Self-rated health (SRH) is considered a relevant and important predictor for major health outcomes in the older population. SRH status may interact with certain factors and change over a person's lifetime. In this study, we sought to characterize profiles of older people over time by constructing prototypical trajectories of the variable of interest, namely SRH. The underlying assumption was that the collection of observed individual trajectories could be efficiently summarized by a smaller set of latent clusters of those trajectories. Data was obtained from the Longitudinal Survey of Health and Living Status of the Elderly in Taiwan, which was conducted between 1989 and 2003 and included five separate waves of survey. A total of 3937 subjects aged 60 or older (2251 males and 1686 females) comprised the major analytic cohort. Latent Class Growth Analysis (LCGA) was used to identify developmental classes of trajectory patterns in SRH. The results showed that during a 14-year period, SRH developed five major longitudinal trajectories. Less than one-third of the older population was able to maintain their formerly good or moderate health status; when change occurred, decline was more likely than improvement. In addition, LCGA indicated that many demographic characteristics, as well as physical and psychological propensities, were associated with poor SRH in the older population. Specifically, these factors played a role in involving baseline SRH level and its trend toward deterioration in later life. Health care professionals must understand the various longitudinal patterns and factors affecting SRH trajectories if they are to develop programs aimed at maintaining the older population's health and well-being.
Copyright © 2011 Elsevier Ireland Ltd. All rights reserved.