This is Archived Content

This content is available for historical purposes only. It may not reflect the current state of science or language from the National Institute on Drug Abuse (NIDA).
View current meetings on


March 27, 2003 - 12:00am to March 28, 2003 - 12:00am
Bethesda, Maryland

Sponsored by: National Institute on Drug Abuse

Meeting Summary

March 27-28, 2003, NIDA sponsored a workshop on Drug Abuse and Suicidal Behavior. Suicidal behavior includes suicidal ideation, suicide attempts, and completed suicide. The workshop focused on what we know about drug abuse and suicidal behavior and what our research needs are. Although suicide is relatively rare, most clinicians who treat drug abusers have encountered suicidal behavior, and suicide is a significant public health issue accounting for more years of productive life lost than AIDS.

This workshop grew out of NIDA's co-funding, along with other NIH institutes and HHS agencies, of an NIMH R13 grant to the University of Rochester Center for the Study and Prevention of Suicide that supports annual consensus conferences on suicide prevention. Staff from NIDA, NIMH, NIAAA, and CDC participated in discussion at the workshop and continue to work collaboratively toward enhancing the scientific knowledge base needed to reduce suicidal behavior.

The first session focused on epidemiology. Dr. Kenneth Conner presented on the risk of suicide associated with substance use disorders that updated 1997 meta-analysis by Harris and Barraclough by searching more recent publications and calculating standardized mortality ratios (SMRs). Women drug users have higher SMRs than men; this is similar to findings on alcoholism and suicide. The studies used for this analysis followed clinical samples not community based samples so there may be issues of generalizability, and nothing can be inferred about causality?substance abuse may be a risk marker rather than a causal factor. Dr. James Wines presented data from two studies of clinical populations, one retrospective and one prospective, that highlighted depressive symptoms, drug use, substance use disorders, and previous suicidal behavior as predictors of suicide attempts. He also reported on MacArthur cohort, a prospective naturalistic study of violence. In this study, drug abusing suicide attempters exhibited higher levels of impulsive attempts than attempters who were not substance abusers. Dr. Rumi Price addressed three topics: whether epidemiology can predict suicide outcomes; whether the predictors of suicide and suicidal behavior are the same; and what the role is of trauma, specifically being a Vietnam Veteran, in suicidal behavior. She presented analyses of suicidal behavior using the National Mortality Follow-back Survey and the National Comorbidity Study and data from the follow-up to the Vietnam Era Study. She concluded that 1) if the dataset is informative, suicide outcome prediction can be quite good; 2) there are common and unique predictors of suicide and suicide attempts and ideation, with substantial gender differences; and 3) Vietnam vets are at higher risk for suicide and suicidal behavior and PTSD plays a critical role in this population. Dr. Roy was the discussant for this session and pointed out that patterns of early childhood trauma might account for the data on higher suicidal behavior in drug abusers but that data was frequently not available as was other data such as measures of personality disorders. He pointed out that there are multiple risk factors including biological factors that need to be considered.

The second session focused on suicide over the course of development. Dr. Ralph Tarter presented work from his NIDA-funded center that has been conducting a longitudinal study of male children of fathers with substance use disorders (high risk group) and male children of fathers without substance use disorders (low risk group) with the aim of elucidating the role of affect and behavior dysregulation in conjunction with biological and environmental factors in the development of substance use and substance use disorders. Dr. Tarter presented data that risk factors measured across development predict the level of risk and rate of development of suicide risk and substance use disorders. Dr. Holly Wilcox presented analyses of whether the risk of suicidal ideation and attempts might be greater for early drug users compared to later users or never users. She presented results from secondary analysis of longitudinal data sets of Baltimore school children assessed from first grade through young adulthood. She found that early inhalant use was associated with later suicide attempts and suggested that early inhalant use may be either a vulnerability marker for suicide attempt or a potentially preventable early causal antecedent. Her study also found associations between youth characteristics such as MDD and aggressive behavior and suicide ideation and attempt. Dr. David Goldston presented issues related to the conceptualization and measurement of suicidal behaviors and their relationship to substance abuse over the course of development. Questions were raised regarding factors that are unique and common to trajectories of suicidal behaviors and substance use, the heterogeneity and potential for subgroups in both the populations of suicidal and substance using individuals, the possibility of bidirectional influences between substance use and suicidality, and the possibility that the relationship between substance use and suicidality may change over time (e.g., from adolescence to young adulthood). Dr. Kenneth Sher, the discussant for this session, highlighted the importance of trait--dispositional characteristics of the individual (e.g., personality, chronic psychopathology), and state--situational factors (e.g., depressive moods, acute intoxication). He suggested that state-trait models that decompose variance in behavior into both trait and state influences might be useful in understanding the relationship between drug abuse and suicidality. He noted that although there are clearly common dispositional contributions to both substance abuse and suicidal behavior (e.g., impulsivity), there also needs to be consideration of effects of the drug itself. Such effects are typically thought of as resulting from acute intoxication but delayed short-term effects such as drug withdrawal and longer term consequences such as role loss and social marginalization also need to be considered as factors in suicidal behavior. He emphasized the need to understand different trajectories of substance use and suicidal behavior and the need for more ecological data.

The third session focused on psychiatric comorbidity and treatment. Dr. Shirley Yen presented on the role of substance abuse as a predictor of suicidal behavior from a longitudinal, naturalistic study of subjects with borderline personality disorder recruited from a variety of treatment settings. Her study found a) that baseline diagnoses of a drug use disorder and borderline personality disorder were predictive of suicide attempts during the follow-up interval, and b) worsening of depression and substance use was predictive of suicide attempts within the next month. Dr. Marsha Linehan summarized the clinical trial literature on treatment of suicidality. While pharmacotherapy trials targeting depression or suicidality directly have not, with the exception of one study, been shown to reduce suicide attempts, both brief problem solving behavioral interventions as well as intensive, longer-term behavioral interventions have been shown to be effective in reducing suicide attempts among highly suicidal individuals. She highlighted the need to conduct larger trials, to stop excluding those at high risk for suicide from standard clinical trials and to include standardized assessments of suicidality across studies. Dr. Gregory Brown presented promising results in reducing subsequent suicide attempts in an urban, low-income sample of who had presented to the university emergency department following a suicide attempt. The majority of subjects were unemployed or disabled, many were homeless, and most were depressed and had a co-occurring substance use disorder. The treatment was a ten session cognitive therapy intervention conducted over a six month period that focused on problem solving, changing maladaptive beliefs and improving problem solving skills and social support, and preventing relapse. Dr. Richard Reis's discussion highlighted issues around defining treatment as usual in the studies of Dr. Linehan and Dr. Brown. In Dr. Yen's study, he asked the question whether clinicians could distinguish major depressive disorder from substance-induced mood and whether treatment approaches would differ. Dr. Reis discussed provision of services for subjects with co-occurring substance use and mental health disorders in the ?real world.? He suggested that since a large proportion of patients treated for addictive disorders have been or are suicidal, drug abuse treatment should include a focus on suicide assessment and prevention. In addition, he noted that research is needed to improve the interface of addiction treatment with suicide interventions.

The fourth session focused on suicide prevention research. Dr. Elaine Thompson presented data on school-based randomized trials of suicide interventions in high-risk youth. The students were selected for indicated prevention based on school records indicating high absenteeism and slipping grade point average. She reported results from a study that compared a small group intervention with a counselor's care crisis intervention program, and usual care, which consisted of a suicide risk assessment followed by the school's typical response in such situations. Both interventions reduced suicidal ideation and depression but didn't impact drug use, but the small group intervention was more effective in reducing anxiety in girls. Other intervention studies are ongoing as is a follow-up of young adults who had received interventions as high- risk high school students. Dr. Eric Caine emphasized the public health need for suicide prevention in the middle years, ages 20-50, which is the age group that accounts for the most years of life lost due to suicide. He described efforts by the Air Force to reduce suicide, the second leading cause of death in the Air Force, and while the Air Force is somewhat unique in having a common culture and well-educated, pre-screened members, lessons from this prevention intervention may be useful in developing other intervention programs for young- and middle adults. The Air Force program targeted the entire community and encouraged service members to seek help for mental health and substance abuse problems and targeted a family of violence related behaviors, including suicide. In his role as discussant, Dr. James Anthony called for large sample epidemiological studies that build experimental intervention research into the design, particularly, the type of low-cost interventions described by Drs. Thompson and Caine.

The fifth session focused on neurobiology and genetics. Dr. Anne Glowinski presented on the utility of twin studies for understanding the genetics of suicidal ideation, attempt, and completed suicide. These studies also afford the possibility of looking at gene environment interactions and how this affects an outcome like suicidal behavior. She described studying monozygotic twins discordant for a particular behavior such as early substance use to understand the unique experiences that could lead to phenotypic differences. She also described use of the children of twins to study gene environment interactions. Dr. Klaus Miczek discussed the use of animal models of impulsive behavior and aggressive behavior to model similar human behavior and to enable correlation real-time measurement of regional neurotransmitter (e.g., serotonin) levels with behavior. His data demonstrated that serotonin levels in prefrontal cortex do not drop as a predispositional factor in aggressive behavior but rather drop once the aggressive encounter has occurred. So serotonin deficits measured in postmortem brain may reflect this drop. Dr. Donald Dougherty described the use of human laboratory measures of impulsivity and their utility for identifying people at immediate or long-term risk for suicidal behavior, and the potential for evaluating new medications or behavioral therapies. He indicated that particular paradigms might be better at measuring trait differences while others might be better at measuring state differences in impulsivity. He demonstrated his computerized behavioral measures of impulsivity that use rapid decision paradigms (immediate and delayed memory tasks and reward choice paradigms), he noted that paradigms need to be tailored to the population being studied, i.e., a paradigm developed for children would need to be modified for use in adults. Dr. Dougherty presented data showing that adolescents with suicidal ideation and adult attempters have increased commission errors on the immediate and delayed memory tasks compared to controls. Dr. Jonathan Mann was the discussant for this session, and he emphasized that suicidal behavior is multifactorial. He presented a stress diathesis model of suicidal behavior that factors in objective states such as life events with subjective states and traits such as substance abuse interacting with reasons for living/suicidal ideation, suicidal planning/intent, impulsivity, and suicidal acts. He suggested that low serotonergic activity is related to aggression/impulsivity and suicidal behavior and described imaging studies that have shown defects in serotonergic function in the ventral prefrontal cortex and of suicide attempters with depression compared to non-attempters with depression. He also discussed the genetic and non-genetic aspects of familial transmission of suicidal behavior and the need for research on gene/environment interactions.

In the final session, each participant was asked to propose an experiment to advance the field. The suggestions included: incorporating suicide prevention into drug abuse treatment, looking especially at the effects of intoxication and withdrawal on suicide behaviors and at studies to understand the mechanism of intervention effects; adding suicidal behavior as an outcome measure in treatment studies; conducting animal studies of impulsive control and behavior regulation, including primate studies and studies in mice; longitudinal studies to understand causal relationships?with interventions added and biologic measures to get at mechanisms; conducting studies of gene/environment interactions; conducting studies of the cultural and macro level factors that influence suicidal behavior; and developing a cadre of young investigators with expertise in substance abuse and suicidology.