Sleep Disturbance Phenotypes and Mental Health in Child Welfare–Involved Adolescents

Investigators used a latent class approach to characterize patterns of sleep disturbance among adolescents involved with the child welfare system, then compared mental health symptom profiles cross-sectionally and again about 18 months later. In adjusted comparisons, the subgroup defined by difficulty maintaining sleep was described as having the highest overall symptom burden, with higher scores across multiple domains relative to adolescents without reported sleep disturbance. The analysis is framed around heterogeneity in sleep complaints rather than a single summed sleep score within an adversity-exposed cohort. Findings are presented as subgroup-linked differences spanning internalizing symptoms, externalizing behaviors, and post-traumatic stress symptoms across two measurement points.

Using the second cohort of the National Survey on Child and Adolescent Well-Being (NSCAW-II), the authors analyzed 1,041 youth aged 11–17 years at baseline in Sleep Disturbance Phenotypes and Mental Health in Child Welfare–Involved Adolescents. Sleep disturbance indicators came from five Youth Self-Report (YSR) items—nightmares, feeling overtired without reason, sleeping less than other kids, sleeping more than other kids, and trouble sleeping—collapsed from the original response scale into dichotomous endorsed/not endorsed variables for latent class analysis. A three-class solution was retained and labeled based on item-response probability patterns: a “no sleep disturbances” subgroup (38%) with low endorsement probabilities across items, a “sleeping more than peers and overtired” subgroup (16%) with higher probability of sleeping more and feeling overtired, and a “trouble maintaining sleep” subgroup (47%) with higher probabilities across multiple sleep-problem items. Class membership was then carried forward to compare adjusted, normed symptom t-scores at baseline and follow-up.

At baseline, the authors report higher adjusted mean t-scores across internalizing, externalizing, and PTSD symptoms in the trouble maintaining sleep subgroup than in the no sleep disturbances subgroup, with pairwise differences summarized alongside standardized effects. For example, adjusted internalizing scores were reported as higher in the trouble maintaining sleep subgroup than the no-disturbance subgroup (62.19 vs 44.13; d = 1.58), and adjusted PTSD scores were similarly higher (61.52 vs 48.35; d = 1.22). The results presentation also notes that additional pairwise contrasts and effect sizes were examined across all three classes for each symptom domain. Overall, baseline symptom differences were described as aligning most consistently with the trouble maintaining sleep phenotype in the adjusted comparisons.

At the approximately 18-month follow-up, the trouble maintaining sleep subgroup was again reported to have higher internalizing, externalizing, and PTSD symptoms than the no sleep disturbances subgroup in adjusted models. In comparisons distinguishing the two symptomatic sleep classes, the trouble maintaining sleep subgroup was reported to have higher PTSD at baseline than the sleeping-more/overtired subgroup, and higher externalizing and PTSD at follow-up. The adjusted models included cumulative childhood adversity and sociodemographic and placement-related covariates, with the authors specifying adjustment for age, gender, race/ethnicity, and placement status (out-of-home care vs living with biological/adoptive parents). The authors also note constraints, including reliance on a small set of self-reported YSR sleep items (dichotomized for analysis) and a two-wave design, which they describe as limiting causal inference about sleep disturbance patterns and symptom change. Taken together, the reported sleep-disturbance phenotypes tracked with differential symptom profiles across two measurement points in this child welfare–involved adolescent cohort.

Key Takeaways:

• The authors report three latent sleep-disturbance subgroups derived from five dichotomized YSR sleep items in a nationally representative child welfare–involved adolescent sample.
• Higher adjusted symptom scores were observed at baseline across internalizing, externalizing, and PTSD measures for the trouble maintaining sleep subgroup compared with the no-disturbance subgroup.
• At follow-up, the trouble maintaining sleep subgroup continued to show higher symptom levels than the no-disturbance subgroup, and the authors report additional contrasts versus the sleeping-more/overtired subgroup for PTSD and externalizing symptoms.