The present study compared the utility of categorical (i.e., diagnostic status) and dimensional (i.e., symptom severity) approaches to measuring posttraumatic stress disorder (PTSD) in predicting future suicide attempts among participants in a nationwide, longitudinal study of U.S. military veterans who were deployed in support of operations in Iraq or Afghanistan after the September 11, 2001, terrorist attacks (9/11) and were enrolled in Veterans Health Administration services (N = 1,649). Following an initial assessment of PTSD symptoms, we assessed for suicide attempts at two subsequent time points (M = 28.74 months, SD = 8.72 and M = 55.11 months, SD = 6.89 following the initial assessment). Between the initial and final assessments, 125 participants (7.58%) made at least one suicide attempt. All categorical and dimensional indicators of PTSD predicted suicide attempts at both time points except the categorical indicator for reexperiencing symptoms. Categorical indicators predicting suicide attempts demonstrated excellent sensitivity but poor specificity and overall accuracy. The point along the continuum at which PTSD symptom severity was most accurate regarding the prediction of future suicide attempts was well above the threshold previously established as indicating a probable diagnosis. Although this score was less sensitive than diagnostic indicators, it demonstrated greater specificity and overall accuracy in predicting future suicide attempts. The present results indicate that veterans whose PTSD symptoms satisfy the diagnostic criteria have a higher risk of future suicide attempts, but this risk appears to be even higher for veterans with symptom levels above the diagnostic threshold.
© 2021 Association for Art History.