Latent Profile Analysis of Mental Health among Chinese University Students: Evidence for the Dual-Factor Model

Abstract

The dual-factor model of mental health has garnered substantial support, positing the necessity of encompassing both negative (e.g., psychological problems) and positive (e.g., well-being) indicators in comprehensive evaluations of people’s mental health. Nonetheless, the nature of the profiles and predictors (such as academic emotions) during four years of university life lack clarity, hampering a profound understanding of mental well-being among university students. This research included 135 items designed to assess an array of depression symptoms, negative emotional experiences, life satisfaction, positive emotional experiences, and academic emotions. First, this research affirmed the applicability of the dual-factor model in the context of Chinese university students (N = 2277) with the utilization of confirmatory factor analysis (CFA). Furthermore, latent profile analysis (LPA) was employed to delineate prevalent constellations of psychological symptoms and subjective well-being within participants. The outcomes unveiled the existence of three distinct clusters: (1) Complete Mental Health, (2) Vulnerable, and (3) Troubled. Third, by employing R3stept on academic emotions and mental health classifications, this study revealed that there were associations between this variable and specific amalgams of psychological symptoms and well-being levels. These findings bear influence on the practice of mental health screening and the identification of individuals necessitating targeted interventions.

Uncovering heterogeneity in mental health changes among first-year medical students

“…Again, in other investigations, only three profiles were identified. For instance, two studies could not detect the symptomatic but content group in their sample [ 31 , 38 ], while another study could not detect the vulnerable group [ 39 ]. The different profile solutions may be due to the different investigated populations and the different mental health indicators used in previous studies.…”

“…Also, studies among college students found support for the dual-factor model revealing variations in (1) hope, gratitude, attention problems, and locus of control [ 25 , 29 ], (2) grit, savoring of positive emotions and self-focused positive rumination [ 29 ], (3) academic achievement, interpersonal connectedness, and physical health [ 30 ], as well as (4) academic emotions [ 31 ] between the sub-groups.…”

“…Since LCA/LPA identifies groups based on the data, it does not necessarily result in four groups, as is the case with theory-driven procedures. In several studies using LPA, only three distinct mental health groups were identified [ 31 , 38 , 39 ]. For example, Jiang et al [ 31 ] and Zhou et al [ 38 ] identified the groups: complete mental health , vulnerable and troubled .…”

“…In several studies using LPA, only three distinct mental health groups were identified [ 31 , 38 , 39 ]. For example, Jiang et al [ 31 ] and Zhou et al [ 38 ] identified the groups: complete mental health , vulnerable and troubled . Whereas Clark & Malecki [ 39 ] identified a different three profile solution corresponding to the groups: complete mental health , symptomatic but content and troubled .…”

Dual-factor Models of Mental Health: A Systematic Review of Empirical Evidence

“…Moreover, a diversity of results was found on age differences. On the one hand, older adults with higher education, income, better cognitive function, and employment seem likelier to belong to the Complete Mental Health group (Jiang & Lu, 2019;Jiang et al, 2023) and less likely to belong to the Troubled group (Xu et al, 2023). On the other hand, older adults seem less likely to belong to the Complete Mental Health group or the Complete Mental Illness group, given that they are more likely to belong to the Moderate Mental Health group (Westerhof & Keyes, 2010).…”