Key points are not available for this paper at this time.
The 1984-1986 Longitudinal Study of Aging (LSOA) was used to investigate self-assessed health as a predictor of 2-year mortality in a subsample of 1,252 persons aged 70 and over. The LSOA sample was screened to exclude individuals reporting a high-risk medical condition or difficulty in instrumental activities of daily living. Logistic regression showed independent predictive effects with higher mortality for age (older), sex (male), less favorable self-rated health, and a family network variable (having no living children or siblings), and sex-specific body mass (highest quintile) was associated with lower mortality. Results therefore support prior studies showing that self-rated health predicts mortality, even in a very healthy elderly subsample, and with a follow-up period that is among the shortest reported to date.
Rakowski et al. (Fri,) studied this question.