Infant attachment and temperament as predictors of subsequent externalizing problems and cardiac physiology


Kim B. Burgess, Department of Human Development, 3304 Benjamin Building, University of Maryland, College Park, MD 20742, USA; Email:


Background: The primary objective of this study was to examine the extent to which both individual child temperament and parent–child relationship quality independently and/or interactively predicted physiological, psychosocial, and behavioral ‘outcomes’. Employing a longitudinal prospective design over three years, statistical associations were found among infant attachment, uninhibited temperament, and 4-year behavioral and physiological functioning that supported a bio-psychosocial model of development.

Method: Three cohorts totaling 140 children and their mothers visited the laboratory for observational assessments of attachment classification at age 14 months (Strange Situation), behavioral inhibition at 24 months, and social behaviors with unfamiliar peers at age 4 years. Cardiac measures of heart rate and respiratory sinus arrhythmia (RSA) were collected at every time point. At age 4 years maternal ratings of child temperament and behaviors were also obtained.

Results: Analyses of variance (ANOVAs) with inhibition group (high, moderate, low) and attachment classification (A, B, C) revealed that the low inhibited group had significantly higher activity level scores and displayed significantly less reticence at 4 years compared to the moderately and highly inhibited groups. Infants who had an avoidant attachment with their mothers had more externalizing problems (aggressive behaviors) at age 4 than either securely or ambivalently attached infants. This predictive relation, however, was qualified by an interaction whereby avoidant attachment and uninhibited temperament together predicted a higher incidence of externalizing behavior problems. Moreover, infants’ avoidant attachment was not concurrently but predictively associated with lower heart rate and high RSA at age 4 years. Therefore, an avoidant mother–child relationship in infancy could influence the development of an underaroused autonomic profile in early childhood.

Conclusions: Consistent with bio-psychosocial models of development, these findings support the contention that both early child temperament and parent–child relationship quality contribute to subsequent psychological/behavioral and physiological functioning.