Longitudinal patterns of mental well-being over four years in a german general population sample: a growth mixture modeling approach

Abstract

Background: Although mental well-being is facing growing challenges against the background of global crises such as climate change, pandemics, and social inequality, little is known about longitudinal patterns of mental well-being in the general population.

Methods: The study is based on self-report data from 1,605 adults aged 18 to 64 years who were proactively recruited at a municipal registration office (M = 31.0 years, SD = 10.8 years). Mental well-being was assessed at baseline, and 3, 6, 12, 36 and 48 months later using the five-item Mental Health Inventory (MHI-5). Covariates were smoking, alcohol consumption, fruit and vegetable intake, moderate-to-vigorous physical activity, age, sex, school education and relationship status. Growth Mixture Modeling was used to identify latent trajectory classes of mental well-being. Multinomial logistic regression was used to test whether class membership is predicted by health behaviors and sociodemographic variables.

Results: Three latent trajectory classes were found. The first class ("stable high", n = 1,251, 78%) showed the highest mental well-being throughout the study with only minor fluctuations over time. The second class (n = 192, 12%) showed a "steadily increasing" trajectory starting with the lowest MHI-5 sum score of the three classes at baseline that increased over time. The third class ("fluctuating", n = 162, 10%) reported a slight increase in mental well-being during the first six months, followed by a steep decline to 36 months, which then increased to almost baseline-level at 48 months. The odds of being classified into the "steadily increasing" or "fluctuating" compared to the "stable high" class were higher for men and younger participants. The odds of belonging to the "steadily increasing" compared to the "stable high" class were higher for participants with more than 12 years of school education and those being in a relationship. Health behaviors did not predict latent trajectory class membership.

Conclusion: Three different temporal patterns of mental well-being were found in a German general population sample, with the majority showing a stable level of high mental well-being over four years. Further research is needed to understand fluctuating patterns and the causal factors influencing mental well-being.

Keywords: Latent trajectory class; Lifestyle behaviors; Longitudinal patterns of mental well-being; MHI-5; Mental health; Mixture modeling.

Fig. 1

Growth Mixture Model for latent trajectory classes of mental well-being. Note. Growth Mixture Model for latent trajectory classes of mental well-being. Mental well-being as repeatedly observed manifest variable over time (in rectangles) is accounted for by linear, quadratic, and cubic growth factors. Heterogeneity in the mental well-being trajectories is captured by a latent class variable that is predicted by sociodemographic variables and health behaviors. The three-step approach for incorporating covariates into mixture models was applied [54]

Fig. 2

Three latent trajectory classes of mental well-being. Note. MHI-5 = Mental Health Inventory