This study will assess early and middle childhood outcomes of an intervention for neglecting parents that was implemented in the children's infancy. We expect that parents who received the Attachment and Biobehavioral Catch-up Intervention in infancy will be more nurturing and will follow children's lead more than parents who received a control intervention, and that children will show better outcomes in attachment, inhibitory control, emotion regulation, and peer relations than children of parents who received the control intervention.

Children were randomly assigned to receive the ABC intervention or a control intervention (DEF) in infancy. These two groups, plus a group of low-risk children, will be studied in early and middle childhood. Of interest will be differences in parent and child outcomes that result from the intervention.

Hypothesis 1: Neglected children whose parents received the ABC intervention and low-risk comparison children will show better inhibitory control than neglected children whose parents received the DEF intervention.

Hypothesis 2: Children in the ABC intervention condition and low-risk comparison children will show better emotion regulation than children in the DEF condition.

Hypothesis 3: Children in the ABC intervention condition and comparison children will show less reactive aggression and less hostile attributional bias than children in the DEF condition.

Hypothesis 4: Children in the ABC condition and comparison children will show more normative cortisol production than children in the DEF condition.

Although we expect that sustained changes in parenting are critical for sustained changes in child behaviors, several alternative models will be tested. First, it is possible that when parents change as a result of the intervention in a child's infancy, there are positive outcomes for children regardless of whether the changes in parenting are sustained. If this is the case, early parenting will mediate the effects of the intervention when controlling for later parenting. Second, if concurrent parenting is what is critical to child functioning, current parenting will mediate intervention effects on child outcomes when controlling for early parenting. Third, longitudinal modeling of both parent and child behaviors allows for analysis of cross-lagged associations using structural equation modeling. Such modeling can examine concurrent and transactional associations between parent and child. We can also examine associations between change at behavioral and biological levels.

Longitudinal modeling will be used to examine models of change in parenting behaviors and how those influence child outcomes.

Inclusion Criteria:

• must have been in earlier randomized clinical trial

Exclusion Criteria:

-