Objective: To explore the dynamic developmental trajectory of problem eating behavior among college students, and to provide evidence for the early prevention and intervention of eating problems. Methods: A cluster sample of 5 372 freshmen from a university in Jiangsu Province were included and followed up for 4 years by using the eating behavior problems scale. Results: The problem eating behavior of college students in Jiangsu Province showed a decreasing after increasing trend(5.17±1.36, 5.39± 1.53 , 5.14±1.42, 4.96±1.37)( F =112.12, P <0.01), and prevalence of problem eating behavior was highest among sophomores. The developmental trajectory of eating behavior problems in college students could be divided into slow decreasing (86.7%) and slow increasing (13.3%) groups. In addition, sex specific developmental trajectory of problem eating behavior was observed, with initial mean level and change speed of female students were signicantly higher than that of male students. Conclusion There are group heterogeneity and sex specific effect in problem eating behavior of college students, which needs specific intervention strategies provided from colleges.

Objective: To examine the longitudinal relationship between academic pressure and Internet addiction of college students and to provide reference for the development of college students mental health. Methods: In October 2016, a longitudinal 4 waves (T1 to T4) of follow up study was conducted among 5 372 college students from a university in Jiangsu, with an interval of 1 year each time. Academic pressure and Internet addiction were derived from the College Students Mental Health Screening Scale. A cross lag analysis was used to explore the relationship between college students academic pressure and Internet addiction. Results: Repeated measures analysis of variance showed that the main effect of time for academic pressure was significant[ F(3,16 113)= 767.28 , P<0.01,η 2=0.13]. The academic pressure of college students showed a downward trend over time, with the highest at T1 stage and the lowest at T4 stage. The main effect of time for college students Internet addiction was significant[ F(3,16 113)= 165.22 , P<0.01, η 2=0.03]. The Internet addiction of college students showed a trend of decreasing after increasing over time, with the highest stage T2 and the lowest stage T4. The cross lag analysis showed that academic pressure of T1 could significantly predict the Internet addiction of T2, and academic pressure of T3 could significantly predict the Internet addiction of T4, and the path coefficients are 0.07 and 0.07( P <0.01). In addition, the pre test (T n ) Internet addiction can significantly predict the post test [T( n +1)] academic pressure, and the path coefficients were 0.06, 0.06, 0.08( P <0.01). Conclusion Academic pressure can partially predict Internet addiction among college students, while Internet addiction can also significantly predict academic pressure. Colleges and universities should strengthen the construction of learning style, reduce the academic pressure of students,and reduce the Internet addiction of college students through various ways, so as to promote the healthy development of college students.

Objective: To explore the dynamic changes of suicidal ideation in college students, and to provide evidence for the early prevention and intervention of suicidal ideation. Methods: A cluster sampling of 5 372 college students from one university in Jiangsu was selected and followed up for four years (2016—2019) using the College Students Mental Health Screening Scale. Results: The detection rates of suicidal ideation among college students in the first to four (T1-T4) stages were 4.36%, 8.99%, 7.32 %, and 5.90%, respectively, showing decreasing after increasing trend. Stage T2 was the turning point, and the trend of change was significantly different ( P <0.01). The detection rate of suicidal ideation of girls in T1 stage was significantly higher than that of boys ( χ 2=10.55, P <0.01). In the T1-T2 stages, the rate of suicidal idea detection among boys increased faster than that of girls, and the rate of boys in T4 stage surpassed that of girls. The detection rate of suicidal ideation in T1-T4 stage of students in big cities was higher than that of students in small and medium sized cities, towns and rural areas, and the difference in stage T4 was very significant ( χ 2=32.11, P <0.01). The contingency table analysis found that the proportion of suicidal ideation at the same time in the two stages before and after was between 18.63%-35.04% ( P <0.01). There was no suicidal ideation in the previous stage, and the proportion of suicidal ideation in the latter stage was between 4.48%-7.80% ( P <0.01). There were 11 participants who had suicidal ideation in the T1-T4 stages, accounting for 0.20% of the total number, and 4.70% of the suicidal ideation number. Conclusion Colleges and universities should pay more attention to suicidal ideation prevention for vulnerable groups in important periods, and grasp the dynamic information of students through a variety of ways.