Social Networks and hypertension in Some rural residents Aged 60-64

Face-to-face interviews were carried out to investigate the relationship between social networks and hypertension in 958 rural residents(males=440, females=518) aged 60-64 of a community-dwelling sample of Dalsung County from April to September in 1996. Eight elements of social network were measured : marital status, regular religious attendance, membership in groups, number of friends, relatives, siblings, children, grandchildren. Hypertensives were defined as meeting at least one of following criteria : hypertension history, systolic blood pressure more than 160 mmHg, diastolic blood pressure more than 95 mmHg. In univariate logistic regression for males, having 1-4 friends vs. none showed odds ratio 0.43 (95% Confidence interval CI 0.19-0.96) and having 2-3, 4 and more than 5 children had reduced prevalence of hypertension with odds ratios 0.21 (95% CI 0.06-0.72), 0.14 (95% CI 0.04-0.49), 0.24 (95% CI 0.07-0.82), respectively when compared with persons without children. In females, there was no elements of social network statistically significant. Having 5-9 grandchildren vs. none showed a marginally significant odds ratio 0.42. In multivariate logistic regression models for males with adjustment for age, education, body mass index, smoking and drinking, number of friends and children showed increased odds ratios and number of close relatives gained a statistically significant odds ratios (0.44-0.50). In females, the adjustment yielded little changes of odds ratios except number of grandchildren which gained a statistically significance. These results suggest that only a certain elements of social network may be associated with reduced risk of hypertension and they may be different between genders in rural resident aged 60-64.