Racial/Ethnic Disparities in Short-Term PM2.5 Air Pollution Exposures in the United States.

Introduction

In 2019, fine particulate matter [, particulate matter (PM) with aerodynamic diameter of ] air pollution was responsible for approximately and 50,000 premature deaths worldwide and in the United States, respectively.[1] Distinct health effects of short-term (i.e., hours-to-days) exposures are well-documented.[2] Although studies have consistently found disparities in long-term exposures for people of color (POC) in the United States,[3-6] racial/ethnic disparities in short-term exposures nationwide have not been investigated. We aimed to advance knowledge by estimating short-term exposure disparities for racial/ethnic groups nationwide and within the U.S. Environmental Protection Agency (U.S. EPA) regions.

Methods

We used 2012–2016 American Community Survey estimates of census tract composition for the four largest racial/ethnic groups in the United States: a) non-Hispanic/Latina/x/o (NH) White, b) Hispanic/Latina/x/o of any race(s), c) NH Black, and d) NH Asian. We also analyzed e) a “POC” group [i.e., people of Hispanic/Latina/x/o ethnicity and/or non-White race(s)].

To estimate short-term exposures for contiguous U.S. census tracts (), we used the high-performing U.S. EPA downscaler model.[7,8] We calculated three exposure estimates of the counts of days during the period 2012–2016 wherein the census tract–level 24-h concentration equaled or exceeded a) [the current 24-h mean World Health Organization (WHO) guideline[9]], b) (2006–2021 24-h mean WHO guideline[9]), and c) [24-h mean National Ambient Air Quality Standard (NAAQS)[10]]. Following methods of Lui et al.,[6] we then calculated the population-weighted mean counts of days during the period 2012–2016 that the concentration was , 25, and for each racial/ethnic group and the total population, nationally and by U.S. EPA region.

Nationwide, we examined two absolute exposure disparity metrics. First, following methods of Lui et al.,[6] we derived the percent difference relative to the total population-weighted mean exposure (PWME) (Table 1, “Total Population” column) for the racial/ethnic groups with the highest and lowest PWME for each threshold {[]}. Second, we calculated the percent difference relative to the total PWME for POC PWME minus NH White PWME. We also calculated two relative PWME disparity metrics for each threshold: a) highest PWME divided by lowest PWME, and b) POC PWME divided by NH White PWME.

Table 1

Population distribution and population-weighted exposure distribution for short-term exposures nationwide and by Environmental Protection Agency (U.S. EPA) region for main racial/ethnic groups and total populations, contiguous United States, 2012–2016.

Demographic	NH White	NH Black	Hispanic/Latina/x/o	NH Asian	People of color	Total population
Nationwide (population)	196,510,991	39,047,978	54,993,788	15,856,777	119,781,899	316,292,890
Proportion of population	62%	12%	17%	5.0%	38%	100%
Days PM_2.5 ≥15 μg/m^3	169.6 (99.7)	197.7 (101.8)	188.5 (132.2)	213.4 (106.7)	192.7 (118.2)	178.3 (107.7)
Days PM_2.5 ≥25 μg/m^3	15.1 (20.4)	15.5 (18.9)	24.9 (36.6)	27.2 (28.3)	21.5 (30.2)	17.6 (24.8)
Days PM_2.5 ≥35 μg/m^3	2.8 (8.2)	2.3 (6.4)	6.0 (16.3)	5.1 (11.9)	4.5 (12.8)	3.4 (10.2)
U.S. EPA region 1 (population)	11,257,770	886,291	1,492,014	650,576	3,411,109	14,668,879
Proportion of population	77% (5.7%)	6.0 (2.3%)	10% (2.7%)	4.4% (4.1%)	23% (2.8%)	100% (4.6%)
Days PM_2.5 ≥15 μg/m^3	94.9 (46.1)	124.7 (47.3)	123.2 (46.6)	109.2 (40.5)	118.1 (46.5)	100.3 (47.2)
Days PM_2.5 ≥25 μg/m^3	3.8 (5.0)	6.6 (6.6)	6.4 (6.3)	4.7 (5.3)	5.9 (6.2)	4.3 (5.4)
Days PM_2.5 ≥35 μg/m^3	0.1 (0.3)	0.3 (0.5)	0.3 (0.5)	0.2 (0.4)	0.3 (0.4)	0.2 (0.4)
U.S. EPA region 2 (population)	16,111,254	3,963,201	5,380,847	2,400,047	12,444,180	28,555,434
Proportion of population	56% (8.2%)	14% (10%)	19% (10%)	8.4% (15%)	44% (10%)	100% (9.0%)
Days PM_2.5 ≥15 μg/m^3	171.2 (62.3)	213.2 (46.5)	218.1 (46.5)	215.5 (41.0)	214.6 (46.6)	190.1 (60.0)
Days PM_2.5 ≥25 μg/m^3	15.3 (12.1)	23.4 (10.4)	25.4 (10.2)	24.1 (8.8)	24.2 (10.3)	19.1 (12.2)
Days PM_2.5 ≥35 μg/m^3	0.8 (1.2)	1.5 (1.4)	1.34 (1.1)	1.3 (0.9)	1.4 (1.2)	1.1 (1.2)
U.S. EPA region 3 (population)	20,839,115	5,231,239	2,296,353	1,329,996	9,648,566	30,487,681
Proportion of population	68% (10.6%)	17% (13%)	7.5% (4.2%)	4.4% (8.4%)	32% (8.1%)	100% (9.6)
Days PM_2.5 ≥15 μg/m^3	213.9 (99.9)	206.2 (86.6)	223.6 (87.8)	219.7 (78.3)	212.0 (86.7)	213.3 (95.9)
Days PM_2.5 ≥25 μg/m^3	19.7 (20.1)	18.7 (15.9)	24.3 (21.0)	21.1 (15.9)	20.3 (17.6)	19.9 (19.4)
Days PM_2.5 ≥35 μg/m^3	2.2 (3.7)	1.8 (3.2)	3.0 (4.3)	2.3 (3.3)	2.2 (3.5)	2.2 (3.6)
U.S. EPA region 4 (population)	39,555,786	13,431,505	7,631,465	1,453,499	24,043,772	63,599,558
Proportion of population	62% (20%)	21% (34%)	12% (14%)	2.3% (9.2%)	38% (20%)	100% (20%)
Days PM_2.5 ≥15 μg/m^3	132.1 (80.3)	150.9 (83.9)	90.5 (79.8)	144.1 (89.0)	130.1 (87.2)	131.4 (83.0)
Days PM_2.5 ≥25 μg/m^3	4.0 (4.0)	4.4 (3.9)	2.7 (3.1)	4.5 (3.9)	3.9 (3.8)	4.0 (3.9)
Days PM_2.5 ≥35 μg/m^3	0.9 (1.5)	0.9 (1.2)	0.6 (1.1)	1.0 (1.2)	0.8 (1.2)	0.9 (1.4)
U.S. EPA region 5 (population)	39,209,672	5,835,394	4,088,540	1,671,462	12,933,797	52,143,469
Proportion of population	75% (20%)	11% (15%)	7.8% (7.4%)	3.2% (11%)	25% (11%)	100% (16%)
Days PM_2.5 ≥15 μg/m^3	255.2 (86.7)	317.6 (68.5)	315.8 (83.6)	291.7 (79.8)	307.3 (79.9)	268.1 (85.0)
Days PM_2.5 ≥25 μg/m^3	20.9 (10.3)	29.5 (11.0)	31.3 (12.4)	27.7 (11.5)	29.1 (11.9)	23.0 (11.3)
Days PM_2.5 ≥35 μg/m^3	2.5 (2.7)	3.9 (3.6)	5.6 (3.8)	4.5 (3.6)	4.4 (3.7)	3.0 (3.1)
U.S. EPA region 6 (population)	20,037,701	5,390,567	12,220,905	1,385,262	20,491,134	40,528,835
Proportion of population	49% (10%)	13% (14%)	30% (22%)	3.4% (8.7%)	51% (17%)	100% (13%)
Days PM_2.5 ≥15 μg/m^3	129.6 (64.1)	157.6 (65.9)	125.9 (80.6)	177.4 (68.9)	137.1 (77.4)	133.4 (71.3)
Days PM_2.5 ≥25 μg/m^3	4.6 (3.8)	5.9 (5.0)	6.3 (5.1)	7.9 (4.7)	6.1 (5.0)	5.4 (4.5)
Days PM_2.5 ≥35 μg/m^3	0.1 (0.4)	0.2 (0.4)	0.3 (0.8)	0.2 (0.5)	0.3 (0.7)	0.2 (0.6)
U.S. EPA region 7 (population)	11,290,564	1,046,341	929,532	289,508	2,655,227	13,945,791
Proportion of population	80.1% (5.7%)	7.5% (2.7%)	6.7% (1.7%)	2.1% (1.8%)	19% (2.2%)	100% (4.4%)
Days PM_2.5 ≥15 μg/m^3	158.3 (70.0)	229.1 (90.4)	140.2 (71.3)	182.3 (76.5)	182.9 (89.1)	163.0 (74.7)
Days PM_2.5 ≥25 μg/m^3	11.4 (7.5)	16.8 (8.5)	10.8 (7.2)	14.1 (7.4)	13.6 (8.3)	11.8 (7.7)
Days PM_2.5 ≥35 μg/m^3	2.0 (1.8)	3.7 (2.5)	1.9 (1.6)	2.7 (2.1)	2.7 (2.3)	2.1 (1.9)
U.S. EPA region 8 (population)	8,756,581	276,089	1,673,124	252,659	2,730,518	11,487,099
Proportion of population	76% (4.5%)	2.4% (0.7%)	15% (3.0%)	2.2% (1.6%)	24% (2.3%)	100% (3.6%)
Days PM_2.5 ≥15 μg/m^3	87.3 (68.9)	89.5 (55.7)	97.6 (69.4)	105.3 (67.8)	93.6 (68.3)	88.8 (68.8)
Days PM_2.5 ≥25 μg/m^3	26.0 (31.8)	20.1 (24.9)	27.6 (32.6)	30.7 (34.2)	26.2 (31.7)	26.1 (31.8)
Days PM_2.5 ≥35 μg/m^3	12.0 (17.8)	7.5 (14.4)	11.9 (19.1)	14.0 (20.5)	11.5 (18.5)	11.9 (18.0)
U.S. EPA region 9 (population)	20,051,810	2,660,265	17,736,338	5,697,876	28,133,438	48,185,248
Proportion of population	42% (10%)	5.5% (6.8%)	37% (32%)	12% (36%)	58% (23%)	100% (15%)
Days PM_2.5 ≥15 μg/m^3	170.3 (135.9)	241.9 (152.0)	252.0 (161.5)	242.0 (130.2)	243.6 (154.5)	213.1 (151.4)
Days PM_2.5 ≥25 μg/m^3	30.6 (42.8)	42.7 (43.7)	47.6 (53.6)	43.9 (37.8)	45.3 (49.2)	39.2 (47.2)
Days PM_2.5 ≥35 μg/m^3	8.3 (19.1)	11.0 (20.1)	14.2 (25.7)	10.4 (17.6)	12.7 (23.3)	10.9 (21.8)
U.S. EPA region 10 (population)	9,400,738	327,086	1,544,670	725,892	3,290,158	12,690,896
Proportion of population	74% (4.8%)	2.6% (0.8%)	12% (2.8%)	5.7% (4.6%)	26% (2.7%)	100% (4.0%)
Days PM_2.5 ≥15 μg/m^3	132.5 (59.2)	145.6 (39.6)	138.1 (50.4)	141.3 (39.7)	138.2 (48.3)	134.0 (56.7)
Days PM_2.5 ≥25 μg/m^3	24.5 (16.2)	26.4 (11.5)	26.3 (13.4)	23.3 (11.1)	25.2 (13.1)	24.7 (15.5)
Days PM_2.5 ≥35 μg/m^3	6.4 (6.3)	4.6 (5.1)	6.6 (5.2)	3.8 (4.9)	5.5 (5.4)	6.2 (6.1)

Note: Rows reporting “days , 25, and ” provide mean values with standard deviations in parentheses; rows reporting “proportion of population” provide racial/ethnic group percentages of total populations nationwide and by U.S. EPA region, and, in parentheses, U.S. EPA region racial/ethnic group percentages of corresponding racial/ethnic group populations nationwide; U.S. EPA Region 1 includes Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont; U.S. EPA Region 2 includes New Jersey, New York, and 8 federally recognized Indian Nations; U.S. EPA Region 3 includes Delaware, Maryland, Pennsylvania, Virginia, Washington D.C., West Virginia, and 7 federally recognized tribes; U.S. EPA Region 4 includes Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, and Tennessee; U.S. EPA Region 5 includes Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin; U.S. EPA Region 6 includes Arkansas, Louisiana, New Mexico, Oklahoma, and Texas; U.S. EPA Region 7 includes Iowa, Kansas, Missouri, and Nebraska; U.S. EPA Region 8 includes Colorado, Montana, North Dakota, South Dakota, Utah, and Wyoming; U.S. EPA Region 9 includes Arizona, California, and Nevada; U.S. EPA Region 10 includes Idaho, Oregon, Washington, and 271 native tribes. Hispanic/Latina/x/o includes any race(s). NH, non-Hispanic/Latina/x/o; , fine particulate matter with an aerodynamic diameter of .

For the U.S. EPA regions, we examined two absolute exposure disparity metrics at each threshold, following methods of Lui et al.[6] First, we calculated the normalized regional disparity as the U.S. EPA regional PWME in each racial/ethnic group minus the U.S. EPA regional total PWME, divided by the national total PWME. Thus, we analyzed 30 exposure threshold-U.S. EPA region combinations () and 120 exposure threshold-EPA region-racial/ethnic group combinations (). Second, we calculated the normalized regional disparity for POC vs. NH White as the U.S. EPA regional POC PWME minus the U.S. EPA regional NH White PWME, divided by the national total PWME.

Results

PWMEs for the period 2012–2016 were higher for the Hispanic/Latina/x/o, NH Black, and NH Asian groups, as well as for POC overall, than for the NH White group across the short-term thresholds, with one exception (NH Black ) (Table 1). PWMEs were highest among NH Asian people for days with and and highest among Hispanic/Latina/x/o people for days with . Disparities between racial/ethnic groups with highest vs. lowest exposures were largest for days with [absolute disparity: 109% (3.7 d), relative disparity (ratio): 2.7]; intermediate for days with [69% (12 d), 1.8]; and lowest for days with [25% (44 d), 1.3]. We observed similarly patterned national exposure disparities across the thresholds for POC overall vs. NH White.

Several patterns emerged across U.S. EPA regions (Figure 1, Table 1). First, disparate exposures were present in each EPA region such that one or more racial/ethnic group experienced exposure disparities greater than the national PWME for one or more threshold. Second, the magnitude and directionality of exposure disparities exhibited greater regional variation for days with than for days with and . Third, POC groups within regions were much more likely than NH White people to have greater than regional average exposure, whereas NH White people were more likely than POC groups to have less than regional average exposure (Figure 1). For example, of the 120 exposure threshold-U.S. EPA region-racial/ethnic group combinations, 58% (52 of 90) vs. 3.3% (1 of 30) were above average for the three POC groups vs. the NH White group, respectively. In sum, short-term exposure disparities burdening POC groups and POC overall varied regionally but prevailed across much of the United States.

Figure 1.

U.S. EPA region racial/ethnic disparities in short-term exposure, 2012–2016, showing the difference between (1) NH White vs. U.S. EPA region average, (2) NH Black vs. U.S. EPA region average, (3) Hispanic/Latina/x/o vs. U.S. EPA region average, (4) NH Asian vs. U.S. EPA region average, and (5) People of color vs. NH White for the three short-term exposure measures. (A) Days concentration ; (B) Days concentration ; and (C) Days concentration . Columns 1–4: exposure disparity relative to U.S. EPA region average; calculated as mean exposure for a racial/ethnic group in that U.S. EPA region minus the overall mean for that U.S. EPA region, then divided by the national overall mean. Column 5: exposure disparity for people of color (i.e., people of Hispanic/Latina/x/o ethnicity and/or of any non-White race) minus mean exposure for NH White people, then divided by the national overall mean. Mean values are population-weighted. U.S. EPA regions displayed in white indicate that the disparity is within of the national overall mean. Orange dotted shading indicates that mean exposures are higher than average by more than 5% of the national overall mean (columns 1–4) or that mean exposures are higher for people of color than for NH White people by more than 5% of the national overall mean (column 5). Green solid shading indicates the reverse: Mean exposures are lower than average for that group (columns 1–4), or mean exposures are lower for people of color than for NH White people (column 5), and the disparity is of the national overall mean. Note: Hispanic/Latina/x/o includes any race(s); NH, non-Hispanic/Latina/x/o; , fine particulate matter with an aerodynamic diameter of .

Discussion

This national study of short-term air pollution exposure disparities revealed an environmentally unjust pattern for POC. Nationally, POC overall experienced 23, 6.4, and 1.7 more days with respective concentrations , 25, and than NH White people (during the period 2012–2016). Disparities appear larger for short-term vs. long-term exposures nationwide. Although prior estimates of disparities in long-term exposures () between the most and least exposed racial/ethnic groups were respectively 13%–14% (absolute) and 1.1 (relative),[5,6] we found worsening disparities as the short-term threshold increased from (absolute: 25% to 69% to 109%; relative: 1.3 to 1.8 to 2.7). This suggests that tightening/lowering the NAAQS toward WHO benchmarks might attenuate racial/ethnic exposure disparities. The short-term -exposure disparities we documented for the Hispanic/Latina/x/o and NH Asian groups nationally align with long-term exposure studies.[3-6] Results indicating that NH Black Americans were the least exposed group to days with diverge from long-term exposure studies[3-6] and are attributable to the fact that 61.4% of NH Black Americans resided in U.S. EPA regions 3, 4, and 6 (Southeastern and mid-Atlantic states), where such exposures were generally less frequent (Table 1). Most short-term exposures for POC populations within U.S. EPA regions were above average, and few were below average. Given the health risks of short-term ,[2] these disproportionate exposures presumably have disparate health effects, which should be examined.