Short-Term Associations of Traffic-Related Air Pollution with Cardiorespiratory Outcomes Among Low-Income Residents from a US–Mexico Border Community

Abstract

Exposure to traffic-related air pollution is not merely linked to respiratory health issues but also poses significant risks to cardiovascular well-being. Individuals from lower-income communities residing in high-pollution zones are particularly vulnerable to adverse cardiorespiratory health impacts. Pollutants such as fine particulate matter (PM_2.5 and PM₁₀), nitrogen dioxide (NO₂), and ozone (O₃) are recognized as a leading, yet preventable, contributor to cardiorespiratory diseases. Although research has extensively explored the short-term impact of these pollutants on respiratory health, the immediate effects on cardiovascular outcomes require further study. We explored associations of traffic-related air pollutants with airway inflammation, lung function, and cardiovascular health outcomes (metabolic syndrome [MetS]) collected from a sample of low-income participants (N = 662) from a US–Mexico border county. Airway inflammation was measured using exhaled nitric oxide tests (eNO), while lung function parameters were measured by spirometry. MetS risk factors (waist circumference, blood pressure, triglycerides, HDL, and fasting blood glucose) were also measured. While spirometry measures were negatively associated with air pollutants (p < 0.05), no associations were noted for eNO. We also found positive associations in linear and logistic models between air pollutants and obesity (BMI: p < 0.04; waist: p < 0.03), fasting blood glucose (p < 0.03), and metabolic syndrome (p < 0.04). These findings reaffirm the immediate adverse effects of air pollution on respiratory function and shed light on its broader metabolic consequences. Environmental and neighborhood conditions could potentially influence the associations with obesity. At the same time, the links between fasting glucose and metabolic syndrome might indicate underlying oxidative stress and systemic inflammation.

1. Introduction

1.1. Effects of Air Pollutants on Cardiorespiratory Health

People living in areas with higher exposure to air pollution, compared with those in less polluted areas, have higher mortality rates and stronger associations with cardiorespiratory disease [1,2,3]. Also, those living in areas with high air pollution have an increased likelihood of heart and lung disease, irregular heartbeat, aggravated asthma, and decreased lung function [4,5,6,7]. These findings make it increasingly clear that air pollution is a critical public health issue requiring immediate and sustained intervention.

In recent decades, many cardiorespiratory biomarkers concerning air pollution exposure have been identified [8] and can be considered valuable indices of air pollution exposure, even if not all of them are in the causal pathway for developing a disease [9]. For example, exhaled nitric oxide (eNO) is a biomarker of airway inflammation, an essential determinant of respiratory outcomes and disease [10]. Lung function measured by spirometry offers respiratory markers affected by exposure to air pollutants in healthy adults and those with pre-existing lung disease [11]. Both airway inflammation and decreased lung function have been associated with the effects of traffic-related pollutants in children and adolescents [12,13,14].

There is also evidence of relationships between air pollutants and adverse cardiovascular outcomes. Studies showed that yearly mean concentrations of particulate matter have been associated with higher hospitalization risks, congestive heart failure, and recurrent heart attacks among patients with previous myocardial infarction [15,16]. Recent reviews have highlighted the associations between traffic-related air pollutants and components of metabolic syndrome—a predictor of cardiovascular disease—such as waist circumference, blood pressure, triglycerides, HDL-cholesterol, fasting glucose, and HbA1c, emphasizing the potential metabolic risks posed by air pollution exposure [17,18]. A meta-analysis suggested that short-term exposure to some air pollutants, such as particulate matter, NO₂, and O₃, may increase the risk of hypertension [19]. Another meta-analysis concluded that the strong to moderate evidence linking long-term exposure to traffic-related air pollution (TRAP) with various adverse health outcomes warrants increased attention from the public and policymakers [20].

1.2. Short-Term Exposure Assessments

Studies assessing exposures to TRAP often include exposure to particulate matter (PM_2.5 and PM₁₀), nitrogen dioxide (NO₂), and ozone (O₃), which are considered major preventable causes of respiratory diseases [21]. Research about the short-term effects of exposure to these air pollutants has also linked them with cardio-respiratory mortality [22]. Furthermore, standard study designs have used time-series or cross-sectional analyses to report associations between elevated air pollutant concentrations over short periods (one or several days) and increased cardiovascular mortality and morbidity [23,24]. However, the precise exposure window for biomarkers is not clearly defined and differs according to the study.

Chuang and collaborators [25] applied mixed models to examine the associations between air pollutants and metabolic markers (systolic blood pressure [SBP], diastolic blood pressure [DBP], HDL-cholesterol [HDL], LDL-cholesterol [LDL], fasting blood glucose [FBG], and HbA1c). The exposure variables included PM, NO₂, and O₃ levels on the same day (24 h mean) and 48 to 144 h standards before the day of the health measurements. They observed that increased PM₁₀ was associated with elevated SBP, triglycerides, hbA1c, and reduced HDL. Bell and colleagues [26] estimated exposure to ambient PM_2.5 based on residential addresses using short-term averaging periods on the day of the blood draw and the day before and a moving mean of the previous five days with HDL and found that exposure to air pollution was adversely associated. The MESA study [27], an extensive epidemiological research, assigned a daily exposure measure from the continuous ambient monitoring station (CAMS) nearest to the participants’ residences and constructed five exposure measures: PM_2.5 the day before measurement and mean concentrations over the 2, 7, 30, and 60 days before measurement using MetS as a modifying factor. The authors found that associations were stronger among individuals with MetS compared to those without MetS. Similarly, Zhu and colleagues [28] investigated the acute effects of traffic-related air pollution (TRAP) on the onset of ST-segment elevation myocardial infarction and associated cardiac impairments. This evidence underscores the acute cardiovascular risks posed by TRAP exposure and highlights the importance of considering localized environmental factors, such as urban greenness, in mitigating these risks.

Statistical models that associate air pollution exposure with cardiorespiratory outcomes also vary according to the study. A study among patients with type 2 diabetes in China considered spline and multiple linear regressions between short-term exposure to PM₁₀, SO₂, and NO₂ and total cholesterol (TC), triglycerides (TGs), and HDL cholesterol [29]. A study among Mexican Americans in Southern California used short-term exposure considering up to 58 days of cumulative daily means of PM_2.5 to find associations with lower insulin sensitivity, a lower HDL-to-LDL ratio, higher fasting glucose, and higher total cholesterol [30].

Previous work by our team employed land-use regression models to investigate the relationship between transportation parameters, spatial variations in air pollution, and MetS risk factors in low-income communities within El Paso, Texas [31]. This study demonstrated that proximity to major traffic sources, including highways and international ports of entry, was strongly associated with increased body mass index (BMI), waist circumference, triglycerides, and glucose levels. Using Geographic Information System (GIS) tools, we identified transportation parameters such as street length and traffic density within 500 m and 1000 m impact zones around participants’ residences as significant predictors of air pollution exposure and health outcomes.

Building upon these findings, the current study expands to assess short-term exposure to traffic-related air pollutants, including PM₁₀, PM_2.5, NO₂, and O₃, and their associations with cardiorespiratory outcomes. Leveraging data from CAMSs located near participants’ residences, this study captures pollutant concentrations averaged over 24 h to 96 h periods prior to health assessments. By linking these short-term exposure metrics to markers of airway inflammation, lung function, and MetS, this study advances our understanding of the immediate health effects of traffic-related air pollution while drawing upon the spatial and transportation-related context established in our prior work.

2. Materials and Methods

2.1. Setting, Population, and Sample

We conducted our data analysis using health outcome measurements collected from an epidemiological study entitled “Evidence-based Screening for Obesity, Cardiorespiratory Disease, and Environmental Exposures in Low-income El Paso Households”, funded by the City of El Paso’s Department of Public Health, which included low-income participants from El Paso, TX, measured between September 2014 and January 2020. Participants included were residents living within low-income communities in El Paso County. The participants were recruited from community gathering sites, including housing centers, faith-based organizations, and neighborhood events, to ensure broad coverage of the targeted population. Low-income status was defined based on self-reported income thresholds consistent with local and federal eligibility criteria.

2.2. Cardiorespiratory Outcome Measures

The methodology of the epidemiological study is described elsewhere [31,32], but briefly, biological outcomes were assessed on-site using standardized procedures. Anthropometric measurements were collected using calibrated equipment, including height, weight, and waist circumference. Body mass index (BMI) was calculated as weight in kilograms divided by height in square meters. Metabolic syndrome risk factors, including fasting blood glucose, triglycerides, HDL, and blood pressure, were measured using a portable LDX Analyzer (Abbott) and an automated blood pressure monitor (Omron).

Airway inflammation was assessed via exhaled nitric oxide (eNO) using the NiOX device (Aerocrine), and lung function was evaluated through spirometry using a MicroLoop Spirometer (CareFusion), which measured forced vital capacity (FVC), forced expiratory volume in one second (FEV₁), and peak expiratory flow (PEF). These measures provide objective assessments of respiratory function, with lower values indicating worse outcomes, such as airway obstruction or reduced lung capacity.

2.3. Air Pollutant Data Collection and Validation

Traffic-related air pollutant data for PM₁₀, PM_2.5, NO₂, and O₃ were extracted using publicly available datasets from CAMSs maintained by the Texas Commission on Environmental Quality (TCEQ). Each participant was assigned to the most representative CAMS based on their residential address; if a participant lived near a CAMS with no data or lived far away from another CAMS, they were excluded from the analysis. The CAMSs included in this study were located in areas heavily influenced by vehicular traffic, such as major roadways and international border crossings, where emissions from idling and moving vehicles significantly contribute to air pollution. These locations were specifically selected to capture traffic-related pollutant concentrations in the region (Figure 1). Previous studies conducted in the Paso del Norte region have confirmed that vehicular emissions are a dominant source of PM_2.5 and NO₂ [33,34].

Short-term exposures refer to the mean pollutant concentrations averaged over the 24, 48, 72, and 96-hour periods before each participant’s examination date. Data validation included the following steps: only data from CAMSs with at least 80% valid hourly data over the study period were included; null and outlier values (defined as values exceeding three standard deviations from the mean for each station) were excluded from the dataset; and a sensitivity analysis was conducted to assess the impact of data gaps on the calculated pollutant averages. Atmospheric variables, such as temperature, humidity, and wind speed, were not included in the analysis due to data limitations. Still, we acknowledge their potential impact on both air pollution levels and health outcomes.

2.4. Data Analysis

The continuous variables in this study include respiratory markers (eNO, FVC, FEV₁, PEF) and cardiovascular risk markers (waist circumference, SBP, DBP, TGs, HDL, and FBG). For further statistical analysis, waist circumference, SBP, DBP, TGs, HDL, and FBG were coded as binary variables (high or low) to determine whether a participant has metabolic syndrome (MetS) following the current diagnostic criteria [35].

Summary statistics of participant demographic information and characteristics were calculated. Age was treated as a continuous variable in all statistical models to preserve statistical power and avoid the dilution of effects. Spearman correlation analyses were conducted to explore relationships between outcome variables and outdoor pollutant concentrations. The associations between pollutant metrics and various health outcomes were analyzed using linear regression. Before the correlation and regression analyses, Box-Cox transformation was applied to some of the variables (TGs and FBG) to account for the skewness in the distribution; we also used a log transformation for the eNO measurement.

Logistic regression analyses were used to examine the relationship between categorical variables for the specified outcome (presence or absence of MetS and each risk factor) and concentration levels of air pollutant variables. Regression models were constructed separately for each pollutant of interest. The level of statistical significance was set at p < 0.05 for all tests. We used R (version 3.6.2) for statistical analyses and visualizations (version 4.3.1).

3. Results

3.1. Demographics

A total of 662 subjects were included in this study. Most of the participants were female (84.4%) and Hispanic (98.2%); the subjects had a mean age of 47.8 (±13.8) years with a range of 6-89 years old. The BMI was a mean of 30.6 (±6.6) kg/m², which ranged from 12.66 to 67.65 kg/m², and 81.1% of the participants were overweight (35.2%) or obese (45.9%), whereas 100 participants (15.1%) had a healthy BMI (

Table 1. Summary statistics of participants’ characteristics.

Variable	Min	Q1	Median	Mean	Q3	Max	SD	IQR
Age (years)	6	40	49	47.8	57	89	13.8	17
Weight (kg)	18.1	66.0	76.0	77.4	87.0	164.0	17.9	21.0
Height (cm)	115.0	154.0	158.0	159.0	163.5	185.0	8.2	9.5
BMI (kg/m2)	12.7	26.5	29.7	30.6	34.6	67.7	6.6	8.0

). Subject demographic information and health characteristics are shown in

Table 2. Demographic information of the subjects (N = 662).

Variable		Frequency	% *
Sex	Female	559	84.4
	Male	103	15.6
Education	Middle school	162	24.5
	Elementary school	148	22.4
	High school, no diploma	130	19.6
	High school graduate	86	13.0
	Some college, not completed	54	8.2
	Associate degree	26	3.9
	Bachelor’s degree	23	3.5
	Never attended or kindergarten only	14	2.1
	Master’s, doctoral, or professional degree	2	0.3
Language	Spanish	506	76.4
	Both	126	19.0
	English	21	3.2
Employed	Homemaker	211	31.9
	Employed—part time	146	22.1
	Employed—full time	81	12.2
	Not employed for more than 1 year	56	8.5
	Not employed for less than 1 year	51	7.7
	Self-employed	32	4.8
	Student	28	4.2
	Retired	24	3.6
	Unable to work	18	2.7
Income	USD 0–19,999	559	84.4
	USD 20,000–29,999	50	7.6
	USD 30,000–39,999	9	1.4
	USD 40,000–49,999	3	0.5
	USD 50,000–69,999	2	0.3
	USD 70,000–99,999	2	0.3
Marital Status	Married	232	35.0
	Never married	136	20.5
	Separated	105	15.9
	Divorced	88	13.3
	Widowed	50	7.6
	A member of an unmarried couple	28	4.2
	Civil union	14	2.1
Ethnicity	Hispanic	650	98.2
	Non-Hispanic	8	1.2
Race	White	600	90.6
	Black or African American	10	1.5
	American Indian or Alaska Native	3	0.5
	Asian	2	0.3
Health	Good	253	38.2
	Fair	236	35.6
	Poor	79	11.9
	Very good	50	7.6
	Excellent	21	3.2
Obesity	Obesity	304	45.9
	Overweight	233	35.2
	Healthy	100	15.1

* Distributions might not add up to 100% due to participants preferring not to answer a question.

.

3.2. Air Pollution Measurements

Hourly concentrations at the nearest CAMS locations to the subjects’ residential addresses were averaged over 24 h, 48 h, 72 h, and 96 h exposure window periods for comparisons. The Chamizal station had the highest frequency and was the nearest CAMS relative to the participants’ residential addresses (

Table 3. Spatial distribution of subjects to the nearest CAMS.

Nearest CAMS		Frequency	%
PM2.5	Chamizal	298	45.0
	Ascarate	136	20.5
	UTEP	121	18.3
	Socorro	107	16.2
PM10	Chamizal	391	59.1
	Socorro	147	22.2
	UTEP	124	18.7
NO2	Chamizal	296	44.7
	Ascarate	242	36.6
	UTEP	124	18.7
O3	Chamizal	194	29.3
	UTEP	115	17.4
	Skyline	111	16.8
	Ascarate	87	13.1
	Socorro	82	12.4
	Ivanhoe	73	11.0

). Not all pollutants are measured at all CAMS locations. The means were aggregated to represent prior pollutant exposure until 10 a.m., when health outcomes were measured.

Table 4. Summary statistics for pollutant measurements over various exposure windows.

Pollutant		Min	Q1	Median	Mean	Q3	Max	SD	IQR
PM2.5 (μg/m3)	24 h	1.7	5.0	7.8	8.9	11.1	30.9	5.6	6.1
	48 h	2.8	5.4	7.8	8.6	9.7	27.3	4.9	4.4
	72 h	3.2	5.9	7.9	8.6	9.5	24.1	4.0	3.6
	96 h	3.2	6.0	7.8	8.4	10.0	20.1	3.4	4.0
PM10 (μg/m3)	24 h	7.3	16.0	24.9	31.8	35.2	102.0	24.2	19.3
	48 h	6.2	17.1	24.4	31.3	37.9	84.7	20.0	20.8
	72 h	9.2	17.0	25.0	30.1	39.0	72.0	16.0	21.9
	96 h	8.2	18.7	25.6	29.4	39.4	65.0	13.6	20.7
NO2	24 h	0.7	8.8	12.7	15.0	21.4	34.0	8.2	12.6
(ppb)	48 h	2.4	9.1	13.0	14.6	19.5	31.1	7.0	10.4
	72 h	3.2	10.2	14.1	14.7	18.0	29.7	6.0	7.8
	96 h	5.0	11.0	13.8	14.8	18.8	29.7	5.2	7.8
O3	24 h	6.3	16.5	24.2	25.2	33.1	51.7	10.7	16.6
(ppb)	48 h	7.9	17.2	25.3	25.4	31.8	47.0	9.8	14.6
	72 h	8.9	17.5	24.7	25.1	32.4	48.6	9.5	14.9
	96 h	9.6	17.0	24.5	25.2	32.7	47.4	9.3	15.6

summarizes the descriptive statistics for the pollutant measurements across the study participants. PM_2.5 concentrations over the 24 h window ranged from 1.7 to 30.9 μg/m³, with a mean of 8.9 μg/m³, while PM₁₀ concentrations ranged from 7.3 to 102.0 μg/m³, with a mean of 31.8 μg/m³. The PM_2.5/PM₁₀ ratio was calculated and averaged 0.28 across the various time windows, indicating that coarse particles (PM₁₀) contributed substantially to the overall particulate matter levels in the region. For gaseous pollutants, 24 h averages showed that NO₂ concentrations ranged from 0.7 to 34.0 ppb, with a mean of 15.0 ppb, while O₃ concentrations ranged from 6.3 to 51.7 ppb, with a mean of 25.2 ppb.

3.3. Respiratory Associations

Descriptive statistics for exhaled nitric oxide (eNO) and spirometry measurements are summarized in

Table 5. Descriptive statistics for eNO, FEV₁, FVC, and PEF metrics.

Variable	Min	Q1	Median	Mean	Q3	Max	SD	IQR
eNO (ppb)	4.9	13.0	18.0	21.4	24.0	113.0	14.0	11.0
FEV1 (L)	0.8	2.0	2.3	2.4	2.7	4.9	0.6	0.7
FVC (L)	0.8	2.2	2.6	2.6	3.0	6.0	0.7	0.8
PEF (L/min)	1.6	4.2	5.1	5.3	6.2	11.5	1.7	2.0
FEV1/FVC	0.6	0.9	0.9	0.9	1.0	1.0	0.1	0.1

. The range for eNO was from 4.9 to 113, with a mean of 21.4 (±14) ppb. The FEV₁ ranged from 0.76 to 4.9 L, with a mean of 2.4 (±0.6) L; the FVC ranged from 0.8 to 6 L, with a mean of 2.6 (±0.7) L; and PEF ranged from 1.6 to 11.5 L/min, with a mean of 5.3 (±1.7) L/min.

Figure 2 presents pollutant effect estimates on respiratory outcomes using linear regression models and corresponding p-values. Additional factors are shown in Table S1. Regression analysis showed that short-term pollutant concentrations of PM_2.5 were negatively associated with spirometry measures such as FEV₁: β₁ = −0.011 for 24 h PM_2.5 (p = 0.038), β₁ = −0.014 for 48 h PM_2.5 (p = 0.018), and β₁ = −0.017 for 96 h PM_2.5 (p = 0.032).

PEF was also negatively correlated with PM_2.5 for all exposure time periods (β₁ = −0.048 for 24 h PM_2.5, β₁ = −0.058 for 48 h PM_2.5, β₁ = −0.054 for 72 h PM_2.5, β₁ = −0.068 for 96 h PM_2.5; p < 0.05). We found that the longer the exposure to PM_2.5 concentrations, the greater the decrease in lung function, represented by PEF. The 24, 48 and 96 h means for NO₂ had a negative association with PEF: β₁ = −0.023 for 24 h NO₂ (p = 0.013), β₁ = −0.028 for 48 h NO₂ (p = 0.011), and β₁ = −0.028 for 96 h NO₂ (p = 0.047). Furthermore, using generalized linear regression modeling, we observed a negative association between FEV₁/FVC and 96 h PM_2.5 (β₁ = −0.023, p = 0.040). The ratio was also negatively associated with 24 h NO₂ (β₁ = −0.011, p = 0.020) and 96 h NO₂ (β₁ = −0.019, p = 0.011). However, 24 h ozone data showed a positive correlation with the FEV₁/FVC value (β₁ = 0.008, p = 0.040), possibly due to a negative correlation between NO₂ and O₃.

3.4. Cardiovascular Associations

Descriptive statistics for cardiovascular measurements are presented in

Table 6. Descriptive statistics for cardiovascular syndrome risk factors.

Variable	Min	Q1	Median	Mean	Q3	Max	SD	IQR
BMI (kg/m2)	12.66	26.52	29.69	30.6	34.56	67.65	6.6	8.04
Waist (cm)	49	86	94	95.5	104	151	14.4	18
Systolic blood pressure (mmHg)	74	113	125	128	140.25	211	20.6	27.25
Diastolic blood pressure (mmHg)	35	69	75	76	82	128	11.4	13
Triglycerides (mg/dL)	44.9	107.25	161	186.0	224	650.1	114.7	116.75
HDL (mg/dL)	14.9	40	48	49.7	58	100.1	14.6	18
Fasting blood glucose (mg/dL)	49.9	86.25	94.5	108.7	108	477	46.5	21.75

. The mean BMI was 30.6 (±6.6) kg/m², and the mean waist circumference was 95.5 (±14.4) cm. Blood pressure, on average, was 128/76 (±21/11) mmHg, with a differential blood pressure of 52 (±14) mmHg. The lipid profile measures indicated mean TGs of 186 (±114.7) mg/dL, mean HDL of 49 (±14.6) mg/dL, and an FBG mean of 108.7 (±46.5) mg/dL.

Regression analyses showed that the continuous variables measuring metabolic syndrome risk factors (e.g., waist, HDL, and FBG) were associated with pollutant measurements (Figure 3). Additional factors are shown in Table S2. The relationship between waist and PM_2.5 may be due to a strong correlation observed between BMI and waist (correlation coefficient of 0.8). The 72 h PM_2.5 concentration was positively associated with BMI (β₁ = 0.132, p = 0.042).

We observed a significant relationship between 96 h mean ozone and HDL, showing a positive correlation with β₁ = 0.136 (p = 0.028). The increases in 24/48 h PM_2.5 and PM₁₀ were significantly associated with an increase in the Box–Cox-transformed fasting blood glucose (p < 0.05) but not for the original scale of fasting blood glucose. The transformation of fasting blood glucose was suitable for finding linear relationships with air pollution measurements.

The classification of MetS risk factors (binary outcomes) based on current guidelines is presented in Table S3. Associations between MetS factors and pollutant metrics are summarized in Figure 4, which shows effect estimates using logistic regression models and the corresponding p-values, odds ratios, and 95% confidence intervals of the odds ratios. In logistic regression modeling, we also found that increasing PM_2.5 and NO₂ concentrations were associated with an increased likelihood of having a large waist (p < 0.05 for 48, 72, and 96 h PM_2.5; p < 0.01 for all windows of NO₂ concentration). However, the odds of having a large waist decreased as the ozone level increased (p < 0.05 for 24/48/72/96 h O₃). The ozone increase was also associated with a lower likelihood of having low HDL (p < 0.05 for 24/48/72/96 h O₃), and more exposures to ozone led to a lower odds ratio for having low HDL (0.983 for 24 h O₃, 0.980 for 48 and 72 h O₃, and 0.976 for 96 h O₃).

A higher likelihood of having high fasting glucose was associated with increased PM concentrations, namely 1.034 and 1.037 times higher odds ratios with a 1 µg/m³ increase in 24 and 48 h PM_2.5, respectively (p < 0.05), which is considerable given our PM_2.5 range from 5 to 31 ug/m³. A 1 µg/m³ change in both 24 and 48 h PM₁₀ resulted in a 1.008 times higher odds ratio for high glucose (p < 0.05). The 72 h NO₂ concentration was also a significant factor in predicting high glucose status, showing an increased likelihood of having high glucose as NO₂ increases (odds ratio = 1.027; p = 0.048).

Metabolic syndrome showed significant associations with PM_2.5, NO₂, and O₃. The odds of having MetS were 1.051 times higher with a 1 µg/m³ increase in 96 h PM_2.5 (p = 0.043). The associations of MetS classification with NO₂ concentrations were also positive, showing increased odds ratios of 1.027 (p = 0.021), 1.040 (p = 0.004), and 1.056 (p = 0.001) for 48, 72, and 96 h exposures, respectively. However, the increased ozone was correlated with a decreased likelihood of having MetS (p < 0.05 for all exposure times). Additional associations are shown in Table S4.

4. Discussion

4.1. Principal Findings

The pollutant levels observed in this study highlight significant public health concerns, particularly for low-income residents living near high-traffic areas. These levels frequently exceeded the latest WHO Air Quality Guidelines (2021), which are more stringent than the current U.S. EPA standards. For instance, the mean PM_2.5 concentration recorded during the study period was 8.9 µg/m³, but it sometimes surpassed the WHO’s 24 h guideline of 15 µg/m³ and occasionally exceeded the U.S. EPA standard of 35 µg/m³ for the same period. Similarly, NO₂ concentrations exceeded the WHO’s annual guideline of 10 µg/m³ despite compliance with the less stringent EPA standard of 53 µg/m³.

The PM_2.5/PM₁₀ ratio provides valuable insights into the composition and sources of particulate matter pollution. Based on the averages presented in Table 4, the PM_2.5/PM₁₀ ratio for the study region was consistently below 0.3, indicating that coarse particles (PM₁₀) constitute the dominant fraction of particulate matter. This pattern suggests contributions from natural sources such as windblown dust, soil re-suspension, occasional sandstorm events, and anthropogenic activities like road dust and construction. In contrast, higher ratios are typically associated with finer particles (PM_2.5), primarily emitted from combustion-related anthropogenic sources, such as vehicular emissions and power generation [36].

4.2. Cardiorespiratory Associations

Our study examined the short-term associations (24/48/72/96 h means) of traffic-related air pollutants (PM_2.5, PM₁₀, NO₂, and O₃) with biomarkers of respiratory and cardiovascular disease in a group of participants from low-income communities in El Paso, TX. We found associations between short-term air pollutant concentrations and altered lung function. Furthermore, we found associations with BMI and MetS risk factors such as waist circumference and fasting glucose.

FEV₁ was negatively correlated with mean concentration levels of PM_2.5 (24/48/96 h), indicating a relationship between lung function and ambient PM_2.5 before the measurement. Specifically, this respiratory indicator represents an increased risk for obstructive respiratory diseases (asthma, chronic obstructive pulmonary disease [COPD]). Furthermore, PEF, which is also an indicator of increased risk for asthma and COPD, was negatively correlated not only with PM_2.5 but also with NO₂. However, we did not see an influence of PM₁₀, which might suggest significant health effects were associated with smaller particles that affect the lower respiratory tract and can further cause obstructive respiratory diseases [37].

The exhaled nitric oxide (eNO) results indicated current airway inflammation but did not correlate with our population’s studied air pollutant concentration levels. Even so, measuring eNO is clinically helpful in treating and controlling asthma and may benefit the studied population [38]. Given that our inclusion methods did not ask if a participant had a history of asthma, we recommend that future studies consider the relationships of eNO with air pollution in participants with asthma. We initially hypothesized that obstructive respiratory diseases (e.g., asthma) are more prevalent in our population than restrictive respiratory diseases (e.g., sarcoidosis, lung fibrosis). Our results support this since we determined associations with the FEV₁/FVC ratio, which can differentiate obstructive from restrictive respiratory diseases, as demonstrated by a ratio of 0.7. This ratio indicates lung obstruction and supports the negative correlations between PM_2.5 and NO₂ for different exposure periods.

Although short-term associations with risk factors related to obesity (waist circumference) in linear and logistic models were not expected as part of this study, the relationship was present across most exposure periods. We do not expect a causal effect between short-term exposure to air pollution and obesity. However, the decreasing ranges and similar means of short-term air pollution concentrations could indicate that similar trends can probably be found with medium- or long-term exposure levels. This finding could also be representative of the environmental conditions and neighborhoods where participants live, aligning with a previous study focusing on NO₂, which had similar trends but considered increasing windows of exposure time when comparing short-term and long-term effects [39].

Even though we did not find associations with other metabolic outcomes such as blood pressure or lipid profiles, we did find associations with fasting blood glucose in linear models and increased odds among those with high fasting glucose levels and metabolic syndrome. Possible explanations for these associations include the effects of oxidative stress and inflammation caused by air pollution exposure [40,41,42,43]. Additionally, the role of oxidative and inflammatory potential in particulate matter (PM) has been identified as a critical factor contributing to cardiometabolic derangements [44]. Furthermore, recent findings [45] demonstrate that air pollution mortality rates are significantly influenced by the intersection of race and social class, with marginalized communities experiencing higher exposure levels and associated risks. This aligns with our focus on low-income populations residing in proximity to high-traffic areas and underscores the need for equity-driven policy interventions to mitigate health disparities.

Lastly, this study analyzed low-income residents disproportionately impacted by environmental exposures due to their proximity to pollution sources and limited access to healthcare. While this focus addresses a critical gap in public health research, the lack of data on middle- and high-income residents from the same area limits our ability to examine differential impacts across income groups.

4.3. Comparison with Other Studies

Respiratory outcomes have been associated with air pollution exposure in other epidemiological studies. The Framingham Study found that moderate exposure measured by the EPA’s Air Quality Index for PM_2.5, NO₂, and O₃ was associated with lower FEV₁ considering 24 and 48 h pollutant concentration means before the measurement [46]. A study among 1694 female non-smokers from the Wuhan–Zhuhai region in China found that in a city with high pollution levels, the moving means of PM_2.5, PM₁₀, NO₂, and O₃ exposures were significantly associated with FEV₁ reductions, but also in a city with low-level air pollution, PM₁₀, O₃, and PM_2.5 were significantly associated with reduced FEV₁ [47]. The same study also found associations with FVC; however, we did not find them in our study, which might be due to the relatively lower levels of exposure, as observed in the available CAMS data.

Furthermore, a repeated measures study from Belgium found that an increase in PM₁₀ on the day of the clinical examination was associated with lower FVC, FEV₁, and PEF. Also, an increase in NO₂ was associated with a reduction in PEF on the examination day [24]. In addition, a study of lung function in adults exposed to very low levels of ambient air pollution in Europe did not observe an association between air pollution and longitudinal change in lung function [48]. However, it was observed that an increase in NO₂ exposure was associated with lower levels of FEV₁ and FVC. Moreover, an increase in PM₁₀, but not other PM metrics (PM_2.5, coarse fraction of PM, PM absorbance), was associated with a lower level of FEV₁. The associations were significant in people with obesity.

Regarding metabolic outcomes, Chuang and collaborators [25] observed that increased PM₁₀ was marginally (p < 0.10) associated with elevated systolic blood pressure (24 h) and triglycerides (24 to 120 h) and significantly associated with hemoglobin A1C (72 h) and reduced HDL (24 h) (p < 0.05). They also reported that ozone was associated with diastolic blood pressure (72 and 120 h) and hemoglobin A1C (24, 72, 120 h) and marginally associated with triglycerides and fasting glucose [25]. Unfortunately, their study did not consider PM_2.5 measurements, which showed some associations in our study.

A study conducted in China showed a positive correlation between air pollution (PM₁₀, NO₂, and O₃) and BMI [49], which agrees well with the associations from our study. However, they considered the time window based on long-term exposures using average concentrations within three years instead of short-term exposure. Furthermore, a review that considered 14 short-term effect studies of air pollutants suggested that the consistent pattern of associations among participants with obesity suggests that obesity may negatively modify the impact of PM_2.5 on cardiovascular health [50].

4.4. Strengths and Limitations

The present study considered ambient (outdoor) air pollution measured at nearby CAMSs. However, there could be some variation in the participants’ indoor environments related to pollutant exposure. Although research indicates a direct relationship between ambient and indoor air pollution [51,52,53], a participant’s actual exposure concentration can differ from the surrogate concentration measured at the CAMS.

This study primarily attributes the observed air pollutant levels to traffic-related emissions due to the residential proximity of the participants to high-traffic roadways and the placement of CAMSs near these roadways. For example, the Chamizal and Ascarate monitoring stations are adjacent to heavily trafficked highways and urban centers, representing vehicular emission sources. However, it is important to acknowledge that ambient PM₁₀, PM_2.5, NO₂, and O₃ concentrations may also arise from non-traffic sources, such as industrial activities, construction, and natural processes like windblown dust. While these contributions are likely secondary in the studied region, they may introduce variability in pollutant levels and limit the specificity of traffic-related attributes.

Additionally, while the CAMS data provide robust regional coverage, the reliance on these fixed monitoring stations introduced certain limitations. In some cases, CAMSs were located far from participants’ residential areas, which excluded some individuals from the analysis. Furthermore, the geographic and temporal resolution of the CAMS data may not fully capture localized exposure variations, such as those influenced by the immediate traffic density or microclimates. This limitation underscores the need for future studies to integrate more granular data, such as portable air quality monitors or personal exposure devices, and to consider source apportionment analyses better to characterize the relative contributions of various pollution sources.

Another limitation lies in the availability of pollutant measurements at the CAMSs. For instance, ozone was measured more frequently across the six stations than other traffic-related pollutants, which could introduce measurement bias. Nonetheless, we observed consistent trends across pollutants, suggesting the robustness of our findings despite these discrepancies.

While this study focuses on the overall impact of air pollution on low-income populations, the potential influences of job nature and other sociodemographic factors could not be examined due to the design and data limitations of the larger project. Furthermore, this study primarily attributed PM₁₀, PM_2.5, NO₂, and O₃ levels to mobile sources; it is vital to recognize that other contributors, such as stationary sources (e.g., industrial facilities) and natural processes (e.g., windblown dust), may also influence ambient air pollution levels.

Another significant limitation is the absence of data on participants’ smoking status. This limitation is particularly relevant for respiratory outcomes such as FEV1 and PEF, which can be significantly influenced by smoking history. While this study’s primary focus was on the associations between traffic-related air pollution and health outcomes, future research should collect and incorporate smoking status to provide a more comprehensive understanding of these relationships.

This study’s cross-sectional design also limits the ability to establish causal relationships between air pollution and health outcomes. While significant associations were identified, longitudinal studies are needed to confirm these effects’ temporality and explore cumulative exposures over time.

4.5. Policy Implications

Since these low-income populations are often situated near major roadways, industrial areas, and other pollution sources, urban planning policies can significantly reduce this exposure. Measures such as enforcing zoning laws to maintain buffer zones between residential areas and high-traffic corridors and limiting industrial development in vulnerable neighborhoods are essential. Furthermore, integrating green infrastructure, including urban vegetation and natural barriers, can reduce pollutant concentrations near roadways and improve local air quality, benefiting public health.

The predominance of coarse particles in the study region has important implications for health outcomes. As observed in this study, PM₁₀ tends to deposit in the upper respiratory tract, contributing to respiratory symptoms such as airway inflammation and reduced lung function. Conversely, PM_2.5 penetrates deeper into the respiratory system, reaching alveoli and potentially entering the bloodstream, which may exacerbate systemic inflammation and cardiovascular conditions. While the PM_2.5/PM₁₀ ratio indicates that natural sources may play a substantial role in particulate matter pollution in the region, it does not diminish the contributions of traffic-related emissions. Vehicular activity remains a significant source of fine and coarse particles, particularly in urban areas where heavy traffic is common. Future studies could benefit from incorporating source apportionment analyses to more precisely quantify the relative contributions of natural and anthropogenic sources to PM pollution in this region. Such analyses would provide a more complete understanding of the complex interplay between pollutant sources and health outcomes, allowing for more targeted public health interventions.

Improving air quality monitoring systems is another critical step. The deployment of additional air monitoring stations, particularly in underserved and high-pollution areas, would enable more precise tracking of exposure levels and identification of pollution hotspots. Advances in mobile and wearable air monitoring technologies could complement fixed stations, providing detailed, localized data on individual exposure patterns.

Community-based interventions are equally important in addressing environmental health disparities. Public health campaigns to educate residents about strategies to reduce exposure, such as improving indoor air quality through filtration and ventilation, can empower individuals to take proactive steps. Resources supporting local initiatives, such as neighborhood-level environmental action plans, can foster community engagement and promote sustainable practices.

Finally, the intersection of socioeconomic inequities and environmental exposures underscores the need for policies that address the broader social determinants of health. Improving access to healthcare for vulnerable populations is critical for the early detection and management of air pollution-related illnesses. Additionally, funding research that examines the cumulative effects of TRAP and other stressors in low-income communities can inform more comprehensive and equitable policy solutions.

5. Conclusions

The present study correlated short-term exposure to traffic-related pollutants with respiratory function outcomes related to pulmonary obstruction. Future studies should consider clinical classifications of obstructive respiratory outcomes such as COPD and asthma while considering the effects on FEV₁ and PEF. Notably, our study may be among the first to identify associations between short-term exposure to air pollutants and obesity. While we do not expect this to represent a direct causal relationship, the findings highlight the complex interplay between environmental exposures and metabolic health. The observed associations could reflect cumulative or synergistic effects mediated by socioeconomic and neighborhood characteristics. Future research should explore obesity as a potential outcome of air pollution exposure by employing extended exposure windows (e.g., 7-day, 30-day, or seasonal averages) to assess both medium- and long-term impacts.

Furthermore, statistical models in future studies could be enriched by incorporating geographic parameters such as distance to pollution sources (e.g., highways, industrial zones), street network density, and traffic volumes in residential areas. These variables could provide a more nuanced understanding of how air pollution interacts with the built environment and socioeconomic factors to influence health outcomes. Additionally, incorporating land-use regression or dispersion models could offer enhanced spatial precision in exposure assessments.

Finally, given the implications for public health, future research should evaluate the effectiveness of mitigation strategies, such as urban green spaces, traffic flow alterations, or improved ventilation systems, in reducing the health burden associated with air pollution. Such studies would provide actionable insights for policymakers and urban planners aiming to minimize the adverse effects of environmental exposures on respiratory and metabolic health.