ORIGINAL RESEARCH

Separation Anxiety and Actual Separation Experiences During Childhood in Patients with Panic Disorder
Borwin Bandelow, MD1, Gabriel Álvarez Tichauer, MD2, Corinna Späth, MD3, Andreas Broocks, MD4, Göran Hajak, MD5, Stefan Bleich, MD6, Eckart Rüther, MD7

Objective: The association between separation anxiety in childhood and actual separation experiences during childhood has not yet been investigated in patients with panic disorder.
Methods:  In 115 patients with panic disorder with or without agoraphobia and in 124 control subjects without a history of psychiatric illness, we assessed separation anxiety during childhood, retrospectively, using DSM-IV and ICD-10 criteria and the Separation Anxiety Symptom Inventory (SASI). In addition, actual separation experiences from age 0 to 15 years were assessed, retrospectively.
Results:  A total of 22.6% of the patients and 4.8% of the control subjects fulfilled both DSM-IV and ICD-10 criteria for childhood separation anxiety (c2 =11.8; P < 0.0001). Further, 57.4% of the patients and 37.9% of the control subjects reported actual separation experiences during their childhood (c2 = 9.09, P < 0.003). Separation anxiety and actual separation experiences, however, were independent of each other.
Conclusion:  These results suggest that separation anxiety during childhood is not a consequence of actual traumatic separation experiences in panic disorder patients.

(Can J Psychiatry 2001;948–952)

Key Words:  panic disorder, agoraphobia, separation anxiety disorder


For the first time in 1964, Donald F Klein observed that 16 out of a group of 32 persons with agoraphobia had suffered from separation anxiety during childhood (1). Likewise, in other studies which did not use a healthy control group, high rates of 18% to 54% of separation anxiety were reported (2–6). In most studies that compared patients with panic disorder with a control group, the incidence of separation anxiety and its severity were significantly higher for panic disorder patients than for healthy control subjects (7–11) or surgical outpatients (12,13).

Manuscript received June 2001 and accepted October 2001.
1Professor, Department of Psychiatry and Psychotherapy, University of Göttingen, Germany.
2Research Assistant, Department of Psychiatry and Psychotherapy, University of Göttingen, Germany.
3Research Assistant, Department of Psychiatry and Psychotherapy, University of Göttingen, Germany.
4Assistant Professor, Department of Psychiatry and Psychotherapy, University of Lübeck, Germany.
5Professor, Department of Psychiatry and Psychotherapy, University of Regensburg, Germany.
6Research Assistant, Department of Psychiatry and Psychotherapy, University of Erlangen, Germany.
7Professor, Head, Department of Psychiatry and Psychotherapy, University of Göttingen, Germany.
Address for correspondence:  Prof Dr Borwin Bandelow, Department of Psychiatry, University of Göttingen, von-Siebold-Str, D-37075 Göttingen, Germany
e-mail: bbandel@gwdg.de

Taken as a whole, the evidence provides support for the hypothesis that early separation anxiety is an antecedent of adult panic disorder (14). We see higher rates of separation anxiety in more severely ill patients (for example, those with agoraphobia or with comorbid conditions (3,5,6,9,11,15,16).

Even so, separation anxiety during childhood does not seem to be specific for panic disorder. The rates for separation anxiety for patients with panic disorder were found to be similar to the rates in patients with depression (11,17), generalized anxiety disorder (10,11,18), social phobia (5,11), obsessive–compulsive disorder (5), or specific phobia (19,20). One study found higher rates in panic disorder than in generalized anxiety disorder or other phobic disorders (21).

These retrospective data have their limitations. Prospective studies that investigate the association between childhood separation anxiety and adult panic disorder, however, are unavailable.

Next