Social networks as a predictor of cardiovascular disease incidence and mortality: 10-year observation of the Polish HAPIEE cohort

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): National Sciene Center of Poland The inverse association between social isolation or low social support and cardiovascular disease (CVD) is known but the evidence comes mainly form countries of stable high socio-economic development while research from populations of different social backgrounds remains scarce. Societies of Central and Eastern Europe which underwent profound socio-economic changes followed by the decrease in CVD mortality provide an interesting setting for a study of social determinants of health. The aim of the study was to assess the relationship between social networks and CVD incidence and mortality in Polish urban population. Additionally predictive performance of social networks index was compared with marital status alone. A cohort study with 10-year follow-up was conducted. A random sample of 10,728 residents of our city, aged 45 to 69 was examined. Social networks were assessed at baseline using questions on marital status, participation in groups and organizations as well as the number of friends and relatives with whom participants maintained contact. Points in individual categories were summed up and an index ranging from 0-4 was obtained where higher values meant bigger social networks. Information on new CVD cases was obtained from questionnaires in subsequent phases of the study and confirmed by the review of clinical records. Data on deaths and causes was obtained from the residents’ registry, Central Statistical Office, and participants’ families. The effect of social networks index as well as marital status on the risk of CVD was assessed using Cox proportional hazard models. The observation for incidence included 3,289 men and 3,673 women, respectively (384 and 175 new CVD cases). For CVD mortality 3,595 men and 3,939 women with 156 and 95 CVD deaths were included, respectively. In fully adjusted models no significant associations between social network index and CVD incidence were found. However, marital status was a significant predictor of CVD incidence in men; compared to single or widowed men those married or cohabiting had lower risk of CVD event by 33% (HR=0.67 95%CI=0.49-0.91). Protective effect of bigger social networks against CVD mortality with dose-response effect was observed in men only. Compared to men with the lowest social network index, any increase in the index was related with reduced risk of CVD mortality up to 86% in men with the highest social networks index (HR=0.14 95%CI=0.03-0.62). Lower risk of CVD death was also observed in married or cohabiting men compared to single or widowed participants (HR=0.40 95%CI=0.26-0.61). No significant differences in risk prediction was found between social index and marital status alone. In conclusion, maintaining social networks in middle-aged and older men from the urban population is protective against CVD mortality. However, social network index including information on social participation was not a better predictor of CVD risk than information on marital status alone.