# Early warning model for depression and anxiety among adolescent students: an empirical study based on Guilin

## TL;DR

This study identifies factors influencing depression and anxiety in adolescent students in Guilin and develops a predictive model to detect these conditions early.

## Contribution

The study introduces a validated early warning model for depression and anxiety in adolescents using logistic regression and ROC analysis.

## Key findings

- 17.83% of the training set and 19.13% of the validation set of adolescent students developed depression and anxiety.
- Factors like academic stress, family cohesion (FACESCV), and resilience (CDRISC) significantly influence depression and anxiety in adolescents.
- The nomogram model achieved 80.25% sensitivity and 80.09% specificity in predicting depression and anxiety.

## Abstract

To analyze the occurrence of depression and anxiety in Guilin and verify the early warning model of depression and anxiety in Guilin.

Using multi-layer sampling method, 10 middle schools were selected from September 2022 to September 2023, with a total of 1,245 middle and high school students in an anonymous self-filled questionnaire survey. A total of 1,150 valid young students were obtained as the survey object, which was randomly divided into training set (920 cases) and verification set (230 cases). Using hospital anxiety and depression scale (HADS) to evaluate the occurrence of Guilin adolescent students depression anxiety, in the form of a questionnaire to understand and analyze the factors affecting Guilin adolescent students depression anxiety, establish the influence of Guilin adolescent students depression anxiety prediction model of occurrence and model validation and efficiency evaluation.

In the training set of 920 adolescent students, 164 patients developed depressive anxiety, or 17.83%, and 756 patients did not develop depressive anxiety. In the validation set, among 230 adolescent students, 44 patients developed depressive anxiety, with an incidence of 19.13%, and 186 patients did not develop depressive anxiety. Logistic Regression analysis showed that OR (95%CI) = 3.44 (1.98, 5.97), Chinese version of the Family Adaptation and Cohesion Scale (FACESCV) [OR (95% CI) = 1.67 (1.07, 2.61)], and the Connor-Davidson Resilience Scale (CDRISC) score for adolescent students [OR (95% CI) = 1.83 (1.15, 2.92)]were the factors that affected the occurrence of depression and anxiety in adolescents in Guilin (p < 0.05). The ROC curve results of the training set showed that the nomogram model predicted the sensitivity of depression and anxiety with 80.25%, specificity of 80.09% and AUC of 0.849.

High risk of depression and anxiety, academic stress, FACESCV and CDRISC score are the influencing factors affecting the occurrence of depression and anxiety in adolescent students in Guilin.