Abstract
Substance use disorders (SUDs) in youth are strongly associated with aggression, delinquency, and involvement with the juvenile justice and mental health systems. This article reviews the relationship between aggression and SUDs and discusses evidence-based approaches to assessment and intervention, with a focus on youth in secure settings. While evidence indicates etiological overlap, SUDs also confer risk for aggression and delinquent behavior. SUDs and aggression are each influenced by executive functions that develop as youth transition toward adult roles. Additionally, the effects of substance use on the adolescent brain impair neurocognitive function and increase the risk for aggression and further substance use. In terms of assessment, it is important to identify function and form of aggression in order to understand motives and associations with substance use and to select appropriate interventions. Evidence-based screening and assessment of aggression, substance involvement, and related domains is also critical. In terms of treatment, youth with SUDs tend to be underserved, particularly when they are also involved with the juvenile justice system. Multiple modes of evidence-based treatment for substance use are available. Approaches that address risk factors common to SUDs and aggression across multiple domains (e.g., family therapies) have been found to be most effective but may be difficult to adapt for use in secure settings. Individual therapy approaches also have empirical support and may generally be more practical in secure settings.
Youth substance use disorders (SUDs) are strongly associated with involvement in the juvenile justice and mental health systems (Chassin, 2008; Grisso & Underwood, 2004). School and community samples have yielded SUD prevalence estimates of 6% to 10% (Kandel, Johnson, & Bird, 1999; Rohde, Lewinsohn, & Seely, 1996), compared with 62% to 81% (Aarons, Brown, Hough, Garland, & Wood, 2001; Milin, Halikas, Meller, & Morse, 1991) and 33% to 41% (Aarons et al., 2001; Deas-Nesmith, Campbell, & Brady, 1998; Piazza, 1996) in the juvenile justice and mental health systems, respectively. Consequently, it is important that clinicians working with youth be sensitive to the possibility of SUDs even for youth whose involvement in these systems is not explicitly SUD-related.
Substance use is associated with recidivism (Young, Dembo, & Henderson, 2007), likely reflecting etiological overlap and shared risk factors (Mason & Windle, 2002). Substance use also contributes to behaviors that result in juvenile justice involvement. For example, early use is a strong predictor of violent behavior (J. D. Hawkins et al., 2000). SUD interventions also reduce behavioral risks for juvenile justice involvement (Fisher & Harrison, 2000). This is particularly important in light of evidence that most juvenile-justice-involved youth in need of SUD treatment do not receive it (Nissen, 2007), indicating that improving treatment access within the system may reduce recidivism.
Substance-involved youth are also likely to become involved in the mental health system, and the majority meet criteria for another Axis I disorder (Deas, 2006). There are multiple pathways through which co-occurrence may develop. Some disorders, including mood and conduct disorders, are associated with increased SUD risk (Armstrong & Costello, 2002). Alternatively, SUDs, especially with early onset, may increase the risk of other psychiatric conditions (Armstrong & Costello, 2002; Lamps, Sood, & Sood, 2008). Additionally, there is evidence of neurobiological and genetic overlap between SUDs and other disorders (McQuown, Belluzzi, & Leslie, 2007).
The goals of this article are to provide an overview of the relationship between aggression and substance use in youth and to discuss the implications of this relationship for clinical work in secure settings. Below, we broadly describe aggression and substance use in the context of adolescent social and neurocognitive development, and we discuss assessment of aggression and substance use. These issues are then tied into discussion of evidence-based approaches to treatment of substance abuse and aggression or related problems such as delinquency and involvement with the juvenile justice system.
Developmental Considerations
It is well-established that substance use has negative consequences for the brain (Oscar-Berman & Marinkovic, 2003) but only recently has the impact on the adolescent brain been explored (Squeglia, Jacobus, & Tapert, 2009). Although the adolescent brain may be more resistant to the neurotoxic effects of substance use, drug exposure may interrupt brain development (D. B. Clark, Thatcher, & Tapert, 2008; Crews, He, & Hodge, 2007), potentially leading to cognitive, social, and psychological deficits (Squeglia et al., 2009). Normal adolescent brain development includes changes in efficiency and specialization, via synaptic refinement and myelination (Yakovlev & Lecours, 1967), and changes in prefrontal and limbic systems thought to contribute to normative adolescent impulsivity, risk-taking, and substance use (Casey, Jones, & Hare, 2008; Chambers & Potenza, 2003; Spear, 2000). Adolescent substance use has been linked to subsequent changes in brain volume, white matter integrity, and function and to neurocognitive deficits, including spatial, learning, memory, and inhibitory dysfunction (De Bellis et al., 2000; De Bellis et al., 2005; Medina, Schweinsburg, Cohen-Zion, Nagel, & Tapert, 2006; Squeglia et al., 2009). These deficits may result from substance-related reductions in the volume of important brain structures (e.g., hippocampus) (Squeglia et al., 2009) and may increase the risk for aggression in addition to the risk for further substance use. For example, inhibitory (Blair, 2001; Kirisci, Tarter, Vanyukov, Reynolds, & Habeych, 2004) and executive (Morgan & Lilienfeld, 2000) dysfunction are associated with both SUDs and aggression.
Social factors also play a role in linking aggression and SUDs in adolescence. Following the onset of puberty, there are normative increases in risk-taking (Reyna & Farley, 2006) and social affiliation drives that enhance substance use and delinquency. These changes are particularly problematic for youth who associate with deviant peers; such youth are themselves more likely to engage in substance use, aggression, and other problem behaviors (Allen, Porter, & McFarland, 2006; Sullivan, 2006). Evidence indicates that teens are far more likely to take risks in the presence of peers (Gardner & Steinberg, 2005) or in strong affective states, suggesting that risk taking may partially reflect youth affiliative needs. Additionally, environmental stress from social role transitions fuels emotional distress, further increasing risk for impulsive decision making (Cyders & Smith, 2007) and thus for substance use and aggressive behavior.
Overall, normative brain maturation and social processes during adolescence are associated with increased impulsivity, which in turn increases the risk for both aggression and substance use. These risks are compounded by substance use during adolescence, which appears to be associated with changes in brain structure and consequent changes in neurocognitive abilities that further increase risks for substance use and for aggressive, antisocial behavior.
Assessment of Aggression and Substance Use
Childhood aggression predicts early onset and frequency of substance use in adolescence (Pulkkinen & Pitkanen, 1994), consistent with a common cause or deviance proneness model (Martel et al., 2009). This model appears to operate through both direct and indirect pathways, although underlying mechanisms are not yet clear (Zucker, 2008). Aggression has been classified in terms of both function and form. Function can be classified as either proactive (i.e., calculated and goal-oriented, motivated by external reward) or reactive (i.e., defensive, impulsive responding to threat or frustration) (Dodge & Coie, 1987). Proactive aggression is associated with delinquency and violence in youth, but reactive aggression has been a less consistent predictor (Card & Little, 2006; Fite, Stoppelbein, & Greening, 2009; Raine et al., 2006). Proactive and reactive aggression have been prospectively linked to SUDs via separate pathways (Fite, Colder, Lochman, & Wells, 2008). Proactive aggression predicts substance use directly and via association with delinquent peers, whereas reactive aggression is indirectly associated with peer delinquency and rejection by peers (Fite & Colder, 2007).
Aggression form can be categorized as either direct or relational (Card, Stucky, Sawalani, & Little, 2008). Direct aggression is defined as behavior directed at individuals with the intent to harm (Coie & Dodge, 1998), while relational aggression refers to acts intended to manipulate or damage relationships (Crick & Grotpeter, 1995). Early direct (Swaim, Deffenbacher, & Wayman, 2004) and relational (Herrenkohl, Catalano, Hemphill, & Toumbourou, 2009; Skara et al., 2008) aggression are associated with subsequent substance use. However, substance involvement may be more strongly related to direct aggression in boys and to relational aggression in girls (Skara et al., 2008). Additionally, youth with high levels of both direct and relational aggression appear to have the worst substance use outcomes (Herrenkohl et al., 2009).
These studies highlight the importance of examining relationships of substance involvement with different functions and forms of aggression separately. Assessing these aspects of aggression is crucial for understanding aggressive behaviors and underlying motives, their associations with substance use, and potential targets of intervention. It is also important that assessment of aggression and substance use be empirically based. The use of standardized, evidence-based tools allows for comparisons across settings and aids our understanding of the scope of problems such as aggression and substance involvement (Wasserman et al., 2003). At the individual level, evidence-based assessment may reduce bias that can influence treatment and placement recommendations (Niarhos & Routh, 1992).
A number of factors have been identified as common risks for both aggression and SUDs in youth. Individual common risk factors include impulsivity and risk-taking tendencies (Cooper, Wood, Orcutt, & Albino, 2003; Farrington, 1989) and early manifestations of violence or delinquency, and low academic achievement (Herrenkohl et al., 2000). Importantly, there is evidence that individual risk factors may lead to different types of delinquent or aggressive behavior, depending on sex. Researchers have suggested that early childhood impulsivity is typically the first step on the developmental pathway toward delinquency, substance use, and criminality (Beauchaine, Klein, Crowell, Derbridge, & Gatzke-Kopp, 2009). Impulsivity is about 80% heritable and increases risk for multiple externalizing disorders (Crowell, Beauchaine, & Lenzenwger, 2008; Kendler, Prescott, Myers, & Neale, 2003; Krueger & Markon, 2006; Price, Simonoff, Walderman, Asherson, & Plomin, 2001; Sherman, Iacono, & McGue, 1997); it appears to result from dysfunction in serotonergic and dopaminergic brain systems (Beauchaine et al., 2009). A number of genes that influence serotonergic and dopaminergic neurotransmission are associated with vulnerability to impulsive, aggressive behavior, but in at least some cases (e.g., MAO-A, COMT, 5HTT) the behavioral expression of this vulnerability appears to differ by sex, with males more prone to outward aggression and females to emotional instability and self-injury (Beauchaine et al., 2009; Courtet et al., 2001; Perez et al., 2007).
Family risk factors include poor family management and parental monitoring (J. D. Hawkins et al., 1998), patterns and values that model or normalize violence (Brewer, Hawkins, Catalano, & Neckerman, 1995), and parental SUDs (Brook & Brook, 1992). Children from minority ethnic groups, urban areas, and families with fewer resources are more likely to be exposed to such risks and may also be more vulnerable to developing aggressive behaviors and substance use following exposure (T. L. Brown, Miller, & Clayton, 2004; Burchinal, Roberts, Hooper, & Zeisel, 2000; Wallace & Muroff, 2002; Wilson, Hurtt, Shaw, Dishion, & Gardner, 2009). Notably, however, data indicate that childhood aggression decreases when in competent families that respond adaptively to early aggression (Andreas & Watson, 2009).
Socially, association with antisocial peers strongly predicts both aggression and substance involvement (J. D. Hawkins et al., 1998; Lipsey & Derzon, 1998). These common risk factors influence both substance use and aggression, which in turn exert reciprocal influence on each other (H. R. White, Loeber, Stouthamer-Loeber, & Farrington, 1999). Consequently, clinicians engaged in assessment and treatment of youth aggression and SUDs should focus on common risk factors; interventions that focus on aggression alone or SUDs alone are likely to be less successful (Brewer et al., 1995). In terms of moderators of peer influence, studies have suggested that delinquent behavior among girls is more closely related to peer deviance, whereas in boys this effect is mitigated by their perception of adult monitoring of behavior (O’Donnell, Richards, Pearce, & Romero, in press). Additionally, there is evidence that, relative to rural youth, urban children may be more heavily influenced by aggressive models and deviant peers in terms of behavior outside the home (Hope & Bierman, 1998; Lanza, Rhoades, Nix, & Greenberg, 2010).
Aggression screening
Several measures of aggression form are available (see Table 1), including observer rating scales (K. Brown, Atkins, Osborne, & Milnamow, 1996; Dodge & Coie, 1987) and the self report Reactive-Proactive Aggression Questionnaire (Raine et al., 2006). Two methodological issues are common to these measures. First, they typically find substantial overlap (e.g., r = .5-.7) between proactive and reactive aggression (Little, Jones, Henrich, & Hawley, 2003). Second, while designed to assess function, item wordings from these scales typically include form. Little, Jones et al. (2003) recently developed an instrument designed to distinguish function from form. With this measure, when form is controlled for, the relationship between proactive and reactive aggression is minimal (Little, Jones et al., 2003). In terms of clinical utility, in light of their brevity and good criterion validity, the measures designed to address function would appear to be useful measures for screening for potential aggression problems. However, accurate assessment of both function and form are important in identifying targets for intervention. For example, appropriate interventions for reactive relational aggression and proactive relational aggression may differ. Consequently, the more comprehensive instrument developed by Little et al. may be more useful to clinicians working with youth with identified aggression problems. However, when considering appropriate treatment targets for youth with problems in multiple domains, the comprehensive assessments described below for substance use and related domains are likely to provide the most complete information regarding risk factors for both aggression and substance involvement.
Measures Assessing Form and Function of Aggression
SUD screening
While youth in secure settings are at increased risk for substance use, the purpose of screening and assessment is to identify those needing treatment (Chassin, 2008; National Institute on Drug Abuse, 2006). Current best practices include screening followed by thorough assessment (Chassin, 2008). Screening is most effective at system entry, when all youth are screened, and when standardized instruments are used (Winters & Yifrah, 2008). While most institutions conduct screening, many only screen youth already identified as having substance-related problems, and up to half do not use standardized measures (Snyder & Sickmund, 2006). Informal questions about quantity and frequency of use are not sufficient for screening, and standardized measures are recommended (Winters & Yifrah, 2008). A number of brief, multisubstance screeners are available. For example, the CRAFFT (Knight et al., 1999) assesses dangerous use, consequences of use, using for positive or negative reinforcement, using alone, blackouts, and a desire to cut down, with good sensitivity and specificity for SUDs. Other brief screeners include the Drug Abuse Screening Test (DAST; Gavin, Ross, & Skinner, 1989; Skinner, 1982) and the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST; World Health Organization, 2002).
A limitation of self-report screeners is their reliance on honest responding. One study suggests that up to half of respondents with a positive bioassay for cocaine use denied recent use (McClelland, Teplin, & Abram, 2004). For this reason, current guidelines endorse the inclusion of objective measures (e.g., urinalysis) in screening protocols (NIDA, 2006). Some tools have been developed to indirectly screen for SUDs by assessing related risk factors, including the Problem-Oriented Screening Instrument for Teenagers (Rahdert, 1991a) and the Substance Abuse Subtle Screening Inventory (F. G. Miller, 1997). However, these measures are less useful, given their length and evidence that indirect SUD assessments tend to be less accurate (Feldstein & Miller, 2007). Effective screening may be best accomplished by combining a brief, face valid screener with objective assessment of recent use.
Comprehensive assessment tools
Positive screens should trigger a comprehensive assessment to identify needs and plan interventions. There are a number of tools with good psychometric properties for thoroughly assessing substance involvement and related domains (e.g., psychiatric history and symptoms, family functioning, school and legal status; see Table 2). The Teen version of the Addiction Severity Index, or T-ASI-2 (Kaminer, Bukstein, & Tarter, 1991) is a semistructured interview that assesses 18 substance-related domains, including substance use, service utilization, and social, family, and school functioning (Skinner, 1982), yielding scores on each domain to assist clinicians in identifying treatment targets.
Substance Use Screening and Assessment Measures
The Global Appraisal of Individual Needs (GAIN; Dennis, 1999) is a semistructured interview that assesses recent and lifetime substance use; psychiatric, school, and legal status; risk behavior; living environment; physical health; and other domains. The GAIN can be used to generate diagnoses (Dennis, White, Titus, & Unsicker, 2008) and has been extensively studied, with norms available for a variety of populations (Dennis et al., 2008). Administration requires relatively extensive training and takes up to 2 hours, although shorter forms are available.
A final comprehensive assessment tool is the self-report Personal Experience Inventory, or PEI (Winters, Stinchfield, & Henly, 1989). The PEI core module assesses drug use severity, psychosocial risk, and response distortion; supplemental modules are available for eating disorders, suicide risk, abuse history, and family drug use history. Scoring software provides nonclinical and substance-involved norms. While other comprehensive tools with similarly appropriate psychometric properties are available, the T-ASI, GAIN, and PEI were selected based on ease of administration and accessibility. In addition to providing baseline information on client needs, these measures are useful for periodically assessing response to treatment.
Treatment
Treatment for adolescent SUDs and comorbid problems such as aggression has often been poorly integrated (Lamps et al., 2008). Barriers to successful treatment include poor coordination between delivery systems (E. H. Hawkins, 2009), a relative lack of research into developmentally appropriate interventions (Lysaught & Wodarski, 1996), and a lack of funding specific to SUDs and comorbid disorders (E. H. Hawkins, 2009). Additionally, aggression is associated with poor treatment outcomes (Crowley, Mikulich, MacDonald, Young, & Zerbe, 1998). Consequently, there is increasing interest in interventions that address youth needs across multiple areas, including problem-solving and communication skills, family, mental health, and SUDs (Libby & Riggs, 2005).
When treating youth in secure settings, five general strategies are recommended (Greenwood, 2008). First, intervention should focus on malleable problem behaviors (e.g., problem-solving skills, peer associations, family dysfunction). Second, interventions should be evidence-based and tailored to individual needs. Third, institutional treatment programs should focus on youth at highest risk for relapse and recidivism, who have both the most room for improvement and for whom failure will tend to have more severe consequences (Latessa, Listwan, & Hubbard, 2005). Fourth, consistent implementation of evidence-based interventions is critical to positive outcomes. Finally, interventions are most effective when provided by mental health professionals (Greenwood, 2008).
The use of evidence-based approaches is a key component of treating youth with aggression and SUDs in secure settings. In criminal justice settings in particular, implementation of such interventions can be more difficult because of the systemic focus on public safety and control of offenders, and because providers may not be fully aware of the effectiveness literature. Additionally, SUD programs are often discretionary and therefore are eliminated when budget cuts occur (Chandler, Peters, Field, & Juliano-Bult, 2004). However, all treatments are not equally effective (W. L. White, 2005). Those with demonstrated effectiveness for SUDs are more likely to reduce recidivism and relapse (Chandler et al., 2004; Farabee, Shen, Hser, Grella, & Anglin, 2001) and the incidence of violence during incarceration (Lovell, Allen, Johnson, & Jemelka, 2001). Selected studies of evidence-based interventions for youth substance abuse are shown in Table 3.
Selected Studies of Evidence-Based Treatments for Youth Aggression and Substance Involvement
NOTE: When only odds ratios were reported or derived, effect sizes were estimated as ln(odds ratio)/1.81 (Chinn, 2000); when only a t or F statistic was reported, effect sizes were estimated as d = 2t/√df or d = 2√F/√dferror (Rosenthal, 1991). AM = anger management; BUL = bullying behavior; CC = contact control; CD = Conduct Disorder or externalizing problems; CMg = case management; EnOpt = Enhanced Outpatient Treatment; EOT = end-of-treatment; FBT = Family Behavior Therapy; GT = group therapy; INC = incarcerated offenders; IntT = Interactional Treatment; JO = juvenile offenders; JSO = juvenile sexual offenders; M = months; PA = physical abuse victims; PSY-I = psychiatric inpatients; RT = relaxation training; S = sessions; SS = social services; SUD-I = substance use disorders–inpatient; SUD-O = substance use disorders–outpatient; US = usual services; W = weeks; WLC = wait-list control; y = year.
Family approaches
Family therapy aims to create more adaptive patterns of family interaction, reducing problem behaviors thought to be a result of family dysfunction (Waldron, 1997). Multiple family therapies have been developed for the treatment of youth SUDs, aggression, and delinquency. Those with the most empirical support include Multisystemic Therapy (MST; Henggeler et al., 1991), Functional Family Therapy (FFT; Alexander & Parsons, 1973; Gordon, Graves, & Arbuthnot, 1995), Brief Strategic Family Therapy (BSFT; Szapocznik, Kurtines, Santisteban, & Rio, 1990), and Multidimensional Family Therapy (MDFT; Liddle, 2010). Evidence indicates that family therapy is more effective than individual therapy for youth SUDs and other problem behaviors (Diamond & Josephson, 2005; Slesnick & Prestopnik, 2005).
MST is an intensive, in-home intervention designed as an alternative to secure placement for youth with severe social, emotional, or behavioral problems. Therapists are available to families at all times, and they work with caregivers to adjust the ecological context of problem behaviors (Schoenwald, Ward, Henggeler, Pickrel, & Patel, 1996). A strong evidence base shows that MST is effective for both SUDs and delinquent, aggressive behavior (Henggeler, Melton, & Smith, 1992; Schaeffer & Borduin, 2005). It produces better family, SUD, and criminal justice outcomes than usual services or individual therapy (Henggeler, Clingempeel, Brondino, & Pickrel, 2002; Henggeler, Melton, Brondino, Scherer, & Hanley, 1997; Letourneau et al., 2009) and is more effective and cost-effective than hospitalization or incarceration (Henggeler et al., 1992; Sheidow et al., 2004).
FFT combines the family systems and social learning approaches (Gordon et al., 1995), focusing on increasing engagement and motivation for change before using behavioral techniques to develop healthier interaction patterns (Flicker, Turner, Waldron, Brody, & Ozechowski, 2008). FFT increases family engagement in treatment (Friedman, 1989) and is effective for adolescent SUDs (Lewis, Piercy, Sprenkle, & Trepper, 1990; Stanton & Shadish, 1997) and for delinquency among adolescents with SUDs (Chamberlain & Reid, 1998). FFT is also superior to usual services in preventing criminal activity in adolescents and young adults (Barton, Alexander, Waldron, Turner, & Warburton, 1985; Gordon et al., 1995; Hansson, 1998). Like MST, FFT has been shown to be more cost-effective than secure placement (Aos & Barnoski, 1998; Sexton & Alexander, 2000).
BSFT is a structured, problem-focused therapy lasting 12 to 16 sessions over 3 to 4 months. Therapists observe family interactions and diagnose strengths and weaknesses, with a focus on issues linked to youth problem behaviors (McClelland et al., 2004; Szapocznik, Hervis, & Schwartz, 2003). BSFT techniques include cognitive restructuring and conflict resolution, behavior management, and parenting skills training (Szapocznik, Hervis et al., 2003). BSFT has been shown to reduce substance use and other problem behaviors and to improve family functioning (Santisteban et al., 2003; Szapocznik, Amaro et al., 2003; Szapocznik, Kurtines, Foote, Perez-Vidal, & Hervis, 1983).
MDFT is a manualized, short-term approach that focuses on risk and protective behaviors for adolescent and family functioning (Liddle, 2010). Therapists conduct interventions with youth and caregivers individually and jointly; in joint sessions, family members try new ways of problem solving with active therapist guidance (Liddle, 1999). MDFT has strong empirical support as an effective treatment for adolescent SUDs, outperforming group, residential, and cognitive behavioral comparisons (Liddle & Dakof, 2002; Liddle et al., 2001; Liddle, Dakof, Turner, Henderson, & Greenbaum, 2008; Liddle, Rowe, Dakof, Henderson, & Greenbaum, 2009; Liddle, Rowe, Dakof, Ungaro, & Henderson, 2004). MDFT has recently been adapted for institutional settings as MDFT-Detention to Community (MDFT-DTC; Liddle, Dakof, Henderson, & Rowe, in press), which includes treatment pre- and postrelease. MDFT-DTC has been shown to be more effective than juvenile detention (Liddle et al., in press) and day treatment (Liddle et al., 2006) in terms of treatment participation, patient satisfaction, substance use, and delinquency.
Family therapies are typically designed to be delivered in the home and/or community (Henggeler, 2011; Sexton & Alexander, 2005). Evidence of equal or superior outcomes and superior cost-effectiveness indicate that they should be considered as alternatives to institutional placement and individual therapy when possible. Unlike individual interventions, family therapies are explicitly designed to address risk factors at multiple levels that contribute to substance abuse, which may explain why they have demonstrated better outcomes. Some components of these interventions (e.g., therapist availability, frequent family sessions) may be difficult to fully adapt to secure settings. However, other components (e.g., engaging family in treatment, working to reduce family dysfunction) may be useful approaches for clinicians in such settings. Additionally, some family-based interventions (e.g., MDFT) have been effective when specifically adapted for secure settings.
Cognitive-behavioral therapy (CBT)
CBT for SUDs (Wright, Beck, Newman, & Liese, 1993) is based on the idea that problem behaviors result from maladaptive cognitions (Winters, 1999). Youth who use drugs may have distorted expectancies about the positive effects of drugs and may not consider the negative consequences. Failure to develop problem-solving, social, and self-control skills or adaptive strategies for coping with peer pressure and negative emotions are also thought to contribute to substance use, aggression, and delinquency (Kaminer, Burleson, Blitz, Sussman, & Rounsaville, 1998). CBT aims to correct maladaptive beliefs by examining their rational basis and substituting beliefs that are consistent with adaptive behavior (Winters, 1999). Meta-analytic studies and reviews have shown CBT to effective for SUDs (Dennis et al., 2004a; Dennis et al., 2004b) and delinquency (Lipsey, Landenberger, & Wilson, 2007) in male and female adolescents. The inclusion of anger management and problem-solving components appears to be particularly important in terms of minimizing aggression (Lipsey et al., 2007).
CBT may be the best studied approach with adolescents in secure settings (Bray, 2000). Evidence indicates that the brief (4-10 session) CBT interventions that focus on interpersonal, behavioral, and anger management skills reduce recidivism and substance use among institutionalized and noninstitutionalized youth offenders (Goldstein, Glick, & Gibbs, 1998; Lipsey, Wilson, & Cothern, 2000). CBT is also well-suited for use with youth in secure settings, due to its brevity, structure, and focus on symptom-reduction rather than insight. Additionally, while it is directive, CBT is also collaborative, which may increase its attractiveness and accessibility to adolescents in residential institutions who may experience little control over their lives and environments (Desai et al., 2006).
While CBT addresses certain specific individual risk factors for aggression and SUDs that family therapies do not, it is limited in that it fails to address common risk factors in family and other systems. Consequently, some researchers have combined CBT with family therapy to address youth SUDs. Integrated Family and Cognitive-Behavioral Therapy (IFCBT) is an intensive program that includes one family and two peer group sessions each week (Latimer, Winters, D’Zurilla, & Nichols, 2003). Initial research suggests that IFCBT may be effective in reducing substance abuse as well as aggression and delinquency (Latimer et al., 2003). Another study (Waldron, Slesnick, Brody, Turner, & Peterson, 2001) compared marijuana outcomes in adolescents randomly assigned to either CBT, FFT, CBT+FFT, or a drug education group. Significant reductions in marijuana use were observed across groups, but the FFT, CBT+FFT, and drug education groups used marijuana less frequently than the CBT group. These studies suggest that individual and family therapy approaches can be usefully combined to treat adolescents with substance abuse and other delinquency problems.
Motivational enhancement therapy (MET)
MET is a brief therapy based on motivational interviewing (W. R. Miller & Rollnick, 1991). MET interventions focus on increasing motivation to change by addressing ambivalence. Therapists empathize with clients and collaboratively assist clients in developing discrepancy between their current behaviors and their goals (Feldstein & Ginsburg, 2006). MET has been shown to be an effective treatment for SUDs (Dunn, Deroo, & Rivara, 2001). For example, brief emergency department interventions with adolescents admitted for alcohol-related injuries have been associated with reduced drinking, alcohol-related problems, and risky behavior (Barnett, Monti, & Wood, 2001; Monti et al., 1999). It has also been shown to improve engagement in substance abuse treatment in incarcerated adolescents (Stein, Monti et al., 2006). Although further research is needed, from a theoretical perspective MET is likely to be useful in secure settings, either as a stand-alone treatment or as an adjunct to other approaches (Feldstein & Ginsburg, 2006). It is brief and has demonstrated effectiveness with clients with high levels of anger or hostility (Waldron et al., 2001). Additionally, the collaborative nature of MET may be particularly useful in working with incarcerated adolescents who are resistant to authority (Marlatt & Witkiewitz, 2002).
Contingency management (CM) and community reinforcement (CR)
CM interventions are based on operant conditioning principles, in which behavior is a function of its consequences (Higgins, 1997; Higgins & Silverman, 2008). CM requires that nonabstinence be readily detectible (Higgins, Budney, & Bickel, 1994). Rewards are given for verified abstinence and other targeted behaviors (e.g., nonaggression), but withheld for nonabstinence (Higgins, Alessi, & Dantona, 2002). Abstinent clients may receive vouchers with monetary values that increase with longer abstinence (Higgins et al., 1991) or may draw slips of paper from a bowl which may contain either written reinforcement or a voucher (Petry & Martin, 2002). CM is well-established as an effective treatment for adult SUDs (Higgins & Silverman, 2008; Prendergast, Podus, Finney, Greenwell, & Roll, 2006), and several recent studies have assessed its use for youth SUDs and conduct problems. For example, adolescent outpatient interventions combining CM with CBT and/or motivational enhancement have been reported to reduce marijuana use and externalizing behaviors (Carroll et al., 2006; Kamon, Budney, & Stanger, 2005) and cigarette smoking (Krishnan-Sarin et al., 2006).
Like CM, CR interventions aim to rearrange environmental contingencies so that refraining from problem behavior is more rewarding than engaging in it (S. H. Godley et al., 2001). CR focuses on the interaction between youth and others in their environment, teaching them how to build a positive support system and additional reinforcers, how to replace maladaptive behaviors with newly developed prosocial alternatives, and how to utilize community resources (S. H. Godley et al., 2001). Initial studies suggest that CR for adolescents is an effective community SUD treatment (Dennis et al., 2004a).
Although additional research is needed, studies suggest that MET, CM, and CRA may be effective approaches for adolescent SUDs. Like CBT, they are individual approaches that do not systematically address common risk factors outside the individual and are not preferred over family approaches when either is feasible. However, individual approaches have the advantage of being better-suited to secure environments such as juvenile detention and residential treatment facilities. For example, youth who maintain abstinence or refrain from aggressive behavior can be rewarded with additional visits or other privileges. Similarly, training adolescents to develop and use strategies for more rewarding social interactions in a controlled environment may be a step toward generalizing such behaviors to a less controlled environment following discharge.
Pharmacotherapy
Most SUD treatment studies in the literature involve psychosocial interventions, due to concerns about the efficacy and safety of psychotropic medications in youth and to legal restrictions on use in youth. However, an increasing number of youth are being treated with pharmacotherapy for SUDs (D. Clark, Wood, Cornelius, Bukstein, & Martin, 2003), and a few studies have examined the efficacy of specific agents. Some evidence supports the use of naltrexone for adolescent alcohol dependence (Deas, May, Randall, Johnson, & Anton, 2005), and nicotine replacement therapy and bupropion for adolescent tobacco dependence (Upadhyaya, Deas, & Brady, 2005). Some studies also suggest that opiate replacement may be useful for older adolescents with highly treatment-refractory SUDs (Ebner, Schreiber, & Zierer, 2004; Woody et al., 2008). Overall, more systematic, controlled research is needed to determine the utility of pharmacotherapy for youth SUDs, and considerable caution is warranted in the use of pharmacologic agents approved for adults in younger populations (Toumbourou et al., 2007).
Interestingly, pharmacotherapy for comorbid disorders may also reduce substance use (Waxmonsky & Wilens, 2005). These findings are important in light of research suggesting that more than half of youth with SUDs have comorbid disorders (Armstrong & Costello, 2002) and that those with comorbid disorders tend to have more severe refractory SUDs (Rowe, Liddle, Greenbaum, & Henderson, 2004). The combination of SUDs with internalizing or externalizing psychopathology has been associated with worse outcomes in terms of aggression and physical and mental health (Clingempeel, Britt, & Henggeler, 2008). Consequently, these findings suggest that effective pharmacologic or psychosocial treatments for psychiatric disorders may be an effective way to reduce later substance use and aggression in adolescents.
Summary
Substance involvement is a critical factor in the assessment and treatment of youth aggression and delinquency. Evidence suggests that shared underlying mechanisms contribute to both behaviors and that each reciprocally influences the other. In light of the strong relationship between the two, it is important that SUD status be considered during assessment and treatment of youth aggression and delinquency. Across settings, it is recommended that clinicians use brief, evidence-based tools to screen for aggression- and substance-related problems. Objective measures (e.g., urinalysis) should also be used to screen for substance use. Positive screening results should be followed by comprehensive, evidence-based assessment to more thoroughly identify problem behaviors and intervention targets (Chassin, 2008; Grisso & Underwood, 2004).
Despite high levels of need, adolescents with aggression and comorbid substance use tend to be underserved. Studies demonstrating reduced aggression and criminality associated with SUD treatment (Bernberg & Thorlindsson, 1999; Rossow & Wichstrom, 1999) suggest a need for increased integration and emphasis in this area. Several interventions have been developed that are effective in reducing aggressive, delinquent behaviors and substance use. Treatment is most likely to be effective when it addresses dysfunction and risk factors across multiple domains (e.g., individual, family, school, peer systems).
Family-based therapies focus on reducing dysfunction in family and other systems that is thought to influence problem behaviors. The literature demonstrates that family therapies are superior to other modalities (Diamond & Josephson, 2005) and should be considered first-line treatments for youth SUDs and aggression. However, these interventions were designed for use in the community, and with the exception of MDFT have not been adapted for use in secure or residential settings. Some elements of these approaches may be practical for use in secure settings. Additionally, further research is needed to adapt family-based interventions to these settings. While less effective than family therapy, other treatment approaches have demonstrated effectiveness with institutionalized youth, including CBT, CM, CRA, and MET. Because these interventions can be conducted one-on-one, they are practical for use in secure settings. These approaches may also have more appeal for adolescents, as they allow a measure of control over the environment. This may be particularly attractive for incarcerated adolescents, who may be more resistant to authority and whose environment is subject to stringent controls.
Footnotes
This article is part of a special issue titled “Treatment Considerations for Aggressive Adolescents in Secure Settings,” edited by Calvin M. Langton of the University of Toronto, Canada.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
