A nomogram for predicting non-suicide self-injury in youth depression patients: a longitudinal analysis

Non-suicidal self-injury (NSSI) behavior is a common issue in youth with depression while lacks a prognostic prediction model. This study aims to develop a nomogram for NSSI in youth with depression. A total of 701 patients were included in the analysis based a youth depression cohort. They were further divided into a training set and a validation set. The Lasso regression and binary logistic regression were used to select variables for the nomogram. The nomogram's discrimination, calibration, accuracy, clinical applicability, and generalization involved the use of metrics such as the area under the receiver operating characteristic curve (ROC), Hosmer-Lemeshow (HL) test, Brier score, calibration curve, decision curve analysis (DCA), and internal validation. The Delong test was employed to compare the ROC performance between the two models. These patients had an average age of 19.3+/-3.3 years, and 101 (20.6%) individuals reported NSSI. Five risk factors for NSSI were identified: age, age at first medication, previous instances of NSSI, academic stress score in the Adolescent Self-rating Life Events Check List (ASLEC), and somatic anxiety of the 14 item Hamilton Anxiety Rating Scale (HAMA-14). The training set and validation set of binary logistic regression-based model showed good discrimination (area under the curve (AUC) 0.781, 95% CI: 0.735 ~ 0.827; and 0.757, 95% CI: 0.682 ~ 0.831, respectively), calibration (P = 0.421 and 0.175, respectively), and accuracy (Brier score 0.119 to 0.155 and 0.109 to 0.168, respectively). A prognostic nomogram was developed and validated to assist clinicians in predicting NSSI of youth with depression.