Socio-demographic predictors of cardiovascular mortality in South Africa: cross-sectional analysis of stats-SA data from 2006 - 2011

A Research Report submitted to the Faculty of Health Sciences, University of the Witwatersrand of Johannesburg, in partial fulfilment of the requirements for the Msc degree in the field of Epidemiology and Biostatistics November 2015Background: Worldwide, more people die annually from cardiovascular diseases than from any other cause. Evidence seems to support that these conditions account for nearly 30% of all deaths worldwide, with more than 10 million of deaths occurring among people aged less than 70 years old. Low and middle income countries are the most affected by these conditions, with proportions of more than 80% of cardiovascular diseases deaths occurring in these countries. Although reducing major risk factors such as tobacco use, obesity, hypertension and lifestyle are key steps toward an effective prevention of these man-made diseases, it is also highly recommended to document demographic and socioeconomic factors associated with these deaths. However, considerable uncertainties regarding the epidemiology of CVDs persist in the African continent. Aims: This study aims to investigate the demographic and socioeconomic factors associated with cardiovascular diseases mortality in South Africa from 2006 – 2011. Secondly, the study wanted to document geographic variations of CVDs deaths across South Africa. Methods: Using repeated cross-sectional data from the vital registration system in South Africa, we investigated demographic and socioeconomic factors associated with CVDs deaths in South Africa, using simple logistic regression model, Cox proportional hazard model and Cox frailty, accounting for a random effect at district level. The hazard ratios of the Cox shared frailty model were mapped to show geographical variation across the nine provinces. Results: 3,454,868 deaths were registered at the Department of Home Affairs from 2006 – 2011. Of these deaths, about 500,633 were attributable to CVDs. The final sample contained 598,013 deaths for which relevant data were available. Gender, age, race, education, tobacco use were associated with CVDs death. Males had 6% reduced risk for dying from CVDs (HR: 0.94; 95% C.I: 0.92 – 0.95). The risk for CVDs mortality increased with education level, confirming lifestyle effect among those with tertiary education (HR: 2.19; 95% C.I: 2.11 – 2.28). Smokers had 31% increased risk (HR: 1.31, 95% C.I: 1.29 – 1.33) for dying from CVDs. Most important significant geographic variations were displayed regarding CVDs mortality, with people who lived in Free State and Kwazulu Natal having the highest risk for dying from CVDs (HR: 1.17; 95% C.I: 1.05 – 1.30 and 1.19; 95% C.I: 1.07 – 1.31 respectively) compared to those who lived in Northern Cape. Conclusion: The results of this study highlight important socioeconomic and demographic differentials regarding CVDs mortality and provide supporting arguments for significant geographical variations of CVDs mortality at provincial level