Addiction and Suicide: A Review

Introduction

Worldwide, more than 1 million people commit suicide every year. In 2010, over 38 thousand suicides were reported in the United States at a case rate of 12.1 suicides per 100,000 and over one million attempted suicide (1,2). Suicide has surpassed motor vehicle accidents as the leading cause of injury mortality and is ranked 10th leading cause of death among persons ages 10 years and older (1). There are several risk factors for suicide completion, the most important of which include psychiatric illness, previous suicide attempts, substance misuse, acute interpersonal stressors, partner-relationship disruption, and history of sexual abuse (3,4).

Suicidal thoughts, planning, attempts and completed suicides represent a continuum of suicidal behavior. The lifetime prevalence of suicidal ideation and suicide attempts in the general population of the United States has been estimated to be 13.5% and 4.6% respectively (5), and there is one death by suicide for every 25 attempted suicides (6). Importantly, a previous suicide attempt is the most powerful predictor of eventual death by suicide. Studies have reported 20–25% of suicide victims made suicide attempts in the previous year before death (7). Investigations report a suicide risk of 29 to 54-fold for men and 50 to77-fold for women with previous attempts as compared to the general population (7). Suicide rates were highest during the first 6 months after an attempt. Violent and more severe attempts further increase the risk of future suicide and multiple attempts are correlated with eventual suicide. Although many chemical dependency treatment programs will not accept clients with recent suicidal behavior, up to 40% of patients seeking treatment for substance dependence report a history of suicide attempts (8–10). Compared to the general population, those with alcohol use disorders are almost 10 times more likely to die by suicide and those who inject drugs are about 14 times more likely to commit suicide (11). Substance dependent individuals entering into treatment are at elevated risk for suicide attempts for various reasons. They often enter with depressive symptoms and a number of severe stressors (relationship loss, job loss, health and financial problems) that not only impel them to seek treatment but also put them at higher risk for suicidal behavior. Thus, suicidal behavior is a very significant problem in addiction treatment and should be assessed and treated along with the addiction. For this reason, suicide needs more clinical and research attention in the addictions field.

Behavioral Indicators of Suicide Risk

Personality Traits

Individual who attempt and die by suicide have higher ratings of personality traits such as aggression and impulsivity compared to non-attempters. Sher and colleagues demonstrated that depressed subjects with alcoholism had higher suicidal ideation, lifetime aggression, and impulsivity scores than non-alcoholic depressed patients (12). Alcohol, cocaine, and opiate dependent suicide attempters also show higher ratings of personality traits such as introversion and neuroticism compared to substance dependent patients who did not attempt suicide (9). Beck and Steer (13) prospectively followed 413 patients who were hospitalized for suicide attempts over 5 to 10 years. They reported that the degree of pessimism, hopelessness, and poor perceived social support are significant independent risk factors for suicide attempts and can predict eventual suicide in individuals with major depression and alcoholism. Other personality traits such as resilience and optimism may protect against suicide. The diagnosis of borderline personality disorder is shown to increase suicide risk in those with previous attempts, substance dependence and comorbid depression (14). Externalizing psychopathology defined as co-occurring antisocial personality disorder, impulsivity, aggressiveness and substance dependence also predicts suicide (15).

External (Environmental) Factors and Suicide Risk

Sexual Assault, Childhood Abuse, Addiction and Attempted Suicide

An epidemiological catchment area study reported that 15% of those with a history of sexual assault had attempted suicide (4). In this study by Davidson and colleagues, history of sexual assault was associated with suicide attempts after controlling for sex, age, psychiatric diagnosis or PTSD symptoms. For women, the odds of attempting suicide were 3–4 times greater if the sexual assault occurred before age 16. Victims of childhood physical, emotional and sexual abuse also showed a greater risk to develop substance use disorders compared with people without a history of child abuse (16). In a German case-control study of almost 2,000 alcohol dependent outpatients, more than 35% of female and 6% of male patients reported a history of sexual violence (17). Another cross-sectional study conducted in Hungary on patients with alcohol dependence seeking treatment found that the history of childhood physical abuse in females increased the likelihood of suicide attempts by approximately 15 times while in males, the increase was about two-fold (18). Roy and Janal (18) interviewed almost 500 abstinent alcohol dependent patients in outpatient treatment and found that a history of childhood sexual abuse was significantly correlated with suicide attempts. Another study by Tiet and colleagues reported that men with addictive disorders seeking treatment who had recent sexual abuse were more likely to attempt suicide (OR=4.8)¹⁹ and after controlling for demographic and diagnostic factors, both recent and lifetime sexual abuse were significantly associated with higher likelihood of recent suicide attempt (19).

Interpersonal Relationship Disruption, Financial Troubles and Unemployment

Using psychological autopsy methodology, studies investigating suicide deaths show that a considerable percentage of suicide victims with substance dependence had recent interpersonal loss or conflict (3,20). Murphy and colleagues (20) reported that nearly one third of alcohol dependent individuals who committed suicide had experienced loss of a close relationship within 6 weeks or less of death. Other researchers have also reported that recent partner-relationship disruption, family discord, financial trouble and unemployment were more common in alcoholic suicides as compared to non-alcoholic depressed suicide victims (3).

Biopsychosocial Theories

The Stress-Diathesis Model of suicidal behavior discussed by Mann and colleagues (21), the theory of Distal and Proximal Risk Factors detailed by Hufford (22) and the concept of Predisposing and Precipitating Factors explored by Conner and Duberstein (23) all propose that the probability of suicidal behavior is influenced by both a stressor (equivalent to a precipitating factor or proximal risk factor) and a diathesis (also known as a predisposing or distal risk factor). A stressor is acute or subacute (i.e. onset or exacerbation of psychiatric illness, acute intoxication or psychosocial crisis). The diathesis is defined as a hereditary disposition or acquired susceptibility that enhance vulnerability to suicide but also may refer to a psychosocial or cultural mediator. It may include a combination of factors such as sex, religion, chronic addiction, familial history and genetic contributions, child-hood experiences, personality characteristics, and psychosocial support as well as the availability of lethal methods. The onset of acute precipitating factors (stressors or proximal risk factors) coupled with the predisposing factors (diatheses or distal risk factors) contributes to increase the likelihood of suicidal behavior. In these models, chronic substance dependence is a predisposing factor while acute intoxication or relapse is a precipitating factor. Multiple risk factors predict greater risk for suicidal behavior in patients with addictive disorders.

Addiction and Suicide in Different Populations

Gender Differences

Although males outnumber female suicide victims by almost four to one, the association of suicide with substance use disorders (SUD) in women is markedly stronger. In an extensive meta-analysis of retrospective and prospective cohort studies, Wilcox and colleagues (11) reported a 17-fold difference in standardized mortality ratios for suicide in females with alcoholism compared to a mere five-fold difference in alcoholic males who sought treatment. Female drug users with past history of attempts are reported to have an 87-fold increase in suicides as compared to the general population.

Role of Alcohol in Suicidal Behavior

Alcohol Use Disorders and Suicide

Wilcox (11) reported an almost 10-fold increase in standardized mortality ratios for suicide among those with alcohol use disorders (AUD) in a meta-analysis of 42 different cohort studies. More recently, a prospective cohort study in Denmark following 18,000 people over 26 years found that there was an eight-fold increase in suicide deaths among individuals with an AUD diagnosis compared to those without AUD (24). Results from this large population study looked at the impact of psychiatric illness on suicide and showed that among suicide victims who were psychiatrically hospitalized, alcohol use disorders in middle aged males had the highest contribution to risk of suicide. A meta-analysis of psychological autopsy studies done by Yoshimasu, Kiyohara and Miyashita (25) demonstrated that substance-related disorders are strongly associated with suicide and the association between SUD and suicide in women was stronger than for men. This study lumped alcohol-related with drug-related disorders in calculating the crude odds ratio.

Acute Alcohol Intoxication

Acute intoxication is associated with suicidal behavior. Hufford (22) proposed that the mechanisms of alcohol intoxication precipitating suicidal behavior in both alcoholics and non-alcoholics alike are: (1) increased psychological distress, (2) increased aggressiveness, (3) increased likelihood of translation of ideation into action, (increased impulsivity) and (4) constricted cognition which impairs the use of alternate coping strategies. Cherpitel, Borges and Wilcox (26) reviewed studies on acute use of alcohol and suicide and reported a wide range of results, between 10–69% of suicide victims were positive for alcohol and 10–73% of those who attempted suicide were positive for alcohol use. Bagge and colleagues studied 192 recent suicide attempters who presented to a Level 1 trauma hospital (27). They reported that after controlling for acute drug use and negative life events, individuals are at increased risk for suicide attempts soon after drinking (OR=6.34) and that higher levels of drinking increase risk over lower levels of drinking (OR=6.13) or not drinking at all (OR=16.19) before an attempt. Another recent study by Conner and colleagues using findings from the National Violent Death Reporting System (28) found that among 37,993 decedents who used the three leading methods of suicide (firearm, hanging and poisoning), alcohol was present in over a third of cases. Although acute intoxication is not the same as alcohol abuse or dependence, the risk of suicide attempts or suicide will intensify during times of intoxication in individuals with or without a substance use disorder. This also may be particularly important for treatment providers to consider when relapse occurs in a patient in recovery who has a history of mental illness or previous suicidal behavior.

Risk Factors for Suicide Attempts in Subjects With Alcohol Dependence (AD)

There are various risk factors for suicidal behavior in those with alcoholism. Roy & Janal (8) concluded that gender, family history and childhood sexual abuse history made significant and independent contributions to risk of suicide attempt in those with AD. In this study of veterans in a substance abuse treatment program, 40% of 499 abstinent alcohol-dependent individuals had attempted suicide and they made a total of 485 attempts (mean = 2.44). Higher rates of attempts were associated with females, younger individuals, those with family history of suicidal behavior, those on antidepressants, those with higher measures of aggression and those who began heavy drinking earlier in life. Preuss and colleagues, in a 5-year prospective study of 1237 alcohol-dependent subjects (29) found that factors predicting future suicide attempts in individuals with AD included prior attempts, younger age, being separated or divorced, other drug dependence, substance-induced psychiatric disorders, and indicators of severe AD. Individuals who attempted or died by suicide were found to have a more severe form of alcoholism compared to alcoholics who never attempted suicide. Suicidal behavior is also associated with a heavier drinking pattern, younger age of onset and longer duration of alcoholism (8,29).

Risk Factors for Suicide in Those With Alcohol Dependence

Fifty-six percent of suicide victims in a U.S. psychological autopsy study had AUD (30). Among alcoholics, the greatest risk factor for suicide or a medically severe suicide attempt is having a diagnosis of major depression (3). As discussed earlier, recent partner-relationship disruption is also a highly significant risk factor for suicide in someone with AD (3,20). Other risk factors for suicide in alcoholics include: male gender, older age, unemployment, poor social support and increased levels of aggression and impulsivity (3). comorbid drug abuse (11), recent heavy drinking and severity of alcoholism (23), serious medical illness (3), and suicidal communication (3).

Drug Addiction and Suicide

Opioid Use Disorder and Suicidal Behavior

For drug addiction, association with suicidal behavior is as compelling as for alcohol. Alec Roy interviewed 527 abstinent opiate dependent patients in substance abuse treatment programs and found that 39% had attempted suicide (10). He reported that significantly more people with opiate dependence who had attempted suicide were female (p<.0001) and unemployed (p<.0006). Opiate-dependent suicide attempters were also younger and had significantly more family history of suicide and more childhood trauma as compared to those with opioid dependence who did not attempt suicide. Logistic regression revealed that family history of suicidal behavior, alcohol dependence, cocaine dependence, and antidepressant treatment were significant predictors of attempting suicide (10). Wilcox and colleagues’ meta-review of cohort studies on victims of suicide found that heroin users had a 13.5-fold increase in standardized mortality ratios for suicide as compared to a 10-fold increase for those with AUD (11).

Cocaine Dependence and Suicidal Behavior

Roy reported that 43.5% of abstinent cocaine dependent subjects had attempted suicide (9). Suicide attempters were more likely to be younger, female and report childhood trauma. They were also more likely to have family history of suicidal behavior, history of aggression, antidepressant treatment, and comorbidity with alcohol and opiate dependence. In this study, childhood trauma, comorbidity with alcohol dependence, and treatment with antidepressants were significant predictors of attempting suicide.

Polysubstance Dependence

Landheim et al (31), studied 260 Norwegian subjects with SUD who sought treatment in inpatient and outpatient settings and found that 58% of those with polysubstance dependence reported lifetime suicide attempts compared to 38% of those who were only alcohol dependent. Female polysubstance abusers had the highest frequency of suicide attempts (70%). Wilcox and colleagues, in their meta-analysis found that polydrug use resulted in a near 17-fold increase in suicide rates as compared to the general population (11).

Co-Occurring Mental Illness, Addiction, and Suicide

Major Depressive Disorder (MDD), Addiction, and Suicide

Sher and colleagues (32) studied outpatient subjects diagnosed with MDD and compared those without substance abuse to those with alcohol dependence (AD). Those with MDD plus AD were younger at first hospitalization, initial depressive episode, and first suicide attempt compared to those with MDD alone. Depressed alcoholics reported more previous episodes of MDD as well as more suicide attempts and recent adverse life events. They also had higher lifetime aggression, impulsivity and hostility scores and were more likely to report tobacco dependence, a lifetime history of abuse, and family history of AUD. There is a tremendous age difference in this cohort: Conner, Beautrais and Conwell (33) found that co-morbidity of MDD and AD increased risk of suicide by 4.5 fold in 20-year-olds and 83-fold in 50-year-olds.

Bipolar Affective Disorder (BAD), Addiction, and Suicide

Co-morbid SUD worsens the course of illness and treatment outcome in BAD and is associated with increased violence against self and others. Comtois et al (34) reported that in bipolar patients acutely admitted to urban county psychiatric unit, the presence of SUD doubled and SUD plus substance-induced symptoms tripled the risk for suicidal behavior. In the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), Oquendo and colleagues reported that among individuals with BAD, those with AUD were more than twice as likely to attempt suicide than those without AUD (35) (OR = 2.25). Among 138 bipolar subjects recruited from community referrals, Sublette and colleagues (36) found that SUD was associated with suicide attempts in bipolar I (BAD-I) but not bipolar II (BAD-II) and the presence of both AUD and DUD increased the odds of suicide attempts in a multiplicative fashion. In this study, 97% of BAD-I patients with AUD and DUD, 93% of those with BAD-I and DUD, and 89% of those with BAD-I and AUD had made a suicide attempt. For those with BAD-II who did not have AUD or DUD, 67% had a history of suicide attempts. In addition, higher impulsivity, hostility and aggression scores were associated with both DUD and BAD-I.

Schizophrenia, Addiction, and Suicide

The association of schizophrenia and addiction with suicide is weaker than that of affective disorders. However, Melle and colleagues found that the presence of drug abuse predicted severe suicidality in the first 2 years of treatment in individuals with schizophrenia spectrum psychosis (37). Hawton and colleagues also found that DUD was associated with eventual suicide in those with schizophrenia (OR = 3.21) (38).

PTSD, Addiction, and Suicide

There are several recent studies on the subject of PTSD and suicidal behavior (39,40). PTSD and SUD are frequently comorbid and each is associated with a higher risk of suicide attempts (40) and suicide (39). In regard to PTSD in suicide victims, a recent large population based study (39) done over a 12-year period in Denmark demonstrated that the odds ratio associating PTSD with suicide was 9.8 and the association remained after controlling for all psychiatric and demographic confounders (OR = 5.3). In this study, depression and SUD were the most frequent psychiatric disorders co-occurring with PTSD in suicide victims. Rojas and colleagues (40) examined the link between PTSD, AD and suicide attempts by analyzing the National Co-morbidity Survey-Replication data. They found that among individuals with PTSD who had endorsed suicidal ideation, co-morbid AD was associated with a six-fold increase in the likelihood of attempting suicide. The cross-sectional analysis demonstrated that almost 80% of individuals who had suicidal ideation and met diagnostic criteria for lifetime PTSD and AD had attempted suicide. Since the data are cross sectional, they cannot rule out that the AD followed suicide attempts. Prospective studies are needed to further investigate suicidal behavior in those with current diagnoses of PTSD and AD as well as with drug use disorders.

Personality Disorders, Addiction and Suicidal Behavior

The co-occurrence of personality disorder with addiction is very common in those seeking treatment for SUD. Preuss and colleagues (41) reported that 55% of addiction treatment-seeking alcoholics were diagnosed with a personality disorder. Those with AD plus personality disorders had higher rates of suicide attempts as compared to those with AD without personality disorders.

In this study, 28% of the female subjects with AD were diagnosed with borderline personality disorder (BPD) and 21% of the males had antisocial personality disorder. Subjects with AD plus personality disorders, (in particular BPD) had higher rates of suicide attempts as compared to alcoholics without personality disorders. The association of suicidal behavior with BPD is well known. Kolla and colleagues (42) estimated that patients with BPD comprise from 9 to 33% of all suicides and they concluded that both MDD and SUD present an elevated risk for suicide in those with BPD.

Treatment Implications

Suicide Risk Assessment and Treatment for Individuals With Addictive Disorders

The Center for Substance Abuse Treatment has a best practice guideline for suicide risk assessment and treatment: the Treatment Improvement Protocol (TIP) Series 50 entitled “Addressing Suicidal Thoughts and Behaviors in Substance Abuse Treatment” (43) which can be accessed via the Internet at http://www.kap.samhsa.gov A Companion video for TIP 50 summarizing the recommendations and providing vignettes is also available and can be downloaded from the Substance Abuse and Mental Health Services Administration (SAMHSA) online publication store. TIP 50 is a valuable resource for providers to learn empirically tested assessment methods and treatment plans for suicidality. To address the need for improved strategies to prevent future suicide and suicide attempts in addiction patients, the Preventing Addiction Related Suicide (PARS) module has also been developed and assessed in a recent pilot study (44). PARS was developed to be a group based, secondary suicide prevention module, built to fit into standard intensive outpatient group addiction treatment programs. It provides the addiction group participants with an overview of risk factors for suicide, discusses warning signs of suicide and the relation between suicide and addiction, and reviews steps that one can take to address suicidal thoughts in themselves or others. PARS was designed to enhance knowledge and encourage group discussion about coping with suicidal thoughts. Outcomes were positive in the initial pilot study. The authors reported that significantly more outpatient addiction patients showed help-seeking behaviors when issues of suicide arose compared to 1 month before the PARS. Receiving the PARS was also associated with significant increases in knowledge and decreased maladaptive attitudes about suicide as well as the increased understanding of the clear-cut relationship between addiction and suicide. Finally, addiction treatment counselors who often feel unprepared to deal with suicidal behavior also responded positively to the three hour training in suicide prevention provided by PARS.

There is strong interplay between SUD, psychiatric illness, psychosocial stressors and suicide potential. Given this, all clients evaluated and treated for addiction should also be screened and monitored for suicidality. Addiction treatment providers should be prepared to routinely gather information about client’s suicidal histories, thoughts, and plans and determine if important risk factors are present in order to assess suicidality at various points in treatment. In general, the greater the burden of risk factors, the more likely suicidal behavior will manifest. However, there are also protective factors that mitigate suicide risk and it is important to identify those factors. Personality traits such as resilience and optimism have been reported to be protective against suicide. Psychosocial stability as manifested by having employment, a supportive spouse or family, and having children modifies suicide risk and individuals often cite their religious beliefs as strong reasons to not commit suicide. A treatment plan should be developed to address suicidality in those with multiple risk factors. Treatment includes exploring the natural ambivalence regarding living or dying that any suicidal person will have as well as investigating past suicidal behavior, the extent of current suicidal planning, and whether the patient has the means to execute the plan. Allowing individuals to explore reasons why they want to live and why they haven’t attempted suicide is helpful to remind them why living is important. The provider should assist the suicidal individual in developing safety plans for how to weather suicidal thoughts, prevent more serious actions, and when or where to seek immediate help. The integration of psychiatric assessment and care along with addiction treatment plays an important role in stabilizing and supporting the person with dual diagnosis who has substantial risk for suicide. Medication plays an important role for those with co-occurring disorders and evidence supports the efficacy of pharmacological treatment and cognitive behavioral therapy in conjunction with addiction treatment in preventing suicidal behavior. Patients with addictions and co-occurring disorders often die from causes other than suicide. For example accidental death from overdose is a much more common cause of death than suicide for these individuals. Providers will find themselves faced with patient overdose deaths that may or may not be related to suicide. It can be very difficult to determine the role of suicidal ideation in these cases. However, if suicidality has been assessed and treated prior to the overdose, providers as well as cohort and family survivors will have a much clearer perspective of this unfortunate outcome.

Integration of Addiction and Mental Health Services With Primary Care

Integration of addiction treatment in the mental health arena is a potent intervention for suicidal persons with co-occurring disorders. Addiction treatment by itself has been shown to have a strong effect on reducing the potential for future suicidal behavior. Although no randomized trials of suicide prevention have been done using addiction treatment, large cohort studies have found markedly decreased rates of suicide attempts in addiction treated patients (45). Primary care providers also increasingly find themselves faced with the need to assess, treat or refer individuals with co-occurring disorders and suicidal behavior. Often these areas are neglected with potential disastrous consequences. Much as the chemical dependency and mental health providers should expand and integrate their repertoire of assessment and treatment skills, so do primary care providers who need to know how to assess, triage, manage and refer individuals with addiction and suicidal behavior. Screening, referral and appropriate treatment of co-occurring disorders may well prove to reduce the high prevalence of both attempted suicide and death by suicide.

Summary

Both chronic addiction and acute intoxication will potentiate suicidal behavior. The degree of severity of the additive disorder(s) also impacts the risk for suicidality. Although suicidal behaviors have different risk factors than completed suicide, a history of attempted suicide is a strong predictor of future suicide. Personality traits such as impulsiveness, aggression, pessimism and hopelessness; acute stressors such as disruption of relationships or loss of income; and a history of childhood and sexual abuse all factor into the risk of suicide for an individual with an addictive disorder. Specific mental disorders such as major depression, bipolar disorder, PTSD and borderline personality disorder also strongly impact the risk for suicide and suicide attempts in patients with SUD. As suicide is a rare event, it is difficult to study risk and protective factors. Psychological autopsy studies of suicide deaths, epidemiological studies of different populations and prospective studies of populations at risk for suicide are numerous. Further research studies on risk factors for suicidality in individuals with drug use disorders as well as prospective studies on effectiveness of treatments for suicidal patients with SUD are necessary. Ultimately, the more we understand the complex interplay between these predictors of suicidality and substance use disorders, the more successful we will be in developing secondary prevention models to improve outcomes for our treatment population.