Measuring transitions through stages of exercise: application of latent transition analysis

In this longitudinal study of 504 nursing staff members from seven Dutch nursing homes, which transitions in exercise behaviour had occurred over a 12-mo. period were investigated. For this purpose, latent transition analyses were employed. Latent transition analysis is an extension of Markov chain modelling and of Latent class analysis and has been especially developed to study dynamic latent variables. The data supported the notion that natural change in exercise occurs through various stages and that these are sequential in nature. Latent transition analysis was an effective method for the comparison of various models describing possible transitions over time and for estimation of the likelihood of the occurrence of such transitions.