Avoidance and Resistance in in-Care and at-Risk Children, a Protective Effect

Background: A strong literature on early attachment indicates that the adaptive strategies that young children develop to maintain proximity with inconsistent or rejecting caregivers (i.e. resistant and avoidant attachment) can put them at risk for later mental health difficulties. A relatively small portion of this research base has studied in-care preschoolers. The current study aims to understand how in-care and at-risk preschoolers’ attachment behaviors towards their caregivers predict later mental health outcomes. 

Methods:Participants were the primary caregivers of 120 children (ages 3-6; 60 in foster care and 60 SES-matched community children). Caregivers completed the Parent Attachment Diary (Stovall-McClough & Dozier, 2000) to measure children’s attachment related behaviors. Behaviors were categorized as proximity-seeking, resistant, or avoidant. Two years later, caregivers completed the Parent-Rated Child Behavioral Checklist (PCBC; Achenbach, 1991) measuring problematic child behaviors.   

Results: A preliminary series of multiple linear regression analyses were conducted to determine how foster care versus community status and attachment behaviors predicted problematic behaviors two years later. An initial model containing main effects of foster care status, avoidant behaviors, and resistant behaviors found that both care status B=.42, t=3.91, p<.001 and resistant behaviors B=-.22, t=-2.08, p=.04 significantly predicted delinquency such that being in foster care was related to more delinquency, and resistant behaviors was related to less delinquency. In a second step, both two-way interactions were added.  Avoidance by foster status was not significant (p=.48), though resistance by foster status trended toward significance (p=.07) such that resistant behaviors positively predicted delinquency in community children but not in foster children. The final full model predicted a significant 17.0% of the variance in later delinquent behaviors, F(5,84)=4.44, p<.001.

The same model building procedure was repeated for aggressive behavior. The full model was significant (p<.001) and explained 28.6% of the variance in outcome., In a first step with main effects, foster care status emerged as a significant predictor (B=.57, t=5.65, p<.001) and resistant behaviors emerged as a marginally significant predictor (B=-.19, t=-1.95, p=.06) of aggression, both in the same direction as in the first model. When two-way interactions were added, foster status by resistance emerged as marginally significant (B=.33, t=1.76, p=.08) such that resistant behaviors positively predicted aggression in community children but not in foster children.

While these models offer some insight into the ways in which these attachment behaviors might function across the two groups and over time, a pending Latent Profile Analysis (LPA) will further elucidate these dynamics by revealing (a) how children cluster according to their utilization of different resistant and avoidant behaviors, and (b) how these profiles go on to predict mental health outcomes across the two groups of children.  Results of this pending LPA will be also be reported.

Implications: Our results to date suggest that resistant behaviors towards caregivers in high-risk preschoolers may have a protective role and be predictive of positive outcomes. For at-risk children, and particularly children in foster care, resistant behavior or fussiness towards a caregiver may help them gain caregiver attention and buffer against the effects of environmental stress.