A longitudinal study of characteristics and predictors of perceived instrumental and emotional support among the old-old in Israel.

The structure and function of social networks and the perceived instrumental and emotional support associated with these networks were studied longitudinally among a random sample of 687 Jewish Israelis aged seventy-five to ninety-four, stratified by five-year age groups, place of birth (Europe-America, Asia-Africa, or Israel), and gender, interviewed in 1989-92 and 1993-94. Perceived instrumental support at follow-up was predicted by baseline measures of proximity of at least one child and by four or more weekly contacts with intimate friends or family, whereas perceived emotional support was more broadly based and was predicted by more frequent contacts with children, neighborliness, and having more intimate friends. Changes occurring during the follow-up interval, such as entering long-term care or losing a confidant, were associated with a decrease in perceived emotional support but not with a loss of instrumental support. Emotional support was less replaceable over time than was instrumental support.