Personality disorder traits as predictors of subsequent first-onset panic disorder or agoraphobia

Abstract

Determining how personality disorder traits and panic disorder and/or agoraphobia relate longitudinally is an important step in developing a comprehensive understanding of the etiology of panic/agoraphobia. In 1981, a probabilistic sample of adult (≥18 years old) residents of east Baltimore were assessed for Axis I symptoms and disorders using the Diagnostic Interview Schedule (DIS); psychiatrists reevaluated a subsample of these participants and made Axis I diagnoses, as well as ratings of individual Diagnostic and Statistical Manual of Mental Disorders, Third Edition personality disorder traits. Of the participants psychiatrists examined in 1981, 432 were assessed again in 1993 to 1996 using the DIS. Excluding participants who had baseline panic attacks or panic-like spells from the risk groups, baseline timidity (avoidant, dependent, and related traits) predicted first-onset DIS panic disorder or agoraphobia over the follow-up period. These results suggest that avoidant and dependent personality traits are predisposing factors, or at least markers of risk, for panic disorder and agoraphobia—not simply epiphenomena.

Introduction

Determining how personality disorder traits and panic disorder and/or agoraphobia relate longitudinally is an important step in developing a comprehensive understanding of the etiology of panic/agoraphobia. Cluster C (anxious cluster) personality traits, especially avoidant and dependent traits, are strongly related to anxiety disorders, including panic disorder and agoraphobia, cross-sectionally [1], [2], [3].

Some have argued that avoidant and dependent traits are predisposing factors for (or prodromal factors in) panic/agoraphobia, based on retrospective reports of anxious patients' premorbid personalities [4]. However, relying on retrospective reports introduces potential recall bias, that is, patients with current panic/agoraphobia may have biased memories of their premorbid personality traits.

Others have argued that early panic symptoms probably shape personality, for example, enhancing avoidant and dependent tendencies [5], [6]. Indeed, personality abnormalities can be diminished somewhat by effective treatment of panic disorder and agoraphobia [7], [8], and this may indicate some degree of state-trait confounding in the context of acute psychopathology [9]. Of course, a state-trait confounding explanation presumes that treatment of panic/agoraphobia has no influence on personality traits themselves; this may not be correct [10], [11].

The most informative method for determining whether or not personality disorder traits are risk factors for panic/agoraphobia is to longitudinally relate the former to later first onset of the latter; we know of only one prior study that used this method [12]. [Note that we are using the term risk factor in a broad sense here: that is, a risk factor is an attribute or exposure that is associated with an increased probability of a specified outcome [13], not necessarily a causal factor.] The current study uses general population cohort data to determine whether baseline personality disorder characteristics predict subsequent first onset of panic disorder or agoraphobia over a 13-year period, excluding participants with baseline spontaneous panic attacks or subthreshold panic-like spells from the risk groups. If personality abnormalities are merely epiphenomena of panic/agoraphobia, baseline personality disorder traits should be unrelated to subsequent onset. Because avoidant and dependent traits are strongly related to panic and agoraphobia cross-sectionally, these were the strongest a priori candidates as personality disorder trait risk factors.