BackHome
Refresh: OffView Personal Schedule
Individual Submission Summary
Share...

Direct link:

Multi-Level Analysis of Foster Care as an Alternative to Institutionalization on Biobehavioral Outcomes Across Childhood

Fri, April 9, 4:20 to 5:50pm EDT (4:20 to 5:50pm EDT), Virtual

Abstract

Background: Previous analyses of the Bucharest Early Intervention Project (BEIP; Zeanah et al., 2003) indicate that children randomized to foster care, on average, fare better than those assigned to the care as usual condition (Zeanah et al., 2017). While individual time points and domains have been examined, the overall effect of the foster care intervention on children’s outcomes across domains and developmental periods has not been quantified and would allow a better understanding of the magnitude, breadth, and longevity of the intervention effects. In the current study, we use a multi-level analysis of six waves of follow-up data to identify the cross-developmental effects of the foster care intervention on psychopathology, IQ, physical growth, and EEG relative alpha power.

Method: Children (N=136, 51% female, 6-31 months) residing in institutional care in Bucharest, Romania were randomly assigned to the foster care intervention (foster care group [FCG]; n=68) or the care as u¬¬sual group (CAUG; n=68). Assessments were conducted at ages 30, 40, and 54 months (when the intervention formally ended), and 8, 12, and ~16 years. At each wave, caregivers and/or teachers completed assessments of children’s symptoms of psychopathology. Research staff administered IQ assessments, measured physical growth, and collected resting EEG. We conducted multi-level models testing: (1) the main effect of the intervention on children’s standardized outcomes (3,330 observations) across the domains of IQ, physical growth, and EEG relative alpha-power and (2) the main effect of the foster care intervention (using a conservative intent-to-treat design) on children’s standardized scores across domains of psychopathology (4,421 observations). We examined outcome domain, wave, and biological sex as moderators.

Results: We found an overall effect of the foster care intervention on IQ scores, alpha power, and physical growth when considered jointly across assessment waves (beta=0.11[0.02,0.21], p=.018; Figure 1). The strength of this effect was significantly moderated by outcome domain (F(2,3202.10)=7.65, p<.001), but not by wave or sex. There was a medium and significant effect of the intervention on IQ (beta=0.31[0.17,0.44]) but smaller non-significant effects on alpha power (beta=0.05[-0.07,0.17]) and physical growth (beta=0.08[-0.03,0.18]). Second, there was a significant overall effect of the intervention on children’s symptoms across dimensions of psychopathology and across assessment waves (beta= -0.25[-0.41,-0.08], p=.004). This effect was moderated by the domain psychopathology (F(4,4184.00)=12.38, p<.001), but not by wave or sex. There were significant and medium to large effects of the intervention on disturbed relatedness (reactive attachment disorder, beta=-0.68[-0.90,-0.47]); disinhibited social engagement disorder, beta=-0.33[-0.55,-0.10]; Figure 2), but small and non-significant effects on symptoms of inattention/hyperactivity (beta=-0.08[-0.27, 0.12]), externalizing (beta=-0.17[-0.34,0.02]), and internalizing (beta=-0.14[-0.36,0.07]).

Conclusion: The BEIP has long-term benefits for children’s IQ scores and disturbed relatedness. Specifically, throughout toddlerhood to late adolescence, children who were randomized to foster care evidence consistently higher IQs and lower symptoms of disturbed relatedness than do children in care as usual. Although children randomized to foster care show higher average EEG relative alpha-power and physical growth, and lower psychopathology in other domains than do care as usual children, using this conservative analytic approach effects in these domains were not statistically significant.

Authors