Skip to main content
NCBI home page
VA Author Manuscripts logoLink to VA Author Manuscripts
. Author manuscript; available in PMC: 2024 Nov 1.
Published in final edited form as: Psychol Trauma. 2022 Jun 2;15(8):1315–1323. doi: 10.1037/tra0001271

The transition to civilian life: Impact of comorbid PTSD, chronic pain, and sleep disturbance on veterans’ social functioning and suicidal ideation

Rachel Shor 1, Shelby Borowski 1, Rachel L Zelkowitz 1, Suzanne L Pineles 1,2, Laurel A Copeland 3,4, Erin P Finley 5,6, Daniel F Perkins 7, Dawne Vogt 1,2
PMCID: PMC10231656  NIHMSID: NIHMS1894561  PMID: 35653743

Abstract

Objective:

Trauma-exposed veterans may be more likely to experience PTSD, chronic pain, and sleep disturbance together rather than in isolation. While these conditions are independently associated with distress and impairment, how they relate to social functioning and suicidal ideation (SI) when experienced comorbidly is not clear.

Method:

Using longitudinal data on 5,461 trauma-exposed U.S. veterans from The Veterans Metrics Initiative Study and self-reported disorders, we assessed (a) the extent to which PTSD co-occurs with sleep disturbance and chronic pain (CP); (b) the relationship of PTSD in conjunction with sleep disturbance and chronic pain with later social functioning and SI; and (c) the extent to which social functioning mediates the impact of multimorbidity on SI.

Results:

At approximately 15-months post-separation, 90.5% of veterans with probable PTSD also reported sleep disturbance and/or CP. Relative to veterans without probable PTSD, Veterans with all three conditions (n = 907) experienced the poorest social functioning (B = −0.56, p < .001) and had greater risk for SI (OR = 3.78, p < .001); Social functioning partially mediated the relationship between multimorbidity and SI. However, relative to those with PTSD alone, sleep disturbance and CP did not confer greater risk for SI.

Conclusions:

Although these findings underscore the impact of PTSD on functioning and SI, they also highlight the complexity of multimorbidity and the importance of bolstering social functioning for veterans.

Approximately 200,000 U.S. service members complete their military service each year (GAO, 2019). The first several years following separation can be a vulnerable time as veterans navigate multiple life transitions and reacclimate to their civilian communities. Moreover, a growing body of literature suggests that problems that occur during readjustment confer significant risk for suicidal ideation (SI; e.g., Kline et al., 2011) and suicide mortality (e.g., Hoffmire et al., 2015; Reger et al., 2015). In a sample of veterans who served between 2001 and 2007, suicide rates were highest in the first three years after leaving the military – highlighting the need to better understand risk factors during early reintegration (Kang et al., 2015).

Posttraumatic Stress Disorder (PTSD) is a debilitating disorder that has been found to be associated with increased risk for suicidal thoughts and behaviors (e.g., Panagioti, Gooding & Tarrier, 2012; Pompili et al. 2013). In fact, among veterans, PTSD was one of the most important correlates of SI identified in a machine learning analysis (Gradus et al., 2017). PTSD though rarely occurs in isolation (Panagioti, Gooding & Tarrier, 2012; Lew et al., 2009; DiNapoli et al., 2016). Among co-occurring disorders, insomnia and chronic pain (CP) are especially prevalent (e.g., Greene, Neria, & Gross, 2016; Beck, & Clapp, 2011) and are also strong risk factors for suicidal thoughts and behaviors (e.g., Pigeon, Pinquart & Conner, 2012; Calati et al., 2015). Furthermore, many individuals continue to experience CP and residual sleep problems even after their PTSD is effectively treated (e.g., López et al., 2019). Therefore, sleep disturbance and CP may uniquely and cumulatively contribute to risk for SI (Racine, 2018).

Understanding the consequences of PTSD-multimorbidity is a pressing concern, as there is growing evidence that co-occurring mental and physical health conditions confer additional risk for functional impairment (Pagotto et al., 2015; Kessler et al., 2003; Ouimette et al., 2011; Larkin, et a., 2020) beyond that associated with PTSD alone (Nichter et al., 2019; Manhapra et al., 2021). However, the specific means by which multimorbidity increases risk for poor outcomes are unclear. One hypothesized mechanism by which multimorbidity may lead to increased SI is through declines in social functioning. Poor social functioning is linked with the aforementioned health conditions (e.g., Kintzle et al., 2018; Hom et al., 2017) and strongly associated with reduced social support, which has been theorized to be a key risk factor for suicide (e.g., Durkheim, 1951; Johnson et al., 2011; Joiner, 2005; Van Orden et al., 2010). In addition, cross-sectional research suggests that key facets of social functioning impact the relationship between SI and mental health symptoms, such that when social support is low mental health symptoms are positively associated with a history of SI (e.g., DeBeer et al., 2014; Dutton et al., 2016). These models may help explain why health conditions that erode social functioning are risk factors for suicide; however, only through longitudinal research can we examine whether social functioning mediates the relationship between health and SI in the context of multimorbidity.

This paper provides a preliminary examination of (1) the frequency with which post-9/11 Veteran’s experience multimorbidity in their reporting of PTSD, sleep disturbance, and chronic pain during their transition from service; (2) the impact of multimorbidity on social functioning and SI; and (3) the extent to which social functioning mediates the relationship between multimorbidity and later SI. We hypothesized that more Veterans would concurrently report PTSD and a physical health condition (i.e., sleep disturbance, chronic pain) than PTSD independently (Hypothesis 1). We also hypothesized that Veterans who reported PTSD in conjunction with one or both conditions would report poorer social functioning (Hypothesis 2a), and increased SI (Hypothesis 2b), relative to those without PTSD or with PTSD alone. Finally, we proposed that social functioning would mediate the relationship between multimorbidity and later SI (Hypothesis 3). In this mediation model, we expected that experiencing multiple chronic health conditions would predict poorer social functioning, which in turn would predict an increase in SI.

Method

Participants and Procedures

This analysis drew from The Veterans Metrics Initiative Study (TVMI), a large prospective cohort study of newly separated U.S. military veterans, described in detail elsewhere (Vogt et al., 2018). A nationally representative sample of veterans who had separated from service within the last 90 days and who were U.S. residents were identified through the VA/Department of Defense Identity Repository and recruited in fall 2016. Participants completed six assessments in total and were compensated at each time point. IRB approval was provided by the survey contractor (ICF International) and VA Boston Healthcare System. We examined data from three time points: approximately 15-months, 21-months, and 27-months post-separation, referred to as T1, T2, and T3, respectively. Although data were also collected at 3 and 9-months post-separation, we started our analyses at 15-months post-separation because this was the first time point that PTSD symptom severity was assessed. Given the primary focus on veterans’ experiences of PTSD, the sample for this analysis was limited to those who self-identified as having been exposed to a traumatic event on the abbreviated PCL-5 at T1 (N = 5,461).

At T1, participants had a mean age of 34.99 (SD = 9.44) and 18.4% identified as female. The majority identified as White/non-Hispanic (64.8%), with smaller percentages identifying as other races (Black = 10.9 %, White/Hispanic = 9.8%, multiracial = 6.8 %, Asian = 3.1%, and other = 4.7%). About 36% of participants separated from the Army, 19% Navy, 18% Air Force, and 17% Marine Corps, and 10% from activated reserve forces. Approximately 25% of participants either reported having PTSD or scored above the abbreviated PCL-5 cut-off score of 19 (Price et al., 2016). The average score for social functioning at T2 was 3.88 (SD = 0.79; range 1 to 5). Of the 5,461 Veterans who completed T1, 472 (8.5%) endorsed having thoughts of death or self-harm at T3. Social functioning at T2 and SI at T3 were significantly correlated (rpb = −.25, p <.01).

Measures

Health conditions

Chronic pain, sleep disturbance, and PTSD were measured at T1, T2, and T3, with the Well-Being Inventory (WBI) assessment of self-reported health conditions (Vogt et al., 2018). Veterans were asked to identify ongoing physical or mental/emotional health conditions, illnesses, or disabilities. The WBI includes PTSD, chronic pain, and sleep problem or disorder as possible responses among 12 health conditions. Veterans also had the option to free-write conditions. In addition to the WBI, PTSD was assessed with the PCL-5 abbreviated measure, which has been found to have comparable diagnostic utility to the total-scale PTSD Checklist (Price et al., 2016) but does not include an item assessing trouble falling or staying asleep. Veterans were given the PCL-5 if they endorsed having experienced a traumatic event. Possible PCL-5 scores ranged from 0 to 32. Veterans who did not endorse a traumatic event were coded as 0. Veterans were considered to have PTSD if they either self-reported having PTSD on the WBI or endorsed a score of 19 or higher on the abbreviated PCL-5; (cutoff score indicating probable PTSD, Price et al., 2016). Approximately 87% of the sample reported consistently on both tools assessing PTSD; that is, they reported having (or not having) PTSD on both the WBI and the abbreviated PCL-5. Veterans were grouped into the following mutually exclusive categories based on their responses: (1) no PTSD, (2) PTSD only, (3) PTSD & Sleep Disturbance, (4) PTSD & Chronic Pain, and (5) PTSD, Sleep Disturbance, & Chronic Pain.

Social functioning

Social functioning was assessed at T2 with the 9-item social functioning measure in the WBI (Vogt et al., 2018). This measure captures the extent to which veterans engage in behaviors that contribute to positive relationships with extended family, friends, and within one’s broader community in the past 3 months. An example item is, “how often have you gotten along well with members of your community.” Response options range from 1(never) to 5(most or all of the time). Social functioning scores were averaged; thus, possible scores ranged from 1 to 5, with higher scores indicating higher levels of social functioning. Prior research has demonstrated the reliability and validity of this scale in post-9/11 Veterans (Vogt et al., 2018). Cronbach’s alpha for this sample was very good (α = .89).

Suicidal Ideation (SI)

SI was captured at T3 using item-9 from the Patient Health Questionnaire-9 (PHQ-9; Kroenke, Spitzer, & Williams, 2001), which is used as an initial screen for suicidality in the VA’s Suicide Prevention Program (DVA, 2018), has been used extensively in prior research and clinical practice (e.g., Wisco et al., 2014; Sall et al., 2019; Elbogen et al., 2021), and has been found to have strong sensitivity of SI (Na et al., 2018). Veterans were asked how often in the past two weeks they were bothered by, “thoughts that you would be better off dead, or of hurting yourself.” Response options ranged from 0 (not at all) to 3 (nearly every day). Given that scores remained positively skewed even after treating the variable (e.g., winsorizing, log10 transformation), responses to this item were dichotomized. Veterans who endorsed any suicidal thoughts (scores of 1, 2, or 3) were categorized in the SI group (1). Veterans who endorsed no suicidal thoughts (score of 0) were coded no-SI (0).

Data Analytic Strategy

Analyses were conducted using IBM SPSS Statistics (v25) and the PROCESS macro (v3.4.1; Hayes, 2017). For Hypothesis 1, we examined the proportion of veterans who met study criteria for no PTSD, PTSD only, and PTSD concurrent with CP and or sleep disturbance. We used multiple linear regression to examine impacts of T1 PTSD and associated comorbidities on T2 social functioning (Hypothesis 2a) and logistic regression to examine the impacts of these variables on T3 SI (Hypothesis 2b). We adjusted for variables that were significantly associated with either social functioning or SI at the p < .05 level at T1 (i.e., gender and age). Using PROCESS, we ran a mediation model to test Hypothesis 3 (Hayes, 2017); using sensitivity analyses we also confirmed that there were no/negligible group differences at TI in PTSD, sleep disturbance, chronic pain, and SI between veterans who did and did not complete all three assessments1. The mediation model included health condition groups as the independent variable, social functioning as the mediator, and SI as the dependent variable; adjusting for gender and age. Direct effects in this model were: 1) the influence of health condition on social functioning; 2) the influence of health condition on SI; and 3) the influence of social functioning on SI. Direct effects with SI as the outcome can be interpreted on a log-odds metric and odds ratios (OR) were calculated for these direct effects. Direct effects with social functioning as the outcome can be interpreted as unstandardized coefficients. Indirect effects were also calculated (Hayes, 2017; Kline, 2015) and can be interpreted on a log-odds metric. We used bootstrapping with 5000 samples to examine the significance of the indirect effects. The regression and mediation models were run twice, examining no-PTSD and PTSD-only as reference group, to provide insight into the impact of comorbidity relative to no-PTSD and to compare the relative contribution of co-occurring physical health conditions to PTSD.

Results

Comorbidity of Chronic Health Conditions (H1)

The frequency of comorbid health conditions reported at each timepoint are reported in Table 1. As expected, more veterans reported PTSD and another condition than PTSD alone; about 1 in 6 were identified as potentially having all three conditions. Of those with probable PTSD at T1 (n=1,369), 90.5% reported also having sleep disturbance and/or CP. Similar trends were observed for T2 and T3, with 88.9% and 88.2% of veterans with PTSD reporting an additional health condition, respectively.

Table 1.

Comorbidity of Health Conditions Among Trauma-Exposed Veterans by Time Point

T1 T2 T3
Variable n (%) N=5461 N=4621 N=4191
Health Condition
 No PTSD 4092 (73.7) 3446 (74.9) 3100 (75.8)
 PTSD Only 131 (2.4) 131 (2.8) 117 (2.9)
 PTSD & Sleep Disturbance 150 (2.7) 134 (2.9) 130 (3.2)
 PTSD & Chronic Pain 181 (3.3) 151 (3.3) 153 (3.7)
 PTSD, Sleep Disturbance, & Chronic Pain 907 (16.3) 759 (16.4) 588 (14.4)
Open in a new tab

Note. T1= 15-months, T2= 21-months, and T3= 27-months post-separation

Chronic Health Conditions and Social Functioning (H2a)

Bivariate analyses for covariates are reported in Table 2 and descriptive information for social functioning and SI by health condition can be found in Table 3. In the model that examined the impact of health conditions at T1 on social functioning 6 months later (T2), the total variance explained was 8.4%, F(7, 4466) = 59.62, p <.001. As shown in Table 4, hypothesis 2a was partially supported. Veterans with probable PTSD and at least one additional health condition had significantly worse social functioning compared to the no-PTSD group. Moreover, veterans identified as potentially having all three health conditions had, on average, a social functioning score that was .57 units lower than the no-PTSD group.

Table 2.

Descriptive Statistics and Correlations at 15-months post-Separation

Variables 1 2 3a 4 5b
1. PTSD - - - - -
2. Age −0.19** - - - -
3. Gender −0.13** 0.16** - - -
4. Social Functioning −0.43** 0.06* −0.03 - -
5. SI 0.39** −0.24 −0.02 −0.22** -
Mean (SD) 17.26 (8.14) 35.55 (8.88) 3.43 (0.86)
Percent 79.8% 71.1%
Open in a new tab

Note.

**

Correlation is significant at the 0.01 level (2-tailed).

*

Correlation is significant at the 0.05 level (2-tailed). PTSD = PTSD symptoms as measured by abbreviated PCL-5; Social Functioning = social functioning measure in the WBI; SI = item-9 from the Patient Health Questionnaire-9

a

= percentage of men

b

= percentage of veterans who endorsed SI

Table 3.

Social Functioning and SI by Health Condition (N = 5461)

Variable M(SD) Social
Functioning		 Suicidal Ideationa
T2 T3
yes no yes no
Health Condition
 No PTSD 4.00 (0.72) 3255 (79.3) 178 (37.8) 2927 (79.5) 201 (42.9)
 PTSD Only 3.53 (0.87) 71 (1.7) 32 (6.8) 72 (2.0) 25 (5.3)
 PTSD & Sleep Disturbance 3.67 (0.82) 93 (2.3) 38 (8.1) 77 (2.1) 34 (7.3)
 PTSD & CP 3.78 (0.83) 124 (3.0) 32 (6.8) 113 (3.1) 140 (3.4)
 PTSD, Sleep Disturbance, & CP 3.44 (0.87) 561 (13.7) 191 (40.6) 494 (13.4) 675 (16.3)
Open in a new tab

Note.

a

= number of participants (% of group); CP = Chronic Pain; bolded indicates that a Pearson Chi-Square test found that there is a statistically significant difference at a .05 level between those who do and do not endorse SI at that time point.

Table 4.

Multiple Linear Regression Analyses for Health Condition Predicting Social Functioning (H2a), controlling for age and gender

Model 1a Model 2b
Unstandardized
coefficient		 Standardized
coefficient		 t Unstandardized
coefficient		 Standardized
coefficient		 t
B SE 95% CI β B SE 95% CI β
No PTSD 3.81 0.05 3.71, 3.90 79.17 0.46 0.08 0.31, 0.60 0.25 6.01
PTSD Only −0.46 0.08 −0.60, −.31 −0.09 −6.01 3.35 0.09 3.18, 3.52 39.07
PTSD & SD −0.32 0.07 −0.45, −.19 −0.07 −4.72 0.14 0.10 0.05, 0.43 0.03 1.37
PTSD & CP −0.22 0.06 −0.34, −.10 −0.05 −3.49 0.24 0.10 0.06, 0.33 0.06 2.49
PTSD, SD, & CP −0.57 0.03 −0.63, −.51 −0.27 −18.31 −0.11 0.08 −0.27, 0.05 −0.05 −1.40
Open in a new tab

Note. Bolding indicates statistical significance ≥ 0.05 level (2-tailed). Boot SE=bootstrapped Standard Error; Boot LLCI=bootstrapped lower limit confident interval and Boot UCLI=bootstrapped upper limit confidence interval – these confidence intervals represent the 95% confidence intervals; OR = odds ratio; SD = sleep disturbance; CP = Chronic Pain

a

= Reference group is No PTSD

b

= Reference group is PTSD Only

When running the same model with PTSD-only as the reference group (see Table 4), the no-PTSD and PTSD-CP groups reported significantly better social functioning relative to those with PTSD-only. However, there were no significant differences between the PTSD-only and the PTSD-sleep disturbance group or those with all three conditions.

Chronic Health Conditions and Suicidal Ideation (H2b)

The logistic regression model examining the impact of reported health conditions (T1) on SI a year later (T3), while controlling for age and gender (T1), explained between 7.7% (Cox and Snell R squared) and 15.3% (Nagelkerke R squared) of the variance in SI χ2 (7, N= 3,745) =300.52, p <.001. As shown in Table 5, all veterans with probable PTSD were significantly more likely to endorse thoughts about death or self-harm compared to the no-PTSD group. The two groups with the largest increased likelihood of SI were those with PTSD and sleep disturbance and those with all three conditions. Veterans with reported PTSD and sleep disturbance were 6 times more likely to have thoughts of death or self-harm a year later compared to those without PTSD, and Veterans who reported all three conditions were 6.1 times more likely; thus, hypothesis 2b was partially supported.

Table 5.

Logistic Regression Analyses for Health Condition Predicting Social Functioning (H2b), controlling for age and gender

Model 1a Model 2b
B Boot SE Boot LLCI Boot ULCI B Boot SE Boot LLCI Boot ULCI
No PTSD −0.92 0.23 −1.49 0.26 0.14 0.38
PTSD Only 1.49 0.23 2.64 7.41 0.57 0.33
PTSD & SD 1.79 0.23 3.80 9.37 0.31 0.33 0.71 2.59
PTSD & CP 1.40 0.23 2.60 6.50 −0.10 0.33 0.47 1.75
PTSD, SD, & CP 1.80 0.12 4.80 7.83 −0.17 0.27 0.80 2.31
Open in a new tab

Note. Bolding indicates statistical significance ≥ 0.05 level (2-tailed). Boot SE=bootstrapped Standard Error; Boot LLCI=bootstrapped lower limit confident interval and Boot UCLI=bootstrapped upper limit confidence interval – these confidence intervals represent the 95% confidence intervals; OR = odds ratio; SD = sleep disturbance; CP = Chronic Pain

a

= Reference group is No PTSD

b

= Reference group is PTSD Only

In the same model with PTSD-only as the reference group there were no group differences between PTSD-only and those with additional chronic health conditions. However, veterans without PTSD were significantly less likely to endorse SI than those with only PTSD [OR = 0.23, p < .000].

Mediation Model (H3)

Given that (1) the previous models established that multiple health conditions predicted lower social functioning scores and increased risk for SI, and (2) in preliminary analyses, social functioning was negatively related to SI, our mediation model tested whether social functioning partially mediated the association between health conditions and SI controlling for age and gender (see Figure 1). Consistent with the previous results (H2a and H2b), all PTSD groups reported poorer social functioning and had a higher likelihood of endorsing SI compared to the no-PTSD group. Controlling for health conditions, higher social functioning scores were related to lower likelihood of SI (b=−.74). For every one-unit increase in social functioning, the odds of endorsing SI decreased by 53%.

Figure 1.

Figure 1

Open in a new tab

Mediation Model – No PTSD as reference group

Relative to the no-PTSD group, veterans who reported all three chronic health conditions were 4.3 times more likely to endorse SI as a result of lower social functioning scores. Veterans who reported PTSD were more likely than Veterans without PTSD to endorse SI a year later: PTSD-only group (OR = 3.13), PTSD-sleep disturbance group (OR =5.05), and the PTSD-pain group (OR = 3.55).

When the reference group was changed to PTSD-only (see Figure 2), we found that veterans with no PTSD and those with PTSD-CP experienced better social functioning, and those with no PTSD reported decreased likelihood of SI; there were no differences between PTSD-only, PTSD-sleep disturbance, and the three-condition groups (see Table 6).

Figure 2.

Figure 2

Open in a new tab

Mediation Model –PTSD-Only as reference group

Table 6.

Indirect effects in the mediation model (H3), controlling for age and gender

Model 1a Model 2b
b Boot SE Boot LLCI Boot ULCI OR b Boot SE Boot LLCI Boot ULCI OR
No PTSD −0.36 0.08 −0.54 −0.23 0.70
PTSD Only 0.36 0.08 0.22 0.52 1.44
PTSD & SD 0.25 0.06 0.13 0.38 1.28 −0.12 0.10 −0.31 0.06 0.89
PTSD & CP 0.14 0.06 0.04 0.26 1.16 −0.22 0.09 −0.40 −0.05 0.80
PTSD, SD, & CP 0.41 0.05 0.32 0.51 1.51 0.05 0.08 −0.11 0.20 1.05
Open in a new tab

Note. Bolding indicates statistical significance ≥ 0.05 level (2-tailed). Boot SE=bootstrapped Standard Error; Boot LLCI=bootstrapped lower limit confident interval and Boot UCLI=bootstrapped upper limit confidence interval – these confidence intervals represent the 95% confidence intervals; OR = odds ratio; SD = sleep disturbance; CP = Chronic Pain

a

= Reference group is No PTSD

b

= Reference group is PTSD Only

Discussion

Given the high suicide rate among post-9/11 military veterans (Maynard et al., 2018), a better understanding of the cumulative burden of multimorbidity in this population is essential for appropriate action to be deployed. While some research has examined the role of multiple health conditions in veterans’ risk for suicide, the additive or potential synergistic role of chronic mental and physical health conditions on SI is still unknown. The current study highlights the importance of considering veterans’ experiences with managing multiple chronic health conditions in both research and clinical care, as most veterans reported PTSD in combination with chronic pain or sleep disturbance. This finding is consistent with other research which has documented high prevalence of PTSD multimorbidity among post-9/11 veterans (e.g., Beckham et al., 1997; Lew et al., 2009; Plumb et al., 2014), and builds on that research by documenting the persistence of multimorbidity over time.

Veterans will all three health conditions reported the lowest social functioning relative to those without PTSD. However, as compared to those with PTSD alone, individuals with PTSD plus chronic pain (but without sleep disturbance) and individuals with PTSD plus sleep disturbance (but without chronic pain) did not demonstrate increased risk for poor social functioning. In fact, unexpectedly veterans with PTSD and CP reported slightly better social functioning than those with PTSD alone. This finding may relate to how veterans chose to seek support during their transition out of the military: while substantial research documents pain-invalidation associated with chronic pain (e.g., Nicola et al., 2021), during this period post-separation, veterans may initially be more willing to seek social support regarding somatic pain as it could be perceived by others as more understandable in the context of military service. More research is warranted to explore reasons why PTSD with CP, at least in the first three years post-separation, may be differentially related to social functioning.

The finding that PTSD multimorbidity was predictive of poorer social functioning and higher rates of SI, but that experiencing sleep disturbance and chronic pain did not confer greater risk for SI above that from PTSD, underscores the particularly pernicious effects of PTSD and the complexity of multimorbidity. These findings also highlight the potentially multiplicative effects of PTSD with additional chronic health conditions, and the urgent need to screen for and intervene with veterans identified as having PTSD as they transition out of the military.

While decreased social functioning has been found to be associated with chronic illness and SI, this is the first study to our knowledge to specifically examine the mediating role of social functioning in the relationship between multimorbidity and SI among recently separated veterans. In this study, we found that social functioning partially mediated the relationship between self-reported chronic health conditions and SI relative to veterans without PTSD, suggesting that impaired social functioning is one mechanism by which multimorbidity may increase risk for suicidal thoughts and behaviors. Certainly, the effect of self-identified multimorbidity is complex, and the increased burden associated with different health conditions is likely multifaceted. Thus, our findings emphasize the need for expanded attention to the role of multimorbidity in longitudinal research on veterans’ social functioning and SI.

More research is needed to address limitations of the current study and to replicate these findings in other samples and with other physical and mental health conditions (e.g., TBI, respiratory disorders, depression, substance use disorders). In particular, depression is highly comorbid with both PTSD and SI. Future research should assess the unique impact of Major Depressive Disorder when considering the relationship between multimorbidity and SI.

The primary limitation of this study was the use of single-item, self-report measures of chronic pain, sleep disturbance, and SI, and the fact that PTSD could not be confirmed in this self-report study. Although our broad approach to measuring possible PTSD in this study allowed us to capture information about individuals who report PTSD but whose experiences may not be well-captured by PTSD screeners, as well as those who have symptoms consistent with PTSD but do not view themselves as having this condition, we were not able to confirm whether these individuals had PTSD, which is only possible when clinician-administered measures are used. Future research should use clinician diagnosed PTSD, as well as examine severity of symptoms on functioning. In addition, although the PHQ-9 item 9 has strong sensitivity, research indicates that it has lower specificity relative to other measures of SI (e.g., Na et al., 2018). That is, most people who report thoughts of suicide also endorse thoughts about death and self-harm as assessed by item-9 of the PHQ-9, but many individuals who endorse item-9 will not ultimately engage in suicidal behavior. However, this measure does have the benefit of capturing initial thoughts that may emerge as an individual begins to grapple with difficult life circumstances and can help identify individuals who could benefit from additional support regardless of whether these thoughts translate into suicidal behavior.

Further research is also needed to account for the degree of pain interference as well as type and chronicity of pain on decisions to engage in social activities, feelings of connectedness, and the resulting impact on suicidal behaviors. In addition to providing stronger measures of individual diagnoses, assessment of symptoms and impairment on a continuum would allow for an examination of how these conditions interact with one another to influence veterans’ functioning and SI. Future research should also examine whether these associations vary for subgroups who experience particularly high rates of trauma-exposure, and are known to be at high risk for SI following separation from service (e.g., women, veterans of color, and those without access VA care). The relationship between multimorbidity and SI during other periods post-separation should also be explored. Although risk for maladjustment is higher at 3 years than at 1 year, the first year transitioning to civilian life may pose unique risk. Lastly, because health conditions may have been in place for some time before T1 and could have led to declines in social functioning that preceded the T2 assessment, we elected not to control for T1 social functioning in the current study. As such, research that can identify the onset of health conditions and more definitely confirm the causal impact of health conditions on changes in social functioning is needed.

This study provides several important clinical implications. Foremost is that clinicians treating post-9/11 veterans with PTSD should carefully assess comorbid pain and sleep disturbance. Our findings suggest that not only are recently separated veterans likely to experience at least one of these physical conditions in addition to PTSD, but that multimorbidity predicts poorer social functioning and, in turn, is associated with SI. Given the additional risk for SI associated with poor social functioning, clinicians should be mindful to not only support veterans’ efforts to seek social support, but also actively refer these individuals to veteran support systems (e.g., AmericaServes and Combined Arms). Moreover, as part of care coordination, clinicians need to monitor quality of support received, and integrate social functioning aims into treatment planning. Greater attention to multimorbidity is an important step in providing effective and comprehensive veteran-focused care.

Clinical Impact Statement:

This study determined that social functioning mediated the relationship between PTSD-multimorbidity (PTSD, chronic pain, and sleep disturbance) and suicidal ideation (SI) among 5,461 veterans who had recently separated from the military. PTSD-multimorbidity predicted poorer social functioning which, in turn, increased risk for SI. Given the additional risk for SI associated with poor social functioning, clinicians should be mindful to not only support veterans’ efforts to seek social support, but also monitor the quality of support received and integrate social functioning aims into treatment planning.

Footnotes

The opinions and assertions contained herein are the private views of the authors. No endorsement by any sponsor listed above is intended nor should any such endorsement be inferred.

1

No/negligible group differences between veterans that completed all three assessment and veterans that did not were found for health conditions, social functioning, or suicidal ideation based on T1 scores. There were no statistically significant differences for the following: endorsement of sleep problem or disorders χ2 (1, N = 5,461) = .14, p = .707; endorsement of chronic pain χ2 (1, N = 5,461) = .21, p = .643; PTSD χ2 (1, N = 5,461) = 1.32, p = .251; and social functioning (Completers: M = 3.87, SD = 0.79; Veterans lost to follow-up: M = 3.82, SD = 0.82; p = .111). Although there was a statistically significant difference on SI by attrition status, with veterans who endorsed suicidal ideation at T1 less likely to complete the follow-up assessments χ2 (1, N = 5,461) = 8.15, p = .004, this difference was negligible based on the effect size of .04 (phi), which does not meet the threshold for a small (i.e., clinically meaningful) effect size (phi = .10). Therefore, we do not believe that attrition biased the study results.

References

  1. Ahern J, Worthen M, Masters J, Lippman SA, Ozer EJ, Moos R The challenges of Afghanistan and Iraq veterans’ transition from military to civilian life and approaches to reconnection. PLoS One. 2015;10(7):e0128599. 10.1371/journal.pone.0128599 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Beckham JC, Crawford AL, Feldman ME, Kirby AC, Hertzberg MA, Davidson JRT, & Moore SD (1997). Chronic posttraumatic stress disorder and chronic pain in Vietnam combat veterans. Journal of psychosomatic research, 43(4), 379–389. 10.1016/S0022-3999(97)00129-3 [DOI] [PubMed] [Google Scholar]
  3. Calati R, Bakhiyi CL, Artero S, Ilgen M, & Courtet P (2015). The impact of physical pain on suicidal thoughts and behaviors: meta-analyses. Journal of psychiatric research, 71, 16–32. 10.1016/j.jpsychires.2015.09.004 [DOI] [PubMed] [Google Scholar]
  4. Department of Veterans Affairs. (2018). National Strategy for Preventing Veteran Suicide: 2018–2028. Washington, DC: US Department of Veterans Affairs. [Google Scholar]
  5. DiNapoli EA, Cinna C, Whiteman KL, Fox L, Appelt CJ, & Kasckow J (2016). Mental health treatment preferences and challenges of living with multimorbidity from the veteran perspective. International journal of geriatric psychiatry, 31(10), 1097–1104. 10.1002/gps.4550 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Durkheim E (1951). Suicide: A study in sociology. Glencoe, Illinois: Free Press [Google Scholar]
  7. Elbogen EB, Lanier M, Blakey SM, Wagner HR, & Tsai J (2021). Suicidal ideation and thoughts of self-harm during the COVID-19 pandemic: The role of COVID-19-related stress, social isolation, and financial strain. Depression and anxiety. 10.1002/da.23162 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Finley EP, Bollinger M, Noël PH, Amuan ME, Copeland LA, Pugh JA, Dassori A, Palmer R, Bryan C & Pugh MJV (2015). A national cohort study of the association between the polytrauma clinical triad and suicide-related behavior among US Veterans who served in Iraq and Afghanistan. American journal of public health, 105(2), 380–387. 10.2105/AJPH.2014.301957 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gradus JL, King MW, Galatzer-Levy I, & Street AE (2017). Gender differences in machine learning models of trauma and suicidal ideation in veterans of the Iraq and Afghanistan wars. Journal of traumatic stress, 30(4), 362–371. 10.1002/j.2162-6057.1984.tb00379.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hayes AF (2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford publications. [Google Scholar]
  11. Hoffmire CA, Kemp JE, & Bossarte RM (2015). Changes in suicide mortality for veterans and nonveterans by gender and history of VHA service use, 2000–2010. Psychiatric Services, 66, 959–965. 10.1176/appi.ps.201400031 [DOI] [PubMed] [Google Scholar]
  12. Hom MA, Hames JL, Bodell LP, Buchman-Schmitt JM, Chu C, Rogers ML, Chiurliza B, Michaels MS, Ribeiro JD, Nadroff MR, Winer ES, Lim IC, Rudd MD & Joiner TE (2017). Investigating insomnia as a cross-sectional and longitudinal predictor of loneliness: Findings from six samples. Psychiatry research, 253, 116–128. 10.1016/j.psychres.2017.03.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Johnson J, Wood AM, Gooding P, Taylor PJ, & Tarrier N (2011). Resilience to suicidality: The buffering hypothesis. Clinical psychology review, 31, 563–591. 10.1016/j.cpr.2010.12.007 [DOI] [PubMed] [Google Scholar]
  14. Joiner T (2015). Why people die by suicide. Cambridge, MA, US: Harvard University Press. [Google Scholar]
  15. Kang HK, Bullman TA, Smolenski DJ, Skopp NA, Gahm GA, & Reger MA (2015). Suicide risk among 1.3 million veterans who were on active duty during the Iraq and Afghanistan wars. Annals of epidemiology, 25(2), 96–100. 10.1016/j.annder.2008.09.021 [DOI] [PubMed] [Google Scholar]
  16. Kavalidou K, Smith DJ, & O’Connor RC (2017). The role of physical and mental health multimorbidity in suicidal ideation. Journal of affective disorders, 209, 80–85. 10.1016/j.jad.2016.11.026 [DOI] [PubMed] [Google Scholar]
  17. Kessler RC, Ormel J, Demler O, & Stang PE (2003). Comorbid mental disorders account for the role impairment of commonly occurring chronic physical disorders: results from the National Comorbidity Survey. Journal of occupational and environmental medicine, 45(12), 1257–1266. 10.1097/01.jom.0000100000.70011.bb [DOI] [PubMed] [Google Scholar]
  18. Kintzle S, Barr N, Corletto G, & Castro CA (2018, September). PTSD in US veterans: The role of social connectedness, combat experience and discharge. In Healthcare (Vol. 6, No. 3, p. 102). Multidisciplinary Digital Publishing Institute. 10.3390/healthcare6030102 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kline RB (2015). Principles and practice of structural equation modeling. Guilford publications. [Google Scholar]
  20. Kline A, Ciccone DS, Falca-Dodson M, Black CM, & Losonczy M (2011). Suicidal ideation among National Guard troops deployed to Iraq: The association with postdeployment readjustment problems. The Journal of nervous and mental disease, 199(12), 914–920. 10.1097/NMD.0b013e3182392917 [DOI] [PubMed] [Google Scholar]
  21. Kroenke K, Spitzer RL, & Williams JB (2001). The PHQ-9: validity of a brief depression severity measure. Journal of general internal medicine, 16(9), 606–613. 10.1046/j.1525-1497.2001.016009606.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Larkin J, Foley L, Smith SM, Harrington P, & Clyne B (2020). The experience of financial burden for people with multimorbidity: A systematic review of qualitative research. Health Expectations, 24, 282–295. 10.1111/hex.13166 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lew HL, Otis JD, Tun C, Kerns RD, Clark ME, & Cifu DX (2009). Prevalence of chronic pain, posttraumatic stress disorder, and persistent postconcussive symptoms in OIF/OEF veterans: polytrauma clinical triad. Journal of Rehabilitation Research & Development, 46, 697–702. 10.1682/JRRD.2009.01.0006 [DOI] [PubMed] [Google Scholar]
  24. López CM, Lancaster CL, Wilkerson A, Gros DF, Ruggiero KJ, & Acierno R (2019). Residual insomnia and nightmares postintervention symptom reduction among veterans receiving treatment for comorbid PTSD and depressive symptoms. Behavior therapy, 50(5), 910–923. 10.1016/j.beth.2019.01.006 [DOI] [PubMed] [Google Scholar]
  25. McLay RN, Klam WP, & Volkert SL (2010). Insomnia is the most commonly reported symptom and predicts other symptoms of post-traumatic stress disorder in US service members returning from military deployments. Military medicine, 175(10), 759–762. 10.7205/MILMED-D-10-00193 [DOI] [PubMed] [Google Scholar]
  26. Maynard C, Trivedi R, Nelson K, & Fihn SD (2018). Disability rating, age at death, and cause of death in US veterans with service-connected conditions. Military medicine, 183(11-12), e371–e376. 10.1093/milmed/usy040 [DOI] [PubMed] [Google Scholar]
  27. Na P, Yaramala S, Kim JA, Kim H, Goes F, Zandi P, Vande Voort J, Sutor B, Croarkin P, Bobo WV (2018). The PHQ-9 Item 9 based screening for suicide risk: A validation study of the Patient Health Questionnaire (PHQ) – 9 Item 9 with the Columbia Suicide Severity Rating Scale (C-SSRS). Journal of Affective Disorders, 232, 34–40. 10.1016/j.jad.2018.02.045 [DOI] [PubMed] [Google Scholar]
  28. Nicola M, Correia H, Ditchburn G, & Drummond P (2021). Invalidation of chronic pain: a thematic analysis of pain narratives. Disability and rehabilitation, 43(6), 861–869. 10.1080/09638288.2019.1636888 [DOI] [PubMed] [Google Scholar]
  29. Ouimette P, Vogt D, Wade M, Tirone V, Greenbaum MA, Kimerling R, Laffaye C, Fitt J, & Rosen CS (2011). Perceived barriers to care among veterans health administration patients with posttraumatic stress disorder. Psychological Services, 8(3), 212. 10.1037/a0024360 [DOI] [Google Scholar]
  30. Pagotto LF, Mendlowicz MV, Coutinho ESF, Figueira I, Luz MP, Araujo AX, & Berger W (2015). The impact of posttraumatic symptoms and comorbid mental disorders on the health-related quality of life in treatment-seeking PTSD patients. Comprehensive psychiatry, 58, 68–73. 10.1016/j.comppsych.2015.01.002 [DOI] [PubMed] [Google Scholar]
  31. Panagioti M, Gooding PA, & Tarrier N (2012). A meta-analysis of the association between posttraumatic stress disorder and suicidality: the role of comorbid depression. Comprehensive psychiatry, 53, 915–930. 10.1016/j.comppsych.2012.02.009 [DOI] [PubMed] [Google Scholar]
  32. Pigeon WR, Pinquart M, & Conner K (2012). Meta-analysis of sleep disturbance and suicidal thoughts and behaviors. The Journal of clinical psychiatry, 73(9), 1160–1167. 10.4088/JCP.11r07586 [DOI] [PubMed] [Google Scholar]
  33. Plumb TR, Peachey JT, & Zelman DC (2014). Sleep disturbance is common among servicemembers and veterans of Operations Enduring Freedom and Iraqi Freedom. Psychological services, 11(2), 209. 10.1037/a0034958 [DOI] [PubMed] [Google Scholar]
  34. Pompili M, Sher L, Serafini G, Forte A, Innamorati M, Dominici G, Giovanni M, Lester D Amore M & Girardi P (2013). Posttraumatic stress disorder and suicide risk among veterans: a literature review. The Journal of nervous and mental disease, 201(9), 802–812. 10.1097/NMD.0b013e3182a21458 [DOI] [PubMed] [Google Scholar]
  35. Price M, Szafranski DD, van Stolk-Cooke K, & Gros DF (2016). Investigation of abbreviated 4 and 8 item versions of the PTSD Checklist 5. Psychiatry research, 239, 124–130. 10.1016/j.psychres.2016.03.014 [DOI] [PubMed] [Google Scholar]
  36. Racine M (2018). Chronic pain and suicide risk: A comprehensive review. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 87, 269–280. 10.1016/j.pnpbp.2017.08.020 [DOI] [PubMed] [Google Scholar]
  37. Reger MA, Smolenski DJ, Skopp NA, Metzger-Abamukang MJ, Kang HK, Bullman TA, & Gahm GA (2015). Risk of suicide among US military service members following Operation Enduring Freedom or Operation Iraqi Freedom deployment and separation from the US military. JAMA psychiatry, 72, 561–569. 10.1001/jamapsychiatry.2014.3195 [DOI] [PubMed] [Google Scholar]
  38. Richardson LK, Frueh BC, & Acierno R (2010). Prevalence estimates of combat-related post-traumatic stress disorder: critical review. Australian & New Zealand Journal of Psychiatry, 44(1), 4–19. 10.3109/00048670903393597 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sall J, Brenner L, Milllikan Bell A, & Colston M (2019). Assessment and management of patients at risk for suicide: Synopsis of the 2019 U.S. Department of Veterans Affairs and U.S. Department of Defense clinical practice guidelines. Annals of Internal Medicine, 171 (5), 343–355. 10.7326/M19-0687 [DOI] [PubMed] [Google Scholar]
  40. Stickley A, Koyanagi A, Ueda M, Inoue Y, Waldman K, & Oh H (2020). Physical multimorbidity and suicidal behavior in the general population in the United States. Journal of affective disorders, 260, 604–609. 10.1016/j.jad.2019.09.042 [DOI] [PubMed] [Google Scholar]
  41. US Department of Veterans Affairs. (2019). 2019 National veteran suicide prevention annual report.
  42. U.S. Government Accountability Office (GAO). Transitioning Servicemembers: Information on Military Employment Assistance Centers. Washington, DC: 2019. (GAO-19-438R; ). Retrieved from https://www.gao.gov/assets/700/699769.pdf. [Google Scholar]
  43. Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby EA, & Joiner TE Jr (2010). The interpersonal theory of suicide. Psychological review, 117(2), 575–600. 10.1037/a0018697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Vogt D, Perkins DF, Copeland LA, Finley EP, Jamieson CS, Booth B, Lederer S & Gilman CL (2018). The Veterans Metrics Initiative study of US veterans’ experiences during their transition from military service. BMJ open, 8(6), e020734. 10.1136/bmjopen-2017-020734 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wang HE, Campbell-Sills L, Kessler RC, Sun X, Heeringa SG, Nock MK, Ursano RJ, Jain S & Stein MB (2019). Pre-deployment insomnia is associated with post-deployment post-traumatic stress disorder and suicidal ideation in US Army soldiers. Sleep, 42, 1–9. 10.1093/sleep/zsy229 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Wisco BE, Marx BP, Wolf EJ, Miller MW, Southwick SM, & Pietrzak RH (2014). Posttraumatic stress disorder in the US veteran population: results from the National Health and Resilience in Veterans Study. The Journal of clinical psychiatry, 75, 1338–1346. 10.4088/JCP.14m09328 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES