Predictors of Institutionalization at the Age of 70

Existing knowledge about institutionalization of elderly individuals is mainly based on cross-sectional data and does not give good indications of the cumulative risk of institutionalization. The purpose of this study was to analyze longitudinal data prospectively, to estimate the cumulative risk of ending up in elder care institutions at some point in life. Methods: The study was based on a longitudinal investigation of a random sample of 70-year-old persons living in Gothenburg, Sweden (H70). The sample was followed from age 70 to age 100. Results: The prospective analysis showed that 50% of the subjects eventually moved to some kind of elder care institution. Significantly more women than men relocated although women moved later in life. In a Cox regression model at the age of 70, gender was a strong predictor of future institutionalization together with SES (Socioeconomic Situation), objective health, objective social network and ADL (Activities in Daily Life) functions. Discussion: The proportion of elderly relocating to institutions was significantly higher than proportions generally found in cross-sectional studies. It is noteworthy that, already at age 70 it was possible to find variables that predicted future institutionalization. Our findings confirm that longitudinal analyses give more in depth understanding of institutionalization than cross-sectional designs.