159 - Interaction of Home and School Chaos for Adolescent to Young Adult Mental and Behavioral Health

Publication Number: 1149.159

Cecilia Wada, University of California, Los Angeles David Geffen School of Medicine, Los Angeles, CA, United States; Benjamin P. L.. Meza, University of California, Los Angeles David Geffen School of Medicine, Los Angeles, CA, United States; Van Viet Thuy Nguyen, University of California, Los Angeles David Geffen School of Medicine, San Jose, CA, United States; Yomara Mendez, UCLA DGSOM, Los Angeles, CA, United States; Kulwant Dosanjh, University of California, Los Angeles David Geffen School of Medicine, Los Angeles, CA, United States; Rebecca N. Dudovitz, University of California, Los Angeles David Geffen School of Medicine, Los Angeles, CA, United States; Mitchell D. Wong, University of California, Los Angeles David Geffen School of Medicine, Los Angeles, CA, United States; Jaime La Charite, University of California, Los Angeles David Geffen School of Medicine, Los Angeles, CA, United States

Background: Measures of home and school chaos (environmental disorder) are associated with adverse health outcomes in youth. Prior studies examined home and school chaos separately.

Objective: We examine the interaction between home and school chaos and its relationship with later mental and behavioral health in adolescents and young adults.

Design/Methods: We analyzed survey data from a prospective longitudinal cohort of youth from low-income Los Angeles neighborhoods who applied via lottery to five high-performing charter schools. They were enrolled in the study from 8th to 9th grade (2013-2014) and followed until 3 years post-high school (2022). The parent study included 1270 participants who did and did not "win" the charter school admissions lottery. Study participants attended 150 high schools and were surveyed every 1-2 years. For this study, participants were eligible if they completed measures of home chaos at baseline and school chaos in 10th grade. Exposures were self-reported home and school chaos measures (continuous). To estimate margins, chaos measures were categorized at the median into four groups: low/low, low home/high school, high home/low school, and high/high chaos. The outcomes were self-reported cannabis and alcohol misuse, delinquency, anxiety, depression, general mental health, and stress measured from 10th grade to age 20-22. Outcomes were regressed on home and school chaos, adjusting for repeated measures, clustering, demographics, and grade-point average. An interaction term for home and school chaos was determined, and post-estimation margins were estimated by chaos category.

Results: Among 1,159 students in the analytic sample (53% female, 90% Latinx), the high/high chaos group showed the worst outcomes. Compared to the low/low chaos group, the high/high chaos group had twice the rate of depression (29.6% vs 14.8%) and anxiety symptoms (16.2% vs. 8.4%), twice the cannabis misuse score (1.14 vs. 0.47), and three times the alcohol misuse score (1.42 vs. 0.43). Chaos interacted multiplicatively for anxiety (p-value=0.04), cannabis misuse (p=0.00), and alcohol misuse (p=0.00).

Conclusion(s): Youth reporting chaos in both home and school settings experience disproportionately worse mental and behavioral health outcomes, even after considering each setting independently. Reducing chaos in either setting could improve the health of young people. School chaos may be a scalable target for positive change.

Table 1: Baseline Demographic Characteristics of the Four Chaos Groups


Table 2: Mental Health Outcomes by Chaos Group
# Results are the predicted margins across the four chaos groups from the covariate-adjusted linear hierarchical regression model, including main effects and interaction term with the dichotomized home and school chaos variables.
## The covariate-adjusted linear hierarchical regression model with the main effects and interaction term using the continuous versions of the home and school chaos variables.
^Asked at baseline, F1, F2, F4 surveys
^^Asked at F5, F6, F7 surveys
*p-value < .05; **p-value < .01; ***p-value < .001


Table 3: Behavioral Outcomes by Chaos Group
# Results are the predicted margins across the four chaos groups from the covariate-adjusted linear hierarchical regression model, including main effects and interaction term with the dichotomized home and school chaos variables.
## The covariate-adjusted linear hierarchical regression model with the main effects and interaction term using the continuous versions of the home and school chaos variables.
*p-value < .05; **p-value < .01; ***p-value < .001