Abstract Background With the popularization of higher education and the intensification of talent competition, the academic pressure faced by college students has become normalized and diversified, and the relationship between academic pressure and mental health has become a research focus in the fields of educational management and psychology. Traditional research mostly relies on questionnaire surveys and focuses on static correlations. It has limitations such as data lag and single indicators, and it is difficult to capture the dynamic patterns of the two changes over time. With the advancement of smart campus construction, the accumulation of multi-source big data such as course grades, attendance records, and mental health assessments has provided the possibility to break through traditional research bottlenecks. The study uses campus big data to quantitatively analyze the dynamic relationship and key influencing factors between academic stress and mental health. The study provides directions for colleges and universities to carry out academic adjustments and mental health intervention, and helps college students grow up healthily. Methods The study selected 2863 full-time undergraduates from a certain university as a sample, including 1521 boys and 1342 girls; 789 freshmen, 856 sophomores, 732 juniors, and 486 seniors. Campus big data was collected for 2 academic years (4 semesters), including academic data (credits, grades, attendance, homework submissions) and mental health data (depression and anxiety scale scores, psychological consultation records). Then we construct a comprehensive index of academic stress (including two dimensions: academic burden and performance) and core indicators of mental health (depression, anxiety level, comprehensive rating), and explore the dynamic correlation through Pearson correlation and cross-lagged analysis. Results Research shows that the academic stress index of college students peaks in their sophomore year and falls back in their senior year, with an average score of 58.3 ± 12.6; 18.3% of students have mild or above depression, and 15.7% have mild or above anxiety. There is a significant positive correlation between academic stress and depression and anxiety scores in each semester (r = 0.32, p.001), and stress and psychological problems are significantly intensified before and after exam week. Academic stress in the previous semester can significantly predict mental health status in the next semester (β = 0.23, p001), but the reverse prediction effect is weak. In addition, the correlation strength among female students, low academic performance, and sophomore students is higher (β = 0.41), while good social interaction and regular schedule can buffer the impact of stress. The results show that the impact of academic pressure on mental health is cumulative and significantly affects students' mental health. Discussion Through campus big data analysis, the study clarified that there is a significant positive correlation between college students' academic stress and mental health, and that the strength of the dynamic correlation is divided into groups (such as girls and students with academic difficulties). It also revealed the buffering effect of social support and work and rest patterns, indicating that colleges and universities should carry out special stress management to strengthen students' psychological endurance during peak stress periods, enrich campus social activities, and guide healthy work and rest. Future research can expand the sample scope, include family background, personality traits and other variables, and combine machine learning to build a risk prediction model to improve the accuracy of intervention, thereby providing a basis for colleges and universities to formulate differentiated psychological intervention. Funding: No. 24ZK0103; No. 2024SZA10; No. SCJG24B097.
Han Zou (Sun,) studied this question.