Adolescents Growing In Contexts of Intimate Partner Violence and Housing Insecurity: A Repeated Measures Latent Class Analysis of Predictors and Behavioral Problems

Background and Purpose: Intimate partner violence (IPV) and housing insecurity indicate an unsafe and unstable family environment, leading to child behavioral problems. However, investigation of their chronic co-occurrence and impacts on children’s behavioral problems over time is limited. This study employed a repeated measures latent class analysis (RMLCA) to explore the longitudinal patterns of children’s exposure to IPV (eIPV) and housing insecurity from infancy to school age, potential demographic predictors of class membership, and the relationship between the identified patterns and behavioral problems in adolescence.

Methods: Using data of 1,778 mother-child dyads from the Future of Families and Child Wellbeing Study, we examined the longitudinal patterns of eIPV and housing insecurity. Children’s eIPV was measured based on mothers’ self-reported IPV experiences when children were ages 1, 3, 5, and 9, including physical, psychological, sexual IPV and other controlling behavior. Housing insecurity was assessed by mothers’ self-reports on six items related to unstable housing across the same waves. A shortened version of the Child Behavior Checklist was used to measure adolescents’ behavioral problems at age 15, including internalizing and externalizing problems, anxious/depressive symptoms, withdrawn symptoms, aggressive behavior, and rule-breaking behavior (alphas = .51–.88). RMLCA was employed to explore the patterns of eIPV and housing insecurity in childhood. Then, multinomial logistic regression was conducted to examine the predictors of class membership. Lastly, adolescents’ behavioral outcomes were compared across classes.

Results: The final model includes 4 classes: Class 1, which consisted of close to 30% of the sample, with low probabilities of experiencing both eIPV or housing insecurity in childhood. One-quarter of the sample was in Class 2, insecure housing throughout childhood, with high probabilities of experiencing housing insecurity from infancy to school age. Class 3, eIPV throughout childhood, accounted for 28.8% of the sample, showing high probabilities of eIPV from infancy to preschool age and a moderate probability at school age. The last class, co-occurring eIPV and housing insecurity throughout childhood, had high probabilities of experiencing both eIPV and housing insecurity throughout childhood, consisting of 17.3% of the sample. Mothers’ age at baseline, race/ethnicity, and highest level of education were significant predictors of children’s class membership. Adolescents in Class 1 reported the lowest level of behavioral problems compared to the other three classes. Adolescents in Class 4, co-occurring eIPV and housing insecurity throughout childhood, had worse internalizing problems than Classes 2 and 3.

Conclusions and Implications: This study innovatively used a person-centered approach to identify children’s longitudinal patterns of eIPV and housing insecurity from infancy to school age, along with related behavioral problems. Adolescents with a high probability of either eIPV or housing insecurity reported worse behavioral problems than low-risk adolescents. Adolescents with a high probability of co-occurring eIPV and housing insecurity had worse internalizing problems than those with high probability of either eIPV or housing insecurity alone. Policy and practice should consider the close relationship between eIPV and housing insecurity, develop and utilize screening tools to identify at-risk families, and provide early interventions to protect children from later detrimental outcomes.