Research On The Association Mechanism Of Social Network Types With Health And Health Related Behaviors Among Older Adults In Rural Areas

[Purpose]This study identifies the social network types of rural elderly by improving the measurement contents of social network types,and explores the associations of social network types with the health and health related behaviors of rural elderly.Then,it focuses on analyzing whether the perceived quality of social relations and the social support play a mediating role in the relationship between social network types and health and health related behaviors.Finally,it would clarify the relevant impact paths.[Methods]This study combined the qualitative and quantitative research methods.In qualitative research,this study adopted a cross-sectional study approach to obtain the latest data on the personal sociodemographic characteristics,the social network types,the perceived quality of social relations,the social support,the health,and health related behaviors of rural older adults in six rural villages in Jiayu County,Hubei Province,China.The cluster sampling was conducted,a final sample of 405 respondents providing complete information was obtained.The reliability and validity of each scale were analyzed.Then,K-means cluster analysis was used to identify the social network types of rural elderly.Furthermore,the bivariate association analysis and multiple linear regression analysis were used to explore the association between social network types and health and health behaviors.First,the analyses of variance(ANOVAs)and the multiple linear regression were used to analyze the relationship between social network types and the health,the health behaviors,the perceived quality of the social relations and the social support of the elderly.Second,the bootstrapped mediation analysis based on the PROCESS script were used to analyze whether the perceived quality of social relations and the social support played a mediating role in the relationship between the social network types and the health and health related behaviors.Moreover,whether the health related behaviors played a role in mediating social network types to the health.In order to control for any possible unequal distributions within clusters,the sociodemographic variables for age,gender,marital status,educational level,household income,and chronic/non-chronic diseases were included as predictors,and the diverse network was used as the reference group.In quantitative research,the semi-structured interviews and thematic framework analysis were used to analyze the social network relationships,the health and health related behaviors of the elderly,as well as the main influencing factors,and clarify the experience and difficulties of the relevant government departments for promoting the health of the elderly.[Results]The results of qualitative research showed that:(1)The scale of social network types contains seven items,including the number of living children,the number of close relatives,the number of close friends,the frequency of contact with living children,the frequency of contact with close relatives,the frequency of contact with close friends,and the frequency of social activities.The reliability and validity of this scale met the requirements(Cronbach's?=0.792,KMO=0.794,the Bartlett test of sphericity was significant at the0.001 level).We identified five robust social network types:the diverse type,the restricted type,the family type,the friends type and a specific type—the family restricted type.(2)ANOVAs showed that there was statistical significance in the health and health related behaviors among different social network types(P<0.001).Older people with the diverse and friend networks were significantly associated with lower levels of depressive symptoms,whereas those with the restricted and family-restricted networks were significantly associated with higher levels of depressive symptoms and lower levels of self-rated health.Older people with the restricted and family-restricted networks had worse health related behaviors than those in other networks,especially than those in diverse network.The regression results showed that the social network types were significantly associated with health and health related behaviors(P<0.05,P<0.01,or P<0.001).(3)ANOVAs showed that there was statistical significance in perceived positive relations among different social network types(P<0.001).Older adults with diverse network were associated with better perceived positive relations than those with other network types,while the opposite was true for those older adults with the restricted network.In addition,the perceived negative relations had statistical significance in different social network types(P<0.05),and the diverse network was better than the restricted network.Besides,the social support had statistical significance in different social network types(P<0.001),the diverse network was better than other network types.However,the restricted network was worse than other network types.The family network was better than the family-restricted network,but it was worse than the friend network.The regression results showed that the social network types were significantly associated with the perceived quality of the social relations and the social support(P<0.05,or P<0.001).(4)The perceived quality of the social relations and the social support played a mediating role in the relationship between the social network types and the health:(1)When the level of depressive symptoms was a dependent variable,the direct effect was significant(R~2=0.03,P<0.001).The perceived positive relations(95%CI=-0.49 to-0.15),the perceived negative relations(95%CI=0.02 to 0.24),and the social support(95%CI=-1.17 to-0.59)played a partial mediating effect,and the mediating effects accounted for 23.57%,11.62%and 56.40%,respectively.Additionally,the variables of the social network types,the perceived quality of the social relations and the social support can effectively explain the 68%variation in the level of depressive symptoms.(2)When the level of self-rated health was a dependent variable,the direct effect was significant(R~2=0.03,P<0.001).The perceived positive relations(95%CI=0.01 to 0.08)played a partial mediating effect,and this effect accounted for 27.78%.The variables of the social network types,the perceived quality of the social relations and the social support can effectively explain the 26%variation in the level of self-rated health.(5)The perceived quality of social relations and the social support played a mediating role in the relationship between the social network types and the health related behaviors:(1)When the level of medication adherence was a dependent variable,the direct effect was significant(R~2=0.02,P<0.05).The perceived positive relations(95%CI=-0.42 to-0.03)and the perceived negative relations(95%CI=0.01 to 0.24)played a partial mediating effect,and the effects accounted for 30.74%and 19.67%,respectively.The variables of social network types,the perceived quality of the social relations and the social support can effectively explain the20%variation in the level of the medication adherence.(2)When the level of medication habits was a dependent variable,the direct effect was significant(R~2=0.02,P<0.05).The perceived positive relations(95%CI=-0.44 to-0.07)and the perceived negative relations(95%CI=0.004 to 0.19)played a partial mediating effect,and the effects accounted for31.53%and 11.86%,respectively.These three variables could effectively explain the 22%variation in the level of the medication habits.(3)When the level of health-seeking behavior was a dependent variable,the direct effect was not significant(R~2=0.007,P>0.05).The prceived positive relations(95%CI=0.02 to 0.08),the perceived negative relations(95%CI=-0.03 to-0.001),and the social support(95%CI=0.01 to 0.14)played a mediating role.These three variables can effectively explain the 20%variation in the level of the health-seeking behavior.(4)When the level of the preventive behaviors was a dependent variable,the direct effect was significant(R~2=0.03,P<0.01).In the indirect effect,only the social support(95%CI=0.08 to 0.29)played a part in the mediating effect,and this effect accounted for74.18%.These three variables can effectively explain the 21%variation in the level of the preventive behaviors.(6)The health was affected by the social network types through the mediators of health related behaviors:(1)When the level of depressive symptoms was a dependent variables,the direct effect was significant(R~2=0.11,P<0.001).In the indirect effect,the medication adherence(95%CI=0.002 to 0.02)played a part in mediating effect,and this effect accounted for 9.6%.(2)When the level of self-rated health was a dependent variable,the direct effect was significant(R~2=0.06,P<0.001).In the indirect effect,the medication adherence(95%CI=-0.004 to-0.001)played a mediating role,and the mediating effect accounted for 21.38%.However,the preventive behaviors(95%CI=-0.02 to-0.002)played a masking effect,and the effect accounted for 13.79%.The results of quantitative research showed that:the interview results of rural elderly were mainly divided into four major themes,including the health status,the health behavior status,the social network relationship,and the main influencing factors.Due to the low level of economic,education and medical care,etc.,the level of health for rural elderly was generally low,and the awareness on the importance of the health related behaviors was not enough.Additionally,the social network relationship of the rural elderly was not rich enough and lacked the social support.Although the relevant government departments had taken some measures to promote the health of the elderly,while the effectiveness of the measures would need to be improved.[Conclusions](1)This study improves the measurement content of the social network types and identifies five robust network types:the diverse type,the restricted type,the family type,the friends type and a specific type—the family-restricted type.(2)There are differences in the health,the health related behaviors,the perceived quality of social relations,and the social support of the rural elderly among different social network types.(3)The identified social network types are closely related to the health and health related behaviors of the rural elderly.In addition to directly affect on health and health related behaviors(except for health-seeking behavior),the social network types also indirectly affect them through the mediating effects of the perceived quality of the social relations and the social support.Besides,the social network types also contribute to health through partial health related behaviors.(4)The overall level of the health and health related behaviors of the rural elderly is low.Policy makers need to pay more attention to the important influencing factors of social network relationships(including the social network types,the perceived quality of the social relations and the social support),and the relevant impact mechanisms,when they take the measures to promote the health of the elderly.[Innovation and Deficiency]The innovations of this study are as follows:first,it improves the measurement content of the social network types,and successfully identifies the social network types of the rural elderly,including the diverse type,the restricted type,the family type,the friends type and a specific type—the family-restricted type.Second,it constructes the association mechanism between the social network types and the health and health related behaviors.(1)The social network types directly affect the health and health related behaviors.(2)The social network types also indirectly affect them through the mediating effects of the perceived quality of social relations and the social support.(3)The social network types also contribute to the health through partial health related behaviors.There have some limitations in this study.First,this study uses cross-sectional survey data.In the future,the longitudinal data or the queue tracking research could be used to analyze the causal relationship between the social network types and the health,thus capturing the characteristics and variations of the social networks of the older people at different stages.Second,the survey sample of this paper only includes rural older adults.The urban older adults should be included in the future,and we would compare the differences between the social network types of the urban and rural older adults.In addition,a powerful statistical technique—the structural equation modeling would be adopted to further clarify the mechanism of the impact of the social network types on health.