Clark Cassandra J, Warren Joshua L, Saiers James E, Ma Xiaomei, Bell Michelle L, Deziel Nicole C
Yale School of Public Health, Department of Environmental Health Sciences, 60 College St., New Haven, CT, 06510, USA.
Yale School of Public Health, Department of Biostatistics, 60 College St., New Haven, CT, 06510, USA.
J Expo Sci Environ Epidemiol. 2024 Nov;34(6):990-999. doi: 10.1038/s41370-023-00636-9. Epub 2023 Dec 26.
Residential mobility can introduce exposure misclassification in pediatric epidemiology studies using birth address only.
We examined whether residential mobility varies by sociodemographic factors and urbanicity/rurality among children with cancer.
Our study included 400 children born in Pennsylvania during 2002-2015 and diagnosed with leukemia at ages 2-7 years. Addresses were obtained from state registries at birth and diagnosis. We considered three aspects of mobility between birth and diagnosis: whether a child moved, whether a mover changed census tract, and distance moved. We evaluated predictors of these aspects in urban- and rural-born children using chi-square, t-tests, and regression analyses.
Overall, 58% of children moved between birth and diagnosis; suburban/rural-born children were more likely to move than urban-born children (67% versus 57%). The mean distance moved was 16.7 km in suburban/rural-born and 14.8 km in urban-born movers. In urban-born children, moving between birth and diagnosis was associated with race, education, participation in the Nutrition Program for Women, Infants and Children (WIC), and census tract-level income (all χ p < 0.01). Urban-born movers tended to be born in a census tract with a higher Social Vulnerability Index than non-movers (t-test p < 0.01). No factors were statistically significantly associated with any of the residential mobility metrics in suburban/rural-born children, although the sample size was small.
In this study of a vulnerable population of children with cancer, we found that rural-born children were more likely to move than urban-born children, however, the frequency of movers changing census tracts was equivalent. Mobility in urban-born children, but not rural-born, was associated with several social factors, although the sample size for rural-born children was small. Mobility could be an important source of misclassification depending on the spatial heterogeneity and resolution of the exposure data and whether the social factors are related to exposures or health outcomes. Our results highlight the importance of considering differences in mobility between urban and rural populations in spatial research.
在儿科流行病学研究中，仅使用出生地址可能会导致居住流动性造成暴露误分类。
我们研究了癌症患儿的居住流动性是否因社会人口学因素以及城市/农村属性而异。
我们的研究纳入了2002年至2015年在宾夕法尼亚州出生、2至7岁时被诊断为白血病的400名儿童。出生和诊断时的地址来自州登记处。我们考虑了出生到诊断期间流动性的三个方面：孩子是否搬家、搬家者是否更换普查区以及搬家距离。我们使用卡方检验、t检验和回归分析评估了城市和农村出生儿童这些方面的预测因素。
总体而言，58%的儿童在出生到诊断期间搬家；郊区/农村出生的儿童比城市出生的儿童更有可能搬家（67%对57%）。郊区/农村出生的搬家者平均搬家距离为16.7公里，城市出生的搬家者为14.8公里。在城市出生的儿童中，出生到诊断期间搬家与种族、教育程度、参与妇女、婴儿和儿童营养计划（WIC）以及普查区层面的收入有关（所有χ²检验p<0.01）。城市出生的搬家者往往出生在社会脆弱性指数高于非搬家者的普查区（t检验p<0.01）。在郊区/农村出生的儿童中，尽管样本量较小，但没有因素与任何居住流动性指标有统计学上的显著关联。
在这项针对癌症患儿这一弱势群体的研究中，我们发现农村出生的儿童比城市出生的儿童更有可能搬家，然而，搬家者更换普查区的频率相当。城市出生儿童的流动性与几个社会因素有关，而农村出生儿童则不然，尽管农村出生儿童的样本量较小。根据暴露数据的空间异质性和分辨率以及社会因素是否与暴露或健康结果相关，流动性可能是误分类的一个重要来源。我们的结果强调了在空间研究中考虑城市和农村人口流动性差异的重要性。
