Abstract
<title>Abstract</title> <p>Children in alternative care are formally protected by the state, yet protection does not always provide stable relationships for selective attachment. Drawing on attachment theory, the bioecological model, and differential susceptibility theory, this three-wave, multi-source study examined whether caregiver continuity and caregiver sensitivity explained variation in reactive attachment disorder (RAD) and disinhibited social engagement disorder (DSED) symptoms among children in alternative care in China. The baseline sample included 707 children aged 24 to 72 months from institutional care, family-like group care, family foster care, kinship care, and demographically comparable low-SES community families; 588 children had T3 outcome data. Care type and pre-placement history were assessed at T1, caregiving mechanisms at T2, and RAD/DSED symptoms and socioemotional adjustment at T3. Compared with institutional care, foster care and kinship care predicted lower adjusted RAD and DSED symptoms, whereas family-like group care did not differ significantly after covariate adjustment. Continuity and sensitivity served as partial indirect pathways, especially among children with high negative emotionality or low effortful control. Findings support a relational-translation account: formal protection becomes developmentally meaningful when placements provide repeated, sensitive, child-specific caregiving relationships.</p>