Yunjiao Luo,1,* Yuhao Wang,1,* Yingxue Wang,1 Yihan Wang,1 Na Yan,1 Blen Dereje Shiferaw,1 Louisa Esi Mackay,1 Ziyang Zhang,2 Caiyi Zhang,3 Wei Wang1,3– 5 

1School of Public Health, Xuzhou Medical University, Xuzhou, Jiangsu, People’s Republic of China; 2St. Luke’s College of Nursing, Trinity University of Asia, Quezon, Philippines; 3The Affiliated Xuzhou Oriental Hospital of Xuzhou Medical University, Xuzhou, Jiangsu, People’s Republic of China; 4Key Laboratory of Human Genetics and Environmental Medicine, Xuzhou Medical University, Xuzhou, Jiangsu, People’s Republic of China; 5Engineering Research Innovation Center of Biological Data Mining and Healthcare Transformation, Xuzhou Medical University, Xuzhou, Jiangsu, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Wei Wang, Department of Community and Health Education, School of Public Health, Xuzhou Medical University, 209 Tongshan Road, Huangshan Street, Yunlong District, Xuzhou, Jiangsu, People’s Republic of China, Email weiwang90@163.com Caiyi Zhang, Department of Psychiatry, The Affiliated Xuzhou Oriental Hospital of Xuzhou Medical University, 379 Tongshan Road, Dongdianzi, Yunlong District, Xuzhou, Jiangsu, People’s Republic of China, Email amanzcy@live.cn

Purpose: Suicidal mortality is high in rural areas, and suicidal ideation, an early psychology of suicidal behavior, is particularly important for the early prevention and intervention of suicide. This study aimed to establish a nomogram model to predict high-risk groups among rural adolescents who might develop suicidal ideation.
Patients and Methods: This study conducted a cross-sectional survey of 1900 rural secondary school students in Xuzhou, China. The samples were randomly divided into a training set (1330) and a validation set (570), and a nomogram prediction model was constructed using the potential predictors of suicidal ideation screened from the training set using Lasso-Logistic regression. The model was validated using ROC, calibration, and clinical decision curves.
Results: The reported rate of suicidal ideation among rural adolescents is 18.9%. Lasso-Logistic regression found that emotional abuse, emotional neglect, hostility, subjective sleep quality, daytime dysfunction, withdrawal/escape, and depression were significant risk factors for suicidal ideation. A nomogram was built using the above 7 predictors. The area under the ROC curve (AUC) of our predictive model was 0.821 in the training set and 0.765 in the validation set, with corrected C-indices of 0.817 and 0.764, respectively. Furthermore, the calibration curves demonstrated good alignment with the ideal line (P > 0.05), and the decision curve analysis results indicated positive clinical utility.
Conclusion: The nomogram model constructed in this study may be an effective tool for predicting suicidal ideation in rural middle school students. It helps identify high-risk groups with suicidal ideation and provides more reliable information for the early prevention and intervention of suicide.

Keywords: suicidal ideation, nomogram, adolescent, rural, LASSO