Are Disinhibited Social Behaviors Among Internationally Adopted Children Mediated by the Attachment Environment or by Children's Difficulties with Inhibitory Control?

ABSTRACT

Internationally adopted children show higher rates of disinhibited social behavior (DSB). Two competing explanations for DSB include difficulties in attachment specifically and deficits in inhibitory control more generally. This four-wave longitudinal study (a) documented the persistence of DSB among internationally adopted children and (b) tested the relative contributions of attachment environment versus inhibitory control difficulties in the persistence of DSB. Improvement in DSB was found 3 to 4 years post-adoption for the sample as a whole, although heterogeneity within the sample was present. Results from structural equation modeling and another test of mediation both indicated that the association between early deprivation and DSB was mediated by general difficulties with inhibitory control, not by attachment environment.

KEYWORDS:

• Disinhibited social behavior
• indiscriminate friendliness
• international adoption
• attachment
• inhibition

Funding

This research was supported by research grants from the Calvin Alumni Association, Calvin Board of Trustees, McGregor Fellowship Program, and the Calvin College Science Division.

Notes

1. According to Rutter et al. (2007), 83 children experienced DSB at age 6 and 45 of these (54%) were still demonstrating DSB at age 11. In addition, 6 children who were symptom-free at age 6 experienced symptoms at age 11. Our computation is 51/89 = 57%.

2. According to Kreppner et al. (2010), 15 of the ERA Study children with symptoms at both ages 6 and 11 no longer had symptoms at age 15. Kreppner et al. did not account for the 6 children who were symptom-free at age 6 but demonstrated symptoms at age 11. In addition, one child seems to have fallen out of the sample, because Rutter et al. reported 45 children demonstrating DSB at age 11 but Kreppner et al. computed persistence as 29/44. Our estimation is (51 − 15) / 88 = 41%.

3. Further support for inhibitory control as the more likely mediator of deprivation and DSB was garnered from a series of post hoc analyses in which we employed attachment relationship and inhibitory control ratings at individual waves as substitutes for the averaged-across-waves variables employed as mediators in the focal analyses. In other words, we reran all three models, three times, first using Wave 1 scores as the mediators, then Wave 2 scores, and then Wave 3 scores (always predicting Wave 4 DSB). Wave 2 data yielded an excellent fit of the model to the data (χ² (2) = 1.43, p = .49; RMSEA = .00; CFI = 1.00). More important, the pattern of fit was constant across all three waves: Model 3 (inhibitory control as the only mediator) always fit better than Model 1 (both attachment relationship and inhibitory control tested as mediators), which always fit better than Model 2 (attachment relationship as the only mediator).

4. As with the post hoc SEM analyses, post hoc individual-wave analyses using the Preacher and Hayes (2008) bootstrapping method also revealed that the mediation effect was strongest at Wave 2.