- Email: [email protected]
Psychiatric Comorbidity and the 16‐Month Trajectory of Substance‐Abusing and Substance‐Dependent Juvenile Offenders
Psychiatric Comorbidity and the 16-Month Trajectory of Substance-Abusing and Substance-Dependent Juvenile Offenders JEFF RANDALL, PH.D., SCOTT W. HENGGELER, PH.D., SUSAN G. PICKREL, M.P.H., M.D., AND
MICHAEL J. BRONDINO, PH.D.
ABSTRACT
Objectives: To examine the concurrent correlates of internalizing and externaliz ing disorders among substance-abusing and substance-dependent juvenile offenders and to determine the association between psychiatric comorb idity and psychosocia l functioning of the youths 16 months later. Method: Participants were 118 juvenile offenders meeting DSM-III-R criteria for substance abuse or dependence and their families. A multisource measurement battery was used to assess drug use, criminal activity, family relations , peer relations, school functioning , and out-of-home placements. Results: Comorbidity for externalizing disorders was associated with high rates of antisocial behavior and predicted worse 16month outcomes than substance abuse alone or substance abuse with comorb id internaliZing disorders . For criminal activity and drug use, the presence of internalizing disorders buffered the deleterious effect of externalizing disorders on substance-abusing and substance-dependent juvenile offenders. Conclusions: Even in substance-abusing delinquents, a population already extreme in antisocial behavior, the presence of externalizing disorders indicates high risk for deterioration. J. Am. Acad. Child Ado/esc. Psychiatry, 1999, 38(9) :1118-1124. Key Words: adolescents, substance abuse, comorbidity, delinquency.
In recent years, reviewershave concluded that substanceabusing youths evidence high rates of comorbid psychiatric problems such as depression and conduct disorder (e.g., Lewinsohn et al., 1995; Weinberg et al., 1998) and that substance-abusing delinquents are at especially high risk of co-occurring mental health disorders (Thompson et al., 1996; White and Labouvie, 1994). To a large extent, the purpose of documenting such comorbidity is to raise issues concerning the influence of co-occurring mental health disorders on th e developmental trajectories of substance-abusing and substance-dependent youths as well as regarding the design of intervention models for such youths.
Accepted March 31, 1999. From theMedical University of South Carolina, Charleston. Thisstudywassupported by N IDA grantR01DA lO0079 (ScottW Henggeler; Principal Investigator). Reprint requests toDr.Randall Family ServicesResearch Center, Department of Psychiatryand BehavioralSciences, Medical University ofSouthCarolina, 67 PresidentStreet, P.O. Box250861, Charleston, SC29425; e-mail: [email protected]. 0890-8567/99/3809-1118©1999 by the American Academy of Child and Ado lescent Psychiatry.
1118
In this regard, adolescent substance abuse with comorbid internalizing or externalizing disorders seems to impa ct adolescents differently than substance abuse alone. For example, adolescent substance abuse with comorbid externalizing disorders, such as conduct disorder, predicted high school dropout (Kessler et al., 1995) and failure to complete treatment on an inpatient unit (Kaminer et al., 1992). In contrast, adolescent substance abuse with comorbid internalizing disorders (e.g., affective disorders) predicted completion of treatment on an inpatient unit (Kaminer et al., 1992). Thus, emerging evidence suggests that types of comorbidity may be linked with differential longer-term outcomes for youths with serious substance abuse problem s. The purpose of this study is to expand on these findings in 2 ways. First, we examine the different concurrent correlates of internalizing and externalizing como rbid conditions in a sampl e of substance-abusing and substance-dependent juvenile offenders.The examined correlates pertain to those domains of functioning that have been consistently linked with substance use and criminal activity by adolescents (Henggeler, 1997; Loeber and Hay,
j . AM. ACAD . C H I LD ADO LESC . PSYCHIATRY, 38:9 , SE PTEMB ER 1999
COMORBIDITY AND FUNCTIONING IN OFFENDERS
1997): family functioning, peer relations, and school functioning. The second purpose is to examine the association between psychiatric comorbidity and the psychosocial functioning of the youths 16 months later. Here, we evaluated whether internalizing and externalizing comorbid conditions predicted improvements or declines in the personal and social functioning of the adolescents. METHOD Subjects One hundred eighteen adolescents served as participants and were recruited from the D epartment ofJuvenile Justice (DJJ) of Charleston, South Carolina, during October 1992 to June 1994. To identify juvenile offenders who met diagnostic cri teria for su bstance abuse or dependence, participants were administered the Structured Clinical Interview for DSM-III-R (Spitzer et al., 1 988), which was supplemented with information from the Personal Experiences Inventory (PEl) (Winters and Henly, 1989). Inclusion criteria included the following: (1) DSM~III-R diagnosis of psychoactive substance ab use or dependence (American Psychiatric Association, 1987), (2) age 12 to 17 years, (3) formal or inform al probationary statu s, (4) Charleston County Resident, and (5) at least one parental figure residing with the adolescent . To enhance external validity, adolescents with preexisting mental health , physical health , or intellectual difficulties were included in the study. We excluded adolescents who were formall y involved in substan ce abu se treatment or who had a sibling who was a stu dy participant. Four hundred twenty-three adolescents were screened, and 140 met the inclusionary criteria. One hundred eighteen (84%) adolescents and fam ilies agreed to participate and completed initial assessments. Families that consented to parti cipate did not differ from refusers on key variables including adolescent age, gender, race, self-reported drug use, or arrest history. Research attrition from initial assessments to follow-up assessments 16 months later was modest. One hundred eight (91.5%) adolescents and families completed follow-up assessments 16 months later. Adolescent s and families who completed follow-up assessments did not differ from noncompleters on key demographic, drug use, and arrest history variables. The low research attrition enhances th e internal and external validity of th e findings .
Design and Procedures An assessment battery was administered to adolescents and primary caregivers in the ir homes and was readministered 16 months later. Research procedures were explained to adolescents and primary caregivers, and caregivers provided informed consent and adolescents provided informed assent . If adolescents or caregivers lacked sufficient reading skills, questions from the assessm ent battery were read to them. Each family was reimbursed $75 per assessment for their time .
Measures A mulrimethod (i.e., adolescents ' self-report, primary caregivers' reports, and archival records ) assessment battery was used to tap key correlates of antisocial behavior as well as substance use and criminal activity. Psychiatric Comorbidity. All participants were ad m inistered the Diagnostic Interview Schedule for Children Version 2.3-child repo rt (DISC) (Shaffer, 1992) to generate DSM-III-R diagnoses. For the
purposes of analyses, participants were considered to have a cornorbid externalizing disorder if they had at least one externalizing diagnosis (i.e., conduct disorder, oppositional defiant disorder, or attentiondeficit hyperactivity disorder) and a comorbid int ern alizing disorder if they had at least one internalizing diagnosis (i.e., major depression , dysthymia, overanxious disorder, agoraphobia, social phobia, simple phobia, separation anxiety, panic disorder, obsessive-compulsive disorder, avoidant disorder, or generalized anxiety disorder) . It should be noted that the child self-report version of the DISC may result in the underreporting of symptoms of attention-deficit hyperactivity disorder (Schwab-Sto ne et al.,1996). Criminal Activity and Drug Use. The Self-Report Delinquency Scale (Elliott et al., 1983) is a 53- item inventory used to assess adolescents' criminal behavior. A summary score, the General Delinquency Index, was used to indicate adolescents' total number of criminal acts for the previo us 6 months. Arrest and offense hist ories were extracted from a data base maintained by the DJ]. Severity of offenses was coded using a Seriousness Inde x developed by H anson et al. (1984). The Seriousn ess Index is a 17-point scale with nu merical values for seriousness (e.g., 1 = truancy, 4 = disorderly conduct, 8 = assault/battery, 11 = grand larceny, 13 = un armed robbery, 17 = murder) . The PEl (Winters and Henly, 1989 ) was used to assess the frequ en cy of adolescent substance use for the previous 3 months. Twelve items from the PEl were used to construct 2 scales that examined alcohol/marijuana use and other illicit drug use. The frequ ency of alcohol and marijuana use was summed and truncated at 100. An illicit drug use scale elicited the frequency of nonprescribed use of cocaine , amphetamines, Quaaludes, barbiturates, LSD , other psychedelics, tranquilizers, heroin, other narcotics, and inhalants. Because of the low prevalence of reported illicit drug use, adolescents were categorized as either using any illicit drug or using no illicit drugs during the previous 3 months. FamilyRelations. The Family Adaptab ility and Co hesion Evaluation Scales-III (O lson er al., 1985) is a 20 -item inventory with Likert-type item s ranging from 1 (alm ost never) to 5 (almost always). The inventory was completed by the adolescent and primary caregiver to assess family adaptability and cohesion . Consistent with the recommendations of Henggeler et al. (1991), cohesion and adaptability were treated as linear scores in analyses. The Issues C hecklist (Robin er al., 1977) is a 44 Ci tem inventory on which adoles cents and caregivers identify topics th at are potential sources of disagreem ent and con flict. Represent ative topics incl ude negotiating bed tim e, receiving low grades, spending free time, and talking back to parents. For each conflict topic identified, adolescents and caregivers rated the intensity of th e conflict on a scale rang ing from 1 (calm) to 5 (angry) and estimated how ofren the topic was discussed during the previous m onth. Peer Relations. The Peer Conform ity Inventory (Bern dt, 1979) measures adole scents' proso cial and antisocial peer conformity and con sists of 30 situ ations (10 ant isocial, 10 pro social, and 10 neutral) in which peers suggested some action . The Revised Problem Behavior Checklist (Q uay and Peters on , 1987) is a 77 -item inventory with item scores ranging from 0 (no problem) to 2 (a severe problem). The inventory was completed by the adolescent 's primary caregiver, and the Socialized Aggression subscale was used as an index of association with deviant peers. School Functioning. The School Functioning subscale of the Social Competence scale of the Child Behavior Checklist (Achen bach, 1991) was completed by th e youth and caregiver. . Out-ofHome Placement. The Monthly Service Utilizat ion Survey (H enggeler er al., unpublished manuscript, 1992) is a semistructured int erview used to collect out-of-home placement data on adolescents .
j . AM . ACAD. C H ILD ADOLESC . PSYCHIATRY, 38 :9, SEPTEMBER 19 99
1119
RANDALL ET AL.
The majority of the interviews were conducted face to face. In the rare exceptions when face-to-face meetings were not possible, telephone interviews were conducted. Researchers interviewed primary caregivers at the end of each month to obtain their retrospective verbal reports of service utilization for that month. The survey was completed each month after the adolescent entered the study through the 16-month follow-up . Primary caregivers provided placement data, and correctional placements were confirmed th rough D]] computer records. O ut-of-home placements included any institution-based domicile, such as detent ion centers, jails, psychiatric or substance abuse hospitals, resident ial treatm ent cente rs, and group homes . Placements at a relative's home or at a surrogate parent's home were not considered our-of-home placements.
Data Analyses Correlation analyses, descriptive statistics, X2 tests, and analyses of variance were run on data collected at initial assessment, and multiple regression analyses were conducted on data collected 16 months later. For data collected at initial assessment, correlations were used to examine th e relation ship between measures of anti social behavior and social funct ioning and internalizing and externalizing comorbid conditions; descripti ve statistics, X2 tests, and analyses of variance were used to explore possible differenc es among comorbid conditions on demogr aphic variables and other sample characteristics. For longitudinal data l hierarchic al multiple regressions were conducted to examine the association between psychiatric comorbidity and the psychosocial functioning of the youths 16 months later.
RESULTS
Sample Characteristics
Demographic and juvenile justice information on the study participants is presented in Table 1. A X2 analysis comp aring the comorbid conditions on number of males only (n = 93) indicated a significant effect for number of males (X 2= 19.39; df= 3,114;p < .002), with a higher than
expected frequency of males with neither cornorbid internalizing nor comorbid externalizing diagnoses (observed 36 versus expected 23) and a lower than expected frequency of males with both comorbid internalizing and comorbid externalizing disorders (observed 7 versus expected 23). A X2 analysis comparing the comorbid conditions on number of females only (n = 25) indicated a significant effect for number of females (X 2 = 10.04; df= 3,114; p < .0182). However, for females there was a lower than expected frequency of females with neither comorbid internalizing nor comorbid externalizing diagnoses (observed 1 versus expected 6) and a higher than expected frequency of females with both comorbid internalizing and comorbid externalizing disorders (observed 11 versus expected 6). Results of all other X2 tests and analyses of variance on demographic variables were not significant. Fifty-six percent of participants met criteria for substance abuse and 44% for substance dependence . In addition to meeting cri teria for substance abuse or dependence, 72% of the sample met DSM~IlI-R criteria for one or more psychiatric diagnoses. Information on individual diagnoses for males and females separately and for male s and femal es com bined is presented in Table 2. As shown in Table 1 at initial assessment, 32 participants were classified as having a comorbid externalizing disorder only, 31 as having a comorbid internalizing disorder only, 18 as having both comorbid externalizing and comorbid int ernalizing disorders, and 37 as having neither comorbid externalizing nor comorbid internalizing diagnoses.
TABLE 1 Demographic and Juvenile Justice Characteristics at Initial Assessment Variable D emographics Age (yr) Race (%) African-American White Other Male (%) Female (%) Medium income (thousands) Two-parent household (%) Juvenile justice Prior incarceration (%) Age of first arrest (yr) Mean no. of arrests Mean % of violent offenses Mean severity of offenses
1120
Not Corn orbid (n = 37)
External (n = 32)
Internal (n = 3 1)
Both (n = 18)
Total (N = 118)
15.1
15.2
15.3
15.3
15.2
18.6 11.9 0.8 30.5 0.8 15- 20 16.1
9.3 16.9 0.8 23.7 25-30 19.5
15.2 10.2 0.8 18.6 7.6 10-15 8.5
6.8 8.5 0 5.9 9.3 10-15 5.9
50 47.5 2.5 79 21 15- 20 50
51 14.0 1.84 30 9.94
62 13.9 2.16 16 9.53
35 14.0 1.90 13 9.39
38 14.5 1.67 11 7.28
46.5 14.1 1.89 17.5 9.04
3.4
J. AM. ACAD . CHI LI) AD OL ESC. PSYCHIAT RY. 38: 9. SEP TE MBER 1999
COMORBIDITY AND FUNCTIONING IN OFFENDERS
Correlational Analyses
TABLE 2 Diagnoses at Initial Assessment Diagnosis Conduct disorder Oppositional defiant disorder ADHD Major depression Dysthymia Overanxious disorder Agoraphobia Social phobia Simple phobia Separation anxiety Panie disorder Panie disorder with agoraphobia Obsessive-compulsive disorder Avoidant disorder Mania Generalized anxiety disorder Anorexia Bulimia Vocal or motor tics Transient tic disorder Tourette's disorder Diurnal enuresis Nocturnal enuresis Encopresis
Male
Female
(n = 93)
(n = 25)
27 7
8 5 2 5 2 6 7 10 9 5 1 7 5
2
4 1
4 3 9
7 2 0 3 6 3 3 1 0 0 0 1 0 1 0 0
Note: Values are percentages . ADHD activity disorder.
Both (N = 118)
35 12
4
4
9 3 10 10 19 16 7 1 10 11 7
1
4
4
5 1 2 3 2 1 2 1 0
1 2 3 1 1 1 1 0
= attention-deficit hyper-
The initial assessment data shown in Table 1 also show that the participants represented a sample that was extreme in antisocial behavior. Results of analyses of variance comparing the 4 comorbid conditions on juvenile justice involvement were not significant.
A correlation analysis was conducted to determine the association between comorbidity for externalizing and for internalizing disorders. The association was not significant (r= -O.lO,p < .30) . Additional correlations were conducted to examine the concurrent associations between psychiatric comorbidity (externalizing disorder and internalizing disorder) and the behavioral and social functioning of the adolescents. Externalizing Correlates. As shown in Table 3, comorbidity for an externalizing diagnosis was consistently associated with increased rates of antisocial behaviors and poorer functioning in family, peer, and school contexts. Specifically, comorbidity for an externalizing disorder was associated with higher rates of criminal offending, alcohol and marijuana use, and other illicit drug use. Similarly, comorbidity for an externalizing disorder was associated with increased association with deviant peers, greater conformity to antisocial peer behavior, lower family cohesion, and poorer school functioning. Internalizing Correlates. Comorbidity for an internalizing disorder was associated with both more favorable and lessfavorableaspects of functioning. On the favorable side, internalizing comorbidity was linked with greater conformity to prasocial peer behavior and increased family structure. Internalizing diagnoses were not linked with higher rates of criminal activity or substance use. On the other hand, comorbidity for internalizing disorders was linked with greater association with deviant peers and more family conflict.
TABLE 3 Significant Univariate Associations Between Comorbidity and Concurrent Antisocial Behavior and Social Relations Comorbidity Measures
Respondent
Externalizing .
Internalizing
SRD General Delinquency Alcohol/Marijuana Illicit Drug Use RPBC Deviant Peer Association Peer Antisocial Conformity Peer Prosocial Conformity FACES-III Cohesion FACES-III Adaptability Family Conflicr-Quantiry Family Conflict-Intensity CBCL School Competence CBCL School Competence
Adolescent Adolescent Adolescent Caregiver Adolescent Adolescent Adolescent Caregiver Caregiver Caregiver Adolescent Caregiver
0.39** 0.22* 0.18* 0.16 0.25** -0.07 -0.21 * -0.07 -0.11 -0.15 -0.18* -0.19*
-0.01 0.06 -0.07 0.19* 0.09 0.18* -0.07 -0.21* 0.18* 0.19* 0.08 0.15
Note: SRD = Self-Report Delinquency Scale; RPBC = Revised Problem Behavior Checklist; FACES-III bility and Cohesion Evaluation Scales-III; CBCL = Child Behavior Checklist.
= Family Adapta-
* p < .05; **P < .01.
J. AM.
ACAD. CHILD ADOLESC . PSYCHIATRY, 38:9, SEPTEMBER 1999
1121
RANDALL ET AL.
TABLE 4 Standardized Regression Coefficients and Significance Levels From rhe Significant Multiple Regression Analyses Relating Comorbidity to 16-Month Outcomes Comorbidity Measures
Respondent
External
Internal
Interact
SRD General Delinquency Illicit Drug Use Days in Out-of-Home Placement FACES-III Cohesion CBCL School Comperence
Adolescent Adolescent Caregiver Adolescent Caregiver
0.38** 0.38** 0.26* 0.31* - 0.15
0.2 5* 0.11 -0. 13 0.08 -0.23*
-0.29* -0.33* -0.21 -0.20 0.03
Note: SRD = Self-Report Delinqu ency Scale; FACES-III Child Behavior Checkl ist * p < .05; **P < .01.
= Family Adaptability and Coh esion Evaluation Scales-III; CBCL =
Longitudinal Analyses
The purpose of these anal yses was to determine whether psychiatric comorbidity was differentially associated with outcomes at 16 months after the initial assessment. Hence, hierarchical regression analyses were conducted in which the measures of antisocial behavior (criminal activity, substance use) and social functioning (family, peer, school) served as dependent measures. The first variable entered into the equation was the corresponding baseline variable, dummy variables indicating the presence or absence of internalizing disorders and externalizing disorders were entered next, and the interaction of externalizing and internalizing disorders was entered last. A similar analysis was conducted on the out-of-home placement measure, except a baseline measure of such was not included. Externalizing Disorders. As shown in Table 4, externalizing comorbidity predicted increased rates of general delinquency and illicit drug use 16 months later. These findings, however, were qualified by significant interactions between externalizing and internalizing disorders. For criminal activity, an examination of comorbid condition means suggested that the presence of internalizing disorders buffered the effect of externalizing disorders on sub stance-abusing and substance-dependent juvenile offenders. As indicated in Table 5, participants with both
comorbid internalizing and comorbid externalizing disorders engaged in a lower level of general delinquency than participants with only a comorbid externalizing disorder.The same pattern of interaction effectswas observed for illicit drug use, with a lower percentage of participants with both comorbid internalizing and comorbid externalizing disorders engaging in a lower level of illicit drug use than participants with only a comorbid externalizing disorder. In addition, comorbidity for externalizing disorder predicted high rates of out-of-home placement and increased family cohesion. Internalizing Disorders. As indicated previously, a significant interaction effect eme rged for gen eral delinquency in which youths with internalizing disorders evidenced less increase in criminal activity than youths with comorbid externalizing disorders only. In add ition , internalizing disorder was associated with decreased school functioning (based on caregiver reports) . Number of DISC Diagnoses. To determine whether changes on dependent measures at 16 months were not merely associated with the number of diagno ses at the time of participant recruitment, hierarchical regression analyses were conducted. The first variable entered into the equation was the corresponding baseline dependent measure followed by the number of DISC diagnoses. None of the results were significant. Hence, in general,
TABLES Means of Significant Interacti on Effects From th e Regression Analyses Relating Comorbidity to 16-Month Outcomes Measures SRD Gen eral Delinquency Mean Illicit D rug Use N ote: SRD
1122
Not Comorbid (n = 37)
External (n = 32)
14.2 1.62
60.3 6.3 1
Internal = 3 1)
Both (N = 118)
25.8 3.30
49.9 2.77
(n
=Self-Report Delinqu ency Scale. ] . AM. ACAD . C HI LD AD OLE SC. PSYCHIAT RY, 38:9, SEPT EMB ER 1999
CO MO RBID IT Y AN D FUNCTIONING IN OFFENDERS
the key determinant of deteriorating functioning was comorbidity for externalizing disorder rather than the number of diagnoses per se. DISCUSSION
Youthswith substance abuse and delinquency represent a population that is extreme in antisocial behavior.Yetthe findings show that substance-abusing delinquents with externalizing disorders are even more antisocial than their substance-abusing delinquent counterparts. They engage in higher rates of crime, more alcohol and marijuana use, and more illicit drug use. In addition, substance-abusing delinquents with comorbid externalizing disorders evidence relatively less family cohesion, greater conformity to antisocial peer pressure, and decreased school competence. The longitudinal analysesprovide further evidence of the increased risk posed by a comorbid externalizing disorder. Sixteen months past the initial assessment, such cornorbidity was associated with higher rates of out-ofhome placement and increased general delinquency and illicit drug use. Delinquency and illicit drug use findings were qualified by significant interactions between internalizing and externalizing disorders. An examination of the comorbid condition means suggested a buffering effect. Substance-abusing and substance-dependent juvenile offenders with both comorbid internalizing and comorbid externalizing disorders engaged in lower levels of general delinquency and illicit drug use than substanceabusing and substance-dependent juvenile offenders with only comorbid externalizing disorders. Thus, the comorbidity of externalizing disorders only seems to add a significant risk factor to a population that already has a poor prognosis. In contrast, the presence of internalizing comorbidity for this sample of substance-abusing and substancedependent juvenile offenders suggests somewhat more favorable characteristics than their substance-abusing delinquent counterparts. Concurrent associations showed that youths with comorbid internalizing disorders had greater conformity to prosocial peer behavior and increased family structure. On the other hand, cornorbidity for internalizing disorders was linked with increased association with deviant peers and increased family conflict. These family-related findings (high conflict and increased structure) might reflect functional caregiver responses to the youths' antisocial behavior and association with deviant peers.That is, in light of the serious problems presented by these youths, increased structure and conflict
are expected if the caregivers are invested in the youth's welfare. With regard to the effects of internalizing disorder on the 16-month trajectories of these youths, findings showed that internalizing comorbidity predicted decreased school competence. Overall, comorbidity predicted worse outcomes than substance abuse alone. However, substance abuse with comorbid externalizing disorders onl y predicted even poorer outcomes than substance abuse with comorbid internalizing disorders. These findings are consistent with previous investigations (Kaminer et al., 1992; Kessler et aI., 1995) suggesting that substance-abusing adolescents with comorbid internalizing disorders may present a different clinical profile than substance-abusing adolescents with comorbid externalizing disorders. One possible explanation for these differences is the role of substance abuse in the adolescents' lives. For substanceabusing adolescents with externalizing disorders, substance abuse might be part of a larger picture of problem behavior (lessor, 1987). However, for substance-abusing adolescents with internalizing disorders, substance abuse may represent a way in which these adolescents selfmedicate to attenuate their affective difficulties. Clinical Implications
Clinical differences between adolescent substance abusers with internalizing or externalizing disorders suggest different treatment emphases. For substance-abusing adolescents with externalizing disorders, treatment should probably focus on enhancing parental competence to structure, monitor, supervise, and provide consequences for their adolescent's behavior. In this regard, the key is to identify those aspects of the family and its social ecology that are barriers to effective parental discipline (e.g., parental drug abuse, low social support, lack of skills) and to intervene accordingly (Henggeler et al., 1998). For substance-abusing adolescents with internalizing disorders, treatment might devote increased attention to the development of coping strategies for anxieties and stress. In light of the link between adolescent substance abuse with comorbid internalizing problems such as depression and suicide (Carlson et al., 1991), safety is also an important issue in treating substance-abusing adolescents with comorbid internalizing disorders. Limitations
Three central limitations of the study should be noted. First, the findings pertain only to the)ldolescent self-report
] . AM . ACAD. CHILD ADOLESC . PSYCHIATRY, 38:9, SEPTEMBE R 1999
1123
RANDALL ET AL.
version of the DISC, and, as noted previously, certain youths underreport certain symptoms (Schwab-Stone et al., 1996). This may account for the low prevalence of attention-deficit hyperactivity disorder (i.e., 4%) reported by the sample. Second, substance use was determined by self-report only, not laboratory measures or caretaker reports. Third, the sample sizewas not sufficient to permit subanalyses examining specific types of externalizing and internalizing disorders. Thus, for example, determining which internalizing disorders most strongly buffered the deleterious effects of externalizing comorbidity was not possible.
REFERENCES Achenbach TM (1991), Manualfor the Child Behavior Checklist and 1991 Profile. Burlington: University of Vermont Deparrment of Psychiatry American Psychiatric Association (1987), Diagnostic and Statistical Manualof MentalDisorders, 3rd edition-revised. Washington, DC: American Psychiatric Association Berndt TJ (1979), Developmental changes in conformiry to peers and parents. DevPsychoI15:608-616 Carlson GA, Rich CL, Grayson P, Fowler RC (1991), Secular trends in psychiatric diagnoses of suicide victims.] AffectDisord21:127-132 Elliott DS, Ageton SS, Huizinga D, Knowles BA, Canter RJ (1983), The Prevalence and Incidence of Delinquent Behavior: 1976-1980 (National Youth Survey Project Reporr No. 26). Boulder, CO: Behavioral Research Institute Hanson CL, Henggeler SW; Haefele WF, RodickJD (1984), Demographic, individual, and family relationship correlates of serious and repeated crime among adolescents and their siblings. ] Consult Clin Psychol 52:528-538 Henggeler SW (1997), The development of effective drug abuse services for youth. In: Treating Drug Abusers Effectively, Egerrson JA, Fox DM, Leshner AI, eds. Cambridge, MA: Blackwell, pp 253-279 Henggeler SW; Burr-Harris AW; Borduin CM, McCallum G (1991), Use of the Family Adaptability and Cohesion Evaluation Scales in child clinical research.] AbnormChildPsychoI19:53-63
Henggeler SW; Schoenwald SK, Borduin MC, Rowland MD, Cunningham PB (1998), Multisystemic Treatment of Antisocial Behavior in Children and Adolescents. New York: Guilford [essor R (1987), Problem-behavior theory, psychosocial development, and adolescent problem drinking. Br] Addict 82:331-342 (special issue: Psychology and Addiction) Kaminer Y, Tarter RE, Bukstein OG, Kabene M (1992), Comparison between treatment completers and noncompleters among dually diagnosed substance-abusing adolescents. ] Am Acad Child Adolesc Psychiatry 31: 1046-1049 Kessler nc, Foster CL, Saunder WB, Stang PE (1995), Social consequences of psychiatric disorders, I: educational attainment. Am] Psychiatry 152: 1026-1032 Lewinsohn PM, Gotlib IH, SeeleyJR (1995), Adolescent psychopathology, IV: specificityof psychosocial risk factors for depression and substance abuse in older adolescents.] AmAcad ChildAdolesc Psychiatry 34:1221-1229 Loeber R, Hay D (1997), Key issues in the development of aggression and violence from childhood to early adulthood. Annu RevPsychoI48:371-41O Olson DH, Portner J, Laves Y (1985), Contributions of family and peers to delinquency. Criminology 23:63-79 Quay He, Peterson DR (1987), Manualfor the Revised Problem Behavior Checklist. Coral Gables, FL: University of Miami Robin AL, Kent R, O'Leary KD, Foster S, Prinz R (1977), An approach to teaching parents and adolescents problem-solving communication skills.
Behav Ther8:639-643 Schwab-Stone ME, Shaffer D, Dulcan MK et al. (1996), Criterion validity of the NIMH Diagnostic Interview Schedule for Children Version 2.3 (DISC-2.3).] Am Acad ChildAdolesc Psychiatry 35:878-888 Shaffer D (1992), Diagnostic Interview Schedule for Children, Version 2.3. New York: Columbia University Division of Child Psychiatry Spitzer RL, Williams JBW; Gibbon M, First MB (1988), Structured Clinical Interview for DSM-III-R: Patient Version (SCED-P 6/1/88). New York: New York State Psychiatric Institute, Biometrics Research Deparrment Thompson L1; Riggs PD, Mikulich SK, CrowleyTJ (1996), Contribution of ADHD symptoms to substance problems and delinquency in conductdisordered adolescents.] Abnorm ChildPsychoI24:325-334 Weinberg NZ, Rahdert E, Colliver JD, Glantz MD (1998), Adolescent substance abuse: a review of the past 10 years.] Am Acad ChildAdolesc Psy-
chiatry 37:252-261 White HR, Labouvie EW (1994), Generality versus specificity of problem behavior: psychological and functional differences. ] DrugIss 42:55-74 Winters KC, Henly G (1989), The Personal Experiences Inventory. Los Angeles: Wesrern Psychological Services
Development of Feeding Practices During the First 5 Years of Life. Lawrence D. Hammer, MD, Susan Bryson, MS, W. Stewart Agras, MD
Objective: To understand the transition from breast- and bottle-feeding to solid-feeding and factors that might affect the duration of breast- and bottle-feeding. Design: Cohorr followed up from birth with relatively well-educated, middle-class parents. Setting: Communiry sample recruited from 3 suburban newborn nurseries (a teaching hospital, communiry hospital, and large health maintenance organization). Participants: One hundred ninety-one healthy full-term infants. Measures: Assessment of feeding practices through the ages of complete weaning from breast- and bottle-feeding. Results: More than 90% of participants breast-fed for at least 2 weeks. Infants of older mothers were weaned from the breast later than infants of younger mothers. First-born infants were weaned from the breast earlier than later-born infants. Eighry-four percent of infants bottle-fed at some time during the first year of life. More than 40% of the cohort was still receiving bottles at 24 months of age, 16% at 36 months, and 8% at 48 months. The duration of breast- and bottle-feeding was related to maternal work status; mothers who returned to work during the first 3 months postpartum weaned sooner from the breast and later from the bottle than women who returned to work after 3 months postpartum. Conclusions: The frequency of late bottle-weaning in this well-educated, middle-class cohorr was unexpected and was related to the timing of the mother's return to work. The impact of prolonged bottle-feeding on later growth and adiposity deserves further investigation. Arch Pediatr Adolesc Med 1999;153: 189-194. Copyright 1999, American Medical Association.
1124
j. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 38:9, SEPTEMBER 1999
MICHAEL J. BRONDINO, PH.D.
ABSTRACT
Objectives: To examine the concurrent correlates of internalizing and externaliz ing disorders among substance-abusing and substance-dependent juvenile offenders and to determine the association between psychiatric comorb idity and psychosocia l functioning of the youths 16 months later. Method: Participants were 118 juvenile offenders meeting DSM-III-R criteria for substance abuse or dependence and their families. A multisource measurement battery was used to assess drug use, criminal activity, family relations , peer relations, school functioning , and out-of-home placements. Results: Comorbidity for externalizing disorders was associated with high rates of antisocial behavior and predicted worse 16month outcomes than substance abuse alone or substance abuse with comorb id internaliZing disorders . For criminal activity and drug use, the presence of internalizing disorders buffered the deleterious effect of externalizing disorders on substance-abusing and substance-dependent juvenile offenders. Conclusions: Even in substance-abusing delinquents, a population already extreme in antisocial behavior, the presence of externalizing disorders indicates high risk for deterioration. J. Am. Acad. Child Ado/esc. Psychiatry, 1999, 38(9) :1118-1124. Key Words: adolescents, substance abuse, comorbidity, delinquency.
In recent years, reviewershave concluded that substanceabusing youths evidence high rates of comorbid psychiatric problems such as depression and conduct disorder (e.g., Lewinsohn et al., 1995; Weinberg et al., 1998) and that substance-abusing delinquents are at especially high risk of co-occurring mental health disorders (Thompson et al., 1996; White and Labouvie, 1994). To a large extent, the purpose of documenting such comorbidity is to raise issues concerning the influence of co-occurring mental health disorders on th e developmental trajectories of substance-abusing and substance-dependent youths as well as regarding the design of intervention models for such youths.
Accepted March 31, 1999. From theMedical University of South Carolina, Charleston. Thisstudywassupported by N IDA grantR01DA lO0079 (ScottW Henggeler; Principal Investigator). Reprint requests toDr.Randall Family ServicesResearch Center, Department of Psychiatryand BehavioralSciences, Medical University ofSouthCarolina, 67 PresidentStreet, P.O. Box250861, Charleston, SC29425; e-mail: [email protected]. 0890-8567/99/3809-1118©1999 by the American Academy of Child and Ado lescent Psychiatry.
1118
In this regard, adolescent substance abuse with comorbid internalizing or externalizing disorders seems to impa ct adolescents differently than substance abuse alone. For example, adolescent substance abuse with comorbid externalizing disorders, such as conduct disorder, predicted high school dropout (Kessler et al., 1995) and failure to complete treatment on an inpatient unit (Kaminer et al., 1992). In contrast, adolescent substance abuse with comorbid internalizing disorders (e.g., affective disorders) predicted completion of treatment on an inpatient unit (Kaminer et al., 1992). Thus, emerging evidence suggests that types of comorbidity may be linked with differential longer-term outcomes for youths with serious substance abuse problem s. The purpose of this study is to expand on these findings in 2 ways. First, we examine the different concurrent correlates of internalizing and externalizing como rbid conditions in a sampl e of substance-abusing and substance-dependent juvenile offenders.The examined correlates pertain to those domains of functioning that have been consistently linked with substance use and criminal activity by adolescents (Henggeler, 1997; Loeber and Hay,
j . AM. ACAD . C H I LD ADO LESC . PSYCHIATRY, 38:9 , SE PTEMB ER 1999
COMORBIDITY AND FUNCTIONING IN OFFENDERS
1997): family functioning, peer relations, and school functioning. The second purpose is to examine the association between psychiatric comorbidity and the psychosocial functioning of the youths 16 months later. Here, we evaluated whether internalizing and externalizing comorbid conditions predicted improvements or declines in the personal and social functioning of the adolescents. METHOD Subjects One hundred eighteen adolescents served as participants and were recruited from the D epartment ofJuvenile Justice (DJJ) of Charleston, South Carolina, during October 1992 to June 1994. To identify juvenile offenders who met diagnostic cri teria for su bstance abuse or dependence, participants were administered the Structured Clinical Interview for DSM-III-R (Spitzer et al., 1 988), which was supplemented with information from the Personal Experiences Inventory (PEl) (Winters and Henly, 1989). Inclusion criteria included the following: (1) DSM~III-R diagnosis of psychoactive substance ab use or dependence (American Psychiatric Association, 1987), (2) age 12 to 17 years, (3) formal or inform al probationary statu s, (4) Charleston County Resident, and (5) at least one parental figure residing with the adolescent . To enhance external validity, adolescents with preexisting mental health , physical health , or intellectual difficulties were included in the study. We excluded adolescents who were formall y involved in substan ce abu se treatment or who had a sibling who was a stu dy participant. Four hundred twenty-three adolescents were screened, and 140 met the inclusionary criteria. One hundred eighteen (84%) adolescents and fam ilies agreed to participate and completed initial assessments. Families that consented to parti cipate did not differ from refusers on key variables including adolescent age, gender, race, self-reported drug use, or arrest history. Research attrition from initial assessments to follow-up assessments 16 months later was modest. One hundred eight (91.5%) adolescents and families completed follow-up assessments 16 months later. Adolescent s and families who completed follow-up assessments did not differ from noncompleters on key demographic, drug use, and arrest history variables. The low research attrition enhances th e internal and external validity of th e findings .
Design and Procedures An assessment battery was administered to adolescents and primary caregivers in the ir homes and was readministered 16 months later. Research procedures were explained to adolescents and primary caregivers, and caregivers provided informed consent and adolescents provided informed assent . If adolescents or caregivers lacked sufficient reading skills, questions from the assessm ent battery were read to them. Each family was reimbursed $75 per assessment for their time .
Measures A mulrimethod (i.e., adolescents ' self-report, primary caregivers' reports, and archival records ) assessment battery was used to tap key correlates of antisocial behavior as well as substance use and criminal activity. Psychiatric Comorbidity. All participants were ad m inistered the Diagnostic Interview Schedule for Children Version 2.3-child repo rt (DISC) (Shaffer, 1992) to generate DSM-III-R diagnoses. For the
purposes of analyses, participants were considered to have a cornorbid externalizing disorder if they had at least one externalizing diagnosis (i.e., conduct disorder, oppositional defiant disorder, or attentiondeficit hyperactivity disorder) and a comorbid int ern alizing disorder if they had at least one internalizing diagnosis (i.e., major depression , dysthymia, overanxious disorder, agoraphobia, social phobia, simple phobia, separation anxiety, panic disorder, obsessive-compulsive disorder, avoidant disorder, or generalized anxiety disorder) . It should be noted that the child self-report version of the DISC may result in the underreporting of symptoms of attention-deficit hyperactivity disorder (Schwab-Sto ne et al.,1996). Criminal Activity and Drug Use. The Self-Report Delinquency Scale (Elliott et al., 1983) is a 53- item inventory used to assess adolescents' criminal behavior. A summary score, the General Delinquency Index, was used to indicate adolescents' total number of criminal acts for the previo us 6 months. Arrest and offense hist ories were extracted from a data base maintained by the DJ]. Severity of offenses was coded using a Seriousness Inde x developed by H anson et al. (1984). The Seriousn ess Index is a 17-point scale with nu merical values for seriousness (e.g., 1 = truancy, 4 = disorderly conduct, 8 = assault/battery, 11 = grand larceny, 13 = un armed robbery, 17 = murder) . The PEl (Winters and Henly, 1989 ) was used to assess the frequ en cy of adolescent substance use for the previous 3 months. Twelve items from the PEl were used to construct 2 scales that examined alcohol/marijuana use and other illicit drug use. The frequ ency of alcohol and marijuana use was summed and truncated at 100. An illicit drug use scale elicited the frequency of nonprescribed use of cocaine , amphetamines, Quaaludes, barbiturates, LSD , other psychedelics, tranquilizers, heroin, other narcotics, and inhalants. Because of the low prevalence of reported illicit drug use, adolescents were categorized as either using any illicit drug or using no illicit drugs during the previous 3 months. FamilyRelations. The Family Adaptab ility and Co hesion Evaluation Scales-III (O lson er al., 1985) is a 20 -item inventory with Likert-type item s ranging from 1 (alm ost never) to 5 (almost always). The inventory was completed by the adolescent and primary caregiver to assess family adaptability and cohesion . Consistent with the recommendations of Henggeler et al. (1991), cohesion and adaptability were treated as linear scores in analyses. The Issues C hecklist (Robin er al., 1977) is a 44 Ci tem inventory on which adoles cents and caregivers identify topics th at are potential sources of disagreem ent and con flict. Represent ative topics incl ude negotiating bed tim e, receiving low grades, spending free time, and talking back to parents. For each conflict topic identified, adolescents and caregivers rated the intensity of th e conflict on a scale rang ing from 1 (calm) to 5 (angry) and estimated how ofren the topic was discussed during the previous m onth. Peer Relations. The Peer Conform ity Inventory (Bern dt, 1979) measures adole scents' proso cial and antisocial peer conformity and con sists of 30 situ ations (10 ant isocial, 10 pro social, and 10 neutral) in which peers suggested some action . The Revised Problem Behavior Checklist (Q uay and Peters on , 1987) is a 77 -item inventory with item scores ranging from 0 (no problem) to 2 (a severe problem). The inventory was completed by the adolescent 's primary caregiver, and the Socialized Aggression subscale was used as an index of association with deviant peers. School Functioning. The School Functioning subscale of the Social Competence scale of the Child Behavior Checklist (Achen bach, 1991) was completed by th e youth and caregiver. . Out-ofHome Placement. The Monthly Service Utilizat ion Survey (H enggeler er al., unpublished manuscript, 1992) is a semistructured int erview used to collect out-of-home placement data on adolescents .
j . AM . ACAD. C H ILD ADOLESC . PSYCHIATRY, 38 :9, SEPTEMBER 19 99
1119
RANDALL ET AL.
The majority of the interviews were conducted face to face. In the rare exceptions when face-to-face meetings were not possible, telephone interviews were conducted. Researchers interviewed primary caregivers at the end of each month to obtain their retrospective verbal reports of service utilization for that month. The survey was completed each month after the adolescent entered the study through the 16-month follow-up . Primary caregivers provided placement data, and correctional placements were confirmed th rough D]] computer records. O ut-of-home placements included any institution-based domicile, such as detent ion centers, jails, psychiatric or substance abuse hospitals, resident ial treatm ent cente rs, and group homes . Placements at a relative's home or at a surrogate parent's home were not considered our-of-home placements.
Data Analyses Correlation analyses, descriptive statistics, X2 tests, and analyses of variance were run on data collected at initial assessment, and multiple regression analyses were conducted on data collected 16 months later. For data collected at initial assessment, correlations were used to examine th e relation ship between measures of anti social behavior and social funct ioning and internalizing and externalizing comorbid conditions; descripti ve statistics, X2 tests, and analyses of variance were used to explore possible differenc es among comorbid conditions on demogr aphic variables and other sample characteristics. For longitudinal data l hierarchic al multiple regressions were conducted to examine the association between psychiatric comorbidity and the psychosocial functioning of the youths 16 months later.
RESULTS
Sample Characteristics
Demographic and juvenile justice information on the study participants is presented in Table 1. A X2 analysis comp aring the comorbid conditions on number of males only (n = 93) indicated a significant effect for number of males (X 2= 19.39; df= 3,114;p < .002), with a higher than
expected frequency of males with neither cornorbid internalizing nor comorbid externalizing diagnoses (observed 36 versus expected 23) and a lower than expected frequency of males with both comorbid internalizing and comorbid externalizing disorders (observed 7 versus expected 23). A X2 analysis comparing the comorbid conditions on number of females only (n = 25) indicated a significant effect for number of females (X 2 = 10.04; df= 3,114; p < .0182). However, for females there was a lower than expected frequency of females with neither comorbid internalizing nor comorbid externalizing diagnoses (observed 1 versus expected 6) and a higher than expected frequency of females with both comorbid internalizing and comorbid externalizing disorders (observed 11 versus expected 6). Results of all other X2 tests and analyses of variance on demographic variables were not significant. Fifty-six percent of participants met criteria for substance abuse and 44% for substance dependence . In addition to meeting cri teria for substance abuse or dependence, 72% of the sample met DSM~IlI-R criteria for one or more psychiatric diagnoses. Information on individual diagnoses for males and females separately and for male s and femal es com bined is presented in Table 2. As shown in Table 1 at initial assessment, 32 participants were classified as having a comorbid externalizing disorder only, 31 as having a comorbid internalizing disorder only, 18 as having both comorbid externalizing and comorbid int ernalizing disorders, and 37 as having neither comorbid externalizing nor comorbid internalizing diagnoses.
TABLE 1 Demographic and Juvenile Justice Characteristics at Initial Assessment Variable D emographics Age (yr) Race (%) African-American White Other Male (%) Female (%) Medium income (thousands) Two-parent household (%) Juvenile justice Prior incarceration (%) Age of first arrest (yr) Mean no. of arrests Mean % of violent offenses Mean severity of offenses
1120
Not Corn orbid (n = 37)
External (n = 32)
Internal (n = 3 1)
Both (n = 18)
Total (N = 118)
15.1
15.2
15.3
15.3
15.2
18.6 11.9 0.8 30.5 0.8 15- 20 16.1
9.3 16.9 0.8 23.7 25-30 19.5
15.2 10.2 0.8 18.6 7.6 10-15 8.5
6.8 8.5 0 5.9 9.3 10-15 5.9
50 47.5 2.5 79 21 15- 20 50
51 14.0 1.84 30 9.94
62 13.9 2.16 16 9.53
35 14.0 1.90 13 9.39
38 14.5 1.67 11 7.28
46.5 14.1 1.89 17.5 9.04
3.4
J. AM. ACAD . CHI LI) AD OL ESC. PSYCHIAT RY. 38: 9. SEP TE MBER 1999
COMORBIDITY AND FUNCTIONING IN OFFENDERS
Correlational Analyses
TABLE 2 Diagnoses at Initial Assessment Diagnosis Conduct disorder Oppositional defiant disorder ADHD Major depression Dysthymia Overanxious disorder Agoraphobia Social phobia Simple phobia Separation anxiety Panie disorder Panie disorder with agoraphobia Obsessive-compulsive disorder Avoidant disorder Mania Generalized anxiety disorder Anorexia Bulimia Vocal or motor tics Transient tic disorder Tourette's disorder Diurnal enuresis Nocturnal enuresis Encopresis
Male
Female
(n = 93)
(n = 25)
27 7
8 5 2 5 2 6 7 10 9 5 1 7 5
2
4 1
4 3 9
7 2 0 3 6 3 3 1 0 0 0 1 0 1 0 0
Note: Values are percentages . ADHD activity disorder.
Both (N = 118)
35 12
4
4
9 3 10 10 19 16 7 1 10 11 7
1
4
4
5 1 2 3 2 1 2 1 0
1 2 3 1 1 1 1 0
= attention-deficit hyper-
The initial assessment data shown in Table 1 also show that the participants represented a sample that was extreme in antisocial behavior. Results of analyses of variance comparing the 4 comorbid conditions on juvenile justice involvement were not significant.
A correlation analysis was conducted to determine the association between comorbidity for externalizing and for internalizing disorders. The association was not significant (r= -O.lO,p < .30) . Additional correlations were conducted to examine the concurrent associations between psychiatric comorbidity (externalizing disorder and internalizing disorder) and the behavioral and social functioning of the adolescents. Externalizing Correlates. As shown in Table 3, comorbidity for an externalizing diagnosis was consistently associated with increased rates of antisocial behaviors and poorer functioning in family, peer, and school contexts. Specifically, comorbidity for an externalizing disorder was associated with higher rates of criminal offending, alcohol and marijuana use, and other illicit drug use. Similarly, comorbidity for an externalizing disorder was associated with increased association with deviant peers, greater conformity to antisocial peer behavior, lower family cohesion, and poorer school functioning. Internalizing Correlates. Comorbidity for an internalizing disorder was associated with both more favorable and lessfavorableaspects of functioning. On the favorable side, internalizing comorbidity was linked with greater conformity to prasocial peer behavior and increased family structure. Internalizing diagnoses were not linked with higher rates of criminal activity or substance use. On the other hand, comorbidity for internalizing disorders was linked with greater association with deviant peers and more family conflict.
TABLE 3 Significant Univariate Associations Between Comorbidity and Concurrent Antisocial Behavior and Social Relations Comorbidity Measures
Respondent
Externalizing .
Internalizing
SRD General Delinquency Alcohol/Marijuana Illicit Drug Use RPBC Deviant Peer Association Peer Antisocial Conformity Peer Prosocial Conformity FACES-III Cohesion FACES-III Adaptability Family Conflicr-Quantiry Family Conflict-Intensity CBCL School Competence CBCL School Competence
Adolescent Adolescent Adolescent Caregiver Adolescent Adolescent Adolescent Caregiver Caregiver Caregiver Adolescent Caregiver
0.39** 0.22* 0.18* 0.16 0.25** -0.07 -0.21 * -0.07 -0.11 -0.15 -0.18* -0.19*
-0.01 0.06 -0.07 0.19* 0.09 0.18* -0.07 -0.21* 0.18* 0.19* 0.08 0.15
Note: SRD = Self-Report Delinquency Scale; RPBC = Revised Problem Behavior Checklist; FACES-III bility and Cohesion Evaluation Scales-III; CBCL = Child Behavior Checklist.
= Family Adapta-
* p < .05; **P < .01.
J. AM.
ACAD. CHILD ADOLESC . PSYCHIATRY, 38:9, SEPTEMBER 1999
1121
RANDALL ET AL.
TABLE 4 Standardized Regression Coefficients and Significance Levels From rhe Significant Multiple Regression Analyses Relating Comorbidity to 16-Month Outcomes Comorbidity Measures
Respondent
External
Internal
Interact
SRD General Delinquency Illicit Drug Use Days in Out-of-Home Placement FACES-III Cohesion CBCL School Comperence
Adolescent Adolescent Caregiver Adolescent Caregiver
0.38** 0.38** 0.26* 0.31* - 0.15
0.2 5* 0.11 -0. 13 0.08 -0.23*
-0.29* -0.33* -0.21 -0.20 0.03
Note: SRD = Self-Report Delinqu ency Scale; FACES-III Child Behavior Checkl ist * p < .05; **P < .01.
= Family Adaptability and Coh esion Evaluation Scales-III; CBCL =
Longitudinal Analyses
The purpose of these anal yses was to determine whether psychiatric comorbidity was differentially associated with outcomes at 16 months after the initial assessment. Hence, hierarchical regression analyses were conducted in which the measures of antisocial behavior (criminal activity, substance use) and social functioning (family, peer, school) served as dependent measures. The first variable entered into the equation was the corresponding baseline variable, dummy variables indicating the presence or absence of internalizing disorders and externalizing disorders were entered next, and the interaction of externalizing and internalizing disorders was entered last. A similar analysis was conducted on the out-of-home placement measure, except a baseline measure of such was not included. Externalizing Disorders. As shown in Table 4, externalizing comorbidity predicted increased rates of general delinquency and illicit drug use 16 months later. These findings, however, were qualified by significant interactions between externalizing and internalizing disorders. For criminal activity, an examination of comorbid condition means suggested that the presence of internalizing disorders buffered the effect of externalizing disorders on sub stance-abusing and substance-dependent juvenile offenders. As indicated in Table 5, participants with both
comorbid internalizing and comorbid externalizing disorders engaged in a lower level of general delinquency than participants with only a comorbid externalizing disorder.The same pattern of interaction effectswas observed for illicit drug use, with a lower percentage of participants with both comorbid internalizing and comorbid externalizing disorders engaging in a lower level of illicit drug use than participants with only a comorbid externalizing disorder. In addition, comorbidity for externalizing disorder predicted high rates of out-of-home placement and increased family cohesion. Internalizing Disorders. As indicated previously, a significant interaction effect eme rged for gen eral delinquency in which youths with internalizing disorders evidenced less increase in criminal activity than youths with comorbid externalizing disorders only. In add ition , internalizing disorder was associated with decreased school functioning (based on caregiver reports) . Number of DISC Diagnoses. To determine whether changes on dependent measures at 16 months were not merely associated with the number of diagno ses at the time of participant recruitment, hierarchical regression analyses were conducted. The first variable entered into the equation was the corresponding baseline dependent measure followed by the number of DISC diagnoses. None of the results were significant. Hence, in general,
TABLES Means of Significant Interacti on Effects From th e Regression Analyses Relating Comorbidity to 16-Month Outcomes Measures SRD Gen eral Delinquency Mean Illicit D rug Use N ote: SRD
1122
Not Comorbid (n = 37)
External (n = 32)
14.2 1.62
60.3 6.3 1
Internal = 3 1)
Both (N = 118)
25.8 3.30
49.9 2.77
(n
=Self-Report Delinqu ency Scale. ] . AM. ACAD . C HI LD AD OLE SC. PSYCHIAT RY, 38:9, SEPT EMB ER 1999
CO MO RBID IT Y AN D FUNCTIONING IN OFFENDERS
the key determinant of deteriorating functioning was comorbidity for externalizing disorder rather than the number of diagnoses per se. DISCUSSION
Youthswith substance abuse and delinquency represent a population that is extreme in antisocial behavior.Yetthe findings show that substance-abusing delinquents with externalizing disorders are even more antisocial than their substance-abusing delinquent counterparts. They engage in higher rates of crime, more alcohol and marijuana use, and more illicit drug use. In addition, substance-abusing delinquents with comorbid externalizing disorders evidence relatively less family cohesion, greater conformity to antisocial peer pressure, and decreased school competence. The longitudinal analysesprovide further evidence of the increased risk posed by a comorbid externalizing disorder. Sixteen months past the initial assessment, such cornorbidity was associated with higher rates of out-ofhome placement and increased general delinquency and illicit drug use. Delinquency and illicit drug use findings were qualified by significant interactions between internalizing and externalizing disorders. An examination of the comorbid condition means suggested a buffering effect. Substance-abusing and substance-dependent juvenile offenders with both comorbid internalizing and comorbid externalizing disorders engaged in lower levels of general delinquency and illicit drug use than substanceabusing and substance-dependent juvenile offenders with only comorbid externalizing disorders. Thus, the comorbidity of externalizing disorders only seems to add a significant risk factor to a population that already has a poor prognosis. In contrast, the presence of internalizing comorbidity for this sample of substance-abusing and substancedependent juvenile offenders suggests somewhat more favorable characteristics than their substance-abusing delinquent counterparts. Concurrent associations showed that youths with comorbid internalizing disorders had greater conformity to prosocial peer behavior and increased family structure. On the other hand, cornorbidity for internalizing disorders was linked with increased association with deviant peers and increased family conflict. These family-related findings (high conflict and increased structure) might reflect functional caregiver responses to the youths' antisocial behavior and association with deviant peers.That is, in light of the serious problems presented by these youths, increased structure and conflict
are expected if the caregivers are invested in the youth's welfare. With regard to the effects of internalizing disorder on the 16-month trajectories of these youths, findings showed that internalizing comorbidity predicted decreased school competence. Overall, comorbidity predicted worse outcomes than substance abuse alone. However, substance abuse with comorbid externalizing disorders onl y predicted even poorer outcomes than substance abuse with comorbid internalizing disorders. These findings are consistent with previous investigations (Kaminer et al., 1992; Kessler et aI., 1995) suggesting that substance-abusing adolescents with comorbid internalizing disorders may present a different clinical profile than substance-abusing adolescents with comorbid externalizing disorders. One possible explanation for these differences is the role of substance abuse in the adolescents' lives. For substanceabusing adolescents with externalizing disorders, substance abuse might be part of a larger picture of problem behavior (lessor, 1987). However, for substance-abusing adolescents with internalizing disorders, substance abuse may represent a way in which these adolescents selfmedicate to attenuate their affective difficulties. Clinical Implications
Clinical differences between adolescent substance abusers with internalizing or externalizing disorders suggest different treatment emphases. For substance-abusing adolescents with externalizing disorders, treatment should probably focus on enhancing parental competence to structure, monitor, supervise, and provide consequences for their adolescent's behavior. In this regard, the key is to identify those aspects of the family and its social ecology that are barriers to effective parental discipline (e.g., parental drug abuse, low social support, lack of skills) and to intervene accordingly (Henggeler et al., 1998). For substance-abusing adolescents with internalizing disorders, treatment might devote increased attention to the development of coping strategies for anxieties and stress. In light of the link between adolescent substance abuse with comorbid internalizing problems such as depression and suicide (Carlson et al., 1991), safety is also an important issue in treating substance-abusing adolescents with comorbid internalizing disorders. Limitations
Three central limitations of the study should be noted. First, the findings pertain only to the)ldolescent self-report
] . AM . ACAD. CHILD ADOLESC . PSYCHIATRY, 38:9, SEPTEMBE R 1999
1123
RANDALL ET AL.
version of the DISC, and, as noted previously, certain youths underreport certain symptoms (Schwab-Stone et al., 1996). This may account for the low prevalence of attention-deficit hyperactivity disorder (i.e., 4%) reported by the sample. Second, substance use was determined by self-report only, not laboratory measures or caretaker reports. Third, the sample sizewas not sufficient to permit subanalyses examining specific types of externalizing and internalizing disorders. Thus, for example, determining which internalizing disorders most strongly buffered the deleterious effects of externalizing comorbidity was not possible.
REFERENCES Achenbach TM (1991), Manualfor the Child Behavior Checklist and 1991 Profile. Burlington: University of Vermont Deparrment of Psychiatry American Psychiatric Association (1987), Diagnostic and Statistical Manualof MentalDisorders, 3rd edition-revised. Washington, DC: American Psychiatric Association Berndt TJ (1979), Developmental changes in conformiry to peers and parents. DevPsychoI15:608-616 Carlson GA, Rich CL, Grayson P, Fowler RC (1991), Secular trends in psychiatric diagnoses of suicide victims.] AffectDisord21:127-132 Elliott DS, Ageton SS, Huizinga D, Knowles BA, Canter RJ (1983), The Prevalence and Incidence of Delinquent Behavior: 1976-1980 (National Youth Survey Project Reporr No. 26). Boulder, CO: Behavioral Research Institute Hanson CL, Henggeler SW; Haefele WF, RodickJD (1984), Demographic, individual, and family relationship correlates of serious and repeated crime among adolescents and their siblings. ] Consult Clin Psychol 52:528-538 Henggeler SW (1997), The development of effective drug abuse services for youth. In: Treating Drug Abusers Effectively, Egerrson JA, Fox DM, Leshner AI, eds. Cambridge, MA: Blackwell, pp 253-279 Henggeler SW; Burr-Harris AW; Borduin CM, McCallum G (1991), Use of the Family Adaptability and Cohesion Evaluation Scales in child clinical research.] AbnormChildPsychoI19:53-63
Henggeler SW; Schoenwald SK, Borduin MC, Rowland MD, Cunningham PB (1998), Multisystemic Treatment of Antisocial Behavior in Children and Adolescents. New York: Guilford [essor R (1987), Problem-behavior theory, psychosocial development, and adolescent problem drinking. Br] Addict 82:331-342 (special issue: Psychology and Addiction) Kaminer Y, Tarter RE, Bukstein OG, Kabene M (1992), Comparison between treatment completers and noncompleters among dually diagnosed substance-abusing adolescents. ] Am Acad Child Adolesc Psychiatry 31: 1046-1049 Kessler nc, Foster CL, Saunder WB, Stang PE (1995), Social consequences of psychiatric disorders, I: educational attainment. Am] Psychiatry 152: 1026-1032 Lewinsohn PM, Gotlib IH, SeeleyJR (1995), Adolescent psychopathology, IV: specificityof psychosocial risk factors for depression and substance abuse in older adolescents.] AmAcad ChildAdolesc Psychiatry 34:1221-1229 Loeber R, Hay D (1997), Key issues in the development of aggression and violence from childhood to early adulthood. Annu RevPsychoI48:371-41O Olson DH, Portner J, Laves Y (1985), Contributions of family and peers to delinquency. Criminology 23:63-79 Quay He, Peterson DR (1987), Manualfor the Revised Problem Behavior Checklist. Coral Gables, FL: University of Miami Robin AL, Kent R, O'Leary KD, Foster S, Prinz R (1977), An approach to teaching parents and adolescents problem-solving communication skills.
Behav Ther8:639-643 Schwab-Stone ME, Shaffer D, Dulcan MK et al. (1996), Criterion validity of the NIMH Diagnostic Interview Schedule for Children Version 2.3 (DISC-2.3).] Am Acad ChildAdolesc Psychiatry 35:878-888 Shaffer D (1992), Diagnostic Interview Schedule for Children, Version 2.3. New York: Columbia University Division of Child Psychiatry Spitzer RL, Williams JBW; Gibbon M, First MB (1988), Structured Clinical Interview for DSM-III-R: Patient Version (SCED-P 6/1/88). New York: New York State Psychiatric Institute, Biometrics Research Deparrment Thompson L1; Riggs PD, Mikulich SK, CrowleyTJ (1996), Contribution of ADHD symptoms to substance problems and delinquency in conductdisordered adolescents.] Abnorm ChildPsychoI24:325-334 Weinberg NZ, Rahdert E, Colliver JD, Glantz MD (1998), Adolescent substance abuse: a review of the past 10 years.] Am Acad ChildAdolesc Psy-
chiatry 37:252-261 White HR, Labouvie EW (1994), Generality versus specificity of problem behavior: psychological and functional differences. ] DrugIss 42:55-74 Winters KC, Henly G (1989), The Personal Experiences Inventory. Los Angeles: Wesrern Psychological Services
Development of Feeding Practices During the First 5 Years of Life. Lawrence D. Hammer, MD, Susan Bryson, MS, W. Stewart Agras, MD
Objective: To understand the transition from breast- and bottle-feeding to solid-feeding and factors that might affect the duration of breast- and bottle-feeding. Design: Cohorr followed up from birth with relatively well-educated, middle-class parents. Setting: Communiry sample recruited from 3 suburban newborn nurseries (a teaching hospital, communiry hospital, and large health maintenance organization). Participants: One hundred ninety-one healthy full-term infants. Measures: Assessment of feeding practices through the ages of complete weaning from breast- and bottle-feeding. Results: More than 90% of participants breast-fed for at least 2 weeks. Infants of older mothers were weaned from the breast later than infants of younger mothers. First-born infants were weaned from the breast earlier than later-born infants. Eighry-four percent of infants bottle-fed at some time during the first year of life. More than 40% of the cohort was still receiving bottles at 24 months of age, 16% at 36 months, and 8% at 48 months. The duration of breast- and bottle-feeding was related to maternal work status; mothers who returned to work during the first 3 months postpartum weaned sooner from the breast and later from the bottle than women who returned to work after 3 months postpartum. Conclusions: The frequency of late bottle-weaning in this well-educated, middle-class cohorr was unexpected and was related to the timing of the mother's return to work. The impact of prolonged bottle-feeding on later growth and adiposity deserves further investigation. Arch Pediatr Adolesc Med 1999;153: 189-194. Copyright 1999, American Medical Association.
1124
j. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 38:9, SEPTEMBER 1999