Cost-effective chronic disease prevention in developing countries

说明(申请人提供)：尽管人们认为传染病和营养不良是发展中国家健康的主要威胁，但现实情况是，慢性病的流行正在迅速增加。到2030年，慢性疾病预计将导致发展中国家40%以上的死亡，这促使世界卫生组织在2008年呼吁制定预防和控制慢性病的国家行动计划。这样的计划必须克服巨大的资源限制。由于慢性病与持续性贫困疾病重叠，政策应对措施必须与多种疾病相互作用的风险作斗争。例如，通过减少肥胖的社会政策控制2型糖尿病，必须避免加剧营养不良，特别是在儿童中普遍存在的营养不良。相反，适当针对性的糖尿病预防也可以降低结核病患病率。在像印度这样的国家，糖尿病与活动性结核病感染风险增加的生物学联系估计占总结核病发病率的15%。然而，糖尿病在中等收入群体中更普遍，而结核病在穷人中更常见，这一点并不明显。目前的政策评估方法不足以解决这些问题，因为典型的分析侧重于解决单一疾病的健康益处，因此没有考虑到重要的跨疾病影响。需要一种新的方法，根据其对多种疾病的特定年龄和亚群的影响，评估每种替代政策的生命过程惠益。此K01应用程序概述了这样一种方法。其培训目的是扩大以下领域的知识和技能：(1)用于分析大型家庭调查、非随机自然减员的计量经济和统计技术，以及影响食品价格、消费和营养健康结果的税收等社会政策；(2)生命历程、导致慢性疾病的发展过程，如糖尿病、心脑血管疾病，以及结核病和营养不良等持续性贫困疾病的生物学和流行病学。这将通过相关的课程作业、有指导的定向阅读和研讨会参与来实现。这一K01发展了具体的例子，其逻辑延伸到其他国家和疾病表明其深远的适用性超出了提案的范围，因此具有很高的公共卫生意义。开发的方法虽然适用范围很广，但将应用于印度糖尿病的具体问题。具体的研究目标是：(1)开发2型糖尿病的终生微观模拟模型，以评估亚裔印度人的一级和二级预防干预措施；(2)评估营养政策对慢性病和营养不良的影响；(3)将结核病传播纳入糖尿病模型，以评估糖尿病患病率上升如何通过生物交互作用改变结核病患病率；(4)比较针对患有不同糖尿病和结核病风险状况的亚群的预防政策的益处。为了确定模型的投入，将使用计量回归和统计校准。

DESCRIPTION (provided by applicant): Despite the perception that infectious diseases and malnutrition are the main threats to health in the developing world, the reality is that the prevalence of chronic diseases is increasing rapidly. By 2030, chronic diseases are projected to cause more than 40% of deaths in developing countries, prompting a 2008 call from the World Health Organization to establish national action plans for their prevention and control. Such plans must overcome formidable resource constraints. Because chronic diseases overlap with persistent diseases of poverty, policy responses must contend with interacting risks from multiple diseases. For instance, the control of type 2 diabetes through social policies to reduce obesity must avoid exacerbating malnutrition, especially prevalent in children. Conversely, properly targeted diabetes prevention can also reduce tuberculosis prevalence. In countries like India, diabetes' biological link to increased risk of active tuberculosis infection accounts for an estimated 15% of overall tuberculosis incidence. It is, however, not obvious which groups to target as diabetes is more prevalent in middle income groups while tuberculosis is more common in the poor. Current policy assessment methods are insufficient for these problems, as typical analyses focus on the health benefits of addressing a single disease and, therefore, fail to account for important cross-disease effects. A novel approach is required that evaluates the life course benefits of each policy alternative based on its age- and subgroup-specific impact on multiple diseases. This K01 application outlines such an approach. Its training aims are to expand knowledge and skills in: (1) econometric and statistical techniques for analyzing large household surveys, longitudinal studies with non-random attrition, and social policies such as taxes that impact food prices, consumption, and nutritional health outcomes; (2) the life course, developmental processes leading to chronic diseases such as diabetes, cardiovascular and cerebrovascular diseases as well as the biology and epidemiology of persistent diseases of poverty like tuberculosis and malnutrition. This will be accomplished through relevant coursework, mentored directed reading, and seminar participation. This K01 develops specific examples whose logical extension to other countries and diseases indicates far-reaching applicability beyond the scope of the proposal and, hence, high public health significance. The methods developed, though broadly applicable, will be applied to the specific problem of diabetes in India. The specific research aims are: (1) develop a lifetime microsimulation model of type 2 diabetes to assess primary and secondary prevention interventions in Asian Indians; (2) evaluate the impact of nutrition policies on chronic diseases and malnutrition; (3) integrate tuberculosis transmission into the diabetes model to assess how rising diabetes prevalence alters tuberculosis prevalence via their biological interaction; (4) compare the benefits of prevention policies that target subpopulations with different diabetes and tuberculosis risk profiles. In order to determine model inputs, econometric regression and statistical calibration will be used.