Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 1.
Published in final edited form as: Environ Res. 2021 Feb 4;195:110827. doi: 10.1016/j.envres.2021.110827

Associations between PM2.5 metal components and QT interval length in the Normative Aging Study

Adjani A Peralta a,*, Joel Schwartz a,b,c, Diane R Gold a,c, Brent Coull d, Petros Koutrakis a
PMCID: PMC7987821  NIHMSID: NIHMS1674360  PMID: 33549618

Abstract

Background:

Several studies have found associations between increases in QT interval length, a marker of cardiac electrical instability, and short-term fine particulate matter (PM2.5) exposures. To our knowledge, this is the first study to examine the association between specific PM2.5 metal components and QT interval length.

Methods:

We measured heart-rate corrected QT interval (QTc) duration among 630 participants in the Normative Aging Study (NAS) based in Eastern Massachusetts between 2000 and 2011. We utilized time-varying linear mixed-effects regressions with a random intercept for each participant to analyze associations between QTc interval and moving averages (0–7 day moving averages) of 24-h mean concentrations of PM2.5 metal components (vanadium, nickel, copper, zinc and lead) measured at the Harvard Supersite monitoring station. Models were adjusted for daily PM2.5 mass estimated at a 1 km × 1 km grid cell from a previously validated prediction model and other covariates. Bayesian kernel machine regression (BKMR) was utilized to assess the overall joint effect of the PM2.5 metal components.

Results:

We found consistent results with higher lead (Pb) associated with significant higher QTc intervals for both the multi-pollutant and the two pollutant (PM2.5 mass and a PM2.5 component) models across the moving averages. The greatest effect of lead on QTc interval was detected for the 4-day moving average lead exposure. In the multi-pollutant model, each 2.72 ng/m3 increase in daily lead levels for a 4-day moving average was associated with a 7.91 ms (95% CI: 3.63, 12.18) increase in QTc interval. In the two-pollutant models with PM2.5 mass and lead, each 2.72 ng/m3 increase in daily lead levels for a 4-day moving average was associated with an 8.50 ms (95% CI: 4.59, 12.41) increase in QTc interval. We found that 4-day moving average of copper has a negative association with QTc interval when compared to the other PM2.5 metal components. In the multipollutant model, each 1.81 ng/m3 increase in daily copper levels for a 4-day moving average was associated with an −3.89 ms (95% CI: −6.98, −0.79) increase in QTc interval. Copper’s essential function inside the human body could mediate its cardiotoxicity on cardiac conductivity and explain why we found that copper in comparison to the other metals was less harmful for QTc interval.

Conclusions:

Exposure to metals contained in PM2.5 are associated with acute changes in ventricular repolarization as indicated by QT interval characteristics.

Keywords: Air pollution, Fine particulate matter, QT interval, Metals, Ventricular repolarization

1. Introduction

Epidemiological studies have reported a consistent association between exposure to fine particulate matter (PM2.5) and cardiovascular morbidity and mortality (Brook et al., 2010; Pope III et al., 2004b), but fewer studies have explored which particular PM2.5 components are responsible for the cardiotoxicity (Cakmak et al., 2009; Dales et al., 2020; Shutt et al., 2017). In particular, less is known about the relative toxicity of PM2.5 metal components independently of PM2.5 mass (Cakmak et al., 2014; Eum et al., 2011; Peltier and Lippmann, 2010; Zanobetti et al., 2008).

Irregularities in cardiac repolarization significantly contribute to the production of cardiac arrhythmias (Nattel et al., 2007). Increases in PM2.5 have been associated with increases in markers of autonomic function such as lower heart rate variability (Gold et al., 2000), increased heart rates (Peters et al., 1999) and a greater risk of cardiac arrhythmias (Peralta et al., 2020). Electrocardiogram (ECG) measurements of repolarization such as QT interval provide non-invasive indicators for possible cardiac arrhythmias and help identify patients susceptible to sudden cardiac death (Giuliani et al., 2014; Schwartz and Wolf, 1978). Prolonged QT intervals can predispose an individual to experience a life-threatening type of ventricular arrhythmia called torsades de pointes (Al-Khatib et al., 2003). While a QTc interval greater than 500 ms increases the risk for torsades de pointes, there is no predetermined threshold that is safe from proarrhythmic risk (Al-Khatib et al., 2003; Bednar et al., 2001). A previous study conducted in the NAS found a positive association between sub-chronic and long-term PM2.5 exposure and QTc interval, but did not assess how the individual metal components of PM2.5 could impact cardiac repolarization (Mordukhovich et al., 2016).

The chemical components of PM2.5 can be found both inside and on the surface of the particle. While there are both natural and anthropogenic sources for these chemical components, anthropogenic sources consist of auto vehicle emissions, industrial activity, fossil fuel combustion, burning of fuel oil and smoking byproducts (Garg et al., 2000; Lough et al., 2005; Pappas et al., 2014; Peltier and Lippmann, 2010). Past studies have found that the chemical composition of PM2.5 could contribute to daily average mortality. Lippmann et al. (2006) found that vanadium and nickel, associated with burning of fuel oils, increased the daily average mortality in New York City. Furthermore, Zanobetti et al. (2008) found that nickel significantly modified that association between PM2.5 mass and daily cardiovascular hospital admissions in 26 U.S. cities.

Only a few studies have examined how PM2.5 metal components could affect cardiovascular markers (Jacobs et al., 2012; Williams et al., 2012). Jacobs et al. (2012) found that among older individuals taking antihypertensive medication, vanadium, iron and nickel content in PM2.5 was significantly associated with systolic blood pressure and pulse pressure. Another study on a non-smoking longitudinal adult cohort in Detroit, Michigan reported a positive association with PM2.5 metal components and brachial artery diameter and a negative association with PM2.5 metal components, blood pressure and heart rate. Lead is well known to have toxic effects on the cardiovascular system including increase in blood pressure and the risk of left ventricular hypertrophy (Schwartz, 1991), as well as neurotoxic effects (Lanphear et al., 2005), which may extend to the autonomic nervous system. However, the underlying mechanisms of the association between ambient air pollution and cardiovascular morbidity are only partly understood, particularly for mixtures of air pollution components.

We evaluate whether short-term exposures to PM2.5 metal components are associated with associated with heart rate corrected QT interval (QTc) duration, which is a marker of ventricular repolarization, in the Normative Aging Study cohort. We hypothesize that exposure to a mixture of PM2.5 metal components (vanadium, nickel, copper, zinc and lead) elevates QTc interval and that each individual PM2.5 metal component will increase QTc interval, a marker of ventricular repolarization, among the 563 men living in Eastern Massachusetts. To our knowledge, this is the first study to assess the effects of PM2.5 metal components on ventricular repolarization through changes in QTc interval.

2. Methods

2.1. Study population

The participants in this study included 563 elderly men living in Eastern Massachusetts who are part of the Veterans Affairs Normative Aging Cohort with up to four visits during the period 2000–2012. Inclusion criteria for the initial cohort required no previous history of chronic disease and the ability to participate in at least one onsite A.A. Peralta et al. physical examination and questionnaire every 3–5 years. Previous studies have reported the enrollment and inclusion requirements in more detail (Baja et al., 2010; Bell et al., 1972). In brief, physical examinations and interviews provided information on the participants height and weight to calculate their Body Mass Index (BMI), current medication use and fasting blood samples to assess cholesterol levels (Peters et al., 2012). Smoking and drinking status were obtained from physician administered questionnaires. Diabetic status was assigned based on a physician’s diagnosis of type II diabetes or the reported use of diabetic medication during a study visit. Mean atrial pressure (MAP) was calculated from the systolic and diastolic blood pressures (SBP and DBP) measured by the physician during a site visit.

While there were 630 total participants during this time with at least one QTc measurement, only 563 of them had all the necessary covariates for this analysis. The study was missing information on forty participants on ozone. One participant was missing information on cholesterol and another on race. Six were missing information on education and four on smoking status. We also excluded 15 participants with no information on PM2.5 metal exposure. Between November 14, 2000 and December 21, 2011, these 563 participants came in for a total of 985 study visits.

The Institutional Review Boards of participating institutions, Harvard T.H. Chan School of Public Health, and the Veteran Administration, approved the study protocol and all participants provided written informed consent.

2.2. ECG measurement and analysis

QTc measurements were obtained from electrocardiogram measurements (ECG). These measurements were obtained at the exam site (VA Boston Healthcare System, Boston, MA) for 5–10 min between 05:30 and 14:00 h with a two-channel (five lead) ECG monitor (Trillium 3000; Forest Medical, Inc., East Syracuse, NY) using a sampling rate of 256 Hz per channel (Park et al., 2005). An earlier study provides more detailed report on how the ECG measurements were processed to attain the corrected QT interval values (Mehta et al., 2014). Briefly, the ECG recordings were processed using the Trillium 3000 software to create a Mathcad (Parametric Technology Corporation, Needham, MA) file that includes the QT interval values. Corrected QT values were calculated using Bazett’s formula by only measuring the start of a normal or supraventricular beat to the end of a T wave with sufficient amplitude (Bednar et al., 2001; Mehta et al., 2014). QTc measurements were expressed in milliseconds (ms).

2.3. Air pollution and meteorology variables

We retrieved daily PM2.5 predictions and ozone predications at 1 km × 1 km grid cells in the continental U.S. using a well-validated model incorporating land use, meteorology, chemical transport models, and satellite remote sensing. Three models were trained using a neural network model, a random forest, and gradient boosting, and then ensemble averaged using a geographically weighted regression (Di et al., 2019). Each patient’s residential address was linked to the nearest center of a 1 km × 1 km grid cell for their exposure estimate. We obtained daily minimum and maximum surface meteorological data for temperature and relative humidity at a spatial resolution of 4 km × 4 km from gridMET for the continental U.S. (Abatzoglou, 2013). The minimum and maximum daily measurement was averaged to create a daily temperature or relative humidity value. Afterwards, each participant’s residential address was linked to the nearest center grid cell for their exposure estimate.

2.4. PM metal components

Daily ambient concentrations of the PM2.5 metal components were collected at the Harvard Supersite in Boston, MA during the study period 2000–2011. The Supersite is located on the roof of the Countway library of the Harvard Medical School, which is approximately one mile from the VA examination site where the ECG measurements took place (VA Boston Healthcare System, Boston, MA). The study focused on five PM2.5 metal components chosen a priori: vanadium, nickel, copper, zinc, and lead based on previous literature (Campen et al., 2001; Eum et al., 2011; Lim et al., 2017; Lippmann et al., 2006; Peltier and Lippmann, 2010; Zanobetti et al., 2008). Daily PM2.5 samples were collected on Teflon filters utilizing Harvard Impactors (Koutrakis et al., 1993) and the PM2.5 elements were evaluated with an Energy Dispersive X-ray Fluorescence Spectrometer (Epsilon 5, PANalytical, Almelo, The Netherlands).

2.5. Statistical analysis

We utilized time-varying linear mixed-effects regressions with a random intercept and Bayesian kernel machine regression (BKMR) to analyze associations between QTc interval and moving averages (0–7 day moving averages) of 24-h mean concentrations of PM2.5 metal components (vanadium, nickel, copper, zinc and lead) measured at the Harvard Supersite monitoring station. We report the changes in QTc interval in milliseconds and 95% CI in QTc interval for an interquartile range (IQR) increase in zero to seven-day moving average for each individual PM2.5 metal component.

Bayesian kernel machine regression (BKMR) is a Bayesian approach that uses a regression kernel to consider high order nonlinearities (squares, cubes, etc.) and interactions among a mixture of exposures, and evaluate which form best explains the outcome (Bobb et al, 2015, 2018). This method controls for multicollinearity, non-linear and non-additive effects while adjusting for any relevant covariates and potential confounders. The model parameters are treated as random variables which help identify a) the most relevant components in a mixture, b) dose-response curves for those components, and c) the overall effect of that mixture and interactions between each component. The BKMR results were reported with the 95% posterior credible interval (PCI) with the other exposures fixed at their 50th percentile. Prior to data analysis for the BKMR model, all the continuous variables were logged, centered, and standardized. We utilized 50,000 iterations for the Markov Chains and to generate the predictions and burned the first half of the iterations.

Three different models were used to examine the association between PM2.5 metal components and QTc interval. The first model was a multi-pollutant model where all the individual metal components were included, the second model was a two-pollutant model that consisted of each individual PM2.5 metal components and PM2.5 mass and the third model was the BKMR analysis that included all the individual metal components.

All models were adjusted for potential cofounders that were identified through a literature search (Baja et al., 2010; Mordukhovich et al., 2016). These potential confounders were related to both QTc interval and air pollution exposure and visualized with directed acyclic graphs (Shrier and Platt, 2008). The covariates included were daily PM2.5 mass, daily ozone, age (years), race, maximum years of education, BMI (kg/m2), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta blocker medication, alcohol intake (2 or more drinks per day or less than 2 drinks per day as reference), smoking status (current, former or never as reference), census tract percent of population age 25 years or older with less than a high school diploma, air temperature (°C), relative humidity (%) and seasonality (sine and cosine).

We performed two sensitivity analysis. First, we examined if season would alter the association between the PM2.5 metal components and QTc interval with stratification in the multipollutant model. Second, to further explore the issue of multicollinearity, we assessed if the effect estimates reported in the multipollutant model would change if we excluded PM2.5 metal components that were highly correlated with each other (excluded vanadium and zinc).

Data management and all statistical analyses were conducted using R version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria).

3. Results

The study included 563 VA Normative Aging Study participants who had all the relevant covariates for this analysis. The participants were older males with a mean age (±SD) of 74.1 years ± 6.8 years who were mostly white (96.6%). 81% of the participants lived within 50 kms of the Supersite. Table 1 presents other characteristics of the study patients. Table 2 presents the summary statistics and Spearman’s correlations between the PM2.5 metal components and meteorological measurements for the study period. During this time, the median PM2.5 mass concentration was 8.2 μg/m3 and the highest correlation between PM2.5 metal components was between nickel and vanadium (Spearman correlation coefficient, ρ = 0.84). The PM2.5 metal components were positively correlated with each other while temperature and relative humidity was negatively correlated with some of the PM2.5 metal components (temperature: vanadium, nickel and zinc; relative humidity: copper and lead). The highest negative correlation existed between temperature and nickel (Spearman correlation coefficient, ρ = −0.14).

Table 1.

Baseline characteristics of the 563 study participants in the VA Normative Aging Study during the study period between November 14, 2000 and December 21, 2011.

Characteristics Mean (SD) N (%)
Age (years) 74.1 (6.8)
Race
White 544 (96.6)
Black 13 (2.3)
Hispanic (White) 5 (0.9)
Hispanic (Black) 1 (0.2)
Body mass index (kg/m2) 27.4 (3.9)
Total cholesterol (mg/dL) 175.0 (36.8)
Mean arterial pressure (mmHg) Diabetes 86.8 (11.1)
Yes 111 (19.7)
No 452 (80.3)
Beta blocker medication
Yes 221 (39.3)
No 342 (60.7)
Maximum years of education 15.0 (3.0)
Alcohol intake
<2 drinks per day 461 (81.9)
2+ drinks per day 102 (18.1)
Smoking status
Never smoker 167 (29.7)
Current smoker 28 (5.0)
Former smoker 7776)
Open in a new tab

Table 2.

Summary statistics and Spearman’s correlation coefficients of PM2.5 metal components and meteorological measurements in the VA Normative Aging Study between November 14, 2000 and December 21, 2011.

Summary Statistics Spearman’s correlation coefficients
Mean (SD) Median (IQR) V Ni Cu Zn Pb PM2.5 O3 Temp RH
V (ng/m3) 3.2 (2.9) 2.4 (3.4) 1.00 0.84 0.25 0.55 0.34 0.39 0.02 −0.01 0.05
Ni (ng/m3) 2.9 (3.2) 2.0 (3.1) 1.00 0.28 0.59 0.32 0.33 −0.04 −0.14 0.06
Cu (ng/m3) 3.7 (2.9) 3.5 (3.5) 1.00 0.28 0.20 0.24 0.01 0.10 −0.13
Zn (ng/m3) 13.5 (13.1) 10.6 (9.5) 1.00 0.32 0.33 −0.06 −0.05 0.03
Pb (ng/m3) 5.7 (3.7) 5.2 (4.1) 1.00 0.28 0.03 0.01 −0.01
PM2.5 (μg/m3) 9.9 (6.4) 8.2 (6.7) 1.00 0.19 0.17 0.01
Ozone (μg/m3) 54.0 (23.1) 52.0 (27.0) 1.00 0.34 −0.09
Temp (°C) 10.5 (10.1) 11.1 (16.2) 1.00 −0.09
RH (%) 61.9 (14.8) 65.0 (19.1) 1.00
Open in a new tab

Abbreviations: SD - standard deviation; IQR-interquartile range; V- vanadium; Ni-nickel; Cu-copper; Zn- zinc; Pb-lead; PM2.5- fine particulate matter mass; O3-ozone; Temp-temperature; RH- relative humidity.

Fig. 1 shows the results from the multi-pollutant linear mixed-effects regression model with all the PM2.5 metal components included in the same model adjusted for PM2.5 mass (Multi-pollutant), the two-pollutant models including each PM2.5 metal component and PM2.5 mass (Two-pollutant) and BKMR which included all five PM2.5 metal components. All models consistently showed that lead had the highest statistically significant effect on QTc interval across 2–7 day moving averages.

Fig. 1.

Fig. 1.

Open in a new tab

Changes in milliseconds and 95% CI in QTc interval for an IQR increase in zero to seven day moving average of each PM2.5 metal component. The results are presented in a multi-pollutant model where all the metal components are included in the same model, the two-pollutant model which includes each individual PM2.5 metal component and PM2.5 mass and the BKMR model. The results for BKMR are reported with the 95% posterior credible interval (PCI) with the other exposures fixed at their 50th percentile. All models are adjusted for PM2.5 mass, daily ozone, age (years), race, maximum years of education, BMI (kg/m2), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta blocker medication, alcohol intake (2 or more drinks per day or less than 2 drinks per day as reference), smoking status (current, former or never as reference), census tract percent of population age 25 years or older with less than a high school diploma, air temperature (°C), relative humidity (%) and seasonality (sine and cosine).

The greatest effect of lead on QTc interval was detected for the 4-day moving average lead exposure across all models. In the multi-pollutant model, each 2.7 ng/m3 increase in daily lead levels for a 4-day moving average was associated with an 7.91 ms (95% CI: 3.63, 12.18) increase in QTc interval. In the two-pollutant models with lead and PM2.5 mass, each 2.7 ng/m3 increase in daily lead levels for a 4-day moving average was associated with a 8.50 ms (95% CI: 4.59, 12.41) increase in QTc interval. In the BKMR model, each geometric IQR increase in daily lead levels for a 4-day moving average was associated with a 13.73 ms (95% CI: 5.69, 21.77) increase in QTc interval. The results were similar across all models with Pb providing the most consistent findings.

The multi-pollutant model and the BKMR model both suggest that 4-day moving average of copper has statistically significant negative association with QTc interval when compared to the other PM2.5 metal components.

Fig. 2 shows the estimated joint effect of the PM2.5 metal mixture on QTc interval length when all the predictors are fixed to different percentiles, as compared with when they are all fixed to the 50th percentile, supporting a strong and linear positive association of the whole mixture with increasing QTc interval length across all the moving averages.

Fig. 2.

Fig. 2.

Open in a new tab

Overall joint effect of the PM2.5 metal mixture for 0–7 day moving averages with QTc interval length estimated by Bayesian Kernel Machine Regression (BKMR). This figure compares the estimated change in QTc interval length when all predictors are at a certain quantile with the value when all of them are at their 50th percentile. BKMR models were adjusted for PM2.5 mass, daily ozone, age (years), race, maximum years of education, BMI (kg/m2), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta blocker medication, alcohol intake (2 or more drinks per day or less than 2 drinks per day as reference), smoking status (current, former or never as reference), census tract percent of population age 25 years or older with less than a high school diploma, air temperature (°C), relative humidity (%) and seasonality (sine and cosine).

For the sensitivity analysis, we found that season could alter the association between PM2.5 metal components and QTc interval (Fig. 3). In the fall, we found that lead continues to consistently have a statistically significant association with QTc interval across the 2–7 day moving averages and reports the largest PM2.5 lead association with QTc interval on the 7-day moving average (15.45 ms, 95% CI: 6.70, 24.21). Furthermore, nickel also showed a statistically significant association with QTc interval in the fall and reported the largest effect size for any PM2.5 metal component on the 6-day moving average (18.40 ms, 95% CI: 5.00, 31.80).

Fig. 3.

Fig. 3.

Open in a new tab

Change in QTc interval length and 95% CI for an IQR increase in 0–7 day moving average of each PM2.5 metal component in the multi-pollutant model stratified by season. Model was adjusted for PM2.5 mass, daily ozone, age (years), race, maximum years of education, BMI (kg/m2), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta blocker medication, alcohol intake (2 or more drinks per day or less than 2 drinks per day as reference), smoking status (current, former or never as reference), census tract percent of population age 25 years or older with less than a high school diploma, air temperature (°C), relative humidity (%) and seasonality (sine and cosine).

In the summer, we found a statistically significant negative association between QTc interval and copper for 0, 2–7 day moving averages and a statistically significant positive association between QTc interval and vanadium for the 0, 4–6day moving averages. We did not find any statistically significant associations between any PM2.5 metal component and QTc interval for the winter or spring seasons.

In our second sensitivity analysis, we removed zinc and vanadium to assess if the results were due to multicollinearity with copper and nickel and the results remained consistent (Fig. 4).

Fig. 4.

Fig. 4.

Open in a new tab

Changes in milliseconds and 95% CI in QTc interval for an IQR increase in zero to seven day moving average of each PM2.5 metal component. The results are presented in a multi-pollutant model where all the metal components are included in the same model except for the indicated PM2.5 metalcomponent. The models were adjusted for PM2.5 mass, daily ozone, age (years), race, maximum years of education, BMI (kg/m2), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta blocker medication, alcohol intake (2 or more drinks per day or less than 2 drinks per day as reference), smoking status (current, former or never as reference), census tract percent of population age 25 years or older with less than a high school diploma, air temperature (°C), relative humidity (%) and seasonality (sine and cosine).

4. Discussion

To our knowledge, this is the first study to investigate the short-term effects of particle metal components on QT interval length. In the Normative Aging Study, an Eastern Massachusetts longitudinal cohort, we found that higher exposure to individual PM2.5 metal components even after controlling for PM2.5 mass was associated with QTc interval length, a marker for ventricular repolarization. Specifically, we found that lead was significantly associated with higher QTc interval length across most periods ranging from 2to 7-day moving averages and that copper was associated with a lower QTc interval length. Since these analyses controlled for PM2.5 mass and ozone, this finding indicates that per unit mass, Cu has a smaller effect on increasing QTc, not that it actually reduced QTc. Hence, in comparison to the other four PM2.5 metal components, Cu is relatively less harmful for QT interval.

Our results further suggest that season, in particular the fall and summer, increases respectively the cardiovascular toxicity of nickel and vanadium. On the other hand, we found no significant associations with QT interval and zinc an indicator of traffic-related emissions. Zinc originates from zinc dithiophosphate which is an anti-wear and anti-oxidant produced by tire and brake wear and tailpipe emissions of motor oil (Apeagyei et al., 2011; Blok, 2005; Lough et al., 2005).

4.1. Lead

Lead exposure arises from inhalation of lead particles and ingestion of polluted water and food sources (Rosin, 2009). While the two major sources of lead were phased out (1978, lead-based paint) (Tong, 1990) or banned (1986, gasoline) (Mielke and Reagan, 1998) in the United States, the resuspension of old lead particles from gas and exterior paint continue to release lead into the environment. Lead not only accumulates on the top layer of soil but will remain for generations due to its half-life of approximately 700 years (Semlali et al., 2004). Current motor vehicle sources of lead include brake wear (Apeagyei et al., 2011), motor vehicle wheel weights (Root, 2015), vaporization from hot brake surfaces (Apeagyei et al., 2011; Grigoratos and Martini, 2015) and motor oil combustion (Davis et al., 2001). Lead’s bioavailability and presence in dust particles will have continued implications for public health.

Previous epidemiological studies on lead exposure mainly considered the cardiovascular effects of hypertension and blood pressure (Den Hond et al., 2002; Gambelunghe et al., 2016; Glenn et al., 2006; Lustberg and Silbergeld, 2002). Navas-Acien et al. (2007) conducted a systematic review and concluded that a causal relationship between lead exposure and hypertension, but did not have sufficient evidence to deduce a causal relationship between lead exposure and other clinical cardiovascular outcomes (Navas-Acien et al., 2007). In a study of the effects of bone and blood lead exposure on QTc interval length in the NAS, low-level cumulative exposure to bone lead was associated with a prolonged QTc interval, while no association was found with blood lead (Eum et al., 2011). Specifically, individuals in the lowest tertile of tibia lead compared to the highest tertile had a 7.95 ms (95% CI: 1.42, 14.45) increase in QTc interval and no association was found for blood lead levels. While we cannot directly compare our results with these studies, they support our hypothesis that PM2.5 lead can adversely impact cardiac conductivity through prolonged QT intervals.

4.2. Nickel and vanadium

Sources of nickel and vanadium include the burning of oil residual in office buildings and heavy fuel oil in marine engines (Corbin et al., 2018; Masri et al., 2015). In our study, V and Ni were highly correlated (ρ = 0.84) most likely due to their joint production through oil combustion. Studies conducted in New York City found that during the fall and winter months higher nickel concentrations could be found due to residual fuel oil used for space heating (Peltier and Lippmann, 2010) and that nickel could modify the association between PM2.5 mass and daily cardiovascular hospital admissions (Zanobetti et al., 2008). Consistent with these findings we found a positive association between PM2.5 nickel and vanadium levels and QTc interval. Animal models have found adverse cardiovascular associations with PM2.5 associated vanadium and nickel concentrations (Campen et al., 2001; Lippmann et al., 2006).

4.3. Copper

Sources of copper include brake wear and lining as well as copper additives in motor oil combustion (Cheung et al., 2010; Kaiser, 2003). Copper is an essential metal involved in the function of several enzymes and required for myocardial contractility (Medeiros, 2017). Several studies have reported that copper deficiency impacts atherosclerosis and increases the risk of coronary heart disease (Klevay, 1983; Medeiros, 2017). Furthermore, both epidemiological (Freisinger et al., 2004; Kinsman et al., 1990; Mielcarz et al., 2001) and animal (Allen and Klevay, 1978;[ISP] . Li et al., 2005a,b; Medeiros, 2017) studies have reported associations between copper deficiency and blood pressure changes, hyperlipidemia, and abnormal electrocardiograms. A study in elderly individuals in South Korea found that blood pressure and heart rate variability measures were associated with lead and strontium, but did not find a statistically significant association with nickel, vanadium, zinc, or copper (Lim et al., 2017). Although the association between copper and heart rate variability measures (Standard Deviation of Normal-to-Normal Intervals (SDNN), Root Mean Square of the Successive Differences (RMSSD), low frequency and high frequency) was not statistically significant, their analysis suggests a possible negative association with copper. However, copper is normally obtained by ingestion, not inhalation. On the other hand, inhaled copper and vanadium increased fibrinogen levels, and induced pulmonary vasoconstriction and phosphorylation of ERK1/2 and p38 in vivo (Huang et al., 2003; . Li et al., 2005a,b). Copper’s essential function inside the human body could mediate its cardiotoxicity on cardiac conductivity and explain why we found that copper in comparison to the other metals was less harmful for QTc interval.

4.4. Clinical perspective

From a clinical perspective, the QTc interval provides a noninvasive assessment tool for ventricular repolarization. Prolonged QT intervals can predispose an individual to experience a life-threatening type of ventricular arrhythmia called torsades de pointes (Al-Khatib et al., 2003). It is important to note that there is no predetermined threshold safe from proarrhythmic risk (Al-Khatib et al., 2003; Bednar et al., 2001). In controlled clinical trials, the US Food and Drug Administration (FDA) requires that pharmacologic medication not alter a patient’s QTc interval by more than 5 ms and warns that drugs that alter QTc between 5 and 20 ms have been associated with proarrhythmic risk (U.S. Food and Drug Administration, 2005). Many of the effect estimates reported in this study are greater than 5 ms and have the potential to increase the risk of arrhythmias.

To reduce the risk that the PM2.5 metal components can have on arrythmias, physicians could increase patient awareness especially among the elderly. It is important for physicians to determine their patient’s home environment. For patients that live near or at polluted areas, then suggesting a air purifier for their home or office could reduce their overall exposure levels. For patients that cannot avoid living near highly polluted areas, then physicians could take this into account when considering treatment options. A more aggressive treatment could be necessary if a patient resides in a highly polluted area since that would increase their risk for arrhythmias.

Numerous biological mechanisms have been proposed to explain how acute exposure to air pollution can induce cardiovascular morbidity including elevated levels of reactive oxygen species (Knaapen et al., 2002), endothelial injury and systemic inflammation (Pope III et al., 2016) and altered autonomic activation (Devlin et al., 2003; Park et al., 2005; Pope 3rd et al., 2004a). Several studies have reported associations between fine particle mass and cardiovascular outcomes representing these biological mechanisms, but none specifically address how PM2.5 metal components could contribute to myocardial vulnerability (ventricular arrhythmias and repolarization dynamics).

4.5. Strengths and limitations

One limitation of our study is that the concentrations of the PM2.5 metal components were assigned from the use of a single monitoring site. The PM2.5 metal exposures capture temporal resolution but are not spatially resolved. The reported correlations between the PM2.5 metal components may not hold for all residential locations. It is a source of measurement error in our study and a limitation since we currently do not have a spatial model for PM2.5 metal constituents. This potential for non-differential measurement error in our exposure assessment has been previously described (Dai et al., 2016). It is unlikely that any measurement error in the PM2.5 metal components concentrations is associated with the participant’s ECG readings because the exposure was measured independently from the QTc interval measurements. This non-differential misclassification underestimates the observed associations and bias the results towards the null (Sarnat et al., 2005).

Our study minimized outcome misclassification because the ECG recordings were processed using a specific software (Trillium 3000) instead of relying on manually readings of beat labels and QT intervals. The automated processing reduced the potential of outcome differential measurement errors and inter-technician variability.

Including multiple correlated exposures in the same statistical model can lead to the issue of multicollinearity. A major strength of this study is the use of different statistical approaches to deal with the collinearity issues. In the multipollutant model, the variance (the uncertainty of the estimate) can be large due to collinearity. We were able to show the robustness of the association between lead and copper with QTc interval length with both BKMR and our sensitivity analysis. Use of BKMR, a novel flexible statistical method, allowed us to present the joint effect of the PM2.5 metal mixtures and address multicollinearity and potential non-linear or non-additive effects.

By including each individual PM2.5 metal component and controlling for PM2.5 mass, we gain insight into the differential toxicity of these PM2.5 metal components. If PM2.5 mass was not included, then the reported effect sizes could be capturing the effect of other components. Thus, including PM2.5 along with the individual metal components allows us to capture the differential toxicity of the constituents while controlling for the average effect of PM2.5 (Mostofsky et al., 2012). We conclude that lead on average has the most adverse impact on QTc interval length and that copper compared to the other four PM2.5 metal components is less toxic towards ventricular repolarization.

Our inclusion of various potential confounders and use of both individual and census tract variables to control for socioeconomic status reduces the potential for residual confounding. Since the study population consists mainly of older white males, the results should be interpreted with caution if applying to others such as younger individuals, females, or other racial groups. Future studies could explore the effect of PM2.5 metal components on QTc interval for these other populations.

5. Conclusions

Exposure to PM2.5 metal components can alter an individual’s QTc interval length, a risk factor for arrythmias. In particular, acute exposure to lead increases QTc interval. On the other hand, copper is relatively less harmful towards QTc interval when compared to the other four PM2.5 metal constituents. Nickle and vanadium have a positive significant association with QTc interval during the fall and summer, respectively. By identifying increased toxicity to certain metals, we can advise individuals to avoid the sources of these cardiotoxic metals or identify potential interventions to reduce their proarrhythmic risk.

Funding

USEPA Grant RD 83479801 & RD 83587201; NIH grant T32ES007069 & T32HL098048.

Footnotes

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Abatzoglou JT, 2013. Development of gridded surface meteorological data for ecological applications and modelling. Int. J. Climatol 33, 121–131. 10.1002/joc.3413. [DOI] [Google Scholar]
  2. Al-Khatib SM, LaPointe NMA, Kramer JM, Califf RM, 2003. What clinicians should know about the QT interval. J. Am. Med. Assoc 289, 2120–2127. 10.1001/jama.289.16.2120. [DOI] [PubMed] [Google Scholar]
  3. Allen KGD, Klevay LM, 1978. Cholesterolemia and cardiovascular abnormalities in rats caused by copper deficiency. Atherosclerosis 29, 81–93. [DOI] [PubMed] [Google Scholar]
  4. Apeagyei E, Bank MS, Spengler JD, 2011. Distribution of heavy metals in road dust along an urban-rural gradient in Massachusetts. Atmos. Environ 45, 2310–2323. [Google Scholar]
  5. Baja ES, Schwartz JD, Wellenius GA, Coull BA, Zanobetti A, Vokonas PS, Suh HH, 2010. Traffic-related air pollution and QT interval: modification by diabetes, obesity, and oxidative stress gene polymorphisms in the normative aging study. Environ. Health Perspect 118, 840–846. 10.1289/ehp.0901396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bednar MM, Harrigan EP, Anziano RJ, Camm AJ, Ruskin JN, 2001. The QT interval. Prog. Cardiovasc. Dis 43, 1–45. [DOI] [PubMed] [Google Scholar]
  7. Bell B, Rose CL, Damon A, 1972. The normative aging study: an interdisciplinary and longitudinal study of health and aging. Aging Hum. Dev 3, 5–17. 10.2190/GGVP-XLB5-PC3N-EF0G. [DOI] [Google Scholar]
  8. Blok J, 2005. Environmental exposure of road borders to zinc. Sci. Total Environ 348, 173–190. [DOI] [PubMed] [Google Scholar]
  9. Bobb JF, Claus Henn B, Valeri L, Coull BA, 2018. Statistical software for analyzing the health effects of multiple concurrent exposures via Bayesian kernel machine regression. Environ. Heal 17, 67. 10.1186/s12940-018-0413-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bobb JF, Valeri L, Claus Henn B, Christiani DC, Wright RO, Mazumdar M, Godleski JJ, Coull BA, 2015. Bayesian kernel machine regression for estimating the health effects of multi-pollutant mixtures. Biostatistics 16, 493–508. 10.1093/biostatistics/kxu058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brook RD, Rajagopalan S, Pope CA III, Brook JR, Bhatnagar A, Diez-Roux AV, Holguin F, Hong Y, Luepker RV, Mittleman MA, Peters A, Siscovick D, Smith SC, Whitsel L, Kaufman JD, 2010. Particulate matter air pollution and cardiovascular disease. Circulation 121, 2331–2378. 10.1161/CIR.0b013e3181dbece1. [DOI] [PubMed] [Google Scholar]
  12. Cakmak S, Dales R, Kauri LM, Mahmud M, Van Ryswyk K, Vanos J, Liu L, Kumarathasan P, Thomson E, Vincent R, 2014. Metal composition of fine particulate air pollution and acute changes in cardiorespiratory physiology. Environ. Pollut 189, 208–214. [DOI] [PubMed] [Google Scholar]
  13. Cakmak S, Dales RE, Gultekin T, Vidal CB, Farnendaz M, Rubio MA, Oyola P, 2009. Components of particulate air pollution and emergency department visits in Chile. Arch. Environ. Occup. Health 64, 148–155. [DOI] [PubMed] [Google Scholar]
  14. Campen MJ, Nolan JP, Schladweiler MCJ, Kodavanti UP, Evansky PA, Costa DL, Watkinson WP, 2001. Cardiovascular and thermoregulatory effects of inhaled PM-associated transition metals: a potential interaction between nickel and vanadium sulfate. Toxicol. Sci 64, 243–252. [DOI] [PubMed] [Google Scholar]
  15. Cheung KL, Ntziachristos L, Tzamkiozis T, Schauer JJ, Samaras Z, Moore KF, Sioutas C, 2010. Emissions of particulate trace elements, metals and organic species from gasoline, diesel, and biodiesel passenger vehicles and their relation to oxidative potential. Aerosol Sci. Technol 44, 500–513. [Google Scholar]
  16. Corbin JC, Mensah AA, Pieber SM, Orasche J, Michalke B, Zanatta M, Czech H, Massabò D, Buatier de Mongeot F, Mennucci C, 2018. Trace metals in soot and PM2. 5 from heavy-fuel-oil combustion in a marine engine. Environ. Sci. Technol 52, 6714–6722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dai L, Koutrakis P, Coull BA, Sparrow D, Vokonas PS, Schwartz JD, 2016. Use of the adaptive LASSO method to identify PM2. 5 components associated with blood pressure in elderly men: the Veterans Affairs Normative Aging Study. Environ. Health Perspect 124, 120–125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Dales R, Lee DS, Wang X, Cakmak S, Szyszkowicz M, Shutt R, Birnie D, 2020. Do acute changes in ambient air pollution increase the risk of potentially fatal cardiac arrhythmias in patients with implantable cardioverter defibrillators? Environ. Heal 19, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Davis AP, Shokouhian M, Ni S, 2001. Loading estimates of lead, copper, cadmium, and zinc in urban runoff from specific sources. Chemosphere 44, 997–1009. [DOI] [PubMed] [Google Scholar]
  20. Den Hond E, Nawrot T, Staessen JA, 2002. The relationship between blood pressure and blood lead in NHANES III. J. Hum. Hypertens 16, 563–568. [DOI] [PubMed] [Google Scholar]
  21. Devlin RB, Ghio AJ, Kehrl H, Sanders G, Cascio W, 2003. Elderly humans exposed to concentrated air pollution particles have decreased heart rate variability. Eur. Respir. J 21, 76s–80s. [DOI] [PubMed] [Google Scholar]
  22. Di Q, Amini H, Shi L, Kloog I, Silvern R, Kelly J, Sabath MB, Choirat C, Koutrakis P, Lyapustin A, 2019. An ensemble-based model of PM2. 5 concentration across the contiguous United States with high spatiotemporal resolution. Environ. Int 130, 104909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Eum K-D, Nie LH, Schwartz J, Vokonas PS, Sparrow D, Hu H, Weisskopf MG, 2011. Prospective cohort study of lead exposure and electrocardiographic conduction disturbances in the Department of Veterans Affairs Normative Aging Study. Environ. Health Perspect 119, 940–944. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Freisinger P, Horvath R, Macmillan C, Peters J, Jaksch M, 2004. Reversion of hypertrophic cardiomyopathy in a patient with deficiency of the mitochondrial copper binding protein Sco2: is there a potential effect of copper? J. Inherit. Metab. Dis 27, 67–79. [DOI] [PubMed] [Google Scholar]
  25. Gambelunghe A, Sallsten G, Borńe Y, Forsgard N, Hedblad B, Nilsson P, Fagerberg B, Engström G, Barregard L, 2016. Low-level exposure to lead, blood pressure, and hypertension in a population-based cohort. Environ. Res 149, 157–163. [DOI] [PubMed] [Google Scholar]
  26. Garg BD, Cadle SH, Mulawa PA, Groblicki PJ, Laroo C, Parr GA, 2000. Brake wear particulate matter emissions. Environ. Sci. Technol 34, 4463–4469. 10.1021/es001108h. [DOI] [Google Scholar]
  27. Giuliani C, Agostinelli A, Fioretti S, Nardo FD, Burattini LB, 2014. Abnormal repolarization in the acute myocardial infarction patients: a frequency-based characterization. Open Biomed. Eng. J 8, 42–51. 10.2174/1874120701408010042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Glenn BS, Bandeen-Roche K, Lee B-K, Weaver VM, Todd AC, Schwartz BS, 2006. Changes in systolic blood pressure associated with lead in blood and bone. Epidemiology 538–544. [DOI] [PubMed] [Google Scholar]
  29. Gold DR, Litonjua A, Schwartz J, Lovett E, Larson A, Nearing B, Allen G, Verrier M, Cherry R, Verrier R, 2000. Ambient pollution and heart rate variability. Circulation 101, 1267–1273. 10.1161/01.CIR.101.11.1267. [DOI] [PubMed] [Google Scholar]
  30. Grigoratos T, Martini G, 2015. Brake wear particle emissions: a review. Environ. Sci. Pollut. Res 22, 2491–2504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Huang Y-CT, Ghio AJ, Stonehuerner J, McGee J, Carter JD, Grambow SC, Devlin RB, 2003. The role of soluble components in ambient fine particles-induced changes in human lungs and blood. Inhal. Toxicol 15, 327–342. [DOI] [PubMed] [Google Scholar]
  32. Jacobs L, Buczynska A, Walgraeve C, Delcloo A, Potgieter-Vermaak S, Van Grieken R, Demeestere K, Dewulf J, Van Langenhove H, De Backer H, Nemery B, Nawrot TS, 2012. Acute changes in pulse pressure in relation to constituents of particulate air pollution in elderly persons. Environ. Res 117, 60–67. 10.1016/j.envres.2012.05.003. [DOI] [PubMed] [Google Scholar]
  33. Kaiser J, 2003. Manganese: a high-octane dispute: a debate over the health effects of airborne manganese is heating up as more and more countries begin adding the metal to gasoline.(News). Science (80) 300, 926–929. [DOI] [PubMed] [Google Scholar]
  34. Kinsman GD, Howard AN, Stone DL, Mullins PA, 1990. Studies in Copper Status and Atherosclerosis. [DOI] [PubMed] [Google Scholar]
  35. Klevay LM, 1983. Copper and ischemic heart disease. Biol. Trace Elem. Res 5, 245–255. [DOI] [PubMed] [Google Scholar]
  36. Knaapen AM, Shi T, Borm PJA, Schins RPF, 2002. Soluble metals as well as the insoluble particle fraction are involved in cellular DNA damage induced by particulate matter. In: Oxygen/Nitrogen Radicals: Cell Injury and Disease. Springer, pp. 317–326. [PubMed] [Google Scholar]
  37. Koutrakis P, Sioutas C, Ferguson ST, Wolfson JM, Mulik JD, Burton RM, 1993. Development and evaluation of a glass honeycomb denuder/filter pack system to collect atmospheric gases and particles. Environ. Sci. Technol 27, 2497–2501. 10.1021/es00048a029. [DOI] [Google Scholar]
  38. Lanphear BP, Hornung R, Khoury J, Yolton K, Baghurst P, Bellinger DC, Canfield RL, Dietrich KN, Bornschein R, Greene T, 2005. Low-level environmental lead exposure and children’s intellectual function: an international pooled analysis. Environ. Health Perspect 113, 894–899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Li Y, Wang L, Schuschke DA, Zhou Z, Saari JT, Kang YJ, 2005a. Marginal dietary copper restriction induces cardiomyopathy in rats. J. Nutr 135, 2130–2136. [DOI] [PubMed] [Google Scholar]
  40. Li Z, Carter JD, Dailey LA, Huang Y-CT, 2005b. Pollutant particles produce vasoconstriction and enhance MAPK signaling via angiotensin type I receptor. Environ. Health Perspect 113, 1009–1014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Lim Y-H, Bae H-J, Yi S-M, Park E, Lee B-E, Hong Y-C, 2017. Vascular and cardiac autonomic function and PM2. 5 constituents among the elderly: a longitudinal study. Sci. Total Environ 607, 847–854. [DOI] [PubMed] [Google Scholar]
  42. Lippmann M, Ito K, Hwang J-S, Maciejczyk P, Chen L-C, 2006. Cardiovascular effects of nickel in ambient air. Environ. Health Perspect 114, 1662–1669. 10.1289/ehp.9150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Lough GC, Schauer JJ, Park J-S, Shafer MM, DeMinter JT, Weinstein JP, 2005. Emissions of metals associated with motor vehicle roadways. Environ. Sci. Technol 39, 826–836. 10.1021/es048715f. [DOI] [PubMed] [Google Scholar]
  44. Lustberg M, Silbergeld E, 2002. Blood lead levels and mortality. Arch. Intern. Med 162, 2443–2449. [DOI] [PubMed] [Google Scholar]
  45. Masri S, Kang C-M, Koutrakis P, 2015. Composition and sources of fine and coarse particles collected during 2002–2010 in Boston, MA. J. Air Waste Manag. Assoc 65, 287–297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Medeiros DM, 2017. Perspectives on the role and relevance of copper in cardiac disease. Biol. Trace Elem. Res 176, 10–19. [DOI] [PubMed] [Google Scholar]
  47. Mehta AJ, Kloog I, Zanobetti A, Coull BA, Sparrow D, Vokonas P, Schwartz J, 2014. Associations between changes in city and address specific temperature and QT interval - the VA normative aging study. PloS One 9, e106258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Mielcarz G, Howard AN, Mielcarz B, Williams NR, Rajput-Williams J, Nigdigar SV, Stone DL, 2001. Leucocyte copper, a marker of copper body status is low in coronary artery disease. J. Trace Elem. Med. Biol 15, 31–35. [DOI] [PubMed] [Google Scholar]
  49. Mielke HW, Reagan PL, 1998. Soil is an important pathway of human lead exposure. Environ. Health Perspect 106, 217–229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Mordukhovich I, Kloog I, Coull B, Koutrakis P, Vokonas P, Schwartz J, 2016. Association between particulate air pollution and QT interval duration in an elderly cohort. Epidemiology 27, 284–290. 10.1097/EDE.0000000000000424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Mostofsky E, Schwartz J, Coull BA, Koutrakis P, Wellenius GA, Suh HH, Gold DR, Mittleman MA, 2012. Modeling the association between particle constituents of air pollution and health outcomes. Am. J. Epidemiol 176, 317–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Nattel S, Maguy A, Le Bouter S, Yeh Y-H, 2007. Arrhythmogenic ion-channel remodeling in the heart: heart failure, myocardial infarction, and atrial fibrillation. Physiol. Rev 87, 425–456. 10.1152/physrev.00014.2006. [DOI] [PubMed] [Google Scholar]
  53. Navas-Acien A, Guallar E, Silbergeld EK, Rothenberg SJ, 2007. Lead exposure and cardiovascular disease—a systematic review. Environ. Health Perspect 115, 472–482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Pappas RS, Fresquez MR, Martone N, Watson CH, 2014. Toxic metal concentrations in mainstream smoke from cigarettes available in the USA. J. Anal. Toxicol 38, 204–211. 10.1093/jat/bku013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Park SK, O’Neill MS, Vokonas PS, Sparrow D, Schwartz J, 2005. Effects of air pollution on heart rate variability: the VA normative aging study. Environ. Health Perspect 113, 304–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Peltier RE, Lippmann M, 2010. Residual oil combustion: distributions of airborne nickel and vanadium within New York City. J. Expo. Sci. Environ. Epidemiol 20, 342–350. 10.1038/jes.2009.28. [DOI] [PubMed] [Google Scholar]
  57. Peralta AA, Link MS, Schwartz J, Luttmann-Gibson H, Dockery DW, Blomberg A, Wei Y, Mittleman MA, Gold DR, Laden F, Coull BA, Koutrakis P, 2020. Exposure to Air Pollution and Particle Radioactivity with the Risk of Ventricular Arrhythmias. 10.1161/CIRCULATIONAHA.120.046321. [DOI] [PMC free article] [PubMed]
  58. Peters A, Perz S, Döring A, Stieber J, Koenig W, Wichmann HE, 1999. Increases in heart rate during an air pollution episode. Am. J. Epidemiol 150, 1094–1098. 10.1093/oxfordjournals.aje.a009934. [DOI] [PubMed] [Google Scholar]
  59. Peters JL, Kubzansky LD, Ikeda A, Fang SC, Sparrow D, Weisskopf MG, Wright RO, Vokonas P, Hu H, Schwartz J, 2012. Lead concentrations in relation to multiple biomarkers of cardiovascular disease: the Normative Aging Study. Environ. Health Perspect 120, 361–366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Pope CA 3rd, Hansen ML, Long RW, Nielsen KR, Eatough NL, Wilson WE, Eatough DJ, 2004a. Ambient particulate air pollution, heart rate variability, and blood markers of inflammation in a panel of elderly subjects. Environ. Health Perspect 112, 339–345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Pope CA III, Bhatnagar A, McCracken JP, Abplanalp W, Conklin DJ, O’Toole T, 2016. Exposure to fine particulate air pollution is associated with endothelial injury and systemic inflammation. Circ. Res 119, 1204–1214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Pope CA III, Burnett RT, Thurston GD, Thun MJ, Calle EE, Krewski D, Godleski JJ, 2004b. Cardiovascular mortality and long-term exposure to particulate air pollution: epidemiological evidence of general pathophysiological pathways of disease. Circulation 109, 71–77. [DOI] [PubMed] [Google Scholar]
  63. Root RA, 2015. Analysis of a study of lead wheel weight deposition and abrasion in New Jersey. Water, Air. Soil Pollut 226, 381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Rosin A, 2009. The long-term consequences of exposure to lead. Isr. Med. Assoc. J. IMAJ 11, 689–694. [PubMed] [Google Scholar]
  65. Sarnat JA, Brown KW, Schwartz J, Coull BA, Koutrakis P, 2005. Ambient gas concentrations and personal particulate matter exposures: implications for studying the health effects of particles. Epidemiology 385–395. [DOI] [PubMed] [Google Scholar]
  66. Schwartz J, 1991. Lead, blood pressure, and cardiovascular disease in men and women. Environ. Health Perspect 91, 71–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Schwartz PJ, Wolf S, 1978. QT interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation 57, 1074–1077. [DOI] [PubMed] [Google Scholar]
  68. Semlali RM, Dessogne J-B, Monna F, Bolte J, Azimi S, Navarro N, Denaix L, Loubet M, Chateau C, van Oort F, 2004. Modeling lead input and output in soils using lead isotopic geochemistry. Environ. Sci. Technol 38, 1513–1521. [DOI] [PubMed] [Google Scholar]
  69. Shrier I, Platt RW, 2008. Reducing bias through directed acyclic graphs. BMC Med. Res. Methodol 8, 70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Shutt RH, Kauri LM, Weichenthal S, Kumarathasan P, Vincent R, Thomson EM, Liu L, Mahmud M, Cakmak S, Dales R, 2017. Exposure to air pollution near a steel plant is associated with reduced heart rate variability: a randomised crossover study. Environ. Heal 16, 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Tong STY, 1990. Roadside dusts and soils contamination in Cincinnati, Ohio, USA. Environ. Manage 14, 107–113. [Google Scholar]
  72. U.S. Food and Drug Administration, 2005. E14 Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-antiarrhythmic Drugs. US Food Drug Adm, Washington, DC. [Google Scholar]
  73. Williams R, Brook R, Bard R, Conner T, Shin H, Burnett R, 2012. Impact of personal and ambient-level exposures to nitrogen dioxide and particulate matter on cardiovascular function. Int. J. Environ. Health Res 22, 71–91. 10.1080/09603123.2011.588437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Zanobetti A, Franklin M, Schwartz J, 2008. Fine particulate air pollution and its components in association with cause-specific emergency admissions in 26 U.S. Cities. Epidemiology 19. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES