Looking Back and Forward: Identifying Foster Care Placement History Classes and Assessing Later Mental Health

Background and Purpose: With foster care (FC) administrative records becoming more available and reliable, researchers have begun to explore the association between FC placement characteristics and subsequent mental health problems (E.g., Rubin et al., 2008; Aarons et al., 2010).  These samples consisted of mostly infants and children, and assessed mental health not long after FC entry.  Thus, little is known about associations between placement history and mental health outcomes for older FC youth, considering their full and sometimes lengthy FC involvement.  This study identifies latent classes based on FC placement characteristics and examines associations between the classes and subsequent mental health.

Methods: FC placement records were obtained from the California Department of Social Services and linked to data from Wave 1 of the CalYOUTH study (n=714).  CalYOUTH participants were screened for mental health conditions using the Mini-KID.  A multivariate latent mixture model was used to identify latent classes based on the following FC characteristics: age of entry, removal reason, total years in FC, rate of placement change, modal placement type, and whether youth exited and reentered FC.  Model fit criteria (e.g., Log Likelihood, AIC, BIC) in combination with substantive interpretation were used to select the class solution.  Since a four-class solution had a very high classification rate (entropy=.927), most likely class assignments were used instead of posterior probabilities in regression analyses.  Logistic regression models examined class differences in the likelihood of mental health conditions (depression, mania, externalizing problems, PTSD, substance use, and alcohol use) and suicidality (ideation and attempts).  Covariates included several demographic variables.  Sensitivity analyses with posterior probabilities instead of likely class assignment yielded largely consistent results. 

Results: Results supported a four class solution, which consisted of early leavers (12.2%), early stayers (14.7%), middle welters (22.5%), and late movers (50.5%).  Early leavers entered care young, exited and reentered care, changed placements frequently, and were more likely to experience abuse.  Early stayers also entered care young but remained in care for over twice as long and were less likely to change placements or leave FC.  Middle welters entered halfway between infancy and adolescence, about half left and returned to care, and were average in placement stability.  This group also had high rates of physical abuse and placement in therapeutic foster care.  Late movers entered around age 15, were more likely to be sexually abused, experienced greater placement instability, and were more likely to spend most of their time in FC in group care.  Late movers were also more likely to screen positive on depression, alcohol use, and attempt suicide than some of other classes (p<.05). 

Conclusions and Implications: Many of the youth experienced mental health difficulties, and youth with different FC placement profiles did not differ significantly on most mental health outcomes.  The late movers were the exception, reporting higher rates of depression, alcohol use, and suicide attempts.  Since these youth make up half of the young people in care at 17, youth entering care in late adolescence may be particularly important to screen for mental health problems.