Date of Award


Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)



First Advisor

Frank W. Porell

Second Advisor

Jeffrey A. Burr

Third Advisor

Mark D. Hayward


Past research often considers living alone as a risk factor for older persons. In fact, adverse health outcomes are associated with living arrangement transitions, suggesting a need to consider the dynamic process of living arrangements. Using eight waves of the Health and Retirement Study (1998-2012), this study examines three research questions: (1) Do older Americans’ living arrangements exhibit a pattern of sequence? (2) What are implications of living arrangements at particular older ages on life expectancy? (3) What factors predict transitions out of a living alone arrangement?

The first analysis displays and classifies ordered patterns of living arrangements over time. Baseline living arrangements have a substantial influence on subsequent transitions. Major patterns of women’s living arrangement sequences are more diverse than those for men. In particular, living alone is both the major living arrangements at baseline or intermediately after baseline. These results suggest the importance of the living arrangement status at old ages in relation to subsequent living arrangements over time.

Next, multistate life tables are estimated for calculating life expectancy in total and among distinct living arrangements. While the expected percentage of time spent living alone for men increases with age, about half of women’s total life expectancy at any ages is spent living alone. Older persons living alone do not have shorter life expectancies than those in co-residential living arrangements, particularly among women. This suggests a selection process in which less robust older persons tend to transition to other living arrangements or die at younger ages.

Lastly, discrete-time event history models are used to examine factors associated with transitioning from living alone. Sentinel health events and poorer functional status are associated with an increased risk of death, and, to a lesser extent, a subsequent transition to co-residence or institutionalization. Analyses of transitions from living alone measured concurrently with changes in functional status suggest that many transitions may be immediate reactions to a recent health decline rather than adjustments following a progressive health decline. In either case, subsequent co-residence does not appear to be a common adaptation for many older adults who live alone with increased needs for care.