Biostatistics and Epidemiology Division, Research Triangle Institute (RTI) International, Durham, NC, USA; Children's Environmental Health Initiative, University of Notre Dame, South Bend, IN, USA.
Children's Environmental Health Initiative, University of Notre Dame, South Bend, IN, USA; Department of Applied and Computational Mathematics and Statistics, University of Notre Dame, South Bend, IN, USA.
Environ Res. 2021 Jul;198:111241. doi: 10.1016/j.envres.2021.111241. Epub 2021 Apr 30.
Persistent disparities in academic performance may result from a confluence of adverse exposures accruing disproportionately to specific subpopulations.
Our overarching objective was to investigate how multiple exposures experienced over time affect early childhood educational outcomes. We were specifically interested in whether there were: racial/ethnic disparities in prevalence of adverse exposures; racial/ethnic disparities in associations observed between adverse exposures and early childhood educational outcomes; and interactions between exposures, suggesting that one exposure augments susceptibility to adverse effects of another exposure.
We link geocoded North Carolina birth data for non-Hispanic white (NHW) and non-Hispanic black (NHB) children to blood lead surveillance data and 4th grade end-of-grade (EOG) standardized test scores (n = 65,151). We construct a local, spatial index of racial isolation (RI) of NHB at the block group level. We fit race-stratified multi-level models of reading and mathematics EOG scores regressed on birthweight percentile for gestational age, blood lead level, maternal smoking, economic disadvantage, and RI, adjusting for maternal- and child-level covariates and median household income.
There were marked racial/ethnic disparities in prevalence of adverse exposures. Specifically, NHB children were more likely than NHW children to be economically disadvantaged (80% vs. 40%), live in block groups with the highest quintile of RI (46% vs. 5%), have higher blood lead levels (4.6 vs. 3.7 μg/dL), and lower birthweight percentile for gestational age (mean: 39th percentile vs. 51st percentile). NHB children were less likely to have mothers who reported smoking during pregnancy (11% and 22%). We observed associations between key adverse exposures and reading and math EOG scores in 4th grade. Higher birthweight percentile for gestational age was associated with higher EOG scores, while economic disadvantage, maternal smoking, and elevated blood lead levels were associated with lower EOG scores. Associations observed for NHB and NHW children were generally not statistically different from one another, with the exception of neighborhood RI. NHB children residing in block groups in the highest RI quintile had reading and math scores 1.54 (0.74, 2.34) and 1.12 (0.38, 1.87) points lower, respectively, compared to those in the lowest RI quintile; statistically significant decrements in EOG scores associated with RI were not observed for NHW children. We did not find evidence of multiplicative interactions between exposures for NHB or NHW children.
Key adverse host, environmental, and social exposures accrue disproportionately to NHB children. Decrements in test scores associated with key adverse exposures were often but not always larger for NHB children, but were not significantly different from those estimated for NHW children. While we did not observe interactive effects, NHB children on average experience more deleterious combined exposures, resulting in larger decrements to test scores compared to NHW children.
学术表现的持续差距可能是由于特定亚人群中不成比例地积累了不利暴露的共同作用。
我们的总体目标是研究随着时间的推移，多种暴露如何影响儿童早期的教育成果。我们特别关注是否存在以下情况：不利暴露的发生率在种族/民族之间存在差异；不利暴露与儿童早期教育成果之间观察到的关联在种族/民族之间存在差异；以及暴露之间的相互作用，表明一种暴露会增加另一种暴露的不良影响的易感性。
我们将非西班牙裔白人（NHW）和非西班牙裔黑人（NHB）儿童的地理编码北卡罗来纳州出生数据与血铅监测数据和 4 年级末（EOG）标准化考试成绩（n=65151）相关联。我们构建了一个当地的、以块组为单位的 NHB 种族隔离（RI）空间指数。我们拟合了阅读和数学 EOG 成绩的种族分层多水平模型，这些成绩回归于胎龄的出生体重百分位数、血铅水平、母亲吸烟、经济劣势和 RI，同时调整了母亲和儿童水平的协变量以及家庭中位数收入。
不利暴露的发生率在种族/民族之间存在明显差异。具体而言，NHB 儿童比 NHW 儿童更有可能处于经济劣势（80%比 40%）、居住在 RI 最高五分位数的块组中（46%比 5%）、血铅水平更高（4.6 比 3.7μg/dL），胎龄出生体重百分位数更低（平均值：39 百分位比 51 百分位）。NHB 儿童的母亲在怀孕期间吸烟的比例较低（11%和 22%）。我们观察到关键不利暴露与 4 年级 EOG 阅读和数学成绩之间的关联。胎龄出生体重百分位数较高与 EOG 成绩较高相关，而经济劣势、母亲吸烟和血铅水平升高与 EOG 成绩较低相关。NHB 和 NHW 儿童观察到的关联通常彼此之间没有统计学差异，除了邻里 RI。与 RI 最低五分位数相比，居住在 RI 最高五分位数块组中的 NHB 儿童的阅读和数学成绩分别低 1.54（0.74，2.34）和 1.12（0.38，1.87）分；与 RI 相关的 EOG 分数的统计学显著下降在 NHW 儿童中并未观察到。我们没有发现 NHB 或 NHW 儿童之间暴露的乘法相互作用的证据。
关键的不利宿主、环境和社会暴露不成比例地累积在 NHB 儿童身上。与关键不利暴露相关的考试成绩下降，对于 NHB 儿童来说往往更大，但与对 NHW 儿童的估计没有显著差异。虽然我们没有观察到交互作用，但 NHB 儿童平均经历更多有害的综合暴露，导致与 NHW 儿童相比，考试成绩下降更大。
