PURPOSE: To examine the effects of living in religiously affiliated and unaffiliated neighborhoods on mortality risks above that of individual risk factors, to determine if this effect behaves in a dose-response manner, and to examine the interaction between community wealth and religious affiliation. METHODS: Multilevel modeling of data from the Israel Longitudinal Mortality Study was used to assess mortality differentials based on neighborhood religious affiliation. Data were analyzed for 141,683 individuals aged 45 to 89 years and living in 882 statistical areas. Overall, 29,709 deaths were reported during the 9.5-year follow-up period. RESULTS: After accounting for individual demographic and socioeconomic (SES) characteristics as well as area-SES, men and women living in religiously affiliated neighborhoods had lower mortality rates than those living in unaffiliated areas (odds ratio(men) = 0.75; 95% CI, 0.67-0.84; odds ratio(women) = 0.86; 95% CI, 0.67-0.96). For men, this relationship behaved in a dose-response manner. Furthermore, the beneficial effects on mortality of living in a religiously affiliated area were consistent across age groups, middle-aged and elderly. Lastly, effect modification of area-SES on area-religion was observed for women only, whereby for women living in higher-SES areas, religiosity had no effect on mortality. CONCLUSIONS: The characteristics of one's immediate neighborhood, namely, community wealth and religious affiliation, have valuable health implications that should be included when assessing mortality risks.
PURPOSE: To examine the effects of living in religiously affiliated and unaffiliated neighborhoods on mortality risks above that of individual risk factors, to determine if this effect behaves in a dose-response manner, and to examine the interaction between community wealth and religious affiliation. METHODS: Multilevel modeling of data from the Israel Longitudinal Mortality Study was used to assess mortality differentials based on neighborhood religious affiliation. Data were analyzed for 141,683 individuals aged 45 to 89 years and living in 882 statistical areas. Overall, 29,709 deaths were reported during the 9.5-year follow-up period. RESULTS: After accounting for individual demographic and socioeconomic (SES) characteristics as well as area-SES, men and women living in religiously affiliated neighborhoods had lower mortality rates than those living in unaffiliated areas (odds ratio(men) = 0.75; 95% CI, 0.67-0.84; odds ratio(women) = 0.86; 95% CI, 0.67-0.96). For men, this relationship behaved in a dose-response manner. Furthermore, the beneficial effects on mortality of living in a religiously affiliated area were consistent across age groups, middle-aged and elderly. Lastly, effect modification of area-SES on area-religion was observed for women only, whereby for women living in higher-SES areas, religiosity had no effect on mortality. CONCLUSIONS: The characteristics of one's immediate neighborhood, namely, community wealth and religious affiliation, have valuable health implications that should be included when assessing mortality risks.