The determinants of children nutritional choices & behavioural interventions for food behaviour change.

公共卫生面临的最大压力是非传染性疾病的持续增加(世界卫生组织，2018年)。83.6%的非传染性疾病死亡是由于暴露于我们环境中易受政策干预的因素造成的(RapPaport，2016)。非传染性疾病的一个关键风险因素与与食物有关的行为有关。公共卫生政策在很大程度上是基于教育或知识转让，并没有对营养不良和肥胖产生重大影响。一个关键的挑战是影响肥胖相关行为的因素的复杂性。有效的干预措施需要全面和全面地了解这些因素在驱动与食物有关的行为方面的相互作用。很少有人试图系统地将社会经济、行为、生物和建筑环境数据和分析的不同来源联系起来并进行三角测量，以便为将影响公众健康的变化提供信息。过去的传统方法几乎完全依赖于部分独立的数据集，针对特定人群进行一次性干预，仅源自狭隘的研究假设。然而，在制定具有长期持续影响的行为改变干预措施方面，这方面的使用相当有限。这些方法未能捕捉到塑造个人健康危险行为的行为的动态及其对健康的影响，以及这些行为如何受到与环境的复杂相互作用的影响。我们建议的方法是利用一个独特的纵向数据集，其中包含关于个人健康结果、行为和特征的关联数据，以及捕捉行为的多层次决定因素的环境地图，以更好地了解营养决定因素，并设计、实施和评估斯里兰卡学校食品环境中的行为干预措施。斯里兰卡近29%的成年人超重或肥胖，73%的人口每天没有摄入足够的水果和/或蔬菜(世界卫生组织，2015)。只有14.7%的儿童(一般男孩和女孩)每天吃5份或更多的水果和/或蔬菜(Darfour-Oduo等人，2018年)。此外，患有高血压、胰岛素抵抗和血脂异常等通常由肥胖引起的合并症的儿童不断增加(Wickramasinghe等人，2013年)。同样重要的是以儿童为目标，因为儿童时期的肥胖已被发现影响成人肥胖、人力资本(学校水平和认知技能)和社会结果(社会排斥和耻辱)等终生结果(Bhadoria等人，2015年)。为此，我们将利用全球糖尿病和心血管疾病健康研究单位(GHRU)队列研究中开发的基础设施，该研究一直在为成人收集的现有南亚生物库的约250个监测点收集环境和监测措施。我们建议以儿童数据补充成人监测数据，并将其与个人层面的环境测绘数据联系起来(使用住所和学校位置)，包括每个监测点的食物和体育活动设施的位置、提供和推广，并覆盖学校200米的缓冲区。

The greatest strain on public health is the continued increase in non-communicable diseases (NCDs) (World Health Organization, 2018). 83.6% of NCDs mortality is caused by exposure to factors in our environment that are amenable to policy interventions (Rappaport, 2016). A key risk factor of NCDs relates to food related behaviours. Public health policies, which are in large based on education or knowledge transfer, have not generated a significant impact on poor nutrition and obesity. A key challenge is the complexity of factors that affect obesity related behaviours. Effective interventions require a comprehensive and holistic understanding of the interplay of these factors in driving food related behaviours. Few attempts have been made to systematically link and triangulate different sources of socio-economic, behavioural, biological and exposure to the built environment data and analyses that can inform change that will affect public health. Past traditional approaches almost exclusively relied on partial separate datasets, targeting specific groups of the population with one off interventions, only derived from narrow research hypotheses. As such are however quite limited in use for the development of behaviour change interventions with a sustained impact over time. Those approaches fail to capture the dynamics of behaviours that shape individual's health risky behaviours and their impact on health, and how these are influenced by complex interactions with the environments. Our proposed approach is to leverage a unique longitudinal dataset with linked data on individual's health outcomes, behaviours and characteristics, and environmental mapping capturing the multilevel determinants of behaviours to develop a better understanding of nutrition determinants and design, implement and evaluate behavioural interventions in school food environments in Sri Lanka. Nearly 29% of the adults in Sri Lanka are overweight or obese, 73% of the population do not consume the sufficient servings of fruits and/or vegetables per day (World Health Organization, 2015). Only 14.7% of children (average boys and girls) eat five or more servings of fruit and/or vegetables per day (Darfour-Oduro et al, 2018). Also, there has been a constant increase in children with comorbidities such as high blood pressure, insulin resistance and dyslipidaemia which are often caused by obesity (Wickramasinghe et al., 2013). It is also important to target children as obesity in childhood has been found to impact lifetime outcomes such as adult obesity, human capital (school proficiency and cognitive skills), and social outcomes (social exclusion and stigma) (Bhadoria et al., 2015). To do so we will leverage the infrastructure developed in the Global Health Research Unit for Diabetes and Cardiovascular disease (GHRU) cohort study that has been gathering environmental and surveillance measures within ~250 surveillance sites of an established South Asia Biobank collected for adults. We propose to complement the adult surveillance data with data on children and link it to environmental mapping data at individual level (using residence and school location) including location, provision, and promotion of food and physical activity facilities in each surveillance site and covers 200-meter buffer from the school.