CHILDHOOD ADHD, COMORBIDITY AND RISK FOR LATE ADOLESCENT MENTAL HEALTH FUNCTIONING. Ken Winters1, 1University of Minnesota, Minneapolis, MN United States
Although some exceptions exist, there is a converging body of studies suggesting that ADHD children are probably at increased risk for adolescent and young adult substance use disorder (SUD) and continued problems with ADHD and related externalizing disorders (e.g., Barkley, et al., 2004; Mannuzza, et al., 1998). This risk is conferred by tendencies toward the severity of the childhood ADHD, as well as the presence of a childhood externalizing disorder (ODD, CD). However, most of these longitudinal studies track clinical samples; questions remain about the course of community derived samples whose levels of risk and functioning outcomes may differ from clinical samples.
A population that included 7,231 children in grades one through four in 22 suburban elementary schools from five independent school districts located in the outer ring of the Minneapolis metropolitan area were screened in 1991 (see August, Realmuto, Crosby, and McDonald, 1995). Of these, 318 were identified as having cross-setting disruptive behavior based on the teacher and parent Connors Hyperactivity Index (HI-T, HI-P; > 1.75 SD). This method selected a sample that was 1.75 standard deviation units above the normative mean for both measures. Of these, 125 met DSM-III-R criteria for ADHD alone or ADHD with an externalizing disorder (ODD, CD), based on the DICA-R-Parent at one or more of three assessment time points. An additional 99 low-risk control group of children were identified from the same suburban school system with a HI-T score < 1.1 SD and absent of a major DSM-III-R diagnosis.
We present late adolescent (mean age 19.7) outcomes of the three subject groups (ADHD only, ADHD + externalizing, and controls) based on four assessment waves for whom we have complete data (89% of eligible subjects). The analysis will examine the inter-relationship of 1) childhood diagnostic status, 2) childhood and adolescent utilization of treatment services (i.e., intensive prevention parenting program, pharmacotherapy and mental health counseling), and 3) late adolescent outcomes (e.g., substance use disorders, mental health, and psychosocial functioning).
Preliminary findings indicate reliable elevations in substance use disorders and poorer mental health and psychosocial outcomes for the ADHD + externalizing group when compared to both the ADHD only and comparison groups. There were no differences between the ADHD only and comparison groups. The implications of the findings in light of the community-based ADHD sample will be discussed.
THE FAST TRACK PREVENTION TRIAL AS AN EXPERIMENT TO TEST MODELS OF SUBSTANCE-USE DEVELOPMENT. Kenneth Dodge1, 1Duke University, Durham, NC United States
Conduct disorder and substance use in adolescence are known to co-occur at greater than chance rates, and developmental research has identified several common risk factors. These factors include problems in early parenting, especially inconsistent harsh discipline and inadequate monitoring and supervision; problems in peer relations, especially early peer social rejection and association with deviant peers; and problems in social cognition, especially poor problem-solving skills. What is not known is whether an intervention designed to prevent serious conduct disorder would have a secondary effect on preventing early-onset substance use during adolescence. The Fast Track Prevention Trial is ideally suited to test this hypothesis.
In Fast Track, over 9,000 kindergarten children at four sites in three cohorts were screened using a multiple-gating strategy with teacher and parent ratings to identify 891 children at high risk for conduct disorder. These children were assigned randomly (at the school level) to receive the Fast Track intervention or not. The intervention lasted 10 years, consisting of parent management training through group meetings and home visits, social-cognitive skill training through group meetings and classroom curricula, enhancement of peer relations through coaching, and tutoring in reading skills.
Assessments at the end of grade 9, structured psychiatric interviews using the Diagnostic Interview Schedule for Children (DISC) and self-reports were used to measure conduct disorder and illicit substance use.
As hypothesized, considerable comorbidity was found. Of those youth without diagnosed conduct disorder, 26 percent reported illicit substance use, whereas of those youth with diagnosed substance use, 81 percent reported substance use.
Assignment to receive the Fast Track intervention (or not) was associated with a lower probability of diagnosed conduct disorder during grade 9 for both males (8 percent vs. 13 percent) and females (3 percent versus 7 percent). Also, assignment to receive intervention was associated with lower rates of substance use for both males (30 percent vs. 41 percent) and females (26 percent vs. 37 percent). Thus, the Fast Track intervention, which was designed to address risk factors for conduct disorder, had a discernible effect on preventing substance use during adolescence.
Additional analyses will test the hypotheses that the intervention effect on substance use was mediated by its effects on: 1) the risk factors of parenting, peer relations, social cognition, and academic skills; and 2) the outcome of conduct disorder.
These findings will be discussed in terms of developmental models of substance use in adolescence.
This study is authored by the Conduct Problems Prevention Research Group.
CAN TREATMENT FOR ADOLESCENT MAJOR DEPRESSION PREVENT SUBSEQUENT SUBSTANCE ABUSE?. John Curry1, 1Duke University, Durham, NC United States
Substance use disorders and depressive disorders are two of the three most prevalent forms of psychiatric disturbance in adolescents. Moreover, depression and substance abuse are independent and interactive risk factors for suicide, a leading cause of death in this age range. Over 60% of adolescents who use, abuse, or depend on substances have comorbid psychiatric disorders, most commonly disruptive behavior disorders or depression. Depressed mood raises the risk of substance use becoming abuse, and of abuse becoming exacerbated. In families with parental substance use disorders, child depression conveys a three-fold risk of the child developing a substance use disorder.
For these reasons it is imperative to develop effective treatments for both depression and substance abuse. The Treatment for Adolescents with Depression Study (TADS) is the first adolescent depression treatment study that has systematically assessed substance use and abuse throughout the course of treatment and follow-up. Both diagnostic interview and self-report data on substance use are collected at 6-month intervals. As participants complete TADS´ one-year follow-up, we are recruiting them to engage in an extension follow-up, SOFTAD, the Substance use and other Outcomes Following Treatment for Adolescent Depression study.
In TADS, 439 12-17-year-olds with stable moderate to severe Major Depression were randomly assigned to one of four treatments: fluoxetine (FLX), cognitive behavior therapy (CBT), a combination of FLX and CBT, or clinical management with pill placebo. Acute treatment (stage 1) lasted for 12 weeks. This was followed by 6 weeks of continuation treatment (stage 2), and 18 weeks of maintenance visits at 6 week intervals (stage 3). Subsequently, subjects were followed for one year of naturalistic assessment (stage 4), and those who consent are now followed in SOFTAD for another 4 years of assessment.
To date, only TADS stage 1 efficacy results have been reported. These indicated that combination treatment had the best results in reducing both depression and suicidal ideation. In this paper, the pattern of alcohol and substance abuse during the first three stages of TADS will be presented, as a function of depression treatment efficacy, and the further questions addressed by SOFTAD will be discussed. The hypothesis being tested is that treatment that reduces depression will reduce the risk of subsequent substance abuse, i.e., treatment responders at any stage will be less likely than non-responders to demonstrate substance abuse at the subsequent stage. Secondary analyses will investigate whether the addition of psychosocial treatment significantly adds to the preventive effect for successful pharmacotherapy cases.
INDICATED PREVENTION IN THE CASE OF CHILDREN´S EXPOSURE TO VIOLENCE: DEVELOPING RESEARCH-BASED MODELS TO ADDRESS ACUTE NEEDS OF CHILDREN AND FAMILIES. Abigail Gewirtz1, Robert Murphy2, Christopher Blodgett3, Gerald August1, 1University of Minnesota, Minneapolis, MN United States; 2Duke University, Durham, NC United States; 3Washington State University, Spokane, Spokane, WA United States
Few preventive interventions have been developed that address children´s emotional and behavioral health in the immediate aftermath of their exposure to violence. Nonetheless, children are frequently directly and indirectly exposed to violence in the family and in the community (e.g., Margolin & Gordis, 2004), and, while violence exposure varies tremendously with regard to type, frequency, and severity, it is increasingly recognized as posing significant risks to the health and development of children. This symposium presents data from several sites around the country (Spokane, WA., Minneapolis, MN., and a group of Child Development Community Policing Program sites including New Haven, CT.) These sites are implementing programs partnering first responders and mental health/social service providers to provide indicated prevention services to children and families in the acute aftermath of an incident of family or community violence. Although these programs differ in approach, orientation, and scope, they have common goals, including de-escalation and containment, crisis assessment and referral for needed services; and psycho-education about the impact of exposure to violence on children. In addition, all such programs face the challenges unique to providing services to children and families immediately after violent traumatization occurs. These challenges include: entering dangerous and unstable situations, partnering with other professionals who have diverse and sometimes conflicting roles, and ongoing engagement with families in the aftermath of crisis. Individual paper presentations address core themes: (1) the complex system and service challenges inherent in a system of care initiative delivering acute crisis services in communities (2) the types of populations served, including levels of need, engagement, and history of service receipt, and (3) the challenges for prevention researchers in developing and evaluating program models of acute crisis service delivery.
PREVENTION RESEARCH CHALLENGES IN DEVELOPING MODELS OF ACUTE CRISIS RESPONSE. Abigail Gewirtz1, Amanuel Medhanie2, 1University of Minnesota, Minneapolis, MN United States; 2Tubman Family Alliance, Minneapolis, MN United States
The prevalence of children´s exposure to violence (through witnessing or direct victimization) is considerable, and an increasing body of both cross-sectional and longitudinal evidence points to the deleterious psychological effects on children of such exposure. Despite this, there is a dearth of research on the utility of indicated prevention/crisis outreach efforts in the immediate aftermath of violence. The broad aims of the acute crisis response programs described in this symposium paper include: (i) supporting adaptive coping in the acute aftermath of trauma and de-escalation of maladaptive, acute stress reactions (ii) educating caregivers about the impact of exposure to violence on children and (iii) offering brief assessment and referrals to services (i.e. acting as a gateway to longer-term mental health and social services, and in some instances actually providing those services).
The interest in developing prevention research protocols for such efforts stems from (a) the potential to defray later acuity of child symptoms through early identification and preventive intervention with traumatized/vulnerable children, (b) clinical evidence suggesting that acute response (in the immediate hours or days following a violent event) can significantly enhance families´ subsequent service engagement (and that acute intervention thus may be a key system of care portal), and (c) early evidence suggesting that acute responses show promise for reducing children's subsequent exposure to violence. However, the challenges to prevention research in this context are considerable: ethical issues with regard to research with populations traumatized or in acute crisis, defining the `preventive intervention´ and `treatment dose´ in what is an emerging intervention that may range from crisis to longer term treatment, and assuring the integrity of both primary data collection and secondary databases (e.g. police and other related data sources). This paper addresses the challenges to prevention research of developing and evaluating models of acute crisis outreach services. Sets of cross-sectional and longitudinal evaluation data are presented from five sites around the country implementing acute crisis outreach programs for families exposed to violence with differing evaluation models, and challenges to successfully implementing and developing prevention research models for this purpose are highlighted.
POLICE-MENTAL HEALTH COLLABORATION AND INDICATED PREVENTION FOR CHILDREN EXPOSED TO VIOLENCE. Robert Murphy1, Steven Marans2, Robert Rosenheck3, Steven Berkowitz2, 1Duke University, Durham, NC United States; 2Yale University, New Haven, CT United States; 3VA-NEPEC, West Haven, CT United States
Indicated prevention approaches that draw on a central collaboration of mental health and law enforcement systems in the provision and study of interventions for children exposed to violence present challenges beyond those of typical system of care initiatives. By providing interventions in the immediate aftermath of domestic and community violence, police officers, mental health providers and partners in child protection, court and domestic violence systems collaborate in a context that challenges deeply set assumptions about one another´s legal and ethical obligations to children and families. These collaborations involving direct intervention with children and families, as well as coordination of complex providers systems, in the wake of psychologically overwhelming and physically dangerous events are further complicated by the frequent presence of ongoing risk to involved families and providers alike. This form of preventive response to children exposed to violence unites law enforcement with mental health and other providers at the scenes of criminal and violence events where the importance of a systematic approach is accentuated. Concurrently circumstances related to the event and the law enforcement response may mitigate the capacity of providers to engage in a clinically sound, empirically rigorous protocol that has been designed to assist families and expand knowledge of effective prevention approaches for children exposed to violence.
In this paper, a model of collaboration and joint response involving law enforcement officers and mental health clinicians is presented. The approach is typified by core components of indicated prevention for children exposed to violence, crisis-oriented strategies for child and family stabilization, coordinated responses involving members of each organization, intensive training across disciplines and novel approaches to the study of service delivery strategies. Data from approximately 6000 children, who were subjects of police-mental health consultation for preventive clinical response subsequent to their exposure to violent or criminal circumstances that resulted in a police response, illustrate service delivery and evaluation challenges related to this unique population. Administrative data from multiple national program sites are used to characterize a model of responding to potentially traumatized children that features engagement proximal in time to exposure, titrated levels of service delivery, operationalization of proximal outcomes and methodological challenges related to evaluation in the context of crisis (Murphy, Rosenheck, Berkowitz, & Marans, In Press).
CHARACTERISTICS OF CHILDREN AND FAMILIES IDENTIFIED THROUGH CRISIS OUTREACH FOR VIOLENCE EXPOSURE. Christopher Blodgett1, 1Washington State University, Spokane, WA United States
A substantial literature describes the developmental risks resulting from children´s exposure to violence. This research has primarily relied on access to children and caregivers in treatment and shelter populations. However, families in services are a comparatively small portion of families exposed to violence each year. We know little of the needs of children and caregivers in the general population of children and families exposed to violence. Community-level outreach is needed to describe the scope of trauma as a public health problem and design effective early intervention services.
In the past decade, several communities in the US have adopted crisis outreach to families at the time of violence exposure. This work includes New Haven´s Child Development Community Policing Program (CDCP) and New Orleans´s Violence Intervention Project. This presentation overviews what has been learned from these crisis outreach programs about families´ and children´s needs regardless of their subsequent engagement with formal services. This community-level description provides an important check of what we have learned from treatment and shelter populations about child and family need.
We describe children and families identified through voluntary community crisis outreach efforts across a range of communities in the US including information from three communities implementing CDCP services and two related but distinct programs in Minneapolis MN and Spokane WA. Data from more than 7000 children and 2000 caregivers are described in this presentation. Because of community-specific interventions and assessment strategies, this presentation provides a multiple method description of children and families exposed to violence. After a description of assessment methods across the participant communities, we discuss evidence for several key themes across families and children engaged through outreach immediately following violence exposure. A conceptual framework and descriptive statistics are presented for the nature of the trauma events, symptom presentation in the children, gender and age differences in children´s response to violence, and caregivers´ histories with violence and violence related risk behaviors (e.g., mental health and substance abuse). Individual community differences are discussed as appropriate. We then describe the degree to which these crisis outreach findings agree and diverge from the literature based on shelter and treatment population studies. We suggest implications for the development of crisis outreach and treatment services research. We close with a brief discussion of the research strengths and limitations in description of children and families at the time of crisis.
TWO STUDIES OF WELL-BEING: MEASURING SUICIDE POTENTIAL AND NON-NORMATIVE BEHAVIORS AND RISKY BEHAVIORS
Chair: Celene Domitrovich
IS TEEN VEGETARIANISM A PROBLEM BEHAVIOR?. Cheryl Perry1, 1University of Minnesota, Minneapolis, MN United States
Despite decades of attempts to encourage consumption of more plant-based foods, Americans eat a primarily meat-based diet. Teen vegetarianism is clearly not normative in this society, and often perceived by adults as a sign of rebellion or anti-social behavior. But, is teen vegetarianism really part of a syndrome of problem behaviors such as alcohol and other drug use? We addressed this question in two studies. In the first, 4746 students in 31 schools were surveyed as part of a large study of adolescent eating patterns called Project EAT. The students in the sample were from the Twin Cities metropolitan area in Minnesota and had a mean age of 14.7 years; 48.5% were Caucasian and 49.8% were female. Among these students, 6% (n=262) reported being a vegetarian. Of these 73.7% were female and 47.5% Caucasian. We also asked questions concerning drug use, suicide ideation and attempts, depression, and physical activity. Among all self-reported vegetarians, they were more likely than non-vegetarians to report suicide ideation and attempts, but there were no other differences between the groups in the other behaviors. In addition, vegetarians were more likely to engage in unhealthy (i.e. laxative use) and healthy (i.e. physical activity) weight control behaviors than were non-vegetarians, and were also more likely to meet the Healthy People 2010 dietary guidelines. In a second study, one-on-one interviews, focus groups, and website chat responses from teen vegetarians were analyzed qualitatively. They were asked whether teen vegetarians were more likely to engage in other behaviors such as smoking and other drug use. Although a majority reported no differences between vegetarians and non-vegetarians, a large minority tended to be “strait-edge” and avoided tobacco, alcohol, other drugs, and sex. Teen vegetarianism in these studies was not associated with increased problem behaviors.
MEASURING ADOLESCENT POTENTIAL FOR SUICIDE. Kenneth Pike1, Elaine Walsh1, Elaine Thompson1, 1University of Washington, Seattle, WA United States
Despite leveling of youth suicide rates, suicide risk among adolescents remains a major public health concern, as it is a leading cause of adolescent death. Reaching suicidal youth in an effective and timely way remains a challenge. The design and evaluation of interventions for populations at suicide risk hinges on the development of valid, reliable, and sensitive measures of adolescent suicide potential, an acknowledged gap in this field. The comprehensive multidimensional Measure of Adolescent for Suicide (MAPS) centers on the measurement of three theoretic dimensions: direct suicide risk, related risk and protective factors. A composite suicide risk (SR) factor was created based on identified direct suicide risk factors (attitudes toward suicide, ideation, behavioral preparation, prior attempts, attempt lethality, suicide exposure). The purpose was to examine the SR factorial structure using a large randomly selected sample of high school youth, to assess the factorial structure across gender and ethnicity, to evaluate a composite suicide risk score, and to test the predictive validity of known risk and protective factors on the composite score.
Data were collected using computer-assisted in-person interviews conducted in schools. The sample included over 1500 high school adolescents ages 14-19, half of whom were known to be at suicide risk. The sample was 50% female and diverse—33% White, 29% Hispanic, 14% Black, 10% Asian/Pacific Islander, 8% Native American, and 6% multi-ethnic.
CFA results supported the proposed unidimensional SR factor model (CFI=.91) with prior attempts (λ=.94) and attempt lethality (λ=.88) the strongest indicators. Multiple group comparisons across gender and ethnic groups are underway. Alternative composite score weighting procedures were tested. Correlations between alternative weighting procedures were >.98, suggesting value of a parsimonious equal-weights method. Multiple regression demonstrated the influence of identified risk and protective factors. Importantly, for these youth, number of stressors, depression, anger, violence/victimization, alcohol use, personal control, sense of support and self-esteem were statistically significant SR predictors.
The MAPS computer-assisted assessment is designed specifically for wide-spread community implementation for mental health promotion in a variety of settings (schools, clinics, counseling centers). The assessment process is comprehensive and efficient. These results provide evidence regarding the the instrument's construct validity and cultural appropriateness. Related studies demonstrate multiple ameliorative influences of the assessment process, revealing its promise as a brief suicide preventive/health promotion intervention.