Beyond categorical classifications: the importance of identifying posttrauma symptom trajectories and associated negative outcomes.

work in this area heavily relied on taxometric analysis, a statistical methodology specifically designed to test between categorical and dimensional models of latent variables.7 In fact, studies examining the latent structure of PTSS using taxometric analyses have demonstrated that the underlying latent structure of posttraumatic stress is more consistent with a dimensional rather than a categorical model.5,8,9 Since then, Andersen and colleagues10 as well as others (eg, Berntsen et al,11 Dickstein et al12) have examined the latent structure of PTSS responses using different approaches. Specifically, these prospective studies applied a set of analytic techniques (eg, latent class analysis, growth mixture modeling) to identify empirically validated symptom profiles that, once again, challenged the assumption that responses to potentially traumatic events develop in a fairly homogenous pattern across individuals. More importantly, Madsen et al3 extended this work to show how qualitatively distinct symptom trajectories were differentially linked to suicidal ideation. The studies by Andersen et al10 and Madsen et al3 are particularly unique in that they offer a perspective on the categorical/dimensional debate that encourages an integration of both approaches (see also Kessler,13 Kraemer et al14). They showed that dimensional assessments of PTSS taken at different time points could be statistically transformed into 6 discrete classes or groupings that varied from each other, in regard to both their symptom patterns over time and their associations with suicidal ideation. In doing so, they underscored the utility of combining both models and highlighted several important areas for future work. Research aimed at identifying unique risk and protective factors for distinct trajectory groups and related outcomes may offer additional insight into the etiologic underpinnings of different symptom courses and their relation to adverse outcomes, such as suicidal ideation. This includes consideration of other possible contributing factors to suicidal ideation (eg, depression, hopelessness, optimism) as well as relations between these factors and symptom trajectories. Along these lines, researchers may also consider how dimensional and categorical assessment can be used to advance understanding of etiologic factors—including the construct of trauma itself—as well as underlying psychological and neurobiological processes that may impact the onset and course of symptoms. Important questions remain unanswered regarding dimensional assessment of suicidal ideation (eg, intention, lethality, severity).15 Moreover, it would be Submitted: August 18, 2014; accepted August 21, 2014. Corresponding author: Nazanin H. Bahraini, PhD, VISN 19 MIRECC, 1055 Clermont St, Denver, CO 80220 ([email protected]).