Trauma and Suicide Attempts among Insanity Acquittees

Discussion

To our knowledge, this is the first study of a forensic population that assesses suicidality in relation to trauma history. NGRI patients were found to have a 45 percent prevalence of suicide attempts, comparable with Mitchell et al.’s finding of 46 percent.¹⁰ Both studies indicate that insanity acquittees have higher rates of suicide attempts compared with the general population (0.6%), general psychiatric patients (41%), and prisoners (27%).^11,–,13 More than one in five participants reported greater than one previous suicide attempt. That is 36 times higher than the prevalence of any attempt¹¹ and higher than the proportion of multiple attempters among all attempters in general populations (48% to 31%).³⁵ These findings raise the question of what factors account for the high prevalence of suicide attempts among forensic populations. We hope that the knowledge of the high rates of suicidality will motivate clinicians to do more suicide-related assessments and interventions in forensic populations, and for researchers to further elucidate suicide risk factors (including and beyond what we describe below) and develop specific treatment interventions.

As we hypothesized, trauma was strongly predictive of suicide attempts. Number of traumas and ACEs were found to account for 16 percent of the variance in the number of suicide attempts, more than any other identifiable cause of variance in this study. Several ACEs, including emotional abuse, emotional neglect, sexual abuse, physical neglect, and living with a household member with substance-related problems were significantly associated with a higher number of suicide attempts. Childhood physical abuse was the only trauma-related variable studied that was significantly associated with suicide attempts, increasing the odds of attempting by three times. ACEs are known risk factors for poor adult mental health, suicide attempts, and increased number of suicide attempts,^{18,24,25,36,37} which was corroborated by our findings. According to the interpersonal theory of suicide, interpersonal traumas like abuse may place an individual at risk for suicide when triggering feelings of social isolation, being a burden, and habituation in response to repeated exposure to painful or fear-inducing experiences.³⁸ Although knowledge that traumas can lead to negative sequalae (including suicidality) may motivate clinicians, researchers, and politicians to examine ways to prevent or provide early interventions for ACEs, once trauma survivors reach the criminal justice system, our main task is preventing suicide. Clinicians can review the above factors when assessing risk, and researchers may further assess these risk factors to understand causation, mediation, etc. and determine effective interventions.

Just as has been found in general^27,39 and psychiatric populations,^27,38 we found that, among NGRI patients, women were significantly more likely to attempt suicide than men. Women were also found to experience significantly more ACEs. Because ACEs are suicide risk factors, this could at least partially explain why women were more likely to attempt suicide.⁴⁰ This is consistent with our hierarchical multiple regression. It is also consistent with previous findings that many trauma-related problems, including suicide attempts, are at least partially explained by higher rates of trauma (especially sexual trauma) among women.⁴¹ Therefore, although gender may be used in risk assessment, it is likely that trauma and trauma type are more important indicators for determining risk and for targeting in trauma-focused and trauma-informed interventions.

We found that insanity acquittees with a history of any substance-related problems, especially alcohol-related problems, had significantly more suicide attempts than those without substance-related problems. Those with stimulant-related problems were also found to be three times more likely to attempt suicide, though this was not statistically significant. Previous studies have identified substance^12,22 and, specifically, alcohol-related problems^22,39 as suicide risk factors. Potential explanations include depressed, anxious or suicidal individuals using substances to self-medicate their psychiatric symptoms, substances exacerbating these symptoms, substances causing disinhibition and impaired judgment, or substances easing the feelings of distress and guilt when attempting suicide.⁴² In addition, untreated alcohol use disorder can lead to social withdrawal and social marginalization which may lead to an increased risk for suicide.⁴² Suicide, substances, and trauma likely have complex interactions. For example, in addition to suicide attempts, ACE scores were also significantly associated with substance-related problems and arousal symptoms. Past research has indicated that trauma, particularly in childhood, increases the risk of substance-related problems, which also increases suicide risk.^18,19 Therefore, trauma may lead to suicide attempts, directly and indirectly, as a result of substance-related problems. We recommend clinicians assess substance use history, in general, but also as it relates to suicide risk and, when present, ensure that substance use is addressed as part of a comprehensive treatment plan that addresses mental illness, substance use, suicide, and violence.

Regarding PTSD symptoms, the NACM and arousal clusters were significantly associated with attempting suicide and repeated attempts, respectively. Among individual symptoms, excessive startle and inattention (both from the arousal cluster) and negative emotions significantly correlated with higher number of attempts. This is consistent with studies finding arousal to be the cluster most associated with suicidality,^43,–,45,46 and links between agitation and suicide in inpatient and incarcerated settings.^47,48 Negative emotions and anhedonia (both from the NACM cluster) were the only individual symptoms to significantly correlate with attempting suicide. This is consistent with previous studies that have found these two symptoms to predict suicidal ideation, suicide attempts, and self-injurious behavior.^22,30,49 In our experience, if PTSD symptoms are addressed at all in treatment, the focus tends to be about externalizing symptoms (e.g., recklessness, irritability). Our findings speak to the importance of addressing internalizing symptoms, which most evidence-based trauma-focused therapies do. We encourage clinicians to ensure that trauma-focused psychotherapy modalities are available to forensic patients.

Because the indicated PTSD symptoms were significantly associated with suicide attempts using binary logistic regression, but not in the intercorrelations and hierarchical multiple regression, we theorize that PTSD symptoms are likely a mediator between ACEs and suicidality. For example, among intercorrelations, arousal was significantly associated with ACEs, total traumas, and NACMs, but not suicide attempts. Whether they are direct risk factors or mediators, our findings are consistent with the escape theory, in which those experiencing trauma-related hyperarousal or negative emotions may use suicide as a means of escaping these aversive states.⁵⁰ In addition, or alternatively, considering that negative emotions, anhedonia, and inattention are symptoms of both PTSD and major depressive episodes, it may be that depressive comorbidities are the true suicide risk factor, though that is not consistent with our lack of significant findings in the psychiatric diagnoses analysis.

We did not find any significant correlations between diagnosis and suicide attempts. This may be explained by high rates of serious mental illness across the entire sample. The NGRI population represents an intersection between psychiatric conditions and criminality, both of which are associated with increased suicide risk. Previous studies have identified links between antisocial personality and suicide attempts among prison inmates.^51,52 Although this correlation was nonsignificant in our study, this may be due to the small n, considering that four of five individuals with antisocial personality disorder (80%) attempted suicide.

Limitations of the study include cross-sectional methodology, self-report bias, and a small n of women. As with all survey studies of suicide attempts, we analyzed survivors, which limits conclusions. Cross-sections limit our ability to assess relationships of the timing of suicide attempts, which could have theoretically occurred before or after traumatic events and before or after NGRI adjudication. Although a self-reported survey method might lead to the possibility of participants reporting inaccurate memories of traumas and past suicide attempts, especially since the surveyed population was severely mentally ill, there is no way of independently verifying the reported histories. Although childhood sexual abuse, as assessed by both the THS and ACE Questionnaire, was associated with both suicide attempt prevalence and number of suicide attempts, each of these two latter variables was significantly correlated with childhood physical abuse as assessed with THS but not ACE. It should be noted that those with ACE childhood physical abuse were twice as likely to attempt suicide as those without, but the difference was not found to be statistically significant. As we have reported elsewhere, ACE has a broader definition of childhood physical abuse while THS is narrower, indicating the importance of how questions are asked and that there are divergent opinions about what constitutes a trauma.³¹

In conclusion, we found high rates of suicide attempts among the insanity acquittees in our study population. As expected, suicide attempts were significantly associated with ACEs, number of traumas, substance-related problems, PTSD arousal symptoms, and NACMs. Some factors distinguished those more likely to attempt suicide: being a woman, worse negative emotions, worse anhedonia, and experiencing childhood physical abuse. Among those with a history of suicide attempts, multiple attempters were distinguished by ACEs (especially emotional neglect and abuse, sexual abuse, physical neglect, and household members with substance problems), number of traumas, substance-related problems (especially alcohol), arousal symptoms (especially excessive startle and inattention) and negative emotions. We hope that this first study of suicide and trauma among a forensic population not only increases awareness and leads forensic staff to better assess trauma and suicide histories in forensic populations, but also leads to further research to identify risk and protective factors. This should improve our ability to prevent suicides in this population by better screening for risk factors and intervening when individuals have them and allow us to deliver care that is more trauma-informed.