Separation anxiety in adulthood: A phenomenological investigation

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 administered a semistructured interview and a self-report checklist for adult separation anxiety {ASA) symptoms, 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), agoraphobia (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

global clinical rating and/or endorsement of DSM-IVderived 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 limited 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

EPARATION ANXIETY DISORDER is now 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 agoraphobia (Ag). However, no criteria are specified for a diagnosis of separation anxiety disorder in adulthood, the implication being that the juvenile criteria 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

procedures have prevented clarification of the precise diagnostic outcomes in adulthood.4-6Retrospective 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 remains 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 levels of early separation anxiety are linked specifically to adult PD-Ag. 1° Uncertainty about the long-term outcome of JSAD is increased by methodological 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 attachment theory increasingly have emphasized the continuities between attachment insecurities in early and later life, 14,15with terms like "rejection sensitivity" 16 or "interpersonal sensitivity" 17 used to denote 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

S

From the Psychiatry Research and Teaching Unit, School of Psychiatry, University of New South Wales at Liverpool Hospital, Liverpool, New South Wales, Australia. Address reprint requests to V. Manicavasagar, M. Psychol., Psychiatry Research and Teaching Unit, Level 4, Health Services 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

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 anxiety 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 att 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 affective 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 separation 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. Subjects 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 conducted 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 devised 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 questionnaire 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

MANICAVASAGAR, SILOVE, A N D CURTIS

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 subjects were included who met criteria either according 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 Positive (n = 34)

DSM-IV Negative (n = 10)

Global rating positive (n = 34)

32

2

Global rating negative (n = 10)

2

8

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 discrepancies 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-IVderived 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 construct 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 mentioned 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 symptoms began in childhood and reappeared in adulthood at times of stress or threatened loss. Twelve

ADULT SEPARATION ANXIETY

277

35 N U

M

30

B

E

25

0

20

R

F

S U B

Fig 1. Frequency of item endorsement of current ASA symptoms. ([]} Interview; (+} checklist.

J E C T S

15 10 5 0 1

2 3 4 5

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 juvenileonset 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 diagnosed with JSAD was high compared with normative 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 individual subjects indicated that several had refused job promotions or had deliberately reorganized

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 appeared 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

MANICAVASAGAR, SILOVE, AND CURTIS

278

Table 2. Endorsement Rates of Separation Anxiety Items According to the ASA-SI and ASA-CL

Items in Order of Endorsement

No.

%

No.

%

Item Corre(ation With Total ASA-CLScore

Item 25: Afraid that he/she would not be able to cope if close attachments left Item 23: Worries about close attachments coming to serious harm§ Item 4: Experiences.difficulty in sleeping alone at night§ Item 15: Feels more secure at home with close attachments Item 22: Worries a lot about close attachments leaving Item 19: Suffers from panic attacks when thinking about leaving close attachments or about them leaving§ Item 26: Worries about the intensity of relationships with close attachments Item 14: Becomes very distressed when thinking about being away from close attachments§ Item 24: Worries about possible events that may separate him/her from close attachments§ Item 17: Becomes very upset with change to usual daily routine if it interferes with contact with close attachments Item 1: Anxiety about not speaking to close attachments on the telephone regularly Item 2: Concerned where close attachments are going when separated from them Item 27: Worries that relationships are so close it may cause others problems Item 10: Talks a lot in order to keep close attachments around Item 18: Suffers from panic attacks when separated from close attachments§ Item 16: Becomes very upset when usual routine is disrupted Item 20: Suffers from nightmares or dreams about separation from close attachments§ Item 13: Experiences extreme stress before leaving someone close when going away on a trip§ Item 6: Better able to sleep if he/she can hear the voices of close attachments or voices on the "IV or radio Item 9: Close attachments have mentioned that he/she talks a lot Item 12: Experiences extreme stress when leaving home to go on a long trip§ Item 3: Experiences difficulty in staying away from home for several hours§ Item 7: Tries to avoid being at home alone when close attachments are out§ Item 8: Carries around something in purse or wallet for security or comfort Item 11: Experiences physical symptoms before leaving to go to work or other regular activity§ Item 21: Suffers from nightmares or dreams about separation from home§ Item 5: Sleeps better if the lights are on in the house or bedroom

29 29 26 26 25

81 81 72 72 69

21 24 24 30 21

58 67 67 83 58

.48t .50t .45t .41" .72$

24 23 23 21

67 64 64 58

15 23 21 17

42 64 58 47

.62$ .535 .725 .605

21 21 21 19 18 17 16 15

58 58 58 53 50 47 44 42

15 19 23 12 14 14 9 15

42 53 64 33 39 39 25 42

.67$ .735 .565 .71$ .47t .51t ,565 .61$

15

42

20

56

.43t

13 13 13 12 12 12

36 36 36 33 33 33

14 12 17 9 8 13

39 33 47 25 22 36

,61$ ,22 ,33 .38* .24 .41"

10 6 4

28 17 11

13 6 3

36 17 8

.43t .46t .26

ASA-SI (n = 36)

ASA-CL (n = 35)

* P < .05. t P < .01. $ P < .001. §DSM-IV-derived items.

were unable to determine which comorbid disorder arose first. All subjects with a history of PD without Ag (n = 3) reported that the panic attacks developed in association with separation anxiety symptoms. Only

two of 19 subjects with a history of PD-Ag reported that such symptoms were unrelated to separation anxiety. All subjects with a diagnosis ofAg (n = 2) alone attributed the symptoms of anxiety to leaving Table 4. Chronology and Severity of ASAD, PD-Ag, and MDD (N = 36)

Table 3. JSAD (DSM-IV) in Subjects With ASAD

ASAD (n = 36)

JSAD (from DSM-IV interviewer-rated checklist) Present (n = 22) No. Subject assignment of SA onset Childhood Adulthood SASl score (mean ± SD) * P < .01.

%

22 100 0 0 4.4 ± 1.3"

Absent (n = 14) No.

%

2 14 12 86 3.3 + 1.0

Primary disorder (4 with only ASAD) Secondary or tertiary disorder Unspecified Severity of impact (mean ± SD)

PD-Ag (n = 24)

MDD (n = 25)

No.

%

No,

%

No.

%

28

78

2

8

4

16

6 2

17 6

20 2

83 8

17 4

68 16

6.5 ± 2.1

5.2 ± 2.4

6.2 ± 2.5

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 ratings 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 severe. Furthermore, discussions with subjects indicated that separation anxiety symptoms were experienced 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 distinguishable from PD, Ag, or MDD, and some spontaneously 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 majority 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%), father (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 media publicity, so they may have constituted an atypical group who were inclined to endorse inquiries 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 endorsement, and similar patterns emerged for the ASA-SI and ASA-CL. This suggests that subjects were responding in a discriminatory but consistent manner to items, and adds face validity to the results. Nevertheless, the constraints of the study suggest caution in generalizing the findings prior to confirmatory data being obtained from more systematically sampled populations. In particular, no assessment 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 preliminary data presented, together with the high alpha statistic (.89) obtained for the items of the ASACL, 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 modification, 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 anxieties about the safety or proximity of close attachment 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 anxiety tend to use talking as a proximity-maintaining device analogous to the physical clinging of children. It is noteworthy that the more specific item 10 that related to excessive talking to keep close attachments in proximity yielded a higher correlation (.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 tentatively 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 adulthood. 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 "avoidance" 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 disorders 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" experienced 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 determine 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, particularly 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 diagnoses assigned to them by mental health professionals 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 separation 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 anxiety 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 independent ratings of audiotaped interviews. If such personality assignments were correct, then the concordance between dependent personality disorder 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 between 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 anxiety 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 fluctuations in the symptoms, with periods of freedom from separation anxiety. Their excessive attachment 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 acknowledged 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 relevant personality characteristics. Thus, although preliminary, the present data allow cautious speculation about possible developmental pathways leading to the onset and persistence 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 preparedness 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 endorsement by a recent twin study.3z The investigations that have followed Klein's 7 early study have suggested 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 vulnerable 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 depressive symptoms. Conclusions

The inferences drawn from the present phenomenological 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

MANICAVASAGAR, SlLOVE, AND CURTIS

DSM-IV for diagnosing the juvenile-onset form of the disorder. Furthermore, such symptoms appeared 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 investigations 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

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.

REFERENCES 1. American Psychiatric Association. Diagnostic and Statistical 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, Liebowitz 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 Implications 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 longitudinal 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.

17. Boyce P, Parker G. Development of a scale to measure interpersonal sensitivity. Aust NZ J Psychiatry 1989;23:341351. 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;21:411427. 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. Structured Clinical Interview for DSM-III-R, Patient Edition. Washington, DC: American Psychiatric Press, 1990. 27. Spitzer RL, Williams JBW, Gibbon M, First MB. Structured 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 Henderson S. Evidence for a general neurotic syndrome. Br J Psychiatry 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;85:201206. 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.