空气污染存在明显的环境公正问题。然而，目前对于交通相关空气污染（TRAP）的研究还很少，而这种污染源是由历史上种族主义基础设施政策所驱动的，而孕妇这一易受空气污染影响的人群在其中更具敏感性。这些不平等现象随着时间的推移如何变化尚不清楚，但对于政策具有重要意义。本研究旨在调查过去20年内，按社会人口统计特征的变化如何影响到TRAP暴露情况。该以居民为基础的出生队列研究采用描述性分析方法，研究对象为德克萨斯州在1996年至2016年期间的孕妇。所有具有有效居住地址、社会经济和人口统计数据的孕妇均纳入研究。个体层面的种族和民族、教育和母亲出生地数据从出生证明中提取，而街区层面的家庭收入和历史街区投资不足（即红线政策）数据则通过居住地址进行评估。数据分析时间为2022年6月至2023年6月。主要结果以居住地址为基准评估TRAP的暴露水平，其中包括交通量水平（以500米范围内的总体和卡车行驶里程表示）、一种空间-时间土地利用回归模型测得的二氧化氮（NO2）浓度（即车辆尾气排放）以及在路车辆排放引起的全国空气有毒物质署癌症风险指数。通过社会人口统计指标分析了1996年至2016年期间TRAP暴露的差异。在1996年至2016年期间的德克萨斯州共有7,043,598名孕妇（平均[标准差]孕妇年龄为26.8 [6.1]岁），其中有48%为西班牙裔或拉丁美洲裔、4%为非西班牙裔亚洲裔或太平洋岛民、12%为非西班牙裔非洲裔和36%为非西班牙裔白人。通过研究的所有社会人口统计变量，孕妇的TRAP存在差异。这些差异的绝对水平在1996年至2016年期间有所降低，但相对水平则有所增加：例如，在1996年，非西班牙裔非洲裔孕妇暴露于平均（标准差）15.3（4.1）ppb的NO2，并与非西班牙裔白人孕妇相比暴露于13.5（4.4）ppb的NO2，相较之下，至2016年，非西班牙裔非洲裔孕妇暴露于平均（标准差）6.7（2.4）ppb的NO2，而白人孕妇暴露于平均（标准差）5.2（2.4）ppb的NO2。关于所有社会人口统计特征的交通量存在着大的绝对差异和相对差异，并且这些差异随时间增加。例如，2016年，非西班牙裔非洲裔孕妇暴露于其住处500米范围内的平均（标准差）22,836（32,844）VMT，而非西班牙裔白人孕妇的住处500米范围内有平均（标准差）12,478（22,870）VMT，差异达到了83%。这项出生队列研究发现，在研究的20年期间，空气污染不平等的水平在绝对意义上有所下降，但相对差异仍然存在，并且交通量方面存在较大差异，需要重新关注相关政策。

Air pollution presents clear environmental justice issues. However, few studies have specifically examined traffic-related air pollution (TRAP), a source driven by historically racist infrastructure policies, among pregnant individuals, a population susceptible to air pollution effects. How these disparities have changed over time is also unclear but has important policy implications.To examine changes in TRAP exposure by sociodemographic characteristics among recorded pregnancies over a 20-year period.This population-based birth cohort study used descriptive analysis among pregnant individuals in Texas from 1996 to 2016. All pregnant individuals with valid residential address, socioeconomic, and demographic data were included. Individual-level race and ethnicity, education, and maternal birthplace data were extracted from birth certificates and neighborhood-level household income and historical neighborhood disinvestment (ie, redlining) data were assessed via residential addresses. Data analysis occurred between June 2022 and June 2023.The main outcome, TRAP exposure at residential addresses, was assessed via traffic levels, represented by total and truck-specific vehicle miles traveled (VMT) within 500 m; nitrogen dioxide (no2) concentrations from a spatial-temporal land use regression model (ie, vehicle tailpipe emissions); and National Air Toxic Agency cancer risk index from on-road vehicle emissions. TRAP exposure differences were assessed by sociodemographic indicators over the 1996 to 2016 period.Among 7 043 598 pregnant people (mean [SD] maternal age, 26.8 [6.1] years) in Texas from 1996 to 2016, 48% identified as Hispanic or Latinx, 4% identified as non-Hispanic Asian or Pacific Islander, 12% identified as non-Hispanic Black, and 36% identified as non-Hispanic White. There were differences in TRAP for pregnant people by all sociodemographic variables examined. The absolute level of these disparities decreased from 1996 to 2016, but the relative level of these disparities increased: for example, in 1996, non-Hispanic Black pregnant individuals were exposed to a mean (SD) 15.3 (4.1) ppb of no2 vs 13.5 (4.4) ppb of no2 for non-Hispanic White pregnant individuals, compared with 2016 levels of 6.7 (2.4) ppb no2 for Black pregnant individuals and 5.2 (2.4) ppb of no2 for White pregnant individuals. Large absolute and relative differences in traffic levels were observed for all sociodemographic characteristics, increasing over time. For example, non-Hispanic Black pregnant individuals were exposed to a mean (SD) of 22 836 (32 844) VMT within 500 m of their homes, compared with 12 478 (22 870) VMT within 500 m of the homes of non-Hispanic White pregnant individuals in 2016, a difference of 83%.This birth cohort study found that while levels of air pollution disparities decreased in absolute terms over the 20 years of the study, relative disparities persisted and large differences in traffic levels remained, requiring renewed policy attention.