Examination of Social Exclusion as a Construct and Its Impact on Mortality and Survival Time among Older Adults

Although social exclusion may be an important predictor of mortality, there is a lack of empirical
evidence on the impact of social exclusion on mortality/survival time among older adults in the
U.S. The study reported in this dissertation examined the effects of social exclusion on mortality
and survival time using a nationally representative sample of adults over 50 years of age. The
primary goal was to better understand how social exclusion affects mortality and survival time
among older adults. Data were drawn from the 2010 and 2014 Health and Retirement Study (core
wave, exit wave, and psychosocial supplement) and focused on older adults aged 65 years or older
(8,439 participants). Social exclusion indicators were based on previous conceptual and empirical
frameworks, and a total of 21 possible indicators were measured. On the basis of correlation
analyses, three indicators were excluded due to very low correlation coefficients or statistical
nonsignificance. The remaining indicators nested in four social exclusion domains (material
resources, social, health, and community). Average inter-item correlations supported the reliability
of the remaining 18 indicators, and multiple regression models supported the validity of the four
domains. The results of a Cox regression model revealed that the material resources domain had a
higher influence on mortality than did the other three domains. Six individual indicators (wealth,
car availability, social activity, social membership, self-rated health, and disability) were associated
with mortality, adjusting for demographic variables. Kaplan–Meier survival analyses based on
each indicator showed that all social exclusion indicators were significantly associated with
survival rate with the exception of social support, city/town satisfaction, and neighborhood safety.
The findings from this dissertation indicate that social exclusion affected mortality and survival
time among older adults in the U.S. However, not all indicators were associated with mortality
and survival time. The results of the dissertation imply that policies can be modified to improve
social participation and access to transportation and that multidimensional services or
interventions focusing on the more influential social exclusion indicators are needed (e.g.,
connecting resources, facilitating social connections). Further, there is a need to reconcile the
differences between conceptual and empirical domains/indicators.