Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Feb 1.
Published in final edited form as: Psychol Serv. 2017 Mar 13;15(1):31–39. doi: 10.1037/ser0000127

Differences in Suicide and Death Ideation Among Veterans and Non-Veterans with Serious Mental Illness

Danielle R Jahn 1,2, Anjana Muralidharan 1,2, Amy L Drapalski 1,2, Clayton H Brown 1,2,3, Li Juan Fang 1,2, Alicia Lucksted 1,2
PMCID: PMC5597449  NIHMSID: NIHMS858225  PMID: 28287770

Abstract

Individuals with serious mental illness and veterans are two populations at elevated risk for suicide; however, research has not examined whether veterans with serious mental illness may be at higher suicide risk than non-veterans with serious mental illness. Additionally, overlapping risk factors for suicide in these populations may account for differences in suicide-related outcomes between these groups. Therefore, the aim of this study was to identify differences in death ideation and suicide ideation among veterans and non-veterans with serious mental illness. We also aimed to explore these effects after adjusting for potentially shared risk factors. We found that veterans with serious mental illness reported death ideation and suicide ideation more than twice as often as non-veterans with serious mental illness. After adjusting for demographic, psychiatric, and theory-driven risk factors, the effect of veteran status on death ideation remained significant, though the effect on suicide ideation was no longer significant. Depressive and psychotic symptoms were significant predictors of death ideation; depressive symptoms and hostility were significant predictors of suicide ideation. Clinicians should particularly monitor death ideation and suicide ideation in veterans with serious mental illness, as well as associated clinical risk factors such as depression, psychotic symptoms, and hostility.

Keywords: Risk factors, Psychiatric symptoms, Interpersonal theory of suicide, Schizophrenia, Bipolar disorder

Suicide rates continue to rise in the United States (Centers for Disease Control and Prevention, 2016), suggesting that more work is needed to effectively identify at-risk individuals and reduce suicide risk. Mental health professionals are regularly required to assess and manage suicide risk in a variety of populations, but research is lacking regarding differences in risk levels and risk factors in certain populations. Research regarding these topics may support clinicians in making decisions regarding suicide risk. Therefore, the aim of this study is to provide information regarding suicide risk levels and important risk factors in high-risk groups, to inform mental health professionals’ decisions regarding levels of risk and appropriate management of risk.

Research shows that suicide risk is differentially distributed in the population (American Association of Suicidology, 2015). This means that there are subpopulations in which suicide risk is quite elevated, whereas other subpopulations are at substantially lower risk. Individuals with serious mental illness (SMI; e.g., schizophrenia, schizoaffective disorder, bipolar disorder, major depressive disorder) have significantly higher suicide rates than the general population (e.g., Angst, Angst, & Stassen, 1999; Hawton, Sutton, Haw, Sinclair, & Harriss, 2005; Hor & Taylor, 2010). Therefore, mental health services for individuals with SMI should routinely address suicide risk. There are numerous risk factors that may drive the elevated suicide rate among individuals with SMI, including age (either being younger [particularly for schizophrenia and bipolar disorder] or older), gender (male), depressive symptoms, positive psychotic symptoms, and anxiety or agitation (e.g., Angst et al., 1999; Hawton et al., 2005; Hor & Taylor, 2010).

Similarly, veterans are at elevated suicide risk compared to the general population (VA Office of Suicide Prevention, 2016). Research to date has examined risk factors for suicide among military service members and veterans, and has found that mental health diagnoses are risk factors for suicide (Nock et al., 2014). Clinical guidelines for assessing and managing at-risk individuals have been published by the Department of Veterans Affairs (VA) and Department of Defense (DoD; 2013). These guidelines indicate that important risk factors for suicide among veterans and military service members include age (either older or younger), gender (male), marital status (divorced/widowed/separated), somatic symptoms, impulsivity (e.g., angry outbursts), and symptoms or presence of mental health diagnoses (i.e., mood disorders, anxiety, schizophrenia). These guidelines also note that a sense of belonging and good self esteem, among other variables, are protective against suicide risk in this population (VA & DoD, 2013).

These protective factors align with the interpersonal theory of suicide (Joiner, 2005; Van Orden et al., 2010). This theory states that thwarted belongingness and perceived burdensomeness create the desire for suicide (Joiner, 2005). Thwarted belongingness includes lack of reciprocal caring relationships in which one feels that he or she belongs; perceived burdensomeness is the belief that one is a liability or burden to others and is indicated by low self-esteem and feeling ineffective. The importance of perceived burdensomeness and thwarted belongingness as risk factors for suicide ideation has been found in multiple samples, including older adults and veterans on an inpatient psychiatric unit (e.g., Jahn, Poindexter, & Cukrowicz, 2015; Monteith, Menefee, Pettit, Leopoulos, & Vincent, 2013).

Collectively, these findings indicate that individuals with SMI and veterans are at high risk for suicide-related outcomes (e.g., Angst et al., 1999; Hawton et al., 2005; VA Office of Suicide Prevention, 2016). Additionally, many of the known risk factors for suicide (e.g., age, gender) in these two populations overlap. Furthermore, veterans’ demographic characteristics are different from the United States population (i.e., the veteran population includes a greater proportion of males than the national average and is older than the national average; National Center for Veterans Analysis and Statistics, 2016). Therefore, veterans with SMI may be at particularly elevated suicide risk because of these differences and the potentially additive effects of veteran status and SMI diagnosis. In fact, veterans with diagnoses of schizophrenia, bipolar disorder, or depressive disorders are at significantly higher risk for suicide than veterans without a psychiatric diagnosis (Ilgen et al., 2010); however, research has not compared veterans and non-veterans with SMI in terms of suicide risk. Additionally, no studies have examined potential risk and protective factors (e.g., psychiatric symptoms, demographics, belonging) that may account for differences in suicide-related outcomes between veterans and non-veterans with SMI. Identifying potential explanatory variables is critical for providers of mental health services, as these variables may be important to identify and address to reduce suicide rates.

Therefore, the purpose of this study was to assess differences in suicide ideation and death ideation between veterans and non-veterans with SMI. Suicide ideation is defined as thoughts of wanting to kill one’s self or die by suicide, whereas death ideation is defined as thoughts of death or wanting to die without explicit thoughts of suicide. Research has shown that these constructs are separate, though related, and death ideation may be up to twice as prevalent as suicide ideation (Scocco & De Leo, 2002). We selected suicide ideation as an outcome because it has been consistently linked to death by suicide (e.g., Hor & Taylor, 2010; Wenzel et al., 2011). Additionally, rates of suicide ideation are elevated in both veterans (Smith et al., 2016) and individuals with SMI (e.g., Hor & Taylor, 2010). We chose to also include death ideation as an outcome because research indicates that death ideation is as important as suicide ideation in terms of risk for negative outcomes, such as death by suicide, suicide attempts, and severity of suicide ideation. Szanto et al. (2002) found that individuals with death ideation did not differ significantly from those with suicide ideation in terms of previous suicide attempts, severity of previous suicide ideation, hopelessness, or depressive symptoms. Additionally, both suicide ideation and death ideation have been documented at similar rates in medical records of individuals who died by suicide, upon case review (L. Berman, personal communication, April 28, 2016).

Additionally, we sought to examine whether risk and protective factors accounted for any identified differences in suicide ideation and death ideation. The differences in demographic characteristics between veterans and non-veterans, as well as the overlapping risk factors in these groups, suggest that identified differences by veteran status may be due to other risk or protective factors. Previous research has found strong support for the interpersonal theory of suicide and for depressive and anxious symptoms as risk factors for suicide ideation in both veterans and individuals with SMI. Thus, we expected that depressive symptoms, anxiety, and indicators of thwarted belongingness and perceived burdensomeness would be significant predictors of suicide-related outcomes. We hypothesized that veterans with SMI would report greater suicide ideation and death ideation than non-veterans with SMI. We then conducted exploratory analyses to assess the influence of veteran status on these outcomes after adjusting for other known risk factors (i.e., demographic characteristics, psychiatric symptoms, and indicators of thwarted belongingness and perceived burdensomeness). We anticipated that depressive symptoms, anxiety, and indicators of thwarted belongingness and perceived burdensomeness would be significant predictors of each outcome. We did not generate specific a priori hypotheses regarding the significance of veteran status after adjusting for other risk factors due to a lack of literature on which to base hypotheses.

Methods

Participants

Baseline assessment data from two concurrent randomized controlled trials (RCTs) of BLINDED were utilized in the present study (CITATION BLINDED FOR REVIEW). BLINDED is a psychoeducational intervention which aims to BLINDED among adults with serious mental illnesses (SMI). Baseline data was collected prior to participants’ randomization or assignment to study groups, so the intervention did not bias data collection or responses. Conclusions about causality among the variables of interest cannot be determined because of the cross-sectional nature of the data. Participants (N = 516) were adults with SMI recruited from five community-based psychosocial rehabilitation programs, and mental health programs in three large Veterans Affairs (VA) medical centers. Participants were majority male (73.3%), African-American (51.9%) or Caucasian (39.3%), unmarried (89.1%), and unemployed (92.6%). The mean age was 48.9 years (SD = 11.7) and most participants had at least a high school education (80.4%). Approximately half of the sample was living in a supervised living facility (45.8%). Psychiatric diagnoses included the following: schizophrenia (28.1%), schizoaffective disorder (23.7%), bipolar disorder (including bipolar I disorder and bipolar II disorder; 13.2%), major depressive disorder with psychotic features (6.5%), major depressive disorder (other than with psychotic features; 4.6%), and other psychotic disorders (e.g., unspecified schizophrenia spectrum or other psychotic disorder, 2.2%). A little over half of the participants were veterans (50.6%), which included 247 veterans recruited from the VA sites, and 14 veterans recruited from the community sites.1

Recruitment and Consent Procedures

Participants were recruited through clinician referrals, recruitment flyers posted in participating clinics, and verbal invitation at program community meetings. At the VA sites, participants were also recruited from review of clinic and program rosters; a partial HIPAA waiver was obtained to allow review of charts to confirm eligibility. Potentially eligible participants identified through chart review were approached in-person at appointments or sent letters regarding the study. A trained research assistant screened all interested individuals. Eligible individuals for both studies included consenting consumers between 18 and 80 (for VA) or 90 (for community) years of age diagnosed with SMI, who were willing and able to participate in all aspects of the study. Though the upper age limits differed due to expectations of different funding sources, the oldest participant enrolled at either site was age 76. At the VA sites, eligibility criteria also included a chart diagnosis of schizophrenia, schizoaffective disorder, bipolar I or II disorder, or major depressive disorder with psychotic features. In the community study individual chart diagnosis was not used for eligibility. Rather, the involved programs serve adults meeting BLINDED’s “severely mentally ill priority population” definition, which requires a diagnosis of schizophrenia, bipolar disorder, recurrent major depressive disorder, schizotypal or borderline personality disorder, or another delusional or psychotic disorder, with documented functional impairments (CITATION BLINDED). Exclusion criteria at both sites included a documented history of severe or profound intellectual developmental disorder. Eligible participants provided written informed consent, preceded by a brief assessment to verify study comprehension. All study procedures for the two concurrent RCTs were approved by the appropriate institutional review boards. Procedures to combine baseline data from the two RCTs were approved by the BLINDED FOR REVIEW Institutional Review Board.

Measures

Demographic and Clinical Characteristics

Demographic information, including age (in years), sex (1 = male, 0 = female), veteran status (1 = veteran, 0 = non-veteran), and relationship status (1 = Currently married, 0 = Not married), was collected from all participants via interview. Additionally, information regarding education, race/ethnicity, living situation, and employment status, was also collected. Mental health diagnosis was recorded from the participant’s clinical record. In this study, age, gender, and marital status were used as covariates in the exploratory analyses, and veteran status was used as the predictor. We grouped all SMI diagnoses together in our study to better power analyses, though we were then unable to detect potential differences in the analyses based on diagnosis.

Brief Symptom Inventory (BSI)

The BSI is a multidimensional symptom inventory derived from the Symptom Checklist-90-Revised (Derogatis, 1993). The BSI is comprised of a list of 53 symptoms; participants were asked to rate how distressed they were by each symptom over the past 7 days using a 5-point scale ranging from “Not at All” to “Extremely.”, BSI items load onto nine symptom dimensions (Somatization, Obsession-compulsion, Interpersonal sensitivity, Depression, Anxiety, Hostility, Phobic anxiety, Paranoid ideation, and Psychoticism); raw means for each of these symptom dimensions were calculated. In the calculation of the Depression subscale score, the suicide ideation item was removed. Each of these subscales was then used as a covariate in the exploratory analyses. At initial validation, each subscale evidenced strong internal consistency reliability and two-week test-retest reliability (Derogatis & Melisaratos, 1983). Subscales also had good convergent and divergent validity based on correlations with other established measures (Derogatis & Melisaratos, 1983). In the current sample, Cronbach’s alphas were 0.80 for the somatization subscale, 0.85 for the obsession-compulsion subscale, 0.71 for the interpersonal sensitivity subscale, 0.86 for the depression subscale (with the suicide ideation item removed), 0.84 for the anxiety subscale, 0.73 for the hostility subscale, 0.80 for the phobic anxiety subscale, 0.71 for the paranoid ideation subscale, and 0.70 for the psychoticism subscale.

In the present study, BSI item 39 (“Thoughts of death or dying”) was used as a measure of death ideation. BSI item 9 (“Thoughts of ending your life”) was used as a measure of suicide ideation. Both were included as dependent variables. Responses on these items were dichotomized into no ideation versus any ideation (i.e., “Not at all” versus any endorsement of the symptom). Though the use of single item assessments of suicide ideation and death ideation is a limitation of the study, as is the dichotomizing of responses, many other studies have used single item measures of suicide ideation and found such assessment to be valid. Single items from larger psychiatric symptom inventories have been significantly associated with comprehensive suicide ideation assessments (Deseilles et al., 2012), as well as number of previous suicide attempts (Deseilles et al., 2012) and death by suicide (Brown, Beck, Steer, & Grisham, 2000). Of additional note, it is also possible that participants may have interpreted these single items in different ways; for example, some respondents may have endorsed thoughts of death or dying that related to developmentally appropriate preparations for death with age or diagnosis of a chronic health condition, as opposed to an immediate desire for death. Despite these limitations, single item measures of suicide and death ideation appear to be a valid assessment of these two suicide-related outcomes.

General Self-Efficacy Scale (GSES)

The GSES measures the degree to which the respondent perceives him or herself as capable of attaining goals, overcoming challenges, and performing well on tasks. Respondents rate 8 items on a 5-point response scale; responses are averaged to produce an overall score. The GSES has strong psychometric properties, performing favorably when compared to other measures of the same construct (Chen, Gully, & Eden, 2001; Sherbaum, Cohen-Charash, & Kern, 2006). Cronbach’s alpha was 0.89 in this study, indicating good internal consistency reliability. Self-efficacy was used as an indicator of perceived burdensomeness (a covariate in the exploratory analyses), as Joiner (2005) posited that perceived burdensomeness was the opposite of feeling effective. Though this measure has not been validated in conjunction with the Interpersonal Needs Questionnaire (INQ; Van Orden, Cukrowicz, Witte, & Joiner, 2012), the measure traditionally used to assess perceived burdensomeness, we selected it for use based on its theoretical link to perceived burdensomeness (Joiner, 2005).

Self-Esteem Rating Scale–Short Form (SERS)

The SERS is a self-report scale that measures self-esteem, with high reliability and evidence of construct validity (Nugent, 2004; Nugent & Thomas, 1993). In the present study, we used the short form of the SERS, in which respondents must rate 20 statements regarding how they feel about themselves (e.g., “I feel that I am a very competent person”) on a 7-point response scale. Responses are summed to produce an overall score; higher scores are indicative of higher self-esteem. Internal consistency reliability was strong in this sample (Cronbach’s α = 0.92). Self-esteem was entered as a covariate in the exploratory analyses and was also used as an indicator of perceived burdensomeness, as Van Orden et al. (2010) noted that low self-esteem was an indicator of the self-hatred component of perceived burdensomeness. Again, this measure has not been examined in relation to the INQ perceived burdensomeness subscale (Van Orden et al., 2012), but was selected based on its relationship with perceived burdensomeness posited by Van Orden et al. (2010).

Sense of Belonging Instrument (SOBI)

The SOBI is a 32-item measure with a 4-point response scale. In the present study, we used the 18-item SOBI-P subscale, which measures the psychological experience of belonging with high construct validity and test-retest reliability (Hagerty & Patusky, 1995; Sargent, Williams, Hagerty, Lynch-Sauer, & Hoyle, 2002). SOBI-P scores represented thwarted belongingness, with higher scores indicative of less thwarted belongingness (i.e., greater belonging). Internal consistency reliability, calculated through Cronbach’s alpha, was 0.93. Scores were entered as a covariate in the exploratory analyses. The SOBI-P has not been correlated with the INQ thwarted belongingness subscale but was used as a measure of thwarted belongingness because it assesses the psychological sense of belonging, consistent with Joiner’s (2005) description of thwarted belongingness.

Data Analysis

Two logistic regression analyses were performed to test the primary hypothesis in the present study. In the first analysis, death ideation was the outcome variable; in the second analysis, suicide ideation was the outcome variable. veteran status was entered as the predictor in both analyses. We then conducted exploratory analyses using two series of logistic regressions to examine the influence of veteran status after adjusting for other risk factors. The outcome variables for each series were the same as for the primary hypothesis tests. We then entered sets of predictor variables into the model in the same order for both series of regressions. veteran status was entered as the first predictor, followed by a set of demographic predictors (age, gender, marital status). The third set of variables entered was the nine BSI subscales. Finally, GSES, SERS, and SOBI-P scores were entered in the last step, as indicators of the constructs in the interpersonal theory of suicide. All continuous predictors, with the exception of age, were standardized prior to entry into the model; age was centered at the mean. We did not formally test mediation in this study to identify which variables may mediate relations between veteran status and suicide-related outcomes, as we did not have a specific directional hypothesis regarding potential mediators; such an analysis may be a direction for future research.

Results

Zero-order correlations between variables of interest are presented in Table 1. Descriptive statistics for all variables are displayed for the total sample and by group (veterans versus non-veterans) in Table 2. In our sample, 14.3% of veterans with SMI and 6.7% of non-veterans with SMI endorsed suicide ideation. Additionally, 38.8% of veterans with SMI and 18.0% of non-veterans with SMI reported death ideation.

Table 1.

Correlations and Descriptive Statistics

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Veteran —
2. Gender .32*** —
3. Marital Stat. .19*** .01 —
4. Age .39*** .18*** .11** —
5. Somatiz. .09** −.04 .14** .02 —
6. Obs. Comp. .14** −.01 .12** −.04 .61*** —
7. Interperson. −.02 −.04 .02 −.06 .44*** .64*** —
8. Depression .17*** .00 .06 −.04 .50*** .70*** .69*** —
9. Anxiety .15*** −.02 .12** −.02 .59*** .74*** .65*** .73*** —
10. Hostility .10* −.05 .09 −.13** .50*** .61*** .55*** .58*** .62*** —
11. Phobia .09* −.05 .11** .00 .46*** .61*** .61*** .58*** .74*** .51*** —
12. Paranoia .01 −.03 .11* −.04 .47*** .63*** .65*** .59*** .64*** .57*** .54*** —
13. Psychotic. .14** .03 .03 −.03 .49*** .72*** .67*** .78*** .74*** .56*** .62*** .63*** —
14. Belonging −.10* .01 −.06 −.04 −.24*** −.45*** −.49*** −.59*** −.42*** −.36*** −.33*** −.43*** −.55*** —
15. Self Est. −.06 .02 .01 .02 −.22*** −.48*** −.57*** −.60*** −.46*** −.33*** −.40*** −.41*** −.59*** .66*** —
16. Self Effic. −.08 .03 −.10* −.04 −.17*** −.35*** −.31*** −.41*** −.31*** −.19*** −.27*** −.25*** −.37*** .47*** .57*** —
17. Death id. .23*** .10* .10* .06 .25*** .32*** .27*** .42*** .40*** .26*** .34*** .28*** .41*** −.26*** −.27*** −.23*** —
18. Suicide id. .12** −.01 −.02 .00 .26*** .28*** .24*** .39*** .29*** .31*** .28*** .19*** .29*** −.21*** −.19*** −.15*** .36***
M — — — 48.9 0.79 1.36 1.11 1.18 1.07 0.73 0.81 1.20 1.00 47.58 98.18 3.63 —
SD — — — 11.7 0.74 0.94 0.87 1.02 0.89 0.69 0.86 0.84 0.83 10.25 20.76 0.72 —
Open in a new tab

Note. Veteran: Veteran status (1 = veteran, 0 = Non-veteran). Gender (1 = Male, 0 = Female). Marital Stat.: Marital Status (1 = Currently married, 0 = Unmarried). Somatiz.: BSI Somatization subscale. Obs. Comp.: BSI Obsessive Compulsive subscale. Interperson.: BSI Interpersonal Sensitivity subscale. Depression: BSI Depression subscale (excluding suicide ideation item). Anxiety: BSI Anxiety subscale. Hostility: BSI Hostility subscale. Phobia: BSI Phobia subscale. Paranoia: BSI Paranoia subscale. Psychotic.: BSI Psychoticism subscale: Belonging: SOBI-P score. Self Est.: SERS score. Self Effic.: GSES score. Death id.: Dichotomized score on the BSI death ideation item. Suicide id.: Dichotomized score on the BSI suicide ideation item. For continuous outcomes, Pearson’s r is reported. For dichotomous outcomes, Spearman’s ρ is reported. Mean and standard deviation not provided for dichotomous outcomes.

*

p < .05.

**

p < .01.

***

p < .001.

Table 2.

Demographics and Descriptive Statistics for Total Sample and Subgroups

Total Sample (N = 516) Veterans (n = 261) Non-Veterans (n = 255) Comparison Test
N M ± SD/Percentage n M ± SD/Percentage n M ± SD/Percentage Test Statistic df p
Age 516 48.9 ± 11.7 261 53.3 ± 9.1 255 44.3 ± 12.4 −9.46b 465 <.001
Gender (Male) 516 73.3% 261 87.4% 255 58.8% 53.60a 1 <.001
Marital Status (Married) 515 10.9% 261 16.9% 254 4.7% 19.56a 1 <.001
Belonging 515 47.6 ± 10.2 261 46.6 ± 10.6 254 48.6 ± 9.8 2.21b 513 .028
Self Esteem 506 98.2 ± 20.8 261 97.0 ± 20.2 245 99.4 ± 21.3 1.30b 504 .195
Self Efficacy 515 3.6 ± 0.7 261 3.6 ± 0.7 254 3.7 ± 0.7 1.91b 513 .057
Somatization 515 0.8 ± 0.7 261 0.9 ± 0.8 254 0.7 ± 0.7 −2.15b 513 .032
Obsessive Compulsive 515 1.4 ± 0.9 261 1.5 ± 1.0 254 1.2 ± 0.8 −3.10b 496 .002
Interpersonal Sensitivity 515 1.1 ± 0.9 261 1.1 ± 0.9 254 1.1 ± 0.8 0.41b 513 .679
Depression 515 1.2 ± 1.0 261 1.4 ± 1.1 254 1.0 ± 0.9 −3.91b 503 <.001
Anxiety 515 1.1 ± 0.9 261 1.2 ± 0.9 254 0.9 ± 0.8 −3.47b 506 <.001
Hostility 515 0.7 ± 0.7 261 0.8 ± 0.7 254 0.7 ± 0.6 −2.40b 503 .017
Phobia 515 0.8 ± 0.9 261 0.9 ± 0.9 254 0.7 ± 0.8 −2.00b 508 .047
Paranoia 515 1.2 ± 0.8 261 1.2 ± 0.8 254 1.2 ± 0.8 −0.29b 513 .775
Psychoticism 515 1.0 ± 0.8 261 1.1 ± 0.9 254 0.9 ± 0.7 −3.24b 503 .001
Thoughts of Ending Life 513 259 254 N/Ac -- <.001
 0-Not At All 89.5% 85.7% 93.3%
 1-A Little Bit 6.0% 8.5% 3.5%
 2-Moderately 1.8% 3.5% 0.0%
 3-Quite A Bit 1.6% 1.9% 1.2%
 4-Extremely 1.2% 0.4% 2.0%
Thoughts of Death/Dying 514 260 254 27.43a 4 <.001
 0-Not At All 71.4% 61.2% 81.9%
 1-A Little Bit 15.4% 20.4% 10.2%
 2-Moderately 5.3% 7.3% 3.1%
 3-Quite A Bit 4.9% 6.5% 3.1%
 4-Extremely 3.1% 4.6% 1.6%
Open in a new tab

Note:

a

Chi-Square test.

b

T-test.

c

Fisher exact. For Fisher exact test, statistic is marked N/A (Not applicable), consistent with standard reporting guidelines for this test. df: Degrees of freedom. Belonging: SOBI-P score. Self Esteem: SERS score. Self Efficacy: GSES score. Somatization: BSI Somatization subscale. Obsessive Compulsive: BSI Obsessive Compulsive subscale. Interpersonal Sensitivity: BSI Interpersonal Sensitivity subscale. Depression: BSI Depression subscale (excluding suicide ideation item). Anxiety: BSI Anxiety subscale. Hostility: BSI Hostility subscale. Phobia: BSI Phobia subscale. Paranoia: BSI Paranoia subscale. Psychoticism: BSI Psychoticism subscale.

Death Ideation

In the first regression model, which was unadjusted (see Table 3), veterans had an almost three-fold greater odds of endorsing death ideation compared with non-veterans. Our exploratory analyses indicated that veteran status remained a significant predictor of death ideation, even when all other predictors were entered into the model. Depressive and psychotic symptoms also emerged as significant predictors. Indicators of thwarted belongingness and perceived burdensomeness (i.e., belonging, self-efficacy, and self-esteem) were not significant predictors.

Table 3.

Logistic Regression Results Predicting Death Ideation

Predictor Outcome: Death Ideation (Dichotomous Yes/No)
Model I Model II Model III Model IV
OR (95% CI) Wald χ2 p OR (95% CI) Wald χ2 p OR (95% CI) Wald χ2 p OR (95% CI) Wald χ2 p
Veteran 2.86 (1.91–4.29) 25.80* 0.000 2.78 (1.74–4.42) 18.55* 0.000 1.92 (1.13–3.29) 5.71* 0.017 1.92 (1.12–3.31) 5.61* 0.018
Age 0.99 (0.97–1.01) 0.73 0.393 1.00 (0.98–1.02) 0.04 0.845 1.00 (0.98–1.02) 0.01 0.940
Gender 1.25 (0.75–2.05) 0.73 0.392 1.50 (0.85–2.62) 1.96 0.162 1.55 (0.87–2.75) 2.22 0.136
Marital Stat. 1.45 (0.80–2.63) 1.51 0.218 1.34 (0.67–2.71) 0.68 0.408 1.32 (0.65–2.70) 0.59 0.441
Somatiz. 1.03 (0.78–1.37) 0.05 0.825 1.05 (0.79–1.40) 0.11 0.746
Obs. Comp. 0.79 (0.54–1.09) 1.53 0.217 0.74 (0.50–1.09) 2.28 0.131
Interperson. 0.77 (0.54–1.09) 2.17 0.141 0.75 (0.52–1.08) 2.34 0.126
Depression 1.79 (1.23–2.62) 9.09* 0.003 1.80 (1.19–2.71) 7.80* 0.005
Anxiety 1.39 (0.92–2.12) 2.38 0.123 1.45 (0.94–2.23) 2.81 0.094
Hostility 0.92 (0.68–1.24) 0.32 0.570 0.92 (0.68–1.26) 0.26 0.614
Phobia 1.25 (0.91–1.71) 1.86 0.173 1.21 (0.87–1.68) 1.30 0.255
Paranoia 1.05 (0.76–1.46) 0.10 0.750 1.10 (0.79–1.53) 0.28 0.594
Psychotic. 1.59 (1.08–2.34) 5.55* 0.019 1.54 (1.03–2.31) 4.41* 0.036
Belonging 1.02 (0.74–1.42) 0.02 0.899
Self Esteem 1.00 (0.70–1.43) 0.00 0.998
Self Effic. 0.85 (0.65–1.12) 1.34 0.247
Open in a new tab

Note:

*

p < 0.05

OR: Odds Ratio. CI: Confidence Interval. Veteran: Veteran status. Marital Stat.: Marital Status. Somatiz.: BSI Somatization subscale. Obs. Comp.: BSI Obsessive Compulsive subscale. Interperson.: BSI Interpersonal Sensitivity subscale. Depression: BSI Depression subscale (excluding suicide ideation item). Anxiety: BSI Anxiety subscale. Hostility: BSI Hostility subscale. Phobia: BSI Phobia subscale. Paranoia: BSI Paranoia subscale. Psychotic.: BSI Psychoticism subscale: Belonging: SOBI-P score. Self Esteem: SERS score. Self Effic.: GSES score.

Suicide Ideation

In the second regression analysis (see Table 4), veteran status was a significant predictor of suicide ideation in the unadjusted model, with veterans having over a two-fold greater odds of endorsing suicide ideation as non-veterans. Additionally, veteran status remained a significant predictor of suicide ideation when demographic factors were entered into the model. However when psychiatric symptom subscales were entered into the model, veteran status was no longer a significant predictor of suicide ideation; depressive symptoms and hostility were the only significant predictors. Once again, belonging, self-efficacy, and self-esteem were not significant predictors.

Table 4.

Logistic Regression Results Predicting Suicide Ideation

Predictor Outcome: Suicide Ideation (Dichotomous Yes/No)
Model I Model II Model III Model IV
OR (95% CI) Wald χ2 p OR (95% CI) Wald χ2 p OR (95% CI) Wald χ2 p OR (95% CI) Wald χ2 p
Veteran 2.31 (1.27–4.23) 7.44* 0.006 3.18 (1.58–6.40) 10.51* 0.001 1.81 (0.81–4.05) 2.09 0.148 1.83 (0.81–4.18) 1.77 0.184
Age 0.99 (0.96–1.01) 1.23 0.267 1.00 (0.97–1.03) 0.00 0.992 1.00 (0.96–1.03) 0.00 0.972
Gender 0.68 (0.34–1.34) 1.25 0.264 0.77 (0.36–1.65) 0.46 0.499 0.85 (0.39–1.86) 0.45 0.503
Marital Stat. 0.63 (0.23–1.68) 0.87 0.351 0.35 (0.11–1.11) 3.16 0.076 0.34 (0.10–1.12) 3.15 0.076
Somatiz. 1.39 (0.96–2.00) 3.04 0.081 1.46 (0.87–2.44) 2.51 0.114
Obs. Comp. 0.86 (0.51–1.47) 0.29 0.589 0.84 (0.47–1.51) 0.29 0.593
Interperson. 0.77 (0.48–1.22) 1.26 0.261 0.79 (0.45–1.38) 0.81 0.367
Depression 3.78 (2.19–6.47) 23.07* 0.000 3.66 (2.03–6.59) 18.99* 0.000
Anxiety 0.87 (0.51–1.49) 0.26 0.613 0.90 (0.48–1.67) 0.20 0.653
Hostility 1.62 (1.11–2.37) 6.10* 0.014 1.98 (1.11–3.51) 5.64* 0.018
Phobia 1.53 (1.00–2.34) 3.85* 0.050 1.68 (1.01–2.80) 3.63 0.057
Paranoia 0.66 (0.41–1.06) 3.01 0.083 0.60 (0.34–1.06) 3.07 0.080
Psychotic. 0.82 (0.49–1.37) 0.59 0.442 0.80 (0.42–1.55) 0.35 0.554
Belonging 0.99 (0.94–1.04) 0.04 0.848
Self Esteem 1.01 (0.99–1.03) 0.35 0.555
Self Effic. 0.94 (0.52–1.69) 0.08 0.777
Open in a new tab

Note:

*

p < 0.05

OR: Odds Ratio. CI: Confidence Interval. Veteran: Veteran status. Marital Stat.: Marital Status. Somatiz.: BSI Somatization subscale. Obs. Comp.: BSI Obsessive Compulsive subscale. Interperson.: BSI Interpersonal Sensitivity subscale. Depression: BSI Depression subscale (excluding suicide ideation item). Anxiety: BSI Anxiety subscale. Hostility: BSI Hostility subscale. Phobia: BSI Phobia subscale. Paranoia: BSI Paranoia subscale. Psychotic.: BSI Psychoticism subscale: Belonging: SOBI-P score. Self Esteem: SERS score. Self Effic.: GSES score.

Discussion

Our aim was to identify differences in death ideation and suicide ideation between veterans and non-veterans with SMI, to inform mental health professionals’ decisions regarding levels of suicide risk and appropriate management of risk. We found that there were differences in both suicide ideation and death ideation by veteran status; the rates of suicide ideation and death ideation among veterans with SMI were more than double those among non-veterans with SMI. In our exploratory analyses, veteran status, depressive symptoms, and psychotic symptoms were significant predictors of death ideation even after adjusting for other risk factors. These findings suggest that being a veteran is uniquely associated with the presence of thoughts of death or dying. Depressive and psychotic symptoms are also associated with these thoughts. When examining suicide ideation as an outcome, veteran status was no longer a significant predictor of suicide ideation after adjusting for psychiatric symptoms. Depressive symptoms and hostility were significantly associated with suicide ideation in the fully adjusted model.

These findings are generally consistent with existing literature, which suggests that depressive symptoms and psychotic symptoms are important risk factors for suicide among individuals with SMI and among veterans (e.g., Angst et al., 1999; Hawton et al., 2005; Hor & Taylor, 2010; VA & DoD, 2013). Interestingly, hostility also emerged as a strong predictor of suicide ideation in our analyses, which has not been examined thoroughly in individuals with SMI. However, hostility has been identified as a suicide risk factor in other populations (e.g., Zhang et al., 2012). Additionally, clinical guidelines suggest that impulsive behaviors (e.g., angry outbursts) are risk factors (VA & DoD, 2013). Angry outbursts may be not only an indicator of impulsivity, but also potentially of hostility. However, we also did not find support for the interpersonal theory of suicide (Joiner, 2005; Van Orden et al., 2010). None of the indicators of perceived burdensomeness or thwarted belongingness were significantly associated with suicide ideation or death ideation, which is inconsistent with previous studies (e.g, Jahn et al., 2015; Monteith et al., 2013). The indicators may not have fully assessed the constructs as proposed in the theory, or the theory may not be as relevant to suicide ideation among individuals with SMI.

There are numerous clinical implications of this study that can support clinicians in their suicide risk assessment and decisions regarding levels of risk. Identifying the specific reasons why veterans with SMI reported death ideation more often than non-veterans with SMI was beyond the scope of this study. However, clinicians should assess for death ideation and suicide ideation in both veterans and non-veterans with SMI given the relatively high rates of death ideation and suicide ideation found in this study. It is especially critical to assess for death ideation among veterans with SMI, as they had higher rates of death ideation than non-veterans with SMI. There is a common belief in mental health that death ideation does not convey the same level of risk as suicide ideation. Yet, research suggests that there are not differences between death ideation and suicide ideation in terms of previous suicide attempts or current hopelessness (Szanto et al., 2002), nor in terms of death by suicide (L. Berman, personal communication, April 28, 2016). As such, assessment of death ideation should be done as part of every suicide risk assessment, in addition to a thorough assessment of suicide ideation. The high rates of suicide among veterans and individuals with SMI (Angst et al., 1999; Hawton et al., 2005; Hor & Taylor, 2010; VA Office of Suicide Prevention, 2016) indicate that a comprehensive suicide risk assessment should be done at the initial visit of every client who is a veteran or has an SMI diagnosis. Comprehensive suicide risk assessments should include questions about death ideation (e.g., “Are you having thoughts about your own death or wanting to die?”), and any reports of death ideation should be taken as seriously as suicide ideation (Szanto et al., 2002). Assessment of suicide ideation (e.g., “Are you having thoughts of killing yourself or thinking about suicide?”) should also occur.

Though we were unable to identify reasons for the particularly high rates of death ideation among veterans with SMI, clinicians may benefit from exploring potential reasons for this with individual clients, though additional research is needed to confirm the utility of this approach. It is possible that veteran status may be associated with unique experiences or beliefs that confer greater risk for death ideation. For example, given the focus in the popular media on veteran suicide (e.g., Philipps, 2015), it may be the case that veterans identify with the veterans covered in these media stories, which then increases their own thoughts of death. Alternatively, it may be that military experience itself increases death ideation among veterans. For example, veterans were trained to handle and use firearms during basic training, may have been exposed to combat in which they witnessed death (e.g., seeing other service members or enemy combatants die), and may have considered the possibility of their own death during their military service. Therefore, thoughts of death or wishes to die may occur more often during times of crisis because veterans have habituated to such thoughts. Additionally, death ideation may differentially serve as a coping mechanism for veterans. Veterans may experience death ideation as a way to cope with pain or difficulties more often than non-veterans, perhaps in part due to their previous military experiences. These various reasons may suggest different approaches for addressing death ideation. Future studies could assess the effectiveness of introducing other coping skills that can replace death ideation (e.g., dialetical behavior therapy, which focuses on replacing maladaptive coping strategies with more effective and adaptive coping skills; Linehan, 1993; cognitive restructuring techniques used to reframe thoughts that he or she is similar to veterans who have died by suicide). Similarly, additional work may examine the role of cognitive therapy in reducing risk factors that contribute to their death ideation (Substance Abuse and Mental Health Services Administration, 2014; Tarrier & Taylor, 2014; Young, Rygh, Weinberger, & Beck, 2014) or suicide risk (Brown, Henriques, Ratto, & Beck, 2002), as well as other options to reduce suicide risk, including problem-solving therapy (Stewart, Quinn, Plever, & Emmerson, 2009) and the collaborative assessment and management of suicidality (Jobes, 2006).

Future work should also investigate other clinically-relevant issues regarding suicide risk among veterans and individuals with SMI. Studies should investigate how ongoing assessment of suicide risk influences the identification and monitoring of suicide risk, as well as treatment decisions. For example, a one-time assessment, even if a client denies suicide or death ideation at that initial assessment, may not be sufficient to monitor and manage suicide risk. Additionally, research should explore how direct versus indirect language in assessment affects clients’ responses and determinations of risk level.

The goal of our study was to examine differences in suicide-related outcomes among veterans and non-veterans with SMI, to inform mental health professionals’ decisions regarding levels of suicide risk and appropriate management of risk. Our findings, which suggest that veterans with SMI have higher rates of suicide and death ideation than non-veterans with SMI, indicate that clinicians should particularly monitor and intervene upon these suicide-related outcomes in veterans with SMI, as well as associated clinical risk factors such as depression, psychotic symptoms, and hostility. In sum, it appears that questions about both death ideation and suicide ideation should be part of comprehensive and ongoing suicide risk assessments. Cognitive strategies may be effective for addressing suicide risk directly, as well as the risk factors identified in this study.

Acknowledgments

This work was supported by NIMH 1R01MH090036-01A1 and VA HSRD 1I01HX000279-0 research grants. The authors would like to acknowledge and thank Deborah Medoff, Ph.D. for her review of an earlier version of this manuscript.

Footnotes

1

We examined differences in variables of interest between Veterans recruited from VA sites and Veterans recruited from community sites. No significant differences were found in terms of age, sex, marital status, presence of death ideation, presence of suicide ideation, self-efficacy, or paranoid ideation. Veterans recruited from VA sites had significantly lower self-esteem and sense of belonging, and higher scores on all other psychiatric symptoms (i.e., somatization, obsession-compulsion, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, and psychoticism). However, these differences are likely not clinically meaningful given the small differences. For example, average sense of belonging scores differed by less than 5 points, and the range of this scale is 72 points. Similarly, most psychiatric symptom average scores differed by 0.5 points or fewer, and the range on these scores was 5 points.

References

  1. American Association of Suicidology. USA suicide: 2014 official final data. 2015 Retrieved from http://www.suicidology.org/Portals/14/docs/Resources/FactSheets/2014/2014datapgsv1b.pdf.
  2. Angst J, Angst F, Stassen HH. Suicide risk in patients with major depressive disorder. Journal of Clinical Psychiatry. 1999;60(S2):57–62. Retrieved from https://www.psychiatrist.com/jcp/pages/home.aspx. [PubMed] [Google Scholar]
  3. Brown GK, Beck AT, Steer RA, Grisham JR. Risk factors for suicide in psychiatric outpatients: A 20-year prospective study. Journal of Consulting and Clinical Psychology. 2000;68:371–377. doi: 10.1037/0022-006X.68.3.371. [DOI] [PubMed] [Google Scholar]
  4. Brown GK, Henriques GR, Ratto C, Beck AT. Cognitive therapy treatment manual for suicide attempters. Philadelphia, PA: University of Pennsylvania; 2002. [Google Scholar]
  5. Brown GK, Ten Have T, Henriques GR, Xie SX, Hollander JE, Beck AT. Cognitive therapy for the prevention of suicide attempts: A randomized controlled trial. Journal of the American Medical Association. 2005;294:563–570. doi: 10.1001/jama.294.5.563. [DOI] [PubMed] [Google Scholar]
  6. Centers for Disease Control and Prevention. Web-based injury statistics query and reporting system. 2016 Retrieved from http://www.cdc.gov/injury/wisqars/index.html.
  7. Chen G, Gully SM, Eden D. Validation of a new general self-efficacy scale. Organizational Research Methods. 2001;4:62–83. doi: 10.1177/109442810141004. [DOI] [Google Scholar]
  8. Desseilles M, Perroud N, Guillaume S, Jaussent I, Genty C, Malafosse A, Courtet P. Is it valid to measure suicidal ideation by depression rating scales? Journal of Affective Disorders. 2012;136:398–404. doi: 10.1016/j.jad.2011.11.013. [DOI] [PubMed] [Google Scholar]
  9. Department of Veterans Affairs, & Department of Defense. VA/DoD Clinical Practice Guideline for Assessment and Management of Patients at Risk for Suicide. 2013 Retrieved from http://www.healthquality.va.gov/guidelines/MH/srb/
  10. Derogatis LR. BSI Brief Symptom Inventory: Administration, scoring, and procedures manual. 4. Minneapolis, MN: National Computer Systems; 1993. [Google Scholar]
  11. Derogatis LR, Melisaratos N. The Brief Symptom Inventory: An introductory report. Psychological Medicine. 1983;13:595–605. doi: 10.1017/S0033291700048017. [DOI] [PubMed] [Google Scholar]
  12. Hagerty BMK, Patusky K. Developing a measure of sense of belonging. Nursing Research. 1995;44:9–13. doi: 10.1097/00006199-199501000-00003. [DOI] [PubMed] [Google Scholar]
  13. Hawton K, Sutton L, Haw C, Sinclair J, Harriss L. Suicide and attempted suicide in bipolar disorder: A systematic review of risk factors. Journal of Clinical Psychiatry. 2005;66:693–704. doi: 10.4088/jcp.v66n0604. Retrieved from https://www.psychiatrist.com/jcp/pages/home.aspx. [DOI] [PubMed] [Google Scholar]
  14. Hor K, Taylor M. Suicide and schizophrenia: A systematic review of rates and risk factors. Journal of Psychopharmacology. 2010;24(11, S4):81–90. doi: 10.1177/1359786810385490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Ilgen MA, Bohnert ASB, Ignacio RV, McCarthy JF, Valenstein MM, Kim M, Blow FC. Psychiatric diagnoses and risk of suicide in veterans. Archives of General Psychiatry. 2010;67:1152–1158. doi: 10.1001/archgenpsychiatry.2010.129. [DOI] [PubMed] [Google Scholar]
  16. Jahn DR, Poindexter EK, Cukrowicz KC. Personality disorder traits, risk factors, and suicide ideation among older adults. International Psychogeriatrics. 2015;27:1785–1794. doi: 10.1017/S1041610215000174. [DOI] [PubMed] [Google Scholar]
  17. Jobes DA. Managing suicidal risk: A collaborative approach. New York, NY: Guilford; 2006. [Google Scholar]
  18. Joiner T. Why people die by suicide. Cambridge, MA: Harvard University Press; 2005. [Google Scholar]
  19. Linehan MM. Cognitive behavioral treatment of borderline personality disorder. New York, NY: Guilford; 1993. [Google Scholar]
  20. Lucksted A, Drapalski A, Calmes C, Forbes C, DeForge B, Boyd J. Ending self-stigma: Pilot evaluation of a new intervention to reduce internalized stigma among people with mental illnesses. Psychiatric Rehabilitation Journal. 2011;35:51–54. doi: 10.2975/35.1.2011.51.54. [DOI] [PubMed] [Google Scholar]
  21. Monteith LL, Menefee DS, Pettit JW, Leopoulos WL, Vincent JP. Examining the interpersonal-psychological theory of suicide in an inpatient veteran sample. Suicide and Life-Threatening Behavior. 2013;43:418–428. doi: 10.1111/sltb.12027. [DOI] [PubMed] [Google Scholar]
  22. National Center for Veterans Analysis and Statistics. Profile of veterans: 2014. Data from the American Community Survey. 2016 Retrieved from http://www.va.gov/vetdata/docs/SpecialReports/Profile_of_Veterans_2014.pdf.
  23. Nock MK, Stein MB, Heeringa SG, Ursano RJ, Colpe LJ, Fullerton CS … Army STARRS Collaborators. Prevalence and correlates of suicidal behavior among soldiers: Results from the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS) JAMA Psychiatry. 2014;71:514–522. doi: 10.1001/jamapsychiatry.2014.30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Nugent WR. A validity study of two forms of the Self-Esteem Rating Scale. Research on Social Work Practice. 2004;14:297–294. doi: 10.1177/1049731503257890. [DOI] [Google Scholar]
  25. Nugent WR, Thomas J. Validation of a clinical measure of self-esteem. Research on Social Work Practice. 1993;3:191–207. doi: 10.1177/104973159300300205. [DOI] [Google Scholar]
  26. Philipps D. In unit stalked by suicide, veterans try to save one another. New York Times. 2015 Sep 19; Retrieved from http://www.nytimes.com.
  27. Sargent J, Williams RA, Hagerty B, Lynch-Sauer J, Hoyle K. Sense of belonging as a buffer against depressive symptoms. Journal of the American Psychiatric Nursing Association. 2002;8:120–129. doi: 10.1067/mpn.2002.127290. [DOI] [Google Scholar]
  28. Scocco P, De Leo D. One year prevalence of death thoughts, suicide ideation and behaviours in an elderly population. International Journal of Geriatric Psychiatry. 2002;17:842–846. doi: 10.1002/gps.691. [DOI] [PubMed] [Google Scholar]
  29. Sherbaum CA, Cohen-Charash Y, Kern MJ. Measuring general self-efficacy: A comparison of three measures using item response theory. Educational Psychology Measurement. 2006;66:1047–1063. doi: 10.1177/0013164406288171. [DOI] [Google Scholar]
  30. Smith NB, Mota N, Tsai J, Monteith L, Harpaz-Rotem I, Southwick SM, Pietrzak RH. Nature and determinants of suicidal ideation among U.S. veterans: Results from the National Health and Resilience in Veterans Study. Journal of Affective Disorders. 2016;197:66–73. doi: 10.1016/j.jad.2016.02.069. [DOI] [PubMed] [Google Scholar]
  31. Stewart CD, Quinn A, Plever S, Emmerson B. Comparing cognitive behavior therapy, problem solving therapy, and treatment as usual in a high risk population. Suicide and Life Threatening Behavior. 2009;39:538–547. doi: 10.1521/suli.2009.39.5.538. [DOI] [PubMed] [Google Scholar]
  32. Substance Abuse and Mental Health Services Administration. Anger management for substance abuse and mental health clients: Participant workbook. 2014 Retrieved from http://store.samhsa.gov/product/Anger-Management-for-Substance-Abuse-and-Mental-Health-Clients-Participant-Workbook/SMA14-4210.
  33. Szanto K, Reynolds CF, III, Frank E, Stack J, Fasiczka AL, Miller M, … Kupfer DJ. Suicide in elderly patients: Is active vspassive suicidal ideation a clinically valid. American Journal of Geriatric Psychiatry. 2002;4:197–207. doi: 10.1097/00019442-199622430-00003. [DOI] [PubMed] [Google Scholar]
  34. Tarrier N, Taylor R. Schizophrenia and other psychotic disorders. In: Barlow DH, editor. Clinical handbook of psychological disorders: A step-by-step treatment manual. 5. New York, NY: Guilford; 2014. pp. 502–532. [Google Scholar]
  35. VA Office of Suicide Prevention. Suicide among veterans and other Americans 2001–2014. 2016 Retrieved from http://www.mentalhealth.va.gov/docs/2016suicidedatareport.pdf.
  36. Van Orden KA, Cukrowicz KC, Witte TK, Joiner TE., Jr Thwarted belongingness and perceived burdensomeness: Construct validity and psychometric properties of the Interpersonal Needs Questionnaire. Psychological Assessment. 2012;24:197–215. doi: 10.1037/a0025358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby EA, Joiner TE., Jr The interpersonal theory of suicide. Psychological Review. 2010;117:575–600. doi: 10.1037/a0018697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wenzel A, Berchick ER, Tenhave T, Halberstadt S, Brown GK, Beck AT. Predictors of suicide relative to other deaths in patients with suicide attempts and suicide ideation: A 30-year prospective study. Journal of Affective Disorders. 2011;132:375–382. doi: 10.1016/j.jad.2011.03.006. [DOI] [PubMed] [Google Scholar]
  39. Young JE, Rygh JL, Weinberger AD, Beck AT. Cognitive therapy for depression. In: Barlow DH, editor. Clinical handbook of psychological disorders: A step-by-step treatment manual. 5. New York, NY: Guilford; 2014. pp. 275–331. [Google Scholar]
  40. Zhang P, Roberts RE, Liu Z, Meng X, Tang J, Sun L, Yu Y. Hostility, physical aggression and trait anger as predictors for suicidal behavior in Chinese adolescents: A school-based study. PLoS One. 2012;7(2):e31044. doi: 10.1371/journal.pone.0031044. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES