Biological as well as psychological concepts stress the importance of SA during childhood in the etiology of adult PD and agoraphobia (1,2). Attachment theory claims that SA during childhood is an instinctive response to the experience of separation from an attachment figure (2). Thus recurrent or lasting experiences of separation can be an important factor in the development of separation anxiety disorder as well as panic disorder and agoraphobia. In contrast, Klein postulates a joint neurobiological disturbance, widely independent of interpersonal events, as the origin for these disorders (1). As higher rates of SA and separation events during childhood have been found in patients with PD and agoraphobia when compared with control subjects (3–5), Bowlby’s assumption seems to be supported.
So far, only one study with PD patients investigated the association between SA and separation experiences during childhood and surprisingly found the 2 conditions to be completely independent of each other (6). The authors concluded that, in some patients, SA might have been an antecedent of later PD, whereas in other patients, actual separation experiences might have predisposed PD development in adulthood but that there was little overlap between the groups. From a theoretical perspective, it appears not very plausible that separation events do not trigger SA at least to some degree. Thus there may be an alternative explanation for these data.
Our study aimed to replicate the data from Bandelow and others (6) and to further investigate the association between SA and separation events during childhood. Like Bandelow and others (6), we assumed that retrospective data are susceptible to faulty recall. We speculated that retrospective ratings of childhood SA could be distorted by separation events during adulthood.
Subjects were agoraphobia patients in our Behaviour Therapy Outpatient Unit at the Hamburg University Hospital. Of 51 contacted patients, 42 (82%) participated in the follow-up study. Diagnosis of agoraphobia was made by clinically experienced psychiatrists or psychologists according to DSM-IV criteria (7) after an extended intake interview. Forty patients suffered from PD with agoraphobia. Two patients fulfilled the criteria of agoraphobia without PD. We only included female patients who had completed exposure-in-vivo treatment in our outpatient unit. Follow-up periods differed from 3 to 9 years (mean 6 years, SD 2).
We assessed separation anxiety during childhood retrospectively at follow-up, using the SASI (8), a self-rated questionnaire concerning separation anxiety before age 18 years. Actual separation experiences were recorded retrospectively at follow-up according to patients’ recollection. We applied definition criteria of a previous study (9): death of or separation from mother and (or) father, divorce or de facto separation of parents, death of a sibling, death of any other cohabiting relative, and severe and chronic illness in the child with hospitalization. Duration of separation had to last 6 months or more and had to occur from age 0 to 18 years.
All patients received exposure in-vivo either in therapist-aided group therapy (10) or as a single treatment. In both settings, exposure was based on a flooding paradigm including anxiety and panic management training. Both treatment procedures included 4 therapist-aided and 1 self-conducted exposure-in-vivo session (4 hours per session). Additionally, patients did exposure exercise as homework.
Separation anxiety during childhood measured by SASI in our agoraphobia patients was significantly higher, compared with nonclinical control subjects taken from Bandelow and others (6). SASI scores of patients with PD from the same study (6) were almost identical to ours (Table 1).
The frequency of actual separation events during childhood for our agoraphobia patients was similar to that found in PD patients by Bandelow and colleagues (6). In addition, rates of childhood separation events for our agoraphobia patients were significantly higher than were those for nonclinical control subjects, which Bandelow and others also found (6) (Table 1).
There were no significant differences in childhood SA between patients with or without separation experiences during childhood (Table 2). Again, this is in line with Bandelow and others’ results (6). Patients who had experienced a separation from a spouse during adulthood scored higher on separation anxiety during childhood, but differences failed to be significant. Patients with a separation from a spouse within the follow-up period scored even higher on SA. Their ratings were significantly higher, compared with patients without a separation during follow-up (Table 2).
Childhood separation anxiety plays an important role in etiologic hypotheses for adult PD and agoraphobia (1,2). Whether separation events during childhood may trigger SA, as hypothesized in the attachment theory, has so far only been investigated by Bandelow and others (6). Like previous studies (3–5), the authors found higher rates of childhood SA and actual separation events during childhood, but both conditions were completely independent of each other. Our data replicated these results in women with PD with agoraphobia (except for 2 patients with agoraphobia without PD). The calculated rates of SA and separation events were almost identical. We also did not find any association between the 2 conditions. Because there is no overlap between SA and separation events during childhood, these data appear to contradict an important assumption of the attachment theory (2). However, there may be an explanation for these data, alternative to concluding that both groups are completely independent of each other. We found childhood SA to be associated with separation events during adulthood. Patients who experienced a separation from a spouse scored higher on SA measured by the SASI (8) than did patients without any comparable event. It appeared that this relation became stronger the closer the separations were to the date of the ratings. Given that actual separation events are recalled more accurately than are such emotional states as separation anxiety, we conclude that retrospective ratings of childhood SA were confounded by adult separation events. A possible explanation for this bias is that involved patients rated separation anxiety that they experienced during adulthood instead of childhood SA. It also could be possible that experiencing separation events and separation anxiety during adulthood, especially in the immediate past, activates relevant memories from childhood.
Owing to the retrospective nature of the data presented, these results have to be interpreted with caution. Our own results suggest that memories can be distorted by faulty recalls or systematic biases. Facing the necessary investigation periods of at least 2 decades, warranted prospective studies will be difficult to conduct. Future research should differentiate between separations in the first years of life and in later childhood, since separations in early childhood may be more harmful. It would also be useful to examine the relation between follow-up status, in terms of diagnosis of panic and of agoraphobia, in future studies.
Funding and Support
No funding or support was received for this paper.
1. Klein DF. Anxiety reconceptualized. In: Klein DF, Rabkin J, editors. Anxiety: new research and changing concepts. New York: Raven Press; 1981. p 235–63.
2. Bowlby J. Attachment and loss. Vol 1. New York: Basic Books; 1982.
3. Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LE. Childhood parental loss and adult psychopathology in women. A twin study perspective. Arch Gen Psychiatry 1992;49:109–16.
4. Shear K. Factors in the etiology and pathogenesis of panic disorder: revisiting the attachment-separation paradigm. Am J Psychiatry 1996;153 (Suppl 7):125–36.
5. Silove D, Maniacasagar V, Curtis J, Blaszczynski A. Is early separation anxiety a risk factor for adult panic disorder? A critical review. Compr Psychiatry 1996;37:167–79.
6. Bandelow B, Alvarez-Tichauer G, Spath C, Brooks A, Hajak G, Bleich S, and others. Separation anxiety and actual separation experiences during childhood in patients with panic disorder. Can J Psychiatry 2001;10:948–52.
7. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington (DC): APA; 1994.
8. Silove D, Manicavasagar V, Connell O, Blaszczynski A, Wagner R, Henry J. The development of the separation anxiety symptom inventory (SASI). Aust N Z J Psychiatry 1993;27:477–88.
9. Faravelli C, Webb T, Ambonetti A, Fonnesu F, Sessarego A. Prevalence of traumatic early life events in 31 agoraphobic patients with panic attacks. Am J Psychiatry 1985;142:1493–4.
10. Hand I. Group exposure in vivo for agoraphobics (1974): a multifaceted pilot study and its impact on subsequent agoraphobia research. Behav Cognitive Psychother 2000;28:335–51.
Manuscript received September 2004, revised, and accepted May 2005.
1. Assistant Professor, Department of Psychiatry and psychotherapy, University Hospital Hamburg-Eppendorf, Department of Psychiatry and Psychotherapy, Martinistr 52, D-20246 Hamburg, Germany.
2. Clinical Psychologist, Department of Psychiatry and Psychotherapy, University Hospital Hamburg-Eppendorf, Martinistr 52, D-20246 Hamburg, Germany.
3. Professor, Department of Psychiatry and Psychotherapy, University Hospital Hamburg, Eppendorf—Martinistr 52, D-20246 Hamburg, Germany.
4. Assistant Professor, Department of Psychiatry, University Hospital Zurich, Cullmannstr 8 Ch-8091 Zurich, Switzerland.
Address for correspondence: Dr H Peter, University Hospital Hamburg-Eppendorf, Department of Psychiatry and Psychotherapy, Martinistr 52, D-20246 Hamburg, Germany.
1 | 2
Archives in English | Archives
RCP en français