Assessment of 10-year cardiovascular disease risk in Bangladesh: a comparative analysis of laboratory and non-laboratory-based WHO risk prediction models

Abstract

Introduction. Cardiovascular diseases (CVDs) are a leading cause of death in low-income and middle-income countries. This study aimed to estimate the 10-year CVD risk in Bangladeshi adults using the updated 2019 WHO laboratory-based and non-laboratory-based models, identify key CVD risk determinants and assess agreement between the models.

Methods. We conducted a cross-sectional analysis of nationally representative the WHO STEPwise approach to NCD risk factor surveillance (STEPS) 2018 Bangladesh survey data, involving 2782 adults aged 40-69 years without prior CVDs, using both versions of the updated 2019 WHO 10-year CVD risk prediction models. Logistic regression identified determinants of elevated predicted 10-year CVD risk (≥10%) and model agreement was assessed using Bland-Altman plots, Lin's Concordance Correlation Coefficient (LCCC) and kappa statistics.

Results. In the laboratory-based model, 10.3% (95% CI 8.7% to 12.0%) of the population was at elevated risk (≥10%), compared with 9.2% (95% CI 7.7% to 11.0%) in the non-laboratory-based model. Risk was higher in males (p<0.001) in both models. Elevated risk in the laboratory-based model was associated with residing in Dhaka division (adjusted OR (aOR)=2.21, 95% CI 1.25 to 3.93), homemakers (aOR=0.29, 95% CI 0.16 to 0.51), unemployed/retired (aOR=7.18, 95% CI 3.84 to 13.44), high waist-hip ratio (aOR=2.10, 95% CI 1.36 to 3.25) and elevated low-density lipoprotein cholesterol (aOR=1.94, 95% CI 1.22 to 3.09). The non-laboratory-based model showed similar associations. Bland-Altman analysis (laboratory minus non-laboratory) showed a small mean difference of 0.18 with 95% limits of agreement from -3.15 to 3.51, with proportional bias (males: y=-0.43 + 1.11x; females: y=-0.43 + 1.16x). LCCC indicated high concordance (0.90, 95% CI 0.89 to 0.91), and the kappa indicated substantial agreement (0.69, 95% CI 0.67 to 0.72).

Conclusion. A substantial portion of Bangladeshi adults had elevated predicted 10-year CVD risk, with higher risk in males. Sociodemographic and metabolic factors indicate potential target groups for intervention. The non-laboratory-based model showed high concordance and substantial categorical agreement with the laboratory-based model and may support population screening where laboratory testing is limited, although caution is warranted at higher risk levels.