Well-being profiles in adolescence: psychometric properties and latent profile analysis of the mental health continuum model – a methodological study

Abstract Background: The Adolescent Mental Health Continuum Short Form (MHC-SF) is a psychometrically valid tool to evaluate the domains of subjective well-being, but there is a lack of investigations which could distinguish subgroups with distinct subjective well-being profiles based on this measurement. Therefore, after testing the competing measurement models of the MHC-SF, our main aim was to identify subjective well-being profiles in a large adolescent sample. Methods: On a representative Hungarian adolescent sample (N=1572; 51% girl; mean age was 15.39, SD=2.26) confirmatory factor analyses (CFA) and exploratory structural equation modeling (ESEM) were used to test the factor stucture of the Adolescents MHC-SF. In addition, gender invariance of the best fitting model was also tested. Latent Profile Analyses (LPA) were conducted to reveal distinct subgroups and these profiles were then compared. Results: Results support the bifactor model of MHC-SF: the general and specific well-being factors which were invariant across gender. LPA yielded four subgroups, three of them have been theoretically hypothesized in previous works (i.e. flourishing, moderate mental health, languishing), but an emotionally vulnerable subgroup also emerged. Compared to the languishing group, this new subgroup demonstrated higher scores on prosocial behaviour, but had comparable level of loneliness and internalizing symptoms. Conclusions: Our results suggest that the MHC-SF is a reliable and valid instrument for assessing overall well-being and its components. In addition, the identification of young people to be at risk for low mental health may help us to tailor mental health promotion programs to their special needs. Keywords: Well-being, Positive Mental Health, Mental Health Continuum Model, Adolescents, Confirmatory Factor Analysis, Exploratory Structural Equation Modeling, Gender Invariance, Latent Profile Analyse