Separation Anxiety in Adulthood: A Phenomenological
Investigation
V. Manicavasagar, D. Silove, and J. Curtis
Separation anxiety disorder is well recognized as a
juvenile psychiatric disorder, b u t it appears t o be
rarely diagnosed in adulthood. Drawing on our clinical
impressions and a review of the relevant literature, w e
sought to investigate whether separation anxiety
symptoms could be identified in adulthood. Forty-four
subjects recruited b y a media campaign were adminis-
tered a semistructured interview and a self-report
checklist for adult separation anxiety {ASA) symp-
toms, as well as the Separation Anxiety Symptom
Inventory (SASI), a retrospective measure of early
separation anxiety symptoms. Diagnoses of major
depressive disorder (MDD), panic disorder (PD), agora-
phobia (Ag), and dependent personality disorder were
made using the SCID-P and SCID-II. Thirty-six subjects
met criteria for a putative diagnosis of ASA based on a
EPARATION ANXIETY DISORDER is now
S well established as a juvenile-onset anxiety
disorder, with an estimated prevalence rate of 4%
among children and adolescents.l For the diagnosis
to be made, DSM-IV and ICD-10 both require that
the onset of symptoms occurs before 18 years of
age. DSM-IV further states that separation anxiety
disorder in adulthood is rare and that the diagnosis
should not be made if the symptoms are better
accounted for by panic disorder (PD) or agorapho-
bia (Ag). However, no criteria are specified for a
diagnosis of separation anxiety disorder in adult-
hood, the implication being that the juvenile crite-
ria are adequate. This report explores the possibility
that separation anxiety symptoms may be identified
in adulthood and that they may aggregate to form a
distinct syndrome with symptoms that vary from
the juvenile form according to the developmental
changes accompanying maturation.
There are several lines of evidence that suggest
such a possibility. Longitudinal studies of children
with school phobia have shown high rates of
ongoing psychosocial disability in later life. 2,3
However, limitations in sampling and assessment
From the Psychiatry Research and Teaching Unit, School of
Psychiatry, University of New South Wales at Liverpool Hospi-
tal, Liverpool, New South Wales, Australia.
Address reprint requests to V. Manicavasagar, M. Psychol.,
Psychiatry Research and Teaching Unit, Level 4, Health Ser-
vices Building, Liverpool Hospital, Liverpool NSW, 2170,
Australia.
Copyright © 1997 by WB. Saunders Company
0010-440X/97/3805-0009503.00/0
274 ComprehensivePsychiatry,Vol. 38, No, 5 (September/October), 1997: pp 274-282
global clinical rating and/or endorsement of DSM-IV-
derived criteria. Although most subjects dated the
separation anxiety symptoms to their juvenile years, it
was notable that one third reported the first onset of
separation anxiety symptoms in adulthood. Although
comorbid lifetime anxiety or depressive disorders
were common, the majority of subjects reported t h a t
the separation anxiety symptoms predated other axis
I disorders. Only six subjects (17%) were diagnosed
w i t h dependent personality disorder. Although lim-
ited by the method of sampling, this preliminary study
suggests the need to examine more systematically
whether a form of separation anxiety disorder may
occur in adulthood.
Copyright © 1997by W.B. Saunders Company
procedures have prevented clarification of the pre-
cise diagnostic outcomes in adulthood.4-6Retrospec-
tive studies, on the other hand, have suggested that
juvenile separation anxiety disorder (JSAD) is
linked to the risk of PD, Ag, or both (PD-Ag) in
adulthood, 7-1° but the specificity of that link re-
mains in doubt. 11 For example, a recent study by
Lipsitz et al. ~2 suggested that early separation
anxiety may constitute a nonspecific vulnerability
to a wide range of anxiety disorders in adulthood in
addition to PD. In contrast, a recent community
study supports the hypothesis that heightened lev-
els of early separation anxiety are linked specifi-
cally to adult PD-Ag. 1° Uncertainty about the
long-term outcome of JSAD is increased by meth-
odological limitations of several previous studies,
which include problems such as small sample size,
variability in the retrospective measurement of
JSAD, and use of inappropriate control groups. 13
At the same time, studies arising from attach-
ment theory increasingly have emphasized the
continuities between attachment insecurities in early
and later life, 14,15with terms like "rejection sensitiv-
ity" 16 or "interpersonal sensitivity" 17 used to de-
note forms of attachment anxiety that persist into
adulthood. In his early study, Bowlby TM suggested
that Ag was an expression of separation anxiety
originating in early life that re-emerged under
conditions of interpersonal stress in adulthood.
Such a developmental formulation is consistent
with the observations of several researchers that
some children with severe separation anxiety may
have a constitutional vulnerability that persists
ADULT SEPARATION ANXIETY
t h r o u g h o u t life. 239-21 O n e p o s s i b i l i t y t h e r e f o r e is
that s y m p t o m s o f J S A D m a y p r o g r e s s i n t o a n a d u l t
e q u i v a l e n t o f the disorder, 13 b u t the s y m p t o m s are
o v e r l o o k e d in the clinical setting e i t h e r b e c a u s e
contemporary diagnostic conventions discourage
m a k i n g s u c h a d i a g n o s i s or b e c a u s e s u p e r v e n i n g
s y m p t o m s o f P D or A g o b s c u r e the u n d e r l y i n g
disorder. A d d i n g to the c o m p l e x i t y is the n e e d to
distinguish between symptoms of separation anxi-
ety a n d d e p e n d e n c y as a p e r s o n a l i t y trait. B o w l b y 22
p o i n t e d o u t that d e p e n d e n c y is a p e r v a s i v e a n d
i n d i s c r i m i n a t e t e n d e n c y to rely e x c e s s i v e l y o n
others, w h e r e a s s e p a r a t i o n a n x i e t y refers to a
l i m i t e d array o f c o n c e r n s a b o u t the p r o x i m i t y a n d
safety o f k e y a t t a c h m e n t figures.
T h e p r e s e n t s t u d y r e p r e s e n t s a p r e l i m i n a r y at-
t e m p t to i d e n t i f y s e p a r a t i o n a n x i e t y s y m p t o m s in
a d u l t h o o d . W e a i m e d to d e s c r i b e the p h e n o m e n o l -
ogy, onset, a n d c o u r s e o f s u c h a n x i e t i e s a n d to
s t u d y the p a t t e r n o f c o m o r b i d i t y w i t h o t h e r affec-
tive d i s o r d e r s a n d d e p e n d e n t p e r s o n a l i t y disorder.
B y r e c r u i t i n g s u b j e c t s f r o m the c o m m u n i t y r a t h e r
t h a n f r o m e s t a b l i s h e d a n x i e t y clinics, w e s o u g h t to
r e d u c e the p o s s i b l e c o n f o u n d i n g i n f l u e n c e o f a
p r i o r d i a g n o s i s o n the s u b j e c t s ' a c c o u n t s o f separa-
tion a n x i e t y s y m p t o m s .
METHOD
Subjects
A limited media campaign was undertaken to recruit adults
whose major concerns were anxieties about separation from key
attachment figures. Advertisements were placed in newspapers,
and these led to a series of radio interviews. A brief account of
the typical fears associated with separation anxiety was given
without emphasizing details. The newspaper advertisements
stated that the research team was interested in interviewing
adults who were troubled by excessive anxieties or fears about
being separated from persons close to them.
All respondents were initially screened by telephone. For
inclusion, subjects had to be over 18 years of age and not
suffering from an obvious psychotic or organic disorder. Sub-
jects were then mailed a set of self-report questionnaires
including a consent form approved by the University of New
South Wales Ethics Committee. The consent form emphasized
the voluntary nature of the study and the right of subjects to
withdraw consent at any point. Consenting subjects were asked
to nominate a convenient time and location to be interviewed.
Interviews took place, on average, 2 to 3 weeks after the
questionnaires were returned.
Measures
Symptoms of separation anxiety in adulthood were assessed
by three different methods: (1) a semistructured interview, (2) a
global clinical rating, and (3) a self-report symptom checklist.
An adult separation anxiety (ASA) semistructured interview
275
(ASA-SI) consisting of 27 questions was devised to assess file
extent of the subjects' anxieties about attachments to persons
identified as close to them, with responses focusing on the
preceding 3 months. Items were derived from several sources,
including DSM-IV criteria for JSAD with symptoms modified
to apply more appropriately to adults; a review of the relevant
literature on attachment theory and research; our own additional
clinical impressions, for example, that adults with separation
anxiety "cling" by excessive talking; and the results of a
qualitative study on clinic patients with suspected ASA. In the
preliminary investigation, open-ended interviews were con-
ducted with nine patients referred to us by psychologists and
psychiatrists who were informed in broad terms of the type of
patients we sought to study. Interviews lasted approximately 1.5
to 2 hours and were audiotaped. Content analysis 23 of the tapes
was undertaken to identify the range of symptoms that appeared
to relate to ASA. The qualitative study was terminated when the
content yielded by successive interviews became repetitive,
suggesting that the themes and symptoms documented were
exhaustive. 24
Based on these sources, items were generated in an interview
format. Items were assigned a score of 3 (threshold) when
responses were judged to be positive. Responses of 1 (absent) or
2 (subthreshold) were regarded as clinically unimportant.
DSM-IV criteria for separation anxiety disorder (modified for
adulthood) were reflected in 13 items, with some redundancy in
such items. Following the interview, a global assessment was
made (present or absent) as to whether the respondent suffered
from a clinically significant problem of separation anxiety. This
global rating was made before systematic analysis of the
ASA-SI items. A 27-item self-report checklist (ASA-CL) for
assessing separation anxiety symptoms in adulthood was de-
vised with items identical to those of the ASA-SI. Items on the
ASA-CL were rated on a four-point scale ranging from 0 ("this
has never happened") to 3 ("this happens very often").
Subjects were also asked to complete a retrospective question-
naire to assess the frequency of separation anxiety symptoms
occurring before 18 years of age (the SASI). 25The psychometric
properties of the measure, including its factorial structure,
test-retest and internal reliability, and concurrent validity, have
been reported previously. 25
Subjects were screened for major depressive disorder (MDD),
PD, Ag, and dependent personality disorder using the SCID-P 26
and SCID-I1.27 (Note that these diagnoses were based on
DSM-III-R, since the SCID for DSM-IV was not available at the
time of study.) A brief DSM-IV~zlerived checklist was used to
assess past history of JSAD. A short questionnaire was included
to obtain information about losses or distressing separations
from caregivers in early life. In addition, a patient-rated scale (0,
no impact; 10, maximum disruption) was devised to determine
the extent to which each identified axis I disorder affected the
subject's life-style.
All interviews were conducted by the first author. Most of the
interviews were conducted (and audiotaped) in the subject's
home. In a minority of subjects who lived in another state or
beyond the metropolitan region of Sydney, interviews were
conducted over the telephone.
RESULTS
F i f t y - e i g h t s u b j e c t s r e s p o n d e d to o u r m e d i a
campaign. Fourteen subjects either did not return
276
the questionnaires or were unable to be contacted
for an interview, yielding a final sample of 44.
Reliability Assessment
An interrater reliability of 100% for a diagnosis
of adult separation anxiety disorder (ASAD) by
global clinical assessment was achieved with a
trained psychologist who independently rated 21
randomly chosen audiotaped interviews (13 ASADs
and eight non-ASADs). Another random selection
of 11 tapes (seven ASADs and four non-ASADs)
were rated blindly by a psychiatrist (J.C.), and
interrater agreement was again 100%. Interrater
reliability assessed for 21 audiotapes by the two
independent raters was also 100% for dependent
personality disorder according to the SCID-II.
Comparison of DSM-IV diagnoses on the ASA-SI
(endorsement of -----three of eight criteria) to the
interviewer's global clinical assessments of ASAD
showed a high degree of concordance across the
two diagnostic procedures (K = .74) (Table 1).
The two patients assigned a diagnosis of ASAD
by global clinical assessment but who did not fulfill
DSM-IV criteria endorsed two rather than three of
the necessary descriptors. We thus assumed that the
minor discrepancies in diagnosis across the two
methods reflected "borderline" cases, so all sub-
jects were included who met criteria either accord-
ing to DSM-IV or on the global clinical assessment.
The remainder of this report will focus on this
group. The sample consisted of 36 subjects (10 men
and 26 women) aged 43.0 4- 11.3 years
(mean 4-_ SD). Seventy-five percent of the sample
were married, with most subjects residing with
their spouses only (n = 12) or with spouses and
children (n = 15). One subject (aged 37 years)
resided with her parents. Fifty-six percent of the
sample held university qualifications, and three
(8%) were unemployed.
Phenomenology of ASA
Table 2 reports endorsement rates of ASA-SI and
ASA-CL items, with the pattern of responses
Table 1. Global Rating and DSM-IV Criteria for Assigning
Subjects a Provisional Diagnosis of ASAD
DSM-IV DSM-IV
Positive Negative
(n = 34) (n = 10)
Global rating positive (n = 34) 32 2 toms began in childhood and reappeared in adult-
Global rating negative (n = 10) 2 8
MANICAVASAGAR, SILOVE, A N D CURTIS
further depicted in Fig 1. Items on the ASA-SI were
regarded as positive if assigned a rating of 3
(threshold) by the interviewer. Items on the ASA-CL
were regarded as positive if assigned a score of 2
(fairly often) or 3 (very often) by the subject. The
pattern of responses suggests consistency in the
most and least frequently endorsed items for both
measures of separation anxiety symptoms (Fig 1).
The majority of ASA-SI items were endorsed by at
least one third of the subjects. Of the 14 most
frequently endorsed items, yielding affirmative
responses in at least 50% of the subjects on the
ASA-SI, five were DSM-IV criteria. Where discrep-
ancies occurred between corresponding items of
the two measures, these appeared to relate to
cognitive aspects of anxiety, with respondents
endorsing lower levels of distress on such items on
the ASA-CL (although it should be noted that the
corresponding anchor points differed across the
two measures).
Two items received low endorsement on both the
ASA-SI and ASA-CL (item 5, sleeps better if the
lights are on in the house or bedroom; and item 21,
suffers from nightmares or dreams about separation
from home), one of which (item 21) was a DSM-IV-
derived criterion.
An analysis of the internal reliability of ASA-CL
items revealed a Cronbach alpha of .89, suggesting
that the items together measured a coherent con-
struct of separation anxiety. This finding was
supported when scores for individual checklist
items were correlated with the total ASA-CL score.
Most associations were statistically significant,
with half of the items yielding correlations greater
than .45 (Table 2) and only four items showing low
correlations: item 9 (close attachments have men-
tioned that he/she talks a lot), 12 (experiences
extreme stress when leaving to go on a long trip), 7
(tries to avoid being home alone when close
attachments are out), and 5 (sleeps better if the
lights are on in the house or bedroom). Items 7 and
12 are DSM-IV-derived criteria for separation
anxiety disorder.
Onset, Course, and Impact of ASA
Onset and course. Subjects were asked to
specify whether the onset of separation anxiety
symptoms occurred before or after 18 years of age.
Two thirds (n = 24, 67%) reported that the symp-
hood at times of stress or threatened loss. Twelve
ADULT SEPARATION ANXIETY
35
N
U
M 30
B
E 25
R
0 20
F
S 15
U
B 10
J
E
C 5
T
Fig 1. Frequency of item en-
dorsement of current ASA symp- S 0
toms. ([]} Interview; (+} check- 1 2 3 4 5
list.
subjects (33%) reported that symptoms began de
novo in adulthood, in all instances following a
major traumatic event or loss (death or divorce),
and five of these described the symptoms as
episodic rather than continuous following that
event. These self-reports corresponded with the
interviewer's retrospective diagnosis of juvenile-
onset separation anxiety disorder, with 22 subjects
(61%) being assigned that diagnosis (Table 3).
SASI responses showed significantly higher
scores for subjects diagnosed retrospectively by the
interviewer as having suffered from JSAD versus
those without such a diagnosis (4.4 v 3.3; t = 2.9,
df = 33, P < .01). The mean SASI score (analyzed
using a square root transformation) for those diag-
nosed with JSAD was high compared with norma-
tive community data 25 (mean SASI, 2.8 _+ 1.2) and
similar to that of a sample of 74 adults with a
history of school phobia (mean __+SD, 4.2 -+ 1.0)
reported in a prior study. 2s
Impact of ASA symptoms. Over three fourths
(n = 28, 77%) of subjects assigned a provisional
diagnosis of ASAD reported that separation anxiety
symptoms had substantially affected their life-style
in adulthood. Twenty-five subjects (69%) reported
that such symptoms had affected their relationship
with their partner. Descriptions provided by indi-
vidual subjects indicated that several had refused
job promotions or had deliberately reorganized
277
6 7 8 9 I0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Item number
their work schedule to accommodate anxieties
about possible separations from close attachment
figures. Twenty-four subjects (67%) had sought
treatment for anxiety and/or depression, and eight
(22%) reported that the term "separation anxiety"
had been used at some point in sessions with their
therapist. However, none had been informed by
their therapist that the primary diagnosis was
ASAD or that the treatment offered specifically
focused on that problem.
Comorbidity and Distress Ratings
Thirty-two subjects had a comorbid lifetime
diagnosis of MDD, PD, or Ag. Twenty-four (67%)
had suffered from or were currently suffering from
PD (n = 3), Ag (n = 2), or PD-Ag (n = 19), and 25
(69%) had a history of MDD. Four subjects ap-
peared to have ASAD alone and were not assigned
a diagnosis of any comorbid anxiety or depressive
disorder (Table 4). Notably, 12 (33%) did not report
a history of PD or Ag.
Subjects were asked to order the sequence of
onset of comorbid disorders. Twenty-four (75%) of
32 with comorbid diagnoses stated that ASAD had
preceded the onset of other disorders. Four subjects
reported that a depressive illness had preceded the
separation anxiety symptoms, and two reported that
one of the designated anxiety disorders preceded
the separation anxiety symptoms. Two subjects
278 MANICAVASAGAR, SILOVE, AND CURTIS

Table 2. Endorsement Rates of Separation Anxiety Items According to the ASA-SI and ASA-CL
Item
ASA-SI ASA-CL
Corre(ation
(n = 36) (n = 35)
With Total
Items in Order of Endorsement No. % No. % ASA-CLScore

Item 25: Afraid that he/she would not be able to cope if close attachments left 29 81 21 58 .48t
Item 23: Worries about close attachments coming to serious harm§ 29 81 24 67 .50t
Item 4: Experiences.difficulty in sleeping alone at night§ 26 72 24 67 .45t
Item 15: Feels more secure at home with close attachments 26 72 30 83 .41"
Item 22: Worries a lot about close attachments leaving 25 69 21 58 .72$
Item 19: Suffers from panic attacks when thinking about leaving close attachments or about
them leaving§ 24 67 15 42 .62$
Item 26: Worries about the intensity of relationships with close attachments 23 64 23 64 .535
Item 14: Becomes very distressed when thinking about being away from close attachments§ 23 64 21 58 .725
Item 24: Worries about possible events that may separate him/her from close attachments§ 21 58 17 47 .605
Item 17: Becomes very upset with change to usual daily routine if it interferes with contact
with close attachments 21 58 15 42 .67$
Item 1: Anxiety about not speaking to close attachments on the telephone regularly 21 58 19 53 .735
Item 2: Concerned where close attachments are going when separated from them 21 58 23 64 .565
Item 27: Worries that relationships are so close it may cause others problems 19 53 12 33 .71$
Item 10: Talks a lot in order to keep close attachments around 18 50 14 39 .47t
Item 18: Suffers from panic attacks when separated from close attachments§ 17 47 14 39 .51t
Item 16: Becomes very upset when usual routine is disrupted 16 44 9 25 ,565
Item 20: Suffers from nightmares or dreams about separation from close attachments§ 15 42 15 42 .61$
Item 13: Experiences extreme stress before leaving someone close when going away
on a trip§ 15 42 20 56 .43t
Item 6: Better able to sleep if he/she can hear the voices of close attachments or voices
on the "IV or radio 13 36 14 39 ,61$
Item 9: Close attachments have mentioned that he/she talks a lot 13 36 12 33 ,22
Item 12: Experiences extreme stress when leaving home to go on a long trip§ 13 36 17 47 ,33
Item 3: Experiences difficulty in staying away from home for several hours§ 12 33 9 25 .38*
Item 7: Tries to avoid being at home alone when close attachments are out§ 12 33 8 22 .24
Item 8: Carries around something in purse or wallet for security or comfort 12 33 13 36 .41"
Item 11: Experiences physical symptoms before leaving to go to work or other
regular activity§ 10 28 13 36 .43t
Item 21: Suffers from nightmares or dreams about separation from home§ 6 17 6 17 .46t
Item 5: Sleeps better if the lights are on in the house or bedroom 4 11 3 8 .26
* P < .05.
t P < .01.
$ P < .001.
§DSM-IV-derived items.

were unable to determine which comorbid disorder two of 19 subjects with a history of PD-Ag reported
arose first. that such symptoms were unrelated to separation
All subjects with a history of PD without Ag anxiety. All subjects with a diagnosis ofAg (n = 2)
(n = 3) reported that the panic attacks developed in alone attributed the symptoms of anxiety to leaving
association with separation anxiety symptoms. Only
Table 4. Chronology and Severity of ASAD, PD-Ag,
Table 3. JSAD (DSM-IV) in Subjects With ASAD and MDD (N = 36)
JSAD (from DSM-IV ASAD PD-Ag MDD
interviewer-rated checklist) (n = 36) (n = 24) (n = 25)
Present Absent No. % No, % No. %
(n = 22) (n = 14)
Primary disorder
No. % No. %
(4 with only ASAD) 28 78 2 8 4 16
Subject assignment of SA onset Secondary or tertiary
Childhood 22 100 2 14 disorder 6 17 20 83 17 68
Adulthood 0 0 12 86 Unspecified 2 6 2 8 4 16
SASl score (mean ± SD) 4.4 ± 1.3" 3.3 + 1.0 Severity of impact
(mean ± SD) 6.5 ± 2.1 5.2 ± 2.4 6.2 ± 2.5
* P < .01.
ADULT SEPARATION ANXIETY
home rather than to the fear of experiencing anxiety
symptoms in specific situations away from home.
In total, therefore, 22 (92%) of those with a
comorbid anxiety disorder related the symptoms of
PD, Ag, or both directly to separation anxiety.
Subjects were asked to rate the extent to which
each identified psychiatric disorder caused distress
and/or affected their life-style adversely. The rat-
ings were made on a scale ranging from 0 (no
problem/no distress) to 10 (extremely problematic/
extremely distressing). The mean distress rating for
ASAD was 6.5 +__ 2.1; for PD, PD-Ag, and Ag
combined, 5.2 ___2.4; and for depression, 6.2 _+ 2.5.
Dependent Personality Disorder
According to the SCID-II, six subjects (17%)
assigned a diagnosis of ASAD also fulfilled criteria
for dependent personality disorder. It is noteworthy
that a high percentage (56%) of respondents held
university degrees and most reported being highly
functional socially and occupationally apart from
when the separation anxiety symptoms were se-
vere. Furthermore, discussions with subjects indi-
cated that separation anxiety symptoms were expe-
rienced as ego-dystonic, intrusive, unwanted, and
the major source of their disabling anxiety. Many
were perplexed by the symptoms, recognizing that
they were excessive, unrealistic, and inconsistent
with their general level of confidence in other areas
of life. In addition, the majority were able to
recognize that the separation anxiety symptoms
represented a specific problem that was distinguish-
able from PD, Ag, or MDD, and some spontane-
ously expressed disagreement with clinicians who
had assigned the latter diagnoses to them in the
past. All subjects were anxiously attached to only a
few individuals (usually one or two) and denied
being dependent on persons in their wider social or
work networks. These observations and evidence
from the SCID-II tended to suggest that the major-
ity of subjects were not suffering from a pervasive
personality disorder of the dependent type.
Social and Family Histories
It was common for subjects to report distressing
separations from their mother (n = 16, 44%), fa-
ther (n = 16, 44%), or both (n = 11, 31%) before
the age of 18 years for a period that exceeded 3
months. For separations from the mother, the most
frequently cited reasons were parental illness
(n = 4), marital difficulties (n = 4), or parental
death (n = 3). The most commonly cited reason for
279
distressing separations from the father was work
commitments (n = 6). Individual accounts of many
of the subjects (n = 17, 47%) indicated severely
disruptive childhoods with frequent changes of
school and caretakers and consequent difficulties in
bonding with significant adults.
DISCUSSION
In considering the results, the limitations of the
sampling method need to be acknowledged. The
sample consisted of volunteers responding to me-
dia publicity, so they may have constituted an
atypical group who were inclined to endorse inqui-
ries about symptoms of separation anxiety. On the
other hand, it seemed appropriate to study a
self-identified group in the first instance, since such
a sample was most likely to describe the full range
of putative ASA symptoms. Recruitment from an
anxiety clinic sample would yield patients who
were referred primarily for conventional axis I
disorders (such as PD-Ag or GAD). The patients'
awareness of such diagnoses might thus obscure
their recognition of ASA symptoms. Nevertheless,
to advance the present line of research, it will be
important in the future to examine the prevalence
of ASAD in a clinic population.
The risk of researcher bias in making a diagnosis
of ASAD also needs to be considered. A minority of
subjects were judged on global clinical assessment
not to be suffering from ASAD, suggesting that the
interviewer was being discriminatory in making
assignments. Also, there was close concordance
across the key measures of separation anxiety,
which included the ASA-SI, the global assessment,
the self-report checklist that was completed, on
average, 2 to 3 weeks before the interview, and
interrater assignments made blind on audiotaped
interviews by independent raters. Items of the
ASA-SI did not receive universally high endorse-
ment, and similar patterns emerged for the ASA-SI
and ASA-CL. This suggests that subjects were
responding in a discriminatory but consistent man-
ner to items, and adds face validity to the results.
Nevertheless, the constraints of the study suggest
caution in generalizing the findings prior to confir-
matory data being obtained from more systemati-
cally sampled populations. In particular, no assess-
ment could be made of the capacity of individual
items to discriminate between pathological and
normal levels of separation anxiety, an issue that
will need to be pursued in future studies using
appropriately selected comparison groups.
280
Notwithstanding these limitations, the prelimi-
nary data presented, together with the high alpha
statistic (.89) obtained for the items of the ASA-
CL, suggest that it is possible to identify a coherent
constellation of separation anxiety symptoms in
adults that parallel those seen in children and
adolescents. The array of symptoms reported were
wide-ranging and, following age-appropriate modi-
fication, included the criteria identified by DSM-IV
for JSAD. Of 14 items endorsed by at least 50% of
the subjects according to the ASA-SI, five were
DSM-IV-derived, of which four related to anxi-
eties about the safety or proximity of close attach-
ment figures and one reflected unease at sleeping
alone. Four ASA-CL items yielded moderate or low
levels of endorsement and also failed to show
statistical associations with the total ASA-CL score,
with two of these being DSM-IV criteria (items 7
and 12). Item 7 may have been too restrictive, with
only severely affected subjects actually avoiding
being at home alone. Item 12 may have been too
specific in that it applied only to those who
embarked on long trips. The item inquiring into
talking excessively was based on our own clinical
observations that some adults with separation anxi-
ety tend to use talking as a proximity-maintaining
device analogous to the physical clinging of chil-
dren. It is noteworthy that the more specific item 10
that related to excessive talking to keep close
attachments in proximity yielded a higher correla-
tion (.47, P < .001) with the total ASA-CL score.
Item 5 (sleeping better with the lights on), a
common symptom associated with JSAD, may be
less relevant in adults in whom such a ritual might
prove impractical or embarrassing.
How these findings are interpreted depends to
some extent on the conceptual framework used in
relation to the nosology of anxiety disorders. 13,29,3°
Those who support a unitary concept of "neurosis"
might argue that symptoms of separation anxiety in
adulthood represent one dimension of a general
neurotic diathesis that may manifest in a wide array
of symptoms over the course of a lifetime. 29,3°Such
a perspective would, by implication, discourage
any attempt to delineate ASAD as a distinctive
nosological category within the anxiety disorders.
A further possibility is that symptoms of separation
anxiety represent a nonspecific reaction or coping
pattern that some individuals manifest when they
develop affective disorders such as PD or MDD.
An alternative possibility that is supported tenta-
tively by the present study is that, at least in some
MANICAVASAGAR, SlLOVE, AND CURTIS
instances, it may be justified to assign a primary
diagnosis of separation anxiety disorder in adult-
hood. The results indicate that the majority of
subjects (75%) assigned a provisional diagnosis of
ASAD reported that separation anxiety symptoms
preceded the onset of PD-Ag or MDD and that the
level of distress caused by separation anxiety
symptoms was at least equivalent to that generated
by the comorbid disorders. The majority of subjects
judged that the onset of panic attacks was closely
associated with separation fears, whereas those
who received diagnoses of Ag related their "avoid-
ance" primarily to leaving home rather than to a
fear of having a panic attack or other anxiety
symptoms in specific external situations. In total,
96% of the subjects with comorbid anxiety disor-
ders reported that separation anxiety was directly
associated with the onset of those disorders. This
finding tentatively supports the possibility that
ASAD may precede and also act as a major trigger
for PD-Ag in a subcategory of cases.
Nevertheless, in drawing such an inference,
errors in recall must be acknowledged. The SCID
inquires as to whether respondents "ever" experi-
enced the relevant symptoms, so it is always
possible that the early symptoms of a disorder, for
example, PD arising in childhood, may have been
forgotten by the subject. Hence, attempts to deter-
mine the order of onset of comorbid disorders
retrospectively can only be regarded as tentative. It
also needs to be acknowledged that our diagnostic
assessments did not cover all subcategories of
anxiety and depressive disorders, so the question of
comorbidity requires further evaluation, particu-
larly in relation to disorders such as social phobia
and obsessive-compulsive disorder. However, it is
noteworthy that at a qualitative level, subjects
strongly endorsed the view that the symptoms of
separation anxiety were their central concern, with
many expressing dissatisfaction with prior diag-
noses assigned to them by mental health profession-
als they had consulted in the past.
As might be expected, the majority (two thirds)
of subjects reported suffering from JSAD in their
early years, suggesting a developmental continuity
between early-onset and adult symptoms of separa-
tion anxiety. However, notably, one third of the
subjects reported the onset of separation anxiety
disorder after 18 years of age, suggesting that in
some individuals the disorder can arise for the first
time in adulthood. Juvenile-onset subjects had
extremely high SASI scores that were equivalent to
ADULT SEPARATION ANXIETY
those of a sample of subjects with past school
phobia, 28 whereas adult-onset cases reported levels
of early separation anxiety that approximated those
of normative populations. 25 It is possible that the
adult-onset cases may simply have forgotten their
childhood symptoms of separation anxiety. On the
other hand, it seems feasible that the age of onset of
separation anxiety disorder may follow a normal
distribution, with most cases having an onset in
early life, but some occurring in early adulthood.
Should the latter hypothesis be confirmed, it will
challenge the prevailing view that separation anxi-
ety disorder arises only in the juvenile years.
A further issue is whether the cases we have
tentatively defined as having ASAD may not be
better assigned an axis II disorder, particularly
dependent personality disorder. It is noteworthy
that only a minority of subjects (17%) fulfilled
SCID-II criteria for dependent personality disorder,
with such assignments being confirmed by indepen-
dent ratings of audiotaped interviews. If such
personality assignments were correct, then the
concordance between dependent personality disor-
der and ASAD would seem to be relevant to only a
small percentage of cases. A number of collateral
features also argued against a close identity be-
tween separation anxiety symptoms and features
usually associated with personality disorder. Most
subjects were highly educated and successful in
their work, and they regarded the separation anxi-
ety as a circumscribed problem area in their lives.
Respondents showed sound insight, judging their
separation fears as ego-alien, unrealistic, and a
source of personal suffering. Many reported fluctua-
tions in the symptoms, with periods of freedom
from separation anxiety. Their excessive attach-
ment anxieties related to only a small number of
intimate others, with no evidence of indiscriminate
interpersonal dependency in the wider social or
work domains. Nevertheless, it must be acknowl-
edged that other relevant personality traits (such as
avoidant or borderline) were not assessed, and
further study will thus be necessary to fully clarify
the relationship between A S A symptoms and rel-
evant personality characteristics.
Thus, although preliminary, the present data
allow cautious speculation about possible develop-
mental pathways leading to the onset and persis-
tence of separation anxiety. Bowlby 31 tended to use
the term "separation anxiety" to refer to a construct
of attachment theory rather than to an operationally
defined disorder. Within such a framework, he sug-
281
gested that although separation anxiety should be
regarded as an evolutionarily adaptive mechanism,
attachment insecurities might be intensified by a
number of aberrant parent-child bonding patterns.
Affected children might be left with a long-lasting
vulnerability to interpersonal stress, placing them
at risk to a wide range of anxiety, depressive, and
personality difficulties in later years.
Klein 21 gave further emphasis to the biological
roots of separation anxiety by suggesting that such
insecurities reflected a neurophysiological prepared-
ness to deal with threats to primary bonds. He
suggested that variations in the intensity of such
anxieties may be at least partially under genetic
control, a postulate that has received some endorse-
ment by a recent twin study.3z The investigations
that have followed Klein's 7 early study have sug-
gested that a substantial percentage but not all
patients with PD-Ag report heightened levels of
early separation anxiety. 13 However, how early
separation anxiety leads to the onset of adult
PD-Ag or other affective disorders has remained
unclear. The present data suggest the possibility
that in some individuals, possibly because of a
strong genetic loading and/or ongoing insecurities
in their primary bonds, early separation anxiety
disorder may progress to an adult form of the
equivalent disorder. It seems plausible that vulner-
able individuals may not only suffer a worsening of
the separation anxiety symptoms under conditions
of interpersonal insecurity, but that in many cases
the distress and insecurity thus engendered will
precipitate secondary anxiety, phobic, and depres-
sive symptoms.
Conclusions
The inferences drawn from the present phenom-
enological study can only be tentative, given the
method of sampling used. As noted, volunteer
subjects may have been influenced by the demand
characteristics of the situation, and hence their
responses may have been biased to fulfill the
evident purpose of the study. Furthermore, even if
the data are accurate, they do not provide any
information about the prevalence of ASAD or the
extent to which the volunteers were representative
of a wider population of persons experiencing
separation anxiety symptoms. Nevertheless, the
study was able to identify a group of individuals
who appeared to be suffering from a wide range of
separation anxiety symptoms in adulthood that
included some of the core criteria specified by
282
DSM-IV for diagnosing the juvenile-onset form of
the disorder. Furthermore, such symptoms ap-
peared to be central to the respondents' major
concerns, and in most instances, comorbid PD-Ag
and MDD appeared to be secondary to underlying
separation anxiety symptoms. Thus, further investi-
gations aimed at testing the hypothesis that adults
may experience a primary form of separation
anxiety disorder seem warranted. If future studies
support such a postulate, this may provide an
REFERENCES
1. American Psychiatric Association. Diagnostic and Statisti-
cal Manual of Mental Disorders. Ed. 4. Washington, DC:
American Psychiatric Association, 1994.
2. Flakierska N, Lindstrom M, Gillberg C. School refusal: a
15-20 year follow-up study of 35 Swedish urban children. Br J
Psychiatry 1988;152:834-837.
3. Berg I, Jackson A. Teenage school refusers grow up: a
follow-up study of 168 subjects, ten years on average after
in-patient treatment. Br J Psychiatry 1985;147:366-370.
4. Warren W. Some relationships between the psychiatry of
children and of adults. J Ment Sci 1960;106:815-826.
5. Coolidge J, Brodie R, Feeney B. A ten-year follow-up
study of sixty-six school-phobic children. Am J Orthopsychiatry
1964;34:675-684.
6. Waldron S. The significance of childhood neurosis for
adult mental health: a follow-up study. Am J Psychiatry
1976;133:532-538.
7. Klein D. Delineation of two drug-responsive anxiety
syndromes. Psychopharmacologia 1964;5:397-408.
8. Berg I, Butler A, Hall G. The outcome of adolescent
school phobia. Br J Psychiatry 1976;128:80-85.
9. Berg I. School phobia in the children of agoraphobic
women. Br J Psychiatry 1976; 128:86-89.
10. Silove D, Harris M, Morgan A, Boyce P, Manicavasagar
V, Hadzi-Pavlovic D, et al. Is early separation anxiety a specific
precursor of panic disorder-agoraphobia?: a community study.
Psychol Med 1995;25:405-411.
11. Shear MK. Factors in the etiology and pathogenesis of
panic disorder: revisiting the attachment-separation paradigm.
Am J Psychiatry 1996; 153:125-136.
12. Lipsitz JD, Martin LY, Manuzza S, Chapman TF, Liebow-
itz MR, Klein DF, et al. Childhood separation anxiety disorder in
patients with adult anxiety disorders. Am J Psychiatry 1994;151:
927-929.
13. Silove D, Manicavasagar V, Curtis J, Blaszczynski A. Is
early separation anxiety a risk factor to adult panic disorder?: a
critical review. Compr Psychiatry 1996;37:1-14.
14. Sroufe L. The role of the infant-caregiver attachment in
development. In: Belsky J, Nezworksi T (eds): Clinical Implica-
tions of Attachment. Hillsdale, NJ: Erlbaum, 1988:18-38.
15. Grossmann K, Grossmann K. Attachment quality as an
organizer of emotional and behavioural responses in a longitudi-
nal perspective. In: Parkes C, Stevenson-Hinde J, Marris P (eds):
Attachment Across the Life Cycle. London, UK: Routledge,
1991:93-114.
16. Klein D. Endogenomorphic depression: a conceptual and
terminological revision. Arch Gen Psychiatry 1974;31:447-454.
MANICAVASAGAR, SlLOVE, AND CURTIS
impetus for the development of specific treatments
that will more directly address the problems of
separation anxiety in adulthood.
ACKNOWLEDGMENT
The authors would like to express their gratitude to Shakeh
Momartin for assistance in establishing the interrater reliability
for the measure of adult separation anxiety. We also are grateful
to Bill Paterson for statistical and editorial comments, and to
Professor Donald Klein for helpful comments and advice in
preparing the manuscript.
17. Boyce P, Parker G. Development of a scale to measure
interpersonal sensitivity. Aust NZ J Psychiatry 1989;23:341-
351.
18. Bowlby J. Attachment and Loss. Vol. 2. Separation,
Anxiety and Anger. New York, NY: Penguin Books, 1973:299.
19. Hersov L. Refusal to go to school J Child Psychol
Psychiatry 1960;1:137-145.
20. Tyrer P, Tyrer S. School refusal, truancy and adult
neurotic illness. Psychol Med 1974;4:416-421.
21. Klein D. Anxiety reconceptualized. Early experience
with imipramine and anxiety. Compr Psychiatry 1980;21:411-
427.
22. Bowlby J. Attachment and Loss. Vol. 1. Attachment,
London, UK: Penguin Books, 1969.
23. Bryman A, Burgess R. Analyzing Qualitative Data.
London, UK: Roufledge, 1994.
24. Crabtree B, Miller W. A qualitative approach to primary
care research: the long interview. J Fam Med 1991;23:145-151.
25. Silove D, Manicavasagar V, O'Connell D, Blaszczynski
A, Wagner R, Henry J. The development of the Separation
Anxiety Symptom Inventory (SASI). Aust NZJ Psychiatry
1993;27:477-488.
26. Spitzer RL, Williams JBW, Gibbon M, First MB. Struc-
tured Clinical Interview for DSM-III-R, Patient Edition. Wash-
ington, DC: American Psychiatric Press, 1990.
27. Spitzer RL, Williams JBW, Gibbon M, First MB. Struc-
tured Clinical Interview for DSM-IU-R Personality Disorders
(SCID-II, Version 1.0). Washington, DC: American Psychiatric
Press, 1990.
28. Silove D, Manicavasagar V. Adults who feared school: is
early separation anxiety specific to the pathogenesis of panic
disorder? Acta Psychiatr Scand 1993;88:385-390.
29. Andrews G, Stewart G, Morris-Yates A, Holt E Hender-
son S. Evidence for a general neurotic syndrome. Br J Psychia-
try 1990;157:6-12.
30. Tyrer P, Seivewright N, Ferguson B, Tyrer J. The general
neurotic syndrome: a coaxial diagnosis of anxiety, depression
and personality disorder. Acta Psychiatr Scand 1992;85:201-
206.
31. Bowlby J. A Secure Base: Parent-Child Attachment and
Healthy Human Development. New York, NY: Basic Books,
1988.
32. Silove D, Manicavasagar V, O'Connell D, Morris-Yates
A. Genetic factors in early separation anxiety: implications for
the genesis of adult anxiety disorders. Acta Psychiatr Scand
1995;92:17-24.