Relational threats and school connectedness in adolescent poor mental health using repeated cross-sectional U.S. YRBS data and locked temporal validation

Background School relationships are associated with adolescent mental health, but it is unclear how consistently these associations and prediction-model performance reproduce across survey waves. Methods This repeated cross-sectional secondary analysis used independent 2021 and 2023 U.S. National Youth Risk Behavior Survey samples (17,232 and 20,103 students). Survey-weighted Poisson models estimated associations with past-30-day poor mental health. Absolute-scale interactions and cross-wave tests assessed bullying, sleep, school connectedness, and parental monitoring. A ridge-penalized logistic model and threshold were locked using 2021 data before evaluation in 2023. Results Weighted outcome prevalence was 29.3% and 28.5%. Bullying, cyberbullying, and sleep <8 h were associated with higher prevalence. High school connectedness was associated with lower prevalence (2021 adjusted prevalence ratio 0.65, 95% confidence interval 0.59–0.70; 2023, 0.71, 0.65–0.78); standardized prevalence differences were −12.8 and −9.7 percentage points. Bullying-by-connectedness estimates were too imprecise to resolve buffering. The connectedness cross-wave test had p = 0.052 (Holm-adjusted p = 0.260). In 2023, the locked model had AUROC 0.733, Brier score 0.175, and calibration intercept −0.124. Among gay/lesbian, bisexual, other, and questioning respondents, weighted outcome prevalence was 49.5%–59.8%, whereas 93.2%–99.6% exceeded the fixed threshold; sensitivity was 0.950–1.000 and specificity 0.009–0.087. Conclusion Main association directions reproduced in U.S. school-attending students in grades 9–12, although selection and interaction uncertainty limit interpretation. School connectedness was associated with lower prevalence; buffering remained inconclusive. Calibration drift and subgroup threshold differences do not support individual screening, referral, or resource allocation.

Authors

Institutions

Publication Details

Journal
Frontiers in Public Health
Published
2026-11-18
DOI
https://doi.org/10.3389/fpubh.2026.1970255
Primary Topic
Adolescent Health and Behaviors
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Relational threats and school connectedness in adolescent poor mental health using repeated cross-sectional U.S. YRBS data and locked temporal validation

Liang Liu, Qingfei Chu, Wenfeng Liu, Wanying Zhang
2 links
Frontiers in Public Health
Adolescent Health and Behaviors
article

Relational threats and school connectedness in adolescent poor mental health using repeated cross-sectional U.S. YRBS data and locked temporal validation

Liang Liu, Qingfei Chu, Wenfeng Liu, Wanying Zhang
article en

Abstract

Background School relationships are associated with adolescent mental health, but it is unclear how consistently these associations and prediction-model performance reproduce across survey waves. Methods This repeated cross-sectional secondary analysis used independent 2021 and 2023 U.S. National Youth Risk Behavior Survey samples (17,232 and 20,103 students). Survey-weighted Poisson models estimated associations with past-30-day poor mental health. Absolute-scale interactions and cross-wave tests assessed bullying, sleep, school connectedness, and parental monitoring. A ridge-penalized logistic model and threshold were locked using 2021 data before evaluation in 2023. Results Weighted outcome prevalence was 29.3% and 28.5%. Bullying, cyberbullying, and sleep <8 h were associated with higher prevalence. High school connectedness was associated with lower prevalence (2021 adjusted prevalence ratio 0.65, 95% confidence interval 0.59–0.70; 2023, 0.71, 0.65–0.78); standardized prevalence differences were −12.8 and −9.7 percentage points. Bullying-by-connectedness estimates were too imprecise to resolve buffering. The connectedness cross-wave test had p = 0.052 (Holm-adjusted p = 0.260). In 2023, the locked model had AUROC 0.733, Brier score 0.175, and calibration intercept −0.124. Among gay/lesbian, bisexual, other, and questioning respondents, weighted outcome prevalence was 49.5%–59.8%, whereas 93.2%–99.6% exceeded the fixed threshold; sensitivity was 0.950–1.000 and specificity 0.009–0.087. Conclusion Main association directions reproduced in U.S. school-attending students in grades 9–12, although selection and interaction uncertainty limit interpretation. School connectedness was associated with lower prevalence; buffering remained inconclusive. Calibration drift and subgroup threshold differences do not support individual screening, referral, or resource allocation.

Frontiers in Public HealthVol. 14
Second Affiliated Hospital of Zhejiang University (CN), Zhejiang University (CN)
Good health and well-being
Adolescent Health and Behaviors
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon