The Impact of Changing Diagnostic Criteria in Posttraumatic Stress Disorder in a Canadian Epidemiologic Sample

Background: Since its inclusion in DSM-III, posttraumatic stress disorder (PTSD) has undergone a number of changes in its diagnostic criteria, including the expansion of Criterion A (traumatic stressor), the addition of symptom duration (none specified in DSM-III), and the requirement for impairment or distress (Criterion F, DSM-IV only).

Method: This study examined the impact of changes in PTSD diagnostic criteria using a Canadian PTSD epidemiologic sample. The rates of PTSD and its correlates were evaluated in a nationally representative random sample of 3,006 adults. DSM-III, DSM-III-R, DSM-IV, and ICD-10 criteria were employed. DSM-III, DSM-III-R, and ICD-10 rates were re-evaluated, substituting specific DSM-IV criteria (A-F).

Results: The prevalence rates of lifetime PTSD ranged from 13.4% (DSM-III) to 13.0% (ICD-10) to 12.2% (DSM-III-R) to 9.2% (DSM-IV); all rates differed significantly from each other (P < .001). Regardless of diagnostic criteria, most people reported more than 1-year duration of symptoms, although rates were significantly higher in those with DSM-IV PTSD (68.2%, P < .0001). Rates of comorbid major depressive disorder and alcohol and substance abuse and dependence were also significantly higher (P < .05) using the DSM-IV PTSD criteria, and those with DSM-IV PTSD reported significantly higher rates of help-seeking (P < .001). When Criterion F was added to earlier versions, lifetime PTSD rates became much closer to those obtained using DSM-IV criteria: 10.6% (DSM-III), 10.2% (DSM-III-R), and 9.9% (ICD-10); however, rates fluctuated when operational definitions of Criterion F were modified. DSM-III PTSD was also substantially affected by DSM-IV Criteria A and C.

Conclusions: DSM-IV PTSD may identify a more severe disorder. The addition of the clinical significance criterion (F) appeared to affect the greatest change in prevalence rates. Defining Criterion F as having both clinically significant psychological distress and functional impairment lowered the diagnostic threshold to a greater degree than did either distress or impairment alone. This information may be useful for future revisions of PTSD diagnostic criteria.

J Clin Psychiatry 2011;72(8):1034-1041

Submitted: September 15, 2009; accepted February 2, 2010.

Online ahead of print: May 31, 2011 (doi:10.4088/JCP.09m05700).

Corresponding author: Michael Van Ameringen, MD, FRCPC, Department of Psychiatry and Behavioural Neurosciences, McMaster University, 439-1 Fontbonne Building, St Joseph’s Healthcare Hamilton, 301 James St S, Hamilton, Ontario, Canada L8P 3B5 (vanamer@mcmaster.ca).