30. Development of an early identification and warning model for bipolar disorder IN college students based on positive psychological resources

Abstract Background Bipolar disorder (BD) is a serious mood disorder, and college students have become a high-risk group due to multiple pressures such as academic and interpersonal stress. Early identification is crucial for improving prognosis, but traditional warning models often focus on pathological symptoms and overlook the dynamic interaction between positive psychological resources as protective factors and susceptibility factors. To achieve faster and more comprehensive early identification of bipolar disorder (BD), a university student bipolar disorder early warning model containing positive psychological resources was constructed by combining risk and positive psychological indicators. Methods The study adopts a longitudinal design. Through literature analysis and expert interviews, determine the warning indicator pool, including risk dimensions and positive psychological resource dimensions. Subsequently, a baseline screening was conducted on freshmen at a certain university, and a brief international neuro psychiatric interview was used to confirm BD spectrum diagnosis, and candidate indicator data was collected. And use logistic regression and machine learning methods to screen key predictive variables and construct an early warning model. According to the prediction results, independent samples (n = 652) were divided into high-risk group (n = 89) and non high risk group (n = 563), and a 12-month follow-up validation was conducted. The mood state was evaluated quarterly, and the endpoint event was the first appearance of symptoms that met the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). Results The key predictive variables of the model and their baseline scores comparison in the BD high-risk group and non high risk group are shown in Table 1. As shown in Table 1, the emotional instability and sleep quality index of the high-risk group students who eventually developed BD at baseline were 18.7 ± 3.2 (p<.001) and 12.5 ± 2.9 (p<.001), respectively, significantly higher than those of the non high risk group students without disease risk, indicating higher levels of emotional instability and sleep problems; And the positive rate of BD family history is significantly higher. On the four positive psychological resource indicators, the scores of high-risk group students were significantly lower than those of non high risk group students (p<.001), indicating a general lack of psychological capital, resilience, optimism, and social support. Discussion The integrated model constructed by the research shows good early recognition performance. It incorporates the scarcity of positive psychological resources into the warning framework, effectively complementing traditional risk indicators, and can more comprehensively characterize the psychological characteristics of early BD risk. We hope to conduct cross sample validation among college students from different cultural backgrounds in the future and construct a dynamic monitoring model. Funding No.2023SJSZ1230; No.230828444507263.