Pergamon Soc. Sci. Med. Vol. 46, Nos 4-5, pp. 599-609, 1998 © 1997 Elsevier Science Ltd. All rights reserved PII: S0277-9536(97)00207-4 Printed in Great Britain 0277-9536/97 $19.00 + 0.00 SOCIAL NETWORK TYPE AND HEALTH STATUS IN A NATIONAL SAMPLE OF ELDERLY ISRAELIS* HOWARD LITWIN Paul Baerwald School of Social Work, The Hebrew University, Mount Scopus, 91905-IL Jerusalem, Israel Abstraet--A typology of social support networks was examined in relation to five health measures in a national probability sample of Israelis aged 60 and over, using multiple classification analysis (N = 4214). The procedure revealed that the more resourceful diversified and friend and neighbor net- work types were consistently associated with better scores on measures of basic and instrumental activi- ties of daily living, incontinence, vision and self-rated health, even when controlling for respondents' age, sex and education. The religious family network type, also endowed with considerable support po- tential, tended to correlate with lower health scores. The narrow family focused network had average health ratings or less, and a moderate support capability. The least resourceful network type, the atte- nuated network, was most frequently associated with poor health. © 1997 Elsevier Science Ltd. All rights reserved Key words--support network, health status, elderly, Israel INTRODUCTION Previous research has examined the effect of key sociodemographic characteristics, particularly age, gender and social class, on the health of older people (Linn and Linn, 1980; Broom, 1984; Halpert and Zimmerman, 1986; Victor, 1991; Arber and Ginn, 1993; Pilisuk et al., 1993; Thorslund and Lundberg, 1994). Several other studies have investi- gated the relationship between social support net- works and the health of the elderly, viewing the support network phenomenon as a potential inter- vening variable between sociodemographic back- ground characteristics and health status (Stoller, 1984; Thomas et al., 1985; Hanson and Ostergren, 1987; Weinberger et al., 1987; Mor-Barak and Miller, 1991). Most such research addresses the as- sociation between specific support network measures and health status, for example, marital status (Wyke and Ford, 1992; Kaplan et al., 1993) and the frequency of contact with one's close per- sonal ties (Stueve and O'Donnell, 1989; Mclntosh et al., 1993). It is far less common to find studies addressing the support network factor as a complex phenomenon, that is, as a collection of different social network types, each of which may have a different association with health outcomes (Wenger, 1990). Particularly lacking, in this regard, are stu- dies that address both social network type and the effects of age, gender and social class in relation to health status. *An earlier version of this paper was presented at the 1997 World Congress of Gerontology in Honolulu. The analysis reported upon in this article exam- ines the relationship between different social net- work types and a range of health status measures in a national probability sample of elderly Jews in Israel (N = 4214). The support network types in question were derived in a previous analysis, the method and results of which have been reported in depth elsewhere (Litwin, 1997). The health measures considered in the current study included basic and instrumental activities of daily living, urinary or fecal incontinence, vision impairment, and overall subjective health. In each case, the analysis con- trolled for respondents' age, gender and educational level as the available proxy measure for social class. The main purpose of the study was to examine the unique association of the support network fac- tor, as measured by network type, with the health status of older people. The aim was to address the association of network type and health status inde- pendent of the effects of age, gender and class, which may themselves directly effect both one's health and one's social network type. The study sought to clarify, in particular, whether elderly people with significant health needs were embedded in the kinds of networks that could provide suffi- cient social supports, when required. LITERATURE REVIEW It is widely viewed that social support is related to health status at various points in the life cycle, and particularly in later life. However, there is much less agreement as to the specific aspect of sup- port that accounts for variance in health outcomes. 599