This study provides prospective evidence that behavioral disinhibition (BD) measured in early childhood predicted the presence, five years later, of current disruptive behavior disorders in general and oppositional defiant disorder in particular, even after considering the contribution of parental psychopathology and low socioeconomic status. Moreover, the study suggests that preschool BD represents a risk factor for the development of comorbid mood plus disruptive behavior disorder by middle childhood.
Our findings are consistent with results from the initial wave of this study demonstrating that BD was associated with disruptive behavior disorders and their comorbidity with mood disorders at a mean age of six years (Hirshfeld-Becker et al 2002
). They are also consistent with other prospective studies finding associations between approach behaviors, stimulation-seeking, or behavioral undercontrol in early childhood and later delinquency (Tremblay et al 1994
), aggression (Raine et al 1998
), inattention, hyperactivity and antisocial behaviors (Caspi et al 1995
; Caspi et al 1996
). However, this is the first study to our knowledge to examine prospective outcomes of laboratory-observed preschool BD using diagnoses derived from structured diagnostic interviews in a sample of offspring at risk for psychopathology, as well as the first to examine the risk associated with BD independently from the risk conferred by parental psychopathology.
Children with BD whose parents did not have bipolar disorder also had significantly increased risk for disruptive behavior plus mood disorders. Therefore, we theorize that BD is a marker of motivational or emotional dysregulation which might predispose a child to develop one of several disorders involving behavioral or motivational dysregulation (Nigg 2000
), including ADHD, ODD, conduct disorder, alcohol or substance use disorders, and bipolar disorder (Depue and Iacono 1989
; Iacono et al 1999
). Continued follow-up of these children as they enter adolescence, the period of greatest risk for these disorders, will clarify whether this is indeed the case.
The disinhibited children in our sample did not show significantly higher current (past month) rates of comorbid mood and disruptive behavior disorders. Because mood disorders are episodic, we would expect a lower number of children to have experienced an episode in the past month than over their lifetime. Only 7% of the disinhibited children were in a current mood episode at the time of the follow-up assessment (and all had a concurrent disruptive behavior disorder). Therefore, even though none of the comparison children had current comorbid mood and disruptive behavior disorders, the difference between disinhibited and comparison children only reached trend significance (p=.058).
It is important to underscore that although BD increased the statistical risk of developing comorbid mood and disruptive behavior disorders, the majority of disinhibited children in our sample did not develop these disorders. Therefore, BD is by no means tautological with this comorbid psychopathology. Instead, our findings suggest that BD may be a temperamental risk factor for the subsequent development of these disorders which is measurable very early in development.
An examination of the raw rates of disorder predicted by BD as assessed at different age points (age 21 months, 4 years, or 6 years) suggested that, regardless of the age at which it was assessed, BD was associated at follow-up with an increase in lifetime comorbid mood plus behavior disorder, current disruptive behavior disorder, and current ODD. This result suggests that BD in early childhood predicts adverse outcome in middle childhood regardless of the age at which it is assessed. Whereas further studies with larger samples are needed to confirm this finding, the result does imply that BD may be a marker of risk even in very young children (21-month-olds).
Our study should be viewed in light of several methodological limitations. Because we assessed BD at only a single time point, we do not know whether the children we studied remained stably disinhibited. Our current results suggest that even based on one assessment, BD in early childhood signals increased risk for future dysfunction. On the other hand, our design limited our ability to assess whether stability of BD across early childhood would increase its predictive association with disruptive behavior and comorbid mood disorders. Future studies that assess BD at two or more time points are needed to address this issue.
Additionally, because of the relatively low base rates of mood disorder in this young at-risk sample, our conclusions about the rates of comorbid mood and disruptive behavior disorders were based on a small number of individuals. Although these differences clearly attained significance, we do not have a precise estimate of the strength of the relationship (odds ratio) given that the 95% confidence interval for the odds ratio was 1.21–58.96.
This study was also limited by reliance upon parental report for lifetime diagnoses in children under age twelve. We did not directly interview younger children about their lifetime diagnoses because children under twelve are limited in their ability to understand questions from structured diagnostic interviews (Breton et al 1995
) and to map events in time and have been found to be unreliable in their reports of psychopathology (Edelbrock et al 1985
; Schwab-Stone et al 1994
). In contrast, maternal reports of psychopathology, have shown high reliability, even over a one-year period (Fallon and Schwab-Stone 1994
; Faraone et al 1995
). By relying on parent-reports for diagnoses in children 11 and under and combined parent- and child-reports in children 12 and older, we introduced method variance. However, we opted for this approach, because using only the parent reports for children over 11 might result in our missing important information about mood, conduct, and inattention symptoms among adolescents. Moreover, because these methods were used to the same degree with both the BD and non-BD groups, this method variance cannot confound our analyses. Future waves of data collection, in which all children in the sample will be old enough to be directly interviewed about their symptoms will enable us to confirm our current findings and to explore further hypotheses about the outcome of children with BD. The study was also limited by the homogeneity of the sample, which included a low representation of minority families. Further studies of more ethnically diverse samples are needed to extend the generalizability of findings.
Despite these limitations, our findings suggest that preschool-age BD appears to identify a subset of children at elevated risk to develop comorbid mood and disruptive behavior disorders by middle childhood. Clinically, these results suggest that young children who show signs of BD should be monitored closely for signs of disruptive behavior and mood disorders. They also raise the hypothesis that these children might benefit from early or preventive interventions.