Trauma and Depersonalization During Panic Attacks

To the Editor: Randall D. Marshall, M.D., and colleagues (1) failed to find support for their hypothesized association between the symptoms of depersonalization during panic attacks and a history of childhood trauma. They cautioned that the generalizability of their null findings was limited by the use of a convenience group of patients participating in a pharmacological trial for panic disorder and by the exclusion of patients with comorbid disorders. To address these limitations, we reexamined the relationship between childhood trauma and the presence of depersonalization symptoms during panic attacks using the National Comorbidity Survey (2) public use database. The National Comorbidity Survey is a nationally representative sample of U.S. adults that contains information on the prevalence and correlates of psychiatric disorders.

Adults who met DSM-III-R criteria for panic disorder (N=186) were selected for participation. An appropriate statistical weight from the National Comorbidity Survey, part 2, public use database was used to make the data representative of the general population. Respondents were queried about symptoms typically experienced during their most severe panic attacks. Responses to the depersonalization-related question (“Did you or things around you seem real?”) were used to assign participants to either a depersonalization group (N=109) or a nondepersonalization group (N=77).

Relative to the nondepersonalization group, the depersonalization group more frequently reported a history of serious neglect as a child (N=15, 13.8%, versus N=3, 3.9%, respectively), having been raped before age 16 (N=12, 11.0%, versus N=1, 1.3%), experiencing physical abuse as a child (N=18, 16.5%, versus N=7, 9.1%), and a history of molestation before age 16 (N=17, 15.6%, versus N=9, 11.7%). The differences regarding childhood neglect and rape were statistically significant (χ²=5.0, df=1, p<0.03, and χ²=6.5, df=1, p<0.02, respectively).

It is possible that the relatively high endorsement of depersonalization by those with a history of either neglect or rape may reflect greater severity of panic attacks rather than a more specific tendency to experience depersonalization during panic attacks. However, this alternative explanation was not supported by a series of additional chi-square analyses. Specifically, the other panic attack symptoms were not more frequently endorsed by those with a history of childhood neglect than by those without such a history and were not more frequently endorsed by those with a history of rape than by those without such history.

In contrast to the findings of Dr. Marshall and colleagues (1), the present results suggest there is a link between traumatic childhood events and depersonalization during panic attacks. A large representative sample arguably would provide a more complete picture of the proposed relationship between depersonalization and trauma. We suggest that further research in this area is warranted.