Background: Socioeconomic circumstances during infancy and in early childhood shape later developmental opportunities. Understanding whether inequality in children’s health becomes stronger with age from the evidence of longitudinal data is an important first step in revealing the mechanism by which the intergenerational transmission of poverty takes place and evolves. Objective: This study investigates the 15-year trajectory of inequality in height growth associated with early-life circumstances among children in Vietnam, Peru, Ethiopia, and India. Methods: This study uses the datasets of the Young Lives study project, which is a large-scale crosscountry longitudinal study on childhood poverty conducted in Vietnam, Peru, Ethiopia, and India since 2002. Machine learning approaches are employed to estimate the relationship between early-life circumstances and child height. Results: The inequality in height stemming from the difference in early-life circumstances persists even after children reach early adolescence. The proportion of inequality peaks when children are 5 years old. Our prediction using the random forest model shows that, if we were able to fully compensate for early-life socioeconomic disadvantages, we could increase the lower percentile of the height distribution and reduce the inequality in height at age 15 by half. Conclusions: The results suggest that children from marginalised households should be supported at the earlier developmental stage. Contribution: This study shows the life-course evolution of inequalities in child growth, explores the dynamic link between early-life circumstances and later consequences, and identifies optimal timing in terms of when circumstances and events matter the most.

中文翻译：

---

儿童身高增长中机会不平等的轨迹：来自年轻生命研究的证据

背景：婴儿期和幼儿期的社会经济状况影响了以后的发展机会。从纵向数据的证据了解儿童健康的不平等是否会随着年龄的增长而加剧，这是揭示贫困代代相传的发生和发展机理的重要的第一步。目的：本研究调查了越南，秘鲁，埃塞俄比亚和印度儿童身高增长不均等的15年轨迹，该高度与儿童的早期生活状况有关。方法：本研究使用“年轻生命”研究项目的数据集，该项目是自2002年以来在越南，秘鲁，埃塞俄比亚和印度进行的大规模跨国纵向儿童贫困研究。使用机器学习方法来估计早期生活状况和儿童身高之间的关系。结果：即使在儿童进入青春期后，由于早期生活状况的差异而导致的身高不平等现象仍然存在。当儿童5岁时，不平等的比例达到顶峰。我们使用随机森林模型进行的预测表明，如果我们能够完全弥补早期生命的社会经济劣势，则可以将身高分布的较低百分比提高，并将15岁时身高的不平等程度降低一半。结论：研究结果表明，边缘化家庭的儿童应在早期发育阶段得到支持。贡献：这项研究显示了儿童成长中不平等的人生历程演变，