Panic attacks as risk markers for mental disorders

Andrew J. Baillie*, Ronald M. Rapee

*Corresponding author for this work

Research output: Contribution to journalArticle

57 Citations (Scopus)


Objective: This paper extends previous epidemiological findings linking panic attacks with future episodes of depression and examines whether this relationship is independent of the effects of gender and neuroticism. Methods: Composite International Diagnostic Interview (CIDI) DSM-IV diagnoses from a stratified multi-stage population survey of 10,641 Australian adults were analysed using logistic regression to examine the relationship between lifetime panic attacks, gender, neuroticism and mental disorders. Results: People who experienced full CIDI DSM-IV panic attacks more than 12 months ago were 4 times more likely to meet criteria for current Depressive Disorder than those who reported no attacks. Those with panic attacks in the past 12 months were 13.3 times more likely to report current Depressive Disorders. A similar pattern was also present for non-panic Anxiety Disorders (odds ratio=7.5 for lifetime, but not 12-month panic attacks, and 21.46 for 12-month panic attacks) and for Substance Use Disorders (2.1 and 4.6, respectively) suggesting a broader relationship with psychopathology than previously reported. For each of these groupings of mental disorders, panic attacks accounted for significant variability over and above the effects of gender, neuroticism, and comorbid Anxiety Disorders. Conclusions: Panic attacks are associated with current and future Anxiety, Depressive, and Substance Use Disorders, and this relationship is not solely accounted for by differences in gender and neuroticism.

Original languageEnglish
Pages (from-to)240-244
Number of pages5
JournalSocial Psychiatry and Psychiatric Epidemiology
Issue number3
Publication statusPublished - Mar 2005

Fingerprint Dive into the research topics of 'Panic attacks as risk markers for mental disorders'. Together they form a unique fingerprint.

  • Cite this