Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Feb 1.
Published in final edited form as: Environ Res. 2015 Nov 21;145:9–17. doi: 10.1016/j.envres.2015.10.026

Association between Satellite-based Estimates of Long-term PM2.5 Exposure and Coronary Artery Disease

Laura A McGuinn 1, Cavin K Ward-Caviness 2, Lucas M Neas 3, Alexandra Schneider 2, David Diaz-Sanchez 3, Wayne E Cascio 3, William E Kraus 4, Elizabeth Hauser 4, Elaine Dowdy 4, Carol Haynes 4, Alexandra Chudnovsky 5,6, Petros Koutrakis 5, Robert B Devlin 3
PMCID: PMC4706491  NIHMSID: NIHMS741259  PMID: 26613345

Abstract

Background

Epidemiological studies have identified associations between long-term PM2.5 exposure and cardiovascular events, though most have relied on concentrations from central-site air quality monitors.

Methods

We utilized a cohort of 5,679 patients who had undergone cardiac catheterization at Duke University between 2002-2009 and resided in North Carolina. We used estimates of daily PM2.5 concentrations for North Carolina during the study period based on satellite derived Aerosol Optical Depth (AOD) measurements and PM2.5 concentrations from ground monitors, which were spatially resolved with a 10 × 10 km resolution, matched to each patient's residential address and averaged for the year prior to catheterization. The Coronary Artery Disease (CAD) index was used to measure severity of CAD; scores >23 represent a hemodynamically significant coronary artery lesion in at least one major coronary vessel. Logistic regression modeled odds of having CAD or an MI with each 1 μg/m3 increase in annual average PM2.5, adjusting for sex, race, smoking status and socioeconomic status.

Results

In adjusted models, a 1μg/m3 increase in annual average PM2.5 was associated with an 11.1% relative increase in the odds of significant CAD (95% CI: 4.0%-18.6%) and a 14.2% increase in the odds of having a myocardial infarction (MI) within a year prior (95% CI: 3.7% -25.8%).

Conclusions

Satellite-based estimates of long-term PM2.5 exposure were associated with both coronary artery disease (CAD) and incidence of myocardial infarction (MI) in a cohort of cardiac catheterization patients.

Keywords: coronary disease, epidemiology, myocardial infarction, PM2.5, spatiotemporal analyses

1. Introduction

Air pollution is associated with several adverse health outcomes. Specifically, ambient fine particulate matter ≤2.5 μg/m3 (PM2.5) is associated with increased mortality and increased risk for respiratory and cardiovascular disease (Brook et al., 2010; Ruckerl et al., 2011). The Global Burden of Disease Study 2010 estimates that worldwide over 3.2 million premature deaths and over 74 million years of healthy life lost were attributable to ambient particulate matter pollution, making it one of the top global health risk factors (Lim et al., 2012). Further, an estimated 22% of disability-adjusted life-years for heart disease are attributable to ambient particulate matter pollution. Recently, a World Health Organization report estimated that air pollution exposure contributed to about 6% (3.7 million) of all deaths in 2012, with 40% of those coming from coronary artery disease (CAD) (World Health Organization, 2014). Populations particularly susceptible to the health effects of air pollution include infants and children, older adults, and those with underlying disease, particularly diabetes and cardiovascular disease (Pope, 2014).

A great deal of research has focused on the short-term variation in air pollution exposure, with fewer studies focusing on long-term PM2.5 exposure. More than 100 time-series and case-crossover analyses have demonstrated associations with short term PM2.5 exposure with myocardial infarctions (MIs) and increased risk for hospitalizations and mortality (EPA., 2009). Cohort studies have also identified associations between long-term PM2.5 exposure and MIs (Cesaroni et al., 2014), mortality, increased susceptibility to progression of disease (EPA., 2009; Hoek et al., 2013), and overall reduced life expectancy (Correia et al., 2013). Recent studies have assessed associations between PM2.5 and biomarkers of atherosclerosis (Rivera et al., 2013; Sun et al., 2013). One limitation of these studies is a reliance on exposure estimates based on single, central site monitors within a community.

The CATHeterization GENetics (CATHGEN) cohort described in this manuscript is a large, sequential cohort with participants selected from patients primarily from North Carolina presenting to the Duke University Medical Center Cardiac Catheterization Clinic between 2001 and 2011. In addition to clinical information collected during the catheterization visit and subsequent visits for follow-up treatment or physical examinations, an extensive set of biological information was collected in this cohort (Kraus et al., 2015). This includes genetic, and epigenetic analysis, RNA transcriptomic and micro-RNA analysis, frozen blood samples that are available for extensive screening of soluble vascular biomarkers including metabolomics, and electrocardiographic data for defining cardiac electrophysiology. The large cohort size and rich biologic database make it feasible to associate long-term exposure to PM2.5 with biomarkers that may define pathways by which air pollution contributes to the progression of CAD.

New approaches to assessing daily PM2.5 values use a fusion of satellite-derived aerosol optical density (AOD) data and ground monitoring stations measurements (Chudnovsky et al., 2012; Lee et al., 2012; Lee et al., 2011). In this paper the model was used to estimate daily PM2.5 values throughout NC from 2002-2009, in 10 × 10 km grids. Using this approach, we investigate the association between long-term PM2.5 exposure and CAD severity and recent MIs using a cardiovascular cohort enriched with individuals having coronary artery disease. Because the population at hand is high risk our findings will help identify subgroups that are particularly susceptible to the health effects of air pollution.

2. Methods

2.1 Study population

The CATHGEN cohort is a group of 9334 patients who underwent cardiac catheterization at Duke University from 2001-2011. Participants came primarily from North Carolina and underwent a cardiac catheterization and coronary angiography in order to diagnose and treat coronary artery disease. Participants underwent catheterization procedures on referral from their referring physician and/or were admitted with an appropriate condition, such as a recent MI. Therefore, all CATHGEN participants had a clinical indication for a cardiac catheterization. Clinical information was obtained from an intake questionnaire at the time of catheterization as well as medical records. All subjects received and signed informed consent forms prior to enrollment, and CATHGEN has been approved by and follows all Duke University Institutional Review Board policies. Ground PM2.5 concentration and satellite AOD data were obtained from 2002-2009 and a yearly average was created for each participant using the 365 days of exposure prior to each participant's catheterization date. Therefore, patients were included in the current analysis only if they resided in North Carolina and their catheterization procedure was performed after January 1, 2003 and before December 31, 2009.

Residential addresses were obtained from medical records and geocoded by the Children's Environmental Health Initiative (CEHI) for 8017 (86%) of the 9334 study participants. Of these 8017 individuals, 7118 (76%) resided in North Carolina, and 5679 (61%) had a catheterization that occurred between 2003 and 2009 (thus having complete exposure data available for the year prior to catheterization). Figure 1 shows the residential locations of these individuals. Participants resided in 91 of the 100 NC counties, with Durham County containing the highest percentage (21%).

Figure 1.

Figure 1

Open in a new tab

Residential location of CATHGEN participants.

2.2 Outcome classification

The Coronary Artery Disease (CAD) index was used to measure severity of coronary artery disease. The index ranges from 0 to 100 and is a risk indicator of events due to coronary atherosclerosis. The CAD index takes into consideration the number of diseased coronary vessels (0-3), left anterior descending CAD, number of coronary vessels with 95% occlusion, 75% and 95% proximal left anterior descending coronary artery stenosis, and 75% and 95% left main coronary artery stenosis (Bart et al., 1997).

A binary measure of CAD was constructed and individuals with a CAD index >23 represent a population with at least one hemodynamically significant lesion (>75% luminal stenosis) in one epicardial coronary artery. A CAD index score was not given to 610 study participants because they underwent a therapeutic intervention that precluded a full catheterization (not all coronary arteries were imaged). Therefore, 5069 study participants had complete PM2.5 and CAD outcome data available. We additionally assessed whether participants experienced an MI within a year prior to their catheterization. Participants were considered to be cases for the MI analyses if they had a documented MI in their medical records within a year prior to their catheterization visit. There were 5,679 study participants who had complete exposure and MI outcome data available.

2.3 Exposure assessment

Daily PM2.5 concentrations were predicted at a 10 × 10 km spatial resolution for the state of North Carolina from 2002-2009 using recently developed statistical prediction models (Chudnovsky et al., 2012; Lee et al., 2012; Lee et al., 2011). These models estimate exposure using two main stages. First, satellite-based AOD data that reflect particle abundance in the atmospheric column were used to predict ground-level PM2.5 concentrations for days when satellite data were available (Chudnovsky et al., 2012; Lee et al., 2011). A daily calibration approach using a mixed effects model was then applied to control for the inherent day-to-day variability in the AOD-PM2.5 relationship that depends on a number of time-varying parameters such as particle optical properties, vertical and diurnal concentration profiles and ground surface reflectance. Second, a cluster analysis approach was applied that uses AOD and PM2.5 ground monitoring data to predict PM2.5 concentrations on days when satellite data are not available due to the presence of clouds or snow (Lee et al., 2012). Cluster analysis was applied to every third day data and because cluster assignment was done every third-day, the same cluster classification was applied to the adjacent two days, assuming an identical spatial pattern for three consecutive days. In addition, cluster analysis was performed on PM2.5 concentration differences obtained by subtracting the daily regional PM2.5 concentrations from the respective PM2.5 concentrations. Finally, a cluster-specific PM2.5 prediction model using a generalized additive model (GAM) was developed. For each cluster PM2.5 concentration, values were regressed on the regional PM2.5 levels and the spatial smooth function of latitude and longitude. For both stages model performance was examined by comparing the measured and predicted PM2.5 concentrations using R2 and % mean relative error (MRE) estimates. More detailed information on the prediction model has been described previously (Chudnovsky et al., 2012; Lee et al., 2012; Lee et al., 2011). Satellite PM2.5 exposures were estimated for 1233 10 × 10 km grid cells, and data from 52 EPA air quality monitors in NC were used.

Patients' geocoded addresses were matched to the centroid of the nearest 10 × 10 km grid location based on spatial location and date. The average number of participants assigned to each 10 × 10 km grid cell was 9 (Min=1, Max=423). Study participants provided detailed address history upon study enrollment and addresses for the year prior to the catheterization date were used. PM2.5 predictions were averaged for the year prior to each patient's catheterization date. There were 15 individuals who moved during the year prior to their catheterization visit. For these individuals a weighted average was computed using exposure data from both residential locations. Some participants underwent multiple catheterization events during the study period. For those individuals the most recent catheterization visit was linked with exposure data. We chose the most recent catheterization visit because we had better residential address data for this visit. Figure 2 shows the average of each of the eight annual averages for each 10 × 10 km grid throughout the state.

Figure 2.

Figure 2

Open in a new tab

Annual average PM2.5 (μg/m3) predictions averaged from 2002 to 2009. Dots represent EPA PM2.5 monitoring stations.

2.4 Confounders and effect measure modifiers

We made use of a directed acyclic graph (DAG) in order to create a least biased estimate (Greenland et al., 1999). Covariates of interest include: age, sex, race/ethnicity (non-Hispanic White, African American, and other race/ethnicity), body mass index (BMI), smoking status, median home value, area level attained education, urban/ rural status, county of residence, history of hypertension, and history of diabetes. History of diabetes was defined as a previous physician diagnosis of either type I or type II diabetes. History of hypertension was ascertained during the health and physical exam administered prior to the catheterization procedure by a physician or physician's assistant.

Data from the 2000 U.S. Census was used to characterize each participant's socioeconomic status (SES) and urban/rural status. Participants were assigned to block groups based on their address at catheterization visit. Block-group level educational attainment was defined as the percentage of males and females in the block group without a high school education. Educational attainment was dichotomized for interaction analyses with ≥25% representing low attainment and <25% representing high educational attainment. Block group-level median home value was considered as another potential measure of socioeconomic status. Finally, participants were defined as living in an urban block group if the block group was ≥50% urban.

2.5 Statistical analyses

Logistic regression analysis was used to estimate odds ratios (OR) and 95% confidence intervals (CI) associated with a CAD index >23 or MI for each 1 μg/m3 increase in annual average PM2.5. Models were adjusted for sex, race/ethnicity, smoking status, and area level attained education.

We examined whether associations with PM2.5 were modified by select covariates by including an interaction term between continuous levels of PM2.5 and the potential modifiers of interest. We compared these models with the main effects model without interaction terms. The Likelihood Ratio Test (LRT) was used to assess potential effect modification and a cutoff of p<0.05 was used for significance and p <0.10 was used to indicate potential presence of modification. Modifiers of interest included: age, sex, race, smoking history, BMI, history of diabetes or hypertension, education rank, and urban/rural status.

We performed sensitivity analyses to address potential sources of bias by restricting analyses to those individuals with an annual PM2.5 average below the current EPA annual PM2.5 National Ambient Air Quality Standard (NAAQS) of 12.0 μg/m3 to determine if effects could be seen at levels below the current standard. Additional sensitivity analyses were performed by comparing the results when choosing the first vs. most recent catheterization visit for the 174 individuals who had multiple catheterization visits. We also performed sensitivity analyses by including county of residence, BMI, age at time of catheterization, and median home value in the adjusted models. Finally, additional sensitivity analyses were conducted for the 610 individuals with only a catheterization on select arteries (were not assigned a CAD index score) and examined the PM2.5-MI associations with this subset of participants who had a clinical indication for an immediate intra-coronary artery intervention at the time of catheterization. Statistical analyses were performed using SAS version 9.3 (Cary, NC).

3. Results

The study population was comprised of 5679 participants in total, and 5069 patients who underwent a cardiac catheterization. Study participants were aged 20-93 (mean age 60.8 ± 12.1 years). Catheterization visits were evenly spaced from 2002 to 2009. Table 1 shows the distribution of the study population characteristics. Those patients who were assigned a CAD score >23 represent 44% of the study population and those who had an MI in the year preceding the catheterization visit represent 12% of the population. The majority of the study population was male (61%), non-Hispanic white (73%), and had a history of hypertension (68%). More than 77% of the study participants were either overweight or obese. Demographic and clinical characteristics were similar within the CAD and MI outcome groups.

Table 1. Characteristics of the CATHGEN study population.

PM2.5 (μg/m3)a Total cohort CADb MI in prior yearc
Total 12.4 5,679 2,491 704
Age at time of enrollment in years (mean ± SD) 60.8 ± 12.1 63.3 ± 11.2 59.8 ± 12.4
Gender
 Male 12.4 3,471 (61) 1,798 (72) 467 (66)
 Female 12.4 2,208 (39) 693 (28) 237 (34)
Body mass index (kg/m2)
 <18.5 (Underweight) 12.5 80 (1) 38 (2) 14 (2)
 18.5-24.9 (Normal weight) 12.4 1,187 (21) 485 (20) 151 (21)
 25.0-29.9 (Overweight) 12.4 1,987 (35) 967 (39) 258 (37)
 ≥30.0 (Obese) 12.3 2,399 (42) 991 (40) 281 (40)
 Missing 26 10 0
Raced
 Non-Hispanic white 12.3 4,146 (73) 1,940 (78) 489 (69)
 African American 12.5 1,204 (21) 409 (16) 161 (23)
 Other 12.5 329 (6) 142 (6) 54 (8)
History of smoking
 Yes 12.4 2,664 (47) 1,364 (55) 398 (57)
 No 12.4 3,015 (53) 1,127 (45) 306 (43)
History of diabetes
 Yes 12.4 1,660 (29) 846 (34) 212 (30)
 No 12.4 4,019 (71) 1645 (66) 492 (70)
History of hypertension
 Yes 12.4 3,882 (68) 1,894 (76) 510 (72)
 No 12.4 1,797 (32) 597 (24) 194 (28)
Urban/ rural statuse
 Urban 12.5 3,110 (55) 1,299 (52) 360 (51)
 Rural 12.2 2,569 (45) 1,192 (48) 344 (49)
Educational attainmentf
 Low 12.4 2,290 (40) 1,032 (41) 338 (48)
 High 12.4 3,389 (60) 1,459 (59) 366 (52)
Median home value ($)
 <82,700 12.5 1,400 (25) 675 (27) 248 (35)
 82,700-118,000 12.4 1,403 (25) 613 (25) 166 (24)
 118,000-166,500 12.3 1,436 (25) 638 (26) 159 (23)
 ≥166,500 12.3 1,414 (25) 561 (23) 127 (18)
 Missing 26 4 4
Open in a new tab

Numbers in parentheses are the percent of each group (Total Cohort, CAD, MI in prior year) represented by each characteristic.

a

Annual average PM2.5 level (μg/m3) is based on the average of each participants' PM2.5 exposure for the year prior to their catheterization visit.

b

Binary measure of CAD index (>23 CAD index). The total sample size for the CAD outcome is 5,069.

c

MI in prior year includes those that experienced a MI in the year prior to their catheterization visit.

d

Other race/ethnicity includes Native American, Hispanic, Asian, and unknown.

e

Urban status was defined as living in a block group that was ≥50% urban.

f

Low educational attainment includes those who live in block groups where ≥25% of males and females have less than a high school education.

The mean annual average PM2.5 level for the study population was 12.4 μg/m3 (Table 1), which is slightly above the NAAQS annual PM2.5 limit of 12.0 μg/m3. African American and other race/ethnicity, those living in urban areas, and those residing in low home value areas had slightly higher PM2.5 concentrations. PM2.5 concentrations did not vary significantly among the two outcome groups (Supplementary Table 1). The interquartile range (IQR) of the annual average PM2.5 concentrations for the CATHGEN study population was 0.78 μg/m3 (Min: 8.6 μg/m3, Max: 14.4 μg/m3). Annual PM2.5 averages in NC decreased over time from 12.5 μg/m3 in 2002 to 9.0 μg/m3 in 2009.

We used the CAD index as a measure of coronary atherosclerosis and MIs within the preceding year as effect outcomes. Table 2 presents the adjusted associations between each 1 μg/m3 increase in annual average PM2.5 exposure for these outcomes. We observed an 11% relative increase in odds of having a CAD index >23 (95% CI: 1.04, 1.19) for each 1 μg/m3 increase in annual average PM2.5 compared with a CAD index <23. Associations between PM2.5 and CAD remained when choosing the first catheterization visit instead of the most recent visit for the 174 individuals who had multiple catheterization visits (OR: 1.12, 95% CI: 1.05, 1.20). For each 1 μg/m3 increase in annual average PM2.5 the odds of an MI occurring within the preceding year increased by 14% (95% CI: 1.04, 1.26).

Table 2.

Association between 1 μg/m3 increase in annual average PM2.5a exposure and measures of CAD and MI for the total cohort and those with an annual average PM2.5 exposure of <12.0 μg/m3

Outcome Total cohort <12.0 μg/m3
N Events Odds Ratiob 95% CI N Events Odds Ratiob 95% CI
CAD Index 5,069 2,491 1.11 1.04, 1.19 1,079 501 1.15 1.01, 1.31
MI in prior year 5,679 704 1.14 1.04, 1.26 1,275 124 1.65 1.29, 2.11
Open in a new tab
a

Interquartile range of annual average PM2.5: 0.78 μg/m3.

b

Models are adjusted for race, area level education, gender, and smoking status.

The majority (78%) of the study population resided in an area that had an annual average PM2.5 level slightly above the current NAAQS of 12.0 μg/m3, with 1275 participants residing in an area with an annual average level below the limit. To determine whether effects could be seen at levels below the current EPA annual PM2.5 standard, the analysis was restricted to these 1275 individuals. As seen in Table 2 the association between PM2.5 and CAD index, and PM2.5 and MI in the prior year, was significant at levels below the current NAAQS.

We conducted sensitivity analyses to assess potential sources of bias in the adjusted models. We assessed whether or not adjusting for county of residence, age at time of catheterization, BMI, median home value, or distance to Duke University altered any of the associations with each outcome. As seen in Table 3, the associations between long-term PM2.5 exposure and outcomes did not change significantly after adjustment for these covariates.

Table 3. Sensitivity analyses for the addition of select covariatesa.

CAD Index MI in prior year
Crude Model 1.07 (1.00, 1.14) 1.14 (1.03, 1.25)
Model 0 (Gender, smoking, race) 1.10 (1.03, 1.18) 1.13 (1.02, 1.24)
Model 1 (Added educational attainment) 1.11 (1.04, 1.19) 1.14 (1.04, 1.26)
Model 2 (Added county of residence) 1.11 (1.03, 1.19) 1.17 (1.05, 1.30)
Model 3 (Added age) 1.12 (1.05, 1.20) 1.14 (1.04, 1.26)
Model 4 (Added BMI) 1.11 (1.04, 1.18) 1.14 (1.03, 1.25)
Model 5 (Added median home value) 1.11 (1.04, 1.18) 1.12 (1.02, 1.24)
Model 6 (Added distance to Duke) 1.11 (1.04, 1.19) 1.14 (1.04, 1.26)
Open in a new tab
a

Total N for the CAD outcome is 5069, and 5679 for the MI outcome.

There were 610 individuals who did not receive a CAD score to quantify the severity of atherosclerotic lesions because their CAD was found to be significant enough during catheterization to require immediate medical intervention. These individuals were not included in the CAD index analysis, but were included in the MI analysis. Because they had more severe CAD than the cohort as a whole, we performed a sensitivity analysis to determine whether this subpopulation was driving the MI association of the entire study cohort. Table 4 shows that when this population was removed from the analysis, there was still a significant positive association between those who had an MI in the prior year (OR 1.12, CI: 1.01, 1.23) with each 1 ug/m3 increase in annual average PM2.5 concentration. The odds of having an MI in the prior year increased by 22% (95% CI: 0.86, 1.72) with each 1 μg/m3 in annual average PM2.5, for those 610 individuals.

Table 4.

Association between 1 μg/m3 increase in annual average PM2.5 and measures of myocardial infarction for those with and without a CAD index

With CAD index Without CAD index
Outcome Events Odds Ratioa 95% CI Events Odds Ratioa 95% CI
MI in prior year 666 1.12 1.01, 1.23 38 1.22 0.86, 1.72
Open in a new tab
a

Models are adjusted for race, area level education, gender, and smoking status.

CATHGEN cohort consists of several thousand individuals, so it is possible to explore the contribution of various effect modifiers on the outcomes. Table 5 shows that participants aged 65 years old or greater had higher odds of having a CAD score >23 for each 1 μg/m3 increase in annual average PM2.5 (OR: 1.20, 95% CI: 1.08, 1.32) compared to those participants under the age of 65 (OR: 1.06, 95% CI: 0.97, 1.16), with p=0.08 for modification. The PM2.5-CAD association was also modified by race/ethnicity and history of hypertension.

Table 5. Effect modification of the PM2.5-CAD index (>23) association by select covariates.

Subjects (%) Odds ratio 95% CI LRT p-value
Age (years)
 <65 3066 (60.5) 1.06 0.97, 1.16
 ≥65 2003 (39.5) 1.20 1.08, 1.32 0.08
Gender
 Male 3145 (62.0) 1.12 1.03, 1.21
 Female 1924 (38.0) 1.10 0.98, 1.23 0.77
Racea
 Non-Hispanic White 3703 (73.1) 1.19 1.10, 1.28
 African American 1071 (21.1) 0.87 0.75, 1.00
 Other 295 (5.8) 1.10 0.76, 1.59 0.001
History of smoking
 Yes 2475 (48.8) 1.16 1.06, 1.27
 No 2594 (51.2) 1.07 0.97, 1.17 0.21
Body mass index (kg/m2)
 <18.5 (Underweight) 68 (1.4) 1.66 0.89, 3.13
 <25 (Under/ normal weight) 1030 (20.4) 1.09 0.94, 1.26
 25.0-29.9 (Overweight) 1802 (35.7) 1.18 1.06, 1.32
 ≥30.0 (Obese) 2144 (42.5) 1.05 0.95, 1.16 0.21
History of diabetes
 Yes 1523 (30.0) 1.06 0.94, 1.19
 No 3546 (70.0) 1.14 1.05, 1.24 0.28
History of hypertension
 Yes 3575 (70.5) 1.08 1.00, 1.17
 No 1494 (29.5) 1.26 1.11, 1.45 0.04
Urban/ rural statusb
 Urban 2743 (54.1) 1.13 1.03, 1.23
 Rural 2326 (45.9) 1.11 1.00, 1.22 0.79
Educational attainmentc
 Low 2057 (40.6) 1.10 0.98, 1.23
 High 3012 (59.4) 1.11 1.02, 1.20 0.86
Open in a new tab
a

Other race/ethnicity includes Native American, Hispanic, Asian, and unknown.

b

Urban status was defined as living in a block group that was ≥50% urban.

c

Low educational attainment includes those who live in block groups where ≥25% of males and females have less than a high school education.

Table 6 shows that participants in the underweight (OR: 1.81, 95% CI: 0.64, 5.15) and normal weight (OR: 1.41, 95% CI: 1.11, 1.79) BMI categories had stronger associations between long-term PM2.5 exposure and having an MI in the year prior, though the interaction term did not reach significance. There was modification by diabetes status, with those without diabetes having a stronger association between long-term PM2.5 exposure and having an MI in the year prior (OR: 1.21, 95% CI: 1.07, 1.36).

Table 6.

Effect measure modification of the PM2.5-MI in prior year association by select covariates

Subjects (%) Odds ratio 95% CI LRT p-value
Age (years)
 <65 3442 (60.6) 1.12 0.99, 1.26
 ≥65 2237 (39.4) 1.19 1.01, 1.40 0.55
Gender
 Male 3471 (61.1) 1.13 1.00, 1.27
 Female 2208 (38.9) 1.18 0.99, 1.39 0.67
Race
 Non-Hispanic White 4146 (73.0) 1.18 1.05, 1.33
 African American 1204 (21.2) 1.02 0.84, 1.23
 Othera 329 (5.8) 1.25 0.77, 2.02 0.38
History of smoking
 Yes 2664 (46.9) 1.22 1.07, 1.40
 No 3015 (53.1) 1.06 0.92, 1.22 0.15
Body mass index (kg/m2)
 <18.5 (Underweight) 80 (1.4) 1.81 0.64, 5.15
 <25 (Under/ normal weight) 1187 (21.0) 1.41 1.11, 1.79
 25.0-29.9 (Overweight) 1987 (35.2) 1.05 0.91, 1.23
 ≥30.0 (Obese) 2399 (42.4) 1.10 0.95, 1.28 0.15
History of diabetes
 Yes 1660 (29.2) 1.02 0.87, 1.20
 No 4019 (70.8) 1.21 1.07, 1.36 0.09
History of hypertension
 Yes 3882 (68.4) 1.14 1.02, 1.27
 No 1797 (31.6) 1.17 0.97, 1.41 0.81
Urban/ rural status
 Urban 3110 (54.8) 1.09 0.96, 1.24
 Rural 2569 (45.2) 1.25 1.07, 1.45 0.18
Educational attainmentc
 Low 2290 (40.3) 1.10 0.95, 1.28
 High 3389 (59.7) 1.15 1.02, 1.30 0.65
Open in a new tab
a

Other race/ethnicity includes Native American, Hispanic, Asian, and unknown.

b

Urban status was defined as living in a block group that was ≥50% urban.

c

Low educational attainment includes those who live in block groups where ≥25% of males and females have less than a high school education.

4. Discussion

We investigated the association between long-term PM2.5 exposure and select cardiovascular disease outcomes using a cardiovascular cohort enriched with prevalent coronary artery disease. We used satellite-based predictions to assess long-term PM2.5 exposure for NC residents of the cohort. We observed positive associations between long-term PM2.5 exposure and CAD and measures of prior MI. These associations remained after adjusting for other covariates in sensitivity analyses. We additionally observed effect modification of the PM-CAD index relationship by age, race, and history of hypertension. Future studies will take advantage of the richness of the CATHGEN cohort that includes a large number of CAD events, extensive clinical information, genetic/epigenetic data, and biological phenotyping from blood which can potentially shed light on the how both short- and long-term air pollution exposure contributes to CAD.

Other large population-based cohorts such as ARIC (Kan et al., 2008), WHI (Miller et al., 2007), Nurses Health Study (Puett et al., 2009) and MESA Air (Kaufman et al., 2012) have reported a positive association between air pollutant exposures and clinically important cardiovascular outcomes or increases in biomarkers of atherosclerosis. Yet, in each of these studies the population was free from overt coronary heart disease at the time of enrollment. By contrast the CATHGEN cohort is highly enriched with research subjects with coronary artery disease. Thus, the high prevalence of CAD in this cohort as well as spatial distribution of the cohort offers a unique opportunity to investigate the effect modification of the PM2.5-CAD relation by community characteristics and social stressors. Knowledge of such relationships might be important in understanding some drivers of health disparities and crafting more effective environmental health literacy messaging to public health officials, health care providers and the public to mitigate the public health burden of particulate air pollution, particularly among at-risk populations.

Many of the previous studies on long-term PM2.5 exposure and CVD morbidity have used pollution data from air quality monitoring networks to assign study participants' PM2.5 exposure. Using only data from ground-based monitors can potentially introduce bias because it assumes that everyone residing within a certain radius of a monitor has the same PM2.5 exposure. Further, counties without air pollution monitors tend to be predominantly rural, have greater percentage of older adults, and more poverty than counties with air pollution monitors (Bravo et al., 2012; Miranda et al., 2011) and are therefore not as well studied as urban areas with more monitors. In contrast, satellite data expand spatial coverage to an entire state or region of the country, enhancing the ability to estimate location- and/or subject-specific exposure PM2.5. In this analysis we used PM2.5 concentration data generated by previously developed and published models that fused satellite data with ground based monitoring to create daily measurements of PM2.5 throughout the entire state of NC, with a resolution of 10km × 10km. There was high predictability between modeled and monitored values at sites where monitors were located, which ensures less uncertainty in the resulting estimates (Lee et al., 2012). For this study annual averages were obtained by averaging the individual daily exposures for the 12 months prior to a person's catheterization procedure in the 10 × 10 km grid in which that person resided. Satellite derived predictions can reduce measurement error and exposure misclassification (Kloog et al., 2011; Zeger et al., 2000), and have recently been used in epidemiological studies of air pollution exposure and mortality (Lee et al., 2015), birth outcomes (Hyder et al., 2014; Lakshmanan et al., 2015), and incidence of acute MIs (Madrigano et al., 2013).

By using satellite-derived predictions combined with a diverse population and clinical phenotyping we were able to conduct stratified analyses and assess the effect of potential modifiers of the air pollution-CAD associations. We observed associations between long-term PM2.5 exposure and increased prevalence of CAD and MI in non-Hispanic whites, but not in African Americans. One possible explanation for this finding is that the prevalence of CAD tends to be greater in non-Hispanic whites compared to African Americans (Budoff et al., 2002), while African Americans tend to have higher blood pressure and incidence of strokes than non-Hispanic whites (Gutierrez and Williams, 2014). Additional research is needed to see if there is an association between long-term PM2.5 exposure and more strokes in African Americans.

We also observed stronger differences in PM2.5 associations for CAD among those aged 65 and above. This finding is consistent with previous research that shows stronger associations with PM2.5 exposure for older adults (Shumake et al., 2013). Surprisingly, we observed stronger associations with PM2.5 among individuals without diabetes. A few studies have found similar results (Wilker et al., 2013), though most have found individuals with diabetes to be more susceptible to the effects from air pollution (Dubowsky et al., 2006; O'Neill et al., 2005; Schneider et al., 2010). Finally, we observed stronger associations between PM2.5 and CAD and the MI outcome among underweight individuals, although the interaction term did not reach significance. Previous studies have found underweight individuals to potentially be more susceptible to experiencing severe cardiovascular disease and mortality (Sharma et al., 2014), with a slightly elevated BMI being protective in some cases (Romero-Corral et al., 2006).

Some have suggested that health effects from long-term PM2.5 exposure may differ for urban and rural residents (Correia et al., 2013). Few studies, however, take rural residents into consideration when examining air pollution and CVD outcomes, primarily because of lack of PM2.5 monitors in rural locations. For this analysis we used US Census data to classify participants as rural and urban, depending on which block group they resided in. The use of satellite data allows inclusion of both urban and rural residents. We found stronger associations between PM2.5 and having an MI in the prior year among rural residents, although the interaction term did not reach significance. This finding is consistent with a recent analysis of short-term PM2.5 exposure and hospital admissions (Kloog et al., 2014), which reported associations in rural but not urban residents. Future studies should make an increased effort to include rural residents when considering health effects from air pollution.

The current annual PM2.5 NAAQS is 12.0 ug/m3, but the shape of the dose-response curve near the standard is uncertain. To partially address this issue we removed from the study all measurements that were above the current standard. The findings between long-term PM2.5 exposure and each of the select CVD outcomes remained when restricting the study population to those exposed to long term PM2.5 levels below 12.0 μg/m3.

During the catheterization procedure, 610 individuals underwent a therapeutic catheterization only rather than a complete diagnostic catheterization. Thus, not all of their coronary arteries were imaged and scored, and they were not assigned a CAD index score. We observed stronger associations for the MI outcomes for these 610 individuals. This population is likely at a more critical stage of their disease and it may not be surprising that environmental influences might be stronger.

In the present study, we found positive associations between long-term PM2.5 exposure and severity of atherosclerosis as measured by the CAD index. Our positive findings between long-term PM2.5 exposure and severity of atherosclerosis and measures of MI are broadly consistent with previous studies which have reported associations between long-term PM2.5 and CVD mortality, incidence of MIs, and increased susceptibility to progression of atherosclerosis (Brook et al., 2010; Newman et al., 2015). Previous research has identified associations between long-term PM2.5 exposure and subclinical measures of atherosclerosis such as carotid intima medial thickness (CIMT) and aortic calcification (Allen et al., 2009; Brook and Rajagopalan, 2010; Diez Roux et al., 2008; Kunzli et al., 2005). Although previous findings on long-term PM2.5 exposure and incidence of MIs are inconsistent (Bhaskaran et al., 2009), a recent epidemiological analysis using satellite derived predictions found an association between long-term PM2.5 exposure and incidence of MI (Madrigano et al., 2013).

The current analysis has both strengths and limitations. One of them is the generalizability of the CATHGEN cohort to other clinical settings and geographic regions. CATHGEN was selected from patients at a cardiac catheterization laboratory at Duke University Medical Center. Most individuals were referred for catheterization because of suspected CAD, although about 30% of the CATHGEN participants were not found to have clinically significant CAD. In sensitivity analyses we found that adjusting for county of residence did not alter the associations, thus there was likely not strong bias due to subject selection into the cohort. However, there may be residual confounding present, particularly since we were unable to obtain information on several individual level socioeconomic indicators. Another limitation is that we did not have complete residential history for each participant. We did, however, have a full address history for each participant for the year prior to his or her catheterization, and studies have shown that exposure during the prior year is an adequate surrogate of long-term exposure associated with cardiovascular disease. A final limitation is the relatively coarse spatial resolution of 10 × 10 km. Higher spatial resolution will become available in the near future (1 × 1 km) and reduce exposure error as well produce daily exposure estimates within micro-environments in urban and rural areas (Chudnovsky et al., 2014; Chudnovsky et al., 2013). Finally, analysis of those below the regulatory limit may be mixing secular trends with low pollution areas. Despite these limitations, the current study has several strengths including the large sample size of the CATHGEN cohort, the predictive ability of the CAD index, and the reduction in exposure error due to the satellite-derived predictions.

5. Conclusions

Our findings indicate that long-term PM2.5 exposure was associated with both coronary artery disease and prevalence of MIs in a cohort of cardiac catheterization patients. We have demonstrated that our AOD-based exposure model can be successfully applied to estimate spatially resolved PM2.5 exposures for individual addresses. Since individuals with cardiovascular disease are known to be especially vulnerable to air pollution, understanding the relationship between exposure and adverse events and disease risk in this population would have a greater impact on public health than a similarly-sized study of the general population.

Supplementary Material

1

Highlights.

  • The research looked at long-term PM2.5 exposure and select cardiovascular outcomes.

  • We utilized a cohort of cardiac catheterization patients who resided in NC.

  • Long-term PM2.5 exposure was characterized using satellite-derived estimates.

  • PM2.5 exposure was associated with coronary artery disease and having a recent MI.

Acknowledgments

Funding Sources: This research was supported by intramural research funding by the US EPA. The Health Effects Institute (Research Agreement #4946-RFPA10-3/14-7) and the NIEHS (award number T32ES007018) provided additional funding.

Footnotes

Competing Financial Interests: None.

Disclaimer: The research described in this article has been reviewed by the Environmental Protection Agency and approved for publication. The contents of this article do not necessarily represent Agency policy nor does mention of trade names or commercial products constitute endorsement or recommendation for use.

Human Study Approval (from page 6 of the manuscript): All subjects received and signed informed consent forms prior to enrollment, and CATHGEN has been approved by and follows all Duke University Institutional Review Board policies.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Allen RW, et al. Fine particulate matter air pollution, proximity to traffic, and aortic atherosclerosis. Epidemiology. 2009;20:254–64. doi: 10.1097/EDE.0b013e31819644cc. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bart BA, et al. Clinical determinants of mortality in patients with angiographically diagnosed ischemic or nonischemic cardiomyopathy. J Am Coll Cardiol. 1997;30:1002–8. doi: 10.1016/s0735-1097(97)00235-0. [DOI] [PubMed] [Google Scholar]
  3. Bhaskaran K, et al. Effects of air pollution on the incidence of myocardial infarction. Heart. 2009;95:1746–59. doi: 10.1136/hrt.2009.175018. [DOI] [PubMed] [Google Scholar]
  4. Bravo MA, et al. Comparison of exposure estimation methods for air pollutants: ambient monitoring data and regional air quality simulation. Environ Res. 2012;116:1–10. doi: 10.1016/j.envres.2012.04.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brook RD, Rajagopalan S. Particulate matter air pollution and atherosclerosis. Curr Atheroscler Rep. 2010;12:291–300. doi: 10.1007/s11883-010-0122-7. [DOI] [PubMed] [Google Scholar]
  6. Brook RD, et al. Particulate matter air pollution and cardiovascular disease: An update to the scientific statement from the American Heart Association. Circulation. 2010;121:2331–78. doi: 10.1161/CIR.0b013e3181dbece1. [DOI] [PubMed] [Google Scholar]
  7. Budoff MJ, et al. Ethnic differences in coronary atherosclerosis. J Am Coll Cardiol. 2002;39:408–12. doi: 10.1016/s0735-1097(01)01748-x. [DOI] [PubMed] [Google Scholar]
  8. Cesaroni G, et al. Long term exposure to ambient air pollution and incidence of acute coronary events: prospective cohort study and meta-analysis in 11 European cohorts from the ESCAPE Project. BMJ. 2014;348:f7412. doi: 10.1136/bmj.f7412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chudnovsky A, et al. Fine particulate matter predictions using high resolution aerosol optical depth (AOD) retrievals. Atmospheric Environment. 2014;89:189–198. [Google Scholar]
  10. Chudnovsky AA, et al. Spatial scales of pollution from variable resolution satellite imaging. Environ Pollut. 2013;172:131–8. doi: 10.1016/j.envpol.2012.08.016. [DOI] [PubMed] [Google Scholar]
  11. Chudnovsky AA, et al. Prediction of daily fine particulate matter concentrations using aerosol optical depth retrievals from the Geostationary Operational Environmental Satellite (GOES) J Air Waste Manag Assoc. 2012;62:1022–31. doi: 10.1080/10962247.2012.695321. [DOI] [PubMed] [Google Scholar]
  12. Correia AW, et al. Effect of air pollution control on life expectancy in the United States: an analysis of 545 U.S. counties for the period from 2000 to 2007. Epidemiology. 2013;24:23–31. doi: 10.1097/EDE.0b013e3182770237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Diez Roux AV, et al. Long-term exposure to ambient particulate matter and prevalence of subclinical atherosclerosis in the Multi-Ethnic Study of Atherosclerosis. Am J Epidemiol. 2008;167:667–75. doi: 10.1093/aje/kwm359. [DOI] [PubMed] [Google Scholar]
  14. Dubowsky SD, et al. Environ Health Perspect. 2006. Diabetes, obesity, and hypertension may enhance associations between air pollution and markers of systemic inflammation; pp. 114pp. 992–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. EPA., U S. Integrated Science Assessment for Particulate Matter (Final Report) U.S. Environmental Protection Agency; Washington, DC: 2009. EPA/600/R-08/139F. [PubMed] [Google Scholar]
  16. Greenland S, et al. Causal diagrams for epidemiologic research. Epidemiology. 1999;10:37–48. [PubMed] [Google Scholar]
  17. Gutierrez J, Williams OA. A decade of racial and ethnic stroke disparities in the United States. Neurology. 2014;82:1080–2. doi: 10.1212/WNL.0000000000000237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hoek G, et al. Long-term air pollution exposure and cardio- respiratory mortality: a review. Environ Health. 2013;12:43. doi: 10.1186/1476-069X-12-43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hyder A, et al. PM2.5 exposure and birth outcomes: use of satellite- and monitor-based data. Epidemiology. 2014;25:58–67. doi: 10.1097/EDE.0000000000000027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kan H, et al. Environ Health Perspect. 2008. Prospective analysis of traffic exposure as a risk factor for incident coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study; pp. 116pp. 1463–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kaufman JD, et al. Prospective study of particulate air pollution exposures, subclinical atherosclerosis, and clinical cardiovascular disease: The Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air) Am J Epidemiol. 2012;176:825–37. doi: 10.1093/aje/kws169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kloog I, et al. Assessing temporally and spatially resolved PM2. 5 exposures for epidemiological studies using satellite aerosol optical depth measurements. Atmos Environ. 2011;45:6267–6275. [Google Scholar]
  23. Kloog I, et al. Short term effects of particle exposure on hospital admissions in the Mid-Atlantic states: a population estimate. PLoS One. 2014;9:e88578. doi: 10.1371/journal.pone.0088578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kraus WE, et al. A Guide for a Cardiovascular Genomics Biorepository: the CATHGEN Experience. J Cardiovasc Transl Res. 2015 doi: 10.1007/s12265-015-9648-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kunzli N, et al. Ambient air pollution and atherosclerosis in Los Angeles. Environ Health Perspect. 2005;113:201–6. doi: 10.1289/ehp.7523. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lakshmanan A, et al. Associations between prenatal traffic-related air pollution exposure and birth weight: Modification by sex and maternal pre-pregnancy body mass index. Environ Res. 2015;137:268–77. doi: 10.1016/j.envres.2014.10.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lee HJ, et al. Use of satellite-based aerosol optical depth and spatial clustering to predict ambient PM2.5 concentrations. Environ Res. 2012;118:8–15. doi: 10.1016/j.envres.2012.06.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lee HJ, et al. A novel calibration approach of MODIS AOD data to predict PM2.5 concentrations. Atmos Chem Phys. 2011;11:7991–8002. [Google Scholar]
  29. Lee M, et al. Acute effect of fine particulate matter on mortality in three Southeastern states from 2007-2011. J Expo Sci Environ Epidemiol. 2015 doi: 10.1038/jes.2015.47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lim SS, et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2224–60. doi: 10.1016/S0140-6736(12)61766-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Madrigano J, et al. Long-term exposure to PM2.5 and incidence of acute myocardial infarction. Environ Health Perspect. 2013;121:192–6. doi: 10.1289/ehp.1205284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Miller KA, et al. Long-term exposure to air pollution and incidence of cardiovascular events in women. N Engl J Med. 2007;356:447–58. doi: 10.1056/NEJMoa054409. [DOI] [PubMed] [Google Scholar]
  33. Miranda ML, et al. Making the environmental justice grade: the relative burden of air pollution exposure in the United States. Int J Environ Res Public Health. 2011;8:1755–71. doi: 10.3390/ijerph8061755. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Newman JD, et al. Particulate air pollution and carotid artery stenosis. J Am Coll Cardiol. 2015;65:1150–1. doi: 10.1016/j.jacc.2014.12.052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. O'Neill MS, et al. Diabetes enhances vulnerability to particulate air pollution-associated impairment in vascular reactivity and endothelial function. Circulation. 2005;111:2913–20. doi: 10.1161/CIRCULATIONAHA.104.517110. [DOI] [PubMed] [Google Scholar]
  36. Pope CA., 3rd Ischaemic heart disease and fine particulate air pollution. Heart. 2014 doi: 10.1136/heartjnl-2014-306800. [DOI] [PubMed] [Google Scholar]
  37. Puett RC, et al. Chronic fine and coarse particulate exposure, mortality, and coronary heart disease in the Nurses' Health Study. Environ Health Perspect. 2009;117:1697–701. doi: 10.1289/ehp.0900572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rivera M, et al. Association between long-term exposure to traffic-related air pollution and subclinical atherosclerosis: the REGICOR study. Environ Health Perspect. 2013;121:223–30. doi: 10.1289/ehp.1205146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Romero-Corral A, et al. Association of bodyweight with total mortality and with cardiovascular events in coronary artery disease: a systematic review of cohort studies. Lancet. 2006;368:666–78. doi: 10.1016/S0140-6736(06)69251-9. [DOI] [PubMed] [Google Scholar]
  40. Ruckerl R, et al. Health effects of particulate air pollution: A review of epidemiological evidence. Inhal Toxicol. 2011;23:555–92. doi: 10.3109/08958378.2011.593587. [DOI] [PubMed] [Google Scholar]
  41. Schneider A, et al. Association of cardiac and vascular changes with ambient PM2.5 in diabetic individuals. Part Fibre Toxicol. 2010;7:14. doi: 10.1186/1743-8977-7-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Sharma A, et al. Relationship of body mass index with total mortality, cardiovascular mortality, and myocardial infarction after coronary revascularization: evidence from a meta-analysis. Mayo Clin Proc. 2014;89:1080–100. doi: 10.1016/j.mayocp.2014.04.020. [DOI] [PubMed] [Google Scholar]
  43. Shumake KL, et al. Susceptibility of older adults to health effects induced by ambient air pollutants regulated by the European Union and the United States. Aging Clin Exp Res. 2013;25:3–8. doi: 10.1007/s40520-013-0001-5. [DOI] [PubMed] [Google Scholar]
  44. Sun M, et al. Particulate matter components and subclinical atherosclerosis: common approaches to estimating exposure in a Multi-Ethnic Study of Atherosclerosis cross-sectional study. Environ Health. 2013;12:39. doi: 10.1186/1476-069X-12-39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wilker EH, et al. Long-term exposure to black carbon and carotid intima-media thickness: the normative aging study. Environ Health Perspect. 2013;121:1061–7. doi: 10.1289/ehp.1104845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. World Health Organization. Burden of Disease from Ambient Air Pollution for 2012. 2014 [Google Scholar]
  47. Zeger SL, et al. Environ Health Perspect. 2000. Exposure measurement error in time-series studies of air pollution: concepts and consequences; pp. 108pp. 419–26. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS741259-supplement-1.docx (77.5KB, docx)

RESOURCES