Risk factors for cardiovascular disease and prediction of coronary artery disease in a rural Bangladeshi population

Abstract

Background/Aim: Non-communicable diseases, in particular cardiovascular diseases (CVDs), have become the leading cause of death in Bangladesh. This thesis contributes with new information to the understanding of the epidemiology of CVDs through a series of studies which: review the existing evidence on CVD risk factors in Bangladesh; explore and characterize individual risk factors as well as CVD absolute risk scores in a rural population; and explore options of absolute risk prediction models applicable for this population.

Methods: A systematic review and meta-analysis to explore the prevalence of CVD risk factors followed by three cross-sectional and one case-cohort study were conducted on a peripherally located rural Bangladeshi population. One cohort study (N=66,710) examined demographic and anthropometric risk factors; one case-control study (cases=1170, control=563) examined the clinical and biochemical risk factors of CVDs; and one cross-sectional study estimated the absolute CVD risk of 563 disease-free participants. The case-cohort study (sub-cohort=563, main cohort=52989, 2.5 years of follow-up) examined risk factors for coronary heart diseases (CHDs), and compared the performance of four CVD risk prediction models in the study population.

Findings: The review and meta-analysis suggest that CVD risk is high among Bangladeshi people. Even in the rural population studied there was a substantial prevalence of anthropometric, clinical, and biochemical risk factors of CVDs. On absolute risk scoring, approximately one-fifth of both male and female adults are at moderate or high risk of developing CVDs within 10 years. The 'with' and 'without' cholesterol versions of the CVD risk prediction tool were concordant. On short-term testing for the outcome of coronary artery disease as diagnosed on electrocardiogram, the laboratory and non-laboratory based absolute risk scoring models had similar predictive accuracy.

Conclusion: Our data suggest that rural Bangladeshis are at substantial risk for CVDs. Rigorous methods in study design and analyses are the main strengths, whereas, lack of representativeness in some studies included in the review, the small number of cases, short duration of follow-up, and probable inaccuracy of case ascertainment in the case-cohort study are the main limitations. Despite these limitations, the studies provide important insights into the prevalence of CVD risk and its prediction among rural Bangladeshis and thus it will make a significant contribution towards management and prevention of CVDs in this population.