Population and high-risk based approaches to reducing the burden of cardiovascular disease in low- and middle-income countries

Abstract

Cardiovascular diseases (CVD) are a major global health problem, with rising prevalence due to aging populations and improved survival rates. Preventive measures like lifestyle changes and medications are key. Two strategies to reduce CVD risk are population-based approaches for entire populations and high-risk approaches for individuals. However, the high-risk approach is challenging in low- and middle-income countries (LMICs) due to limited resources. Non-laboratory-based CVD risk prediction tools could help, though their effectiveness is not well understood. A combined approach considering local sociodemographic and environmental contexts may most effectively reduce CVD in LMICs. The thesis aims to compare the performance of non-laboratory-based and laboratory-based equations across different settings and populations. Additionally, it aims to quantify the disparities in CVD risk among disadvantaged groups by identifying variations and social determinants across different settings. The thesis consists of two systematic reviews to compare non-laboratory-based and laboratory-based CVD equations, and four empirical studies to identify the most susceptible groups for CVD risk. Systematic reviews included studies published until March 12, 2024, with 25 studies on the correlation and agreement between laboratory- and non-laboratory-based equations and nine studies on their prediction performance. The results show strong correlations and moderate agreement between laboratory- and non-laboratory-based equations, with substantial hazard ratios for additional predictors significantly altering predicted risk, particularly for individuals with higher or lower levels of these predictors compared to the average. For empirical studies, data from the WHO STEPS national survey in Ethiopia, Brunei Darussalam, Iraq, Nepal, Afghanistan, Jordan, Mongolia, and Sao Tome and Principe include sociodemographic factors, behavioural factors (tobacco and alcohol use, physical inactivity), and biological risk factors (blood pressure, blood glucose, and cholesterol). These data are linked to 30 years (1970-2000) of district-level climate history, with multivariable multilevel regression models and geospatial analysis used to identify individual- and community-level factors associated with ten-year CVD risk. Ten-year CVD risk is significantly higher among study participants who are widowed, divorced, or separated, reside in urban areas, have lower educational status, are retired or unable to work, engage in low physical activity, and live in communities with higher water vapor pressure climates. Conversely, lower ten-year CVD risks were associated with communities in inherently hotter climates. Hotspot areas for mean ten-year CVD risk were observed across various regions of Ethiopia, rather than being limited to a specific area. In addition to these hotspots, regions with higher annual water vapor pressure (humidity) were linked to higher ten-year CVD risk, offering insights for targeted CVD prevention strategies in more humid regions. The findings of this thesis indicate the validity of affordable and feasible CVD risk equations made for resource-limited settings, alongside the possible integration of non-laboratory-based methods within their healthcare systems. Moreover, the thesis identifies vulnerable groups susceptible to CVD risk disparities, including sociodemographic, behavioural, geographical, and environmental factors, and informs targeted prevention strategies