Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: Suicide Life Threat Behav. 2021 Apr;51(2):237–246. doi: 10.1111/sltb.12699

Substance use, current criminal justice involvement, and lifetime suicidal thoughts and behaviors history: The moderating role of thwarted belonging

Sean M Mitchell 1,2, Sarah L Brown 3, Tuğba Görgülü 2,4, Kenneth R Conner 2,5, Marc T Swogger 2
PMCID: PMC8128077  NIHMSID: NIHMS1697135  PMID: 33876498

Abstract

Objective:

Criminal justice involvement, substance use, and suicide pose significant public health concerns; however, the unique and synergistic effects of these experiences among high-risk individuals remain understudied. We hypothesized positive main effects for alcohol-related severity, drug-related severity, current criminal justice involvement, and thwarted belonging (TB) on suicide ideation history (SIH) and suicide attempt history (SAH) and that TB would moderate these associations.

Method:

We report on cross-sectional analyses of self-report assessments completed by 824 adult residential substance use patients.

Results:

Multinomial logistic regression analyses indicated that as alcohol- and drug-related severity increases, the probability of SIH and SAH increases; however, TB was only associated with a higher SIH probability. Significant two-way interactions (current criminal justice involvement*TB; alcohol-related severity*TB) indicated that (1) those high in TB with current criminal justice involvement were more likely to report a SIH than those without current criminal justice involvement; and (2) those low in TB and alcohol-related severity had the lowest SAH probability, whereas those low in TB and high in alcohol-related severity had the highest SAH probability.

Conclusion:

The unique and combined effects of interpersonal and contextual risk factors may improve suicide risk conceptualization and assessment, and allow for tailored treatments for this high-risk population.

Keywords: substance use, thwarted belonging, suicide ideation, suicide attempt, criminal justice system

1 |. Introduction

The suicide rate in the United States has increased by 33% over the past two decades (Hedegaard et al., 2018). Recent data indicate that the suicide rate increased 1.4% from 2017 to 2018, and suicide remains the 10th leading cause of death in the United States (Xu et al., 2020). This is a serious public health problem that requires a focus on high-risk groups who disproportionately contribute to the increase and a multipronged prevention approach that involves multiple systems and service providers. Individuals who engage in substance use are a particularly high-risk population (Conner et al., 2019; Darvishi et al., 2015). The interplay between interpersonal and contextual factors may be especially salient to individuals who use substances; however, this has remained largely uninvestigated. Thus, our study tested the associations between feelings of belonging, substance use (including drug- and alcohol-related severity), current criminal justice (CJ) involvement, and lifetime suicidal thoughts and behaviors history among residential substance use patients. Such work is necessary to identify key synergistic effects of risk factors for suicidal thoughts and behaviors for individuals who cross multiple systems and agencies (i.e., substance use treatment and the CJ system) to best coordinate suicide risk assessment and management among this high-risk group.

CJ involvement and substance use often co-occur and contribute to a significant public health burden, maintaining and exacerbating the interrelated crises of poverty, trauma, and disease (Csete et al., 2016). Meta-analytic findings indicated that the odds of criminal offending are approximately three to four times greater among substance users than non-substance users (Bennett et al., 2008). In addition, individuals who meet the criteria for substance dependence or abuse are overrepresented in correctional facilities. For example, approximately 58% of state prisoners and 63% of jail inmates meet the criteria for substance use disorders prior to incarceration compared to the 5% of individuals in the general population (Bronson et al., 2017). This problem is not unique to correctional facilities as substance use is approximately two to three times more prevalent among those involved in community correction (i.e., probation and parole) than those who are not (Feucht & Gfroerer, 2011). Taken together, CJ involvement is likely a prevalent concern among residential substance use patients, which creates an opportunity for coordinated care that would reduce public health burden.

Residential substance use patients may be at particularly elevated risk for suicide, given these multiple contextual factors (i.e., CJ involvement and substance use severity) that have been linked to suicidal thoughts and behaviors. Both CJ involvement (Cook, 2013; Cook & Davis, 2012) and substance severity (Conner et al., 2019; Darvishi et al., 2015) are robust risk factors for suicidal thoughts and behaviors. Psychological autopsies of 76 suicide decedents in New York correctional facilities revealed that 95% of the prisoners who died by suicide had a substance use history (Kovasznay et al., 2004). Moreover, prisoners and probationers who attempted suicide also had significant histories of substance use and used a greater number of different substances than individuals without a suicide attempt history (Hakansson et al., 2010). Therefore, CJ involvement and substance use severity are noteworthy correlates of suicidal behaviors and suicide deaths. However, research is needed to elucidate how substance use severity and CJ involvement are uniquely linked to suicidal thoughts and behaviors among residential substance use patients, which could inform the public health intervention need for these high-risk individuals.

In addition to substance use and CJ involvement, interpersonal factors are likely relevant to suicide risk among residential substance use patients. Thwarted belonging (TB; i.e., a lack of reciprocal caring relationships and feelings of loneliness) and social connectedness have been linked to suicidal thoughts and behaviors in several contemporary theories of suicide (i.e., the interpersonal theory of suicide [Joiner, 2005; Van Orden et al., 2010], the three-step theory [Klonsky & May, 2015], and the integrated motivational–volitional model of suicidal behavior [O’Connor, 2011]). TB is likely particularly relevant for the conceptualization of suicide risk among individuals with substance use problems and those with CJ involvement. For example, conflictual or terminated relationships predict suicide death among individuals with substance dependence (Yuodelis & Ries, 2015). Likewise, individuals involved in the CJ system with a suicide attempt history are more likely to have relationship problems and associates who participate in substance use (Hakansson et al., 2010). Similarly, psychiatric inpatients who reported elevated TB and more time spent with criminal associates most likely reported “extreme” suicide ideation distress (Mitchell et al., 2019). Thus, TB may interact with substance use severity and current CJ involvement, and produce synergistic risk for suicidal thoughts and behaviors among residential substance use patients.

In sum, existing literature indicates that substance use severity and CJ involvement are commonly experienced in combination (e.g., Bennett et al., 2008) and therefore are relevant risk factors for suicidal thoughts and behaviors among residential substance use patients (e.g., Hakansson et al., 2010; Yuodelis & Ries, 2015). TB may be particularly relevant for residential substance use patients with high substance use severity and who have CJ involvement (e.g., Mitchell et al., 2019; Yuodelis & Ries, 2015). However, research has not examined the unique and synergistic effects of these risk factors in relation to suicidal thoughts and behaviors. This is especially concerning among residential substance use patients for whom public health crises intersect—substance use, physical and mental health condition comorbidity, the influx of physical and mental health conditions in the CJ system, and suicide (Al-Rousan et al., 2017; Csete et al., 2016). This work may inform assessment, intervention, and prevention at multisystemic levels.

We aimed to test these relations among individuals receiving residential substance use treatment to inform assessment and intervention for this high-risk group. We hypothesized positive main effects for alcohol- and drug-related severity, current CJ involvement, and TB on lifetime suicide ideation and attempts history (i.e., lifetime suicide-related history). We also hypothesized that TB would moderate alcohol- and drug-related severity, and current CJ involvement in relation to lifetime suicide-related history. Specifically, we hypothesized that these relations would be strengthened at higher levels of TB.

2 |. Methods

2.1 |. Participants

Out of the 827 participants recruited, the final sample in this study was 824 residential substance use patients (591 men and 233 women; ages 18 to 70 years, M = 39.06, SD = 11.28). Participants were identified as non-Hispanic White (n = 480, 58.3%), non-Hispanic Black (n = 287, 34.8%), Hispanic White (n = 26, 3.2%), Native American (n = 15, 1.8%), Hispanic Black (n = 8, 1%), and “other” race/ethnicity (n = 8, 1%). On average, participants reported 12.19 years of education (SD = 1.97). Most participants were single/never married (n = 419, 50.8%) and unemployed during the past 30 days (n = 416, 50.5%), and reported an income less than $20,000 per year (n = 514, 65.7%). Diagnostic data were not available, as these participants were part of a brief screening procedure for a larger study.

2.2 |. Measures

2.2.1 |. Interpersonal needs questionnaire (INQ)

The INQ (Van Orden et al., 2012) TB scale was used in the current study where higher scores indicate greater TB. Participants respond to 9 self-report items with responses ranging from 1 (not at all true for me) to 7 (very true for me). Therefore, scores range from 9 to 63. The INQ has demonstrated strong psychometric properties (e.g., Van Orden et al., 2012). We did not administer the INQ perceived burdensomeness items. In our study, the internal consistency for the INQ-TB scale was α = 0.80.

2.2.2 |. Patient health questionnaire-9 (PHQ-9)

The PHQ-9 (Spitzer et al., 1999) is a 9-item self-report screening assessment for depressive symptoms during the past two weeks. Participants respond to items with responses ranging from 0 (not at all) to 3 (nearly every day). Higher scores indicate more severe depressive symptoms. The PHQ-9 has demonstrated strong psychometric characteristics (e.g., Dum et al., 2008). We excluded the suicide ideation item (item 9) from scoring, given that lifetime suicide ideation and attempt history are our primary outcome. Therefore, scores ranged from 0 to 24. The internal consistency of the items used in this study was α = 0.87.

2.2.3 |. Alcohol use disorder identification test (AUDIT)

The AUDIT (Babor et al., 1992) is a 10-item self-report screening assessment of alcohol-related severity during the past 12 months. Each item is rated from 0 to 4, which yields a total score ranging from 0 to 40. Higher scores indicate greater alcohol-related severity. The AUDIT is a valid and reliable assessment tool (e.g., Donovan et al., 2006). In our study, the internal consistency for the scale was α = 0.92.

2.2.4 |. Drug abuse screening test (DAST)

The DAST (Cocco & Carey, 1998) includes 10 yes/no self-report items to screen drug-related severity (excluding alcohol and tobacco) during the past 12 months. Scores range from 0 to 10, where higher scores indicate greater drug-related severity. The DAST has demonstrated strong reliability and validity (see Yudko et al., 2007). In our study, the internal consistency for the DAST was α = 0.81.

2.2.5 |. Current criminal justice involvement

Current CJ involvement was assessed using a self-report yes/no item, asking, “Are you currently facing legal charges, or currently on probation or parole?” Most participants reported current CJ involvement (n = 474, 57.5%).

2.2.6 |. Lifetime suicide-related history

Lifetime suicide attempt history was assessed using a self-report yes/no item, asking, “Have you ever tried to kill yourself or attempt suicide?” This item has demonstrated strong test–retest reliability among substance use patients (Conner et al., 2007). Lifetime suicide ideation history was assessed using a self-report yes/no item, asking, “Have you ever seriously thought about committing suicide?” (Kessler et al., 1999). These items were used to place participants into three mutually exclusive categories of lifetime suicide-related history: no lifetime suicide ideation or attempt history (n = 454, 55.1%); lifetime suicide ideation without a suicide attempt history (n = 168, 20.4%); and lifetime suicide attempt with or without suicide ideation history (n = 202, 24.5%).

2.3 |. Procedure

The necessary institutional review boards approved all study procedures. Researchers recruited participants from four residential substance use treatment facilities in upstate New York. After a brief announcement about the study, interested patients met individually with the research study staff for a screening session. Only English-speaking adult patients were eligible to participate in the screening session. After providing informed consent, participants completed self-report assessments, and they were compensated with a $10 gift card. Some participants (n = 200) went on to complete a more in-depth assessment battery. We only utilized the screening data; therefore, we do not provide further detail about the additional assessments.

2.4 |. Data analysis plan

We conducted multinomial logistic regression analyses using SPSS version 25. The criterion variable, lifetime suicide-related history, was multicategorical (reference group was no lifetime suicide-related history). We mean-centered all continuous predictor variables before analysis. We entered drug-related severity scores, alcohol-related severity scores, TB scores, and current CJ involvement (coded 1 = current CJ involvement, 0 = no current CJ involvement) as main effect variables in relation to lifetime suicide-related history. Then, we added the hypothesized two-way interactions as predictor variables to the model (i.e., alcohol-related severity × TB, drug-related severity × TB, and current CJ involvement × TB). We conducted simple slope analyses to probe statistically significant interactions.

3 |. Results

3.1 |. Data screening and preparation

Little’s missing completely at random (MCAR) test indicated that data were not MCAR (χ2[480, N = 825] = 564, p = 0.005); however, given the small number of missing values (1.25%), expectation-maximization imputation was deemed an acceptable approach to impute continuous missing data values. Of the original 827 participants, we excluded two participants from analyses because they were missing responses on the lifetime suicide-related history variable. Additionally, we identified and winsorized 7 univariate outliers (i.e., scores greater than ±3.29 SD from the mean) for the years of education variable. We also identified one participant as a multivariate outlier using Mahalanobis distance and removed them from the analyses. Therefore, we used the final sample of 824 participants in the analyses.

We considered additional covariates based on bivariate associations with the criterion variable, lifetime suicide-related history. Depressive symptoms (OR = 1.11, p < 0.001), sex (OR = 0.52, p < 0.001), and years of education (OR = 0.82, p < 0.001) were associated with no lifetime suicide-related history vs. lifetime suicide attempt history. Depressive symptoms were also associated with no lifetime suicide-related history vs. lifetime suicide ideation history (OR = 1.07, p < 0.001). Other demographic variables (e.g., ethnicity and age) were not associated with the criterion variable (p > 0.05). Therefore, we included depressive symptoms, sex, and years of education as covariates in the analyses. See Table 1 for correlations and descriptive statistics.

TABLE 1.

Bivariate correlations and descriptive statistics

1 2 3 4 5 6 7
1. TB ----
2. Drug 0.17** ----
3. Alcohol 0.08* −0.04 ----
4. CJ† 0.07* 0.03 0.17** ----
5. Depression 0.35** 0.26** 0.22** 0.12** ----
6. Education −0.05 −0.12** −0.03 −0.02 .00 ----
7. Sex† 0.08* 0.00 −0.02 −0.14** −.06 .02 ----
M 30.24 5.95 20.65 ---- 11.35 12.19 ----
SD 10.63 2.78 11.81 ---- 6.28 1.97 ----
Observed Range 9–63 0–10 0–40 0–1 0–24 3–19 0–1
Open in a new tab

Abbreviations: Alcohol, Alcohol Use Disorder Identification Test (AUDIT) alcohol-related severity score; CJ, Current Criminal Justice Involvement (coded 0 = no involvement, 1 = involvement); Depression, Patient Health Questionnaire-9 score (with the suicide ideation item removed); Drug, Drug Abuse Screening Test (DAST) drug-related severity score; Education, years of education; Sex, self-reported sex (coded 0 = women; 1 = men); TB, Interpersonal Needs Questionnaire Thwarted Belongingness score.

†

Point-biserial correlation between dichotomous and the other continuous variables or phi coefficient between two dichotomous variables;

*

p < 0.05;

**

p < 0.01.

3.2 |. Primary analyses

First, we tested the main effects (Nagelkerke R2 = 0.20; −2LL = 1,491.30; χ2[14, N = 825] = 152.23, p < 0.001). As seen in Table 2, there were significant main effects of drug-related severity (OR = 1.07, p = 0.040), alcohol-related severity (OR = 1.03, p = 0.002), and TB (OR = 1.03 p = 0.001) in association with no lifetime suicide-related history vs. lifetime suicide ideation history, adjusting for the other covariates. There were also significant main effects of drug-related severity (OR = 1.24, p < 0.001) and alcohol-related severity (OR = 1.02, p = 0.033), but not TB, in association with no lifetime suicide-related history vs. lifetime suicide attempt history, adjusting for the other covariates. Current CJ involvement was not significantly associated with the criterion variable, adjusting for the other covariates. These results indicate that as alcohol- and drug-related severity increases, individuals are more likely to report a lifetime suicide ideation and attempt history; however, increased TB is only associated with an increased likelihood of a lifetime suicide ideation history.

TABLE 2.

Multinomial logistic regression results testing the main effects of and interactions between thwarted belongingness, drug-related severity, alcohol-related severity, and current criminal justice involvement predicting suicide-related history status after adjusting for additional covariates

OR 95% CI
Criterion Variable Predictor Variable Logit SE Wald χ2 p OR Lower Upper
Suicide Intercept −1.09 0.23 22.65 <0.001
Ideation Sex −0.07 0.22 0.10 0.750 0.93 0.61 1.43
Education −0.02 0.05 0.24 0.624 0.98 0.89 1.08
Depression 0.04 0.02 4.87 0.027 1.04 1.00 1.07
Drug 0.07 0.04 4.21 0.040 1.07 1.00 1.15
CJ 0.27 0.19 1.86 0.173 1.30 0.89 1.91
Alcohol 0.03 0.01 9.22 0.002 1.03 1.01 1.04
TB 0.03 0.01 10.15 0.001 1.03 1.01 1.05
Suicide Intercept −0.69 0.21 10.70 0.001
Attempt Sex −0.63 0.20 10.19 0.001 0.53 0.36 0.78
Education −0.17 0.05 12.29 <0.001 0.84 0.76 0.93
Depression 0.07 0.02 17.50 <0.001 1.07 1.04 1.11
Drug 0.21 0.04 29.26 <0.001 1.24 1.15 1.34
CJ 0.32 0.19 2.76 0.097 1.38 0.94 2.00
Alcohol 0.02 0.01 4.54 0.033 1.02 1.00 1.03
TB 0.02 0.01 2.68 0.102 1.02 1.00 1.03
Suicide Intercept −1.09 0.23 22.36 <0.001
Ideation Sex −0.06 0.22 0.08 0.783 0.94 0.61 1.44
Education −0.02 0.05 0.16 0.686 0.98 0.89 1.08
Depression 0.04 0.02 5.21 0.022 1.04 1.01 1.07
Drug 0.07 0.04 3.94 0.047 1.07 1.00 1.15
CJ 0.21 0.20 1.16 0.282 1.24 0.84 1.82
Alcohol 0.03 0.01 8.37 0.004 1.03 1.01 1.04
TB 0.00 0.01 0.04 0.833 1.00 0.98 1.03
Drug × TB 0.00 0.00 0.31 0.578 1.00 0.99 1.01
Alcohol × TB 0.00 0.00 0.13 0.719 1.00 1.00 1.00
CJ × TB 0.05 0.02 6.06 0.014 1.05 1.01 1.09
Suicide Intercept −0.71 0.21 11.22 0.001
Attempt Sex −0.65 0.20 10.53 0.001 0.52 0.35 0.77
Education −0.18 0.05 12.43 <0.001 0.84 0.76 0.93
Depression 0.07 0.02 16.68 <0.001 1.07 1.04 1.11
Drug 0.22 0.04 30.71 <0.001 1.25 1.16 1.35
CJ 0.33 0.19 2.90 0.088 1.39 0.95 2.03
Alcohol 0.02 0.01 5.42 0.020 1.02 1.00 1.04
TB 0.01 0.01 0.68 0.410 1.01 0.98 1.04
Drug × TB 0.004 0.00 1.139 0.286 1.004 0.997 1.011
Alcohol × TB −0.002 0.00 4.321 0.038 0.998 0.997 0.999
CJ × TB 0.004 0.02 0.05 0.824 1.004 0.969 1.040
Open in a new tab

Note: The reference group was no suicide-related history;

Abreviations: Alcohol, Alcohol Use Disorder Identification Test (AUDIT) alcohol-related severity score; CJ, Current Criminal Justice Involvement (coded 0 = no involvement, 1 = involvement); Depression, Patient Health Questionnaire-9 score (with the suicide ideation item removed); Drug, Drug Abuse Screening Test (DAST) drug-related severity score; Education, years of education; Sex, self-reported sex (coded 0 = women; 1 = men); TB, Interpersonal Needs Questionnaire Thwarted Belongingness score.

To test moderation effects, we added the two-way interactions to the model (Nagelkerke R2 = 0.21; −2LL = 1,477.79; χ2[20, N = 825] = 165.74, p < 0.001). As seen in Table 2, there was a significant current CJ involvement by TB interaction in association with no lifetime suicide-related history vs. lifetime suicide ideation history (OR = 1.05, p = 0.014). There was also a significant alcohol-related severity by TB interaction in association with no lifetime suicide-related history vs. lifetime suicide attempt history (OR = 0.99, p = 0.038). Simple slope analyses were conducted to probe these significant interactions.

We tested the association between current CJ involvement and no lifetime suicide-related history vs. lifetime suicide ideation history at different TB levels. This association was not significant at low TB (−1 SD; OR = 0.75, p = 0.333) nor at average TB (M; OR = 1.24, p = 0.282). However, this association was significant at high TB (+1 SD; OR = 2.03, p = 0.009), indicating the relation between current CJ involvement and no lifetime suicide-related history vs. lifetime suicide ideation history became stronger and positive as TB increased. As depicted in Figure 1, individuals with current CJ involvement were more likely to report a lifetime suicide ideation history than those without current CJ involvement only when they also reported higher TB.

FIGURE 1.

FIGURE 1

Open in a new tab

The simple slope analysis for the significant interaction between current criminal justice (CJ) involvement and thwarted belongingness (TB) predicting no suicide-related history vs. suicide ideation history, adjusting for additional covariates. This figure shows the estimated probability of reporting a suicide ideation history for those with and without current CJ involvement at low (−1 SD), average (M), and high (+1 SD) levels of TB. *p < 0.05

We also tested the association between alcohol-related severity and no lifetime suicide-related history vs. suicide attempt history different TB levels. This association was significant at low TB (−1 SD; OR = 1.04, p = 0.004) and average TB (M; OR = 1.02, p = 0.020) but not at high TB (+1 SD; OR = 1.002, p = 0.842). This indicates the relation between alcohol-related severity and no lifetime suicide-related history vs. lifetime suicide attempt history became weaker as TB increased. As seen in Figure 2, those with low TB and alcohol-related severity had the lowest estimated probability of a lifetime suicide attempt history, whereas those with low TB and high alcohol-related severity had the highest estimated probability. These results indicate that the potential protective role of feelings of belonging dissipates as individuals’ alcohol-related severity increased.

FIGURE 2.

FIGURE 2

Open in a new tab

The simple slope analysis for the significant interaction between alcohol-related severity (AUDIT scores) and thwarted belongingness (TB) predicting no suicide-related history vs. suicide attempt history, adjusting for additional covariates. This figure shows the association between AUDIT scores and suicide attempt history status at average (M), high (+1 SD), and low (−1 SD) levels of TB. This figure was created using the coefficients from mean-centered predictors, including the mean-centered AUDIT scores; however, the non-mean-centered AUDIT scores are presented on the x-axis

4 |. Discussion

In the current study, we tested the moderating role of TB on CJ involvement, alcohol-related severity, and drug-related severity in relation to lifetime suicide-related history among a high-risk group of individuals receiving residential substance use treatment. Indeed, approximately 45% of the current sample reported a lifetime suicide ideation or suicide attempt history. Of note, 25% reported a lifetime suicide attempt history, which is substantially higher than the general population (2.7%; Nock et al., 2008). Therefore, this high-risk group requires additional attention given that they fall at the intersection of multiple public health crises: substance use, physical and mental health condition comorbidity, the influx of physical and mental health conditions in the CJ system, and suicide (Al-Rousan et al., 2017; Csete et al., 2016). Focusing our attention on high-risk individuals who are involved in multiple systems (i.e., substance use treatment and the CJ system) will enable us to best coordinate suicide risk conceptualization, assessment, and treatment.

In line with previous literature (e.g., Yuodelis & Ries, 2015) and our hypotheses, alcohol- and drug-related severity were associated with an increased likelihood of a lifetime suicide ideation and suicide attempt history. Interestingly, TB was only associated with an increased likelihood of lifetime suicide ideation history. This is consistent with suicide theories that emphasize TB primarily in relation to suicidal desire, whereas other suicide capability factors transition an individual from suicidal desire to suicidal behaviors (Klonsky & May, 2015; O’Connor, 2011; Van Orden et al., 2010). Contrary to our hypotheses and previous findings that suggest CJ involvement is associated with suicide risk (e.g., Cook, 2013), current CJ involvement was not uniquely associated with suicide-related history outcomes. These findings suggest alcohol- and drug-related severity and TB, rather than current CJ involvement more generally, appear to be independently linked to suicidal thoughts and behaviors histories. CJ involvement, alcohol- and drug-related severity, and TB appear to be unique but co-occurring experiences among substance use residential patients; therefore, intervening on only one of these experiences within one system may not be sufficient to mitigate suicide risk and distress.

In addition to examining whether these interpersonal and contextual risk factors are uniquely associated with lifetime suicide-related history, we examined the combinatory effects of multiple risk factors. We hypothesized that increased TB would strengthen the positive associations between alcohol- and drug-related severity, current CJ involvement, and suicide-related history. As hypothesized, there was a significant interaction between current CJ involvement and TB. Specifically, those who reported current CJ involvement and high TB were more likely to have a lifetime suicide ideation history; however, for those who reported low or average TB, CJ involvement was not associated with a lifetime suicide ideation history. This suggests that individuals are more likely to report a lifetime suicide ideation history when they are CJ-involved and perceive that they do not belong with others or society. Additionally, even average levels of belonging buffer against the negative effects of stress related to CJ involvement. Our findings are congruent with previous findings that CJ-involved individuals with a suicide attempt history are more likely to have relationship problems (Hakansson et al., 2010), and psychiatric inpatients who reported elevated TB and more time spent with criminals were most likely to also report “extreme” suicide ideation distress (Mitchell et al., 2019).

Further, we found a significant interaction between alcohol-related severity and TB predicting a lifetime suicide attempt history; however, the nature of this interaction was contrary to our hypotheses. Specifically, individuals high in TB evidenced the weakest relation between alcohol-related severity and lifetime suicide attempt history. Individuals low in TB and who endorsed the lowest alcohol-related severity score had the lowest likelihood of a lifetime suicide attempt history (17%); however, individuals low in TB and who had the highest alcohol-related severity score had the highest likelihood of a lifetime suicide attempt history (47%). Additionally, among those high in TB, the estimated lifetime suicide attempt history probability stayed relatively stable, between 35% and 37%, regardless of the alcohol-related severity score. In other words, greater alcohol-related severity in the context of feelings of belonging is associated with elevated risk for a lifetime suicide attempt history. One possible explanation for these findings is that current supportive social relationships (low TB) in the context of greater alcohol-related severity may strain social relationships and produce additional social risk factors, such as feelings of perceived burdensomeness (i.e., feelings of self-hate and liability to others) and humiliation (e.g., Van Orden et al., 2010), that elevate distress and increase suicide risk. Another possibility is that those who have supportive relationships (low TB) but also have higher alcohol-related severity may have lowered inhibitions and impulsivity due to their alcohol use (e.g., Conner et al., 2006; Pompili et al., 2010) that increases suicide risk despite having another protective factor (e.g., low TB). These possibilities should be directly tested in future work. Notably, those high in TB evidenced a stable, but elevated, likelihood of a suicide attempt history across alcohol-related severity scores when examined in the context of multiple risk factors. This suggests that lacking social connection may dampen the impact of alcohol-related severity on suicide attempts, but the risk for suicide attempts is elevated. This finding is preliminary and requires replication.

Identifying risk factors for suicidal thoughts and behaviors is crucial, especially for understudied, high-risk populations within multiple systems and agencies. However, given the current findings, we should continue to move toward considering combinations of risk factors to understand the complex process of suicide risk and identify individuals who are at elevated risk. This is especially important for those who engage in high-risk behaviors, such as substance use, and are more likely to be involved in separate systems that may not have the resources to address all of an individual’s needs (e.g., Al-Rousan et al., 2017). Incorporating synergistic effects of various risk factors that are particularly relevant to specific high-risk populations (e.g., residential substance use patients) may improve our ability to predict and understand suicide and improve risk management, maximizing public health import.

The current findings have important implications for the conceptualization and implementation of psychosocial interventions for individuals receiving residential substance use treatment that ameliorates these public health concerns. Although TB may have an independent association with suicide ideation history, the role of TB in association with suicide attempt history appears to be at least partially conditional on alcohol-related severity. These findings highlight the importance of emphasizing the assessment of TB among CJ-involved individuals and individuals with greater alcohol-related severity. When working with CJ-involved individuals, clinicians should consider interventions aimed at increasing positive reciprocal relationships to reduce the risk of suicide ideation. However, it may be crucial to consider the context in which social relationships are beneficial or potentially harmful in order to guide clinical intervention. For example, the literature indicates greater amounts of time spent with criminals when one has elevated TB may actually increase the risk of suicide ideation distress (Mitchell et al., 2019); thus, promoting prosocial and healthy relationships and connections with others is likely key. Additionally, system-level programs may be especially helpful to CJ-involved individuals. For example, transitioning from incarceration to the community (which may still involve community supervision) can be very stressful. Re-entry programming can help individuals reintegrate into society, improve their continuity of care, and help them regain a sense of connection with others and society (Woods et al., 2013). Our findings support the concept behind these programs; that is, coordinated continuity of care as individuals navigate between systems will likely benefit high-risk individuals at the crossroads of multiple public health crises, and these programs are well-positioned to assess and assist in managing suicide risk.

The current findings also highlight the importance of both relapse prevention and addressing social relationships among individuals with high alcohol-related severity. Clinicians should consider interventions aimed at reducing alcohol- and drug-related severity and preventing relapse (e.g., motivational interviewing, cognitive behavioral therapy, and dialectical behavior therapy; Dimeff & Linehan, 2008; Swogger et al., 2016), while also focusing on building positive social relationships and mending existing relationships (e.g., behavioral couples therapy; McGovern & Carroll, 2003; O’Farrell & Schein, 2011). This is important given that interpersonal stressful life events have been linked to increased risk for suicide attempts among adults with an alcohol use disorder (Conner et al., 2012). Addressing additional social risk factors that may result from supportive relationships, such as perceived burdensomeness, may be necessary to reduce suicide risk among individuals with high alcohol-related severity; however, this postulation requires further empirical investigation.

Despite the merits of our study, there are also limitations. The current study used cross-sectional and self-report data, and the suicide ideation and attempts variable was historical; therefore, we cannot make conclusions about temporal associations or causality. Furthermore, our findings may not generalize to other populations or other individuals who engage in alcohol and drug use given the geographical location, demographic homogeneity (largely non-Hispanic White men), and English-speaking inclusion criterion. We also did not have diagnostic data and could not incorporate this potentially relevant information into our analyses. Therefore, replication of our work is essential.

Considering the need for research focused on dynamic risk factors, future research should assess the prospective role of the variables we examined, and their synergistic effects longitudinally. Given generalizability concerns related to our findings, future research should also replicate our work with more diverse samples in other regions. The current findings suggest that belonging may not be protective in the context of high alcohol-related severity; however, future research is needed to identify other potential factors and mechanisms (e.g., perceived burdensomeness, humiliation, perceived stigma, impulsivity, and diagnostic criteria) that may explain the elevated suicide attempt history among individuals currently high in alcohol-related severity. Future research may also consider alternative measures of substance use and specific substances, as well as more detailed assessments of CJ involvement (e.g., types of offenses).

Suicide is a serious public health problem that is increasing and requires rigorous study. The current findings provide preliminary evidence for the unique and synergistic effects of various factors that span multiple public health crises in association with suicide-related experiences among individuals receiving residential substance use treatment. Additional work that focuses on elucidating the unique and combined effects of interpersonal and contextual risk factors has the potential to prevent suicide by advancing research and suicide risk conceptualization that can be used to improve assessment, tailor treatment, and better coordinate care across multiple systems for high-risk populations.

Acknowledgments

This work was supported by grants from the National Institute of Mental Health (T32 MH020061; L30 MH120575; R01 MH115922), the National Institute on Alcohol Abuse and Alcoholism (R01 AA016149), and the Scientific and Technological Research Council of Turkey to Tugba Gorgulu (1059B191800009).

References

  1. Al-Rousan T, Rubenstein L, Sieleni B, Deol H, & Wallace RB (2017). Inside the nation’s largest mental health institution: A prevalence study in a state prison system. BMC Public Health, 17, 342. 10.1186/s12889-017-4257-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Babor TF, De la Fuente JR, Saunders J, & Grant M (1992). The alcohol use disorders identification test: Guidelines for use in primary health care (WHO Publication No. 92.4) World Health Organization. [Google Scholar]
  3. Bennett T, Holloway K, & Farrington D (2008). The statistical association between drug misuse and crime: A meta-analysis. Aggression and Violent Behavior, 13, 107–118. 10.1016/j.avb.2008.02.001 [DOI] [Google Scholar]
  4. Bronson J, Stroop J, Zimmer S, & Berzofsky M (2017). Drug use, dependence, and abuse among State prisoners and jail inmates, 2007–2009. Retrieved from https://www.bjs.gov/content/pub/pdf/dudaspji0709.pdf
  5. Cocco KM, & Carey KB (1998). Psychometric properties of the drug abuse screening test in psychiatric out-patients. Psychological Assessment, 10, 408–414. 10.1037/1040-3590.10.4.408 [DOI] [Google Scholar]
  6. Conner KR, Bridge JA, Davidson DJ, Pilcher C, & Brent DA (2019). Meta-analysis of mood and substance use disorders in proximal risk for suicide deaths. Suicide & Life-Threatening Behavior, 49, 278–292. 10.1111/sltb.12422 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Conner KR, Britton PC, Sworts LM, & Joiner TE (2007). Suicide attempts among individuals with opiate dependence: The critical role of belonging. Addictive Behaviors, 32, 1395–1404. 10.1016/j.addbeh.2006.09.012 [DOI] [PubMed] [Google Scholar]
  8. Conner KR, Hesselbrock VM, Schuckit MA, Hirsch JK, Knox KL, Meldrum S, & Soyka M (2006). Precontemplated and impulsive suicide attempts among individuals with alcohol dependence. Journal of Studies on Alcohol, 67, 95–101. 10.15288/jsa.2006.67.95 [DOI] [PubMed] [Google Scholar]
  9. Conner KR, Houston RJ, Swogger MT, Conwell Y, You S, He H, Gamble SA, Watts A, & Duberstein PR (2012). Stressful life events and suicidal behavior in adults with alcohol use disorders: Role of event severity, timing, and type. Drug and Alcohol Dependence, 120, 155–161. 10.1016/j.drugalcdep.2011.07.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cook T (2013). Recent criminal offending and suicide attempts: A national sample. Social Psychiatry & Psychiatric Epidemiology, 48, 767–774. 10.1007/s00127-012-0567-9 [DOI] [PubMed] [Google Scholar]
  11. Cook TB, & Davis MS (2012). Assessing legal strains and risk of suicide using archived court data. Suicide & Life-Threatening Behavior, 42, 495–506. 10.1111/j.1943-278X.2012.00107.x [DOI] [PubMed] [Google Scholar]
  12. Csete J, Kamarulzaman A, Kazatchkine M, Altice F, Balicki M, Buxton J, Cepeda J, Comfort M, Goosby E, Goulão J, Hart C, Kerr T, Lajous AM, Lewis S, Martin N, Mejía D, Camacho A, Mathieson D, & Beyrer C (2016). Public health and international drug policy. Lancet, 387, 1427–1480. 10.1016/S0140-6736(16)00619-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Darvishi N, Farhadi M, Haghtalab T, & Poorolajal J (2015). Alcohol-related risk of suicidal ideation, suicide attempt, and completed suicide: A meta-analysis. PLoS One, 10, e0126870. 10.1371/journal.pone.0126870 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dimeff LA, & Linehan MM (2008). Dialectical behavior therapy for substance abusers. Addiction Science & Clinical Practice, 4, 39–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Donovan DM, Kivlahan DR, Doyle SR, Longabaugh R, & Greenfield SF (2006). Concurrent validity of the Alcohol Use Disorders Identification Test (AUDIT) and AUDIT zones in defining levels of severity among out-patients with alcohol dependence in the COMBINE study. Addiction, 101, 1696–1704. 10.1111/j.1360-0443.2006.01606.x [DOI] [PubMed] [Google Scholar]
  16. Dum M, Pickren J, Sobell LC, & Sobell MB (2008). Comparing the BDI-II and the PHQ-9 with outpatient substance abusers. Addictive Behaviors, 33, 381–387. 10.1016/j.addbeh.2007.09.017 [DOI] [PubMed] [Google Scholar]
  17. Feucht TE, & Gfroerer J (2011). Mental and substance use disorders among adult men on probation or parole: Some success against a persistent challenge. Retrieved from https://www.ncjrs.gov/pdffiles1/nij/235637.pdf
  18. Hakansson AA, Bradvik LL, Schlyter FF, & Berglund MM (2010). Factors associated with the history of attempted suicide: A criminal justice population examined with the Addiction Severity Index (ASI). Crisis: The Journal of Crisis Intervention and Suicide Prevention, 31, 12–21. 10.1027/0227-5910/a000008 [DOI] [PubMed] [Google Scholar]
  19. Hedegaard H, Curtin SC, & Warner M (2018). Suicide mortality in the United States, 1999–2017 (Report No. 330) Retrieved from https://www.cdc.gov/nchs/data/databriefs/db330-h.pdf [PubMed]
  20. Joiner T (2005). Why people die by suicide. Harvard University Press. [Google Scholar]
  21. Kessler RC, Borges G, & Walters EE (1999). Prevalence of and risk factors for lifetime suicide attempts in the National Comorbidity Survey. Archives of General Psychiatry, 56, 617–626. 10.1001/jama.296.17.2112 [DOI] [PubMed] [Google Scholar]
  22. Klonsky ED, & May AM (2015). The Three-Step Theory (3ST): A new theory of suicide rooted in the “ideation-to-action” framework. International Journal of Cognitive Therapy, 8, 114–129. 10.1521/ijct.2015.8.2.114 [DOI] [Google Scholar]
  23. Kovasznay B, Miraglia R, Beer R, & Way B (2004). Reducing suicides in New York state correctional facilities. Psychiatric Quarterly, 75, 61–70. 10.1023/B:PSAQ.0000007561.83444.a4 [DOI] [PubMed] [Google Scholar]
  24. McGovern MP, & Carroll KM (2003). Evidence-based practices for substance use disorders. The Psychiatric Clinics of North America, 26, 991–1010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Mitchell SM, Cukrowicz KC, Roush JF, Brown SL, Alquist JL, Bolaños AD, Morgan RD, & Poindexter EK (2019). Thwarted interpersonal needs and suicide ideation distress among psychiatric inpatients: The moderating role of criminal associates. International Journal of Offender Therapy and Comparative Criminology, 63, 2138–2156. 10.1177/0306624X19842027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Nock MK, Borges G, Bromet EJ, Alonso J, Angermeyer M, Beautrais A, Bruffaerts R, Chiu WT, de Girolamo G, Gluzman S, de Graaf R, Gureje O, Haro JM, Huang Y, Karam E, Kessler RC, Lepine JP, Levinson D, Medina-Mora ME, … De Graaf R (2008). Cross-national prevalence and risk factors for suicidal ideation, plans and attempts. The British Journal of Psychiatry, 192, 98–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. O’Connor RC (2011). The integrated motivational-volitional model of suicidal behavior. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 32, 295–298. 10.1027/0227-5910/a000120 [DOI] [PubMed] [Google Scholar]
  28. O’Farrell TJ, & Schein AZ (2011). Behavioral couples therapy for alcoholism and drug abuse. Journal of Family Psychotherapy, 22, 193–215. 10.1080/08975353.2011.602615 [DOI] [Google Scholar]
  29. Pompili M, Serafini G, Innamorati M, Dominici G, Ferracuti S, Kotzalidis GD, Serra G, Girardi P, Janiri L, Tatarelli R, Sher L, & Lester D (2010). Suicidal behavior and alcohol abuse. International Journal of Environmental Research and Public Health, 7, 1392–1431. 10.3390/ijerph7041392 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Spitzer RL, Kroenke K, & Williams JB (1999). Validation and utility of a self-report version of PRIME-MD: The PHQ primary care study: Primary care evaluation of mental disorders: Patient health questionnaire. Journal of the American Medical Association, 282, 1737–1744. 10.1001/jama.282.18.1737 [DOI] [PubMed] [Google Scholar]
  31. Swogger MT, Conner KR, Caine ED, Trabold N, Parkhurst MN, Prothero LM, & Maisto SA (2016). A test of core psychopathic traits as a moderator of the efficacy of a brief motivational intervention for substance-using offenders. Journal of Consulting and Clinical Psychology, 84, 248–258. 10.1037/ccp0000065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Van Orden KA, Cukrowicz KC, Witte TK, & Joiner TE (2012). Thwarted belongingness and perceived burdensomeness: Construct validity and psychometric properties of the Interpersonal Needs Questionnaire. Psychological Assessment, 24, 197–215. 10.1037/a0025358 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby EA, & Joiner TE (2010). The interpersonal theory of suicide. Psychological Review, 117, 575–600. 10.1037/a0018697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Woods LN, Lanza AS, Dyson W, & Gordon DM (2013). The role of prevention in promoting continuity of health care in prisoner reentry initiatives. American Journal of Public Health, 103, 830–838. 10.2105/AJPH.2012.300961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Xu J, Murphy SL, Kochanek KD, & Arias E (2020). Mortality in the United States, 2018 (Report No., 355) Retrieved from https://www.cdc.gov/nchs/data/databriefs/db355-h.pdf
  36. Yudko E, Lozhkina O, & Fouts A (2007). A comprehensive review of the psychometric properties of the Drug Abuse Screening Test. Journal of Substance Abuse Treatment, 32, 189–198. 10.1016/j.jsat.2006.08.002 [DOI] [PubMed] [Google Scholar]
  37. Yuodelis FC, & Ries RK (2015). Addiction and suicide: A review. The American Journal on Addictions, 24, 98–104. 10.1111/ajad.12185 [DOI] [PubMed] [Google Scholar]

RESOURCES