How do prenatal and postnatal circumstances interact in shaping health? An interdisciplinary approach using quasi-experiments

有很强的理由认为，出生前和出生后的环境在塑造健康方面存在交互作用，但这一点在人类身上几乎没有研究。我们开发了一个研究这种影响的框架，并在一系列关于特定产前/出生后相互作用的研究中对此进行了严格的测试。我们的研究结合了医学/流行病学和计量经济学-统计学方法。产前阶段对健康具有长期影响。胎儿编程理论描述了在关键发育阶段暴露在环境环境中如何引发适应，从而导致有机体生理的永久性变化。对人类的研究集中在广泛的不利环境，如饥荒、污染和有毒物质、压力和非饥荒营养不足，并显示出负面的长期健康影响。动物研究表明，生物体不仅在生理上适应不利环境。也就是说，在产前，生物体具有一定的表型可塑性，使它们能够适应环境，例如通过表观遗传变化。当产前环境可以作为对出生后环境的正确预测时，这可能是有益的。在这个意义上，这种预测性的适应性反应(PARs)使物种能够迅速适应它可能面临的环境的变化。然而，如果妊娠期间遇到的环境与后代所遇到的出生后环境有很大的不同，则妊娠结局可能会产生有害的影响。人类早期生活影响的研究通常集中在不利的产前环境如何导致不利的晚年健康结果，通常忽略了与出生后环境的潜在交互作用。这可能是因为不可能在人类身上做实验：研究人员通过利用划分清楚、相对较短的-明显不利的-历史时期，如饥荒来解决这个问题。这种情况不太可能在以后的生活中出现类似的情况，因此足以确定产前环境造成了长期的健康损害，而不需要研究出生前和出生后的情况是否匹配或不匹配。我们的项目是一名应用计量经济学家和流行病学家/医学研究人员之间的跨学科合作，他们专门从事自然/准实验在健康领域的应用，他们已经成功地合作。我们将使用观测数据结合自然实验和自然实验来研究产前/出生后的相互作用，自然实验导致出生前环境的准随机变化，自然实验导致出生后环境的准随机变化。我们包括五个研究项目，这些项目共同为我们的研究框架提供了坚实的测试。我们将使用这些项目来展示产前/产后互动是否以及如何塑造人类健康，并相应地进一步完善这一框架。

There are strong reasons to expect that prenatal and postnatal circumstances interact in shaping health, but this has hardly been studied in humans. We develop a framework for studying such effects and rigorously test this in a series of studies on specific prenatal/postnatal interactions. Our research combines medicine/epidemiology with an econometric-statistical approach.The prenatal phase is known to have long-run implications on health. Fetal programming theory describes how exposures to environmental circumstances during critical developmental phases set off adaptations that result in permanent changes to an organism’s physiology. Research on humans has focused on a wide range of adverse circumstances such as famines, pollution and toxic substances, stress, and non-famine nutritional deficits and showed negative long-run health effects.Animal studies demonstrate that organisms adapt physiologically not only in response to adverse circumstances. I.e. prenatally, organisms have a certain phenotypic plasticity that allows them to adapt to their environment, e.g. via epigenetic changes. This can be beneficial when prenatal circumstances serve as correct predictions for postnatal environments. In this sense, such Predictive Adaptive Responses (PARs) allow a species to quickly adapt to variations in circumstances it may be confronted with. If the environment encountered during gestation however differs strongly from the postnatal environment encountered by the offspring, PARs may have harmful effects.Human early-life effects studies usually focused on how adverse prenatal circumstances lead to adverse later-life health outcomes, generally ignoring potential interactions with postnatal circumstances. This may be due to the impossibility to do experiments on humans: researchers solved this by utilizing clearly demarcated, relatively short – clearly adverse – historical periods, such as famines. Such circumstances were unlikely to be followed by similar circumstances later in life, so that sufficed to establish that the prenatal environment caused long-run health damage, without a need to study whether it makes a difference whether pre- and postnatal circumstances match or mismatch.Our project is an interdisciplinary cooperation between an applied econometrician, who is specialized in the application of natural/quasi experiments in the health domain, and an epidemiologist/medical researcher, who have already been working together successfully. We will study prenatal/postnatal interactions using observational data by combining natural experiments that lead to quasi-random variation in prenatal circumstances with natural experiments that lead to quasi-random variation in postnatal circumstances. We include five research projects, that together provide a solid test of our research framework. We will use these projects to demonstrate whether and how prenatal/postnatal interactions shape human health, and to accordingly further refine this framework.