Early Risk Factors and Pathways in the Development of Early Disruptive Behavior Problems
Recently, there has been a growing interest in identifying the early antecedents of disruptive behavior problems. The rationale for this interest is clear. Early externalizing problems, particularly those involving aggressive and destructive behavior, have been found to be relatively stable beginning as early as age 2 (Cummings, Iannotti, & Zahn-Waxler, 1989; Fagot, 1984; Olweus, 1979), and predictive of more serious antisocial behavior during the school-age period and beyond (Robins, 1966; Robins, West, & Herjanic, 1975; Tremblay, Pihl, Vitaro, & Dobkin, 1994). More serious forms of antisocial behavior have been highly resistant to intervention among school-age children and adolescents (Kazdin, 1995), and financially and emotionally costly to problem youngsters, their families, and society (Loeber, 1982). Because developmental trajectories leading to adaptive or maladaptive outcome begin in the period from infancy to preschool, this period is considered critical to development (Campbell, 1995). Therefore, a growing number of researchers have begun to trace the precursors of early disruptive behavior problems in young children (Campbell, 1994; Gardner, 1989; Lyons-Ruth, Alpern, & Repacholi 1993).
In reviewing the literature on risk factors leading to childhood externalizing problems, it appears that several types may be implicated. Research on preschool and school-age children has demonstrated that children with externalizing disorders are more likely to show an increased likelihood of difficult temperaments (Bates, Maslin, & Frankel, 1985) and dysfunctional attachment relationships (Erickson, Sroufe, & Egeland, 1985), higher rates of parental maladjustment (Shaw, Vondra, Dowdell Hommerding, Keenan, & Dunn, 1994; Zahn-Waxler et al., 1988), and impoverished family sociodemographic characteristics (Rutter, 1978; Sanson, Oberklaid, Pedlow, & Prior, 1991). In previous studies of childhood disruptive behavior problems using socioeconomically heterogeneous samples, it has been shown that risk factors from multiple domains are found more often in families from lower socioeconomic backgrounds (Richman, Stevenson, & Graham, 1982; Rutter, 1978). However, because multiple risk factors tend to cluster in the same low SES environments, it has been difficult to disentangle the exact pathways leading from risk to the development of disruptive behavior problems. Relatively few studies have been conducted to examine how the interrelationships among these risk factors unfold in the first five years of children's lives within the context of a predominantly low SES context (Erickson et al., 1985). In the present study, risk factors were ascertained during the first and second years of life, and included infant characteristics, maternal adjustment and resources, and sociodemographic measures of family adversity.
Risk Factors during Infancy
Child Effects
Research on infant characteristics associated with later externalizing problems has been primarily focused on difficult temperament and attachment security. Several investigators have examined the relation between early temperamental difficulty and later behavioral adaptation, beginning with Thomas, Chess, and Birch's (1968) ground breaking study, and continuing more recently with the studies of Earls and Jung (1987) and Sanson and colleagues (1991). Though the conceptualization of temperament remains controversial, and there is continued debate about the meaning and measurement of temperamental attributes after the first year of life (i.e., the influence of biological versus environmental influences), there is some consensus that negative emotionality, in the form of fussiness and irritability, are at the core of a difficult temperament during infancy. It is believed that difficult temperament may influence the course of later externalizing problems directly through a relation to later oppositional and aggressive behavior (Graham, Rutter, & George, 1973), or indirectly through its effects on attitudes and behaviors of caregivers (Bates, 1980). Studies examining the direct effects of difficult temperament on later externalizing behavior problems in the preschool and school-age periods have shown modest to moderate relations (Maziade, Cote, Bernier, Boutin, & Thivierge, 1989; Sanson et al., 1991).
Infant attachment security also has been found to be related to later externalizing problems, particularly among samples of high-risk children (Erickson et al., 1985; Shaw & Vondra, 1995). In a recent longitudinal study using a high-risk sample, the disorganized insecure classification was strongly predictive of preschool-age aggressive behavior (Lyons-Ruth et al., 1993), a relation which has persisted at age 7 (Lyons-Ruth, Easterbrooks, Davidson, Cibelli, & Bronfman, 1995). The measurement of infant attachment security has been operationalized and studied using the Strange Situation (Ainsworth & Wittig, 1969). Though it remains controversial whether infant attachment security is more of a reflection of the infant's temperament (i.e., proneness to distress and proximity seeking, see Kagan, 1984), or the history of the caregiver's responsiveness towards the infant (Sroufe, 1983), past research has shown that it places children in high-risk environments at increased risk for preschool disruptive behavior problems.
Parental Adjustment and Resources
Several researchers have attempted to identify parental attributes such as antisocial behavior and psychopathology, that are related to the development of children's early disruptive behavior. Some studies were based on similar work with school-age children and adolescents regarding antisocial personality disorder (Robins et al., 1975). Although relations between parental antisocial personality and behavior problems in younger children rarely have been examined, parents' antisocial behavior has been shown to precede older children's behavior problems (Robins et al., 1975). In an earlier follow-up of the present cohort, above-average scores on maternal self-report of aggression and suspiciousness, and below-average scores of social desirability during infancy were related to child externalizing problems at age 3 (Shaw et al., 1994). In addition to antisocial personality attributes, associations between child behavior problems and other forms of parental psychopathology, such as unipolar and bipolar depression, have been explored. Zahn-Waxler and colleagues (1988) followed a small group of children from families in which one parent had been hospitalized for bipolar disorder prior to the children's birth. Compared to a matched control group, target children were more likely to have a variety of psychiatric problems at age 6, particularly conduct disorder. Investigators comparing young children with disruptive behavior problems to normal controls have found that mothers of children with behavior problems report more depressive symptomatology (Mash & Johnston, 1983), and these differences in externalizing problems persist at follow-up (Campbell, March, Pierce, Ewing, & Szumowski, 1991; Webster-Stratton, 1990).
Children from families with few and/or pressured parental resources also have shown increased risk for externalizing problems. Factors such as marital dissatisfaction (Jouriles et al., 1991), parental conflict (Emery, 1988; Block, Block, & Gjerde, 1986), and perceived parenting hassles (Crnic & Greenberg, 1990), all indicate significant associations with child externalizing problems. Similarly, social support dissatisfaction from sources outside of the nuclear family (Crnic, Greenberg, Ragozin, Robinson, & Basham, 1983) also has been found to add unique variance to the prediction of child psychopathology, particularly in interaction with other risk factors (Crockenberg, 1981).
Sociodemographic Risk Factors
Families with low socioeconomic status are more likely to experience the child and parent risk factors noted above (e.g., higher rates of insecure attachment, higher rates of maternal depression), as well as other stressors that are intertwined with low socioeconomic status. These include single-parent status, overcrowding in the home, and low family income. Children already at risk because of parental psychopathology and/or poor family functioning appear to be especially vulnerable to the effects of such stressors (Compas, Howell, Phares, Williams, & Giunta, 1989; Pianta, Egeland, & Sroufe, 1990; Rutter, 1981), each of which have been associated with a greater risk of maladaptive outcome for children. However, in relation to the development of childhood antisocial behavior, Loeber and Dishion's (1983) meta-analysis indicated that effects of sociodemographic factors such as single-parent status, family income, and parental education, exert consistent but modest effects in comparison to child or parent risk factors.
In summary, previous research on risk factors leading to early child externalizing problems has shown the importance of child, parent, and to a lesser extent, socioeconomic factors. However, few investigations have compared the predictive validity of these individual variables over time.
While examining relations between risk factors and later externalizing problems is important, so is examining the multiple pathways children travel on their way to developing clinically-elevated rates of disruptive behavior problems. Much of what we know about the predictors and correlates of early disruptive behavior suggests that there are multiple pathways to the same destination (Richters & Cicchetti, 1993), one of the reasons this special section is concerned with the issue of equifinality. Despite the fact that many of these same child, parent, and sociodemographic risk factors tend to be found in the same high-risk environments, it behooves researchers to examine the exact and different trajectories children take to reach similar destinations. Relatedly, comparing the trajectories between those who do and do not develop clinically-elevated problems permits us to explore risk factors shared by children with externalizing disorders. Thus, our second goal was to investigate both the similarities and differences in pathways children take during the first two years who go on to develop clinically-elevated externalizing problems at preschool.
Method
Subjects
A sample of 100 low SES mothers and infants was recruited from the Allegheny County Health Department's Women, Infants, and Children (WIC) Nutritional Supplement Program, which provides nutritional services to low-income families. Mothers were between 17 and 36 years of age at the time of recruitment, 60% were single, 40% were African-American, and 73% were unemployed. Marital status was not restricted due to the considerable instability within this population. See Shaw and colleagues (1994) for additional information about the demographic characteristics of the sample.
Procedures
Mothers of infants six to eleven months of age were recruited to participate in on-site screening interviews at WIC offices, for which they received $5. Mothers who agreed to the screening interview were then recruited to participate in a longitudinal study of infant development and were told that they would be paid $15.00 for each visit, plus $10 for transportation costs. Those who agreed to participate were scheduled for a University visit within two weeks of their infant's first birthday. Of the 144 who were asked to take part in the study, 129 (89.6%) agreed to participate, but only 100 (69.4%) completed the 12-month assessment. Of 100 subjects seen when infants were 12 months old, 89 participated in the 18- and 24-month laboratory assessments months. However, due to errors in the videotaping of assessments, sample sizes for specific videotaped measures were slightly less than the total sample, particularly at the 12-month assessment. At age 3, 82 mothers returned completed questionnaire reports on their children's behavior problems. No significant differences were found when demographic characteristics of families who completed the 12-, 18-, 24-, and 36-month assessments (82 of the 100) and those who did not (18 of the 100) were compared. A similar comparison was made between families who completed the age 4 3/4 assessment to those who did not, with the same result.
The laboratory assessments conducted at 12, 18, and 24 months lasted approximately two hours. They consisted of seven components: a 15-minute free-play situation, a high-chair (12 months only) or clean-up task (18 and 24 months only), a three-minute period with no toys in the room (Martin, 1981), four (12 months) or three (18 and 24 months) teaching tasks (based on Matas, Arend & Sroufe, 1978), the Strange Situation (Ainsworth & Wittig, 1969), and a 5-minute free play. Mothers and infants took a 15-minute snack break prior to the Strange Situation procedure.
A staff member read aloud all questionnaires to the mother, with the exception of the Beck Depression Inventory and the Bates Infant Characteristics Questionnaire. These instruments were completed by mothers during tasks in which time they were asked to simultaneously respond to their children's needs (e.g., crying, talking; see Shaw et al., 1994 for more information on these tasks). The infant remained in the mother's sight throughout the visit, and the mother was asked to attend to her child as she normally would.
Measures
The measures were grouped into three categories: (1) child characteristics, (2) parental adjustment and personal resources, and (3) sociodemographic factors.
Child Characteristics
Infant Characteristics Questionnaire (ICQ). The ICQ assesses temperamental characteristics from which factor scores have been constructed indexing infant behavior (Bates, Freeland & Lounsbury, 1979). For the present investigation, the Difficulty factor was used, given its relation to preschool behavior problems and comparability of items from 12 to 24 months. Since the item structure of the Difficulty factor changes at 6, 13, and 24 months, the seven-item 6-month factor was used at the screening, the nine-item 13-month factor was used at the 12-month assessment, and the 7-item 24-month factor was used at the 18- and 24-month assessments. The instrument meets customary psychometric standards for maternal reports on infant temperament, has the advantage of brevity, and has shown longitudinal relations with preschool behavior problems (Bates et al., 1985). The ICQ was administered four times during the child's first two years. Two composite scores were created based on theoretical and empirical grounds (r was equal to or greater than .51 for composited scores, p<.0001). "Year 1" scores refer to the averaged recruitment and 12-month assessment ratings, and "Year 2" scores refer to the averaged 18- and 24-month ratings.
Strange Situation (SS) Procedure. The measure of attachment security employed in this investigation was the SS (Ainsworth & Wittig, 1969), carried out at 12 and 18 months. All SSs were coded into one of four categories (A, B, C, D) according to the procedures described by Ainsworth,Blehar, Waters, and Wall (1978) and by Main and Solomon (1987). So that the data of the present cohort could be compared with previous studies examining later disruptive behavior, three different methods were used to compare attachment security classifications (Erickson et al., 1985; Lyons-Ruth et al., 1993; Shaw & Vondra, 1995). This included comparing: (1) secures (B) versus all insecures (A, C, D), (2) disorganized (D) versus all other categories (A, B, C), and (3) all individual categories (A, B, C, D). Three coders, blind to scores for maternal behavior, were trained to reliability and tested for interrater agreement using two different sets of attachment assessments, one set from the lab of J. Belsky and a second set from the lab of A. Sroufe. The instructor and principal coder was also trained in D classifications by D. Cicchetti, using tapes from his lab. Interrater agreement on major classifications ranged from 80% to 100% with a mean of 86% for the test assessments, and 80% for all combinations of raters using a random set of 10 attachment assessments (representing all major classifications) from the current study.
Traditionally, the SS is administered at the beginning of an assessment to heighten the novelty of the situation for the infant and to prevent other activities from affecting the child's behavior during the assessment of infant attachment. Theoretically, this assumption is difficult to dispute, particularly in the absence of empirical data. In this study, the SS was administered during the second half of the laboratory visit, following a 15-minute break. To test whether the integrity of the SS may have been affected by antecedent procedures, we administered the procedure at the start of the 12-month assessment with a completely independent cohort. This group, which was recruited approximately two years later from the same WIC sites, did not differ significantly from the present group on any demographic characteristics. The proportion of infants in each of the four attachment categories (A, B, C, D) was not significantly different between the two cohorts (Shaw et al., 1995). These data do not necessarily imply that the children of the present study did not act differently in the SS as a result of being exposed to a series of tasks, some of which are affectively provocative. However, with respect to changing rates of attachment classifications, we have found no evidence to suggest this would be the case.
Child Behavior Checklist for Ages 2-3 and 4-16 (CBCL) (Achenbach, 1991, 1992). These two versions of the CBCL are the most widely used behavior checklists of childhood behavior problems. In the present study, mothers completed the age 2-3 version of the CBCL when children were age 3 and the 4-16 version when children were 4 3/4 years old. Both the age 2-3 and 4-16 parent-versions yield broad- and narrow-band factors indicative of externalizing and internalizing problems. However, because we were interested in pathways leading to early disruptive behavior problems, we decided to limit outcome to two scales, the broad-band Externalizing and the narrow-band Aggression factors. To examine both the full range and clinically-elevated scores, CBCL outcome was examined using both continuous and clinical cut-off scores. We used t-scores at or above 63 (90th percentile) to indicate clinically-significant scores. Separate norms were computed by sex for the age 4-16 version of the scales (no sex differences were found before age 4). Both the Externalizing and Aggression factors were normed separately by age group. Research with the parent version is based on diagnosable psychiatric disturbance according to Diagnostic and Statistical Manual-III (DSM-III) criteria (Edelbrock & Costello, 1988).
Maternal Adjustment and Resources
Personality Research Form (PRF). Items from three factors of the PRF (Jackson, 1989), Aggression, Defendence (i.e., suspiciousness), and Desirability, were administered to mothers at the screening interview and 18-month assessment to provide information on these personality characteristics. Each factor is represented by 16 items, for a total of 48 true-false questions. These factors were selected based on their ability to discriminate stable secure (secure ---> secure) from unstable secure (secure ---> insecure) infant attachment classifications from 12 to 18 months of age (Egeland & Farber, 1984). Internal consistency reliabilities for the three factors range from .72 to .87, while test-retest reliabilities range from .84 to .87 (Jackson & Morf, 1973). In the present study between the screening and 18-month assessments (a 7-12 month period), test-retest correlations were .59, .45, and .59 (p < .001 for all) for Aggression, Defendence, and Desirability, respectively. Ratings from the recruitment administration are termed "Year 1 personality risk" scores, and 18-month assessment ratings are referred to as "Year 2 personality risk" scores.
Beck Depression Inventory (BDI). The BDI, a well-established and widely-used measure of depressive states (Beck & Beamesderfer, 1974; Beck, Ward, Mendelon, Mock, & Erbauch, 1961; Reynolds & Gould, 1981), was administered at the screening, 12-, 18-, and 24-month assessments. Split-half reliability of the scale is high (.86 to .93). Because the BDI also was administered four times during the course of the study, we decided to form two composites based on assessments in the first and second years of life, in the same manner as the ICQ. Empirically, this was justifiable given that correlations between composited BDI scores were equal to or greater than .53, p <.0001. Scores for the recruitment and 12-month administrations were summed and averaged, reflecting maternal depressive symptoms in the first year of the infant's life (termed "Year 1" below), as were scores for the 18- and 24-month administrations (termed "Year 2" below), reflecting BDI scores during the infant's second year. Previous research has identified maternal depression to be associated with externalizing behavior problems in early childhood (Shaw et al., 1994; Zahn-Waxler, Iannotti, Cummings, & Denham, 1990). Instructions for the BDI were altered to include the subject's symptomatology during the last six months (instead of two weeks) in order to give a more stable indication of affect between assessments.
Parental Criminality. As part of the Background Information Questionnaire (see below), mothers were asked about criminal behavior committed by either of the infant's parents living at home at the 12-month assessment.
General Life Circumstances Questionnaire (GLS). Maternal social stress and support was evaluated with the GLS at the 12-, 18, and 24-month assessments, using the sum of two factors, Amount and Satisfaction (Crnic et al., 1983). Mothers were asked to indicate the amount of support received from neighbors, friends or family, organized groups, and intimate relationships, as well as their level of satisfaction with that support. Ratings were made on a four-point scale. Responses on the GLS from mothers of infants have demonstrated substantial concurrent relations to other measures of perceived stress and social support (Crnic et al., 1983). Year 1 scores were based on the 12-month assessment; Year 2 scores were the average of the 18- and 24-month scores.
Marital Adjustment Test (MAT). Maternal satisfaction with her marital or significant other relationship was assessed using the short form of the MAT at the 12-month laboratory visit (Locke & Wallace, 1959). This measure has proven successful in discriminating harmonious and disturbed marriages (Locke & Wallace, 1959; Rosenbaum & O'Leary, 1981), and predicting children's behavior problems (Emery & O'Leary, 1982). In the event of a recent separation, mothers were instructed to report on that period within the last year when she and her partner were living together. In those instances where mothers were not married, they were asked to rate their closest intimate relationship, including a live-in boyfriend, girlfriend, relative, or current dating partner, substituting the word "relationship" or "close relationship" for "marriage." When the relationship was non-sexual in nature, the items concerned with sex relations and relationships with in-laws were omitted. This strategy is responsive to the fact that many of the study's mothers were single (60%), and allowed for the provision of important information on whichever close relationship mothers considered to have primacy.
Child-Rearing Disagreements Scale (CRD). The CRD is a 21-item measure of common topics concerning child-rearing disagreements, based on interviews with parents over the course of several years (Jouriles et al., 1991). It was administered at the 24-month assessment. In comparison with general marital satisfaction scales, the CRD has been correlated with a greater variety of child behavior problems and predicts significant variance in behavior problems after accounting for non-child disagreements and child exposure to marital conflict. The response format is a 6-point Likert scale with a Cronbach alpha of .86. The CRD was administered in a similar manner to the MAT to account for contextual variability of childrearing arrangements among study participants. In those instances where mothers were not married, they were asked to complete the questionnaire if there was a co-parent who was responsible for rearing the child with them, including live-in boyfriends, ex-husbands, mothers, or other live-in relatives and friends.
Parenting Daily Hassles (PDH). The PDH is a 20-item measure of everyday events parents encounter with children. Parents rate the frequency of occurrence on a 4-point scale and how hassled they felt by the event on a 5-point scale (Crnic & Greenberg, 1990). The PDH was administered in the present study at the 24-month assessment. The sum of two factors, frequency and intensity was used in analyses. Cronbach alphas of .81 and .90 have been demonstrated for the frequency and intensity factors, respectively. Research with the PDH has demonstrated strong relations with child behavior outcome over and above those accounted for by more global life stresses (Crnic & Greenberg, 1990).
Sociodemographic Factors
Background Information. This questionnaire, administered at the screening and the 12-, 18-, and 24-month assessments, included items regarding demographic information and data on the home environment. Information pertaining to family income, parental education and occupation, single-parent status, and overcrowding were derived from this questionnaire. Socioeconomic status (SES) was derived from parental education and occupation at 12 and 18 months using Hollingshead's (1975) four-factor index of social status. For single parents, only mother's occupational status and educational attainment were used to calculate SES.
Results
Results are presented in four stages: (1) descriptive statistics among independent and dependent variables; (2) relations among risk factors and univariate relations among risk factors and CBCL scores at age 5; (3) multivariate prediction of continuous and clinically-elevated CBCL scores from risk factors; (4) group comparisons of risk factors between those children with and without clinically-elevated externalizing problems at age 5.
Descriptive Statistics
Means and standard deviations for all continuous variables are presented in Table 1, as well as cutoff scores used in later analyses. Case numbers for individual measures vary according to attrition at the follow-up assessments and inappropriateness of the MAT and CRD scales for all subjects. Scores for the different-age versions of the Difficulty factor of the ICQ are comparable to published norms (Bates et al., 1979). BDI scores also are comparable to other samples of low-income mothers (Christopoulos et al., 1987). Achenbach CBCL t-scores at age 3 are similar to published norms, whereas scores on the Externalizing factor at age 5 were .4 a standard deviation greater than those reported in the standardization sample (Achenbach, 1991, 1992).
__________________________________
Insert Table 1 about here
__________________________________
Individual Relations Among Specific Risk Factors and With Disruptive Behavior Problems
Relations among individual risk factors at Year 1 and at Year 2 are presented in Table 2A below using Pearson and point-biserial (for dichotomous predictor variables) correlations. Though the number of relations attaining statistical significance was beyond that expected by chance (23 of 128 or 18%), very few relations were significant across child, parent, or sociodemographic domains. For example, among Year 1 and 2 risk factors 3 of 60 (5%) correlations were significant between child risk factors and those from parental and sociodemographic domains, a percentage that would be expected by chance. Among parent factors, maternal depressive symptomatology was concurrently related to social support, marital satisfaction, childrearing disagreements, parenting daily hassles, and personality risk at both Year 1 and Year 2. Similarly, among sociodemographic risk factors, significant associations were found among SES, single-parent status, and family income.
In Table 2B, Pearson and point-biserial correlation coefficients are presented between individual predictor variables and the CBCL Externalizing and Aggression factors. Child's sex was included as a predictor variable to examine its effects on early disruptive behavior problems. Year 1 significant correlates of both age 5 Aggression and Externalizing behavior included insecure attachment (i.e., secures vs. all insecures), disorganized attachment (i.e., disorganized versus all other attachment classifications), and maternal personality risk. In addition, low social support and low marital satisfaction were predictive of Externalizing problems only. At Year 2, difficult temperament and several parent variables were predictive of both Aggression and Externalizing problems, including childrearing disagreements, parenting hassles, and personality risk.
__________________________________
Insert Tables 2A-B about here
__________________________________
To examine relations between individual attachment classification groups and age 5 CBCL factors, a MANCOVA was used to predict continuous CBCL scores, and a series of Chi Square analyses were conducted to examine relations with clinical cutoff scores. In the MANCOVA, the four attachment groups served as the independent variable, child's sex was the covariate, and continuous scores of the age 5 Aggression and Externalizing factors were the dependent variables. No attachment or sex effects, or attachment by sex interactions were significant for the age 5 Aggression or Externalizing factors.
Four Chi Square analyses were computed to examine relations between attachment security classification at 12 and 18 months and clinically-elevated CBCL factor scores. No significant effects were found between 18-month attachment status and Aggression, or between 12- or 18-month attachment status and clinically-elevated Externalizing. However, the chi square was significant between attachment classification at 12 months and clinically-elevated Aggression, X2 (3, 74) = 8.15, p < .05. Results are displayed in Figure 1. Sixty percent of infants classified as insecure-disorganized at 12 months had scores in the clinical range on the Aggression factor, as compared with 17% of secure, 31% of insecure-avoidant, and 38% of insecure-resistant infants.
__________________________________
Insert Figure 1 about here
__________________________________
Multivariate Relations between Risk Factors and Age 5 Behavior Problems
Multiple Regression Analyses
In order to examine the effects of individual risk factors from a multivariate perspective, a series of multiple regression procedures were computed using risk factors across domains. In the first series of regressions the age 5 CBCL Aggression and Externalizing factors served as dependent variables. The format of these regressions involved the stepwise-elimination entry (with a p < .05 criterion for variable retention) of Year 1, then Year 2 risk factors selected from those showing significant associations with age 5 CBCL Aggression and Externalizing (see Table 2A). A stepwise procedure was used because there was a concern with conserving power and there was no a priori theoretical rationale for the entry of risk factors other than the child's age. Interaction terms were entered after individual predictor variables. Child's sex was not entered as a control variable in these analyses to conserve power and because no significant effects were found for sex in any of the MANCOVA analyses.
In the prediction of age 5 CBCL Aggression, the overall model was significant, F (3, 72) = 7.94, p < .001. Year 1 disorganized attachment (R2 change = .11, p < .003), Year 2 infant difficulty (R2 change = .07, p < .02), and the disorganized attachment x infant difficulty interaction (R2change = .06, p < .02) contributed significant variance to the prediction of age 5 Aggression scores. Follow-up analysis of the interaction between Year 1 disorganized attachment and Year 2 infant difficulty indicated a difference on Aggression of approximately two standard deviations between those infants with disorganized attachments at 12 months and Year 2 difficulty scores above the sample median, and those with only one or no risk factor present (F = 8.27, p < .001). The six children with disorganized attachment at 12 months and above-median scores on the Year 2 difficulty factor had a mean t-score of 72.0 (99th percentile) on the Aggression factor of the CBCL at age 5, versus a score of 59.7 for those infants only with above-median difficulty (n = 34), 57.5 for those with only disorganized attachment (n = 4), and 58.1 for those children with neither risk factor present (n = 33). To place these findings in context, Achenbach (1991) reports a mean t-score of 54 and a standard deviation of 6 for the Aggression factor.
The overall model also was significant in predicting age 5 Externalizing scores, F(3, 73) = 3.90,p < .02). Both Year 1 social support (R2 change = .08, p < .02) and Year 2 infant difficulty (R2change = .06, p < .04) contributed unique variance to the prediction of age 5 Externalizing problems. The interaction between social support and infant difficulty was nonsignificant.
__________________________________
Insert Figure 2 about here
__________________________________
Logistic Regression Analyses
To examine the predictive value of risk factors across domains for clinically-elevated CBCL factor scores, a series of logistic regression analyses (LRA) was computed using age 5 Aggression and Externalizing as dependent variables. For these analyses, independent variables were selected based on correlational analyses with clinical cutoff-scores (p < .05, see Table 2B). A stepwise procedure was used to eliminate nonsignificant predictors.
The overall model for predicting clinically-elevated age 5 Aggresion was significant, X2 (4) = 24.41, p < .001. Year 1 insecure attachment (Wald = 4.60, p < .04, odds ratio = 3.80), Year 1 maternal personality risk (Wald = 7.11, p < .01, odds ratio = 8.40), Year 2 childrearing disagreements (Wald = 4.39, p < .04, odds ratio = 1.03), and Year 2 overcrowding (Wald = 4.01, p< .06, odds ratio = .09) contributed unique variance to the prediction of clinically-elevated Aggression. In the equation for Externalizing, the overall model was significant, X2 (2) = 10.04, p< .01, but only Year 1 disorganized attachment status (Wald = 5.67, p < .02, odds ratio = 6.00) and Year 2 maternal personality risk (Wald = 5.80, p < .02, odds ratio = 5.25) were found to contribute significant variance. In both of the regression equations involving Aggression and Externalizing, no interactions among predictor variables were found to be significant.
Pathways Analysis
In order to trace the developmental trajectories of children who demonstrated clinically-elevated CBCL externalizing problems at age 5, subjects were divided twice into two groups: those who did and did not exhibit t-scores at or above 63 (90th percentile) on the CBCL Aggression and Externalizing factors. Groups were compared on individual child, parent, and socioeconomic risk factors at ages 1 and 2. Because it was assumed that children with clinically-elevated problems would show higher levels of early risk, one-tailed tests were used. However, due to the small sample of the clinically-elevated groups (i.e., n = 22) and the number of tests computed, these analysis should be considered exploratory, and need to be interpreted with caution. Results are presented in Tables 3A-B.
Children with clinically-elevated Aggression at preschool were more likely than peers with sub-clinical scores to have insecure, particularly disorganized, attachments at 12 months, more difficult temperaments during Year 2, greater disagreements over childrearing during Year 2, higher rates of maternal personality risk during Years 1 and 2, and higher rates of reported daily hassles at age 5. Unexpectely, children with clinically-elevated rates of Aggression were significantly more likely to come from homes of higher family income and lower rates of overcrowding at Year 2.
A similar, but weaker pattern was found for risk factors in comparing groups of children with and without clinically-elevated Externalizing problems at age 5. Children with serious Externalizing problems were more likely to have had disorganized attachments and increased maternal personality risk at 12 months, and higher rates of parental childrearing disagreements at Year 2.
__________________________________
Insert Tables 3A-B about here
__________________________________
Next, a descriptive set of analyses are presented in Table 4 to compare the predictive utility of individual pathways leading to the development of clinically-elevated disruptive behavior problems at age 5. This within-group approach permitted an examination of true-positive and false-positive cases associated with individual risk factor groups, as well as the percentage of cases correctly identified from the total pool of children with clinically-elevated scores. In order to establish cutoff scores for the continuously scaled factors of infant difficulty and childrearing disagreements, subjects rated at or above one standard deviation above the sample mean were used (see Table 1 for cutoff scores). Additionally, in order to present some context for the predictive validity of individual risk factors, the age 3 version of the same CBCL factor was employed as a predictor using the same clinical-cutoff score as at age 5 (t-score > 63). The age 3 CBCL factors were used for comparison because of the stability previously demonstrated for disruptive behavior beginning as early as age 2 (Cummings et al., 1989; Olweus, 1979).
In examining the accuracy of the pathways leading to clinically-elevated Aggression, several risk factors showed predictive validity comparable to the CBCL Aggression clinical cutoff score (t-score > 63) assessed at age 3 (true/false positive percentage = 60/40). In particular, having elevated rates of childrearing disagreements at age 2 (true/false positive percentage = 67/33) or disorganized attachment classification at age 1 (true/false positive percentage = 60/40) placed children at similar or greater risk for clinically-elevated Aggression at age 5. In addition, the childrearing disagreement risk factor identified a greater percentage of the total cases with clinical Aggression scores than previous Aggression, 36 vs. 33%, respectively.
The same descriptive data are presented in Table 4 for risk factors that discriminated clinically-elevated CBCL Externalizing at age 5, with the age 3 Externalizing factor (cutoff t-score > 63) used for comparitive purposes. In general, risk factor groups better discriminated the prediction of clinically-elevated Aggression compared to clinically-elevated Externalizing problems. Though the identification of true positives for risk factors was between 42-50% for the three risk factor groups (i.e., Year 1 disorganized attachment, Year 2 personality risk and childrearing disagreements), with the successful identification of between 23-29% of the total cases, this paled in comparison to the 70% true positive rate and the 47% of total cases identified using groups defined based on age 3 Externalizing problems.
__________________________________
Insert Table 4 about here
__________________________________
Finally, a descriptive analysis was computed to examine pathways from ages 1 and 2 that were associated with clinically-elevated disruptive behavior problems at age 5. Our goal was to identify pathways that were mutually exclusive to the one taken by infants with D attachment in Year 1 and infant difficulty in Year 2; thus because of the larger sample size of children left with clinically-elevated Aggression versus Externalizing scores (16 vs. 11), we restricted the analysis to children who would go on to show clinically-elevated Aggression. Of these 16 non-D children, there was not a sizable subset who shared the precise same trajectory at Years 1 and Year 2. However, nine of the 16 showed an avoidant or resistant insecure attachment at 12 months, and seven of the nine went on to show one or more of the following risk factors at Year 2: infant difficulty (n = 5), maternal personality risk (n = 4), or childrearing disagreements (n = 3). Interestingly, of the four children who showed a disorganized attachment at 12 months, but failed to demonstrate clinically-elevated Aggression at age 5, none were reported to have any of these risk factors present at Year 2.
Discussion
The present study examined relations between early risk factors from child, parent, and sociodemographic domains and the development of early disruptive behavior problems. In addition, pathways leading to clinically-elevated externalizing problems were investigated. Overall, as expected, relations between risk factors from the child and parent domains were more consistently associated with age 5 disruptive behavior problems in the hypothesized direction than those from the sociodemographic domain. When continuous versions of the CBCL factors were used, risk factors during infancy that were associated with the prediction of both age 5 CBCL Aggression and Externalizing included disorganized attachment, difficult temperament, maternal personality risk, parenting hassles, and childrearing disagreements. When multivariate analyses were employed to control for the effects of multicollinearity among predictor variables, only disorganized attachment at Year 1 and infant difficulty at Year 2 provided unique variance to the prediction of age 5 Aggression. Furthermore, it was found that all six children who followed a pathway from disorganized attachment at Year 1 to perceived difficulty at Year 2 developed clinically-elevated aggressive problems at age 5. Children with other types of insecure attachments at Year 1 also appeared at heightened risk for developing later disruptive behavior problems when during the second year other child or parent risk factors were present. When multivariate techniques were explicitly applied to the issue of discriminating clinically-elevated CBCL scores, the following risk factors accounted for unique variance: insecure attachments (particularly disorganized ones), maternal personality risk, and childrearing disagreements. Finally, a descriptive analysis of individual pathways indicated that Year 1 disorganized attachment, and Year 2 maternal personality risk and childrearing disagreements placed children at a 54-67% probability of showing serious aggressive problems at age 5.
The results confirm and extend several previous investigations that have examined the relation between specific risk factors and early disruptive behavior problems. In the present study consistent support was found for the longitudinal effects of disorganized attachment status, maternal personality risk, and childrearing disagreements. The relation between disorganized attachment and preschool disruptive behavior problems replicates the findings of Lyons-Ruth and colleagues (1993), who found that 44% of those classified as disorganized insecure between 12 and 18 months showed elevated rates of hostile behavior at age 5, as compared to a rate of 9% for those classified as secure. In the present study, 60% of those categorized as disorganized at 12 months demonstrated clinically-elevated CBCL Aggression scores in contrast to 17% of secure infants.
In examining the pathways leading to age 5 disruptive behavior, results indicated that children who exhibited a disorganized attachment at Year 1 and who were perceived by mothers as difficult temperamentally at Year 2 were at particular risk for showing clinically-elevated aggression at age 5. The six children who showed disorganized status at 12 months and difficult temperament at Year 2 had mean t-scores at the 99th percentile on the CBCL Aggression factor, approximately 2 standard deviations above those who had only disorganized status or above-median difficulty, or neither. In contrast, the four D infants who demonstrated age-5 Aggression scores within the normal range followed a trajectory that did not include the perception of temperamental difficulty during the second year. These results suggest that though children with disorganized attachments are at increased risk for a pathway leading to disruptive behavior problems, this pathway needs to be potentiated by the perception of difficulty by mothers during the second year of life. The nature of this "difficulty" remains unclear. Few D infants were seen as difficult by mothers during Year 1 and the ICQ composite for all subjects showed considerable stability from Year 1 to 2 (r = .64, p < .0001). However, during the second year, maternal perceptions of temperament are based on an increased repertoire of child behaviors by virtue of the substantive gains in cognitive and physical maturity during the second year (i.e., toddling and talking, see Shaw & Bell, 1993). These increases in physical and cognitive skills may change perceptions of difficulty among a subset of infants. In support of this idea, stability of infant difficulty was found to be higher among mothers with secure (r = .74, p < .001) versus disorganized (r = .38, ns) infants at 12 months. Though these relations discriminated statistical significance in the stability of difficulty between secure and D infants, due in part to the small number of D infants, the magnitude of the difference between correlations was not significant using a Fisher's-Z test.
A similar, albeit less specific pathway towards preschool disruptive problems also was identified for a subset of infants, involving an insecure avoidant or insecure resistant attachment at 12 months and the presence of one or more child or parent risk factors during the second year. These results suggest that infants with non-D types of insecure attachments also are at risk for showing later aggressive problems when during the second year other child or parent risk factors are present. At a broader level, the findings also highlight how the cumulative impact of multiple stressors across time during the infancy period can influence the developmental trajectory of child behavior.
The results also were supportive of previous research on parental conflict and disruptive behavior problems (Block, Block, & Gjerde, 1986; Emery, 1988). Though it was not possible to test its predictive power on repeated occasions, it was one of only two parent factors to be consistently related to age 5 disruptive behavior. Among families with elevated childrearing disputes at Year 2, there was a 67% probability of clinically-elevated aggressive problems at age 5. Also in accord with the previous research (Jouriles et al., 1991), a specific index of childrearing disagreements was more consistently related to behavior problems than a global scale of marital adjustment. General marital adjustment was modestly related to continuous age 5 Externalizing problems in the present study, but the magnitude of its effects were less consistent than those found for childrearing conflicts in the prediction of CBCL Aggression.
The other parent risk factor that showed predictive value was maternal personality risk. Previous research with this sample (Shaw & Vondra, 1993) and others (Egeland & Farber, 1984) has shown the same personality factors to differentiate infant attachment security, and CBCL Externalizing problems at age 3 (Shaw et al., 1994). This finding, along with the longitudinal relations found for maternal depressive symptomatology and social support are in accord with previous research which suggests that the main effects of factors that directly affect maternal functioning within the family exert more influence on child behavior problems than those whose influence is not so proximal, such as socioeconomic adversity (Richman et al., 1982). Richman and colleagues (1982) found similar results. Relationship-oriented variables (e.g., quality of the marriage, maternal criticism) within the family were the strongest predictors of age 3 behavior problems, and these relations were higher among low SES families.
An unexpected finding was the pattern of correlations across risk factor domains (see Table 2A). Though within parent and within sociodemographic domains, relations were significant approximately 50% of the time, this was not the case for relations within the child domain, nor across domains. One explanation for the results across domains may be the sample's sociodemographic characteristics. Compared to other studies examining the early antecedents of behavior problems (Richman et al., 1982; Campbell, 1994), SES was relatively homogeneous. By selecting a low-income sample, a ceiling was placed on the variability of socioeconomic factors. This may partially explain why relatively few predictor or outcome factors were related to socioeconomic risk factors: 6 of 64 relations with parent and child risk factors were significant (9%), and 1 of 18 relations with CBCL factors were significant in the predicted direction (5%). On the other hand, 47% of correlations among parent and child risk factors and CBCL factors were significant. The modest number of relations found between child and parent domains was not totally unexpected. For instance, findings from other high-risk samples exploring the relation between infant attachment security and unipolar maternal depression have been inconsistent. In some cases, no effects between secure and all insecure groups are found (Radke-Yarrow, Cummings, Kuczynski, & Chapman, 1985). While in others, differences are evident only between specific insecure groups (i.e., anxious-ambivalent) and secures (Lyons-Ruth, Zoll, Connell, Grunebaum, & Grunebaum, 1990). Readers are referred to Cohn and Campbell (1992) for more discussion about this issue.
Two limitations of the study need to be addressed. First, a response bias may partially account for the pattern of results. With the exception of the assessment of infant attachment security, maternal report was the sole source of data. Though there is the possibility that mothers who report higher rates of maternal personality risk (i.e., high aggression and defendence, low social desirability), childrearing disagreements, low social support, and difficult temperament would report higher rates of disruptive behavior problems 3-4 years later, the pattern of results is not generally consistent with such an explanation. Also, if a response bias were operating, it is unclear why similar relations were not found between maternal depressive symptoms and disruptive behavior. Maternal depression consistently has been found to inflate relations with child disruptive behavior (Fergusson, Lynskey, & Horwood, 1993), but relations with disruptive behavior were inconsistent and modest at best. Second, maternal personality risk was the only risk for which significant associations across time and CBCL factors were found, though the stability of the BDI and the PRF were comparable over time in the present study.
A second limiting factor was the sample size. The sample was small by epidemiological standards, particularly when examining the role of sex. Investigators have noted that boys demonstrate greater vulnerability to the effects of psychosocial adversity at early ages than girls (Keenan & Shaw, in press), just as boys are at greater risk for neurodevelopmental disorders such as autism, learning disabilities, and attention-deficit hyperactivity disorder (American Psychiatric Association, 1987). Sex effects were not found in the present study, though it is possible that with a larger sample, such effects may surface. Similar caution should be noted for the findings regarding the relations between disorganized attachment status and later aggressive behavior. Though our work replicates and extends the findings of others (Lyons-Ruth et al., 1993), the results are based only on 10 subjects.
In summary, the results of the present study provide important information on independent and interactive pathways leading to the emergence of age 5 disruptive behavior. In regard to the issue of equifinality, the results also provide data on alternative pathways leading to early disruptive behavior problems. Children on trajectories towards serious externalizing problems are likely to have insecure, particularly disorganized, attachments in the first year. However, during the second year, for infants previously identified as insecure, multiple pathways were evident that were later associated with clinically-elevated disruptive problems at preschool.
References
Achenbach, T.M. (1992). Manual for the Child Behavior Checklist 2/3 and 1992 profile. Burlington, VT, University of Vermont Department of Psychiatry.
Achenbach, T.M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991 profiles. Burlington, VT: University of Vermont Department of Psychiatry.
Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns of attachment. Hillside, N.J.: Lawrence Erlbaum.
Ainsworth, M.D.S., & Wittig, D. (1969). Attachment and exploratory behavior of one-year-olds in a strange situation. In B.M. Foss (Ed.), Determinants of infant behavior (Vol. 4). London: Metheun.
American Psychiatric Association (1987). Diagnostic and Statistical Manual (3rd Edition-Revised). Washington DC.
Bates, J.E. (1980). The concept of difficult temperament. Merrill-Palmer Quarterly, 26, 299-319.
Bates, J.E., Freeland, C.A. & Lounsbury, M.L. (1979). Measurement of infant difficultness. Child Development, 50, 794-803.
Bates, J., Maslin, C., & Frankel, K. (1985). Attachment security, mother-child interaction, and temperament as predictors of behavior-problem ratings at age 3 years. In I. Bretherington and E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50, Nos. 1-2, 167-193.
Beck, A.T., & Beamesderfer, A. (1974). Assessment of depression: The Depression Inventory. In P. Pichot (Ed.), Psychological measurement in psychopharmacology: Modern problems in pharmacopsychiatry (Vol. 7). Basel, Switzerland: Kanger.
Beck, A.T., Ward, C.H., Mendelon, M., Mock, J.E., & Erbaugh, J.K. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561-571.
Block, J.H., Block, J., & Gjerde, P.F. (1986). The personality of children prior to divorce: A prospective study. Child Development, 57, 827-840.
Campbell, S.B. (1995). Behavior problems in preschool children: A review of recent research. Journal of Child Psychology and Psychiatry, 36, 113-149.
Campbell, S.B. (1994). Hard-to-manage preschool boys: Externalizing behavior, social competence, and family context at two-year-follow-up. Journal of Abnormal Child Psychology, 22, 147-166.
Campbell, S., March, C., Pierce, E., Ewing, L, & Szumowski, E. (1991). Hard-to-manage preschool boys: Family context and the stability of externalizing behavior. Journal of Abnormal Child Psychology, 19, 301-310.
Christopoulos, C., Cohn, D.A., Shaw, D.S., Joyce, S., Sullivan-Hanson, J., Kraft, S., & Emery, R.E. (1987). Children of abused women: I. Adjustment at time of shelter residence. Journal of Marriage and the Family, 49, 611-619.
Cohn, J.F., & Campbell, S.B. (1992). Influence of maternal depression on infant affect regulation. In D. Cicchetti and S. Toth (Eds.), Rochester Symposium on Developmental Psychopathology, Vol. 4 (pp. 105-130). Hillsdale, NJ: Erlbaum.
Compas, B., Howell, D., Phares, V., Williams, R, & Giunta, C. (1989). Risk factors for emotional/behavioral problems in young adolescents: A prospective analysis of adolescent and parental stress and symptoms. Journal of Consulting and Clinical Psychology, 57, 732-740.
Crnic, K.A., & Greenberg, M.T. (1990). Minor parenting stresses with young children. Child Development, 61, 1628-1637.
Crnic, K.A., Greenberg, M.T., Ragozin, A.S., Robinson, N.M., & Basham, R.B. (1983). Effects of stress and social support on mothers and premature and full-term infants. Child Development, 57, 209-217.
Cummings, E.M., Iannotti, R. J., & Zahn-Waxler, C., (1989). Aggression between peers in early childhood: Individual continuity and developmental change. Child Development, 72, 887-895.
Earls, F., & Jung, K.G. (1987). Temperament and home environment characteristics as causal factors in the early development of childhood psychopathology. Journal of the American Academy of Child Psychiatry, 26, 491-498.
Edelbrock, C., & Costello, A. J. (1988). Convergence between statistically derived behavior problem syndromes and child psychiatric diagnoses. Journal of Abnormal Child Psychology, 16, 219-231.
Egeland, B. & Farber, E. (1984). Infant-mother attachment: Factors related to its development and changes over time. Child Development, 55, 753-771.
Emery, R.E. (1988). Marriage, divorce, and children's adjustment. Beverly Hills, CA: Sage.
Emery, R.E., & O'Leary, K.D. (1982). Children's perceptions of marital discord and behavior problems of boys and girls. Journal of Abnormal Child Psychology, 10, 11-24.
Erickson, M., Sroufe, L.A., & Egeland, B. (1985). The relationship between quality of attachment and behavior problems in preschool in a high risk sample. In I. Bretherington and E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50, Nos. 1-2, 147-166.
Fagot, B. (1984). The consequences of problem behaviors in
toddler children. Journal of Abnormal Child
Psychology, 12, 385-395.
Fergusson, D.M., Lynskey, M.T., & Horwood, L.J. (1993). The effect of maternal
depression on maternal ratings of child behavior. Journal
of Abnormal Child Psychology, 21,
245-270.
Gardner, F. (1989). Inconsistent parenting: Is there evidence for a link with children's conduct problems? Journal of Abnormal Child Psychology, 17, 223-233.
Graham, P., Rutter, M., & George, S. (1973). Temperamental characteristics as predictors of behavior disorders in children. American Journal of Orthopsychiatry, 43, 328-339.
Hollingshead, A.B. (1975). Four factor index of social status. New Haven, CT: Department of Sociology, Yale University.
Jackson, D.H. (1989). Personality research form manual (3rd Edition). New York: Research Psychologists Press.
Jackson, D.H. & Morf, M. (1973). An empirical evaluation of factor reliability. Multivariate Behavioral Research, 8, 439-459.
Jouriles, E.N., Murphy, C. M., Farris, A.M., Smith, D. A., Richters, J.E., & Waters, E. (1991). Marital adjustment, parental disagreements about child rearing, and behavior problems in boys: Increasing the specificity of the marital assessment. Child Development, 62, 1424-1433.
Kagan, J. (1984). The nature of the child. New York: Basic Books.
Kazdin, A.K. (1995). Conduct disorders in childhood and adolescence. Thousand Oaks, CA: Sage.
Keenan, K., & Shaw, D.S. (in press). Developmental and social influences on young girls' behavioral and emotional problems. Psychological Bulletin.
Locke, H.J., & Wallace, K.M. (1959). Short
marital-adjustment and prediction tests: Their reliability and validity. Marriage and Family Living, 21, 251-255.
Loeber, R. (1982). The stability of antisocial and delinquent childhood behavior. Child Development, 53, 1431-1446.
Loeber, R., & Dishion, T.J. (1983). Early predictors of male delinquency: A review. Psychological Bulletin, 94, 68-99.
Lyons-Ruth, K., Alpern, L., & Repacholi, B. (1993). Disorganized infant attachment classification and maternal psychosocial problems as predictors of hostile-aggressive behavior in the preschool classroom. Child Development, 64, 572-585.
Lyons-Ruth, K., Easterbrooks, M.A., Davidson Cibelli, C.E., & Bronfman, E. (1995). Predicting school-age externalizing symptoms from infancy: Contributions of disorganized attachment strategies and mild mental lag. Presented at the biennial convention of the Society for Research in Child Development, Indianapolis, IN.
Lyons-Ruth, K., Connell, D.B., Grunebaum, H., & Botein, S. (1990). Infants at social risk: Maternal depression and family support services as mediators of infant development and security of attachment. Child Development, 61, 85-98.
Main, M., & Solomon, J. (1987). Discovery of an insecure disorganized/disoriented attachment pattern: Patterns, findings and implications for the classification of behavior. In M. Yogman & T.B. Brazelton (Eds.), Affective development in infancy (pp. 95-124). Norwood, NJ: Ablex.
Martin, J. (1981). A longitudinal study of the consequences of early mother-infant interaction: A microanalytic approach. Monographs of the Society for Research in Child Development, 46.
Mash, E., & Johnston, C. (1983). Parental perceptions of child behavior problems, parenting self-esteem, and mothers' reported stress in younger and older hyperactive and normal children.Journal of Consulting and Clinical Psychology, 51, 86-99.
Matas, L., Arend, R.A. & Sroufe, L.A. (1978). Continuity of adaptation in the second year of life: The relationship between quality of attachment and later competence. Child Development,49, 547-556.
Maziade, M., Cote, R., Bernier, H., Boutin, P., Thivierge, J. (1989). Significance of extreme temperament in infancy for clinical status in pre-school years. British Journal of Psychiatry, 154, 544-551.
Olweus, D. (1979). Stability of aggressive patterns in males: A review. Psychological Bulletin,86, 852-875.
Pianta, R.C., Egeland, B., & Sroufe, L.A. (1990). Maternal stress and children's development: Prediction of school outcomes and identification of protective factors. In J. Rolf, A. Masten, D. Cicchetti, K. Nuechterlein, & S. Weintraub (Eds.), Risk and protective factors in the development of psychopathology (pp. 215-235). New York: Cambridge University Press.
Radke-Yarrow, M., Cummings, E.M., Kuczynski, L, & Chapman, M. (1985). Patterns of attachment in two- and three-year-olds in normal families and families with parental depression. Child Development, 56, 884-893.
Reynolds, W.M. & Gould, J.W. (1981). Marital psychometric investigation of the standard and short form Beck Depression Inventory. Journal of Consulting and Clinical Psychology, 49, 63-71.
Richman, M., Stevenson, J., & Graham, P.J. (1982). Preschool to school: A behavioral study. London: Academic Press.
Richters, J.E., & Cicchetti, D. (1993). Mark Twain meets DSM-III-R: Conduct disorder, development, and the concept of harmful dysfunction. Development and Psychopathology, 5, 5-30.
Robins, L.N. (1966). Deviant children grow up. Baltimore, MD: Williams & Wilkins.
Robins, L., West, P., & Herjanic, B., (1975). Arrests and delinquency in two generations: A study of black urban families and their children. Journal of Child Psychology and Psychiatry, 16, 125-140.
Rosenbaum, A., & O'Leary, K.D. (1981). Marital violence: Characteristics of abusive couples. Journal of Consulting and Clinical Psychology, 12, 328-329.
Rutter, M. (1981). Stress, coping, and development: Some issues and some questions. Journal of Child Psychology and Psychiatry, 22, 323-356.
Sanson, A., Oberklaid, F., Pedlow, R., & Prior, M. (1991). Risk indicators: Assessment of infancy predictors of pre-school behavioural maladjustment. Journal of Child Psychology and Psychiatry, 32, 609-626.
Shaw, D.S., & Bell, R.Q. (1993). Developmental theories of parental contributors to antisocial behavior. Journal of Abnormal Child Psychology, 21, 493-518.
Shaw, D.S., Keenan, K., Owens, E., Winslow, E.B., Hood, N., & Garcia, M. (1995). Developmental Precursors of Externalizing Behavior Among Two Samples of Low-Income Families: Ages 1 to 5. Presented at the biennial meeting of the Society for Research in Child Development, Indianapolis, IN.
Shaw, D.S., & Vondra, J.I. (1995). Infant attachment security and maternal predictors of early behavior problems: A longitudinal study of low-income families. Journal of Abnormal Child Psychology, 23, 335-357.
Shaw, D.S., & Vondra, J.I. (1993). Chronic family
adversity and infant attachment security. Journal
of Child Psychology and Psychiatry, 34,
1205-1217.
Shaw, D.S., Vondra, J.I., Dowdell Hommerding, K., Keenan, & Dunn, M. (1994). Chronic
family adversity and early child behavior problems: A longitudinal study of low income
families. Journal of Child Psychology and
Psychiatry 35, 1109-1122.
Sroufe, L.A. (1983). Infant-caregiver attachment and patterns of adaptation in preschool. In M. Perlmutter (Ed.), Development and policy concerning children with special needs: Minnesota symposium on child psychology (Vol. 16). Hillsdale, NJ: Erlbaum.
Thomas, A., Chess, S., & Birch, H.H. (1968). Temperament and behavior disorders in children. New York: New York University Press.
Tremblay, R.E., Pihl, R.O., Vitaro, F., & Dobkin, P.L. (1994). Predicting early onset of male antisocial behavior from preschool behavior. Archives of General Psychiatry, 51, 732-739.
Webster-Stratton, C. (1990). Long-term follow-up of families with young conduct problem children: From preschool to grade school. Journal of Clinical Child Psychology, 19, 144-149.
Zahn-Waxler, C., Iannotti, R.J., Cummings, E. M., & Denham, S. (1990). Antecedents of problem behaviors in children of depressed mothers. Development and Psychopathology, 2, 271-292.
Zahn-Waxler, C., Mayfield, A., Radke-Yarrow, M., McKnew, D., Cytryn, L, & Davenport, Y. (1988). A follow-up investigation of offspring of bipolar parents. American Journal of Psychiatry,145, 506-509.
Author Notes
This study was supported by the following organizations within the University of Pittsburgh: the Mental Health Clinical Research Center for Affective Disorders, the Office of Child Development, the Central Research Development Fund, the School of Education in conjunction with the Buhl Foundation, and the Faculty of Arts and Sciences. Collection of age 5 data were supported by Grant No. 34528 from the National Institute of Mental Health to Daniel S. Shaw. The age 5 data were collected by Dr. Kate Keenan as part of her doctoral dissertation. Dr. Keenan is now at the Department of Psychiatry, University of Chicago. Requests for reprints should be sent to the first author at the following address: Department of Psychology, Clinical Psychology Center, 604 Old Engineering Hall, 4015 O'Hara Street, University of Pittsburgh, Pittsburgh, PA, 15260. We would like to extend our appreciation to study participants for letting us watch and learn about family development.
Figure 1.
Relations between attachment security classification at 12 months and clinically-elevated
CBCL Aggression at age 5.
Figure 2. Interaction between Year 1 disorganized attachment security and Year 2 infant difficulty in predicting CBCL Aggression at age 5.