Abstract
Objective:
This study compared rates of suicide attempt (SA) and ideation (SI) during the first 5 years after traumatic brain injury (TBI) among Veterans and Service Members (V/SMs) in the Veterans Affairs (VA) and the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) Model Systems national databases to each other and to non-Veterans (non-Vs) in the NIDILRR database.
Setting:
21 NIDILRR and 5 VA TBI Model Systems (TBIMS) inpatient rehabilitation facilities in the U.S.
Participants:
Participants with TBI were discharged from rehabilitation alive, had a known military status recorded (either non-V or history of military service), and successful 1-, 2-, and/or 5-year follow-up interviews completed between 2009–2021. The Year 1 cohort included 8,737 unique participants (8,347 with SA data and 3,987 with SI data); the Year 2 (7,628 participants) and Year 5 (4,837 participants) cohorts both had similar demographic characteristics to the Year 1 cohort.
Design:
Longitudinal design with data collected across TBIMS centers at 1, 2, and 5 years post-injury.
Main Outcomes and Measures:
History of SA in past year and SI in past 2 weeks assessed by the Patient Health Questionnaire-9 (PHQ-9). Patient demographics, injury characteristics, and rehabilitation outcomes were also assessed.
Results:
Full sample rates of SA were 1.9%, 1.5%, and 1.6%, and rates of SI were 9.6%, 10.1%, and 8.7% (respectively at Years 1, 2, and 5). There were significant differences among groups based on demographic, injury-related, mental/behavioral health, and functional outcome variables. Characteristics predicting SA/SI related to mental health history, substance use, younger age, lower functional independence, and greater levels of disability.
Conclusions:
Compared to participants with TBI in the NIDILRR system, higher rates of SI among V/SMs with TBI in the VA system appear associated with risk factors observed within this group, including mental/behavioral health characteristics and overall levels of disability.
Keywords: traumatic brain injury, suicide, suicidal ideation, suicide attempt, Veterans and Service Members, non-Veterans
Introduction
Individuals with traumatic brain injury (TBI) are at elevated risk for suicidal ideation (SI), suicide attempts (SA), and death by suicide.1–3 Individuals with TBI are 2–4 times more likely to die by suicide than the general population.4–6 Reports of lifetime rates of SA after TBI have ranged from 8%–26%, with 1.6% of individuals reporting SA in the past year.1,7,8 Reported rates of SI after TBI range from 7–33%, with 25% of individuals experiencing SI in the first year after injury.9–11 Variability in methodological approaches to investigating suicide-related outcomes after TBI (e.g., measures/assessments, definition of terms, characterization of participants) have yielded mixed findings regarding risk factors, but typically implicate a range of sociocultural, behavioral and physical health, injury-related, and functional variables.1,12–14 Pre-injury history of depression and/or SA increase risk of post-TBI suicidal thoughts and behaviors; post-injury modifiable risk factors include comorbid depression, anxiety, and substance use disorders.13,15,16 Veterans with TBI are 1.55–2.19 times more likely to die by suicide than those without TBI.17,18 Further, among post-9/11 Veterans, the highest rates of suicide and accidental deaths occur among those with service-related moderate and severe TBI19, and history of TBI has also been linked to elevated risk for SI among Veterans and Service Members (V/SMs) with post-deployment diagnosis of major depression.20 Post-traumatic stress disorder (PTSD), the largest mediating factor in the association between TBI and suicide among V/SMs, can further worsen risk of self-directed violence and negatively influence potential recovery from injury and illness.21
The Veterans Affairs (VA) health care system prioritizes initiatives to assess, identify, and prevent both TBI and suicidal behavior.22 Recognizing that V/SMs may be served in a variety of health care systems and practice settings, the VA’s National Strategy for Preventing Veteran Suicide prioritizes research regarding risk and protective factors for suicide among V/SMs across multiple sectors, implementing treatment and support services, and use of predictive analytics to identify individuals at heightened risk.23 These priorities are underscored by a recent review of suicide-related risk factors among people with TBI that emphasized the challenges associated with comparing studies conducted with either military or non-Veteran (non-V) samples.13 The review’s authors highlight that military samples tend to include participants primarily with a history of mild TBI, whereas non-V samples represent a broader range of injury severities; moreover, considering these samples separately limits understanding of how demographic and behavioral health factors might differentially contribute to suicide risk across groups.13
These gaps in the literature can be addressed by leveraging the combined TBI Model Systems (TBIMS) National Databases of both the VA Polytrauma Rehabilitation Centers (PRCs) and the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) to examine factors associated with suicide risk among a well-characterized cohort of people with a history of TBI. Combining data from these two studies will allow for a comparison of V/SMs served in multiple sectors (i.e., PRCs in the VA healthcare system and inpatient rehabilitation centers [IRFs] in non-V healthcare systems) with a broad range of injury severities. This exploratory study had two aims: (1) to compare rates of SI and SA cross-sectionally during the first 5 years after TBI among V/SMs enrolled in the VA PRC or the NIDILRR TBIMS studies to each other and to non-Vs in the NIDILRR study, and (2) to examine risk factors that might explain observed differences in rates of SA/SI across groups, including sociodemographic variables, pre-injury mental health history, injury-related variables, functional recovery outcomes, and behavioral health functioning at follow-up.
Method
Participants
Institutional Review Board approval to conduct research with human participants was obtained by each of the 26 sites that contributed data to this project. Participants who met the following criteria were selected from the NIDILRR and VA PRC TBIMS national databases: (a) discharged from rehabilitation alive, (b) known military status recorded in the NIDILRR database, and (c) successful 1-, 2-, and/or 5-year follow-up interview completed between 01Jul2009 and 30Sep2021. Non-missing data were required for each outcome (SA and SI) for a participant to be included in the respective analysis. Three primary cohorts were identified for cross-sectional analyses, representing participants interviewed at each follow-up period. Sample sizes within each cohort depended on the number of participants with non-missing outcome data; thus, there were six total cohorts across the three follow-up years and two outcomes (see eFigure 1). The Year 1 cohort included 8,737 participants (8,347 with SA data and 3,987 with SI data). The Year 2 cohort included 7,628 eligible participants (7,237 with SA data and 3,521 with SI data). The Year 5 cohort included 4,937 eligible participants (4,707 with SA data and 2,505 with SI data).
Measures
The two outcome measures for this study were SA and SI. For SA, participants (or their proxy) were asked at follow up if they had attempted suicide in the past year (yes or no). To assess SI, participants (not their proxy) were administered the Patient Health Questionnaire-9 (PHQ-9)24 at follow-up interview with the introduction “Over the past 2 weeks, how often have you been bothered by the following problems...?” Scores of 1 (several days) to 3 (nearly every day) on the 9th item “thoughts that you would be better off dead, or of hurting yourself in some way” were coded as yes for SI and scores of 0 (not at all) as no. Cumulative SA and SI outcomes at Year 2 and Year 5 (yes or no) were coded as yes if the participant indicated SA (or SI) at Years 1 or 2 and Years 1, 2, or 5, respectively, and no otherwise, even those with only one follow-up time point. Of note, the PHQ-9 was only administered directly to participants (i.e., data were not obtained by proxy report, as was done with SA) and this instrument was not administered to participants in the NIDILRR database between Oct2013 and Oct2017, resulting in high missing data counts of SI for these cohorts.
Our primary independent group variable indicated if the participant was a non-V enrolled in the NIDILRR study, a V/SM enrolled in the NIDILRR study (i.e., a V/SM who completed inpatient rehabilitation in a NIDILRR health care system), or a V/SM enrolled in the VA PRC study (i.e., a V/SM who completed inpatient rehabilitation in the VA/PRC health care system). Additional participant characteristics extracted from medical records or reported by patient/proxy include age at injury, sex (male or female), minority status (non-Hispanic white or not), violence as a cause of injury, highest level of education at follow-up (greater than high school or not), employment status at follow-up (competitively employed or not), marital status at follow-up (married or not), pre-injury history of mental health treatment or psychiatric hospitalization (yes or no), and pre-injury history of SA (yes or no). Injury recovery and rehabilitation characteristics included measures of functional independence at follow-up (Functional Independence Measure [FIM] Motor and Cognitive domain scores)25, severity of injury indicated by duration of posttraumatic amnesia (based on Departments of Defense and Veterans Affairs26 consensus-based classification as mild [0 days], moderate [1–7 days], or severe [>7 days]), illicit or non-prescription drug use in the year prior to follow-up (yes or no), alcohol use in the 30 days prior to follow-up (coded using Centers for Disease Control classifications27 as abstaining, light, moderate, heavy), and Glasgow Outcome Scale – Extended (GOSE)28 at follow-up (coded as vegetative/severe disability, moderate disability, or good recovery).
Statistical analyses
All analyses were conducted using SAS v.9.4 assuming a significance level of α=0.05 unless otherwise indicated. We conducted analyses for each outcome and within each follow-up cohort separately using the same process discussed below. First, we described the sample characteristics and assessed for significant differences in these characteristics among groups (NIDILRR non-V, NIDILRR V/SM, PRC V/SM) using ANOVA, Wilcoxon-Rank sum tests, and chi-square tests. Next, we estimated SA and SI rates at each cross-sectional time point using frequency counts and percentages, and these (unadjusted) rates were compared among groups using chi-square tests. The unadjusted effect sizes were estimated using logistic regression to compare groups and were quantified as odds ratios (ORs), expressing the odds of SA (or SI) versus no SA (or no SI) for one group relative to another. We then examined the bivariate relationships between each of the participant characteristics and each outcome using chi-square tests and logistic regression to identify additional characteristics associated with differential rates in SA or SI. Finally, we fit multiple logistic regression models to assess for differences in SA and SI rates between groups after controlling for all characteristics found to differ significantly among groups and/or have a significant bivariate relationship with the outcome. Whenever possible, we used the concurrent version of a covariate (e.g., drug or alcohol use) when examining SA or SI rates at a particular time point rather than the baseline version of the covariate. However, many variables only had a baseline version (e.g., pre-injury mental health treatment or hospitalization) and no parallel follow-up version. To keep the models consistent across outcomes and follow-up years for comparative purposes, we included a participant characteristic in all cohort models if it was different among groups or bivariately associated with either outcome for any cohort. From these models, we estimated adjusted odds ratios for the V/SMs group variable and the covariates along with 95% confidence intervals. A Bonferroni correction was used to adjust for multiple comparisons.
Results
Sample characteristics and group comparisons
We compared the characteristics of each group (NIDILRR non-Vs, NIDILRR V/SMs, and PRC V/SMs) across follow-up cohorts (Years 1, 2, and 5) among both outcomes (SA/SI). The sample characteristics by group for the Year 1 SA and SI cohorts are summarized in Table 1. The Year 2 and Year 5 cohorts for both outcomes had similar distributions as the Year 1 cohorts and are presented as eTable 1 and eTable 2, respectively.
Table 1:
Distribution of Sample Characteristics and Differences among Groups for Year 1 SA and SI Cohorts
| Suicide Attempts Year 1 (N = 8347) | Suicidal Ideation Year 1 (N = 3987) | |||||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Characteristic | NIDILRR Non-Vs (N = 6590) |
NIDILRR V/SM (N = 963) |
PRC V/SM (N = 794) |
p-value | NIDILRR Non-Vs (N = 2942) |
NIDILRR V/SM (N = 431) |
PRC V/SM (N = 614) |
p-value |
|
| ||||||||
| Age at Injury, Mean (SD) | 42.5 (19.3) | 60.1 (18.9) | 36.7 (16.0) | <0.001 * | 40.9 (18.5) | 58.3 (18.5) | 37.2 (16.5) | <0.001 * |
|
| ||||||||
| Sex, % Male | 69.9% | 97.4% | 94.0% | <0.001 * | 70.0% | 97.2% | 93.2% | <0.001 * |
|
| ||||||||
| Minority, % Yes | 38.1% | 23.6% | 34.4% | <0.001 * | 35.4% | 23.0% | 35.8% | <0.001 * |
|
| ||||||||
| Education Level at FU, % > HS | 49.6% | 61.9% | 61.6% | <0.001 * | 51.9% | 64.9% | 64.2% | <0.001 * |
|
| ||||||||
| Employment Status at FU, % Employed | 27.8% | 22.2% | 23.4% | <0.001 * | 32.1% | 25.3% | 27.4% | 0.003 * |
|
| ||||||||
| Marital Status at FU, % Married | 31.2% | 55.3% | 38.4% | <0.001 * | 32.2% | 54.4% | 38.7% | <0.001 * |
|
| ||||||||
| Cause of Injury, % Violent | 9.4% | 6.3% | 22.3% | <0.001 * | 8.6% | 6.5% | 21.4% | <0.001 * |
|
| ||||||||
| Pre-Injury Mental Health, % Yes | 23.0% | 18.5% | 36.1% | <0.001 * | 23.9% | 16.4% | 33.9% | <0.001 * |
|
| ||||||||
| Pre-Injury Suicide Attempt, % Yes | 5.2% | 3.3% | 8.7% | <0.001 * | 5.3% | 3.5% | 8.4% | 0.002 * |
|
| ||||||||
| FIM Motor at FU, Median (IQR) | 89 (81, 91) | 88 (80, 91) | 89 (81, 91) | <0.001 * | 89 (85, 91) | 89 (83, 91) | 90 (86, 91) | <0.001 * |
|
| ||||||||
| FIM Cognitive at FU, Median (IQR) | 32 (29, 34) | 33 (30, 34) | 32 (28, 34) | 0.029 * | 33 (30, 34) | 33 (31, 34) | 33 (31, 34) | 0.039 * |
|
| ||||||||
| PTA Russell Classification, % | <0.001 * | <0.001 * | ||||||
| Mild (0 days) | 8.4% | 15.8% | 22.3% | † | 8.8% | 17.9% | 26.3% | † |
| Moderate (1–7 days) | 18.2% | 19.0% | 10.8% | † | 20.2% | 19.1% | 12.4% | † |
| Severe (> 7 days) | 73.5% | 65.2% | 66.9% | † | 71.0% | 62.9% | 61.2% | † |
|
| ||||||||
| Drugs at FU, % Yes | 12.6% | 5.9% | 7.5% | <0.001 * | 14.0% | 6.7% | 8.6% | <0.001 * |
|
| ||||||||
| Drinking Category at FU, % | 0.208 ns | 0.015 * | ||||||
| Abstaining | 62.6% | 63.1% | 62.1% | 58.0% | 60.6% | 54.3% | ||
| Light | 19.7% | 18.6% | 19.8% | 23.5% | 18.8% | 24.0% | † | |
| Moderate | 13.6% | 15.6% | 15.0% | 14.5% | 18.6% | 17.9% | † | |
| Heavy | 4.1% | 2.7% | 3.2% | 3.9% | 2.1% | 3.8% | † | |
|
| ||||||||
| GOSE at FU, % | <0.001 * | <0.001 * | ||||||
| Vegetative/Severe Disability | 33.1% | 32.4% | 35.9% | 23.4% | 24.9% | 24.7% | ||
| Moderate Disability | 31.9% | 27.8% | 38.8% | † | 37.4% | 29.4% | 44.8% | † |
| Good Recovery | 35.0% | 39.8% | 25.3% | † | 39.2% | 45.7% | 30.5% | † |
Abbreviations: SA, suicide attempt; SI, suicidal ideation; NIDILRR, National Institute on Disability, Independent Living, and Rehabilitation research; Non-V, Non-Veteran; V/SM, Veteran / Service Member; PRC, Polytrauma Rehabilitation Center; FU, follow-up; HS, high school; PTA, post-traumatic amnesia; FIM, Functional Independence Measure; GOSE, Glasgow outcome scale – extended
statistically significant at α = 0.05 for continuous variables and global tests for categorical variables
not statistically significance at α = 0.05 for continuous variables and global tests for categorical variables
cell chi-square values > 2 indicating levels with large differences contributing to global differences
Within each cohort, there were significant group differences for nearly all characteristics. The V/SMs in the NIDILRR and PRC groups were predominately male and at follow up had higher education and lower levels of employment than non-Vs. The V/SMs in the NIDILRR group tended to be older, and the V/SMs in the PRC group tended to be younger. The non-Vs had the highest rates of substance use at Year 1 follow-up; however, beyond Year 1 there were no differences among groups regarding alcohol use. Based on duration of PTA, the non-Vs had the highest rates of severe TBI; the V/SMs in the PRC group had the highest rates of mild TBI and lowest rates of moderate TBI. The PRC V/SMs had the highest rates of injury sustained by violence. Regarding functional outcomes, the PRC V/SMs had the highest rates of moderate disability and lowest rates of good recovery, as measured by the GOSE. The NIDILRR V/SMs had the lowest rates of moderate disability and the highest rates of good overall recovery.
Rates of SA and SI across groups at each follow-up year
Irrespective of group, the rates of SA were 1.87% at Year 1, 1.46% at Year 2, and 1.59% at Year 5. Similarly, the rates of SI were 9.63% at Year 1, 10.05% at Year 2, and 8.74% at Year 5. The distributions of SA and SI by group at each cross-sectional year is presented in Figure 1 (and Table 2), along with cumulative SA and SI rates for Years 2 and 5. Based on chi-square tests, there was a significant difference in rates of SA among groups at Year 5, and there were significant differences in rates of SI at each follow-up year, with PRC V/SMs having the highest rates.
Figure 1.
e Figure 1. SA and SI Rates within Cohort by Year
Table 2:
Bivariate Relationships between Sample Characteristics and Suicide Attempts and Suicidal Ideation Outcomes
| Suicide Attempts | Suicidal Ideation | |||||
|---|---|---|---|---|---|---|
|
| ||||||
| Year 1 | Year 2 | Year 5 | Year 1 | Year 2 | Year 5 | |
|
| ||||||
| Overall (Baseline Rate) | 1.87 | 1.46 | 1.59 | 9.63 | 10.05 | 8.74 |
|
| ||||||
|
Percent
(95% CI) |
Percent
(95% CI) |
Percent
(95% CI) |
Percent
(95% CI) |
Percent
(95% CI) |
Percent
(95% CI) |
|
|
| ||||||
| Group | ns | ns | * | * | * | * |
| NIDILRR Non-V | 1.93 (0.16, 2.26) |
1.43 (1.12, 1.74) |
1.38 (1.01, 1.76) |
9.08 (8.04, 10.11) |
9.37 (8.25, 10.49) |
7.45 (6.24, 8.65) |
| NIDILRR V/SM | 1.04 (0.40, 1.68) |
1.36 (0.56, 2.16) |
1.47 (0.39, 2.56) |
8.12 (5.54, 10.70) |
8.33 (5.32, 11.34) |
9.17 (5.34, 13.01) |
| PRC V/SM | 2.39 (1.33, 3.46) |
1.82 (0.88, 2.77) |
3.14 (1.67, 4.60) |
13.36 (10.66, 16.05) |
13.84 (11.11, 16.58) |
13.67 (10.53, 16.80) |
|
| ||||||
| Categorical Characteristics | Percent | Percent | Percent | Percent | Percent | Percent |
|
| ||||||
| Sex | ns | ns | ns | ns | ns | ns |
| Male | 1.81 | 1.33 | 1.54 | 9.91 | 10.24 | 8.88 |
| Female | 2.05 | 1.89 | 1.75 | 8.75 | 9.46 | 8.32 |
|
| ||||||
| Minority | * | * | ns | ns | * | ns |
| Yes | 2.36 | 1.91 | 1.57 | 10.24 | 11.50 | 10.19 |
| No | 1.60 | 1.22 | 1.58 | 9.31 | 9.27 | 8.01 |
|
| ||||||
| Education at FU | * | * | ns | ns | ns | * |
| HS or less | 2.39 | 1.85 | 1.66 | 10.48 | 10.95 | 10.22 |
| > HS | 1.38 | 1.11 | 1.55 | 8.96 | 9.27 | 7.87 |
|
| ||||||
| Employment at FU | ns | * | ns | * | * | * |
| Employed | 1.80 | 0.81 | 1.25 | 6.08 | 7.48 | 4.91 |
| Not Employed | 1.90 | 1.69 | 1.78 | 11.22 | 11.33 | 11.56 |
|
| ||||||
| Marital Status at FU | * | * | ns | * | * | * |
| Married/Other | 0.93 | 0.65 | 1.28 | 7.76 | 7.46 | 5.80 |
| Single/Divorced/Widowed | 2.37 | 1.89 | 1.77 | 10.64 | 11.44 | 10.49 |
|
| ||||||
| Cause of Injury | ns | ns | * | ns | ns | ns |
| Violent | 2.69 | 1.56 | 2.86 | 10.46 | 10.54 | 10.88 |
| Not Violent | 1.78 | 1.46 | 1.44 | 9.56 | 10.00 | 8.48 |
|
| ||||||
| Pre-Injury Mental Health | * | * | * | * | * | * |
| Yes | 3.40 | 2.52 | 3.45 | 16.39 | 15.53 | 12.64 |
| No | 1.41 | 1.15 | 1.08 | 7.42 | 8.23 | 7.50 |
|
| ||||||
| Pre-Injury Suicide Attempt | * | * | * | * | * | * |
| Yes | 10.02 | 4.23 | 6.76 | 26.24 | 25.68 | 22.41 |
| No | 1.40 | 1.31 | 1.35 | 8.55 | 9.16 | 8.00 |
|
| ||||||
| PTA Russell Classifications | ns | ns | ns | ns | ns | * |
| Mild TBI (0 days) | 1.58 | 1.00 | 1.65 | 8.22 | 9.64 | 8.97 |
| Moderate TBI (1–7 days) | 1.49 | 1.26 | 1.53 | 8.24 | 8.03 | 4.70 |
| Severe TBI (> 7 days) | 2.02 | 1.56 | 1.58 | 10.33 | 10.83 | 9.18 |
|
| ||||||
| Drugs at FU | * | * | * | * | * | * |
| Yes | 5.24 | 3.81 | 3.92 | 16.06 | 18.74 | 13.85 |
| No | 1.42 | 1.10 | 1.17 | 8.75 | 8.63 | 7.63 |
|
| ||||||
| Drinking Category at FU | * | ns | * | ns | * | * |
| Abstaining | 1.82 | 1.44 | 1.63 | 9.44 | 10.12 | 9.30 |
| Light | 1.63 | 1.19 | 1.48 | 9.66 | 8.88 | 7.72 |
| Moderate | 1.50 | 1.30 | 0.92 | 8.36 | 9.43 | 6.53 |
| Heavy | 4.47 | 2.68 | 3.69 | 15.07 | 16.85 | 14.12 |
|
| ||||||
| GOSE at FU | * | * | * | * | * | * |
| Veg/Severe Disability | 2.35 | 1.67 | 2.28 | 14.33 | 16.22 | 16.89 |
| Moderate Disability | 2.22 | 2.08 | 1.84 | 12.02 | 11.75 | 10.65 |
| Good Recovery | 1.15 | 0.50 | 0.96 | 4.17 | 5.09 | 3.29 |
|
| ||||||
| Continuous Characteristics | Odds Ratio | Odds Ratio | Odds Ratio | Odds Ratio | Odds Ratio | Odds Ratio |
| Age at Injury, 10 year increase | 0.677 * | 0.752 * | 0.673 * | 0.915 * | 0.888 * | 0.882 * |
| FIM Motor at FU, 5 unit increase | 1.012 ns | 1.032 ns | 1.031 ns | 0.873 * | 0.878* | 0.881 * |
| FIM Cognitive at FU, 5 unit increase | 0.870 * | 0.858 ns | 0.704 * | 0.554 * | 0.510 * | 0.429 * |
Abbreviations: SA, suicide attempt; SI, suicidal ideation; NIDILRR, National Institute on Disability, Independent Living, and Rehabilitation research; Non-V, Non-Veteran; V/SM, Veteran / Service Member; PRC, Polytrauma Rehabilitation Center; FU, follow-up; HS, high school; PTA, post-traumatic amnesia; FIM, Functional Independence Measure; GOSE, Glasgow outcome scale – extended
statistically significant at α = 0.05 for continuous variables and global tests for categorical variables
not statistically significance at α = 0.05 for continuous variables and global tests for categorical variables.
The unadjusted odds ratios comparing groups are presented in the upper parts of Tables 3 and 4; the cross-sectional data for each follow-up year are presented in Table 3, and the cumulative data for Years 2 and 5 are presented in Table 4. At Year 1, PRC V/SMs had greater odds of SA compared to NIDILRR V/SMs (OR=2.34), however no differences among groups were significant after controlling for multiple comparisons. At Year 5, PRC V/SMs had greater odds of SA compared to NIDILRR non-Vs (OR=2.31). The cumulative odds of SA by Year 5 were also greater for PRC V/SMs compared to NIDILRR non-Vs (OR=1.56) but did not remain significant after adjusting for multiple comparisons.
Table 3:
Unadjusted and Adjusted Odds Ratios for Suicide Attempts and Suicidal Ideation
| Suicide Attempts | Suicidal Ideation | ||||||
|---|---|---|---|---|---|---|---|
|
| |||||||
| Year 1 | Year 2 | Year 5 | Year 1 | Year 2 | Year 5 | ||
|
| |||||||
| Unadjusted Effect | Comparison | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) |
|
| |||||||
| Group | NIDILRR V/SM vs NIDILRR Non-V | 0.53 (0.28, 1.02) |
0.95 (0.50, 1.79) |
1.07 (0.48, 2.37) |
0.89 (0.61, 1.28) |
0.88 (0.58, 1.33) |
1.26 (0.77, 2.05) |
| PRC V/SM vs NIDILRR Non-V | 1.25 (0.77, 2.03) |
1.28 (0.72, 2.27) |
2.31 b (1.32, 4.03) |
1.54 b (1.19, 2.01) |
1.55 b (1.19, 2.03) |
1.97 b (1.43, 2.70) |
|
| PRC V/SM vs NIDILRR V/SM | 2.34 a (1.08, 5.05) |
1.35 (0.61, 2.98) |
2.16 (0.89, 5.27) |
1.74 b (1.15, 2.65) |
1.77 b (1.12, 2.79) |
1.57 (0.92, 2.67) |
|
|
| |||||||
| Adjusted Effect | Comparison | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) |
|
| |||||||
| Group | NIDILRR V/SM vs NIDILRR Non-V | 1.36 (0.65, 2.83) |
1.90 (0.89, 4.05) |
2.54 (0.99, 6.50) |
1.25 (0.82, 1.91) |
1.09 (0.67, 1.78) |
1.69 (0.88, 3.24) |
| PRC V/SM vs NIDILRR Non-V | 1.42 (0.77, 2.61) |
0.72 (0.27, 1.89) |
1.75 (0.75, 4.06) |
1.41 (0.99, 2.00) |
1.37 (0.94, 1.98) |
1.10 (0.64, 1.88) |
|
| PRC V/SM vs NIDILRR V/SM | 1.04 (0.43, 2.54) |
0.38 (0.12, 1.20) |
0.69 (0.22, 2.16) |
1.13 (0.68, 1.86) |
1.25 (0.71, 2.20) |
0.65 (0.30, 1.39) |
|
|
| |||||||
| Covariate | OR | OR | OR | OR | OR | OR | |
|
| |||||||
| Sex | Female vs Male | 1.37 | 1.50 | 1.39 | 0.88 | 0.94 | 1.10 |
|
| |||||||
| Minority | Minority vs Non-Minority | 1.16 | 1.55 | 1.10 | 1.03 | 1.27 | 1.34 |
|
| |||||||
| Education at FU | ≤ HS vs > HS | 1.56 b | 1.14 | 0.99 | 1.02 | 0.98 | 1.03 |
|
| |||||||
| Employment at FU | Not Employed vs Employed | 0.68 | 1.71 | 0.85 | 1.18 | 0.85 | 1.41 |
|
| |||||||
| Marital at FU | Not Married vs Married | 1.12 | 1.20 | 0.62 | 1.25 | 1.08 | 1.67 b |
|
| |||||||
| Cause | Not Violent vs Violent | 1.21 | 1.71 | 0.79 | 1.40 | 1.38 | 1.14 |
|
| |||||||
| Pre-Injury Mental Health | Yes vs No | 1.50 | 1.59 | 1.76 | 1.72 b | 1.45 b | 0.93 |
|
| |||||||
| Pre-Injury Suicide Attempt | Yes vs No | 4.20 b | 2.10 b | 3.11 b | 2.41 b | 2.65 b | 2.39 b |
|
| |||||||
| Age at Injury | 10 Year Increase | 0.73 b | 0.75 b | 0.60 b | 0.95 | 0.91 b | 0.92 |
|
| |||||||
| FIM Motor at FU | 5 Unit Increase | 1.09 b | 1.09 | 1.17 b | 0.94 b | 0.91 b | 1.03 |
|
| |||||||
| FIM Cognitive at FU | 5 Unit Increase | 0.84 | 0.88 | 0.56 b | 0.69 b | 0.57 b | 0.57 b |
|
| |||||||
| PTA | Mild vs Moderate | 1.56 | 1.11 | 0.55 | 1.06 | 1.23 | 2.41 a |
| Mild vs Severe | 1.58 | 1.11 | 0.81 | 0.94 | 0.94 | 1.21 | |
| Moderate vs Severe | 1.01 | 1.00 | 1.46 | 0.89 | 0.76 | 0.50 a | |
|
| |||||||
| Drugs at FU | Yes vs No | 2.19 b | 2.30 b | 2.00 b | 1.56 b | 2.11 b | 1.72 b |
|
| |||||||
| Drinking Category at FU | Light vs Abstaining | 0.91 | 0.82 | 0.86 | 1.28 | 1.04 | 1.12 |
| Moderate vs Abstaining | 0.74 | 0.89 | 0.47 | 1.26 | 1.10 | 0.95 | |
| Heavy vs Abstaining | 2.11 a | 1.65 | 2.40 | 1.31 | 2.25 b | 1.40 | |
| Moderate vs Light | 0.81 | 1.09 | 0.54 | 0.98 | 1.06 | 0.84 | |
| Heavy vs Light | 2.31 a | 2.01 | 2.78 a | 1.02 | 2.16 b | 1.25 | |
| Heavy vs Moderate | 2.87 b | 1.85 | 5.14 b | 1.04 | 2.04 a | 1.48 | |
|
| |||||||
| GOSE at FU | V/SD vs GR | 2.31 b | 2.83 b | 1.55 | 2.23 a | 1.53 | 2.53 b |
| MD vs GR | 1.88 b | 3.05 b | 1.62 | 2.68 a | 1.68 b | 1.91 b | |
| V/SD vs MD | 1.23 | 0.93 | 0.96 | 0.83 | 0.91 | 1.33 | |
Abbreviations: SA, suicide attempt; SI, suicidal ideation; NIDILRR, National Institute on Disability, Independent Living, and Rehabilitation research; Non-V, Non-Veteran; V/SM, Veteran / Service Member; PRC, Polytrauma Rehabilitation Center; FU, follow-up; HS, high school; PTA, post-traumatic amnesia; FIM, Functional Independence Measure; GOSE, Glasgow outcome scale – extended
indicates statistical significance at α = 0.05
indicates statistically significant at α = 0.05 and after Bonferroni adjustment (α = 0.0167 for Group, PTA, and GOSE and α = 0.0083 for Drinking Category).
Table 4:
Unadjusted and Adjusted Odds Ratios for Cumulative Suicide Attempts and Suicidal Ideation
| Suicide Attempts | Suicidal Ideation | ||||
|---|---|---|---|---|---|
|
| |||||
| Year 2 | Year 5 | Year 2 | Year 5 | ||
|
| |||||
| Unadjusted Effect | Comparison | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) |
|
| |||||
| Group | NIDILRR V/SM vs NIDILRR Non-V | 0.77 (0.47, 1.26) |
0.89 (0.53, 1.51) |
0.96 (0.70, 1.32) |
1.15 (0.82, 1.63) |
| PRC V/SM vs NIDILRR Non-V | 1.23 (0.81, 1.86) |
1.56 a (1.05, 2.33) |
1.40 b (1.12, 1.75) |
1.46 b (1.14, 1.87) |
|
| PRC V/SM vs NIDILRR V/SM | 1.60 (0.87, 2.92) |
1.75 (0.94, 3.24) |
1.46 b (1.02, 2.10) |
1.27 (0.86, 1.87) |
|
|
| |||||
| Adjusted Effect | Comparison | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) |
|
| |||||
| Group | NIDILRR V/SM vs NIDILRR Non-V | 1.53 (0.87, 2.69) |
1.88 a (1.01, 3.50) |
1.15 (0.78, 1.69) |
1.44 (0.93, 2.23) |
| PRC V/SM vs NIDILRR Non-V | 0.97 (0.53, 1.77) |
1.08 (0.57, 2.02) |
1.19 (0.87, 1.63) |
0.79 (0.52, 1.20) |
|
| PRC V/SM vs NIDILRR V/SM | 0.63 (0.29, 1.36) |
0.57 (0.25, 1.30) |
1.03 (0.65, 1.63) |
0.55 a (0.32, 0.96) |
|
|
| |||||
| Covariate | OR | OR | OR | OR | |
|
| |||||
| Sex | Female vs Male | 1.25 | 1.35 | 0.98 | 0.87 |
|
| |||||
| Minority | Minority vs Non-Minority | 1.47 b | 1.26 | 1.17 | 1.08 |
|
| |||||
| Education at FU | ≤ HS vs > HS | 1.24 | 1.12 | 1.01 | 1.05 |
|
| |||||
| Employment at FU | Not Employed vs Employed | 1.19 | 1.05 | 0.98 | 1.25 |
|
| |||||
| Marital at FU | Not Married vs Married | 1.44 | 0.98 | 1.13 | 1.32 |
|
| |||||
| Cause | Not Violent vs Violent | 1.69 | 1.10 | 1.23 | 1.22 |
|
| |||||
| Pre-Injury Mental Health | Yes vs No | 1.62 b | 1.68 b | 1.59 b | 1.02 |
|
| |||||
| Pre-Injury Suicide Attempt | Yes vs No | 2.94 b | 2.76 b | 2.73 b | 2.90 b |
|
| |||||
| Age at Injury | 10 Year Increase | 0.78 b | 0.67 b | 0.92 b | 0.90 b |
|
| |||||
| FIM Motor at FU | 5 Unit Increase | 1.09 b | 1.10 b | 0.93 b | 1.01 |
|
| |||||
| FIM Cognitive at FU | 5 Unit Increase | 0.87 | 0.62 b | 0.60 b | 0.67 b |
|
| |||||
| PTA | Mild vs Moderate | 1.32 | 1.39 | 1.27 | 1.48 |
| Mild vs Severe | 1.27 | 1.37 | 1.01 | 0.92 | |
| Moderate vs Severe | 0.96 | 0.98 | 0.79 | 0.62 a | |
|
| |||||
| Drugs at FU | Yes vs No | 1.68 b | 1.83 b | 1.83 b | 1.12 |
|
| |||||
| Drinking Category at FU | Light vs Abstaining | 1.00 | 0.92 | 0.99 | 1.02 |
| Moderate vs Abstaining | 0.86 | 0.73 | 1.36 a | 1.00 | |
| Heavy vs Abstaining | 1.59 | 1.30 | 2.19 b | 1.57 | |
| Moderate vs Light | 0.86 | 0.79 | 1.38 a | 0.98 | |
| Heavy vs Light | 1.60 | 1.41 | 2.22 b | 1.55 | |
| Heavy vs Moderate | 1.85 | 1.79 | 1.61 a | 1.57 | |
|
| |||||
| GOSE at FU | V/SD vs GR | 1.90 | 1.00 | 1.42 | 3.52 b |
| MD vs GR | 2.19 b | 1.84 b | 1.56 b | 2.52 b | |
| V/SD vs MD | 0.87 | 0.54 a | 0.91 | 1.40 | |
Abbreviations: SA, suicide attempt; SI, suicidal ideation; NIDILRR, National Institute on Disability, Independent Living, and Rehabilitation research; Non-V, Non-Veteran; V/SM, Veteran / Service Member; PRC, Polytrauma Rehabilitation Center; FU, follow-up; HS, high school; PTA, post-traumatic amnesia; FIM, Functional Independence Measure; GOSE, Glasgow outcome scale – extended
indicates statistical significance at α = 0.05
indicates statistically significant at α = 0.05 and after Bonferroni adjustment (α = 0.0167 for Group, PTA, and GOSE and α = 0.0083 for Drinking Category).
The PRC V/SMs had significantly greater odds of SI compared to NIDILRR non-Vs at Years 1, 2, and 5 (ORs=1.54–1.97) and compared to NIDILRR V/SM at Years 1 and 2 (ORs=1.74–1.77). The cumulative odds of SI were greater for PRC V/SMs compared to NIDILRR non-Vs (OR= .40) and NIDILRR V/SMs (OR=1.46) by Year 2 and for PRC V/SMs compared to NIDILRR non-Vs (OR=1.46) by Year 5. No other differences in SA or SI outcomes were identified among groups.
Risk factors associated with SA/SI outcomes
We examined sociodemographic, injury/recovery-related, rehabilitation and behavioral health variables that might explain differences in SA/SI outcomes across groups. The bivariate relationships between each predictor variable and SA/SI outcomes are presented in Table 2, along with the rate of SA/SI endorsement observed within each categorical characteristic across cohorts. Among the continuous variables, younger age at injury and lower (worse) FIM Cognitive score were associated with increased odds of both SA and SI, while lower (worse) FIM Motor score was associated with increased odds of SI only. For the categorical variables, the observed SA and SI rates in each grouping of the predictor are compared to the overall SA and SI rates (representing the null rate) using chi-square tests. We found that minority status, lower levels of education, not being employed, not being married, violent etiologies, pre-injury mental health history, pre-injury SA history, drug use, moderate to heavy drinking, and moderate-severe disability on GOSE were associated with increased rates in SA or SI 1-, 2- or 5-years post-injury. Pre-injury mental health history, pre-injury SA history, drug use, and GOSE were consistent predictors of SA and SI outcomes across all years. Sex was not associated with SA or SI rates.
Adjusted relationships between group SA/SI outcomes
To test whether the observed difference in SA/SI rates across groups persisted after controlling for key SA/SI predictor variables, all multivariable logistic regressions were adjusted for the following: sex, minority status, education, employment, marital status, cause of injury, pre-injury mental health history, pre-injury SA history, age at injury, FIM Motor and Cognitive scores, PTA classification, drug use, drinking category, and GOSE score. The adjusted odds ratios comparing SA and SI among groups after controlling for this set of participant characteristics are summarized in Tables 3 and 4 (cumulative). After controlling for participant characteristics, there were no significant differences among NIDILRR and PRC groups in SA or SI outcomes at 1, 2, or 5 years. The cumulative odds of SA by Year 5 were higher for NIDILRR V/SMs compared to NIDILRR non-Vs (OR=1.88), and the cumulative odds of SI by Year 5 was lower for PRC V/SMs compared to NIDILRR V/SMs (OR=0.55); however, these differences were not significant after controlling for multiple comparisons.
The adjusted relationships between the covariates and outcomes are also summarized in Tables 3 and 4 (cumulative). In the multivariate model, pre-injury mental health history, pre-injury SA history, younger injury age, lower FIM Cognitive scores, drug use, and moderate-severe disability on GOSE tended to predict greater odds of SA and SI across one or more follow-up points. Higher FIM Motor scores tended to be positively associated with greater SA odds and negatively associated with greater SI odds. Minority status, lower levels of education, not being married, and heavy drinking were associated with increased SA or SI odds at some years. Sex, employment status, violent etiology, and PTA were not associated with increased SA or SI odds at any of the years.
Discussion
The current study compared rates and predictors of SA and SI during the first 5 years post-TBI among non-Vs and V/SMs based on whether they had received rehabilitation services in either a VA PRC or NIDILRR health care system. V/SMs with a history of TBI in the VA PRC system endorsed higher rates of SI across all follow-up years and higher rates of SA at Y5, which is generally consistent with previous literature indicating higher rates of SA and SI among Veterans than the general population29–32 and higher risk for SI among V/SMs with TBI.33,34 There were significant differences among groups based on demographic, injury-related, mental/behavioral health, and functional outcome variables. After controlling for these variables, there were no appreciable differences in risk for SA and SI among groups across time. The characteristics that consistently predicted SA and SI among people with TBI suggest a risk pattern related to mental health history, drug and alcohol use, younger age, lower functional independence with cognitive tasks, and greater levels of disability. The V/SMs served in the VA PRC system of care more commonly fit this pattern, suggesting these individuals with TBI have experiences associated with SA/SI that are fundamentally different from those receiving care in the predominately non-V inpatient rehabilitation facilities comprising the NIDILRR Model Systems.
Among our sample of people with a history of TBI—irrespective of military history or rehabilitation setting—rates of SA in the past year ranged from 1.46% to 1.87% across follow-up years, and rates of SI around the time of follow-up ranged from 8.74% to 10.05%. These rates are higher than those typically observed among the general population, as indicated by a recent surveillance report of suicidal thoughts and behaviors among adults in the United States estimating that 0.6% attempted suicide in the last year and 4.3% experienced SI.35 Overall, our findings are consistent with prior research indicating that SI and suicidal behaviors among people with TBI can be linked to demographic, medical, and psychosocial factors including premorbid and concurrent psychiatric illness, age at injury, and poor post-injury psychosocial functioning.3,36,37 There have been mixed findings in previous research regarding the association between suicide-related outcomes following TBI and measures of either functional performance or the extent of extracranial injury.8,38–40 Although lower FIM Motor and Cognitive scores were associated with SI across all groups at each time point, the clinical differences in median scores between groups were difficult to discern, suggesting that concurrent measures of overall disability might be more useful to consider in clinical practice and future research regarding suicide-related outcomes and TBI.
The observed increased risk of SI among V/SMs, particularly those served at VA PRCs, may be linked to psychiatric comorbidities (e.g., PTSD and substance use) and the complex needs of those with active duty, combat experiences, and injuries related to violence.32,41–43 When comorbid with TBI, PTSD can perpetuate other risk factors for SA and SI, including drug abuse44 and poor functional recovery.21 Unfortunately, information regarding history of PTSD was not available for individuals with TBI served in the NIDILRR health care systems. Given the known mediating effect of PTSD on the association between TBI and suicide45 and the percentage of V/SMs in the PRC cohort who had been deployed to a combat zone, PTSD may be a unique and particularly important risk factor among V/SMs receiving care in the VA PRC system.
Limitations
There are limitations related to study procedures. First, participants were recruited only among individuals with access to inpatient rehabilitation programs. Although mild TBI was experienced by some participants, these cases were a small proportion of the overall sample, are over-represented in the VA system, and may not be representative of patients with mild TBI who do not require inpatient services. Second, the NIDILRR TBIMS question used to identify V/SM status asked if a participant has a history of military service, which could include service without a history of active-duty orders (e.g., National Guard). Differences in military service and experiences between the two V/SM groups could exist and may account for some of the observed differences; unfortunately, we are limited in making these comparisons, as there are several aspects of military service history captured among V/SMs in the PRC TBIMS national database that are not collected among V/SMs in the NIDILRR database. Third, the grouping of V/SMs was based on TBI Model Systems enrollment and is not an optimal proxy for healthcare system utilization. Lastly, the methods of SA/SI data collection (i.e., SA could be proxy reported, and SI could only be self-reported) have advantages and disadvantages. Individuals may have attempted suicide without telling others, limiting the value of proxy report; however, the use of self-report for ideation restricted those analyses to individuals who were cognitively capable of responding.
There are also limitations regarding analytic approaches used with these data. First, despite a large sample size, our data included low prevalence outcomes (especially SA) and—in some instances—low prevalence covariates, which limited our options to explore how relationships between predictors and outcomes vary by study group. Second, in our effort to understand the unique variation in SA/SI risk across groups, we selected potential covariates that have historically been associated with suicide-related outcomes in previous literature; then, to compare models across both outcomes and all follow-up years, we chose to consistently adjust for covariates that either differed between the groups or had significant bivariate association with SA/SI outcomes. Resultingly, we adjusted for a combination of baseline and concurrent variables. Although many of the concurrent variables represent potentially modifiable treatment targets (e.g., substance use) or otherwise meaningful aspects of the lived experience of a person with TBI, they might also mediate the relationship between groups and SA/SI and warrant caution in interpreting our results.
Conclusion
These findings underscore the critical importance of assessing mental/behavioral health history and screening for suicidal thoughts and behaviors among all people with a history of TBI throughout the continuum of post-acute care rehabilitation services. Across cohorts in this study, greater levels of disability following TBI tended to confer higher risk of SA and SI than did indicators of initial injury severity. Clinicians are encouraged to focus on the overall functional outcomes at the point of service rather than the injury history itself when assessing potential for suicide risk among people with TBI. For V/SMs receiving mental or other health care services in the VA system, consistent screening for a history of TBI would help to identify individuals whose risk for SA/SI could be exacerbated by their injury. Health care providers serving V/SMs with a history of TBI outside the VA system could also benefit from familiarity with services dedicated to V/SMs, including crisis intervention, mental health services, and referral pathways into the VA polytrauma care system. To better understand the associations between TBI and suicide risk, longitudinal studies such as the TBIMS must expand their focus to a broader range of biopsychosocial factors that could influence suicide risk and develop the infrastructure to assess such factors with higher resolution at more frequent intervals.
Supplementary Material
Disclosure of Funding:
The views expressed in this manuscript are those of the authors and do not necessarily represent the official policy or position of the Defense Health Agency, Department of Defense, or any other U.S. government agency. This work was prepared under Contract HT0014-22-C-0016 with DHA Contracting Office (CO-NCR) HT0014 and, therefore, is defined as U.S. Government work under Title 17 U.S.C.§101. Per Title 17 U.S.C.§105, copyright protection is not available for any work of the U.S. Government. For more information, please contact dha.TBICOEinfo@mail.mil.
UNCLASSIFIED.
The contents of this publication were developed under grants from the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR): Virginia Commonwealth University TBI Model System (#90DPTB0005); TBI Model Systems National Data and Statistical Center (#90DPTB0018); Texas TBI Model System of TIRR (#90DPTB0016); Spaulding/Harvard TBI Model System (#90DPTB0011); a Center within the Administration for Community Living (ACL), Department of Health and Human Services (HHS). The contents of this publication do not necessarily represent the policy of NIDILRR, ACL, HHS, and you should not assume endorsement by the Federal Government.
Footnotes
Conflict of Interest:
None
References
- 1.Simpson G, Tate R. Suicidality after traumatic brain injury: demographic, injury and clinical correlates. Psychol Med. 2002;32(4):687–697. doi: 10.1017/s0033291702005561 [DOI] [PubMed] [Google Scholar]
- 2.Wasserman L, Shaw T, Vu M, Ko C, Bollegala D, Bhalerao S. An overview of traumatic brain injury and suicide. Brain Inj. 2008;22(11):811–819. doi: 10.1080/02699050802372166 [DOI] [PubMed] [Google Scholar]
- 3.Wadhawan A, Stiller JW, Potocki E, et al. Traumatic Brain Injury and Suicidal Behavior: A Review. Journal of Alzheimer’s Disease. 2019;68(4):1339–1370. doi: 10.3233/JAD-181055 [DOI] [PubMed] [Google Scholar]
- 4.Fazel S, Wolf A, Pillas D, Lichtenstein P, Långström N. Suicide, fatal injuries, and other causes of premature mortality in patients with traumatic brain injury: a 41-year Swedish population study. JAMA Psychiatry. 2014;71(3):326–333. doi: 10.1001/jamapsychiatry.2013.3935 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Harrison-Felix CL, Whiteneck GG, Jha A, DeVivo MJ, Hammond FM, Hart DM. Mortality over four decades after traumatic brain injury rehabilitation: a retrospective cohort study. Arch Phys Med Rehabil. 2009;90(9):1506–1513. doi: 10.1016/j.apmr.2009.03.015 [DOI] [PubMed] [Google Scholar]
- 6.Madsen T, Erlangsen A, Orlovska S, Mofaddy R, Nordentoft M, Benros ME. Association Between Traumatic Brain Injury and Risk of Suicide. JAMA. 2018;320(6):580–588. doi: 10.1001/jama.2018.10211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Silver JM, Kramer R, Greenwald S, Weissman M. The association between head injuries and psychiatric disorders: findings from the New Haven NIMH Epidemiologic Catchment Area Study. Brain Inj. 2001;15(11):935–945. doi: 10.1080/02699050110065295 [DOI] [PubMed] [Google Scholar]
- 8.Klyce DW, Perrin PB, Fisher LB, et al. Identifying group-based patterns of suicidal ideation over the first 10 years after moderate-to-severe TBI. J Clin Psychol. 2022;78(5):877–891. doi: 10.1002/jclp.23282 [DOI] [PubMed] [Google Scholar]
- 9.Mackelprang JL, Bombardier CH, Fann JR, Temkin NR, Barber JK, Dikmen SS. Rates and Predictors of Suicidal Ideation During the First Year After Traumatic Brain Injury. Am J Public Health. 2014;104(7):e100–e107. doi: 10.2105/AJPH.2013.301794 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fisher LB, Pedrelli P, Iverson GL, et al. Prevalence of suicidal behaviour following traumatic brain injury: Longitudinal follow-up data from the NIDRR Traumatic Brain Injury Model Systems. Brain Inj. 2016;30(11):1311–1318. doi: 10.1080/02699052.2016.1195517 [DOI] [PubMed] [Google Scholar]
- 11.Tsaousides T, Cantor JB, Gordon WA. Suicidal ideation following traumatic brain injury: prevalence rates and correlates in adults living in the community. J Head Trauma Rehabil. 2011;26(4):265–275. doi: 10.1097/HTR.0b013e3182225271 [DOI] [PubMed] [Google Scholar]
- 12.Bahraini NH, Simpson GK, Brenner LA, Hoffberg AS, Schneider AL. Suicidal Ideation and Behaviours after Traumatic Brain Injury: A Systematic Review. Brain Impairment. 2013;14(1):92–112. doi: 10.1017/BrImp.2013.11 [DOI] [Google Scholar]
- 13.McIntire KL, Crawford KM, Perrin PB, et al. Factors Increasing Risk of Suicide after Traumatic Brain Injury: A State-of-the-Science Review of Military and Civilian Studies. Brain Inj. 2021;35(2):151–163. doi: 10.1080/02699052.2020.1861656 [DOI] [PubMed] [Google Scholar]
- 14.Dennis JP, Ghahramanlou-Holloway M, Cox DW, Brown GK. A guide for the assessment and treatment of suicidal patients with traumatic brain injuries. J Head Trauma Rehabil. 2011;26(4):244–256. doi: 10.1097/HTR.0b013e3182225528 [DOI] [PubMed] [Google Scholar]
- 15.Awan N, DiSanto D, Juengst SB, et al. Evaluating the Cross-Sectional and Longitudinal Relationships Predicting Suicidal Ideation Following Traumatic Brain Injury. J Head Trauma Rehabil. Published online July 22, 2020. doi: 10.1097/HTR.0000000000000588 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fisher LB, Curtiss JE, Klyce DW, et al. Using Machine Learning to Examine Suicidal Ideation After TBI: A TBI Model Systems National Database Study. Am J Phys Med Rehabil. Published online June 8, 2022. doi: 10.1097/PHM.0000000000002054 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Brenner LA, Ignacio RV, Blow FC. Suicide and traumatic brain injury among individuals seeking Veterans Health Administration services. J Head Trauma Rehabil. 2011;26(4):257–264. doi: 10.1097/HTR.0b013e31821fdb6e [DOI] [PubMed] [Google Scholar]
- 18.Hostetter TA, Hoffmire CA, Forster JE, Adams RS, Stearns-Yoder KA, Brenner LA. Suicide and Traumatic Brain Injury Among Individuals Seeking Veterans Health Administration Services Between Fiscal Years 2006 and 2015. J Head Trauma Rehabil. 2019;34(5):E1–E9. doi: 10.1097/HTR.0000000000000489 [DOI] [PubMed] [Google Scholar]
- 19.Howard JT, Stewart IJ, Amuan M, Janak JC, Pugh MJ. Association of Traumatic Brain Injury With Mortality Among Military Veterans Serving After September 11, 2001. JAMA Netw Open. 2022;5(2):e2148150. doi: 10.1001/jamanetworkopen.2021.48150 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wilks CR, Morland LA, Dillon KH, et al. Anger, social support, and suicide risk in U.S. military veterans. Journal of Psychiatric Research. 2019;109:139–144. doi: 10.1016/j.jpsychires.2018.11.026 [DOI] [PubMed] [Google Scholar]
- 21.Fonda JR, Fredman L, Brogly SB, McGlinchey RE, Milberg WP, Gradus JL. Traumatic Brain Injury and Attempted Suicide Among Veterans of the Wars in Iraq and Afghanistan. Am J Epidemiol. 2017;186(2):220–226. doi: 10.1093/aje/kwx044 [DOI] [PubMed] [Google Scholar]
- 22.Olenick M, Flowers M, Diaz VJ. US veterans and their unique issues: enhancing health care professional awareness. Adv Med Educ Pract. 2015;6:635–639. doi: 10.2147/AMEP.S89479 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Department of Veterans Affairs Office of Mental Health and Suicide Prevention CM. National Strategy for Preventing Veteran Suicide. Accessed August 15, 2022. https://www.mentalhealth.va.gov/suicide_prevention/docs/Office-of-Mental-Health-and-Suicide-Prevention-National-Strategy-for-Preventing-Veterans-Suicide.pdf
- 24.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–613. doi: 10.1046/j.1525-1497.2001.016009606.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Linacre JM, Heinemann AW, Wright BD, Granger CV, Hamilton BB. The structure and stability of the Functional Independence Measure. Arch Phys Med Rehabil. 1994;75(2):127–132. [PubMed] [Google Scholar]
- 26.The Management of Concussion-mild Traumatic Brain Injury Working Group. VA/DoD Clinical Practice Guidlines for the Management of Concussion-Mild Traumatic Brain Injury (MTBI). Department of Veterans Affairs and Department of Defense; 2016. Accessed August 26, 2019. https://www.healthquality.va.gov/guidelines/Rehab/mtbi/mTBICPGFullCPG50821816.pdf
- 27.Centers for Disease Control. CDC - BRFSS. Behavioral Risk Factor Surveillance System. Published 2003. Accessed August 29, 2022. https://www.cdc.gov/brfss/index.html
- 28.Wilson JT, Pettigrew LE, Teasdale GM. Structured interviews for the Glasgow Outcome Scale and the extended Glasgow Outcome Scale: guidelines for their use. J Neurotrauma. 1998;15(8):573–585. doi: 10.1089/neu.1998.15.573 [DOI] [PubMed] [Google Scholar]
- 29.Kessler RC, Berglund P, Borges G, Nock M, Wang PS. Trends in Suicide Ideation, Plans, Gestures, and Attempts in the United States, 1990–1992 to 2001–2003. JAMA. 2005;293(20):2487–2495. doi: 10.1001/jama.293.20.2487 [DOI] [PubMed] [Google Scholar]
- 30.Nock MK, Kessler RC. Prevalence of and risk factors for suicide attempts versus suicide gestures: Analysis of the National Comorbidity Survey. Journal of Abnormal Psychology. 2006;115(3):616–623. doi: 10.1037/0021-843X.115.3.616 [DOI] [PubMed] [Google Scholar]
- 31.SIMPSON G TE R. Suicidality after traumatic brain injury: demographic, injury and clinical correlates. Psychological Medicine. 2002;32(4):687–697. doi: 10.1017/S0033291702005561 [DOI] [PubMed] [Google Scholar]
- 32.Fonda JR, Fredman L, Brogly SB, McGlinchey RE, Milberg WP, Gradus JL. Traumatic Brain Injury and Attempted Suicide Among Veterans of the Wars in Iraq and Afghanistan. American Journal of Epidemiology. 2017;186(2):220–226. doi: 10.1093/aje/kwx044 [DOI] [PubMed] [Google Scholar]
- 33.Greer N, Sayer NA, Spoont M, et al. Prevalence and Severity of Psychiatric Disorders and Suicidal Behavior in Service Members and Veterans With and Without Traumatic Brain Injury: Systematic Review. J Head Trauma Rehabil. 2020;35(1):1–13. doi: 10.1097/HTR.0000000000000478 [DOI] [PubMed] [Google Scholar]
- 34.Ashrafioun L, Pigeon WR, Conner KR, Leong SH, Oslin DW. Prevalence and correlates of suicidal ideation and suicide attempts among veterans in primary care referred for a mental health evaluation. J Affect Disord. 2016;189:344–350. doi: 10.1016/j.jad.2015.09.014 [DOI] [PubMed] [Google Scholar]
- 35.Ivey-Stephenson AZ. Suicidal Thoughts and Behaviors Among Adults Aged ≥18 Years — United States, 2015–2019. MMWR Surveill Summ. 2022;71. doi: 10.15585/mmwr.ss7101a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.McIntire KL, Crawford KM, Perrin PB, et al. Factors Increasing Risk of Suicide after Traumatic Brain Injury: A State-of-the-Science Review of Military and Civilian Studies. null. 2021;35(2):151–163. doi: 10.1080/02699052.2020.1861656 [DOI] [PubMed] [Google Scholar]
- 37.Simpson G, Tate R. Suicidality in people surviving a traumatic brain injury: Prevalence, risk factors and implications for clinical management. null. 2007;21(13–14):1335–1351. doi: 10.1080/02699050701785542 [DOI] [PubMed] [Google Scholar]
- 38.Dreer LE, Tang X, Nakase-Richardson R, et al. Suicide and traumatic brain injury: a review by clinical researchers from the National Institute for Disability and Independent Living Rehabilitation Research (NIDILRR) and Veterans Health Administration Traumatic Brain Injury Model Systems. Curr Opin Psychol. 2018;22:73–78. doi: 10.1016/j.copsyc.2017.08.030 [DOI] [PubMed] [Google Scholar]
- 39.Perrin PB, Klyce DW, Fisher LB, et al. Relations Among Suicidal Ideation, Depressive Symptoms, and Functional Independence During the 10 Years After Traumatic Brain Injury: A Model Systems Study. Arch Phys Med Rehabil. 2022;103(1):69–74. doi: 10.1016/j.apmr.2021.07.790 [DOI] [PubMed] [Google Scholar]
- 40.Kesinger MR, Juengst SB, Bertisch H, et al. Acute Trauma Factor Associations With Suicidality Across the First 5 Years After Traumatic Brain Injury. Archives of physical medicine and rehabilitation. 2016;97(8):1301–1308. doi: 10.1016/j.apmr.2016.02.017 [DOI] [PubMed] [Google Scholar]
- 41.Ropacki S, Nakase-Richardson R, Farrell-Carnahan L, Lamberty GJ, Tang X. Descriptive Findings of the VA Polytrauma Rehabilitation Centers TBI Model Systems National Database. Archives of Physical Medicine and Rehabilitation. 2018;99(5):952–959. doi: 10.1016/j.apmr.2017.12.035 [DOI] [PubMed] [Google Scholar]
- 42.McGarity S, Nakase-Richardson R, Lamberty GJ, et al. Poster 100 The VA PRC Traumatic Brain Injury Model System Program. Archives of Physical Medicine and Rehabilitation. 2013;94(10):e45. doi: 10.1016/j.apmr.2013.08.223 [DOI] [PubMed] [Google Scholar]
- 43.Sayer NA, Cifu DX, McNamee S, et al. Rehabilitation Needs of Combat-Injured Service Members Admitted to the VA Polytrauma Rehabilitation Centers: The Role of PM&R in the Care of Wounded Warriors. PM&R 2009;1(1):23–28. doi: 10.1016/j.pmrj.2008.10.003 [DOI] [PubMed] [Google Scholar]
- 44.Alway YJ, Gould K, McKay A, Johnston L, Ponsford J. The Evolution of Posttraumatic Stress Disorder following Moderate to Severe Traumatic Brain Injury. J Neurotrauma. Published online July 15, 2015. doi: 10.1089/neu.2015.3992 [DOI] [PubMed] [Google Scholar]
- 45.Bahraini NH, Breshears RE, Hernández TD, Schneider AL, Forster JE, Brenner LA. Traumatic brain injury and posttraumatic stress disorder. Psychiatr Clin North Am. 2014;37(1):55–75. doi: 10.1016/j.psc.2013.11.002 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.