Differences in Internalizing and Externalizing Symptoms Based on Latent Classes of Sleep Disturbances Among Child Welfare-Involved Adolescents

Background and Purpose: Sleep health is an important aspect of adolescent health and development. Sleep disturbances (e.g., difficulty falling asleep, night wakings, less sleep efficiency) can contribute to increased risk for psychopathology. Sleep disturbances and childhood adversity share similar neurobiological pathways, and thus it is important to understand sleep disturbances among youth exposed to adversity. Child welfare-involved youth is one population in particular who is at increased risk for exposure to childhood adversity and subsequent sleep disturbances and psychopathology. Among maltreated youth, estimates suggest that 10-54% of youth experience sleep disturbances and as many as 49% of child welfare-involved youth have at least one psychiatric diagnosis. Adolescence is a developmental period characterized by normative sleep disturbances; however, for maltreated adolescents, disrupted sleep is hypothesized to be a “second hit” that increases risk for psychopathology. It is thus important to better understand sleep disturbances and their associations with psychopathology among maltreated adolescents. The current study had two main aims: (1) to identify subgroups of youth based on patterns of sleep disturbances and (2) to examine differences in internalizing and externalizing symptoms based on subgroup membership.

Methods: The current study used data from the second cohort of the National Survey on Child and Adolescent Wellbeing (NSCAW), a nationally representative survey of youth in the U.S. child welfare system. The analytic sample included 1,004 youth ages 11-17 years (M = 13.62 years, SD = 1.87; 58.5% female; 55.4% from minoritized racial/ethnic backgrounds). We conducted latent class analysis to identify patterns of sleep disturbances and used the BCH approach to include internalizing and externalizing symptoms as distal outcomes, adjusting for covariates (age, gender, out-of-home care placement, race/ethnicity, childhood adversity). Standard fit criteria were used to select the best fitting latent class model.

Results: We identified three subgroups of sleep disturbances based on their probability of endorsing items: high nightmares only (Class 1, 38%), high sleep more than most/low sleep less than most (Class 2, 16%), and multiple sleep disturbances (Class 3, 47%). The multiple sleep disturbances subgroup had high probability of endorsing nightmares, being tired, sleeping less, and having trouble sleeping. Adjusting for covariates, only the multiple sleep disturbances subgroup had internalizing and externalizing scores that met borderline clinical thresholds. The multiple sleep disturbances internalizing (66.63 vs. 41.57) and externalizing scores (64.67 vs. 50.05) were higher than the high nightmares only subgroup. No other pairwise comparisons were significant.

Conclusions and Implications: The findings indicate that almost half of adolescents involved with the child welfare system (47%) experienced multiple sleep disturbances and were at increased risk for borderline clinically significant psychopathology. However, our study relied on cross-sectional data, so the causal connection between sleep disturbances and psychopathology cannot be established. It is important that future research using longitudinal methods examine how sleep disturbances contribute to psychopathology (and how psychopathology contributes to sleep disturbances) over time. Overall, our findings emphasize the role of sleep disturbances in psychopathology and the need for intervention efforts focused on promoting sleep health for maltreated adolescents.