Examining the relation between posttraumatic stress disorder and suicidal ideation in an OEF/OIF veteran sample

Abstract

This study examined the relation between posttraumatic stress disorder (PTSD) and suicidal ideation among U.S. military veterans deployed during Operation Enduring Freedom and/or Operation Iraqi Freedom. Specific aims included investigation of (1) whether PTSD was associated with suicidal ideation after controlling for combat exposure and history of suicide attempt(s), (2) whether PTSD was associated with suicidal ideation absent a co-occurring depressive disorder (MDD) or alcohol use disorder (AUD), (3) whether co-occurring MDD or AUD increased risk of suicidal ideation among those with PTSD and (4) whether PTSD/MDD symptom clusters were differentially associated with suicidal ideation. Results pointed to unique effects associated with prior suicide attempt(s), PTSD and MDD. PTSD-diagnosed participants with co-occurring MDD or AUD were not significantly more likely to endorse suicidal ideation than PTSD-diagnosed participants without such comorbidity. The ‘emotional numbing’ cluster of PTSD symptoms and the ‘cognitive-affective’ cluster of MDD symptoms were uniquely associated with suicidal ideation.

Introduction

A report from the Institute of Medicine (2008) concluded that there is compelling empirical evidence of a link between deployment to a war zone and an elevation of suicide risk subsequent to that deployment. It is likely that this link is largely mediated by compromised emotional-behavioral health, as suicide has long been understood to occur most often in the context of psychiatric illness (Barraclough et al., 1974, Fontana and Rosenheck, 1995, Henriksson et al., 1993, Shafii et al., 1985). Exposure to wartime stressors may precipitate a range of aversive psychological sequelae, prominent among which are symptoms of posttraumatic stress disorder (Hoge et al., 2004, Schlenger et al., 1992). Estimates of the prevalence of posttraumatic stress disorder (PTSD) in combat veterans range considerably. With respect to deployments associated with Operations Enduring Freedom and Iraqi Freedom (OEF/OIF), surveys have found that 11–30% of returning veterans meet screening criteria for PTSD (Hoge et al., 2004, Lapierre et al., 2007, Milliken et al., 2007).

Previous studies have found PTSD to be significantly associated with elevated suicide ideation and suicide attempts (e.g., Davidson et al., 1991, Jakupcak et al., 2009, Marshall et al., 2001, Sareen et al., 2007). A number of questions remains regarding the relation between PTSD and suicidality, however, several of which were tackled by the study presented here. The specific aims of this investigation, conducted with a sample of OEF/OIF veterans, were to consider (a) whether PTSD is associated with suicidal ideation after first accounting for veterans’ prior suicide attempts and combat history, (b) whether PTSD is associated with suicidal ideation absent the diagnosis of a co-occurring major depressive disorder (MDD) or alcohol use disorder (AUD), (c) whether PTSD-diagnosed participants with co-occurring MDD or AUD are more likely to report suicidal ideation than PTSD-diagnosed participants who do not manifest such comorbidity and (d) whether, among PTSD-diagnosed participants, there are distinct clusters of symptoms differentially associated with suicidal ideation.

PTSD commonly co-occurs with other psychiatric illnesses. Epidemiological findings suggest that the DSM-IV Axis I psychiatric diagnoses most frequently comorbid with PTSD are MDD and AUD (Keane and Wolfe, 1990, Kessler et al., 1995, Stewart, 1996). Like PTSD, both MDD and AUD have been found to be robust predictors of elevated suicidality (Borges et al., 2000, Henriksson et al., 1993, Liu et al., 2009, Yen et al., 2003). Evidence that PTSD regularly co-occurs with disorders that are themselves associated with suicidality raises two questions. First, does PTSD predict increased suicidality absent a comorbid diagnosis of MDD or AUD? Second, are persons diagnosed with PTSD and a co-occurring MDD or AUD at greater risk of suicidality than PTSD-diagnosed persons who do not concurrently struggle with MDD and/or AUD?

There are limited data that speak directly to the question of whether suicidality is heightened among PTSD-diagnosed persons who do not manifest other clinical risk factors (e.g., MDD/AUD). Instead, investigators interested in the link between PTSD and suicidality have typically employed statistical approaches that partial out effects associated with other mental health disorders, such that a relation between PTSD and suicidality may be said to exist after “controlling”/“adjusting” for variance in suicidality that may be explained by other disorders included in the analysis (e.g., Marshall et al., 2001, Sareen et al., 2005). With respect to the second question, which asks whether there is an incremental risk of suicidality associated with comorbidity among those diagnosed with PTSD, more data are available. Across four studies (Ferrada-Noli et al., 1998, Jakupcak et al., 2009, Shalev et al., 1998, Tarrier and Gregg, 2004), participants diagnosed with both PTSD and MDD did not endorse significantly greater suicidal ideation than participants with PTSD only, and in one study (Jakupcak et al., 2009), participants diagnosed with both PTSD and an AUD (alcohol abuse) did not endorse significantly greater suicidal ideation than participants with PTSD only.²

Among persons diagnosed with PTSD, frequency and intensity of specific symptoms of PTSD vary markedly. Similarly, although it is common for PTSD-diagnosed persons to endorse symptoms of MDD, there is considerable variability in the type and severity of depressive symptomatology experienced. Taxometric work points to distinct clusters of both PTSD and MDD symptoms; the present report considers whether these symptom clusters are differentially associated with suicidal ideation.

With respect to PTSD, factor analyses most often suggest four symptom clusters, which describe, in turn, symptoms related to (1) the re-experiencing of the trauma, (2) the experiential and behavioral avoidance of trauma-related stimuli, (3) a general numbing of responsiveness and (4) hyperarousal (Asmundson et al., 2000, Cox et al., 2008, King et al., 1998, McDonald et al., 2008). Factor analyses of depressive symptomatology suggest that symptoms of MDD cluster into cognitive-affective and somatic-vegetative symptoms, respectively (Beck et al., 1996a, Dozois et al., 1998, Steer et al., 1999).

Several studies have examined the relation of PTSD or MDD symptom clusters (or specific symptoms) to suicidality among persons diagnosed with PTSD, with mixed results. Hendin and Haas (1991) reported that the experience of “persistent guilt” contributed uniquely to the prediction of both suicidal ideation and suicide attempts in a sample of PTSD-diagnosed Vietnam combat veterans; no specific symptoms of PTSD (per DSM-III-R criteria) were uniquely predictive of suicidality. Bell and Nye (2007), also studying PTSD in Vietnam combat veterans, focused exclusively on the relative import of the three PTSD symptom clusters delineated in DSM-IV, and found that only re-experiencing symptoms were uniquely, positively predictive of suicidal ideation. Finally, Tarrier and Gregg (2004) reported evidence of univariate relations between suicidality and the PTSD symptom clusters of re-experiencing and hyperarousal in a sample of PTSD-diagnosed civilians, but neither of these symptom clusters was uniquely associated with suicidality in multivariate analyses.

In order to address the questions raised above, the present study examined (1) the relation of PTSD to suicidal ideation, with specific attention to the import of psychiatric comorbidity and (2) the relation of PTSD/MDD symptom clusters to suicidal ideation among PTSD-diagnosed participants, with specific attention to those symptom clusters identified robustly across recent taxometric work. Note that determination of participants’ status vis-à-vis each psychiatric disorder of focus (i.e., PTSD; MDD; AUD) was made using a psychometrically well-established structured clinical interview that was developed consistent with DSM-IV-TR criteria; many of the studies referenced above determined participants’ diagnostic status using questionnaire data (e.g., Jakupcak et al., 2009, Marshall et al., 2001) or DSM-III-R diagnostic criteria (e.g., Hendin and Haas, 1991, Shalev et al., 1998, Tarrier and Gregg, 2004).

It was hypothesized that PTSD would be associated with increased risk of suicidal ideation after accounting for participants’ history of combat exposure and history of suicide attempts. It was hypothesized that PTSD would be associated with increased risk of suicidal ideation even absent a co-occurring MDD or AUD diagnosis. Moreover, consistent with the extant, albeit limited, research referenced above, dually-diagnosed participants were not expected to be at increased risk of suicidal ideation relative to participants diagnosed only with PTSD. With respect to questions regarding the relative strength of relations between symptom clusters and suicidal ideation among PTSD-diagnosed participants, no a priori hypotheses were generated.