Elsevier

Psychiatry Research

Volume 267, September 2018, Pages 455-460
Psychiatry Research

Association of suicidal ideation with trajectories of deployment-related PTSD symptoms

https://doi.org/10.1016/j.psychres.2018.06.034Get rights and content

Highlights

  • Trajectories of PTSD symptoms in Reserve Component soldiers is heterogeneous.

  • Resilience, recovery, late onset and chronic trajectories were identified.

  • Late onset and chronic trajectories were at higher risk for suicidal ideation.

Abstract

This study examined the association between suicidal ideation and posttraumatic stress disorder (PTSD) symptom trajectories in a nationally representative sample of United States Reserve Component soldiers. PTSD symptoms related to a traumatic event during the most recent deployment were assessed in four annual waves in 2010–2013 among 682 Reserve Component soldiers. Latent Growth Mixture Modeling (LGMM) was used to examine the longitudinal trajectories of PTSD symptoms. The association between the PTSD trajectories and suicidal ideation at waves 2 to 4 was examined in logistic regression analyses. Four trajectories were identified: resilience (73.0%), recovery (11.7%), late onset (11.6%) and chronic (3.6%). Pairwise comparisons demonstrated significant differences between trajectories in risk of suicidal ideation. Among the chronic trajectory group, 50.9% reported suicidal ideation (25.8% late onset group; 11.3% recovery group; 4.0% resilience group). After controlling for baseline characteristics, the late onset and chronic trajectory groups were more likely to have suicidal ideation than the resilience and recovery trajectories, respectively. Findings suggest the late onset and chronic trajectories of PTSD symptoms are associated with higher risk of suicidal ideation. They support the importance of follow-up assessment of suicide risk even among individuals with low PTSD symptoms at homecoming.

Graphical abstract

Note. Baseline interviews were conducted between January and July of 2010 and three follow-up annual waves from 2011–2013. The average years since the traumatic event were 3.84 years at Wave 1.

Image, graphical abstract
  1. Download : Download high-res image (129KB)
  2. Download : Download full-size image

Introduction

As in many countries, the Reserve Component of the United States (US) Armed Forces plays an important role in national defense. The US's Reserve Component is comprised of approximately 1.1 million service members, which accounts for 45% of the military's overall strength (NATO, 2004). The Reserve Component has played a pivotal role in recent conflict, with reservists accounting for approximately 40% of Operation Enduring Freedom (OEF) and Operation Iraqi Freedom (OIF) deployed forces (Tanielian and Jaycox, 2008). Compared to the Active Component, Reserve Component personnel may be at a higher risk for posttraumatic stress disorder (PTSD; e.g., Milliken et al., 2007, Schell and Marshall, 2008, Smith et al., 2008, Thomas et al., 2010), which has generally (but not always, e.g., Conner et al., 2014) been found to be associated with higher risk for suicidal ideation in military samples (Ramsawh et al., 2014). Recent work has found heterogeneous trajectories of PTSD symptoms over time following a deployment-related traumatic event among Reserve Component personnel (Fink et al., 2017, Sampson et al., 2015). However, there have been no studies that have assessed the association of PTSD trajectories with suicidal ideation among US military populations. Suicide risk is associated with suicide ideation, plans, and attempts in the military (Nock et al., 2013), as well as in civilian populations (Kessler et al., 2005). Suicide ideation nearly always precedes suicide attempts or completed suicide (Nock et al., 2008a). Therefore suicide ideation is an often examined clinical risk factor for suicide risk (Nock et al., 2008b). As evidence shows an increase in suicide attempts in the US army in the past decade (Schoenbaum et al., 2014, Ursano et al., 2015), examining suicidal ideation by longitudinal development of PTSD symptoms may provide information on suicide prevention in military populations.

Research on the course of PTSD symptoms can be traced to Bonanno's (2004) conceptual model for patterns of disruption in normal functioning following potentially traumatic events or interpersonal loss. Bonanno hypothesized four trajectories: resilience, a trajectory of healthy functioning with minimal impairment; recovery, a trajectory of gradual improvement from an acute reaction; late onset, a trajectory of worsening symptoms; and chronic, a trajectory of consistent dysfunction. Using latent growth mixture modeling (LGMM, Muthén and Muthén, 2000), prototypical trajectories can be empirically estimated. In analyses stratified by single and multiple deployers among US military personnel deployed to Iraq and Afghanistan, similar prototypical patterns of PTSD symptoms were found: 83.1–84.9% of resilience, 8.0–8.5% of recovery, 4.5–6.7% of late onset, and 2.2% of chronic or high-improving trajectories (Bonanno et al., 2012).

Although mixed findings are reported in the number and shape of trajectories in military studies, all have demonstrated heterogeneity in the course of deployment-related PTSD symptoms. For example, three trajectories (resilience, new-onset, and pre-existing) of PTSD symptoms (from pre- to 8 months post-deployment) were identified among the US male Marines deployed to Iraq and Afghanistan (Boasso et al., 2015). Eekhout et al. (2016) assessed PTSD symptoms from pre- to 5 years after deployment among Dutch military personnel and identified three trajectories, i.e., resilience, recovery, and late onset. Recently, Sampson et al. (2015) examined post-deployment PTSD symptoms across four years in Ohio National Guard members. They identified four trajectories which were ordered by increasing severity, including two low- consistent trajectories - resistant (42%) and resilient (42%), a constant mild (12%) and a chronic (5%) trajectory. Fink et al. (2017) found that trajectories for deployment-related and civilian-related traumatic events were similar in a nationally representative sample of Reserve Component personnel, including low-consistent (71% for military trauma and 74% for civilian trauma), borderline-stable (23% and 19%), and chronic (7% and 7%).

There is an apparent gap in the literature as few studies have examined the association between PTSD trajectories and suicidal ideation. Madsen et al. (2014) examined suicidal ideation as an outcome of six derived trajectories of PTSD symptoms from pre- to 2.5 years post-deployment among Danish soldiers deployed to Afghanistan. The trajectories included low-stable (78%), low-fluctuating (8%), late onset (6%), mild distress (4%), and two small sized trajectories characterized as fluctuating over time, i.e., distressed-improving (3%) and relieved-worsening (2%). Compared to the low-stable group, suicidal ideation at the last assessment was higher in the late onset and relieved-worsening trajectories. Their results demonstrated the association of suicidal ideation with PTSD trajectories in a longitudinal perspective and highlighted that suicidal ideation may vary across trajectories. However, there were two noteworthy limitations to the study. First, an index traumatic event was not measured in the first five waves. Without a measure of an index traumatic event, the PTSD symptoms may not be deployment-related. Among those who reported an index traumatic event at time six, about 15% of the traumatic events were not related to their deployment. Therefore, without tracking the index traumatic event, the PTSD symptoms may reflect different traumatic events across time. Second, the relieved-worsening trajectory, one of the two trajectories that were found associated with higher risk for suicidal ideation, contributing only 2% of the sample. Even though this trajectory may represent an interpretable pattern, the small size makes it challenging to conduct analyses such as pairwise comparisons between trajectories.

The current study examines the association between PTSD trajectories and suicidal ideation in a nationally representative sample of Reserve Component of the US Armed Forces. First, PTSD symptoms were assessed over a four-year period and applied a LGMM to identify heterogeneous trajectories. An index traumatic event was measured and tracked to ensure the PTSD symptoms were deployment-related and were linked to the same traumatic event over time. Then, the association between the PTSD trajectories and suicidal ideation were examined. Relevant baseline characteristics were explored and controlled as covariates in the prediction of suicidal ideation by PTSD trajectories.

Section snippets

Sample and procedures

Data were collected from a representative sample of 2,003 Reserve Component personnel serving as of June 2009 using a stratified random sampling procedure (described in detail elsewhere, e.g., Russell et al., 2015, Ursano et al., 2016). The sample was weighted to be representative of the Reserve Component population in the year of data collection with respect to military branch and characteristics including pay grade, age, gender, and race.

Baseline interviews were conducted between January and

Sample and deployment characteristics

Comparison between the 682 participants in the analytic sample and the 179 participants who had missed waves 2 to 4 showed that missingness was not related to gender, race, marital status, or severity of baseline PTSD symptom. However, the group who only participated at baseline was younger (Mean = 33.7, SD = 9.3), had lower level of education (Mean = 13.8, SD = 1.9), and consisted of more enlisted soldiers (84.8%) relative to those in the analytic sample.

Sample characteristics are reported in

Discussion

Prior evidence supports the association of PTSD with suicidal ideation, but few studies have examined the relationship between suicidal ideation and PTSD trajectories. The current study identified differences in suicidal ideation across different trajectories. Controlling for baseline characteristics, elevated risk for suicidal ideation was associated with membership in the late onset and chronic trajectories compared to the resilience and recovery trajectories.

PTSD symptoms were assessed from

Acknowledgments

This research was funded by Department of Defense grants W81XWH-08-2-0204, W81XWH-08-2-0650, MH 082729; and National Institute of Mental Health grant 5R01MH082729-05.

The opinions expressed in the manuscript are those of the authors and, therefore, do not necessarily reflect the views of the Department of Defense, the Uniformed Services University of the Health Sciences, or the Center for the Study of Traumatic Stress.

Conflicts of interest

The authors declare that they do not have any conflicts of interest.

References (40)

  • S. Clark et al.

    Relating latent class analysis results to variables not included in the analysis

  • DicksteinB.D. et al.

    Heterogeneity in the course of posttraumatic stress disorder: trajectories of symptomatology

    J. Trauma. Stress

    (2010)
  • D.S. Fink et al.

    Trajectories of posttraumatic stress symptoms after civilian or deployment traumatic event experiences

    Psychol. Trauma.

    (2017)
  • M.J. Gray et al.

    Psychometric properties of the life events checklist

    Assessment

    (2004)
  • R.C. Kessler et al.

    Trends in suicide ideation, plans, gestures, and attempts in the United States, 1990–1992 to 2001–2003

    JAMA

    (2005)
  • L.A. King et al.

    Deployment risk and resilience inventory: a collection of measures for studying deployment-related experiences of military personnel and veterans

    Mil. Psychol.

    (2006)
  • K. Kroenke et al.

    The PHQ-9: validity of a brief depression severity measure

    J. Gen. Intern. Med.

    (2001)
  • Y. Lo et al.

    Testing the number of components in a normal mixture

    Biometrika

    (2001)
  • T. Madsen et al.

    Postdeployment suicidal ideations and trajectories of posttraumatic stress disorder in Danish soldiers: a 3-year follow-up of the USPER study

    J. Clin. Psychiatry

    (2014)
  • C.S. Milliken et al.

    Longitudinal assessment of mental health problems among active and reserve component soldiers returning from the Iraq war

    JAMA

    (2007)
  • Cited by (10)

    • The evolution of PTSD symptoms in serving and ex-serving personnel of the UK armed forces from 2004 to 16: A longitudinal examination

      2023, Journal of Psychiatric Research
      Citation Excerpt :

      By identifying clinical phenotypes that may not be directly observable, LGVM methods can help identify distinct patterns of change over time and identify subgroups of individuals that could not be discovered based on cross-sectional data or basic descriptive analysis alone (Mori et al., 2020). Military research has followed suit with PTSD trajectory analyses published in the US (Boasso et al., 2015; Bonanno et al., 2012; Dickstein et al., 2010; Donoho et al., 2017; Fink et al., 2017; Nash et al., 2015; Orcutt et al., 2014; Polusny et al., 2017; Porter et al., 2017; Sampson et al., 2015; Schultebraucks et al., 2021; Wang et al., 2018), Denmark (Andersen et al., 2014), the Netherlands (van der Wal et al., 2019), Israel (Karstoft et al., 2013; Solomon et al., 2021), Australia (Dell et al., 2022) and the UK AF (Palmer et al., 2019). Military studies employing trajectory analyses confirm that most personnel do not meet thresholds of probable PTSD; yet, when reviewing the classes each study identified, worsening symptom classes were more frequently occurring across the trajectory models compared to improving symptom classes.

    • Associations between PTSD symptoms and suicide risk: A comparison of 4-factor and 7-factor models

      2020, Journal of Psychiatric Research
      Citation Excerpt :

      As previous studies have revealed differential associations between more refined dimensional models of PTSD and functional outcomes (Tsai et al., 2012) and neurobiological markers (Pietrzak et al., 2013), identification of more refined PTSD symptom clusters and their relationships with suicide risk have both clinical and etiological significance. Moreover, most studies focused on specific occupations, such as the military population (Wang et al., 2018), veterans (Elbogen et al., 2018), or firefighters (Boffa et al., 2017). To the best of our knowledge, no study has investigated the associations between PTSD symptoms and suicide risk using a national survey database.

    View all citing articles on Scopus
    View full text