Mental health issues are increasingly prevalent among college students, impacting their academic performance and overall well-being. While previous research has examined overall trends, this study addresses the need to understand the diverse developmental trajectories of depression, anxiety, and stress within this population. The transition to college involves significant life changes, including increased academic pressure, independence from family, and navigating new social dynamics. These factors can contribute to mental health challenges, but not all students experience these challenges in the same way. The study utilizes a person-centered approach, focusing on individual differences in mental health development, rather than solely on average trends. Understanding these diverse trajectories is crucial for developing effective prevention and intervention strategies tailored to specific student needs. The study utilizes data from the Beijing College Students Panel Survey (BCSPS), a longitudinal dataset offering a unique opportunity to track mental health changes over four consecutive academic years. The sophomore year is chosen as a potential turning point in college life due to its significance in students' adaptation to the college environment and establishment of academic and career goals.

Existing research has explored mental health trajectories across various age groups, including adolescents and adults. Studies using growth mixture models (GMM) have identified distinct trajectories of anxiety and depression, demonstrating significant individual heterogeneity. In the context of college students, previous research primarily focused on overall changes in mental health, noting increased strain during the first year and potential improvement in later years. However, these studies largely overlooked the heterogeneity within the student population. This study bridges this gap by specifically examining the developmental trajectories of depression, anxiety, and stress among college students, accounting for individual differences using a person-centered approach. Prior research also highlighted various internal (gender, personality, lifestyle) and external (family background, peer relationships) factors that influence mental health. These factors, such as gender, personality (extroversion), BMI, sleep hours, family background (hometown location, parental education, siblings), and relationships with classmates were considered potential predictors of trajectory class membership in this study.

This longitudinal study utilized data from the Beijing College Students Panel Survey (BCSPS), employing a stratified, multistage, probability proportional to size sampling method to minimize sampling bias. The study included 2473 students enrolled in 2008, tracking their mental health over four years. The Depression Anxiety Stress Scales-42 (DASS-42) was used to measure depression, anxiety, and stress levels annually. The DASS-42 is a widely used and validated self-report measure, assessing the severity of symptoms over the past week. Additional variables collected at baseline included gender, extroversion personality, BMI, sleep hours, relationship with classmates, hometown location, number of siblings, and parents' education levels. Missing data were handled through appropriate techniques (not explicitly stated in the paper). A piecewise growth mixture model (PGMM) was employed in Mplus 8.3 to identify distinct trajectories of depression, anxiety, and stress, using the sophomore year as a change point. Model fit indices (AIC, BIC, SABIC, entropy, LMR-LRT, B-LRT) were used to determine the optimal number of trajectory classes for each mental health measure. The number of latent classes was iteratively increased until the model provided an adequate fit to the data. A multinomial logistic regression model in Stata 16.0 was used to identify factors associated with membership in each trajectory class, comparing each subtrajectory to the “low and stable” class as a reference.

The PGMM analysis revealed four distinct trajectories for depression: low and stable (79.34%), increasing (8.61%), decreasing then stable (8.25%), and increasing then decreasing (3.80%). For anxiety and stress, five trajectories were identified: low and stable, increasing, increasing then decreasing, decreasing then stable, and decreasing and high. The decreasing and high class was unique to anxiety and stress, showing an initial decline in symptoms followed by relatively high levels persisting throughout the remaining years of college. Multinomial logistic regression analysis identified several factors associated with trajectory class membership. For depression, hometown location, low sleep hours, and relationship with classmates were significantly associated with the decreasing then stable class. High BMI and high sleep hours were associated with the increasing class. For anxiety, males were more likely to be in the increasing class, while extroversion and father's education were associated with the decreasing then stable class. High sleep hours were also a significant predictor of the increasing class. For stress, hometown location and low sleep hours were associated with the decreasing then stable class, and hometown location and high sleep hours predicted the increasing class. The decreasing and high class for both anxiety and stress showed low sleep hours as a significant predictor.

The findings support the heterogeneity in mental health development during college, highlighting the limitations of focusing solely on average trends. The identified trajectories suggest different patterns of mental health challenges across the four years of college, with the sophomore year potentially representing a turning point for some students. The significant role of various factors, both internal (e.g., sleep, BMI, personality) and external (e.g., social support, family background), emphasize the multi-faceted nature of college student mental health. The study's findings have implications for prevention and intervention strategies, emphasizing the need for targeted interventions based on individual student profiles. Further research should focus on understanding the mechanisms underlying the identified trajectories, exploring longitudinal relationships between these variables and specific outcomes, and validating these findings across different college settings and cultural contexts. The study emphasizes the need for personalized interventions targeting various subgroups of college students, including those at high risk and those exhibiting initial declines followed by persistent difficulties.

This study identified distinct developmental trajectories of depression, anxiety, and stress among college students, highlighting the heterogeneity within this population. Different factors were associated with these trajectories, suggesting the need for tailored interventions. Future research should focus on replicating these findings in diverse populations and explore the underlying mechanisms that drive these trajectories. The identification of distinct trajectories emphasizes the need for personalized support and intervention strategies aimed at preventing mental health challenges and promoting overall well-being.

The study's generalizability may be limited by its focus on college students in Beijing. The use of self-report measures could introduce bias. Some variables were measured with a single question, limiting their depth and potentially affecting the interpretation of results. Future studies should address these limitations by utilizing larger, more diverse samples and employing multiple data collection methods to mitigate self-report biases. More comprehensive assessments of social support and other relevant factors may also yield a deeper understanding of the developmental trajectories.