 months of age as were control dogs with TPA ≤ 30° in both limbs. Control dogs were 6.8 times (95% confidence interval, 2.03 to 22.42) as likely to have been regularly engaged in working or hunting before 1 year of age as were case dogs. There was a significant (P < 0.001) difference between control and case groups in regard to owner perception of factors that may have contributed to CCLD; owners of 10 case dogs and 26 control dogs reported that high activity level could potentially have contributed to CCLD in their dogs, and owners of 15 case dogs and 3 control dogs reported that trauma may have contributed to CCLD in their dogs. There were no significant differences between groups with regard to any of the remaining variables examined (feeding regimen prior to 6 months of age, P = 0.06; feeding frequency, P = 0.78; feeding of supplements, P = 0.051; veterinarian assessment of patient as underweight, P = 0.16; veterinarian assessment of patient as overweight, P = 0.43; other medical conditions, P = 0.11; history of any physical trauma, P > 0.99; number of other pets in the household, P = 0.7; time spent each day on controlled leash walks, P = 0.45; time spent each day in off-leash activity, P = 0.68; time spent each day in unobserved activity, P = 0.49; amount of agitation in response to strangers or environmental changes, P = 0.31; and overall activity level during the first year of life, P = 0.97).

Age at the onset of hind limb lameness was significantly (P < 0.001) lower for case dogs than for control dogs; mean ± SD age at the onset of hind limb lameness was 22.9 ± 17.2 months for case dogs and 58.9 ± 35.6 months for control dogs. Forty of the 58 (69%) case dogs had bilateral hind limb lameness, and 10 (17%) had unilateral hind limb lameness (information was not available for the remaining 8 [14%] dogs). Thirty-four of the 58 (59%) control dogs had bilateral hind limb lameness, and 16 (27%) had unilateral hind limb lameness (information was not available for the remaining 8 [14%] dogs). Case dogs were 1.9 times (95% confidence interval, 0.76 to 4.69) as likely to have bilateral hind limb lameness as were control dogs, but this difference was not significant.

Discussion

Results of the present study indicated that neutering before 6 months of age was a significant risk factor for excessive TPA among large-breed dogs with CCLD. Dogs with TPA ≥ 35° in 1 or both hind limbs were 3 times as likely to have been neutered before 6 months of age as were control dogs. Dogs with TPA ≥ 35° in both hind limbs were 13.6 times as likely to have been neutered before 6 months of age as were dogs with TPA ≤ 30° in both hind limbs.

Traditionally, most veterinarians in the United States have neutered dogs at 6 to 8 months of age.^11-14 The effects of gonadectomy in animals prior to this age (so-called early-age gonadectomy) have been intensively investigated,^12,13,15-18 and it has been shown that prepubertal gonadectomy alters the histologic appearance of the growth plate, resulting in widening of the growth plate and increased longitudinal bone growth.^15,17 Gonadectomy as late as 7 months of age results in delayed growth plate closure in male and female dogs,¹⁵ and this delayed closure may render growth plates more susceptible to trauma and subsequent physeal fractures.^15,19-21 In cats, spontaneous femoral capital physeal fractures have been suggested to be a result of delayed physeal closure related to early-age neutering.²¹ A previous study¹³ found an increased incidence of hip dysplasia among dogs neutered at an early age, and the authors speculated that this increased incidence may have resulted from altered hip joint conformation secondary to increased bone length.

In dogs, excessive TPAs are most likely a result of alterations in the growth of the proximal tibial physis.⁴ Proposed pathophysiologic mechanisms include trauma, imbalance between the quadriceps femoris and gastrocnemius muscles, reduced blood supply, breed predisposition, overgrowth of the cranial part of the proximal tibial physis secondary to an increased blood supply, and developmental tibial deformity.^4,9,10 Overt trauma seemed an unlikely causal factor in that, in previous studies,^4,9,10 many dogs had no known history of trauma and most were affected bilaterally. Isolated dogs with an excessive TPA resulting from fracture of the proximal tibial physis have been reported²²; however, these dogs typically are unilaterally affected and frequently have a history of trauma. In the present study, only 4 dogs had a TPA ≥ 35° in 1 limb and TPA ≤ 30° in the contralateral limb. The only dog in the case group that was sexually intact was also the only dog in the study population with a known history of trauma. Together, these findings suggest that although selected dogs may develop an excessive TPA in a single limb as a result of physical trauma, early-age gonadectomy likely plays a larger role, particularly in dogs with excessive TPAs in both hind limbs.

Excessive TPA may be a result of subclinical, premature closure of the caudal aspect of the proximal tibial physis, increased growth of the cranial aspect of the physis, or a combination of these 2 mechanisms. Both of these proposed mechanisms could be related to lack of gonadal hormones secondary to early-age gonadectomy. Early-age gonadectomy results in delayed physeal closure and subsequent increased longitudinal bone growth^15,17 and, thus, could account for increased growth of the cranial aspect of the proximal tibial physis. Alternatively, delayed closure and the associated histologic alteration of the physis may make it more susceptible to injury that could lead to premature closure of the caudal aspect of the physis.^15,17,19-22

In the present study, age at the onset of hind limb lameness was significantly lower for case dogs than for control dogs, suggesting that excessive TPA may have contributed to earlier degeneration of the cranial cruciate ligament. No significant differences were detected between the case and control groups with regard to feeding regimen, body weight, history of other medical problems, number of other pets in the household, activity level, or agitation in response to strangers or environmental changes. Dogs in the control group were more likely to have been used for working or hunting before 1 year of age, and a higher proportion of their owners perceived that high activity level may have contributed to CCLD. Interpreting these findings is difficult, as dogs with excessive TPA were more likely to be affected early in life and, therefore, were less likely to be exercising at a high level. Results may also have been influenced by selection bias because most control dogs were from Colorado, where outdoor activities are common. Owners of case dogs perceived trauma as a contributing factor to lameness more frequently than did owners of control dogs. However, when owners were directly asked whether their dog had a history of trauma, no significant difference was identified between groups. Likely, owner education regarding the degenerative nature of CCLD may have affected the responses to this question. Alternatively, dogs with excessive TPA may be more predisposed to traumatic rupture secondary to increased forces acting on the ligament as a result of the high TPA.

Results of the present study indicate that further research is needed on the effects of early-age gonadectomy in dogs, particularly with respect to risk of CCLD. Veterinarians and owners should be aware that earlyage gonadectomy appears to increase the risk of excessive TPA. However, this risk needs to be communicated in the context that excessive TPA remains an infrequent finding and that the benefits of early-age gonadectomy in terms of population control are substantial.