Next Article in Journal
Characteristics of Meteorological Droughts Across Different Climatic Zones in Benin
Previous Article in Journal
Relationships Among Atmospheric Suspended Particulates with Different Sizes: A Case Study of Chongqing City
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 16
Issue 5
10.3390/atmos16050610
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Disparities in Fine Particulate Matter Air Pollution Exposures at the US–Mexico Border: The Intersection of Race/Ethnicity and Older Age

by
Timothy W. Collins
1,*,
Colby M. Child
2,
Sara E. Grineski
3 and
Mathilda Scott
3
1
School of Environment, Society, and Sustainability, Center for Natural & Technological Hazards, University of Utah, Salt Lake City, UT 84112, USA
2
Center for Natural & Technological Hazards, University of Utah, Salt Lake City, UT 84112, USA
3
Department of Sociology, Center for Natural & Technological Hazards, University of Utah, Salt Lake City, UT 84112, USA
*
Author to whom correspondence should be addressed.
Atmosphere 2025, 16(5), 610; https://doi.org/10.3390/atmos16050610
Submission received: 17 April 2025 / Revised: 6 May 2025 / Accepted: 14 May 2025 / Published: 17 May 2025
(This article belongs to the Section Air Quality)
Browse Figures
Versions Notes

Abstract

Environmental justice research in the United States (US) documents greater air pollution exposures for Hispanic/Latino vs. non-Hispanic White groups. EJ research has not focused on the intersection of race/ethnicity and older age nor short-term fine particulate matter (PM2.5) exposures. We address these knowledge gaps by studying US metropolitan area census tracts within 100 km of the US–Mexico border, a region with serious air quality issues. We use US Census American Community Survey data to construct sociodemographic variables and Environmental Protection Agency Downscaler data to construct long-term and short-term measures of PM2.5 exposure. Using multivariable generalized estimating equations, we test for differences in PM2.5 exposures between census tracts with higher vs. lower proportions of older Hispanic/Latino residents and older non-Hispanic White residents. The results indicate that as the proportion of the Hispanic/Latino population ≥ 65 years of age increases, long-term and short-term PM2.5 exposures significantly increase. In contrast, as the proportion of the non-Hispanic White population ≥ 65 years of age increases, changes in long-term and short-term PM2.5 exposures are statistically non-significant. These findings illuminate how race/ethnicity and older age intersect in shaping PM2.5 exposure disparities and may inform efforts to mitigate air pollution exposures for overburdened people along the US–Mexico border.

1. Introduction

Air pollution poses significant health risks to populations worldwide, with fine particulate matter 2.5 microns or less in diameter (PM2.5) being a key contributor to these risks. Globally, an estimated 8 million premature deaths annually (12% of total premature deaths) are attributable to air pollution, with PM2.5 accounting for 97% of those premature deaths [1]. In addition to causing premature mortality, air pollution exposure is linked to a broad range of health problems [2,3]. Not only is air pollution exceptionally harmful, but the literature on environmental justice (EJ) has shown that it is distributed inequitably across social groups, with marked exposure inequities in the United States (US) occurring based on racial/ethnic minority status and low socioeconomic status [4,5,6,7,8,9,10,11].
Recent research has shown that racial/ethnic minorities and older adults in the United States and elsewhere exhibit heightened vulnerability to the health risks of air pollution and PM2.5 exposures in particular [12,13,14,15,16,17,18,19,20,21]. Despite these unequal risks, research on environmental justice (EJ) has largely focused on racial/ethnic disparities in air pollution exposure and overlooked the intersection of race/ethnicity and age in shaping differential exposures to PM2.5. An emphasis on this intersection aligns with calls to conduct intersectional EJ studies that seek to clarify how different dimensions of social inequalities combine to shape risk [22]. Quantitative intersectional EJ research has focused primarily on documenting how air pollution exposures increase based on the intersection of disadvantaged racial/ethnic status with foreign nativity (vs. US birth), specific ancestries (e.g., Dominican Hispanic/Latino, Ethiopian Black, and Middle Eastern White), English-language deficiency (vs. proficiency), and lower (vs. higher) socioeconomic status (e.g., [23,24]), neglecting the intersection of race/ethnicity with older age.
Only two EJ studies on air pollution have explored the intersection of race/ethnicity and older age. The first study was conducted in the US–Mexico border location of El Paso County, TX [25]. It was found that the risk of a neighborhood being highly exposed to carcinogenic air pollution (at or above the 90th percentile in exposure locally) was nearly four times higher if the neighborhood was in the highest (vs. the lowest) quartile for the percentage of Hispanic/Latino residents who were ≥65 years of age. Conversely, neighborhoods in the highest (vs. lowest) quartile for the percentage of non-Hispanic White residents who were ≥65 years of age were less likely to be highly exposed to carcinogenic air pollution [25]. A study in Harris County similarly looked at “within racial/ethnic group” differences in cancer risk from air pollution based on age and found greater risks for tracts with higher concentrations of older adults (≥65) among both the Hispanic/Latino and non-Hispanic White populations [26].
Like the study by Collins et al. [25], other quantitative EJ studies conducted on the US–Mexico border have tended to focus on one metropolitan area [27,28,29]. To our knowledge, just one study has examined the entirety of the border and found that increased concentrations of renter-occupants, Hispanic/Latino residents, households with middle to high socioeconomic status, and foreign-born US citizens were associated with elevated risks from long-term carcinogenic air pollution, and that an older age was not associated with risk [30]. Despite the most prominent risks it poses, no prior studies have investigated PM2.5 pollution disparities along the entirety of the US–Mexico border, a critical area of research given the unique challenges and vulnerabilities faced by communities there. The US–Mexico border faces air pollution problems because of the heavily trafficked cross-border points of entry and the failure or reluctance of governments and industries to meet air quality standards [7,30,31,32,33,34].
Nearly all EJ studies of air pollution have examined disparities in long-term exposures, overlooking potential disparities in exposures to short-term air pollution (see [5] for a review). This is problematic since short-term increases in PM2.5 are also associated with serious negative health impacts [35,36,37,38]. Two nationwide US studies found consistent racial/ethnic disparities across measures of long-term and short-term PM2.5 exposures, with disparities being larger for measures of short-term (and more acute) vs. long-term (and less acute) PM2.5 exposures [5,6]. The racial/ethnic groups with the greatest long- and short-term PM2.5 exposures nationally across the United States included Hispanic/Latino, non-Hispanic Asian, and non-Hispanic Black people, as well as people of color (i.e., those of Hispanic/Latino and/or any non-White race) overall. The non-Hispanic White group exhibited the lowest long- and short-term PM2.5 exposures [5,6].
This study addresses the limitations of prior work and advances knowledge on this topic by examining separate and intersectional associations of neighborhood-level racial/ethnic and older age statuses with long-term and short-term exposures to PM2.5 air pollution in US metropolitan areas along the US–Mexico border, areas with well-documented PM2.5 pollution problems [7,31,32,33,34]. We ask two research questions: What are the associations of race/ethnicity and age with long-term and short-term exposures to PM2.5? How do race/ethnicity and older age intersect to influence long-term and short-term exposures to PM2.5?

2. Materials and Methods

2.1. Study Area

The study area encompasses metropolitan US Census tracts within 100 km of the US–Mexico border (Figure 1a). Census tracts were included in our analysis if they were located in a metropolitan area core (i.e., rural–urban commuting area (RUCA) primary code 1 [39]) and had a total population of 500 or greater and at least 100 residents ≥ 65 years of age. We focused on tracts in US metropolitan areas along the US–Mexico border because those areas have well-documented PM2.5 pollution problems [7,31,32,33,34]. We excluded census tracts with less than 500 residents because tracts with such small populations are outliers. Given our analytic focus on an older age group—as well as intersectional groups defined by race/ethnicity and older age—we excluded census tracts with fewer than 100 residents ≥ 65 years of age because estimates for small count groups have high error in the ACS at the tract level [40]. We also excluded tracts with missing data for analysis variables. With those criteria, 294 census tracts were excluded, leaving 1208 in our analyses. We used census tracts as our analysis unit because they represent the finest resolution for which reliable sociodemographic data (independent variables) are available from the US Census and because they are the unit for which daily PM2.5 estimates are available from the US Environmental Protection Agency (EPA) Downscaler (dependent variables).

2.2. Dependent Variables

We developed three measures of PM2.5 at the census tract level that represent three dependent variables in our analysis, with one measuring long-term exposure and two gauging the frequency of short-term exposures. Specifically, we used the (i) mean daily average concentration of PM2.5 in μg/m3 in the 2012–2016 period as a measure of long-term exposure. Our short-term exposure measures include counts of days from 2012 to 2016 for which the tract-level 24 h mean estimated PM2.5 concentration equaled or exceeded (ii) 25 μg/m3 and (iii) 35 μg/m3. Those two short-term exposure measures are based on policy-relevant thresholds established by the World Health Organization (WHO) guidelines [41] and USEPA National Ambient Air Quality Standards (NAAQS) [42] applicable to the period of study (2012–2016). The WHO’s 24 h mean PM2.5 guideline from 2006 to 2021 was 25 μg/m3; the USEPA’s 24 h mean PM2.5 NAAQS from 2012 to the present is 35 μg/m3.
These three measures were derived from the USEPA Downscaler, which provides estimates of daily mean (average) PM2.5 concentrations in μg/m3 for each census tract in the United States [43]. The Downscaler PM2.5 concentration estimates were obtained from a Bayesian space-time fusion mode developed by Berrocal et al. [44]. The Downscaler model combines data from the gridded atmospheric model, also known as the Community Multi-Scale Air Quality Model or CMAQ, and the air pollution point measurements from the National Air Monitoring Stations (NAMS/SLAMS). By explicitly accounting for the spatiotemporal dependence in PM2.5 concentrations, the Downscaler model performs well compared to direct PM2.5 observations [45,46,47]. Downscaler data have been widely employed in air pollution epidemiology and EJ studies (e.g., [5,6,48,49,50]). Archived Downscaler data with descriptive files are available from the USEPA [43]. We obtained Downscaler daily PM2.5 concentration estimates at the locations of census tract centroids.
We used Downscaler data for the years 2012–2016 to match the time frame of the American Community Survey’s 5-year estimates used for our independent variables. PM2.5 concentrations exhibit substantial interannual variability, justifying our use of a 5-year period [51]. For all three measures, we downloaded daily mean PM2.5 concentration estimates for all US census tracts from 2012 to 2016 and bound the data to the tracts of our study area described above. Next, we used those data to calculate the long-term and short-term PM2.5 exposure measures described above for the 2012–2016 study period. Figure 1 presents maps for each of the PM2.5 measures.

2.3. Independent Variables

We obtained Hispanic/Latino ethnicity by race data from the American Community Survey (ACS) for the years 2012–2016, which we used to calculate mutually exclusive race/ethnicity variables for the proportions of the census tract population that were Hispanic/Latino as well as non-Hispanic White, Black, American Indian, Asian, Pacific Islander, or other/multi-race. While each of the non-Hispanic groups excludes the Hispanic/Latino population, we heretofore do not refer to those groups as “non-Hispanic”.
To examine census tract-level exposure disparities based on older Hispanic/Latino and older White compositions, we used data on age by race/ethnicity to create two cross-classified variables that gauge the proportions of each tract’s Hispanic/Latino population and White population that were 65 years of age or older. We focus on the intersection of Hispanic/Latino vs. White race/ethnicity with older age in our analysis for three reasons. First, the Hispanic/Latino vs. White comparison captures the most salient dimension of racial/ethnic inequality along the US–Mexico border [25,30]. Second, the Hispanic/Latino and White populations are the only racial/ethnic groups that are sufficiently large to analyze across the US–Mexico border study area. Third, these are the only two populations for which age composition is distinguished by the ACS based on Hispanic/Latino ethnicity and race (i.e., ACS data for age do not distinguish (non-)Hispanic/Latino ethnicity from non-White races).
Median household income and the proportion of renter-occupied housing units were included to represent socioeconomic status (SES). We also included median household income squared because of the potential for a non-linear relationship with air pollution exposure [52]. Proportion renter-occupancy was included because it is representative of accumulated wealth and stability. A population density variable (total tract population divided by the area of the tract (sq. miles)) was also included in the models. Tabular ACS data and a shapefile of census tract boundaries, which were used to create the independent variables, were downloaded from the National Historical Geographic Information System website (https://www.nhgis.org/, accessed on 11 January 2025). Finally, to further control for the effects of urban context, our models adjust for tract clustering based on the county of location and eight categories of age of housing stock.

2.4. Analysis Approach

We undertook our analysis in four steps. First, we mapped the dependent variable and key independent variables and calculated descriptive statistics for all analysis variables. Second, following the methods described by Liu et al. [53] and Collins and Grineski [6], we calculated population-weighted mean PM2.5 long-term (2012–2016 average) and short-term (counts of days with average concentrations ≥ 25 and ≥35 μg/m3) exposures for the total population and our focal populations separately within each of the four US border states (CA, AZ, NM, and TX). Our focal populations include people who are Hispanic/Latino (all ages), Hispanic/Latino ≥ 65 years of age, and White ≥ 65 years of age. For comparative purposes, we also calculated the population-weighted mean PM2.5 exposures for White (all ages) people. Population-weighted means enable us to describe unadjusted exposures for the total and focal populations. We calculated these by (a) multiplying the group-specific population (for the total population and for each focal group separately) in each census tract by the corresponding tract PM2.5 value (for the long-term and short-term measures separately); (b) summing the group-specific * PM2.5 measure-specific values separately for all census tracts; (c) summing the group-specific population values separately for all census tracts; and (d) dividing the values calculated in step (b) by the values calculated in step (c) separately for each group and PM2.5 measure. We performed these steps separately for metropolitan census tracts in each border state (CA, AZ, NM, and TX).
Third, we performed a multivariable analysis using generalized estimating equations (GEEs). GEEs are useful for EJ research because they accommodate non-normally distributed variables and adjust for clustering [52]. We fit nine separate GEE models with the same model specifications. We tried all possible combinations of probability distributions and link functions to find the combination yielding the best model fit. The gamma distribution and logarithmic link function fit best for all models. The nine models consist of three subsets of three models. Each subset includes a separate model for each of the three PM2.5 measures (dependent variables). The first subset of three models includes race/ethnicity, SES, and population density variables as predictors of each dependent variable. The second subset of three models substitutes the “proportion Hispanic/Latino ≥ 65” variable for the “proportion Hispanic” variable and includes a new variable, “proportion White.” The “proportion Hispanic/Latino ≥ 65” variable is interpretable in reference to the proportion of the Hispanic/Latino population < 65 years of age, adjusting for the other variables in the model and clustering. The third subset of three models substitutes the “proportion White ≥ 65” variable for the “proportion Hispanic/Latino ≥ 65” variable and the “proportion Hispanic” variable for the “proportion White” variable. The “proportion White ≥ 65” variable in those three models is interpretable in reference to the proportion of the White population that is <65 years of age.
Fourth, to clarify the effects of key independent variables on long-term and short-term PM2.5 exposures in the GEEs, we calculated estimated marginal (EM) means. EM means provide value estimates of the dependent variables based on specified values for independent variables of interest, adjusting for covariates, clustering, and all other GEE specifications. Specifically, we calculated EM means for each PM2.5 measure at every 10th percentile (from 0th to 100th) for each independent variable that exhibited a statistically significant association. This included independent variables representing the proportions of the census tract population that were Hispanic/Latino, Black, Asian, ≥65, and Hispanic/Latino ≥ 65. EM mean values for those independent variables of interest were predicted at every 10th percentile while fixing each other covariate (i.e., independent variable) in the GEEs at its mean. We used IBM SPSS Statistics v25 to conduct the analyses.

3. Results

3.1. Univariate Results

The three dependent variables mapped in Figure 1 show similar general patterns, although the long-term exposure variable (i.e., 2012–2016 mean PM2.5) displays more gradual spatial variation, while the two short-term pollution exposure variables (i.e., the number of days with a PM2.5 concentration ≥ 25 μg/m3 or ≥35 μg/m3 from 2012 to 2016) have more concentrated high values. The highest values are in the census tracts of the largest metropolitan areas along the border (San Diego, Imperial Valley, El Paso, and Brownsville), except for the count of days with a PM2.5 concentration ≥ 35 μg/m3, for which tracts in Brownsville have zero values. Within large metropolitan areas, the values are the highest in urban centers and tend to decrease moving outward. While San Diego census tracts have the highest long-term PM2.5 concentrations, El Paso tracts exhibit the greatest number of days with PM2.5 concentrations exceeding the two short-term thresholds (i.e., 24 h mean ≥ 25 μg/m3 or ≥35 μg/m3). Although California’s Imperial Valley is not a very large population center, metropolitan tracts there exhibit high PM2.5 concentrations across all three variables, which is partly attributable to the nearby Mexican metropolis of Mexicali.
Maps of focal independent variables are displayed in Figure 2. Census tracts with the highest proportion of Hispanic/Latino residents were concentrated most heavily in Texas. Tracts in New Mexico, Arizona, and California had a lower proportion of Hispanic/Latino individuals except for small clusters in San Diego, the Imperial Valley, and Tucson. This pattern changes for the proportion of Hispanic/Latino individuals ≥ 65, with higher proportions being observed in tracts in all border states. For the proportion of White individuals ≥ 65, the pattern appears more random, with higher proportion tracts spread across the border region.
Descriptive statistics for each analysis variable are shown in Table 1. Our long-term air pollution dependent variable, the mean PM2.5 (concentration in μg/m3), ranged from 5.735 to 9.916 with a mean of 8.675. Our first short-term exposure variable, the number of days with a 24 h mean PM2.5 concentration ≥ 25 µg/m3, ranged from 0 to 20 with a mean of 7.010. Our second acute pollution variable, the number of days with a PM2.5 concentration ≥ 35 µg/m3, had values ranging from 0 to 6 and a mean of 0.950. For our focal independent variables, the proportion of Hispanic/Latino individuals ranged from 0.014 to 1 with a mean of 0.546, the proportion of Hispanic/Latino individuals ≥ 65 had a range of 0–0.694 and a mean of 0.092, and the proportion of White individuals ≥ 65 ranged from 0 to 1 with a mean of 0.234.

3.2. Population-Weighted Mean PM2.5 Exposure Results

Table 2 reports the population-weighted mean exposure values for each of the three dependent variables for the total population and focal groups by US border state. Differences in population exposures between states for all three dependent variables were notable. California had the highest average total population exposure to long-term PM2.5 (>9 µg/m3), followed by Texas (>8 µg/m3), New Mexico (>7 µg/m3), and Arizona (>6 µg/m3). The same pattern held between states for the two short-term PM2.5 exposure variables. Differences between subgroups of focus in exposure to each measure of PM2.5 within each state were relatively small but consistent. Specifically, Hispanic/Latino populations had higher exposure, while White populations had lower exposure to PM2.5 than the total population. Similarly, we found that the Hispanic/Latino population ≥ 65 had higher exposure, while the White population ≥ 65 had lower exposure to PM2.5 than the total population.

3.3. Multivariable GEE Model Results

Table 3 includes results from the nine GEE models. For the base model variables, we found that increases in the proportions of the population that were Hispanic/Latino, Black, Asian, and ≥65, as well as housing units that were occupied by renters, were associated with significant increases in PM2.5 values across the three dependent variables. This was with the exceptions of the proportion ≥ 65 for mean PM2.5 and the proportion of Black individuals for the number of days with a PM2.5 concentration ≥ 35 µg/m3, both of which exhibited positive but statistically non-significant associations.
Exponentiated beta values (see Exp(B) column, Table 3) provide percentage increases or decreases in the dependent variable range based on the independent variable to assist with the relative and comparative interpretation of the GEE results. As the proportion of Hispanic/Latino residents increased by one standard deviation, the mean PM2.5 increased by 1.5% (B = 0.015; 95% CI = 0.008 to 0.022), the number of days with a PM2.5 concentration ≥ 25 µg/m3 increased by 23.4% (0.210; 0.123 to 0.297), and the number of days with a PM2.5 concentration ≥ 35 µg/m3 increased by 38.9% (0.329; 0.200 to 0.457). As the proportion of Black individuals increased, the mean PM2.5 increased by 0.6% (0.006; 0.003 to 0.008), and the number of days with a PM2.5 concentration ≥ 25 µg/m3 increased by 5.5% (0.053; 0.020 to 0.087). As the proportion of Asian individuals increased, the mean PM2.5 increased by 0.7% (0.007; 0.006 to 0.009), the number of days with a PM2.5 concentration ≥ 25 µg/m3 increased by 7.3% (0.070; 0.052 to 0.088), and the number of days with a PM2.5 concentration ≥ 35 µg/m3 increased by 12.7% (0.120; 0.080 to 0.159). As the proportion of the total population ≥ 65 years of age increased, the number of days with a PM2.5 concentration ≥ 25 µg/m3 increased by 0.2% (0.057; 0.019 to 0.094), and the number of days with a PM2.5 concentration ≥ 35 µg/m3 increased by 5.8% (0.076; 0.031 to 0.120). As the proportion of renter-occupied housing units increased, the mean PM2.5 increased by 0.8% (0.008; 0.004 to 0.012), the number of days with a PM2.5 concentration ≥ 25 µg/m3 increased by 21.8% (0.197; 0.141 to 0.253), and the number of days with a PM2.5 concentration ≥ 35 µg/m3 increased by 24.7% (0.221; 0.162 to 0.280). An isolated finding is that as the proportion Pacific Islander increased, the number of days with a PM2.5 concentration ≥ 25 µg/m3 increased by 1.3% (0.013; 0.002 to 0.024).
Additionally, we found a significant and positive association between median household income and each of the PM2.5 measures, such that a higher income was associated with increased exposure. The effect was non-linear in the case of the number days with a PM2.5 concentration ≥ 35 µg/m3, such that there was a slight decline in that short-term exposure measure at the highest levels of median household income. When adjusting for other variables in the GEEs, the population density control had relatively weak positive associations with the PM2.5 measures.
Continuing with Table 3 in three models using the older Hispanic/Latino variables, we found that a standard deviation increase in the proportion of Hispanic/Latino residents ≥ 65 years of age was associated with a 0.1% increase in the mean PM2.5 (0.001; 0.000 to 0.002), a 5.9% increase in the number of days with a PM2.5 concentration ≥ 25 µg/m3 (0.058; 0.032 to 0.084), and a 5.1% increase in the number of days with a PM2.5 concentration ≥ 35 µg/m3 (0.050; 0.016 to 0.083). The results for the proportion of Hispanic/Latino individuals ≥ 65 are in reference to the proportion of the Hispanic/Latino population < 65 years of age. We also found an inverse relationship between the proportion of White residents and PM2.5 values. That is, as the proportion of White individuals increased, the mean PM2.5 exposure decreased by 1.1% (−0.011; −0.017 to −0.006), the number of days with a PM2.5 concentration ≥ 25 µg/m3 decreased by 86.1% (−0.150; −0.22 to −0.080), and the number of days with a PM2.5 concentration ≥ 35 µg/m3 decreased by 78.6% (−0.241; −0.341 to −0.142). We found that the associations between the proportion of White residents ≥ 65 years age and PM2.5 exposures were not statistically significant. The results for the proportion of White individuals ≥ 65 variable are in reference to the proportion of White population < 65.
The EM means help describe the associations for the race/ethnicity, older age, and race/ethnicity by older age variables, exhibiting statistically significant associations with the PM2.5 measures. Figure 3 graphs the EM means for those variables (derived from the GEE results in Table 3), holding the other independent variables at their means. The results show significant spikes in exposure for all three PM2.5 exposure dependent variables past the 90th percentile for the proportions of individuals who are Black, Asian, ≥65 years of age, and Hispanic/Latino ≥ 65 years of age variables. As the proportion of Hispanic/Latino individuals increased, there was a steady increase in PM2.5 exposure for all three dependent variables. As the proportion of White individuals increased, there was a steady decrease in PM2.5 exposure for all three dependent variables.

4. Discussion

In response to our first research question—What are the associations of race/ethnicity and age with long-term and short-term exposures to PM2.5?—we found that greater proportions of Hispanic/Latino and Asian residents in US census tracts along the Mexican border were associated with worse PM2.5 air pollution with respect to long-term exposure (5-year mean) and short-term exposure (measured based on exceedances of the WHO and NAAQS 24 h mean standards applicable during the study period). A greater proportion of Black residents was associated with worse long-term exposure as well as short-term exposure using the WHO measure but not the NAAQS measure. We found that a higher proportion of Pacific Islander residents was associated with greater short-term exposure based on the WHO measure only. Conversely, we found that a greater proportion of White residents were associated with less exposure to PM2.5 with respect to both long-term and short-term exposure. While these results generally correspond with the findings from the EJ literature, the only prior study of the entirety of the US side of the US–Mexico border found an increased risk to carcinogenic air pollution based on a greater Hispanic/Latino composition but not based on Black, Asian, or Pacific Islander composition [30]. Prior studies of racial/ethnic disparities in PM2.5 exposures in the US have consistently documented disproportionate exposures for Hispanic/Latino, Black, and Asian Americans [5,6,8,9], as we have here, but few have examined the Pacific Islander population. The pattern of disparate exposure to short-term PM2.5 exposure that we found based on Pacific Islander neighborhood composition across the metropolitan US–Mexico border is generally influenced by this population’s relatively high residential composition within the largest border cities (San Diego, CA, and El Paso, TX), which have worse air pollution problems. Future EJ research on the Pacific Islander population is needed to advance knowledge of their air pollution exposure disparities along the US–Mexico border and elsewhere in the US.
In terms of age, we found that an increasing proportion of older adults was associated with greater short-term PM2.5 exposure (based on both the WHO’s and NAAQS’s measures), but not long-term exposure. Relationships between older age and air pollution have not been extensively examined in the literature on EJ. The only prior study of air pollution exposure disparities across the US–Mexico border found no disparities based on older age [30]. Our findings of greater PM2.5 exposures for older adults along the US–Mexico border, however, align with the results from studies of long-term carcinogenic air pollution in El Paso [25] and long-term PM10 air pollution in Phoenix [54]. Given the vulnerability of older adults to the health effects of air pollution exposure (which we discuss below) and the relative lack of prior studies examining disparate exposures based on older age, age-based disparities should be more thoroughly investigated in future EJ research.
With respect to our second research question—How do race/ethnicity and older age intersect to influence long-term and short-term exposures to PM2.5?—we found that older age intersects with Hispanic/Latino status, but not White status, to increase long- and short-term exposures to PM2.5. These findings suggest that Hispanic/Latino status amplifies the risks of PM2.5 exposure in older age populations, while Whiteness protects against such risks. The EM means graphed in Figure 3 clarify how increases until around the 90th percentile in the tract proportion of the Hispanic/Latino population ≥ 65 years of age are associated with small and gradual increases in PM2.5 exposures, with PM2.5 exposures then spiking among tracts above the 90th percentile for the proportion of the Hispanic/Latino population ≥ 65. This helps contextualize the GEE result by revealing that census tracts with very high proportions of older Hispanic/Latino individuals tend to experience dramatically elevated PM2.5 exposures. This pattern holds across the long- and short-term measures of PM2.5 exposure. This is similar to the bivariate pattern found in El Paso, Texas [25]. It is somewhat different from the pattern observed in Harris County, Texas, where increasing older (vs. younger) neighborhood composition—within both the Hispanic/Latino and White populations—was associated with significantly greater carcinogenic air pollution [26]. We speculate that the divergence between our results and those of Loustaunau and Chakraborty [26] is partly attributable to differences between the geographic contexts and the air pollutants examined.
Based on our findings and the interpretation of similar patterns by Collins et al. [25], we can speculate how Hispanic/Latino vs. White status intersects with older age to shape inequality in long- and short-term exposures to PM2.5. In metropolitan contexts along the US side of the US–Mexico border, racial/ethnic status has intersected in complex ways with older age to shape divergent relationships with PM2.5 exposures for older Hispanic/Latino vs. older White people. First, Hispanic/Latino ethnicity is associated with significantly increased PM2.5 exposure, while Anglo Whiteness is associated with significantly decreased PM2.5 exposure. Along the metropolitan US–Mexico border, higher proportions of Hispanic/Latino people tend to concentrate in central city census tracts and tracts most proximate to the Mexican border (Figure 2). In contrast, the highest proportions of White individuals are located in suburban (and relatively affluent) census tracts and tracts further from the border. In particular, the higher proportions of Hispanic/Latino people in barrios adjacent to city centers and nearer the border (and the lower proportions of non-Hispanic White people residing in those contexts) underlie the PM2.5 exposure disparities we found. The pattern of (poorer) racial/ethnic minority groups concentrating in city centers and (more affluent) White Anglo individuals residing in suburbs is common in the US. Pulido [55] argues that this pattern is a product of the ideology of White privilege that has structured urbanization in the US, particularly following World War II, and underpins contemporary race-based environmental injustices.
Moreover, our results indicate that Hispanic/Latino ethnicity has a multiplicative influence on PM2.5 exposure, while Anglo Whiteness protects against PM2.5 exposure disparities based on older age. In other words, the results indicate that disadvantages associated with older age amplify PM 2.5 exposures for Hispanic/Latino people along the metropolitan US–Mexico border but not for White people. Specifically, a process of aging in place within central city barrios along the border contributes to disparate PM2.5 exposures for census tracts with high proportions of older Hispanic/Latino people. In contrast, non-Hispanic White residents along the metropolitan US–Mexico border are more likely to concentrate in suburbs, where PM2.5 exposures are reduced. Thus, the socio-spatiality of aging along the metropolitan US–Mexico border may reinforce disparate PM2.5 exposures for older Hispanic/Latino people and alleviate exposures for older White people. We presume that, if border communities more aggressively market themselves as national retirement destinations (as Tucson, AZ; Las Cruces, NM; San Diego, CA; and McAllen, TX, have successfully done), the PM2.5 exposure disparities we observed between (aging) Hispanic/Latino and (retiring) White people will widen since new retirement communities will almost exclusively locate in suburban rather than central city areas. More research is needed to determine whether the patterns found here—in which Hispanic/Latino ethnic status intersects with older age as a risk factor while Anglo Whiteness intersects as a protective factor—derive more from general processes that shape environmental injustices in the US or more from regional particularities.
In terms of the public health significance of these results, >60% of premature deaths from all environmental/occupational risks worldwide are attributable to PM2.5 air pollution exposure according to the Global Burden of Disease Study [1]. And recent research has firmly established that small increases in PM2.5 at relatively low exposure levels—which characterize the exposure disparities and levels we observed here—have larger effects on negative health outcomes than small increases in PM2.5 at relatively high levels of exposure [12,56,57,58,59,60,61]. Additionally, the disparate PM2.5 exposures we found based on Hispanic/Latino (vs. White), older (vs. younger), and older Hispanic/Latino statuses likely operate to compound negative health outcomes due to the heightened vulnerability to PM2.5 exposures among racial/ethnic minorities and older people, which is well documented [12,13,14,15,16,17,18,19,20,21].
It is important to note that this study has limitations. First, we used aggregated (not individual-level) population data and measures of PM2.5 at the census tract level. Future research should analyze data for the population at an individual level, including measures of hyperlocal air pollution exposure. Second, our study is cross-sectional and cannot address causality. Future analyses employing longitudinal data are needed to evaluate causes of the exposure disparities we found. Third, our analysis cannot clarify whether the disparate exposures we documented contribute to health inequities; future research that integrates health outcome data and examines those linkages is needed. Fourth, because our analysis focuses exclusively on the US–Mexico border region, we cannot ascertain whether the novel patterns we found (e.g., significantly amplified long-term and short-term PM2.5 exposures for older Hispanic/Latino people but not older White people) exist elsewhere. Thus, future work should assess whether these patterns are present in other regions and at a national level. Lastly, we could not include indoor, workplace, and commuting exposures in our analyses as data from the ACS only identify location of residence, and available air pollution data with complete population coverage only measure outdoor exposures. Future research addressing these limitations is needed to advance fundamental knowledge of social disparities in air pollution exposures.

5. Conclusions

In sum, our findings regarding disparities based on race/ethnicity, older age, and the intersection of race/ethnicity and age in long- and short-term exposures to PM2.5 advance our understanding of EJ issues at the US–Mexico border. Practical interventions to reduce these exposure disparities should focus on neighborhoods with relatively high PM2.5 exposures and high compositions of social groups experiencing disparate exposures, including neighborhoods home to high concentrations of Hispanic/Latino, Asian, and Black people, as well as older Hispanic/Latino people in particular.
With that in mind, our analysis is useful for screening geographic areas of special concern where high levels of PM2.5 intersect with high concentrations of such residents. For example, Figure 1 and Figure 2 reveal high exposures to long- and short-term PM2.5 as well as a high composition of older Hispanic/Latino people in central El Paso (TX) census tracts. In those central El Paso neighborhoods, a substantial component of PM2.5 emissions is attributable to heavily trafficked cross-border transportation routes spanning several major international ports of entry [32,62]. In El Paso, collaborative efforts at the international, federal, and local levels could most effectively reduce transportation emissions at border crossings in order to reduce the exposure disparities that we documented here. Additionally, since the economic health cost per capita of PM2.5 exposure for the older population is ~10 times that for the younger population [21], such strategies could also focus on targeting enhanced health-care services for older Hispanic/Latino people in central El Paso neighborhoods, many of whom experience barriers to health-care access. Similar targeted interventions could seek to reduce exposure disparities in other Hispanic/Latino-majority US–Mexico border locations, such as the Imperial Valley (CA), where urban air pollution emissions (some from nearby areas, such as Mexicali, Mexico), thermal inversions, and dust storms exacerbated by the shrinking Salton Sea lead to amplified long- and short-term PM2.5 exposures. Future research and practice should more systematically seek to identify areas where targeted interventions can most effectively address air pollution exposure disparities along the US–Mexico border and elsewhere.

Author Contributions

Conceptualization, T.W.C. and C.M.C.; methodology, T.W.C.; formal analysis, T.W.C. and C.M.C.; resources, T.W.C. and S.E.G.; data curation, C.M.C. and T.W.C.; writing—original draft preparation, T.W.C., C.M.C., S.E.G., and M.S.; writing—review and editing, T.W.C., C.M.C., S.E.G., and M.S.; visualization, C.M.C. and T.W.C.; supervision, T.W.C. and S.E.G. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All data used in this analysis are publicly available. The publicly archived datasets analyzed are available from https://www.epa.gov/hesc/rsig-related-downloadable-data-files (accessed on 11 April 2025) (USEPA Downscaler data for PM2.5 exposure measures), https://www.nhgis.org/ (accessed on 11 January 2025) (US Census Bureau American Community Survey data from the National Historical Geographic Information System), and https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes (accessed on 11 April 2025) (USDA rural–urban community area code data on metropolitan area core census tracts). The analyzed data will be made available upon request.

Acknowledgments

C.M.C. acknowledges the University of Utah Undergraduate Research Opportunities Program, which supported his contribution to this research.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. IHME (Institute for Health Metrics and Evaluation). Global Burden of Disease Study 2021 Results. 2021. Available online: http://ghdx.healthdata.org/gbd-results-tool (accessed on 11 April 2025).
  2. Manisalidis, I.; Stavropoulou, E.; Stavropoulos, A.; Bezirtzoglou, E. Environmental and health impacts of air pollution: A review. Front. Public Health 2020, 8, 14. [Google Scholar] [CrossRef] [PubMed]
  3. Tavella, R.A.; Penteado, J.O.; de Lima Brum, R.; da Silva Bonifácio, A.; Martin, M.C.S.; Saes-Silva, E.; Brum, A.N.; Buffarini, R.; Filho, W.L.F.C.; Adamatti, D.F.; et al. An exploratory study on the association between air pollution and health problems (ICD-10) with an emphasis on respiratory diseases. Atmos. Pollut. Res. 2025, 16, 102377. [Google Scholar] [CrossRef]
  4. Brulle, R.J.; Pellow, D.N. Environmental justice: Human health and environmental inequalities. Annu. Rev. Public Health 2006, 27, 103–124. [Google Scholar] [CrossRef]
  5. Collins, T.W.; Grineski, S.E.; Shaker, Y.; Mullen, C.J. Communities of color are disproportionately exposed to long-term and short-term PM2.5 in metropolitan America. Environ. Res. 2022, 214, 114038. [Google Scholar] [CrossRef]
  6. Collins, T.W.; Grineski, S.E. Racial/ethnic disparities in short-term PM2.5 air pollution exposures in the United States. Environ. Health Perspect. 2022, 130, 087701. [Google Scholar] [CrossRef] [PubMed]
  7. Colmer, J.; Hardman, I.; Shimshack, J.; Voorheis, J. Disparities in PM2.5 air pollution in the United States. Science 2020, 369, 575–578. [Google Scholar] [CrossRef]
  8. Jbaily, A.; Zhou, X.; Liu, J.; Lee, T.H.; Kamareddine, L.; Verguet, S.; Dominici, F. Air pollution exposure disparities across US population and income groups. Nature 2022, 601, 228–233. [Google Scholar] [CrossRef] [PubMed]
  9. Tessum, C.W.; Paolella, D.A.; Chambliss, S.E.; Apte, J.S.; Hill, J.D.; Marshall, J.D. PM2.5 polluters disproportionately and systemically affect people of color in the United States. Sci. Adv. 2021, 7, eabf4491. [Google Scholar] [CrossRef]
  10. Mohai, P.; Pellow, D.; Roberts, J.T. Environmental justice. Annu. Rev. Environ. Resour. 2009, 34, 405–430. [Google Scholar] [CrossRef]
  11. Nadybal, S.; Grineski, S.; Collins, T.; Castor, A.; Flores, A.; Griego, A.; Mullen, C.; Rubio, R. Environmental justice in the US and beyond: Frameworks, evidence, and social action. In Geographies of Behavioral Health, Crime, and Disorder; Lersch, K., Chakraborty, J., Eds.; Springer: New York, NY, USA, 2020; pp. 187–209. [Google Scholar]
  12. Bowe, B.; Xie, Y.; Yan, Y.; Al-Aly, Z. Burden of cause-specific mortality associated with PM2.5 air pollution in the United States. JAMA Netw. Open 2019, 2, e1915834. [Google Scholar] [CrossRef]
  13. Di, Q.; Wang, Y.; Zanobetti, A.; Wang, Y.; Koutrakis, P.; Choirat, C.; Dominici, F.; Schwartz, J.D. Air pollution and mortality in the Medicare population. N. Engl. J. Med. 2017, 376, 2513–2522. [Google Scholar] [CrossRef] [PubMed]
  14. Geldsetzer, P.; Fridljand, D.; Kiang, M.V.; Bendavid, E.; Heft-Neal, S.; Burke, M.; Thieme, A.H.; Benmarhnia, T. Disparities in air pollution attributable mortality in the US population by race/ethnicity and sociodemographic factors. Nat. Med. 2024, 30, 2821–2829. [Google Scholar] [CrossRef] [PubMed]
  15. Josey, K.P.; Delaney, S.W.; Wu, X.; Nethery, R.C.; DeSouza, P.; Braun, D.; Dominici, F. Air pollution and mortality at the intersection of race and social class. N. Engl. J. Med. 2023, 388, 1396–1404. [Google Scholar] [CrossRef]
  16. Ma, Y.; Zang, E.; Opara, I.; Lu, Y.; Krumholz, H.M.; Chen, K. Racial/ethnic disparities in PM2.5-attributable cardiovascular mortality burden in the United States. Nat. Hum. Behav. 2023, 7, 2074–2083. [Google Scholar] [CrossRef] [PubMed]
  17. Power, M.C.; Weisskopf, M.G.; Alexeeff, S.E.; Coull, B.A.; Spiro, A., III; Schwartz, J. Traffic-related air pollution and cognitive function in a cohort of older men. Environ. Health Perspect. 2011, 119, 682–687. [Google Scholar] [CrossRef]
  18. Pun, V.C.; Kazemiparkouhi, F.; Manjourides, J.; Suh, H.H. Long-term PM2.5 exposure and respiratory, cancer, and cardiovascular mortality in older US adults. Am. J. Epidemiol. 2017, 186, 961–969. [Google Scholar] [CrossRef]
  19. Spiller, E.; Proville, J.; Roy, A.; Muller, N.Z. Mortality risk from PM2.5: A comparison of modeling approaches to identify disparities across racial/ethnic groups in policy outcomes. Environ. Health Perspect. 2021, 129, 127004. [Google Scholar] [CrossRef]
  20. Weuve, J.; Puett, R.C.; Schwartz, J.; Yanosky, J.D.; Laden, F.; Grodstein, F. Exposure to particulate air pollution and cognitive decline in older women. Arch. Intern. Med. 2012, 172, 219–227. [Google Scholar] [CrossRef]
  21. Yin, H.; Brauer, M.; Zhang, J.; Cai, W.; Navrud, S.; Burnett, R.; Howard, C.; Deng, Z.; Kammen, D.M.; Schellnhuber, H.J.; et al. Population ageing and deaths attributable to ambient PM2.5 pollution: A global analysis of economic cost. Lancet Planet. Health 2021, 5, e356–e367. [Google Scholar] [CrossRef]
  22. Malin, S.A.; Ryder, S.S. Developing deeply intersectional environmental justice scholarship. Environ. Sociol. 2018, 4, 1–7. [Google Scholar] [CrossRef]
  23. Liévanos, R.S. Race, deprivation, and immigrant isolation: The spatial demography of air-toxic clusters in the continental United States. Soc. Sci. Res. 2015, 54, 50–67. [Google Scholar] [CrossRef] [PubMed]
  24. Rubio, R.; Grineski, S.; Collins, T.; Morales, D.X. Ancestry-based intracategorical injustices in carcinogenic air pollution exposures in the United States. Soc. Nat. Resour. 2020, 33, 987–1005. [Google Scholar] [CrossRef] [PubMed]
  25. Collins, T.W.; Grineski, S.E.; Chakraborty, J.; McDonald, Y.J. Understanding environmental health inequalities through comparative intracategorical analysis: Racial/ethnic disparities in cancer risks from air toxics in El Paso County, Texas. Health Place 2011, 17, 335–344. [Google Scholar] [CrossRef] [PubMed]
  26. Loustaunau, M.G.; Chakraborty, J. Vehicular air pollution in Houston, Texas: An intra-categorical analysis of environmental injustice. Int. J. Environ. Res. Public Health 2019, 16, 2968. [Google Scholar] [CrossRef]
  27. Grineski, S.E.; Collins, T.W. Environmental injustices in transnational context: Urbanization and industrial hazards in El Paso/Ciudad Juárez. Environ. Plan. A 2010, 42, 1308–1327. [Google Scholar] [CrossRef]
  28. Lara-Valencia, F.; Declet-Barreto, J.; Keys, E. Spatial equity and transportation hazard along cross-border trade corridors: The case of Ambos Nogales. J. Borderl. Stud. 2008, 23, 1–18. [Google Scholar] [CrossRef]
  29. Lara-Valencia, F.; Harlow, S.D.; Lemos, M.C.; Denman, C.A. Equity dimensions of hazardous waste generation in rapidly industrialising cities along the United States–Mexico border. J. Environ. Plan. Manag. 2009, 52, 195–216. [Google Scholar] [CrossRef]
  30. Rubio, R.; Grineski, S.; Collins, T. Carcinogenic air pollution along the United States’ southern border: Neighborhood inequities in risk. Environ. Res. 2022, 212, 113251. [Google Scholar] [CrossRef]
  31. Kelly, K.E.; Jaramillo, I.C.; Quintero-Núñez, M.; Wagner, D.A.; Collins, K.; Meuzelaar, H.L.; Lighty, J.S. Low-wind/high particulate matter episodes in the Calexico/Mexicali region. J. Air Waste Manag. Assoc. 2010, 60, 1476–1486. [Google Scholar] [CrossRef]
  32. Lauer, F.T.; Mitchell, L.A.; Bedrick, E.; McDonald, J.D.; Lee, W.-Y.; Li, W.-W.; Olvera, H.; Amaya, M.A.; Berwick, M.; Gonzales, M.; et al. Temporal–spatial analysis of US–Mexico border environmental fine and coarse PM air sample extract activity in human bronchial epithelial cells. Toxicol. Appl. Pharmacol. 2009, 238, 1–10. [Google Scholar] [CrossRef]
  33. Ordieres, J.B.; Vergara, E.P.; Capuz, R.S.; Salazar, R.E. Neural network prediction model for fine particulate matter (PM2.5) on the US–Mexico border in El Paso (Texas) and Ciudad Juárez (Chihuahua). Environ. Model. Softw. 2005, 20, 547–559. [Google Scholar] [CrossRef]
  34. Quintana, P.J.; Dumbauld, J.J.; Garnica, L.; Chowdhury, M.Z.; Velascosoltero, J.; Mota-Raigoza, A.; Flores, D.; Panagon, N.; Gamble, J. Traffic-related air pollution in the community of San Ysidro, CA, in relation to northbound vehicle wait times at the US–Mexico border Port of Entry. Atmos. Environ. 2014, 88, 353–361. [Google Scholar] [CrossRef]
  35. Chen, C.; Zhu, P.; Lan, L.; Zhou, L.; Liu, R.; Sun, Q.; Ban, J.; Wang, W.; Xu, D.; Li, T. Short-term exposures to PM2.5 and cause-specific mortality of cardiovascular health in China. Environ. Res. 2018, 161, 188–194. [Google Scholar] [CrossRef] [PubMed]
  36. Fajersztajn, L.; Saldiva, P.; Pereira, L.A.A.; Leite, V.F.; Buehler, A.M. Short-term effects of fine particulate matter pollution on daily health events in Latin America: A systematic review and meta-analysis. Int. J. Public Health 2017, 2, 729–738. [Google Scholar] [CrossRef]
  37. Orellano, P.; Reynoso, J.; Quaranta, N.; Bardach, A.; Ciapponi, A. Short-term exposure to particulate matter (PM10 and PM2.5), nitrogen dioxide (NO2), and ozone (O3) and all-cause and cause-specific mortality: Systematic review and meta-analysis. Environ. Int. 2020, 142, 105876. [Google Scholar] [CrossRef]
  38. Yu, W.; Xu, R.; Ye, T.; Abramson, M.J.; Morawska, L.; Jalaludin, B.; Johnston, F.H.; Henderson, S.B.; Knibbs, L.D.; Morgan, G.G.; et al. Estimates of global mortality burden associated with short-term exposure to fine particulate matter (PM2.5). Lancet Planet. Health 2024, 8, e146–e155. [Google Scholar] [CrossRef]
  39. USDA (United States Department of Agriculture). 2010 Rural-Urban Commuting Area Codes. 2019. Available online: https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes/ (accessed on 11 April 2025).
  40. Spielman, S.E.; Folch, D.; Nagle, N. Patterns and causes of uncertainty in the American Community Survey. Appl. Geogr. 2014, 46, 147–157. [Google Scholar] [CrossRef]
  41. WHO (World Health Organization). Air Quality Guidelines for Particulate Matter, Ozone, Nitrogen Dioxide and Sulfur Dioxide: Global update 2005—Summary of Risk Assessment; World Health Organization: Copenhagen, Denmark, 2006. [Google Scholar]
  42. USEPA (United States Environmental Protection Agency). Revised Air Quality Standards for Particle Pollution and Updates to the Air Quality Index (AQI). Office of Air Quality Planning and Standards. 2012. Available online: https://www.epa.gov/sites/production/files/2016-04/documents/2012_aqi_factsheet.pdf (accessed on 11 April 2025).
  43. USEPA (United States Environmental Protection Agency). Fused Air Quality Surface Using Downscaling (FAQSD) Files, Downscaling Output Files. 2025. Available online: https://www.epa.gov/hesc/rsig-related-downloadable-data-files (accessed on 11 April 2025).
  44. Berrocal, V.J.; Gelfand, A.E.; Holland, D.M. A spatio-temporal downscaler for output from numerical models. J. Agric. Biol. Environ. Stat. 2010, 15, 176–197. [Google Scholar] [CrossRef]
  45. Berrocal, V.J.; Guan, Y.; Muyskens, A.; Wang, H.; Reich, B.J.; Mulholland, J.A.; Chang, H.H. A comparison of statistical and machine learning methods for creating national daily maps of ambient PM2.5 concentration. Atmos. Environ. 2020, 222, 117130. [Google Scholar] [CrossRef]
  46. Jin, X.; Fiore, A.M.; Civerolo, K.; Bi, J.; Liu, Y.; Van Donkelaar, A.; Martin, R.V.; Al-Hamdan, M.; Zhang, Y.; Insaf, T.Z.; et al. Comparison of multiple PM2.5 exposure products for estimating health benefits of emission controls over New York State, USA. Environ. Res. Lett. 2019, 14, 084023. [Google Scholar] [CrossRef]
  47. Kelly, J.T.; Jang, C.; Timin, B.; Di, Q.; Schwartz, J.; Liu, Y.; van Donkelaar, A.; Martin, R.V.; Berrocal, V.; Bell, M.L. Examining PM2.5 concentrations and exposure using multiple models. Environ. Res. 2021, 196, 110432. [Google Scholar] [CrossRef] [PubMed]
  48. Bravo, M.A.; Ebisu, K.; Dominici, F.; Wang, Y.; Peng, R.D.; Bell, M.L. Airborne fine particles and risk of hospital admissions for understudied populations: Effects by urbanicity and short-term cumulative exposures in 708 US counties. Environ. Health Perspect. 2017, 125, 594–601. [Google Scholar] [CrossRef] [PubMed]
  49. Grineski, S.E.; Renteria, R.A.; Collins, T.W.; Bakian, A.V.; Bilder, D.; VanDerslice, J.A.; Fraser, A.; Gomez, J.; Ramos, K.D. PM2.5 threshold exceedances during the prenatal period and risk of intellectual disability. J. Expo. Sci. Environ. Epidemiol. 2024, 34, 861–867. [Google Scholar] [CrossRef]
  50. Son, J.Y.; Lane, K.J.; Miranda, M.L.; Bell, M.L. Health disparities attributable to air pollutant exposure in North Carolina: Influence of residential environmental and social factors. Health Place 2020, 62, 102287. [Google Scholar] [CrossRef] [PubMed]
  51. O’Dell, K.; Ford, B.; Fischer, E.V.; Pierce, J.R. Contribution of wildland-fire smoke to US PM2.5 and its influence on recent trends. Environ. Sci. Technol. 2019, 53, 1797–1804. [Google Scholar] [CrossRef]
  52. Collins, T.W.; Grineski, S.E.; Morales, D.X. Sexual orientation, gender, and environmental injustice: Unequal carcinogenic air pollution risks in greater Houston. Ann. Am. Assoc. Geogr. 2017, 107, 72–92. [Google Scholar] [CrossRef]
  53. Liu, J.; Clark, L.P.; Bechle, M.J.; Hajat, A.; Kim, S.-Y.; Robinson, A.L.; Sheppard, L.; Szpiro, A.A.; Marshall, J.D. Disparities in air pollution exposure in the United States by race/ethnicity and income, 1990–2010. Environ. Health Perspect. 2021, 129, 127005. [Google Scholar] [CrossRef]
  54. Pope, R.; Wu, J.; Boone, C. Spatial patterns of air pollutants and social groups: A distributive environmental justice study in the phoenix metropolitan region of USA. Environ. Manag. 2016, 58, 753–766. [Google Scholar] [CrossRef]
  55. Pulido, L. Rethinking environmental racism: White privilege and urban development in Southern California. Ann. Assoc. Am. Geogr. 2000, 90, 12–40. [Google Scholar] [CrossRef]
  56. Brauer, M.; Brook, J.R.; Christidis, T.; Chu, Y.; Crouse, D.L.; Erickson, A.; Hystad, P.; Li, C.; Martin, R.V.; Meng, J.; et al. Mortality–air pollution associations in low exposure environments (MAPLE): Phase 2. Res. Rep. Health Eff. Inst. 2022, 2022, 212. Available online: https://pmc.ncbi.nlm.nih.gov/articles/PMC9556709/pdf/hei-2022-212.pdf (accessed on 11 April 2025).
  57. Brunekreef, B.; Strak, M.; Chen, J.; Andersen, Z.J.; Atkinson, R.; Bauwelinck, M.; Bellander, T.; Boutron, M.-C.; Brandt, J.; Carey, I.; et al. Mortality and morbidity effects of long-term exposure to low-level PM2.5, BC, NO2, and O3: An analysis of European cohorts in the ELAPSE project. Res. Rep. Health Eff. Inst. 2021, 2021, 208. Available online: https://pmc.ncbi.nlm.nih.gov/articles/PMC9476567/pdf/hei-2021-208.pdf (accessed on 11 April 2025). [PubMed]
  58. Burnett, R.; Chen, H.; Szyszkowicz, M.; Fann, N.; Hubbell, B.; Pope, C.A., III; Apte, J.S.; Brauer, M.; Cohen, A.; Weichenthal, S.; et al. Global estimates of mortality associated with long-term exposure to outdoor fine particulate matter. Proc. Natl. Acad. Sci. USA 2018, 115, 9592–9597. [Google Scholar] [CrossRef] [PubMed]
  59. Dominici, F.; Zanobetti, A.; Schwartz, J.; Braun, D.; Sabath, B.; Wu, X. Assessing adverse health effects of long-term exposure to low levels of ambient air pollution: Implementation of causal inference methods. Res. Rep. Health Eff. Inst. 2022, 2022, 211. Available online: https://pmc.ncbi.nlm.nih.gov/articles/PMC9530797/pdf/hei-2022-211.pdf (accessed on 11 April 2025). [PubMed]
  60. Pope, C.A., III; Burnett, R.T.; Krewski, D.; Jerrett, M.; Shi, Y.; Calle, E.E.; Thun, M.J. Cardiovascular mortality and exposure to airborne fine particulate matter and cigarette smoke: Shape of the exposure-response relationship. Circulation 2009, 120, 941–948. [Google Scholar] [CrossRef]
  61. Yazdi, M.D.; Wang, Y.; Di, Q.; Requia, W.J.; Wei, Y.; Shi, L.; Sabath, M.B.; Dominici, F.; Coull, B.; Evans, J.S.; et al. Long-term effect of exposure to lower concentrations of air pollution on mortality among US Medicare participants and vulnerable subgroups: A doubly-robust approach. Lancet Planet. Health 2021, 5, e689–e697. [Google Scholar] [CrossRef]
  62. Chen, L.W.A.; Tropp, R.J.; Li, W.W.; Zhu, D.; Chow, J.C.; Watson, J.G.; Zielinska, B. Aerosol and air toxics exposure in El Paso, Texas: A pilot study. Aerosol Air Qual. Res. 2012, 12, 169–179. [Google Scholar] [CrossRef]
Atmosphere 16 00610 g001
Figure 1. Census tract-level maps of the study area and dependent variables in the 2012–2016 period. (a) The study area, (b) mean PM2.5 exposure (long-term), (c) the number of days with a PM2.5 concentration ≥ 25 μg/m3 (short-term), and (d) the number of days with a PM2.5 concentration ≥ 35 μg/m3 (short-term). Note: To facilitate visualization, the map includes pollution values for non-metropolitan core area tracts, which are excluded from the statistical analyses (n = 1502).
Figure 1. Census tract-level maps of the study area and dependent variables in the 2012–2016 period. (a) The study area, (b) mean PM2.5 exposure (long-term), (c) the number of days with a PM2.5 concentration ≥ 25 μg/m3 (short-term), and (d) the number of days with a PM2.5 concentration ≥ 35 μg/m3 (short-term). Note: To facilitate visualization, the map includes pollution values for non-metropolitan core area tracts, which are excluded from the statistical analyses (n = 1502).
Atmosphere 16 00610 g001
Atmosphere 16 00610 g002
Figure 2. Census tract-level maps of focal independent variables in 2012–2016 period. (a) Proportion of Hispanic/Latino individuals, (b) proportion of Hispanic/Latino population ≥ 65 years of age, and (c) proportion of non-Hispanic White population ≥ 65 years of age. Note: To facilitate visualization, map includes values for non-metropolitan core area tracts, which are excluded from statistical analyses (n = 1502).
Figure 2. Census tract-level maps of focal independent variables in 2012–2016 period. (a) Proportion of Hispanic/Latino individuals, (b) proportion of Hispanic/Latino population ≥ 65 years of age, and (c) proportion of non-Hispanic White population ≥ 65 years of age. Note: To facilitate visualization, map includes values for non-metropolitan core area tracts, which are excluded from statistical analyses (n = 1502).
Atmosphere 16 00610 g002
Atmosphere 16 00610 g003
Figure 3. Estimated marginal means of the PM2.5 measures based on significant race/ethnicity, older age, and race/ethnicity by older age variables in percentiles in 2012–2016 period (n = 1208 census tracts). Note: Values are derived from Table 3 and were predicted by holding other independent variables in generalized estimating equations at their means.
Figure 3. Estimated marginal means of the PM2.5 measures based on significant race/ethnicity, older age, and race/ethnicity by older age variables in percentiles in 2012–2016 period (n = 1208 census tracts). Note: Values are derived from Table 3 and were predicted by holding other independent variables in generalized estimating equations at their means.
Atmosphere 16 00610 g003
Table 1. Descriptive statistics for variables analyzed for 2012–2016 period (n = 1208 census tracts).
Table 1. Descriptive statistics for variables analyzed for 2012–2016 period (n = 1208 census tracts).
Analysis VariableMin.–Max.Mean (Std. Dev.)
Dependent Variables
Mean PM2.5 (μg/m3)5.735–9.9168.675 (1.023)
Number of days with PM2.5 ≥ 25 µg/m30–207.010 (3.983)
Number of days with PM2.5 ≥ 35 µg/m30–60.950 (1.189)
Independent Variables
Prop. Hispanic/Latino ≥ 65 years of age0.000–0.6940.092 (0.057)
Prop. White ≥ 65 years of age0.000–1.0000.234 (0.167)
Prop. Hispanic/Latino0.014–1.0000.546 (0.323)
Prop. White (non-Hispanic)0.000–0.9180.336 (0.275)
Prop. Black (non-Hispanic)0.000–0.3730.031 (0.044)
Prop. American Indian (non-Hispanic) a0.000–0.6460.005 (0.029)
Prop. Asian (non-Hispanic)0.000–0.7170.058 (0.095)
Prop. Pacific Islander (non-Hispanic) b0.000–0.0740.002 (0.007)
Prop. Other/Multi-Race (non-Hispanic)0.000–0.1120.021 (0.021)
Prop. Total Population ≥ 65 Years of Age0.011–0.7530.143 (0.077)
Median Household Income (2016, USD)11,497–185,72456,144.073 (29,464.540)
Prop. Renter Occupancy0.034–0.9450.428 (0.216)
Population Density (People per Sq. Mile)7.740–50,077.9185544.455 (4820.609)
a Includes Alaska Native individuals. b Includes Native Hawaiian individuals.
Table 2. Metropolitan population-weighted mean PM2.5 exposures by focal group and US–Mexico border state, 2012–2016.
Table 2. Metropolitan population-weighted mean PM2.5 exposures by focal group and US–Mexico border state, 2012–2016.
Focal Group by StateMean PM2.5 (μg/m3)Number of Days PM2.5 ≥ 25 µg/m3Number of Days PM2.5 ≥ 35 µg/m3
California
Total Population9.4708.2390.822
White9.4387.8460.756
Hispanic/Latino9.4798.6020.925
White ≥ 65 years9.4327.7450.751
Hispanic/Latino ≥ 65 years9.4659.1021.071
Arizona
Total Population6.4891.8380.451
White6.3421.7710.373
Hispanic/Latino6.6311.9140.541
White ≥ 65 years6.2841.6890.533
Hispanic/Latino ≥ 65 years6.6001.9120.602
New Mexico
Total Population7.1994.2731.490
White7.0183.2981.260
Hispanic/Latino7.2774.7361.587
White ≥ 65 years6.9113.0781.160
Hispanic/Latino ≥ 65 years7.2544.6771.542
Texas
Total Population8.1665.5560.961
White8.0866.2991.353
Hispanic/Latino8.1735.4120.890
White ≥ 65 years8.1435.7010.994
Hispanic/Latino ≥ 65 years8.1755.7081.033
Table 3. Results of generalized estimating equations predicting PM2.5 exposure measures for metropolitan US–Mexico border census tracts, 2012–2016 (n = 1208 census tracts).
Table 3. Results of generalized estimating equations predicting PM2.5 exposure measures for metropolitan US–Mexico border census tracts, 2012–2016 (n = 1208 census tracts).
(a) Mean PM2.5 (μg/m3) b(b) Number of Days PM2.5 ≥ 25 µg/m3 c(c) Number of Days PM2.5 ≥ 35 µg/m3 c
Parameter aB (95% CI)Exp(B)pB (95% CI)Exp(B)pB (95% CI)Exp(B)p
Base Model Variables
Intercept2.093 (2.060, 2.125)8.108<0.0011.871 (1.716, 2.026)6.493<0.0010.063 (−0.210, 0.337)1.0650.649
Prop. Hispanic/Latino0.015 (0.008, 0.022)1.015<0.0010.210 (0.123, 0.297)1.234<0.0010.329 (0.200, 0.457)1.389<0.001
Prop Black0.006 (0.003, 0.008)1.006<0.0010.053 (0.020, 0.087)1.0550.0020.027 (−0.017, 0.072)1.0280.230
Prop. American Indian−0.003 (−0.009, 0.002)0.9970.242−0.034 (−0.095, 0.027)0.9670.2770.012 (−0.048, 0.071)1.0120.697
Prop. Asian0.007 (0.006, 0.009)1.007<0.0010.070 (0.052, 0.088)1.073<0.0010.120 (0.080, 0.159)1.127<0.001
Prop. Pacific Islander0.000 (−0.001, 0.001)1.0000.5570.013 (0.002, 0.024)1.0130.016−0.019 (−0.045, 0.007)0.9810.147
Prop. Other/Multi-Race0.000 (−0.004, 0.005)1.0000.850−0.006 (−0.054, 0.042)0.9940.794−0.060 (−0.131, 0.011)0.9420.096
Prop. Total Pop. ≥ 65 years0.002 (−0.001, 0.005)1.0020.1820.057 (0.019, 0.094)1.0580.0030.076 (0.031, 0.120)1.0790.001
Med. Household Income0.015 (0.0062, 0.003)1.0160.0130.471 (0.303, 0.639)1.602<0.0010.458 (0.250, 0.665)1.580<0.001
Med. Household Income (sq.)−0.006 (−0.014, 0.002)0.9940.119−0.186 (−0.305, −0.067)0.8300.002−0.126 (−0.256, 0.004)0.8810.058
Prop. Renter Occupancy0.008 (0.004, 0.012)1.008<0.0010.197 (0.141, 0.253)1.218<0.0010.221 (0.162, 0.280)1.247<0.001
Population Density0.000 (−0.001, 0.002)1.0000.8410.041 (0.001, 0.081)1.0420.0440.041 (0.002, 0.081)1.0420.039
Older Hispanic/Latino Variables d
Prop. Hispanic/Latino ≥ 65 e0.001 (0.000, 0.002)1.0010.0310.058 (0.032, 0.084)1.059<0.0010.050 (0.016, 0.083)1.0510.004
Prop White e−0.011 (−0.017, −0.006)0.989<0.001−0.150 (−0.22, −0.080)0.861<0.001−0.241 (−0.341, −0.142)0.786<0.001
Older White Variable f
Prop. White ≥ 65 g0.002 (0.000, 0.004) 0.0600.025 (−0.004, 0.055) 0.0930.044 (−0.003, 0.090) 0.066
Notes: a Each independent variable is standardized. b Model (a) uses an exchangeable correlation matrix, gamma distribution, and log link function and adjusts for clustering by county and age of housing stock; reference for race/ethnicity variables is Prop. White; reference for Prop. Total Pop. ≥ 65 years is Prop Total Pop. < 65 years; reference for Prop. Renter Occupancy is Prop Owner Occupancy. c Models (b) and (c) use exchangeable correlation matrix, negative binomial distribution, and log link function and adjust for clustering by county and housing age. d Separate model results with Prop. Hispanic/Latino replaced by Prop. Hispanic/Latino ≥ 65 and Prop. White variables and otherwise identical specifications to base model; associations for Percent Renter Occupancy, Median Household Income (quadratic), and Population Density are same in terms of direction and significance. e Reference: Prop. Hispanic/Latino < 65 years of age. f Separate model results including Prop. White ≥ 65 variable and otherwise identical specifications to base model; associations for Percent Renter Occupancy, Median Household Income (quadratic), and Population Density are same in terms of direction and significance. g Reference: Prop. White < 65 years of age.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Collins, T.W.; Child, C.M.; Grineski, S.E.; Scott, M. Disparities in Fine Particulate Matter Air Pollution Exposures at the US–Mexico Border: The Intersection of Race/Ethnicity and Older Age. Atmosphere 2025, 16, 610. https://doi.org/10.3390/atmos16050610

AMA Style

Collins TW, Child CM, Grineski SE, Scott M. Disparities in Fine Particulate Matter Air Pollution Exposures at the US–Mexico Border: The Intersection of Race/Ethnicity and Older Age. Atmosphere. 2025; 16(5):610. https://doi.org/10.3390/atmos16050610

Chicago/Turabian Style

Collins, Timothy W., Colby M. Child, Sara E. Grineski, and Mathilda Scott. 2025. "Disparities in Fine Particulate Matter Air Pollution Exposures at the US–Mexico Border: The Intersection of Race/Ethnicity and Older Age" Atmosphere 16, no. 5: 610. https://doi.org/10.3390/atmos16050610

APA Style

Collins, T. W., Child, C. M., Grineski, S. E., & Scott, M. (2025). Disparities in Fine Particulate Matter Air Pollution Exposures at the US–Mexico Border: The Intersection of Race/Ethnicity and Older Age. Atmosphere, 16(5), 610. https://doi.org/10.3390/atmos16050610

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop