Small Area Geographic Estimates of Cardiovascular Disease Risk Factors in India

Abstract

Importance: With an aging population, India is facing a growing burden of cardiovascular diseases (CVDs). Existing programs on CVD risk factors are mostly based on state and district data, which overlook health disparities within macro units. Objective: To quantify and geovisualize the extent of small area variability within districts in CVD risk factors (hypertension, diabetes, and obesity) in India. Design, Setting, and Participants: This cross-sectional study analyzed nationally representative data from the National Family Health Survey 2019-2021, encompassing individuals aged 15 years or older, for hypertension (n = 1715895), diabetes (n = 1807566), and obesity (n = 776023). Data analyses were conducted from July 1, 2022, through August 1, 2023. Exposures: Geographic units consisting of more than 30000 small areas, 707 districts, and 36 states or Union Territories across India. Main Outcomes and Measures: For primary outcomes, CVD risk factors, including hypertension, diabetes, and obesity, were considered. Four-level logistic regression models were used to partition the geographic variability in each outcome by state or Union Territory (level 4), district (level 3), and small area (level 2) and compute precision-weighted small area estimates. Spatial distribution of district-wide means, within-district small area variability, and their correlation were estimated. Results: The final analytic sample consisted of 1715895 individuals analyzed for hypertension (mean [SD] age, 39.8 [17.3] years; 921779 [53.7%] female), 1807566 for diabetes (mean [SD] age, 39.5 [17.2] years; 961977 [53.2%] female), and 776023 for obesity (mean [SD] age, 30.9 [10.2] years; 678782 [87.5%] women). Overall, 21.2% of female and 24.1% of male participants had hypertension, 5.0% of female and 5.4% of men had diabetes, and 6.3% of female and 4.0% of male participants had obesity. For female participants, small areas (32.0% for diabetes, 34.5% for obesity, and 56.2% for hypertension) and states (30.0% for hypertension, 46.6% for obesity, and 52.8% for diabetes) accounted for the majority of the total geographic variability, while districts accounted for the least (13.8% for hypertension, 15.2% for diabetes, and 18.9% for obesity). There were moderate to strong positive correlations between district-wide mean and within-district variability (r = 0.66 for hypertension, 0.94 for obesity, and 0.96 for diabetes). For hypertension, a significant discordance between district-wide mean and within-district small area variability was found. Results were largely similar for male participants across all categories. Conclusions and Relevance: This cross-sectional study found a substantial small area variability, suggesting the necessity of precise policy attention specifically to small areas in program formulation and intervention to prevent and manage CVD risk factors. Targeted action on policy-priority districts with high prevalence and substantial inequality is required for accelerating India's efforts to reduce the burden of noncommunicable diseases..