Predicting mortality among older adults in Europe employing SHARE longitudinal data

Georgia Verropoulou, University of Piraeus

The study aims at assessing the relative importance of objective/specific self-reported versus subjective/general health indicators in predicting mortality among older adults and at exploring the potential value of the Global Activity Limitation Indicator (GALI), a recently validated subjective measure of functional limitations, as predictor of death. Longitudinal data from two waves (2004 and 2006-07) of the Survey of Health, Ageing and Retirement in Europe were employed. The sample comprises 17,941 persons aged 50+ at baseline, representing 11 European countries. Associations between covariates and mortality were estimated by sex using Cox’s proportional hazards regression models. Most objective and subjective indicators of health are strong and independent predictors of mortality. Subjective indicators add information on aspects of health and disability beyond the objective ones. There are disparities by sex; Self-Rated Health (SRH) is significant among males only while GALI is significant among women. A combination of objective and subjective measures is more efficient in predicting mortality than objective or subjective indicators alone. GALI emerges as a strong predictor of mortality. This matter deserves further exploration; observation over longer periods of time would help to clarify associations and reinforce conclusions.