ABEL, P. ELIZABETH; PHD
 
                        EMORY UNIVERSITY, 1987
 
                        SOCIOLOGY, GENERAL (0626)
 

                         Stressful life events have been associated with the onset of illness. Many studies have explored social
                         networks as a means to reduce the effects of stress on health. Social networks have been found to have
                         a direct and a buffering effect on physical or mental health. These studies often used an institutionalized
                         sample focusing on mental illness due to stress. A national non-institutionalized sample (N = 3025) was
                         used for this research. The relationship among social networks, stress and physical health status were
                         explored. Social networks were measured using the Berkman Social Network Index. Physical Health
                         Status measurements were based on those used in a longitudinal study on the effects of stress and
                         social networks on mortality in Alameda county in the 1970's. The stressful life events measure included
                         eight events with weighted scores similar to those used by Dohrenwend and Associates in 1978.
                         Analyses were done for male and females separately. Select variables were entered into regression
                         equations to determine the change in the R-square. Those variables that demonstrated a significant
                         contribution remained in the model (p < = .05). Although highly significant the amount of variation in
                         physical health status explained by each variable was minimal. Social networks were observed to have a
                         main effect on physical health status for males. There was a gender difference in that social networks did
                         not demonstrate any main effect for women. No social network buffering effect was noted for males or
                         females. Health habits, income, education and age were controlled. Health habits did not contribute to
                         variation in physical health status for men but did for women. When health habits were controlled for
                         women social networks demonstrated a negative influence where previously they had been positive.
                         Social-exchange theory was used as a theoretical explanation to account for the social network gender
                         differences. The fact that women did not believe social networks influenced stress or physical health
                         status merits further investigation. The effect of health habits has implications for methodological
                         improvements in studies that consider physical or mental health as the dependent variable. The fact that
                         social networks does not appear to be very important in predicting physical health status when stressful
                         life events are considered may be due to the temporal nature of these variables. These findings have
                         implications for methodological and theoretical development in the area of social network research.
 

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603