Table 1 Illustrative papers that used race/ethnicity in relation to air pollution exposure or effects on health
From: Inferential challenges when assessing racial/ethnic health disparities in environmental research
Study | Aims | Outcome | Air pollutants | How race/ethnicity variables were used |
|---|---|---|---|---|
R/E as a confounder | ||||
Nobles et al., 2019 [50] | Impact of air pollution on fetal growth restriction | Physician diagnosed fetal growth restriction | SO2, O3, NOX, NO2, CO, PM10, PM2.5 | As a confounder, maternal R/E was included in models with maternal age, race/ethnicity, pre-pregnancy body mass index, smoking, alcohol, parity, insurance, marital status, asthma and temperature. |
McGuinn et al., 2019 [51] | Impact of air pollution on cardiovascular disease risk | Lipoprotein levels | PM2.5 | As a confounder, R/E was included in models with age, sex, history of smoking, area-level education, urban/rural status, body mass index, and diabetes. |
Bragg-Gresham et al., 2018 [52] | Impact of air pollution on the prevalence of diagnosed chronic kidney disease in US medicare population | Chronic kidney disease | PM2.5 county level | As a confounder, R/E was included in models with age, sex, hypertension, diabetes, and urban/rural status. |
Ng et al., 2017 [53] | Impact of air pollution on birth weight | Term low birth weight | PM2.5 | As a confounder, maternal R/E was included in models with maternal age, maternal education, gestational age, year of birth, gestational apparent temperature exposure, and percentage of households below poverty line at the ZCTA level Also as an effect measure modifier (multiplicative scale) |
Gray et al. 2014 [54] | Impact of air pollution and SES variables on birth outcomes | Low Birth Weight, Preterm Birth | O3 and PM2.5 | As a confounder; R/E was included in models with maternal education, maternalage at delivery, and census tract-level median household income |
Chen et al. 2015 [55] | Impact of air pollution on brain volumes in older women | Cognitive decline (measures of gray matter and normal appearing white matter) | PM2.5 | As a confounder; analysis used a staged modelling approach where minimally adjusted models were adjusted for R/E and other covariates (not SES) and more fully adjusted models included both R/E and SES (education, family income, and employment status) and other covariates. Additional analyses restricted to non-Hispanic White women. |
R/E as a EMM | ||||
Leiser et al., 2019 [56] | Effects of air pollution on spontaneous pregnancy loss | Spontaneous pregnancy loss | PM2.5, NO2, O3 | As an effect measure modifier (multiplicative scale, in a case crossover design) |
Laurent et al., 2016 [57] | Impact of air pollution on birth outcomes | Low birth weight | PM, NO2, O3 | As an effect measure modifier (multiplicative scale) |
Delfino et al. 2014 [58] | Impact of air pollution on asthma and R/E as a vulnerability factor | Asthma-related hospital morbidity | Traffic-related air pollution | As an effect measure modifier (multiplicative scale, in a case crossover design) |
Strickland et al. 2014 [59] | Impact of air pollution on children’s asthma and R/E as a vulnerability factor | Emergency department for asthma or wheeze among children 2 to 16 years of age | CO, NO2, PM2.5, O3 | As an effect measure modifier (multiplicative scale). Heterogeneity tests were conducted. |
R/E as a main exposure | ||||
Grineski & Collins, 2018 [60] | Disparities in exposure to neurotoxicants in US public schools | Air pollution | neurotoxicants from US Environmental Protection Agency’s National Air Toxics Assessment (NATA). | As the main exposure of interest, adjusting for school district effects. |
Tonne et al., 2018 [61] | Inequalities in air pollution exposure by socio-economic status and racial/ethnic groups | Air pollution | PM2.5, NO2 | As the main exposure of interest, adjusting for age, age squared, ethnicity, household income, area-level income deprivation, and a random effect for household. |
Kravitz-Wirtz et al. 2016 [62] | Inequalities in air pollution exposure by racial/ethnic groups | Air pollution | NO2, PM2.5, and PM10 | As the main exposure of interest, adjusting for age, family size, income, employment, housing tenure, metropolitan level segregation and industrial share |
Jones et al. 2014 [63] | Inequalities in air pollution exposure by racial/ethnic group and racial residential segregation | Air pollution | PM2.5 and NOx | As the main exposure of interest, adjusting for education, annual family income, and neighborhood median family income |
Jones et al. 2015 [64] | Decomposition of the total effect between R/E and intima-media thickness using air pollution exposure as a mediator. | Intima-media thickness | PM2.5 and NOx | As the main exposure of interest, adjusting for age, education, annual family income, smoking status, pack-years of smoking, BMI, diabetes status, systolic blood pressure, total and HDL cholesterol, antihypertensive medication use and statin use. |