Moffitt's theory regarding two types of adolescent antisocial behavior wasinvestigated using a prospective, longitudinal study of normal and abnormal development in aprimarily low socioeconomic status, ethnically diverse sample. Results supported the presence ofan early-onset/persistent (EOP) group and an adolescence-onset (AO) group. Groups weremost reliably and significantly distinguished by indices of socioemotional history within the first3 years, but no significant differences were found on early measures of temperament orneuropsychological functioning. EOPs scored significantly lower than other groups on measuresof neuropsychological functioning only during late childhood and adolescence,suggesting that the declines in verbal functioning that have been so reliably found in this andother samples of early-starting antisocial adolescents are progressive and consequent to adverseexperience. In adolescence, AOs were significantly more likely to report high levels ofinternalizing symptoms and life stress, suggesting that AO antisocial behavior is not a benignphenomenon. Implications of these findings for etiologic theories of adolescent antisocialbehavior are discussed.

This paper presents single-subject analyses of newly developing attachment relationships in10 foster infant–caregiver dyads. Using a diary methodology, at least 2 months of dailydata were provided by foster parents on infants' attachment behaviors. Foster infantattachment was also assessed using the Strange Situation. Foster mother state of mind regardingattachment was measured using the Adult Attachment Interview. For eight infants, diary datarevealed predominant patterns of attachment behavior emerging within 2 months of placement.In most cases, diary data predicted Strange Situation classifications. Both Strange Situation anddiary data indicated that the three children placed in foster care before 12 months of age withfoster parents having primary or secondary autonomous states of mind were classified as havingsecure attachments. The five children placed after 12 months of age showed predominantlyinsecure attachment behavior in the diary and were classified as insecure in the Strange Situation.Contingency analyses of behavioral sequences reported in the diary revealed that foster parentstended to complement their foster childrens' attachment behaviors.

Eighty-two mother–infant dyads, comprising women with psychiatric disorder andindividually matched controls, were followed up over the children's 1st year of life. Themothers with mental illness consisted of two subgroups: first, 25 severely mentally ill motherswho had been admitted to a psychiatric unit with their infants; and second, 16 mothers from acommunity sample meeting research diagnostic criteria for unipolar, nonpsychotic depression.With the exception of six dyads in the in-patient group, observations were made of themother–infant interaction and the quality of the infant–mother attachmentrelationship at 12 months. The nature and course of the mothers' illness was alsodocumented. Although few residual symptoms of maternal mental illness were detected at 1 yearpostpartum, interactional disturbances were evident among the case group dyads. A strongassociation was revealed between infant–mother attachment quality and maternaldiagnosis: a manic episode of illness in the postpartum period was related to security in theattachment relationship, and psychotic or nonpsychotic depression was related to insecurity.Concurrent patterns of mother–infant interaction provided support for this finding.

Eighteen children with specific language impairment (SLI), from 6 to 8 years of age, werecompared with 9 control children matched on age and nonverbal ability (CA controls) and with 9younger control children of comparable language level (LA controls). Half of the SLI group wererated on a teacher checklist as having pragmatic difficulties: these were referred to as thepragmatic language impairment (PLI) group; the remainder were the typical (SLI-T) group.Children's responses to adult soliciting utterances were compared. All children usuallyresponded to conversational solicitations, but children in the PLI group were more likely thancontrol children to give no response, and they also made very little use of nonverbal responses,such as nodding. Nonverbal responding was closely related to the quality of children'sresponses. Children who failed to use nonverbal responses also had a relatively high level ofpragmatically inappropriate responses that were not readily accounted for in terms of limitedgrammar or vocabulary. This study lends support to the notion that there is a subset of thelanguage-impaired population who have broader communicative impairments, extending beyondbasic difficulties in mastering language form, reflecting difficulty in responding to andexpressing communicative intents. The analytic methods developed for this project have promisefor the study of pragmatic difficulties in other clinical groups.

The influences of three risk factors (insecure attachment, maternal depressive symptoms, andeconomic risk) on children's depressive symptomatology were examined. Subjects were85 children from a range of income levels; all subjects were between 7 and 9 years of age.Children's depressive symptomatology was assessed with the Dimensions of DepressionProfile for Children and Adolescents. Results indicated that children at high economic risk weremore likely to exhibit depressive symptomatology than were children at low economic risk.Security of attachment was significantly related to children's depressive symptomatology.In addition, a significant relation was noted between children's depressivesymptomatology and the depressive symptomatology of their mothers. Contrary to expectations,no significant relation was found between maternal depressive symptomatology and security ofattachment. A multiple regression analysis revealed that security of attachment, maternaldepressive symptomatology, and economic risk accounted for 47% of the variability inchildren's depression scores. Secure attachment served as a buffer; economic risk wasassociated with depressive symptoms only among insecurely attached children.

The validity of a developmentally based life-stress model of depression was evaluated in 88clinic-referred youngsters. The model focused on (a) the role of child–environmenttransactions, (b) the specificity of stress–psychopathology relations, and (c) theconsideration of both episodic and chronic stress. Semistructured diagnostic and life-stressinterviews were administered to youngsters and their parents. As predicted, in the total samplechild depression was associated with interpersonal episodic and chronic stress,whereas externalizing disorder was associated with noninterpersonal episodic andchronic stress. However, the pattern of results differed somewhat in boys and girls. Youngsterswith comorbid depression and externalizing disorder tended to experience the highest stresslevels. Support was obtained for a stress-generation model of depression, wherein childrenprecipitate stressful events and circumstances. In fact, stress that was in part dependent onchildren's contribution distinguished best among diagnostic groups, whereas independentstress had little discriminative power. Results suggest that life-stress research may benefit fromthe application of transactional models of developmental psychopathology, which consider howchildren participate in the construction of stressful environments.

The Relational Psychotherapy Mothers' Group (RPMG), a developmentally informed,supportive psychotherapy designed to serve heroin-addicted mothers with children up to 16 yearsof age, aims at addressing psychosocial vulnerabilities, and facilitating optimal parenting, amongat-risk mothers. We present preliminary evidence on the efficacy of RPMG as an “addon” treatment in comparison with standard methadone counseling alone. At the end of the24-week treatment period, mothers receiving RPMG plus standard methadone counselingdemonstrated lower levels of risk for child maltreatment, greater involvement with their children,and more positive psychosocial adjustment than women who received methadone counselingalone. Children of RPMG participants also reflected fewer problems in multiple areas. At 6months posttreatment, RPMG recipients continued to be at a relative advantage, although themagnitude of group differences was often attenuated. Notably, urinalyses indicated that RPMGmothers showed greater improvements in levels of opioid use over time than comparisonmothers.