Characteristics of compulsive tail chasing and associated risk factors in Bull Terriers (original) (raw)

Tail chasing is a repetitive behavior that is expressed as slow to rapid circling with the dog's attention directed toward the tail or rapid spinning in tight circles with no apparent focus on the tail. Within the same dog, these 2 forms of expression (slow, focused; rapid, unfocused) may be expressed interchangeably. Slow chasing focused on the tail often precedes rapid unfocused spinning bouts. In its most advanced stage, tail chasing is a debilitating and potentially life-threatening behavioral condition.

Classic tail chasing, which is frequently bidirectional, is not caused by organic problems such as a brain tumor, local irritation, or other medical conditions, and neither is it typically an attention-seeking behavior. Many dogs chase their tail when separated from their owners, and when fully engaged in the behavior, the dogs appear dissociated from their environment and resistant to any form of interruption. They are often unresponsive to their owner's commands when in this state, many shun their owner's attention, and some become aggressive when attempts to interrupt them are made. Bull Terriers that are punished for tail chasing will often remove themselves to a location remote from the owner to engage in the behavior. Although tail chasing occurs in a variety of breeds, it is most commonly observed in Bull Terriers and German Shepherd Dogs.[1,2](#ref1 ref2)

The disorder has previously been attributed to opioid-mediated stereotypy3 or a seizure-related neurologic syndrome phenomenon.[4,5](#ref4 ref5) The seizure-related neurologic syndrome hypothesis suggests a putative association between tail chasing and episodic aggression, trance-like behavior, hyperactivity, sound sensitivity, and fear responses and phobias. It has also been hypothesized that this neurologic syndrome has some features in common with another disease in Bull Terriers, lethal acrodermatitis.6 In addition to having dermatologic problems, dogs with lethal acrodermatitis may have hydrocephalus and characteristic behavioral signs, such as aggression and prolonged staring. Recent studies[2,7–9](#ref2 ref7 ref8 ref9) investigating the clinical signs, development, and response to pharmacological treatment of tail chasing in dogs support a compulsive etiology similar to human obsessive-compulsive disorder. The purpose of the study reported here was to define and evaluate characteristics of tail chasing in Bull Terriers and any association with other physical and behavioral characteristics as a preliminary step toward future investigation of the inheritance of tail chasing in Bull Terriers.

Materials and Methods

Study participants—Dog owners for the study were solicited through clients of Tufts University Cummings School of Veterinary Medicine, the Bull Terrier Club of America, Bull Terrier Welfare Foundation, local dog shows, Silverwood National Bull Terrier Specialty show, and the Bull Terrier Neurological Disorder website. Convenience sampling methods were used to increase the number of dogs included in the study. Dogs were assigned to either the affected or unaffected group on the basis of the presence or absence of tail chasing as described by owners. All dogs included in the study were examined by the owner's local veterinarian or evaluated at Tufts University Cummings School of Veterinary Medicine. Most owners of Bull Terriers with tail-chasing behavior had contacted a veterinarian for treatment advice prior to participating in the study. Owners of dogs with daily tail-chasing behavior typically requested medication to treat the condition because of the severity and disruptive nature of the behavior. No medical conditions associated with the onset or propagation of tail chasing were reported. Owners completed a research questionnaire that was designed to identify those dogs affected with compulsive tail chasing as opposed to other conditions.

Survey—Owners of dogs with tail-chasing behavior were sent a survey designed to solicit information regarding the appearance of tail chasing, age of onset, frequency of bouts, duration of bouts, and eliciting triggers. Owners were asked to provide their best estimate for age of onset and frequency and duration of tail chasing bouts on the basis of their history of living with the dog in the home. Questions regarding various physical and behavioral conditions were also included in the survey. As new behavior patterns emerged over the 16-year duration of data gathering, 2 additional questions were added to assess the degree of interference that the dog's tail chasing caused regarding the dog's quality of life and the owner's relationship with the dog. Subsequent study participants were asked to report how the amount of time the dog spent tail chasing interfered with its normal daily activities and relationship with the owner. The owner was asked to score 0 for no interference, 1 for slight interference, 2 for mild to moderate interference, 3 for definite interference that was still manageable, and 4 for interference that incapacitated every aspect of the dog's or owner's life. When responses were unclear, owners were contacted directly for clarification.

For comparative purposes, sex and neuter status as well as behavioral data were collected for all Bull Terriers in the study regardless of whether they had tail-chasing behavior. Owner-directed aggression was assessed by having owners complete a previously published checklist indicating interactions in which their dog might challenge them, including but not limited to resource guarding, routine handling, and mild restraint.10 Owner-directed aggression included behaviors of growling, lift lip, snapping, or biting. Episodic aggression was defined as recurrent, unprovoked attacks directed toward people, other animals, or objects in the household. The attacks are explosive, violent, sudden, and unpredictable, with little to no provocation and no typical warning signals. The behavior is out of character for the dog's normal demeanor. An awake but peaceful dog may have a transition state, often of quite short duration, in which the eyes glaze, followed by an attack.

Statistical analysis—The survey was designed to evaluate dogs on a large set of binary random variables, among these being tail-chasing behavior (yes or no), sex (male or female), coat color (white or other color), and a list of other behavioral variables. Data were analyzed in accordance with the concept of log-linear models.11 By use of this approach, tail chasing was considered as just another observed binary random variable (ie, tail chasing was not considered as caused by the action of other variables). The count of dogs was evaluated in cross-sectional tables to evaluate whether tail-chasing behavior is observed independent of other behaviors, sex, or coat colors. The modeled variable was thus a count, not the presence or absence of tail chasing; the count of dogs was classified by sex, coat color, tail chasing status, and all other binary behavior observations.

The first step in this analysis was to estimate the correlation between all the assembled elements of the survey. Being binary characters, the polychoric correlation12 was computed by use of a software program of the public domain.a Once computed, these correlations were used to ascertain those behaviors and characters that were most closely correlated with tail chasing. Nine such characters were found to have estimates of correlations with tail chasing, and with each other, to warrant further investigation. However, 7 behavioral variables (fly-snapping, shadow and light chasing, fly-snapping and shadow chasing combined, flank sucking, owner aggression 1 interaction, owner aggression 2 interaction, and deafness) were found to have poorly estimated correlation coefficients (ie, SEs well above 1.0) and were excluded from further analysis.

The remaining 10 variables formed the basis of a series of log-linear analyses. All possible sets of 8 variables, where tail chasing was kept in each set (this being the variable of principle interest), were evaluated. This was done to balance the interest in estimating unknown effects against the preponderance of empty cells given the limited number of dogs observed across the many subclasses. The initial analyses fit all 8 variables, including models for 2- and 3-way interactions. In each setting, nonsignificant variables were removed and a suitable submodel was chosen through comparisons of the Akaike Information Criterion and the residual deviance. In repeating this process over the 9 subsets of variables, a subset of variables and their interactions were arrived at, which were consistently found to be associated with each other. All computations were performed with the general linear model function of the R programing language with the dependent variable of counts and a Poisson model with and without consideration of overdispersion.a

Results

Data were collected on 333 Bull Terriers, 145 dogs with tail-chasing behavior, and 188 unaffected dogs; however, not all dogs had information recorded for each descriptive trait. Because some owners did not answer certain questions, the total number of dogs for every variable varied.

Sex and neuter status—A total of 169 female and 164 male Bull Terriers were included in the analysis. Reproductive status was not reported for 6 dogs; reproductive status was known for 144 dogs with tailchasing behavior and 183 unaffected dogs. Thirty-one percent (45/144) of dogs with tail-chasing behavior were neutered males, and 18% (33/183) of unaffected dogs were neutered males. Thirteen percent (19/144) of dogs with tail-chasing behavior were sexually intact females, and 23% (42/183) of unaffected dogs were sexually intact females. Thirty-one percent (44/144) of dogs with tail-chasing behavior were spayed females, and 34% (62/183) of unaffected dogs were spayed females. Twenty-five percent (36/144) of dogs with tail-chasing behavior were sexually intact males, and 25% (46/183) of unaffected dogs were sexually intact males.

Age of onset—Age of onset for tail chasing was known or confidently estimated for 61% (89/145) of dogs with tail-chasing behavior. The median age of onset was 6 months. The range of age of onset was 2 months to 6 years of age.

Frequency and duration of bouts—Frequency (daily tail chasing or less than daily tail chasing) was recorded for 109 of the 145 dogs with tail-chasing behavior. Of these 109 dogs, 74% (81/109) chased their tail daily while 26% (28/109) chased their tail less than daily. Information on the duration (< 2 minutes, 2 to 30 minutes, > 30 minutes) of tail chasing was available for 67% (97/145) of the dogs with tail-chasing behavior. The average duration of a tail-chasing bout was > 30 minutes for 28% (27/97) of the dogs. Average tail-chasing bouts ranged from 2 to 30 minutes for 29% (28/97) of the dogs. Average tail chasing bouts that lasted < 2 minutes were reported for 43% (42/97) of the dogs. For 94 dogs with tail-chasing behavior, both frequency and duration data were available for cross tabulation. Of the 74 dogs that chased their tails daily for which duration data were also available, 70% (52/74) did so for 2 to 30 minutes or > 30 minutes. By comparison, of the 20 dogs that did not chase their tails daily for which duration data were also available, only 10% (2/20) did so for 2 to 30 minutes or > 30 minutes (χ2 = 23.8; P < 0.001).

Ability to interrupt—Owners of 99 dogs completed the question regarding the ease with which they could interrupt their dog's tail-chasing behavior. Eighty-seven (88%) owners reported that they could interrupt their dog from tail chasing. However, 48% (42/87) of these owners reported that the dog would immediately or within minutes resume tail chasing, indicating that the interruption was only temporary. Thirty-four percent (30/87) of owners indicated that the dog would either not resume or resume tail chasing at a later time, suggesting that these dogs were more easily interrupted. Seventeen percent (15/87) of owners reported that whether the dog resumed tail chasing following an interruption depended on the situation. If the dog was bored or stressed at the time, it would likely resume immediately to within a few minutes. If the dog's attention could be redirected onto another activity, the tail chasing could be successfully interrupted. Ten percent (10/99) of owners could not interrupt their dog's tail chasing, and 2 owners (2%) did not attempt to interrupt the behavior.

Triggers—As part of the survey, owners were asked an open-ended question regarding what conditions triggered the initial onset and continued elicitation of their dog's tail-chasing behavior. A total of 129 owners responded, but many owners listed > 1 trigger for tail chasing, resulting in 239 reported triggers. Triggers were grouped into 9 general categories. Thirty-one percent (73/239) of triggers for tail chasing involved situations that increased the dog's level of arousal or frustration. Examples included owner departures and returns, visitors, food, presence of other dogs, aggressive interactions with other dogs, exposure to moving cars and bicycles, being released from a crate, going outside, the owner cooking, or general excitement. New, unpredictable or restrictive environments were submitted as triggers for 18% (43/239) of dogs. This category included crate confinement, new locations, closed areas, room corners, and return to indoors. Lack of mental or physical stimulation and insufficient interaction with the owner or conspecific was listed as a trigger for 15% (35/239) of dogs. Sensitivity to sound was also cited as a trigger for tail chasing for 10% (25/239) of dogs. Specific sounds reported included loud noises, rain on the roof and windows, running water, vacuum, hair dryer, lawn mower, microwave bells, and other household beeping-type alarms. Physical conditions associated with the onset and continued elicitation of tail chasing were cited for 8% (20/239) of dogs and included first or second estrus cycle, diarrhea, defecation, flatulence, hunger, fatigue, allergies, and tail injury. Ten percent (23/239) of owners reported no discernible trigger associated with tail chasing and viewed the behavior as spontaneous. Six percent (14/239) of owners reported that their dog chased their tail at certain times of the day or in response to other situations. A change in the dog's social group was associated with tail chasing in 2% (4/239) of dogs. The sight of certain objects (eg, broom or bag) triggered tail chasing in 1% (2/239) of dogs.

The development of tail-chasing behavior differed among individuals, varying from a gradual to a sudden onset. For some dogs, the onset of tail-chasing behavior occurred suddenly with no apparent trigger, whereas for other dogs, the onset coincided with exposure to identifiable psychological, physiologic, or environmental triggers that were interpreted as increasing anxiety or arousal levels.

Interference with dog's quality of life and owner's relationship with dog—Fifty-one owners responded to the question of how tail chasing interfered with their dog's normal activities. Twenty-seven percent (14/51) of owners responded that the behavior occurred with such high frequency and duration that it was negatively affecting the dog's ability to function to the point of incapacitating every aspect of its life. Eighteen percent (9/51) of owners reported definite interference but that the behavior was still manageable. Sixteen percent (8/51) reported that the interference was slight (5/51) to mild to moderate (3/51). Thirty-nine percent (20/51) reported that the dog's tail-chasing behavior posed no interference.

Fifty-two dog owners responded to the question of how much the dog's tail-chasing behavior interfered with the relationship with their dog. Twenty-five percent (13/52) of owners reported that their dog's tail-chasing behavior posed an incapacitating interference in the relationship with their dog. Another 25% (13/52) of owners reported that the tail-chasing behavior posed a definite but manageable interference. Mild to moderate interference was reported by 8% (4/52) of owners. Twelve percent (6/52) of owners found that their dog's tail-chasing behavior caused a slight interference in the relationship with their dog, whereas 31% (16/52) reported no interference.

Injuries—Dog owners were not specifically asked to report injuries until later in the study. Fifteen dogs incurred tail injuries as a direct result of biting the tail during tail-chasing bouts, 7 of which necessitated amputation because of the severity of the injuries. Six dogs sustained injuries to the head and tarsi in the form of abrasions and cuts from banging into walls and furniture while tail chasing. Two owners reported that their dog's digital pads and nails were severely worn because of excessive tail chasing. Eight dogs that did not respond to treatment were euthanatized because of excessive tail chasing. Another 8 dogs were euthanatized because of tail chasing in conjunction with owner-directed aggression.

Degrees of expression—On the basis of clinical impression data from owner reports regarding frequency, duration, interruptability, and degree of disruption to the dogs' normal functioning and the owners' relationship with the dogs, 2 subsets of tail-chasing behavior were identified that differed in degree of expression for the Bull Terrier population of this study: clinical and subclinical tail-chasing behavior.

Based on differences in frequency of expression only, dogs with clinical tail-chasing behavior (81/109) were those dogs that underwent multiple tail-chasing bouts on a daily basis. They often appear dissociated from the environment and unresponsive to owner commands. These dogs were not easily interrupted from tail chasing, and many had signs of anxiety or aggression if restrained. If the owner was able to interrupt the dog from tail chasing, usually by restraint or redirection onto a preferred toy or object, the dog typically resumes immediately upon release or removal of the object. On the basis of owner reports, the duration of individual tail-chasing bouts for dogs with clinical tail-chasing behavior generally ranged from 60 seconds to > 2 hours. Dogs with clinical tail-chasing behavior often chased their tail daily, usually to the exclusion of other normal activities, with tail-chasing behavior occupying an owner-estimated 30% to 80% of the dog's daily time budget. The behavior was disruptive for both the dog and owner and in many instances was viewed by the owner as incapacitating or definitely interfering with the dog's normal functioning (23/51 [45%]) and relationship with the owner (26/52 [50%]). Owners reported that the dog often chased its tail rather than interact with their human companions or other dogs and commented that their dog was no longer a good companion. These dogs were not reliably responsive to training commands and could not be exercised, as they chased their tail rather than walking on leash or chasing a ball. A few dogs were reported to continue to chase their tail while they urinated and defecated. Some dogs with clinical tail-chasing behavior mutilated their tail, sustained tail fractures while tail chasing, had weight loss (either because they chased their tail rather than eat or the excessive physical activity resulted in weight loss), had signs of exhaustion, and had abraded foot pads, all indicators of the severity of the behavior problem.

By definition, dogs with subclinical tail-chasing behavior (28/109) had an apparently milder form of tail chasing that occurred typically 1 or 2 times/wk or 1 or 2 times/mo, usually in response to specific, predictable stimuli. The episodes were comparatively short in duration, usually < 3 minutes, and were easily interrupted by owners via distraction techniques. These dogs did not appear dissociated from their environment, responded to owner commands while tail chasing, and did not immediately resume tail chasing once they had been interrupted. Their tail-chasing behavior was not reported to interfere with the dog's normal functioning or its relationship with the owner. However, some dogs with subclinical tail-chasing behavior later developed clinical tail-chasing behavior in response to changes in their environment or physiologic condition.

Fear responses and phobias—A total of 60 Bull Terriers, both dogs with tail-chasing behavior (n = 36) and unaffected dogs (24), were described by owners as having phobias of common objects or situations in their environment. Some owners reported > 1 trigger for their dog's fearful behavior. Phobias were divided into 3 categories: natural environment, social, and situational. Natural environment phobias included the following: vacuums (17/60); loud noises (8/60) including vehicles, mechanical and machinery noise, and a noise-making doll; water (9/60) including hose water, bath water, and rain; stairs (2/60); heights (1/60); thunderstorms (1/60); slippery floors (5/60); mopping, sweeping, and raking (9/60); reflective surfaces (3/60); novel household objects in environment (5/60); common household objects out of place (5/60); moving or stationary automobiles (2/60); fear of own shadow (1/60); fear of outdoors in daytime (1/60); steel doors (2/60); and doorways (1/60). Four dogs had social phobias in the presence of people dressed in white (1/60), strangers (1/60), people speaking loudly (1/60), and any human contact (1/60). Ten dogs had situational phobias involving the veterinarian's office (3/60), grooming shop (1/60), pet store (1/60), show ring (1/60), elevators (1/60), crates (1/60), walks (1/60), and being away from home (1/60).

Owner-directed aggression—From the total of 333 dogs, 14 dogs with tail-chasing behavior expressed owner-directed aggression in ≥ 5 types of interactions with owners. Five unaffected dogs expressed owner-directed aggression in ≥ 5 situations.

Episodic aggression—Twenty-two dogs had episodic aggression. Sixteen dogs with episodic aggression were dogs with tail-chasing behavior, and 6 were unaffected dogs. Ninety-five percent (20/21) of dogs with episodic aggression (not observed for 1 dog with episodic aggression) went directly from a sleep state to an attack state. All episodes were < 60 seconds in duration as described by owners. Forty-one percent (9/22) of owners also marked situations or interactions on the owner-directed aggression checklist that resulted in aggression. In all 9 instances, aggression was reported to be triggered by only 1 or 2 specific interactions.

Association of tail chasing with other variables—For determination of what physical and behavioral characteristics were correlated with tail chasing, 10 variables were evaluated (Table 1). Correlations (and their associated SEs) among the 10 variables were considered in the analysis. Increased risk for developing tail chasing was not associated with coat color; dogs with white or other color coats appeared at equal risk. A history of tonic clonic seizures, deafness, or skin allergies also was not associated with increased risk of developing tail chasing, and neither was a history of shadow chasing, flysnapping, flank sucking, or noise sensitivity.

Table 1—

Polychoric correlation values (above the diagonal [bolded 1.000 values]) and their SEs (below the diagonal) among behavioral and explanatory characteristics in 333 Bull Terriers (145 dogs with tail-chasing behavior and 188 unaffected dogs).

Variables Sex Coat Tailch Tr Phobia Noise Ownagg EA Seizures Skin
Sex 1.000 0.085 0.193 –0.200 –0.032 –0.045 0.171 –0.110 –0.182 –0.003
Coat 0.008 1.000 0.052 –0.005 0.135 –0.016 0.267 –0.006 –0.359 0.353
Tailch 0.007 0.008 1.000 0.107 0.296 0.172 0.377 0.374 0.273 0.156
Tr 0.008 0.009 0.009 1.000 0.204 0.350 0.158 0.155 0.453 0.167
Phobias 0.010 0.010 0.009 0.011 1.000 0.465 0.215 0.352 0.070 0.147
Noise 0.010 0.010 0.009 0.009 0.008 1.000 0.339 0.644 0.465 0.029
Ownagg 0.019 0.018 0.016 0.020 0.022 0.018 1.000 0.265 0.151 –0.003
EA 0.017 0.019 0.015 0.019 0.017 0.009 0.032 1.000 0.542 0.333
Seizures 0.032 0.037 0.030 0.024 0.042 0.025 0.063 0.029 1.000 0.277
Skin 0.008 0.007 0.007 0.009 0.010 0.010 0.020 0.015 0.030 1.000

However, on the basis of polychoric correlations, several behaviors were aggregated (Table 1). Tail chasing occurred frequently with owner-directed aggression and episodic aggression (polychoric correlations of 0.377 and 0.374, respectively). Similarly, the behavior of noise sensitivity occurred frequently with episodic aggression (polychoric correlation of 0.644).

A log-linear model including terms for tail chasing, episodic aggression, trance-like behavior, phobia, sex, and owner-directed aggression (≥ 5 types of interactions), along with all possible 2-way interactions, was developed (Table 2); this model captured those variables that provided the best fit to the observed counts. Not presented is the model with all possible 3-way interactions, none of which proved to be significantly different from zero, a result that was repeated across all models and submodels evaluated. Other submodels not presented included selected 3-way interactions; however, at no time were any of these terms significantly different from zero.

Table 2—

Parameter estimates and their SEs for the log-linear model without consideration of overdispersion including counts of tail chasing, episodic aggression, trance-like behavior, sex, owner-directed aggression, and phobia, and all possible 2-way interactions for 145 Bull Terriers with tail-chasing behavior.

| Parameter | Estimate | SE | t value* | Pr. > |t|† | | --------------- | ------------ | ------ | ------------------------------ | ----------------- | | Intercept | 1.908 | 0.278 | 6.854 | 0.000 | | Tailch | 0.035 | 0.005 | 7.512 | 0.000 | | Ownagg | –0.005 | 0.317 | –0.016 | 0.987 | | EA | –2.871 | 0.545 | –5.271 | 0.000 | | Tr | –3.493 | 0.720 | –4.850 | 0.000 | | Sex | –1.189 | 0.441 | –2.698 | 0.010 | | Phobia | –0.104 | 0.358 | –0.292 | 0.772 | | Tailch × Ownagg | 0.009 | 0.008 | 1.141 | 0.260 | | Tailch × EA | 0.604 | 0.219 | 2.752 | 0.009 | | Tailch × Tr | 1.057 | 0.303 | 3.487 | 0.001 | | Tailch × Sex | 0.078 | 0.021 | 3.730 | 0.001 | | Tailch × Phobia | 0.005 | 0.007 | 0.735 | 0.466 | | Ownagg × EA | 0.334 | 0.564 | 0.593 | 0.557 | | Ownagg × Tr | –0.226 | 0.535 | –0.422 | 0.675 | | Ownagg × Sex | 0.282 | 0.302 | 0.935 | 0.355 | | Ownagg × Phobia | –0.220 | 0.263 | –0.835 | 0.408 | | EA × Tr | 1.209 | 0.778 | 1.555 | 0.127 | | EA × Sex | 0.620 | 0.556 | 1.116 | 0.271 | | EA × Phobia | –0.001 | 0.529 | –0.001 | 0.999 | | Tr × Sex | 0.883 | 0.505 | 1.749 | 0.088 | | Tr × Phobia | 0.101 | 0.438 | 0.231 | 0.818 | | Sex × Phobia | 0.257 | 0.302 | 0.851 | 0.399 |

Episodic aggression and trance-like behavior, along with sex, had significant interactions with tail chasing (Table 2). There was an 8% increase in tail-chasing males (exp 0.078 = 1.081). Positive parameter estimate for the Poisson model demonstrates that the observed frequency of male dogs that chase their tail, as well as male dogs that have had either episodic aggression or trance-like behavior, is more common than can be explained by a model of independent occurrence of these characteristics (ie, parameter estimates of 0.078, 0.620, and 0.883 vs −1.189, −2.871, and −3.493, respectively). In fact, the Akaike information criterion for the model with only main effects of tail chasing, episodic aggression, trance-like behavior, phobia, sex, and owner-directed aggression was 200.82, whereas the Akaike information criterion for the log-linear model was 184.84. Accordingly, the log-linear model provided a better explanation for the observed counts of these traits than that of an independent (ie, no interaction) model. Tail-chasing behavior was more commonly found with episodic aggression and trance-like behavior than what would be expected given the probability of these 3 behaviors taken individually. However, the failure of a 3-way interaction to be significant, or for the episodic aggression and trance-like behavior interaction to be significant (P = 0.127), suggested that tail chasing was segregated with each of these aberrant behaviors, but that episodic aggression and trance-like behavior were unrelated.

Discussion

To our knowledge, the study reported here represents the largest study of tail chasing in Bull Terriers to date. Tail chasing has been documented as a form of canine compulsive disorder.[2,7–9](#ref2 ref7 ref8 ref9) Some of the previous assertions about this purported canine compulsive disorder were confirmed and more closely detailed in the present study, while new findings, in particular, increased male susceptibility and associations of tail chasing with episodic aggression and trance-like (staring) behaviors, were made. These conclusions were based on significant interaction terms for these behaviors with tail chasing. Interestingly, there were initial models for this study in which owner-directed aggression and phobias were found to have a significant interaction with tail chasing. However, these initial exploratory models did not include terms for episodic aggression or trance-like behavior. Accordingly, overall interpretation of the final model developed in this study suggests that although owner-directed aggression and phobias had some impact on tail chasing, trance-like behavior and episodic aggression associate more strongly with tail chasing. This general observation was also supported in the polychoric correlation values (Table 1).

Our clinical perception was that many Bull Terriers with tail-chasing behavior had mild owner-directed aggression (as distinct from violent episodic aggression), although statistical analysis indicated only a loose association. A clinical explanation for the association of tail chasing with owner-directed aggression may be that increased anxiety, frustration, or conflict associated with the performance of tail chasing or the owners' attempts to interrupt tail chasing lower the threshold for the dogs' aggressive response toward their owners.

The weak association between compulsive tail chasing and phobic conditions is explicable if tail-chasing compulsion, like human obsessive-compulsive disorder, is regarded as an anxiety disorder.13 According to the diagnostic manual of the American Psychiatric Association,13 various anxiety-type disorders, including specific phobia, social phobia, and panic disorder, are comorbid with obsessive-compulsive disorder. Findings of the study reported here suggest that a similar association of tail-chasing compulsion and anxiety-type disorders exists for Bull Terriers and support the biological homology concept of canine and human compulsive behavior.

Although tail chasing in dogs is commonly described as a compulsive disorder or partial seizure disorder,[2,4,7–9](#ref2 ref4 ref7 ref8 ref9) findings of the present study lead to another possibility. Males had a slight (8%) but significantly greater risk for developing tail chasing than females (Table 2). Furthermore, tail chasing in Bull Terriers is closely associated with episodic aggression and trancelike behavior. In terms of the cluster of clinical signs and manifestations of tail chasing, it is speculated that this syndrome in Bull Terriers may have features in common with autism in humans. Autism is also more common in males, is associated with explosive aggression, trance-like staring, and involves repetitive movements and self-injurious behavior.[14–16](#ref14 ref15 ref16) In addition, autism is characterized by autonomy, impaired social interactions, and obsession with objects.[16–18](#ref16 ref17 ref18) Many owners of Bull Terriers with tail-chasing behavior describe their dogs as asocial, somewhat withdrawn, and abnormally preoccupied with objects, such as balls or sticks. Indeed, many owners use objects to redirect their dog from tail chasing, and the dog responds to the distraction with similar intensity.

A final possible explanation for the relationship between tail chasing, trance-like behavior, and episodic aggression is that all stem from underlying complex partial seizures. Bull Terriers with tail-chasing behavior have been shown to have epileptiform activity on electroencephalographic recordings.4 Furthermore, violent episodic aggression in dogs has previously been described as a seizure-related problem,[19,20](#ref19 ref20) and trance-like behavior may represent a form of partial seizure in which consciousness is altered but not lost. It is noteworthy that epileptic seizures are reported in 4% to 32% of humans with autism.[21,22](#ref21 ref22)

The present study provides detailed phenotypic and developmental information about tail chasing in dogs and illustrates some intriguing parallels with human obsessive-compulsive disorder and possibly autism. Compulsions are typically time-consuming in clinical tail chasing and obsessive-compulsive disorder, and the behavior seems to be performed in response to and in order to alleviate stress. If a dog with tail-chasing behavior or a human with obsessive-compulsive disorder is physically prevented from engaging in a compulsion, the result is mounting anxiety or tension.13 Repeating actions in an excessive or unreasonable manner is a way that obsessive-compulsive disorder presents in humans and is the primary manifestation of tail chasing in dogs.7 In addition, the onset of tail chasing typically occurs in young adulthood shortly before or just after puberty. Early-life onset is also a feature of both obsessive-compulsive disorder and autism in humans.[23,24](#ref23 ref24)

As with obsessive-compulsive disorder, a familial pattern of expression of tail chasing has been described.7 Obsessive-compulsive disorder and tail chasing affect so-called occupational or normal daily functioning and, as shown in this study, have a detrimental effect on social activities and relationships with others. Both disorders can be disruptive to overall functioning and may lead to self-injury. Furthermore, there is comorbidity between tail chasing and various phobias, as occurs in obsessive-compulsive disorder. That said, over 10% of autistic children have noise phobias and clinically diagnosable problems of an anxiety-related disorder, and in addition, autism is familial in expression, with an increased risk among siblings.24 Dogs with tail-chasing behavior seem to be of an anxious disposition and have been shown to respond to the same types of medications.[2,25](#ref2 ref25) Although there are many parallels between canine compulsive tail chasing and human obsessive-compulsive disorder, if tail chasing in Bull Terriers is more closely related to autism, a new channel of translational research could be pursued relative to this common and extremely debilitating condition.

References