Growth Trajectories of Peer Norms, Self-efficacy and Condom Use Behavior Among Sexually Active Chinese Men Who Have Sex with Men: Latent Class Analysis and Growth Mixture Modeling

Abstract

Data from a randomized controlled trial in 2015 were used to estimate the growth trajectories of peer norms, self-efficacy, and condom use behavior, and to identify associated sociodemographic and behavioral factors among a sample of 804 Chinese men who have sex with men (MSM). Latent class analysis and growth mixture modeling were conducted using Mplus. Two growth trajectories were estimated for each outcome variable with good model fit. The growth trajectories of peer norms were related to age (β = −0.066, p < 0.05). The growth trajectories of self-efficacy were related to age (β = 0.057, p < 0.01) and using a condom during first sexual encounter with another man (β = 0.777, p < 0.001). The growth trajectories of condom use behavior were related to income (β = 0.366, p < 0.01) and having casual male partners (β = −1.016, p < 0.001). Predictors for the growth factors within each latent class were also estimated. For subsets of MSM who are older, richer, used a condom during their first sexual encounter with another man, and do

Identifying longitudinal patterns of CPAP treatment in OSA using growth mixture modeling: Disease characteristics and psychological determinants

“…Our approach differed from previous analyses in two ways. First, we incorporated growth mixture modeling to examine patterns in longitudinal data with three repeated measures to identify classes or subgroups within a population, which were empirically derived, not rationally or qualitatively derived; that is, we allowed the GMM statistical algorithm to produce the latent profiles ( 43 ). Second, we associated the trajectories with multidimensional disease manifestations of OSA patients, which helped to further understand the potential characteristics behind the complex adherence behavior.…”