(PS13-71) A Latent Class Analysis of Emotional Responses After Trauma Predicting Suicidal Ideation and Behavior One Year Later

Posttraumatic stress disorder (PTSD) is a concern within the military given the prevalence rates among military personnel and the association between PTSD and suicide. Negative affect may be a key factor impacting the relationship between PTSD and suicide given that it plays an etiological role in PTSD development and emotion dysregulation has been identified as a risk factor for suicide.

Latent analyses (LCA) were used to identify subgroups of cognitive-affective states among previously deployed military (n = 626) with probable PTSD. Participants were patients recruited from six military primary care clinics across the US. Participants were eligible for the parent study if they were (1) at least 18 years old, (2) eligible for healthcare services from the Military Health System, and (3) able to complete informed consent. Participants were included in this study’s analyses if they (a) reported exposure to a potentially traumatic event on the Life Events Checklist and (b) screened positive for PTSD on the Primary Care PTSD Screen for DSM-IV. Using these criteria, 626 of the 2744 enrolled patients (22.4%) were included in the present analyses.

A LCA was conducted to identify distinct patterns of emotional reactivity using observed indicators (guilt, anger, and anxiety/fear) to identify homogeneous patterns (i.e., “classes”). Chi-square were conducted to examine group differences between classes regarding suicidal ideation and behavior one year after baseline.

Results indicated a 4-class model was the best fit for the data. Classification probabilities for the most likely class membership were 0.84 - 0.92, indicating acceptable separation of classes. Differences in emotional reactivity after trauma served as the basis for developing descriptive labels: Class 1 was labeled “low symptomatology” due to the lowest scores across indicators, Class 2 was labeled “anxiety, fear, and guilt” due to low anger scores, Class 3 was labeled “guilt and anger” due to low anxiety/fear reaction, and Class 4 was labeled “anxiety/fear, guilt, and anger” due to elevations on all indicator variables.

Chi-square results indicated significant differences between the classes on suicidal ideation χ² (3, N = 626,  p < .001) and suicidal behavior χ² (3, N = 626,  p < .001). Results were further examined through odds ratios and comparing classes to the “anxiety/fear, guilt, and anger” class. Regarding suicidal ideation, participants in the “anxiety/fear, guilt, and anger” class were at 1.95 increased odds compared to the “low symptomatology” class, 1.41 increased odds compared to than the “anxiety, fear, and guilt” class, and 5.27 increased odds compared to the “guilt and anger” class.  Regarding suicidal behavior, participants in the “anxiety/fear, guilt, and anger” class were at 2.31 increased odds compared to the “low symptomatology” class, 1.08 increased odds compared to than the “anxiety, fear, and guilt” class, and 8.63 increased odds compared to the “guilt and anger” class.

Results indicate that there are distinct profiles of emotional reactivity within PTSD and that these meaningful profiles are independently associated with future suicidal ideation and behavior. Further, the combination of guilt and anxiety/fear emotions may be a unique risk factor for suicide.