Content uploaded by Nicholas Dodman
Author content
All content in this area was uploaded by Nicholas Dodman on Mar 11, 2014
Content may be subject to copyright.
460 Scientific Reports: Original Study JAVMA, Vol 219, No. 4, August 15, 2001
SMALL ANIMALS
B
y definition, separation anxiety is severe distress
when an individual is distanced from other group
members, but in canine behavioral terminology this
term is most often restricted to dogs that become upset
when separated from their owner. Separation anxiety is
1 of the most common canine behavior problems and is
diagnosed in 20 to 40% of dogs referred to animal
behavior practices in North America.
1,2
The most com-
mon complaints are destructive behavior directed at the
home, self-inflicted trauma, inappropriate elimination,
and excessive vocalization (whining, barking, or howl-
ing) only in the owner’s absence.
2-6
Problems that occur
when the owner is absent represent 1 of the principal
causes for the breakdown of the human-companion
animal bond and lead to surrender of numerous dogs to
shelters.
7,8
In a survey
9
of factors associated with canine
relinquishment to a humane society, 3 of the 10 most
commonly reported problems were consistent with sep-
aration anxiety. Within 6 months of adoption from a
shelter in southern England, 22.3% of dogs had at least
1 behavior indicative of separation anxiety.
10
There are a large number of differential diagnoses
for separation-related problems that do not have a basis
in the anxiety of separation
1-3
; therefore, the condition
can be difficult to diagnose correctly. The potential
causes of separation anxiety are numerous, may not be
exclusive, and may overlap.
11
These causes include
pathologic overattachment to the owner, negative early
experiences such as too early separation from their
dam, a traumatic experience while alone, and a change
in family circumstances. Dogs may also have a genetic
predisposition to develop the condition, because they
have been bred to be socially dependent, devoted, and
infantile.
12
Little has been published in peer reviewed journals
regarding potential risk factors or behaviors associated
with canine separation anxiety; as a result, much of the
information available on separation anxiety is found in
review articles or conference abstracts.
1,3,10,13,14
McCrave
3
reported that dogs with separation anxiety were likely
to be of mixed breed, be a stray or from a shelter, fol-
low their owners excessively, and display increased
greeting behavior, although there was no association
with spoiling activities such as allowing the dog to
sleep on the owner’s bed, feeding the dog from the
table, or taking the dog on errands. In a prospective
study
14
of 36 dogs with separation anxiety and 64 dogs
with other behavioral problems, the only significant
finding was that dogs with separation anxiety were
more likely to have increased greeting behavior, and
many followed their owner excessively. Significant dif-
ferences between these 2 groups of dogs were not
detected in sex, age at referral, age at acquisition,
source, number of adult humans in the home, a recent
move, or how long the dog was owned. In a prospec-
tive study
10
of shelter dogs performed 6 months after
adoption, an association was found between separation
anxiety and the presence of a cat in the home and
attachment to 1 person. Factors not associated with
separation anxiety included sex, neuter status, amount
of time at the shelter, number of previous homes, and
the presence of other dogs in the home.
The objectives of the study reported here were to
determine potential risk factors and behaviors associ-
ated with separation anxiety and develop a practical
index that may aid in the diagnosis of canine separa-
tion anxiety.
Materials and Methods
Case selection—A retrospective case-control study of
200 dogs with separation anxiety and 200 control dogs with
other behavioral problems was performed, using records
from the Tufts University School of Veterinary Medicine
Risk factors and behaviors associated
with separation anxiety in dogs
Gerrard Flannigan, DVM, MSc, and Nicholas H. Dodman, BVMS, DACVB, DACVA
Objectives—To determine potential risk factors and
behaviors associated with separation anxiety and
develop a practical index to help in the diagnosis of
separation anxiety in dogs.
Design—Case-control study.
Animals—200 dogs with separation anxiety and 200
control dogs with other behavior problems.
Procedures—Medical records were reviewed for sig-
nalment, history of behavior problems, home envi-
ronment, management, potentially associated behav-
iors, and concurrent problems.
Results—Dogs from a home with a single adult
human were approximately 2.5 times as likely to have
separation anxiety as dogs from multiple owner
homes, and sexually intact dogs were a third as likely
to have separation anxiety as neutered dogs. Several
factors associated with hyperattachment to the
owner were significantly associated with separation
anxiety. Spoiling activities, sex of the dog, and the
presence of other pets in the home were not associ-
ated with separation anxiety.
Conclusions and Clinical Relevance—Results do
not support the theory that early separation from the
dam leads to future development of separation anxi-
ety. Hyperattachment to the owner was significantly
associated with separation anxiety; extreme following
of the owner, departure cue anxiety, and excessive
greeting may help clinicians distinguish between
canine separation anxiety and other separation-relat-
ed problems. (
J Am Vet Med Assoc
2001;219:
460–466)
From the Department of Clinical Studies, School of Veterinary
Medicine, Tufts University, North Grafton, MA 01536. Dr.
Flannigan’s present address is Carolina Veterinary Specialists, 501
Nicholas Rd, Greensboro, NC 27409.
00_11_0494.QXD 10/11/2005 4:15 PM Page 460
JAVMA, Vol 219, No. 4, August 15, 2001 Scientific Reports: Original Study 461
SMALL ANIMALS
Behavior Clinic (from either in-clinic or fax services), to
determine potential risk factors and behaviors associated
with canine separation anxiety. Records of dogs with separa-
tion anxiety from January 1996 to December 1998 were
included in this study. Each client filled out an 8-page behav-
ioral history form prior to clinical evaluation. Diagnosis was
based on either a behavioral resident’s or a board certified
behaviorist’s judgment of the history in consultation with the
client or referring veterinarian. For a diagnosis, dogs needed
to have signs of anxiety (destruction, inappropriate elimina-
tion, or excessive vocalization) that occurred only while the
owner was absent. Only separation-related problems in
which other differential diagnoses could not be established
were included in the analysis. Signs indicative of hyperat-
tachment (excessive following, predeparture cue anxiety, and
excessive greeting behavior) were not necessary for a diagno-
sis. Control cases were collected at random from other
behavior-related cases, with the provision that owners filled
out the separation anxiety section of the history form and
met matching criteria. Because a large number of associated
factors were under study, control cases were matched with
study cases by 2 factors: the type of behavior service used and
the approximate time of year of referral or evaluation.
Variables—For all dogs, factors collected included sig-
nalment, history of the behavioral problem, structure of the
home environment, management, behaviors potentially asso-
ciated with separation anxiety, and potential concurrent
behavior problems. The dependent variable was the presence
or absence of a diagnosis of separation anxiety. Independent
variables included: age of onset (years; some owners could
not determine age of onset, because the problem was present
when the dog was acquired); sex (male vs female); reproduc-
tive status (sexually intact vs neutered); age at referral
(years); age at acquisition (years); breed (entered as individ-
ual breeds [including the closest cross when a mixed breed]
and separately as either purebred or mixed breed); weight
(kilograms and pounds); gender of the owners in the home
(male, female, or both); number of adults in the home (1
adult vs multiple adults; ≤ 2 adults vs > 2 adults [children
older than 15 years were considered adults]); number of chil-
dren in the home; number of previous owners (none vs ≥ 1);
source (breeder, family or friends, pet store, shelter, rescue
organization, or veterinary hospital; and shelter, rescue orga-
nizations, and veterinary hospitals vs all others); number of
dogs and cats in the household (each as a continuous vari-
able and as present or absent); dog allowed on the bed, fed
from the table, or taken to obedience class (yes or no); recent
changes in the home at the time of onset of the problem
(multiple or recent move, divorce, work schedule change,
less exercise than normal, new dog in the household,
increased traveling by the owner without the dog, loss of a
canine housemate, new infant, return from being with a con-
formation handler, accidental confinement, or previous
owner died); and recent changes versus no recent changes.
Several behaviors were presented to the owner on a
chart within the history form recorded as not present, mild,
moderate, or severe. These behaviors included following the
owner, increased greeting behavior > 2 to 3 minutes’ dura-
tion, anxiety at the noise of keys, anxiety when the owner put
on shoes or coat, destruction only in the absence of the
owner, elimination only in the absence of the owner, barking
or whining following the owner’s departure, decreased activ-
ity after the owner departs, anorexia following owner depar-
ture, excessive salivation in the owner’s absence, and signs of
depression after the owner departs. Some clients used the
form as a sliding scale and checked 2 sections (eg, mild and
moderate) or put a check mark on the line between the cate-
gories. When queried about this practice, clients indicated
that their dog’s behavior fell in between the categories; there-
fore, numeric values were assigned as follows: not present, 0;
mild, 1; mild to moderate, 2; moderate, 3; moderate to
severe, 4; severe, 5. If the owner was unsure or left the sec-
tion blank, the data were not included in the analysis. The
factors were also recorded as binary variables (absent to mild
vs moderate to severe).
Because multiple behaviors should be evaluated for con-
firmation of a diagnosis of separation anxiety in dogs, an
index was devised in an attempt to aid veterinarians in the
determination of this behavior problem. The index included
extreme following behavior, departure cue anxiety (at the
sound of the owner’s keys or when the owner put on shoes or
coat), and excessive greeting on a scale in which absence of
the behavior = 0, mild = 1, mild to moderate = 2, moderate =
3, moderate to severe = 4, and severe = 5.
Statistical analyses—Data were analyzed statistically by
use of computer software.
a
Factors were compared between
affected and control dogs by use of either a rank sum 2-sam-
ple (Mann-Whitney) test for continuous variables or a
Pearson χ
2
test for heterogeneity or independence for cate-
goric variables. For the Pearson χ
2
test, an odds ratio (OR)
was used to determine the strength of associations; to main-
tain an overall significance value of P < 0.05, the adjusted
alpha level for each test was based on the alpha level of the
overall study divided by the number of factors under study
(Bonferroni correction).
15
As a result, significance for each
test was P < 0.002 for the categoric variables. In some
instances, information on the history form was incomplete;
therefore, some factors were analyzed with data from < 200
cases. Because a potential cause of separation anxiety is
removal from the dam at an early age, the age at which the
dog was acquired was analyzed in the subpopulation of dogs
obtained by the present owner prior to 12 weeks of age to
determine whether there was an association between separa-
tion anxiety and early acquisition (at < 7 weeks of age vs
those acquired between 7 and 12 weeks of age).
Results
Control dogs—A wide variety of behavior prob-
lems were diagnosed in control dogs. Aggression was
the most common category of behavior problem found
in control dogs, with dominance to people (29.9% of
reported problems) and territorial aggression (13.9%)
most prominent among categories of aggression. Fears
and phobias constituted 11% of problems reported in
control dogs, whereas compulsive disorders constitut-
ed slightly > 10% of reported problems. Behaviors such
as inappropriate elimination, excessive vocalization,
and self-induced trauma reported in dogs with separa-
tion anxiety were also observed in control dogs.
However, in control dogs, these behaviors were evident
in the presence of the owner as well as when the dog
was left alone.
Although control dogs were only matched with
affected dogs by type of behavioral service and approx-
imate date of referral to the behavior clinic, the 2
groups of dogs were remarkably similar. Mean weight,
age at referral, age when acquired, number of children
in the home, and number of other pets in the home
were not significantly different between groups. In
both groups, approximately a quarter of dogs had the
behavior problem when acquired (41/159 affected vs
44/182 control dogs). Distribution between sexes was
also similar; 58% (116/200) of control dogs and 60%
(120/200) of dogs with separation anxiety were male.
Separation anxiety developed at any age in these dogs,
00_11_0494.QXD 10/11/2005 4:15 PM Page 461
462 Scientific Reports: Original Study JAVMA, Vol 219, No. 4, August 15, 2001
SMALL ANIMALS
but mean age of onset for affected dogs was signifi-
cantly (P = 0.029) greater than for other behavior prob-
lems that developed in control dogs (Fig 1).
Undesirable behaviors—The most common com-
plaint (71.7%) in dogs with separation anxiety was
destruction in the home (Table 1). Excessive vocaliza-
tion was reported in 61.2% of dogs, and inappropriate
elimination was reported in 28.1%. A small percentage
of control dogs had destructive behavior, inappropriate
elimination, and increased vocalization only in the
owners’ absence. These undesirable behaviors were
never the owners’ primary complaint at referral, nor
did the behavioral clinician believe that the signs con-
stituted a diagnosis of separation anxiety.
Factors associated with separation anxiety—
There were 2 potential risk factors that had a signif-
icant (P < 0.002) association with the incidence of
separation anxiety in this population of dogs. More
than three quarters (76.5%) of dogs with separation
anxiety lived in homes that had multiple adults (Fig
2); however, dogs referred to the behavior clinic and
kept by a single owner were approximately 2.5 times
Figure 1—Distribution of age of onset for behavior problems in dogs with separation
anxiety (solid bars; n = 118) and control dogs with other behavior problems (open
bars; 138). Approximately 25% of dogs in both groups had the behavior problem
when obtained.
Table 1—Incidence of behaviors used for diagnosis of separation anxiety in control dogs with vari-
ous behavior problems and dogs with separation anxiety
Control dogs Dogs with separation anxiety
Moderate Owner Moderate Owner
No. of to severe unsure No. of to severe unsure
Behavior dogs* (%)* (%) dogs* (%)* (%)
Dog attributes
Follow owner 200 64.0 0 199 83.8 0.5
Increased greeting 200 31.0 0 199 62.8 0.5
Signs of angst at sound
of keys 197 26.9 1.5 197 62.5 1.5
Signs of angst when
owner puts on coat
or shoes 199 32.1 0.5 197 69.5 1.5
After owner departs
Decreased activity 153 34.0 23.5 139 33.8 30.5
Signs of depression 154 16.2 23.0 139 59.0 30.5
Will not eat 170 20.6 15.0 161 46.6 19.5
Hypersalivation/vomiting/
diarrhea 199 0 0.5 197 19.3 1.5
Undesirable behavior†
(only in owner’s absence)
Destruction 198 3.0 1.0 198 71.7 1.0
Elimination 195 1.5 2.5 192 28.1 4.0
Vocalization 191 5.2 4.5 178 61.2 11.0
*Excludes data for which owners were unsure.
†Some dogs had more than 1 undesirable behavior.
00_11_0494.QXD 10/11/2005 4:15 PM Page 462
JAVMA, Vol 219, No. 4, August 15, 2001 Scientific Reports: Original Study 463
SMALL ANIMALS
as likely to have separation anxiety (P < 0.002; OR,
2.364).
Overall, 91% of dogs in this study were neutered
(86.5% of control dogs and 95.5% of affected dogs).
Prior to analysis of the data, we hoped that potential
associations of separation anxiety between spayed
females, castrated males, and their sexually intact
counterparts could be evaluated, but sample size of
sexually intact dogs was too small; therefore, the data
were pooled. Sexually intact dogs (male or female)
were > 3 times less likely to have separation anxiety (P
< 0.002; OR, 3.312) than neutered dogs.
Other potentially associated factors—For several
factors, association with separation anxiety approached
significance but did not meet the significance require-
ments of this study (P < 0.002). Dogs from shelters or
rescue groups and those adopted from veterinary hos-
pitals or found abandoned more commonly (P = 0.017;
OR, 1.663) had separation anxiety than dogs from other
sources (breeders, family or friends, and pet stores col-
lectively); 41.1% (79/192) of dogs with separation anx-
iety versus 29.6 % (58/196) of control dogs were from
shelter-type environments. Only 16% of owners could
recall a change in the home when separation anxiety
initially developed, whereas 10% of owners of control
dogs had an environmental change when their dog’s
particular problem developed (P = 0.01; OR, 2.162).
Recent changes included divorce, multiple or recent
moves, change in work schedule, the dog receiving less
exercise, increased traveling by the owner without the
dog, introduction of a new dog, loss of canine or human
housemate, a new infant in the home, and accidental
confinement. Dogs with separation anxiety were less
likely to have attended an obedience class than dogs
with other behavior problems, although the difference
between groups was small (P = 0.02; OR, 1.614).
Purebred dogs were predominant in affected and
control groups. Separation anxiety was reported in 69
mixed-breed dogs and 131 purebred dogs of 56 breeds.
In comparison, control dogs comprised 45 mixed-
breed dogs and 155 purebred dogs representing 62
breeds. German Shepherd Dogs and German Shepherd
Dog crosses, followed by Labrador Retrievers and
Labrador Retriever crosses, were the most common
breeds in both groups. In the separation anxiety group,
the subsequent order of breed incidence was Golden
Retrievers, English Springer Spaniels, and English
Cocker Spaniels. In the control group, the subsequent
order of breed incidence was English Springer
Spaniels, English Cocker Spaniels, and Doberman
Pinschers. During the time of this study, these breeds
were among the most popular breeds, as judged by use
of American Kennel Club registration data.
16
Overall,
more mixed-breed dogs had separation anxiety than
did purebred dogs, although this difference was not
significant (P = 0.008; OR, 1.814).
Factors not associated with separation anxiety—
Several factors were not associated with separation
anxiety. Sex of dog, age when acquired, presence of
other pets in the home, gender of the owner, and hav-
ing at least 1 previous owner were not associated with
separation anxiety. Spoiling activities such as allowing
the dog on the owner’s bed or feeding the dog from the
table also were not associated with separation anxiety.
Acquiring a dog from a pet shop was not significantly
associated with an increased incidence of separation
anxiety. Approximately half of the study dogs (101
affected dogs and 105 control dogs) were acquired by
the owner before 12 weeks of age. In this subpopula-
tion, there was no statistical association between
acquiring the puppy at a young age (< 7 weeks of age)
and the future development of separation anxiety.
Noise phobias—Many dogs in this study had
some fear of noises, but noise phobia was significantly
more common in dogs with separation anxiety; almost
Figure 2—No. of adults in the homes of dogs with separation anxiety (solid bars) and
control dogs with other behavioral problems (open bars; n = 200/group).
00_11_0494.QXD 10/11/2005 4:15 PM Page 463
464 Scientific Reports: Original Study JAVMA, Vol 219, No. 4, August 15, 2001
SMALL ANIMALS
half (97/200 [48.5%]) of dogs with separation anxiety
were fearful of noises, whereas less than a third
(63/200 [31.5%]) of control dogs had a similar fear.
Other behaviors associated with separation anxi-
ety—Dogs with separation anxiety had other behaviors
more commonly than control dogs did (P < 0.001;
Table 1). Dogs with separation anxiety were 3 times
more likely to follow their owner excessively, com-
pared with control dogs (OR, 2.936), and almost 4
times more likely to have excited greeting behavior of
duration > 2 to 3 minutes (OR, 3.76). Dogs with sepa-
ration anxiety were 4.5 to 5 times more likely to be
anxious at the time of departure. After the owner
departed, affected dogs were > 7 times more likely to
have signs of depression (OR, 7.432) and > 3 times
more likely to not eat while their owner was absent
(OR, 3.364).
Diagnostic index—Following analysis of diagnos-
tic index data, cut-off value for a diagnosis was deter-
mined to be 10, using the described scale. On the basis
of this index, a dog that the owner believed had severe
problems in all 3 behavioral categories would have an
index of 15. Approximately three quarters (75.2%) of
dogs with separation anxiety had an index of 10 or
greater, whereas only 28.9% of control dogs had an
index of 10 or greater. Dogs that met this criterion were
7.6 times more likely to have separation anxiety than
control dogs were (P < 0.001; OR, 7.653).
Although highly associated with separation anxi-
ety, behaviors that occur after the owner leaves such as
signs of depression and anorexia were not included in
the index, because approximately a quarter of owners
in both groups were unable to confirm these behaviors,
because they may have occurred when other family
members were also absent. Repeated hypersalivation,
vomiting, and diarrhea when the owner left the home
were not a commonly expressed behavior in dogs with
separation anxiety and were found in only 19.3% of
these dogs. However, when present, these behaviors
were an excellent supportive sign of separation anxiety,
because not a single control dog had these behaviors.
Discussion
Retrospective case-controlled studies may be
extremely useful for uncovering potential causal fac-
tors and behaviors associated with primary behavioral
problems if bias is not introduced in the history of
exposure to the factors of interest or in the selection of
cases and controls.
17
Because we depended on the
owner to fill out the history form, we believed that the
information given was accurate. The proportion of var-
ious behavior diagnoses among the control dogs was
similar to published data for other animal behavior
clinics when separation anxiety was excluded. A strong
association between 2 factors lends support to but does
not prove causation
17
; there could be a third factor
impacting both components, uncertainty regarding
which factor was present first, or a factor that was pre-
sent in an attempt to treat the behavior problem.
Principal complaints at referral for dogs with sep-
aration anxiety in our study were destruction, exces-
sive vocalization, and inappropriate elimination, in
that order. In a prospective study
10
of dogs adopted
from a shelter, a slightly greater proportion of dogs
were destructive, compared with results of our study
(84 vs 71.4%, respectively), and a smaller proportion
had excessive vocalization (41 vs 61.2%) and inappro-
priate elimination (18 vs 28.1%), although differences
may be a consequence of the source of the dogs.
Because our study had a large population of dogs
acquired prior to 12 weeks of age (approx 50%), per-
haps this had an effect on the type of complaint report-
ed at referral. Podberscek et al
18
reported the problems
at referral of 49 dogs in a study of the effectiveness of
clomipramine as an adjunct to behavior modification;
the proportions of dogs with destruction (73.5%),
excessive vocalization (69.4%), and inappropriate
elimination (42.8%) were quite similar to those in our
study. In contrast, in a study
14
of 36 dogs with separa-
tion anxiety, vocalization was the most common
behavior (90%), followed by destruction (80%) and
inappropriate elimination (55%).
In the study reported here, sex distribution of dogs
with separation anxiety was similar to that reported in
other studies
1,6,16,18-20
and that of control dogs with other
behavior problems. Many studies investigating sex dis-
tribution among dogs with behavior problems (separa-
tion anxiety or others) have found that approximately
60% of the dogs were male. The common conclusion
by other authors is that male dogs are more prone to
behavior problems. However, an alternate explanation
is that males are more severely affected than females,
prompting owners to seek treatment from a behavior
specialist in higher numbers.
To the authors’ knowledge, this is the first study to
reveal a significant association between separation anx-
iety and a dog being kept by a single adult owner.
Similar to other studies, affected dogs typically had > 1
owner in the home. Other studies
1,10,18
that found no
association between the number of adults in the home
and separation anxiety had < 50 affected dogs. Because
a typical home has > 1 adult, our large sample size (n
= 200) may have been essential to uncover this associ-
ation.
For the period of this study, neutering pet dogs
was a common practice in the canine population
referred to the Tufts University Behavior Clinic. Only 1
North American study
20
has examined this factor with
respect to separation anxiety; no association was
found, but this may be another factor that requires a
large sample of affected dogs to detect an association,
because of canine population dynamics. Separation
anxiety represents a lower proportion of behavior
referrals to European behaviorists, which corresponds
to a higher proportion of sexually intact dogs.
11,19
Other investigators
3,a,b
have detected a much
stronger relationship between source of the dog and
canine separation anxiety than was detected in our
study, demonstrating as much as a 3-fold greater inci-
dence of separation anxiety in strays or dogs from shel-
ters. Although separation anxiety is often reported as a
sequela to a sudden change in the household,
1,5,11,21
a
direct association has been considered in only 1
report
14
and was not found to be significant. We found
that few owners could remember a change in the home
00_11_0494.QXD 10/11/2005 4:15 PM Page 464
JAVMA, Vol 219, No. 4, August 15, 2001 Scientific Reports: Original Study 465
SMALL ANIMALS
that coincided with initiation of separation anxiety, but
it was more common in affected dogs than in control
dogs.
Voith et al
22
and Takeuchi et al
20
found that behav-
ior problems such as separation anxiety were unrelated
to obedience training. However, Clark and Boyer
23
found that obedience training improved the relation-
ship between owner and dog and detected less separa-
tion anxiety in obedience-trained dogs, compared with
dogs with no obedience instruction. In our study, obe-
dience training may have been less common in dogs
with separation anxiety, compared with control dogs,
because of the type of behavior problems in the control
dogs, rather than as a protective effect induced by obe-
dience training. Because aggression was the most com-
mon behavioral problem in the control group, obedi-
ence training may have been used by owners more
commonly in these dogs in an attempt to curb the con-
dition.
Whether separation anxiety is more common in
mixed-breed dogs, compared with purebred dogs, dif-
fers among studies. In 3 studies
3,14,20
comparing dogs
with separation anxiety and dogs with other behavior
problems, mixed-breed dogs were more likely to have
separation anxiety, but in a further study,
13
separation
anxiety was more common in purebred dogs. In anoth-
er study,
b
distribution between purebred and mixed-
breed dogs did not differ between dogs with separation
anxiety and those referred for medical or surgical prob-
lems. Unfortunately, results of our study did not
resolve the debate.
In agreement with our study, other authors have
not found an association between canine separation
anxiety and spoiling activities,
3
sex of the dog,
10,14
age at
referral,
14
and presence of other dogs in the home.
10,14
In
contradiction to our findings, McCrave
3
did not find an
association between separation anxiety and thunder-
storm phobia. However, Overall
24
found that 40% of
dogs with a noise phobia also had separation anxiety
and that 8% of dogs with separation anxiety had thun-
derstorm phobia. Results of our study indicated that a
fear of noises may be more common in dogs than pre-
viously observed and that almost half of dogs with sep-
aration anxiety have this fear.
Although there is no evidence to support the
claim, a commonly cited potential cause of canine sep-
aration anxiety is early separation from the dam.
11,12
Several authors have advised that puppies should not
be separated from the dam and littermates until 7 to 8
weeks of age,
4,5
because development of social attach-
ment in the puppy is dependent on early experience
with littermates between the ages of 5 and 7 weeks.
25
Compared with dogs that are acquired from other
sources, we found that separation anxiety was not
more common in dogs that are acquired from pet
stores, although such dogs may have been separated
from their dam and littermates at an early age.
26
In
addition, puppies that were acquired by the owner at 6
weeks of age or younger (no matter what the source)
were no more likely to develop separation anxiety than
were dogs acquired between 6 and 12 weeks of age.
The results of our study do not support the theory that
early separation from the dam and littermates leads to
the development of separation anxiety. Perhaps pup-
pies acquired at a young age are equally susceptible to
behavior problems in general.
Hyperattachment to the owner has been cited
most commonly and is believed to be a necessary com-
ponent of the disorder.
3,27,c
However, at least 1 author
4
has questioned it as a necessary and sufficient condi-
tion for a diagnosis, because many dogs with the dis-
order lack elements of extreme attachment. It is also
argued that owner behaviors considered as spoiling
activities such as feeding the dog from the table or
allowing the dog to sleep on the owner’s bed are not
more common in affected dogs,
22
possibly because
these behaviors need not lead to overattachment of the
dog to the owner. Evidence seen within our study sup-
ports the contention that hyperattachment is a compo-
nent of separation anxiety. In addition, separation anx-
iety was significantly associated with dogs dwelling in
homes in which there was only 1 adult. Therefore, we
believe that use of the diagnostic index aids in the diag-
nosis of separation anxiety.
The index was designed as a diagnostic tool.
Because some dogs without hyperattachment to the
owner may have separation anxiety, the index should
not be used to the exclusion of good clinical judgment.
The index should be applied when a dog has behaviors
suggestive of separation anxiety such as moderate to
severe destruction, inappropriate elimination, or
excessive vocalization taking place only in the owners’
absence. For dogs with behaviors that result in a diag-
nostic index value less than the cut-off value of 10, we
suggest the clinician consider another cause for the
problem behavior.
a
Analytical Software, version 4.1, Statistix, Tallahassee, Fla.
b
Voith VL, Ganster D. Separation anxiety: review of 42 cases (abstr).
Appl Anim Behav Sci 1993;37:84–85.
c
Dehasse J. The systematic approach to companion animal problem
behaviour (abstr), in Proceedings. 1st Int Conf Vet Behav Med
1997;223.
References
1. Voith VL, Borchelt PL. Separation anxiety in dogs. In: Voith
VL, Borchelt PL, eds. Readings in companion animal behavior.
Trenton, NJ: Learning Systems, 1996;124–139.
2. Simpson BS. Canine separation anxiety. Compend Contin
Educ Pract Vet 2000;22:328–337.
3. McCrave EA. Diagnostic criteria for separation anxiety. Ve t
Clin North Am Small Anim Pract 1991;21:247–255.
4. Overall KL. Clinical behavioral medicine for small animals. St
Louis: Mosby Year Book Inc, 1997;5–8, 209–250, 323–343, 393–400,
457–458.
5. Landsberg GM, Hunthausen W, Akerman L. Handbook of
behavior problems of the dog and cat. Oxford, UK: Butterworth-
Heinemann, 1997;13,27–45, 85, 102–104.
6. King JN, Simpson KL, Overall KL, et al. Treatment of sepa-
ration anxiety in dogs with clomipramine: results from a prospective,
randomized, double-blind, placebo-controlled, parallel-group,
multi-center clinical trial. Appl Anim Behav Sci 2000;67:255–275.
7. van der Borg JAM, Netto WJ, Planta DJU. Behavioural test-
ing of dogs in animal shelters to predict problem behaviour. Appl
Anim Behav Sci 1991;32:237–251.
8. Houpt KA, Honig SU, Reisner IR. Breaking the human-com-
panion animal bond. J Am Vet Med Assoc 1996;208:1653–1659.
9. Miller DD, Staats SR, Partlo C, et al. Factors associated with
the decision to surrender a pet to an animal shelter. J Am Vet Med
Assoc 1996;209:738–742.
00_11_0494.QXD 10/11/2005 4:15 PM Page 465
466 Scientific Reports: Original Study JAVMA, Vol 219, No. 4, August 15, 2001
SMALL ANIMALS
10. McBride EA, Bradshaw JW, Christians A, et al. Factors pre-
disposing dogs to separation problems, in Proceedings. 29th Int
Congr Int Soc Appl Ethol 1995;103–105.
11. Askew HR. Treatment of behavior problems in dogs and cats.
Oxford, UK: Blackwell Science, 1996;49–64, 96, 216–227.
12. Serpell J, Jagoe JA. Early experience and the development of
behaviour. In: Serpell J, ed. The domestic dog: its evolution, behaviour,
and interactions with people. Cambridge, UK: Cambridge University
Press, 1995;79–102.
13. Voith VL, Goodloe L, Chapman B, et al. Comparison of
dogs for behavior problems by source of dog, in Proceedings. Am Vet
Med Assoc Annu Meet 1993;54.
14. Voith VL. Profiles of dogs with separation anxiety and treat-
ment approaches, in Proceedings. Annu Meet Am Vet Med Assoc
1994;64.
15. Lehner PN. Handbook of ethological methods. 2nd ed.
Cambridge, UK: Cambridge University Press, 1996;353.
16. American Kennel Club. The top 50 registered AKC breeds.
Available at: http://www.akc.org/breeds/top50.cfm. Accessed Aug 1,
2000.
17. Martin SW, Meek AH, Willeberg P. Veterinary epidemiology:
principles and methods. Ames, Iowa: Iowa State University Press,
1987;159–167.
18. Podberscek AL, Hsu Y, Serpell JA. Evaluation of
clomipramine as an adjunct to behavioural therapy in the treatment
of separation-related problems in dogs. Vet Rec 1999;145:365–369.
19. Association of Pet Behaviour Counsellors. Data from annu-
al review of cases (1994–1999). Available at: http://
apbc.org.uk/data.htm. Accessed May 14, 2000.
20. Takeuchi Y, Ogata N, Houpt KH, et al. Differences in back-
ground and outcome of three behavior problems in dogs. Appl Anim
Behav Sci 2001;70:297–308.
21. Beaver BV. Canine behavior: a guide for veterinarians.
Philadelphia: WB Saunders Co, 1999;184–187.
22. Voith VL, Wright JC, Danneman PJ. Is there a relationship
between canine behavior problems and spoiling activities, anthropo-
morphism, and obedience training? Appl Anim Behav Sci
1992;34:263–272.
23. Clark GI, Boyer WN. The effects of dog obedience training
and behavioural counselling upon the human-canine relationship.
Appl Anim Behav Sci 1993;37:147–159.
24. Overall KL. Neuroanatomy and neurochemistry of fears,
anxiety and aggression, in Proceedings. Annu Meet Am Vet Med
Assoc 1998;548–551.
25. Scott JP, Stewart JM, DeGhett VJ. Separation in infant dogs:
emotional response and motivational consequence. In: Scott JP,
Senay E, eds. Separation and depression: clinical and research aspects
(AAAS No. 94). Washington, DC: American Association for the
Advancement of Science, 1973:3–32.
26. Mugford RA. Canine behavioural therapy. In: Serpell J, ed.
The domestic dog: its evolution, behaviour, and interactions with people.
Cambridge, UK: Cambridge University Press, 1995;139–152.
27. Appleby D, Heath S. Behaviour therapy: a need for critical
evaluation, in Proceedings. 1st Int Conf Vet Behav Med 1997;10–13.
00_11_0494.QXD 10/11/2005 4:15 PM Page 466
