Predicting Survival of Older Adults
ABSTRACT: This talk explores various "non-conventional" predictors of short-term survival. Social scientists have been arguing that a simple question asking respondents to rate their overall health in several categories (a question often labeled "self-rated health") is a strong predictor of downstream health and survival. But can external assessors - such as physicians or interviewers - provide as good or perhaps better ratings than individuals themselves? How well do biological markers predict survival compared with a broad range of self-reported health information including the presence of disease and disability? Does information on changes in biomarker values over time enhance mortality prediction over a one-time measure? Among an extensive set of predictors of survival in four countries, what matters the most? How does telomere length compare with more conventional predictors? These questions will be addressed using data from recent national longitudinal surveys.#4816