Leveraging a natural experiment to estimate the causal impact of blood sugar on dementia and cognition in India

点击翻译按钮获取中文摘要

Project Summary The global burden of dementia is growing substantially as low- and middle-income countries (LMICs) undergo demographic transitions resulting in rapidly aging populations. Understanding the factors that affect dementia prevalence in LMICs such as India and identifying cost-effective interventions that may curb the onset of dementia and cognitive decline is thus an important public health and policy priority. Evidence from several studies indicates that diabetes increases the risk of developing cognitive impairment and dementia, but few studies have used population-representative data from LMICs experiencing rapidly increasing diabetes rates. Furthermore, interventions that seek to change diets and reduce diabetes risk by targeting individuals directly have had mixed success and are costly to implement at scale. In contrast, policy-level solutions such as sugar-sweetened beverage taxes or regulation of food supply (e.g., caps on added sugars) have the potential to influence dietary patterns at scale, especially in low-income settings where resources are scarce. There is a vital need to identify large-scale policy solutions for reducing diabetes and dementia in LMICs. This project will use population-representative data on dementia in India to accomplish three aims. First, we will study the association between dementia and diabetes, a known risk factor, in a low-income setting with a high prevalence of diabetes. We will also study heterogeneity in this relationship by dietary practice and by sex. Second, we will evaluate a scalable public health intervention that may reduce HbA1c levels and diabetes rates: the delivery of parboiled rice through the Indian government’s Public Distribution System, which provides roughly one-third of the total grain consumed in the country. To accomplish this goal, we will leverage detailed data on the quasi-random distribution of parboiled rice in Kerala, a state with high rice consumption, as well as data on HbA1c and blood glucose from three data sources including the Harmonized Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD). Finally, we will leverage the quasi-random delivery of parboiled rice to study the causal effect of diabetes on dementia and cognitive decline using an instrumental variables approach.