bsessive-compulsive disorder (OCD) in pet dogs
and cats is usually recognized because of a com-
pulsive component (ritualistic, stereotypic behaviors).
Obsessive-compulsive behaviors in dogs can include
those characterized by circling, tail chasing, flank suck-
ing (particularly in Doberman Pinschers), fence run-
ning, fly biting, self mutilation, hair or air biting, pica,
pacing or spinning, staring and vocalizing, some aggres-
sions, self-directed vocalizing, and fabric sucking or
chewing. In cats, self-mutilation, excessive grooming,
tail chasing, and wool or fabric sucking or chewing are
also signs of OCD.
Because behaviors seen in OCD are
often normal behaviors performed in an inappropriate,
excessive, or out-of-context manner,
particularly important in elucidating whether the
patient truly has OCD. The purpose of this retrospec-
tive study was to determine clinical features, response
to treatment, and outcome in dogs and cats with OCD.
Criteria for Selection of Cases
All dogs and cats seen at the Behavior Clinic of the
University of Pennsylvania from January 1989 through
December 2000 were assessed for OCD as part of a
thorough history and by use of a standardized ques-
Inclusion criterion was a diagnosis of OCD
made on the basis of finding repetitive, stereotypic
motor, locomotory, grooming, ingestive, or hallucino-
genic behaviors that occurred out of context or in fre-
quency or duration in excess of that required to accom-
plish the ostensible goal or in a manner that interfered
with the animal’s ability to function in its social envi-
As a result of the standardized screening
of all patients, the diagnosis of OCD was made for
some patients for whom the affiliated behaviors were
not the clients’ primary complaint, and in a few ani-
mals the diagnosis of OCD was secondary to another
primary condition (eg, obsessive-compulsive spinning
and barking that is particularly performed during
bouts of separation anxiety).
From the medical records, patients were classified
by breed; weight; sex and neuter status; age at neuter-
ing; age at onset of OCD; source; number of humans,
cats, and dogs in the household; training or schooling;
duration of treatment; class of OCD (hallucinatory,
vocalization, locomotory, grooming or self-mutilation,
ingestion [pica or coprophagy])
; whether other
dogs or cats in the household had the same or similar
behavior and if so, which animals; concurrent behav-
ioral diagnoses; medication used and duration; extent
of behavior modification; and outcome.
For assessment of animal age and duration of
OCD, actual birth dates were used when known. If the
day was unknown but the month of birth was known,
the patient was assigned a birth day of 15. If both
month and day were unknown, the patient was
assigned a birth date of June 30 of the year suspected
to be the birth year.
JAVMA, Vol 221, No. 10, November 15, 2002 Scientific Reports: Retrospective Study 1445
Clinical features and outcome in dogs and cats
with obsessive-compulsive disorder:
126 cases (1989–2000)
Karen L. Overall, VMD, PhD, DACVB, and Arthur E. Dunham, PhD
Objective—To determine clinical features and out-
come in dogs and cats with obsessive-compulsive
Animals—103 dogs and 23 cats.
Procedures—Records of patients with OCD were
analyzed for clinical features, medication used, extent
of behavior modification, and outcome.
Results—Most dogs affected with OCD had been
obtained from breeders. Male dogs significantly out-
numbered females (2:1). Female cats outnumbered
male cats by 2:1 in a small sample. Most affected
dogs lived in households with 2 or more humans
and other dogs or cats, and had some formal train-
ing. Client compliance with behavior modification
was high. A combination of behavior modification
and medication resulted in a large decrease in inten-
sity and frequency of OCD in most animals.
Clomipramine was significantly more efficacious for
treatment in dogs than was amitriptyline. Only 1 dog
and 1 cat were euthanatized because of OCD during
Conclusions and Clinical Relevance—OCD in dogs
does not appear to be associated with lack of training,
lack of household stimulation, or social confinement.
In cats, OCD may be associated with environmental
and social stress. Obsessive-compulsive disorder
appears at the time of social maturity and may have
sporadic and heritable forms. With appropriate treat-
ment (consistent behavior modification and treatment
with clomipramine), frequency and intensity of clinical
signs in most dogs and cats may decrease by > 50%.
Success appears to depend on client understanding
and compliance and the reasonable expectation that
OCD cannot be cured, but can be well controlled.
J Am Vet Med Assoc
From the Biology Department (Dunham) and the Center for
Neurobiology and Genetics-Psychiatry Department, School of
Medicine (Overall), University of Pennsylvania, Philadelphia,
Data entry was supported by Novartis Animal Health, Greensboro,
NC; medication for some patients was provided by CIBA, Summit,
NJ, and Novartis Animal Health, Greensboro, NC.
The authors thank Len Rhone and Ken Mullin for technical assis-
Address correspondence to Dr. Overall.
Sources of animals included stray or found, breed-
er (serious or show), breeder (backyard), Society for
the Prevention of Cruelty to Animals or humane shel-
ter, breed rescue service, newspaper adoption adver-
tisement (not breeder), pet store, friend, and other.
Training or schooling categories included no school,
trained by client, puppy kindergarten, group lessons
(basic), group lessons (advanced), private trainer at
house, and private trainer (sent to trainer).
The general behavior modification scheme for dogs
consisted of 3 basic steps designed to stop unwanted
behaviors and reward preferable behaviors that were
directly competitive with the undesirable behaviors. In
step 1, clients were to cease even unintentional reward
for the undesirable behavior. In step 2, clients were
asked to follow a passive behavior modification pro-
gram designed to teach dogs to sit quietly, look at the
person from whom they were seeking attention or with
whom the interaction was occurring, and wait to take
the cues about the appropriateness of their behavior
from those people (ie, Protocol for Deference
). In step
3, clients were requested to begin active behavior mod-
ification designed to teach the dogs to relax when they
would otherwise be engaged in the behavior, in
exchange for which the dog was given a food or play
reward (ie, Protocol for Relaxation: Tier 1
For clients with feline patients, steps 1 and 3 were
recommended; step 2 was modified to recommend that
the client only interact with the cat when the cat was
calm. Cats were not required to sit in step 3, although
this was encouraged, but the clients were advised to
engage the cat in an enjoyable behavior that was direct-
ly competitive with the undesirable behavior.
Behavior modification was assessed on the basis of
client responses and clinician opinion.
For some of the patients, medication had been pre-
scribed by the referring veterinarian. When the first
patient in this study was treated, treatment with tricyclic
antidepressants (TCAs; eg, clomipramine and specific
serotonin reuptake inhibitors [SSRIs]) was cost pro-
hibitive (> $10/d). Clomipramine is now affordable and
available in a canine formulation.
Accordingly, the TCA
(amitriptyline) was the first drug of choice for all
patients with conditions that primarily involved anxiety
and were enrolled in the early portion of the study.
Clomipramine was only used instead of amitriptyline if
amitriptyline was not efficacious or caused gastrointesti-
nal disorders. Because of its histamine-1 receptor antag-
onist properties, the TCA doxepin was the first drug of
choice for conditions that primarily involved pruritus,
followed by amitriptyline.
Other medications were
occasionally prescribed as dictated by alterations in the
patients’ behaviors, and for some animals, combinations
of medications were both more efficacious and more
cost-effective for clients, as described elsewhere.
When the senior author was awarded a grant pro-
viding clomipramine at no cost, clomipramine was used
as the first drug of choice for treating OCD, as it has
been in humans. This circumstance allowed us to retro-
spectively compare the relative effects of amitriptyline
and clomipramine for patients that met the criteria for
OCD. If a patient was administered amitriptyline and
acceptable results were obtained (marked decrease in
intensity and frequency of OCD behaviors), administra-
tion of the drug was maintained and treatment was con-
sidered successful. If a patient was treated initially with
amitriptyline and acceptable results were not obtained,
clomipramine was administered instead; if acceptable
results were obtained with clomipramine, treatment
with amitriptyline was considered a failure and treat-
ment with clomipramine was considered a success.
Patients for whom treatment with clomipramine failed
were likewise treated with another medication.
were used consistently and
included amitriptyline (1 mg/kg [0.45 mg/lb], PO, q 12
h for 10 days; if no change in behavior was detected,
the dosage was increased to 2 mg/kg [0.91 mg/lb] for
10 days; if still no change was detected, the medication
was changed [if treatment was efficacious it was con-
tinued for a minimum of 1 month]); doxepin (3 to 5
mg/kg [1.4 to 2.3 mg/lb], PO, q 8 h to q 12 h for a min-
imum of 1 month); clomipramine (1 mg/kg, PO, q 12
h for 14 days, then 2 mg/kg, PO, q 12 h for 14 days,
then 3 mg/kg, PO, q 12 h for 1 month, for a minimum
of 2 months of treatment [the gradual increase in
dosage was intended to minimize gastrointestinal dis-
orders]); and the SSRIs, sertraline and fluoxetine (1
mg/kg, PO, q 24 h for 2 months initially). Treatment
was then continued at the minimum effective dose nec-
essary to control the behavior.
All patients received full physical and laboratory
evaluation prior to treatment. Any nonspecific derma-
tologic, medical, or neurologic signs potentially associ-
ated with OCD were evaluated by the veterinarians
trained in the respective specialty prior to behavioral
treatment. When warranted, patients also received var-
ious nonroutine diagnostic procedures (eg, assessment
of thyroid function and magnetic resonance spec-
troscopy of the brain).
Eight clinicians were involved in evaluation of the
dogs reported here. All clinicians adhered to the same
diagnostic criteria, and at least 2 clinicians reviewed each
diagnosis. Although many records contained detailed out-
come data, quality of the description of the behaviors
depended on clinician expertise versus resident expertise,
whether the behaviors were videotaped, client capability
and motivation in keeping data logs, and knowledge
accrued by managing increasing numbers of these dogs.
Accordingly, outcomes were broadly grouped and statisti-
cally treated as class variables to evaluate older and newer
cases equivalently. Response variables for outcome were
small decrease in intensity (≤ 50%), large decrease in
intensity (> 50%), small decrease in frequency (≤ 50%),
large decrease in frequency (> 50%), the behavior stopped
totally, no change, small increase in intensity (< 50%),
large increase in intensity (> 50%), small increase in fre-
quency (< 50%), large increase in frequency (> 50%),
died, euthanatized because of OCD, euthanatized for
other reasons, placed, other, and unknown. Patients clas-
sified with unknown outcomes were those lost to follow-
up. For patients with concurrent behavioral diagnoses,
alterations in intensity and severity reported here pertain
only to OCD.
Data were analyzed by use of the log-likelihood
ratio test, and relevant nonparametric tests as indicat-
ed for categoric data.
1446 Scientific Reports: Retrospective Study JAVMA, Vol 221, No. 10, November 15, 2002
One hundred three dogs met inclusion criteria,
including 5 sexually intact females, 26 spayed females,
20 sexually intact males, and 52 neutered males.
Twenty-three cats (1 sexually intact female, 14 spayed
females, 1 sexually intact male, and 7 neutered males)
that met the diagnostic criteria had sufficiently com-
plete records to be included in the study.
Category of OCD—The most common category of
OCD in dogs was that associated with grooming or
self-mutilation, followed by OCD involving locomo-
tion and OCD involving signs of hallucinations (Fig 1).
For cats, the most common category of OCD involved
grooming or self-mutilation (Fig 2).
There was no significant association (log likeli-
hood ratio test) between sex and neuter status catego-
ry of affected dogs and the category of OCD they had.
However, compared with the sex and neuter data for
the entire veterinary teaching hospital canine popula-
tion during the study period (n = 47,473), male dogs
were significantly over-represented in the OCD popu-
lation (log likelihood ratio test statistic, G
13.256 [P < 0.05]). In addition, neutered males were
over-represented in the OCD population, compared
with non-neutered males (log likelihood ratio test sta-
= 36.641 [P ≤ 0.001]).
Breed—The population of dogs with OCD com-
prised 18 mixed-breed dogs; 12 German Shepherd
Dogs; 7 Rottweilers; 6 Golden Retrievers; 5 each of
Dalmatians, Labrador Retrievers, and Lhasa Apsos; 3
each of Doberman Pinschers, Poodles, Soft-Coated
Wheaten Terriers, and English Springer Spaniels; 2
each of American Pit Bull Terriers, Bulldogs, Great
Danes, Miniature Schnauzers, and Cocker Spaniels;
and 1 dog each of 22 other breeds. This distribution
did not differ significantly from that in the overall
canine population of the Behavior Clinic or of the vet-
erinary teaching hospital. Further examination of the
most common breeds of dogs (mixed breeds, German
Shepherd Dogs, Rottweilers, Dalmations, and
Bulldogs) revealed interesting patterns among the
breeds. The specific manifestations of the OCD
appeared to be associated with the tasks for which the
dogs were developed. Dogs of herding breeds often had
excessive tail chasing. Dogs of guarding breeds and
those selected for intense focus and tenacity (eg,
Dalmatians, Rottweilers, and German Shepherd Dogs)
often had signs of hallucinations. All Rottweilers and
all but 1 Dalmatian had signs of hallucinations. Nine of
12 German Shepherd Dogs chased their tails.
The cat population consisted of 14 Domestic
Shorthairs, 6 Siamese, 1 Devon Rex, 1 Russian Blue,
and 1 Bengal. Siamese cats commonly had pica involv-
ing sucking, chewing, or ingestion of fabrics; 2 of the 6
Siamese cats in the study chewed, sucked, or ingested
fabric; and 1 ingested electric cords. All cats that
ingested fabric were Siamese, but neither cats that
licked plastic substrate or had sucking behavior cate-
gory were. The Bengal cat had over-grooming and
urine marking, both anxiety-related conditions. The
small number of cats we evaluated precluded statistical
comparisons among breeds. The behaviors associated
with OCD appeared after trauma in 3 cats and after
changes in feline or human social circumstances or
relationships in 7 cats. The changes in the feline or
human relationships were also often associated with
intercat aggression or elimination abnormalities (9
cats). Grooming or self-mutilation involved 16 of the
cats, 3 of which had a previous diagnosis of hyperes-
thenia and 3 of which had a diagnosis of atopy.
Age of onset—Age of onset was known or confi-
dently estimated for 95 of the 103 dogs. Mean ± SD age
at onset was 20.3 ± 23.7 months; however, the fre-
quency distribution of age at onset was highly skewed
(skewness, 3.214). Age of onset for cats was 28.2 ±
30.1 months (n = 21), and the frequency distribution
JAVMA, Vol 221, No. 10, November 15, 2002 Scientific Reports: Retrospective Study 1447
Figure 2—Frequency (%) distribution of categories of OCD in 23
Figure 1 for key.
Figure 1—Frequency (%) distribution of categories of obsessive-
compulsive disorder (OCD) in 103 dogs. M = Self-mutilation
(grooming). L = Locomotor (spinning, chasing). H = Hallucinatory.
P = Pica. V = Vocalization. K = Licking. C = Coprophagy. S =
Sucking. D = Digging. O = Other.
was less skewed (skewness, 1.14) than that of the dog
population. Median age of onset for cats and dogs was
12 months, indicating that half of the patients devel-
oped signs of OCD by the age of 1 year.
Source—Almost 60% of the dogs in this study
originated from either serious show breeders (54.4%)
or backyard breeders (4.9%). These categories were
self-assigned by the dog owners, so the low percentage
of backyard breeders may have been an underestimate.
Only 9% of dogs came from pet stores, whereas 11%
came from humane shelters.
Six of the cats came from friends, 5 came from
either serious show breeders or backyard breeders, 5
were found or were stray, 3 came from humane shel-
ters, and 4 came from miscellaneous sources.
Size of human and pet households—Only 18
(17.5%) of the dogs in this study lived in single-human
households; 46 (44.7%) lived in households with 2
humans, and 39 (37.7%) lived in households with 3 or
more humans. More than half of the dogs in this study
had other dogs in the household, and almost 30% had
cats in the same household.
Most of the cats lived in households with 2 or
more humans (n = 18). Fifteen cats lived with other
cats; 11 lived in 2-cat households, and 4 lived in house-
holds with 3 or more cats. Few cats (n = 3) with OCD
had dogs in their household.
Training for dogs—Most dogs in this study had
what is considered to be basic training of some kind; 6
(5.9%) attended puppy kindergarten, 20 (19.6%)
attended basic group lessons, 10 (9.8%) attended
advanced group lessons, 17 (16.7%) had a private
trainer come to their house, and 3 (2.9%) were sent to
a private trainer.
Client compliance with treatment—Only 3% of
canine clients admitted to performing none of the
behavior modification. For 17% of the patients, com-
pliance data were unavailable. Sixty-three percent of
the clients either complied extensively (38%) or con-
sistently (25%). Twenty percent complied intermittent-
ly. Client compliance for treatment in cats was also
Treatment outcomes for dogs and cats were calcu-
lated (Fig 3 and 4). Percentage frequencies of dogs
with large decreases in intensity and dogs with large
decreases in frequency of behavioral problems were
significantly associated (Cochran-Mantel Haenzel test
= 59.87; df = 1; P < 0.001). Other out-
comes included unknown (14 dogs were lost to follow-
up), small increase in intensity but a large decrease in
frequency of episodes (n = 1), death unrelated to OCD
(1), euthanasia for OCD (1), and euthanasia for rea-
sons unrelated to OCD (8).
Efficacy of medication—Of the 103 dogs in this
population, 84 were treated with 1 or more drugs.
Nineteen clients declined a drug treatment option for
their dog. There were sufficient data only to compare
the relative success rates for amitriptyline and
clomipramine (Table 1). The observed success rate for
clomipramine (0.83) was significantly greater than that
(0.59) for amitriptyline (log likelihood ratio test statis-
= 6.03; P < 0.05).
Duration of treatment—Dogs for which complete
information was available (n = 80) were treated for a
mean of 14.1 months, a median of 12 months, and a range
of 6 to 78 months. Cat data were insufficient for analysis,
but the range of continuous treatment was 2 to 4 months.
All cats from which medication was withdrawn (n = 9)
relapsed and drug treatment was reinstated.
Affected relatives—Clients knew whether rela-
tives were also affected with some form of OCD for
only 30 of the 103 dogs in this study; 15 dogs had
1448 Scientific Reports: Retrospective Study JAVMA, Vol 221, No. 10, November 15, 2002
Figure 3—Outcomes for 103 dogs treated for OCD. SDI = Small
decrease in intensity of OCD (≤ 50%). LDI = Large decrease in
intensity (> 50%). SDF = Small decrease in frequency of OCD
behavior (≤ 50%). LDF = Large decrease in frequency (> 50%).
S = Stopped OCD behavior completely. NC = No change.
Figure 4—Outcomes for 20 cats treated for OCD.
affected relatives and 15 did not. Of the 15 dogs with
affected relatives, 7 had 1 known affected relative (3
male littermates, 3 sires, and 1 dam), 7 had 2 known
affected relatives (female littermate-male littermate,
sire-male offspring, sire-other, grandsire-other, dam-
male littermate, dam-second degree relative), and 1
had 3 known affected relatives. Only 3 clients with cats
knew about affected relatives, and only 2 of these defi-
nitely had an affected relative (1 sire, 1 offspring).
Concurrent behavioral diagnoses—Seventy-seven
(74.8%) dogs had concurrent behavioral conditions, as
determined via published diagnostic criteria.
three (32.0%) dogs met the criteria for a diagnosis of
attention-seeking behavior (eg, excessive solicitation
and neediness), 29 (28.2%) met the criteria for a diag-
nosis of dominance or impulse-control aggression, 23
(23.3%) met the criteria for a diagnosis of separation
anxiety, and 16 (15.5%) met the criteria for a diagnosis
of generalized anxiety disorder. Thirty other miscella-
neous behavioral and medical diagnoses were also
noted. There was no association between duration that
the dog had been affected before treatment and the
number of concurrent behavioral diagnoses (r = 0.13;
P = 0.23). Because many dogs in this study had multi-
ple concurrent behavioral diagnoses, the sample sizes of
each diagnostic combination were insufficient to deter-
mine whether the associations were random, as has
been performed elsewhere.
The data were also insuffi-
cient to assess whether dogs that had been affected
longer had more intense or more frequent signs of
OCD, compared with dogs affected for a short period.
In contrast with dogs, only 9 of 23 (39.1%) cats
met the criteria for concurrent behavioral diagnoses, a
difference that was significant (log likelihood ratio test,
= 10.04; P < 0.05). In contrast with dogs, most of
the concurrent behavioral diagnoses were associated
with elimination disorders; 3 cats also sprayed, 2 cats
urine-marked without spraying, 1 cat marked with
feces, and 1 cat had a substrate aversion to the litter.
Only 1 cat met the criteria for separation anxiety, a
major concurrent diagnosis in dogs.
Stereotypic behaviors may or may not be associat-
ed with OCD.
Differential diagnoses for other
behavioral conditions in which the nonspecific signs
associated with OCD can occur include environmental
causes, management, humane considerations (eg,
those involving understimulation, neglect, or excessive
confinement), separation anxiety, attention-seeking
behavior, generalized anxiety, and hyperactivity.
Obsessive-compulsive disorder in all species is
characterized by repetitive, ritualistic behaviors, in
excess of any required for normal function, the execu-
tion of which interferes with normal, daily activities
and functioning. Inherent in this description is a
behavior that is exaggerated in form as well as dura-
tion. The diagnostic criteria employed here have dual
advantages. First, they permit separation of nonspecif-
ic signs from diagnostic criteria. This allows the non-
specific signs to be used to evaluate changes in the con-
dition and to further describe populations afflicted
with different manifestations of the condition.
Second, they do not require definition or assessment of
underlying motivational states, which is difficult to
accomplish in a meaningful manner in other species.
It appears that dogs, as do humans, may perceive
that their behaviors are abnormal and control their
behaviors to the extent that the behavior is performed
only minimally, or not at all, in the presence of others.
Dogs who flank suck or tail chase may, after frequent
reprimands and corrections, remove themselves from
view and then commit the behavior elsewhere. Upon
approach, the behavior ceases, only to begin again
when no one is watching or when the animal removes
itself from view. Results of this study support the exis-
tence of this evasive behavior pattern. If the desire to
perform the behavior is present, despite restraint
because of punishment, training, or physical incarcer-
ation, the condition is present. The key is that if such
control is removed and the animal can commit the
behavior, it will commit the behavior. Ignoring this
crucial point will result in underdiagnosis of OCD and
underestimation of its frequency in canine and feline
The presence of this and other cognitive compo-
nents suggests that the problem is rooted at a higher
level than the behavior alone may indicate (ie, a
Doberman Pinscher may be flank sucking, but not
because anything is wrong with its flank). Such exam-
ples support the contention that obsessions are a valid
component of OCD. We evaluate obsessions in
humans by asking them about repetitive, invasive
It is inappropriate to apply a criterion (eg,
assessment that relies on a verbal response) to 1
species that has a divergent phylogeny (eg, nonverbal)
that prohibits the use of that tool or criteria.
extent to which the patients in our study focused on
their behaviors, avoided those who sought to interfere
with them, and were avoided by clinically normal or
unaffected canine and feline housemates strongly sug-
gests that a cognitive component was present, albeit
difficult to assess.
Obsessive-compulsive disorder in humans fre-
quently appears in adolescence, at the onset of social
JAVMA, Vol 221, No. 10, November 15, 2002 Scientific Reports: Retrospective Study 1449
Table 1—Drug treatment and evaluation of success versus fail-
ure in 89 dogs and 20 cats with obsessive-compulsive disorder
No. of No. of Probability
Drug Successes Failures of success
32 22 0.593
30 6 0.833
8 1 0.889
1 1 0.500
1 0 1.000
3 0 1.000
1 2 0.333
1 0 1.000
1 0 1.000
1 0 1.000
3 0 1.000
*Used in combination with clomipramine to treat signs of panic. †Used for
pruritus associated with lesion caused by grooming in cats. ‡Used for pruri-
tus associated with lesions caused by grooming in dogs. §For treatment of
putative intervertebral disk disease. IIUsed in combination with amitriptyline
(2 animals) and clomipramine (1 animal) for signs of panic.
maturity, and continues through midlife. In dogs and
cats, OCD also appears during this indistinct period of
social maturity (range for dogs, 12 to 36 months;
mean, approx 18 to 24 months; range for cats, 24 to 48
months; mean, approx 30 to 36 months)
and if left
untreated, whether by behavioral or pharmacologic
intervention, it worsens. Given the relatively early age
at which this condition develops and the probability of
profound deterioration when left untreated, young ani-
mals should be routinely screened for OCD and treat-
ed appropriately early. Dogs and cats from families
with a history of OCD should be carefully watched for
its appearance, albeit possibly in a different form than
that of their relatives.
In this study, 10 of the 23 affected cats had signs of
their particular form of OCD after some physical trau-
ma or social upheaval, and the OCD in these cats may
have occurred concomitant with intercat aggression or
elimination problems. Siamese cats were ranked as the
second most common breed in this study. Although
this does not differ substantially from their rank in the
overall hospital population (3), it is dramatically dif-
ferent from the breed rank in our Behavior Clinic pop-
ulation (22), suggesting that when a Siamese cat is
evaluated because of a behavioral problem, the behav-
ior is likely associated with OCD. Siamese cats were
most often involved in ingestion of fabric, supporting
other findings regarding increased prevalence of OCD
in Oriental-breed cats,
but there were too few mem-
bers of each breed to reach breed-related conclusions.
It is interesting that the only Bengal cat in the study
population had self-mutilation and urine marking.
These are both anxiety-related conditions and may
have some association with the relatively recent
domestication of this breed. Most cats affected with
OCD had self-mutilation or excessive grooming. No
cats were reported to have signs of hallucinating; how-
ever, hallucinations may have been associated with tail
chasing. Most owners of these cats reported that the
cats acted as if something was on or near the cat’s tail
and that the cat was either trying to chase this entity or
escape it. Accordingly, feline hallucinations may not
have been adequately identified in this study.
Unlike cats, few dogs had OCD after trauma or
social-situational distress or upheaval, and few had
concurrent behavioral diagnoses involving elimination
or social relationships with other dogs. These data sug-
gest that the behavioral characteristics, neuroanatomic
regions affected, and molecular and neurogenetic
mechanisms of OCD may differ for dogs and cats. Two
dogs had OCD after physical trauma. In both dogs, the
trauma consisted of abusive training (hanging by a
choke collar). That 2% of this population of dogs with
OCD was subject to such abuse should be of concern
to all veterinarians.
One pet-store dog had profound coprophagia, sug-
gesting that at some point coprophagia may have rep-
resented a nutritional strategy. Of the 103 dogs, few
(approx 10%) had a putative neurologic disorder,
physical condition, or potentially painful disorder
associated with OCD, which could either be primary or
secondary to OCD. One dog had a diagnosis of irrita-
ble bowel syndrome, a diagnosis that may be simply a
nonspecific sign of an anxiety-related condition. This
finding supports the hypothesis that OCD in dogs is
based in some primary neurochemical or neurogenetic
dysfunction, and that the mechanisms driving OCD
may differ between dogs and cats.
Obsessive-compulsive disorder affects at least 2%
of the human population, and this is believed to be an
Some forms of OCD have a familial
; however, most instances of
human OCD appear to be sporadic. It is important to
recognize that the development of specific animal
breeds and the practice of inbreeding within those
breeds suggest that the prevalence of OCD in dogs
could be higher than that reported for humans.
On the basis of client interviews and complaints,
OCD may be familial in Great Danes, German Short-
haired Pointers, German Shepherd Dogs, Bull
Jack Russell Terriers, Dalmatians, Bouvier de
Flanders, Salukis, Cairn Terriers, Basset Hounds, and
Soft-Coated Wheaten Terriers.
The strong correla-
tions between canine breeds and forms of OCD we
detected strongly suggest a genetic basis for OCD,
albeit, in part as the result of genetic limitations and
subsequent potential decreases in genetic heterogene-
ity associated with breed.
As is true for humans, first-degree relatives usual-
ly have a different form of OCD than the proband,
which supports the hypothesis of a heritable, neuro-
chemically variable basis for OCD. That 50% of the
dogs in this study for which familial data were known
had a relative affected with some form of OCD strong-
ly suggests 2 important points: certain breeds of dogs
appear to have a high prevalence of OCD, perhaps
higher than that in the human population, and a larg-
er proportion of canine relatives are affected than
appears true for humans. This strongly suggests a
genetic component for OCD in dogs.
Results of recent studies indicate that OCD in
humans is the result of dysfunction of genes involving
neurochemical and intracellular regulatory systems.
Similar complex regulatory systems that have a genet-
ic, heritable basis have also been reported for dogs
and may be involved in OCD.
Human OCD has been postulated to be caused by
aberrant serotonin metabolism.
ment has been directed at affecting serotonergic metabo-
lism; pharmacological agents used for treatment are fair-
ly specific and affect subclasses—primarily the 5-HT
class—of serotonin receptors. Neuropharmacologic
approaches to treatment have sought to address such reg-
ulatory abnormalities by augmenting serotonin through
the use of TCAs and SSRIs.
The key to the success of
the specific TCAs and SSRIs over other classes of med-
ication is that they use the same second messenger sys-
tems and transcription pathways that are used to develop
cellular memory (ie, learning).
While the best design for comparing drug efficacy
is a prospective, placebo-controlled, double-blind
study, other statistical comparisons can be made if cri-
teria for switching medications are consistent, as was
the case here. Such techniques are underused in veteri-
nary medicine, which is unfortunate since the funding
and large enrollment pools required for the former are
1450 Scientific Reports: Retrospective Study JAVMA, Vol 221, No. 10, November 15, 2002
seldom available. As is true for humans, dogs with
OCD respond well to the TCA, clomipramine,
and to the SSRI, fluoxetine.
Results of the study
reported here indicated that clomipramine was superi-
or for treatment of OCD in dogs, compared with
amitriptyline. The difference in efficacy is likely asso-
ciated with specificity for the serotonin 1A subtype (5-
) receptor of the parent compound and at least 1
of the intermediate metabolites that acts as a SSRI.
This specificity may also be responsible for the success
of fluoxetine in the treatment of some forms of
although our sample size was not sufficient to
test this hypothesis.
Adverse effects appear rare in canine patients; the
most common adverse effects have been gastrointesti-
Use of TCAs is contraindicated in ani-
mals with a history of urinary retention and severe,
uncontrolled cardiac arrhythmias.
It is surprising that amitriptyline was at all effica-
cious (32/54 animals) in the treatment of OCD in our
study. Although not the drug of choice for treatment of
OCD in humans, this relatively nonspecific TCA may
sufficiently decrease nonspecific anxiety so that
patients can learn to change their behavior. At present,
we have no way to evaluate how advanced OCD is
when it is diagnosed in dogs and cats. Early signs may
go unnoticed. The success associated with treatment
with amitriptyline may reflect that some animals have
less severe forms of OCD and fewer comorbid diag-
noses. No animals with long-term OCD and multiple
concurrent behavioral diagnoses improved when treat-
ed with amitriptyline alone in this study.
In our study, 74.8% of the canine patients, but only
39.1% of the feline patients, had concurrent behavioral
diagnoses. When considered in light of the relative role
apparently played by environmental factors in OCD for
these 2 species and the divergent evolutionary and
domestication histories of dogs and cats, caution is
urged in assuming that OCD is mechanistically the
same in these 2 species. This finding could be fortu-
itous and lead us to a clearer understanding of a con-
dition that is multifactorial and has both genetic and
sporadic forms. Furthermore, the high rate of concur-
rent primary anxiety disorders (attention-seeking
behavior, separation anxiety, and generalized anxiety
disorder) in dogs affected with OCD suggests that as 1
anxiety-related condition progresses, other conditions
may represent manifestations of additional underlying
neurochemical and molecular changes.
Few dogs or cats in our study had complete ces-
sation of behaviors associated with OCD, although
most clients complied with the treatment protocols;
however, none became worse as a result of treatment
and only 1 dog and 1 cat were euthanatized because
of OCD. Treatment with medication (clomipramine)
and behavior modification was extremely successful;
however, mean canine treatment time was 14.1
months, and more than half of all canine patients
were treated for > 12 months. This is important
because the label instructions for clomipramine for
treatment of separation anxiety suggest a shorter
treatment period. There is considerable variation
between individual dogs in plasma concentrations. In
dogs, clomipramine reaches steady-state concentra-
tions in 3 to 5 days, peak plasma concentrations are
attained in approximately 1 to 3 hours, the half-life of
the parent compound is 1 to 16 hours, and the half-
life of the active intermediate metabolites is 1 to 2
These data suggest that dogs may require
higher dosages or more frequent dosing than do
humans treated with such medications. Should the
medication be discontinued, relapses occur in many
Signs also worsened or became more pro-
nounced in stressful or anxiety-inducing circum-
stances for some patients in our study. Use of
clomipramine to treat OCD in animals is extralabel
usage in the United States.
Available from corresponding author upon request.
Clomicalm, Novartis Animal Health, Greensboro, NC.
SAS, SAS Institute Inc, Cary, NC.
Amitriptyline, Stuart, Wilmington, Del.
Clomipramine, Ciba-Geigy, Summit, NJ.
Fluoxetine, Eli Lilly & Co, Indianapolis, Ind.
Buspirone, Mead Johnson Pharmaceuticals, division of Bristol-
Myers Squibb Co, Princeton, NJ.
Alprazolam, Upjohn, New York, NY.
Hydroxyzine, Roering, New York, NY.
Doxepin, Roering, New York, NY.
Sertraline, Pfizer Animal Health, Groton, Conn.
Pentazocine, Winthrop, NC.
Diazepam, Roche, New York, NY.
1. Luescher UA, McKeown DB, Halip J. Stereotypic or obses-
sive-compulsive disorders in dogs and cats. Vet Clin North Am Small
Anim Pract 1991;21:401–413.
2. Overall KL. Recognition, diagnosis, and management of
obsessive-compulsive disorders. Canine Pract 1992;17(2):
3. Overall KL. Recognition, diagnosis, and management of
obsessive-compulsive disorders. Canine Pract 1991;17(3):25–27.
4. Overall KL. Recognition, diagnosis, and management of
obsessive-compulsive disorders. Canine Pract 1992;17(4):39–43.
5. Hewson CJ, Luescher UA, Ball RO. The use of chance cor-
rected agreement to diagnose canine compulsive disorder: an
approach to behavioral diagnosis in the absence of a ‘gold standard’.
Can J Vet Res 1999;63:201–206.
6. Hewson CJ, Luescher UA, Ball RO. Measuring change in the
behavioral severity of canine compulsive disorder: the construct
validity of categories of change derived from two rating scales. Appl
Anim Behav Sci 1998;60:55–68.
7. Overall KL. Use of clomipramine to treat ritualistic stereo-
typic motor behavior in three dogs. J Am Vet Med Assoc 1994;205:
8. Overall KL. Clinical behavioral medicine for small animals. St
Louis: Mosby Year Book Inc, 1997;209–250, 404, 410–423,
9. Overall KL. Self-injurious behavior (SIB) and obsessive-
compulsive disorder in domestic animals. In: Shuster L, Dodman N,
eds. Psychopharmacology of animal behavior disorders. Boston:
Blackwell Science, 1998;222–252.
10. Overall KL. Dogs as “natural” models of human psychi-
atric disorders: assessing validity and understanding mechanism.
Prog Neuropsychopharmacol Biol Psychiatry 2000;24:727–276.
11. Hewson CJ, Luescher A, Parent JM, et al. Efficacy of
clomipramine in the treatment of canine compulsive disorder.
J Am Vet Med Assoc 1998;213:1760–1766.
12. Gupta M, Gupta AK, Ellis CN. Antidepressant drugs in der-
matology. Arch Dermatol 1987;123:647–652.
13. Overall KL. Pharmacological treatment in behavioral medi-
cine: the importance of neurochemistry, molecular biology, and
mechanistic hypotheses. Vet J 2001;62:9–23.
14. Shanley KS, Overall KL. Psychogenic dermatoses. In: Kirk
JAVMA, Vol 221, No. 10, November 15, 2002 Scientific Reports: Retrospective Study 1451
RW, Bonagura JD, eds. Kirk’s current veterinary therapy XI. Small ani-
mal practice. Philadelphia: WB Saunders Co, 1992;552–558.
15. Overall KL, Dunham AE, Frank D. Frequency of nonspecif-
ic clinical signs in dogs with separation anxiety, thunderstorm pho-
bia, and noise phobia, alone or in combination. J Am Vet Med Assoc
16. Mojtabai R, Rieder RO. Limitations of the symptom-orient-
ed approach to psychiatric research. Br J Psychiatry 1998;173:
17. American Psychiatric Association. DSM-IV. Diagnostic and
statistic manual of mental disorders. 4th ed. Washington, DC: APA
18. Seksel K, Lindeman MJ. Use of clomipramine in the treat-
ment of anxiety-related and obsessive-compulsive disorders in cats.
Aust Vet J 1998;76:317–321.
19. Robins LN, Helzer JE, Weisman MM. Lifetime prevalence of
specific psychiatric disorders in three sites. Arch Gen Psychiatry
20. Karno M, Golding I, Sorenson S, et al. The epidemiology of
obsessive compulsive disorder in five U.S. communities. Arch Gen
21. Flamment M, Whittaker A, Rapoport J, et al. Obsessive com-
pulsive disorder in adolescence: an epidemiological study. J Am Acad
Child Adolesc Psychiatry 1988;27:764–771.
22. Pauls DL, Alsobrook JP, Goodman W, et al. A family study of
obsessive-compulsive disorder. Am J Psychiatry 1995;152:76–84.
23. Nestadt G, Samuels JF, Bienvenu OJ, et al. A family study of
obsessive-compulsive disorder. Arch Gen Psychiatry 2000;57:
24. Grados MA, Riddle MA, Samuels JF, et al. The familial phe-
notype of obsessive-compulsive disorder in relation to tick disorders:
the Hopkins OCD family study. Biol Psychiatry 2001;50:559–565.
25. Moon-Fanelli AA, Dodman NH. Description and develop-
ment of compulsive tail chasing in terriers and response to
clomipramine treatment. J Am Vet Med Assoc 1998;212:1252–1257.
26. Nestadt G, Lan T, Samuels JF, et al. Complex segregation
analysis provides compelling evidence for a major gene underlying
obsessive-compulsive disorder (OCD) and heterogeneity by gender.
Am J Hum Genet 2000;67:1611–1616.
27. Greer JM, Capecchi MR. Hoxb8 is required for normal
grooming behavior in mice. Neuron 2002;33:23–34.
28. Mignot E. A commentary on the neurobiology of the
hypocretin/orexin systems. Neuropsychopharmacology 2001;25:
29. deBoer T, Stoof JC, van Duijn H. The effects of convulsant
and anticonvulsant drugs on the release of radio labeled GABA, glu-
tamate, noradrenaline, serotonin, and acetylcholine from rat cortical
slices. Brain Res 1982;253:153–160.
30. Jacobs BL, Wilkinson LO, Fornal CA. The role of brain sero-
tonin: a neurophysiologic perspective. Neuropsychopharmacology
31. Murphy DL. Neuropsychiatric disorders and the multiple
human brain serotonin receptor subtypes and subsystems.
32. Altemus M, Pigott T, Kalogeras K, et al. Abnormalities in the
regulation of vasopressin and corticotropin releasing factor secretion
in obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:
33. Thoren P, Asberg M, Cronholm B. Clomipramine treatment
of obsessive-compulsive disorder. Arch Gen Psychiatry 1980;37:
34. Flament MF, Rappoport JL, Berg CJ. Clomipramine treat-
ment of childhood obsessive-compulsive disorder. A double-blind
controlled study. Arch Gen Psychiatry 1985;42:977–983.
35. Ananth J. Clomipramine: an antiobsessive drug. Can J
36. Zohar J, Insel TR, Zohar-Kadouch RC, et al. Serotonergic
responsivity in obsessive compulsive disorder; effects of chronic
clomipramine treatment. Arch Gen Psychiatry 1988;45:167–172.
37. Perse T. Obsessive-compulsive disorder: a treatment review.
J Clin Psychiatry 1988;49:48–55.
38. McTavish D, Benfield P. Clomipramine: an overview of its
pharmacological properties and a review of its therapeutic use in
obsessive-compulsive behavior and panic attack. Drugs 1990;39:
39. Blier P, deMontigny C, Chaput Y. A role for the serotonin
system in the mechanism of action of antidepressant treatments: pre-
clinical evidence. J Clin Psychiatry 1990;51(suppl):14–20.
40. Duman RS. Novel therapeutic approaches beyond the sero-
tonin receptor. Biol Psychiatry 1998;44:324–335.
41. Duman RS, Heninger GR, Nestler EJ. A molecular and cel-
lular theory of depression. Arch Gen Psychiatry 1997;54:597–606.
42. Wynchank D, Berk M. Behavioural changes in dogs with
acral lick dermatitis during a 2 month extension phase of fluoxetine
treatment. Hum Psychopharmacol Clin Exp 1998;13:435–437.
43. Wynchank D, Berk M. Fluoxetine treatment of acral lick
dermatitis in dogs: a placebo-controlled randomized double blind
trial. Depress Anxiety 1998;8:21–23.
44. Overall KL. Animal behavior case of the month. J Am Vet
Med Assoc 1995;206:629–632.
45. Dodman NH, Donnelly R, Shuster L, et al. Use of fluoxetine
to treat dominance aggression in dogs. J Am Vet Med Assoc 1996;209:
46. Overall KL. Rational use of behavioral pharmacology. Vet Clin
North Am Small Anim Pract 1997;27:637–665.
47. Murphy D, Pato M, Pigott T. Obsessive-compulsive disorder:
treatment with serotonin-selective uptake inhibitors, asapirones, and
other agents. J Clin Psychopharmacol 1990;10:91S–100S.
48. Tollefson GD, Rampey AH, Potvin JH, et al. A multicenter
investigation of fixed-dose fluoxetine in the treatment of obsessive-
compulsive disorder. Arch Gen Psychiatry 1994;51:559–567.
49. King J, Simpson B, Overall KL, et al. Treatment of separa-
tion anxiety in dogs with clomipramine. Results from a prospec-
tive, randomized, double-blinded, placebo-controlled clinical trial.
J Appl Anim Behav Sci 2000;67:255–275.
50. Reich MR, Ohad DG, Overall KL, et al. Electrocardiographic
assessment of antianxiety medication in dogs and correlations with
serum drug concentration. J Am Vet Med Assoc 2000;216:1571–1575.
51. Lueger RJ, Lutz W, Howard KI. The predicted and
observed course of psychotherapy for anxiety and mood disorders.
J Nerv Ment Dis 2000;188:127–134.
52. Hewson CJ, Conlon PD, Luescher UA, et al. The pharmaco-
kinetics of clomipramine and desmethylclomipramine in dogs: para-
meter estimates following a single oral dose and 28 consecutive daily
oral doses of clomipramine. J Vet Pharmacol Ther 1998;21:214–222.
53. King JN, Maurer MP, Altmann B, et al. Pharmacokinetics of
clomipramine in dogs following single-dose and repeated-dose oral
administration. Am J Vet Res 2000;61:80–85.
1452 Scientific Reports: Retrospective Study JAVMA, Vol 221, No. 10, November 15, 2002