Anterior cruciate ligament (ACL) injury occurs two to eight times more frequently in human females than in males.2 The prevalence of ACL injury in the general population is approximately 0.38 per 1000 people.20 Intrinsic9-11 and extrinsic9-11 factors likely affect susceptibility to ACL injury. Gender differences in human ACL injury have been associated with phases of the menstrual cycle3,17,19,23 and with specific neuromuscular traits such as medial knee collapse during landing from a jump.8,18
Numerous studies have shown that the distribution of human ACL injury is not uniform across the phases of the menstrual cycle.4,19,24,27,28 Several of these studies identified the late luteal, menstrual, or early follicular phases as times when women are at greatest risk for ACL injury.4,19,24 Two studies identified at-risk times to be the late luteal and ovulatory phases, as determined by urinary levels of sex hormones.27,28 Differences in study design, mechanism of ACL injury, and study populations are the likely source of variability in defining specific at-risk times for ACL injury. Despite the variable results of these diverse studies, however, none found a uniform distribution of ACL injury across the menstrual cycle, suggesting the presence, absence, or combination of sex hormones has an uncharacterized effect on ACL injury.
Studies on how gender affects human ACL injury are complicated by extrinsic differences such as sport, shoe wear, surface type, coaching, and gender social biases. Examination of a nonhuman animal model with a high prevalence of ACL injury would eliminate many extrinsic factors that complicate human studies. Such a model would enable a more focused analysis of specific intrinsic differences such as gender, the presence or absence of sex hormone, or effect of the estrus cycle on ACL injury.
Anterior cruciate ligament rupture is a relatively common cause of hind limb lameness in animals treated in veterinary practices.6,12,26 Two studies have yielded conflicting results on the effects of gender, ovariohysterectomy, and orchiectomy on ACL injury in dogs.6,26 The study populations were dogs from practices in university veterinary schools. Such populations may represent a selection bias as compared with all dogs because university practices are tertiary referral centers that treat dogs with more complex problems. In contrast, the population of dogs seen in a private veterinary practice might more accurately reflect the general dog population. Furthermore, examining records from one surgeon’s practice would eliminate variability derived from differences in care provided by different veterinarians.
We tested the hypothesis that ovariohysterectomy or orchiectomy increases the prevalence of canine ACL injury. We did logistic regression analysis on the data to determine if the prevalence of ACL injury holds across dog sizes and breeds. The results revealed an increased prevalence of ACL injury in animals that had gonadectomy. Our findings suggest that the dog may be a useful model for studying the effects of hormones on the material and molecular properties of the ACL, and in turn, the relationship of these properties to the risk of ACL injury.
MATERIALS AND METHODS
The records of all dogs seen and evaluated during a 2-year period in a one-surgeon orthopaedic veterinary practice were reviewed retrospectively. Each dog was identified and was counted as one unit of observation regardless of the number of visits. No dogs were excluded from the general population in the clinic. Inclusion criteria for the ACL-injured group were a history of hind-limb lameness, a positive anterior drawer examination by the orthopaedic veterinarian, and an observed torn ACL at the time of surgery. Breed, gender, ovariohysterectomy status, and orchiectomy status were recorded for all dogs.
Prevalence was calculated for sexually intact males, sexually intact females, ovariohysterectomized females, and orchiectomized males by identifying the total number of dogs with ACL ruptures in the groups. To test whether gender was associated with the probability of ACL rupture, we used three logistic regression models, because the dependent variable is dichotomized either as having or not having ACL rupture. The first model was a univariate logistic regression, with a single correlate gender, that had four categories: sexually intact female; sexually intact male; ovariohysterectomized female; and orchiectomized male. The first category served as the comparison group in the regression. Whether sexually intact males, ovariohysterectomized females, and orchiectomized males, respectively, had higher probabilities than sexually intact females of having ACL ruptures was tested by comparing the estimated odds ratios (OR), which were calculated by dividing the probability in the test group by the probability in sexually intact females. The significance of the differences of the OR was determined by using a threshold p value of 0.05. The second logistic regression model investigated the gender difference, controlling for breed. Finally, we estimated a third logistic model controlling for dog sizes. Because the actual dog height was not recorded in the charts and a dog’s weight varies with time, we used a generalized method to group dogs, as defined by Verhoef-Verhallen.25 A dog’s size was categorized as large, medium, or small, based on mean breed height at the withers (the ridge between the shoulder bones). Dogs with a height less than 40 cm were deemed small breeds and dogs with a height greater than 60 cm were deemed large.25 Dogs whose sizes could not be determined for various reasons (for example, mixed breeds) were assigned to a fourth category, missing size, for the size variable.
Among the 3218 dogs included in this study, the overall prevalence of ACL rupture was 3.48%. Table 1 shows the prevalence of ACL rupture for sexually intact females, sexually intact males, ovariohysterectomized females, and orchiectomized males in our sample. Compared with the other three gonadal gender categories, ovariohysterectomized females had the greatest prevalence of ACL rupture (5.15%), and sexually intact males had the least prevalence (2.09%). In addition, prevalence of ACL rupture in ovariohysterectomized females and orchiectomized males was 2.1-fold greater than in sexually intact animals.
Gonadectomy increased the prevalence of ACL injury. The probability of injury rate in ovariohysterectomized females was greater (OR = 2.19; p = 0.005) than in sexually intact females (Table 2). Also, in the postestimation test, ovariohysterectomized females were more likely (OR = 2.51; p = 0.0007) than sexually intact males to have ACL injuries.
This prevalence of ACL injury occurred across the sizes and breeds of dogs. When breed was considered a confounding factor, sexually intact females, orchiectomized males (OR = 2.04; p = 0.04), and ovariohysterectomized females (OR = 2.57; p = 0.002) were more likely to have ACL ruptures (Table 2). In addition, the postestimation test showed that ovariohysterectomized females had greater (OR = 2.64; p = 0.0002) probabilities of sustaining ACL injures than sexually intact males; differences (p = 0.0143) between sexually intact males and orchiectomized males also were seen. No difference was found, however, between sexually intact females and sexually intact males, between sexually intact males and orchiectomized males, or between orchiectomized males and ovariohysterectomized females. When dogs were controlled for size, 229 (7.12%) were large, 1053 (32.72%) were medium, 1509 (46.89%) were small, and 427 (13.27%) had an undetermined size. Using size as a control model, ovariohysterectomized females were more likely (OR = 2.27; p = 0.003) than sexually intact females to have ACL ruptures. The postestimation tests indicated that probability of ACL injuries was less (OR = 0.51; p = 0.039) among sexually intact males than among orchiectomized males, and the probability of ACL injuries was less (OR = 0.38; p = 0.0005) in sexually intact males than among ovariohysterectomized females. Furthermore, dog size contributed to the probability of ACL rupture. In particular, compared with large dogs, medium (OR = 0.38; p = 0.001) and small dogs (OR = 0.38; p = 0.001) were less likely to sustain ACL injuries. However, no difference (p = 0.99) was found between medium and small dogs. The probability of injury between medium- and small-sized dogs did not differ. Furthermore, postestimation tests were done to identify differences among the three noncomparison groups (sexually intact males, orchiectomized males, and ovariohysterectomized females).
Human ACL injury is an important clinical problem. Some effects of sex hormones, gender, and other factors on injury risk could be determined best by studying appropriate, nonhuman animal models. The dog is a potentially good model animal, but previous studies on ACL injuries in dogs have produced conflicting results.6,26 To clarify the results and limit the biases inherent in these previous studies, we examined the effect of gonadectomy, breed, and size of breed on the prevalence of ACL injuries in dogs treated at one veterinarian’s community-based practice.
Specific risk factors for ACL injury can be studied in the dog because many extrinsic factors that may affect ACL injury either do not apply or can be minimized. For example, shoe size, foot surface, coaching, and social biases are irrelevant in the dog. Other factors, such as animals being bred to run, may be minimized by large sample sizes and breed comparisons. Consequently, the effect of intrinsic differences, such as gender, hormone status, or gonadectomy on prevalence of ACL injury can be assessed more accurately.
In our study, gonadectomy increased the prevalence of ACL injuries across sizes and breeds of dogs of either gender. Gonadectomy in dogs produces a hormonal change, enabling direct comparison of four groups of animals (two with high levels of sex hormones and two with low levels) that presumably run, jump, and turn in similar, uncontrolled settings. The finding that gonadectomy in dogs is associated with increased prevalence of ACL injury supports the idea that sex hormones affect the ACL’s material properties, size, or both.
Our key findings are similar to results from two previous studies.6,26 Whitehair et al26 retrospectively reviewed more than 10,000 dogs with ACL injuries from 23 veterinary school clinics. They reported that neutered and spayed dogs had a higher prevalence of ACL injuries than sexually intact dogs, a result confirmed by our study. They also reported an increased prevalence of injuries in sexually intact female dogs compared with male dogs, a result not seen in our sample population. Duval et al6 used a case-controlled design examining 200 dogs younger than 2 years with ACL injuries. Similar to our findings, their results identified an increased risk of ACL injuries in dogs that were ovariohysterectomized and orchiectomized, but no differences in risk among dogs that were sexually intact. The contradictory findings may derive from differences in study design (case-control versus retrospective review), study population selection bias (academic versus private setting), or targeted age groups (younger than 2 years versus all-ages group).
One potential source of bias in this study is that only dogs whose owners sought care for lameness were included. This potential source of bias is inherent in all retrospective studies of clinic records. Of particular concern is the possibility that some dogs in our region (eg, strays or pets of negligent owners) never receive veterinary care of any kind, including for spaying, neutering, or treatment for lameness. If such animals were a large segment of the dog population in our region, it would tend to exaggerate the apparent prevalence of ACL injury in spayed and neutered animals in our study because a population of nonspayed, nonneutered, lame dogs was not represented. However, one also must consider that our region includes a substantial population of highly prized, nonspayed, nonneutered dogs specifically bred for show and for working, especially hunting. Owners of such animals typically seek veterinary care routinely, and especially for conditions that are highly symptomatic, such as ACL injury. Consequently, this population might be over-represented in our study, and thereby would tend to exaggerate the apparent prevalence of ACL injury in nonspayed, nonneutered animals, in opposition to the effects of the stray and neglected population. Regardless, both of these populations are likely to be small in comparison to the overall dog population in our region, and therefore unlikely to affect significantly the conclusions of this study.
Comparisons of mechanism of injury between the dog and human are difficult. The anatomy of the dog knee (stifle joint) is different from that of the human knee. The active range of motion of the dog stifle is 46°-142°,14 whereas that of the human knee is 0°-160°. Dogs have a four-legged gait, so the dynamics of running and cutting are different from humans. These differences suggest the mechanism of injury in dogs is not the same as in humans. Nevertheless, the mechanism of injury is likely to be similar, and most, or all, of the external biases affecting ACL injury introduced by human behavior are minimized in the dog.
The effect of ovariohysterectomy and orchiectomy on the dog is complex and results in various body changes in the animal far greater than an effect seen solely from the absence of one sex hormone. For example, bone density decreases secondary to the reduction of the influences of sex hormones on calcium deposition into the skeleton. Despite the effect on bone, none of the injured animals had a ligament-bone avulsion at the time of surgery. Whether ovariohysterectomy or orchiectomy affect weight gain is controversial, but one study showed no significant increase.21,22 Additionally, ovariohysterectomy or orchiectomy could affect the behavior of the dog.21,22 However, an ovariohysterectomized or orchiectomized dog most likely would become less aggressive, less active, or both, probably decreasing the risk for ACL injury.21,22 Although we cannot predict all of the effects that could result from canine ovariohysterectomies or orchiectomies, many of the effects are likely to decrease injury risk.
The effect of decreased sex hormones on the ACL of the dog is not known. However, alterations in sex hormones may affect the size, shape, or material properties of the ACL. Anterior cruciate ligament remodeling is affected by loads placed on the ligament (Wolff’s Law). The remodeling of many soft tissues, including the ACL,1,5,7,13,15,16 also is influenced by sex hormones. In humans, elevated levels of estrogen decrease collagen production of the ACL in tissue culture. Additionally, expression of an ACL remodeling gene is increased, on average, in human females. Gender differences in ligament remodeling presumably would affect the size, shape, or internal structure of the ACL, thereby affecting its ability to sustain loads without rupturing. Therefore, it is possible that by altering ligament remodeling, a decrease in sex hormones would affect the predisposition of the ACL to injury.
Ovariohysterectomy and orchiectomy increased the prevalence of ACL injuries across sizes and breeds of dogs. These findings have established an alternative model by which the effect of gonadal gender on ACL injury can be studied, and are consistent with the view that ACL injury in humans is, in part, related to gender differences in hormone production and presentation.
We thank the staff at the Bozeman Animal Clinic for obtaining the records on the animals. We also thank Dr. Thomas Neal.
1. Amiel D, Ishizue KK, Harwood FL, Kitabayashi L, Akeson WH: Injury of the anterior cruciate ligament: The role of collagenase in ligament degeneration. J Orthop Res 7:486-493, 1989.
2. Arendt EA: Orthopaedic issues for active and athletic women. Clin Sports Med 13:483-503, 1994.
3. Arendt EA, Agel J, Dick RW: Anterior cruciate ligament injury patterns among collegiate men and women. J Athl Train 34:86-92, 1999.
4. Arendt EA, Bershadsky B, Agel J: Periodicity of noncontact anterior cruciate ligament injuries during the menstrual cycle. J Gend Specif Med 5:19-26, 2002.
5. Dahlberg L, Friden T, Roos H, Lark MW, Lohmander LS: A longitudinal study of cartilage matrix metabolism in patients with cruciate ligament rupture: Synovial fluid concentrations of aggrecan fragments, stromlysin-1 and tissue inhibitor of metalloproteinase-1. Br J Rheumatol 33:1107-1111, 1994.
6. Duval JM, Budsberg SC, Flo GL, Sammarco JL: Breed, sex, and body weight as risk factors for rupture of the cranial cruciate ligament in young dogs. J Am Vet Med Assoc 215:811-814, 1999.
7. Foos MJ, Hickox JR, Mansour PG, Slauterbeck JR, Hardy DM: Expression of matrix metalloprotease and tissue inhibitor of metalloprotease genes in human anterior cruciate ligament. J Orthop Res 19:642-649, 2001.
8. Griffin LY, Agel J, Albohm MJ, et al: Noncontact anterior cruciate ligament injuries: Risk factors and prevention strategies. J Am Acad Orthop Surg 8:141-150, 2000.
9. Haycock CE, Gillette JV: Susceptibility of women athletes to injury: Myths vs reality. JAMA 236:163-165, 1976.
10. Ireland ML: Special concerns of the female athlete. Sports Injuries 13:153-188, 1994.
11. Knapik JJ, Bauman CL, Jones BH, Harris JM, Vaughan L: Preseason strength and flexibility imbalances associated with athletic injuries in female collegiate athletes. Am J Sports Med 19:76-81, 1991.
12. Korvick DL, Pijanowski GJ, Schaeffer DJ: Three-dimensional kinematics of the intact and cranial cruciate ligament-deficient stifle of dogs. J Biomech 27:77-87, 1994.
13. Lohmander LS, Roos H, Dahlberg L, Hoerrner LA, Lark MW: Temporal patterns of stromelysin-1, tissue inhibitor, and proteoglycan fragments in human knee joint fluid after injury to the cruciate ligament or meniscus. J Orthop Res 12:21-28, 1994.
14. Marsolais GS, McLean S, Derrick T, Conzemius MG: Kinematic analysis of the hind limb during swimming and walking in healthy dogs and dogs with surgically corrected cranial cruciate ligament rupture. J Am Vet Med Assoc 222:739-743, 2003.
15. Matrisian LM: Matrix metalloproteinase gene expression. Ann N Y Acad Sci 732:42-50, 1994.
16. Meikle MC, Hembry RM, Holley J, et al: Immunolocalization of matrix metalloproteinases and TIMP-1 (tissue inhibitor of metalloproteinases) in human gingival tissues from periodontitis patients. J Periodontal Res 29:118-126, 1994.
17. Moller-Nielsen J, Hammar M: Women’s soccer injuries in relation to the menstrual cycle and oral contraceptive use. Med Sci Sports Exerc 21:126-129, 1989.
18. Myklebust G, Engebretsen L, Braekken IH, et al: Prevention of anterior cruciate ligament injuries in female team handball players: A prospective intervention study over three seasons. Clin J Sport Med 13:71-78, 2003.
19. Myklebust G, Maehlum S, Holm I, Bahr R: A prospective cohort study of anterior cruciate ligament injuries in elite Norwegian team handball. Scand J Med Sci Sports 8:149-153, 1998.
20. Nielsen AB, Yde J: Epidemiology of acute knee injuries: A prospective hospital investigation. J Trauma 31:1644-1648, 1991.
21. Salmeri KR, Bloomberg MS, Scruggs SL, Shille V: Gonadectomy in immature dogs: Effects on skeletal, physical, and behavioral development. J Am Vet Med Assoc 198:1193-1203, 1991.
22. Salmeri KP, Olson PN, Bloomberg MS: Elective gonadectomy in dogs: A review. J Am Vet Med Assoc 198:1183-1192, 1991.
23. Schneikert J, Peterziel H, Defossez PA, et al: Androgen receptors-Ets protein interaction is a novel mechanism for steroid hormone-mediated down-modulation of matrix metalloproteinase expression. J Biol Chem 271:23907-23913, 1996.
24. Slauterbeck JR, Fuzie SF, Smith MP, et al: The menstrual cycle, sex hormones, and anterior cruciate ligament injury. J Athl Train 37:275-280, 2002.
25. Verhoef-Verhallen E: The Dog Encyclopedia. Buffalo, NY, Firefly Books Ltd 1997.
26. Whitehair JG, Vasseur PB, Willits NH: Epidemiology of cranial cruciate ligament rupture in dogs. J Am Vet Med Assoc 203:1016-1019, 1993.
27. Wojtys EM, Ashton-Miller JA, Huston LJ: A gender-related difference in the contribution of the knee musculature to sagittal-plane shear stiffness in subjects with similar knee laxity. J Bone Joint Surg 84A:10-16, 2002.
28. Wojtys EM, Huston LJ, Lindenfeld TN, Hewett TE, Greenfield ML: Association between the menstrual cycle and anterior cruciate ligament injuries in female athletes. Am J Sports Med 26:614-619, 1998.