The structure of physical health status. Comparing proxies and self-respondents.

Because gerontological studies often need to rely on the use of proxy respondents, the comparability of proxy and self-respondents is of particular interest. However, it is often impossible to evaluate response agreement between proxy and self-respondents because the necessary data are not available. This study addresses the problem by using a model of health status to evaluate differences between proxy and self-respondents regarding the conceptualization of health status. The model included three dimensions of physical health: chronic illness, functional limitations, and subjective health. Three groups of respondents who differ regarding proxy status and/or physical health status were compared. A subset of matched self-respondents (n = 146) was selected to be comparable in objective health status to respondents who have proxies (n = 140). The third group consisted of physically healthier self-respondents (n = 1,425). Data came from the Study of Well-Being of Older People in Cleveland, OH. No support was found for the expectation that proxies and self-respondents would weight objective health information differently when providing summary statements of subjective health. The results suggest that proxies rely on a conceptualization similar to self-respondents when providing information about another person's health.