It is interesting, but perhaps not surprising, that living alone seems to be a stronger predictor of mortality in younger persons than older persons. In older persons, living alone may be a proxy for very different phenomena that have opposing impacts on mortality. On the one hand, persons who live alone are more likely to have limited social support, and limited social support increases the risk for mortality. On the other hand, an older person who lives alone is more likely to have good functional status, particularly independence in the basic activities of daily living that are required to live without assistance. Functional independence is a powerful predictor of survival in older persons. In epidemiologic studies, these 2 factors may cancel each other out, leading to a null impact on mortality.
Covinsky KE. The Differential Diagnosis of Living AloneComment on “Living Alone and Risk of Mortality in Older, Community-Dwelling Adults”. JAMA Intern Med. 2013;173(4):321. doi:10.1001/jamainternmed.2013.1617