Environmental Pollution xxx (2017) 1e6

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Particulate and gaseous pollutants on inflammation, thrombosis, and

autonomic imbalance in subjects at risk for cardiovascular disease*
Szu-Ying Chen a, b, Chang-Chuan Chan c, Ta-Chen Su c, d, *
a
Division of Surgical Intensive Care, Department of Critical Care Medicine, E-Da Hospital, I-Shou University, Kaohsiung, Taiwan
b
Department of Nursing, Fooyin University, Kaohsiung, Taiwan
c
Institute of Occupational Medicine and Industrial Hygiene, College of Public Health, National Taiwan University, Taipei, Taiwan
d
Department of Internal Medicine and Cardiovascular Center, National Taiwan University Hospital, Taipei, Taiwan

a r t i c l e i n f o a b s t r a c t

Article history: This study examined effects of short-term urban air pollution exposures on inflammation, thrombosis,
Received 18 May 2016 and autonomic imbalance in subjects at risk for cardiovascular disease (CVD). We enrolled 61 patients
Received in revised form with multiple CVD risk factors and measured high sensitive C-reactive protein (hs-CRP), fibrinogen, D-
14 January 2017
dimer, and heart rate variability (HRV) indices. Two health examinations for each participant were
Accepted 15 January 2017
Available online xxx
performed during December 2002 through September 2003. Changes in inflammation and thrombotic
markers and HRV indices with exposures to PM2.5, organic carbon (OC), elemental carbon (EC), sulfur
dioxide (SO2), nitrogen dioxide (NO2), and carbon monoxide (CO) at 1- to 3-day lags were analyzed using
Keywords:
Air pollution
mixed models. The results showed inflammatory and thrombotic markers increased with 1- to 3-day
Particulate matter lagged PM2.5 components and gaseous pollutants exposures. hs-CRP maximally increased 0.19 [95%
Inflammation confidence interval (CI): 0.07e0.31] and 0.15 (95% CI: 0.05e0.24) mg/L for an interquartile range (IQR) of
Thrombosis 1-day lagged SO2 (2.3 ppb) and CO (0.5 ppm), respectively. D-dimer maximally increased 1.05 (95% CI:
Heart rate variability 0.13e1.75), 0.72 (95% CI: 0.09e1.21), 0.92 (95% CI: 0.13e1.50), and 0.90 (95% CI: 0.07e1.61) mg/dL for an
Cardiovascular disease IQR of 1-day lagged OC (3.9 mg/m3), EC (2.0 mg/m3), SO2, and NO2 (13.4 ppb), respectively. The HRV
indices, including low frequency, very low frequency, and the ratio of low frequency to high frequency
decreased 19.8 (95% CI: 4.4e32.7), 12.9 (95% CI: 0.8e23.4), and 17.6 (95% CI: 5.4e28.2)% for an IQR of 1-
day lagged PM2.5 (20.2 mg/m3), respectively. Our findings demonstrated PM2.5 components and gaseous
pollutants exert prolonged inflammatory and thrombotic reactions, while PM2.5 exert an immediate
autonomic imbalance.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction of fine particulate matter (PM2.5) and chemical constituents as well

A large body of epidemiological evidence has demonstrated the
causal relationship between exposures to ambient particulate
matter (PM10) and cardiovascular disease (CVD) (Brook et al., 2010).
The PM10-related cardiovascular effects have been postulated in
association with several potential biological pathways, including
systemic inflammation, thrombotic reaction, and autonomic ner-
vous system imbalance (Brook et al., 2010; Franchini and Mannucci,
2011; Franchini et al., 2012; Nelin et al., 2012). However, the effects
*
This paper has been recommended for acceptance by David Carpenter.
* Corresponding author. Department of Internal Medicine and Cardiovascular
Center, National Taiwan University Hospital, No. 7, Chung-Shan South Rd, Taipei,
10020, Taiwan.
E-mail address: tachensu@ntu.edu.tw (T.-C. Su).
as gaseous pollutants on these hypothesized biomechanisms were
less addressed and the results were inconsistent in previous
studies. Several studies reported the positive associations between
PM2.5 and systemic inflammation markers, including C-reactive
protein (CRP) and interleukin-6 (IL-6), especially in the elderly
subjects or subjects at risk of CVD (Chuang et al., 2007; Dabass et al.,
2015; Delfino et al., 2008; Pope et al., 2004; Dubowsky et al., 2006);
whereas other studies failed to demonstrate such relationships
(Sullivan et al., 2007; Diez-Roux et al., 2006; Folino et al., 2009; Liu
et al., 2007; Pope et al., 2004; Ruckerl et al., 2007; Sullivan et al.,
2007; Zeka et al., 2006). Regarding air pollution-related thrombotic
relation, associations between PM2.5 and chemical compositions,
including black carbon, nitrate, and sulfate with increased fibrin-
ogen level were observed across several studies (Chuang et al.,
2007; Ruckerl et al., 2007; Zeka et al., 2006); however, other

http://dx.doi.org/10.1016/j.envpol.2017.01.037
0269-7491/© 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Chen, S.-Y., et al., Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance
in subjects at risk for cardiovascular disease, Environmental Pollution (2017), http://dx.doi.org/10.1016/j.envpol.2017.01.037
2 S.-Y. Chen et al. / Environmental Pollution xxx (2017) 1e6
studies did not observe the associations of PM2.5 with fibrinogen, D-
dimer, or prothrombin time, in individuals with comorbid diseases
(Delfino et al., 2008; Steinvil et al., 2008; Sullivan et al., 2007).
Although many studies reported short-term PM2.5 exposures were
associated with decreases in either in time-domain or frequency-
domain heart rate variability (HRV) indices (Adar et al., 2007;
Chuang et al., 2005, 2007; de Hartog et al., 2009; Ebelt et al.,
2005; Park et al., 2005; Pope et al., 2004; Schwartz et al., 2005).
However, other studies still failed to observe such relationships
(Dales, 2004) or even reported increased standard deviation of
normal-to-normal intervals (SDNN), root mean square of successive
differences (rMSSD), and high frequency (HF) power with PM ex-
posures (Riediker et al., 2004; Wheeler et al., 2006). The hetero-
geneity of PM2.5 components and the susceptibility of study
subjects may result in the diverse findings between studies. Be-
sides, the publication bias must be considered to inflate the re-
ported inflammation, thrombosis, and decreased HRV. Therefore,
we conducted a sub-cohort study to examine the effects of short-
term exposures to PM2.5 and chemical constituents as well as
gaseous pollutants on inflammation, thrombosis, and autonomic
dysfunction in the elder subjects at risk for CVD.
2. Methods
2.1. Study population
The study subjects were selected from 267 subjects who had
past history of physician-diagnosed coronary artery disease (CAD)
or at least two of CVD comorbid diseases, such as hypertension,
diabetes mellitus (DM), or hyperlipidemia, and participated in a
cardiovascular promotion program in 2002. Two health examina-
tions for each participant, first was performed during December,
2002 and February, 2003 and follow-up was completed during
AprileSeptember, 2003. Here, hypertension was defined as
measured systolic blood pressure
140 mm Hg and/or diastolic
blood pressure
90 mm Hg, and/or history of physician-diagnosed
hypertension and use of anti-hypertensive agents. DM was defined
as fasting glucose over 126 mg/dL and/or history of DM with
management. Hyperlipidemia was defined as total cholesterol over
240 mg/dL or triglyceride over 200 mg/dL or use of lipid-lowering
agents. After excluding the subjects who did not reside in Taipei
City, a total of 61 subjects were finally recruited in this study. This
study was approved by the Institutional Review Board of National
Taiwan University Hospital. All participants provided informed
consent upon recruitment into the study.
2.2. Health data
All of study subjects received two repeated assessments about
one year apart during the study period, including clinical interview,
self-reported questionnaire, venous blood sampling, and HRV
measurement. The participants' age, sex, body mass index (BMI),
and smoking status were obtained from the clinician interview and
self-reported questionnaire. Blood biochemistry data, including
blood sugar, total cholesterol, and triglyceride levels, were collected
following a 10-h overnight fast. Biochemistry data were analyzed in
the central laboratory of the National Taiwan University Hospital by
enzymatic methods using an automatic multi-channel chemical
analyzer (Hitachi 7450, Hitachi Corp., Tokyo, Japan). Serum high
sensitive-CRP (hs-CRP) was measured by the chemiluminescent
enzyme-labeled immunometric assay (Immulite C-Reactive Pro-
tein, Diagnostic Products Co., Los Angeles). Fibrinogen was
measured by the clotting method of Clauss, using STA®-Fibrinogen
Kits (Diagnostica, Stago). Plasma D-dimer was measured by
INNOVANCE* D-Dimer (Siemens Healthcare Diagnostics Products
Please cite this article in press as: Chen, S.-Y., et al., Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance
in subjects at risk for cardiovascular disease, Environmental Pollution (2017), http://dx.doi.org/10.1016/j.envpol.2017.01.037
GmbH, Marburg, Germany), a particle-enhanced, immunoturbidi-
metric assay.
After blood sample collection and a 15-min rest for each
participant, we immediately performed the resting electrocardio-
gram (ECG) examination, in the sitting position, during daytime
(8:00 a.m. to 12:00 p.m.) using a PacerCorder three-channel device
(model 461A; Del Mar Medical Systems LLC, Irvine, CA) with a
sampling rate of 250 Hz (4 ms). ECG tapes were processed using a
Delmar Avionics model Strata Scan 563 (Irvine, CA). A complete 5-
min segment of N-N interval was taken for HRV analysis. The time-
domain measurements of HRV included SDNN and rMSSD between
adjacent normal-to-normal intervals. The frequency-domain mea-
surements of HRV included very low frequency (VLF:
0.003e0.04 Hz), low frequency (LF: 0.04e0.15 Hz), high frequency
(HF: 0.15e0.40 Hz), and LF: HF ratio, which were calculated by
Welch's averaged periodogram of the normal-to-normal intervals.
The details of translating ECG wave complexes to HRV indices is
given in the previous study (Chuang et al., 2005).
2.3. Environmental data
The data of particulate pollutants, including PM2.5 and chemical
constituents (organic carbon (OC) and elemental carbon (EC)) and
gaseous pollutants, including sulfur dioxide (SO2), nitrogen dioxide
(NO2), and carbon monoxide (CO), were collected from a fixed-site
air quality monitoring station, Sinjhuang Supersite, which is part of
the Taiwan Environmental Protection Agency (EPA) air quality
monitoring network. The Sinjhuang Supersite Station is located in
Sinjhuang Sports Park. Toward north of the monitoring station is
the main road and the southwestern side is the green boulevard
and Sinjhuang Baseball Field. The pollution around the station is
primarily due to vehicles on workdays and normal activities on
holidays, which are representative of typical pattern of city life.
PM2.5 were measured by the tapered element oscillating micro-
balance (TEOM) 1400a monitor (R&P, Albany, NY, USA), equipped a
sample equilibrium system Nafion dryer. OC was measured by the
Rupprecht & Patashnick 5400 (R&P, Albany, NY, USA) at 340  C. EC
measurements were obtained by subtracting the OC amount from
that of total carbon. The instruments used for gaseous pollutants
were an ultraviolet fluorescence analyzer (Ecotech 9850, Blackburn,
Victoria, Australia) for SO2, a chemiluminescence analyzer (Ecotech
9841, Blackburn, Victoria, Australia) for NO2, and a non-dispersed
infrared absorption analyzer (APMA-360 CE, Irvine, CA, USA) for
CO. The scheduled quality control procedures included daily zero
and span checks, biweekly precision checks, quarterly multi-point
calibration, and data validation. The 24-h average concentrations
of PM2.5, OC, EC, SO2, and NO2 and 1-h daily maximum concen-
tration of CO 1- to 7-days preceding the health examination was
matched to each participant. To account for meteorological factors
on the outcome variables, we also collected the data of ambient
temperature and relative humidity from the Sinjhuang Supersite
Station.
2.4. Statistical analysis
Because there were two repeated measures for each of study
subject, we used generalized linear mixed-effect models with a
random intercept for study subject and a constant correlation for
each subject's two measurements to estimate the changes in
inflammation marker (hs-CRP), coagulation factors (D-dimer and
fibrinogen), and HRV indices (SDNN, rMSSD, VLF, LF, HF, and LF: HF
ratio) in response to each particulate and gaseous pollutants at 1- to
7-day lags. We applied the variable selection technique, 10%
change-in-estimate criterion, to select the potential confounding
factors. Covariates including age, sex, BMI, smoking, history of
 S.-Y. Chen et al. / Environmental Pollution xxx (2017) 1e6 3

hypertension and CAD, weekday for health examination, and 24-h Table 2
average temperature before the examination, were finally identi- Air pollution and meteorological data during the study period.

fied in sequence as adjustment variables. All of the analyses were Mean SD Minimum Maximum IQR
performed using SAS software (Version 9.1.3; SAS Institute, Cary, PM2.5 (mg/m3) 41.4 21.6 9.8 101.0 20.2
NC). The changes in biochemical parameters were expressed as the OC (mg/m3) 0.9 2.1 23.4 5.9 3.9
mean values or percent changes and 95% confidence intervals (CIs) EC (mg/m3) 2.7 2.4 0.2 17.1 2.0
for an interquartile range (IQR) increase in air pollution in different SO2 (ppb) 2.9 2.5 0.1 17.3 2.3
NO2 (ppb) 22.2 11.1 0.9 62.6 13.4
lagged days.
CO (ppm) 0.9 0.5 0.1 3.6 0.5
Temperature ( C) 22.0 5.5 11.1 33.3 3.0
3. Results Relative humidity (%) 56.7 16.2 26.7 88.5 29.4

3.1. Characteristics of study subjects

mg/L for an IQR of SO2 (2.3 ppb) and CO (0.5 ppm) at 1-day lag,
The basic characters of 61 study subjects are summarized in respectively. For coagulation factors, the fibrinogen levels increased
Table 1. All participants in the study were aged between 45 and 88 with exposures to CO at 1- to 3-day lags, with the greatest increase
years, the mean age was 62.7 ± 9.3 years and 54.1% were males. to 13.68 (95% CI: 0.37, 27.72) mg/dL for an IQR increase in CO
Among participants, 32 subjects (52.4%) with a BMI greater than (0.4 ppm) at 1-day lag. D-dimer levels increased with exposures to
25 kg/m2 and 32.8% were current smokers. 67.2%, 50.8%, 45.9%, and OC, EC, SO2, and NO2, with the greatest increase in 1.05 (95% CI:
31.1% of study subjects had a history of hypertension, hyperlipid- 0.13, 1.75), 0.72 (95% CI: 0.09, 1.21), 0.92 (95% CI: 0.13, 1.50), and 0.90
emia, DM, and CAD, respectively. The mean values of blood markers (95% CI: 0.07, 1.61) mg/dL for an IQR increase in OC (3.9 mg/m3), EC
for the study subjects over the study period were 0.44 ± 0.66 mg/L (2.0 mg/m3), SO2 (2.3 ppb), and NO2 (13.4 ppb) at 1-day lag,
for hs-CRP, 369.6 ± 65.7 mg/dL for fibrinogen, and 4.8 ± 3.5 mg/dL respectively. The changes of hs-CRP, fibrinogen, or D-dimer were all
for D-dimer. insignificant with exposures to air pollution at 4-day to 7-day lags.
The results were shown in the Supplemental Table 2.
3.2. Environmental data
3.4. Heart rate variability
Table 2 shows air pollution and meteorological data during the
study period. The averaged concentrations of SO2, NO2, and CO Table 4 shows the percent changes on HRV indices with expo-
were below the accepted National Ambient Air Quality Standards of sures to air pollution at 1- to 3-day lags. The frequency-domain of
the Taiwan EPA. However, the range of PM2.5 was relatively wide HRV indices, VLF, LF and LF/HF ratio, decreased 12.9% (95% CI:
and the daily average concentration was higher than 35 mg/m3, the 0.8e23.4%), 19.8% (95% CI: 4.4e32.7%), and 17.6% (95% CI:
National Ambient Air Quality Standards of Taiwan for 24-hr average 5.4e28.2%) for an IQR increase of PM2.5 (20.2 mg/m3) at 1-day lag,
concentration of PM2.5. The descriptive statistics of 3-day averages respectively. The time-domain of HRV index, rMSSD, borderline
of air pollution and meteorological conditions prior to health ex- decreased 7.8% (95% CI: 0.1e15.1%) with exposures to PM2.5 at 1-
amination for each of study subjects were shown in Supplemental day lag. HRV indices did not changed by exposures to PM2.5
Table 1. We found that the study subjects exposed higher PM2.5, OC, chemical components or gaseous pollutants.
and EC concentrations and lower ambient temperature and relative
humidity in the first examination compared with the second
4. Discussion
examination.
4.1. Principal findings and comparison with previous research
3.3. Inflammatory and thrombotic reactions
This study takes the advantage of repeated measurements of
Table 3 summaries the estimated changes and 95% CI on health and exposure data to better demonstrate of the casual
inflammation and coagulation markers with exposures to air relationship between exposures and health outcomes. The study
pollution at 1- to 3-day lags. The hs-CRP levels increased with ex- results suggest that urban particulate and gaseous pollutants may
posures to SO2 and CO at 1- to 3-day lags, with the greatest in- jointly impact cardiovascular health in subjects with CVD risk fac-
creases of 0.19 (95% CI: 0.07, 0.31) mg/L and 0.15 (95% CI: 0.05, 0.04) tors through multi-biological pathways. Our findings indicate that
both particulate chemical components and gaseous pollutants,
including OC, EC, SO2, NO2, and CO, may trigger prolonged in-
Table 1
Summary of basic characters of 61 study subjects.
flammatory and thrombotic responses for 3 days in vulnerable
subjects, while fine particulates immediately disturb the autonomic
Characteristics All subjects (N ¼ 61)
balance for 1 day. The estimates of response to air pollution
Age, years 62.9 ± 9.3 observed in vulnerable subjects are stronger than that reported by
BMI, kg/m2 25.6 ± 3.3 Chuang et al. (2007) which focused on young healthy college
Fasting sugar, mg/dL 101.3 ± 19.8
Triglyceride, mg/dL 166.4 ± 117.3
students.
Total cholesterol, mg/dL 211.6 ± 53.6 Studies on associations between gaseous pollutants and sys-
hs-CRP, mg/L 0.44 ± 0.66 temic inflammation were rarely addressed. Only a cross-sectional
Fibrinogen, mg/dL 369.6 ± 65.7 study observed SO2 was associated with increased CRP levels
D-dimer, mg/dL 4.8 ± 3.5
(Khafaie et al., 2013). This study reconfirmed hs-CRP levels
Male sex, n (%) 33 (54.1)
Current smoking, n (%) 29 (32.8) increased with short-term exposures to gaseous pollutants of SO2
Hypertension, n (%) 41 (67.2) and CO in susceptible subjects. However, we did not find changes in
Diabetes, n (%) 28 (45.9) hs-CRP levels with exposures to PM2.5 or chemical constituents.
Hyperlipidemia, n (%) 31 (50.8) Epidemiological and human controlled studies also reported
Coronary artery disease, n (%) 19 (31.1)
inconsistent findings in CRP levels in response to PM. Some studies

Please cite this article in press as: Chen, S.-Y., et al., Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance
in subjects at risk for cardiovascular disease, Environmental Pollution (2017), http://dx.doi.org/10.1016/j.envpol.2017.01.037
4 S.-Y. Chen et al. / Environmental Pollution xxx (2017) 1e6

Table 3
Estimated changes and 95% CI in inflammatory and coagulation markers for an interquartile range increase of air pollution at 1- to 3-day lags.

Pollutant Lag days hs-CRP (mg/L) Fibrinogen (mg/dL) D-dimer (mg/dL)

PM2.5 1-day lag 0.04 (0.15, 0.07) 3.33 (16.63, 9.61) 0.32 (0.08, 0.85)
2-day lag 0.07 (0.18, 0.04) 1.85 (14.04, 10.35) 0.42 (0.03, 0.93)
3-day lag 0.09 (0.26, 0.08) 1.48 (18.85, 15.89) 0.65 (0.03, 1.36)
OC 1-day lag 0.04 (0.19, 0.10) 5.91 (11.09, 22.92) 1.05 (0.13, 1.75)
2-day lag 0.08 (0.22, 0.07) 10.72 (8.13, 29.57) 0.85 (0.03, 1.61)
3-day lag 0.06 (0.17, 0.05) 12.57 (0.74, 26.24) 0.42 (0.05, 0.98)
EC 1-day lag 0.00 (0.11, 0.11) 2.96 (8.87, 14.78) 0.72 (0.09, 1.21)
2-day lag 0.02 (0.12, 0.08) 6.28 (4.44, 17.00) 0.45 (0.00, 0.88)
3-day lag 0.02 (0.12, 0.08) 10.35 (0.74, 21.44) 0.29 (0.07, 0.76)
SO2 1-day lag 0.19 (0.07, 0.31) 5.17 (8.87, 19.22) 0.92 (0.13, 1.50)
2-day lag 0.14 (0.06, 0.21) 2.59 (6.65, 11.83) 0.63 (0.10, 1.01)
3-day lag 0.14 (0.08, 0.21) 4.44 (3.70, 12.57) 0.45 (0.04, 0.81)
NO2 1-day lag 0.06 (0.08, 0.19) 14.41 (4.44, 33.26) 0.90 (0.07, 1.61)
2-day lag 0.04 (0.04, 0.12) 8.87 (2.22, 19.96) 0.56 (0.05, 0.99)
3-day lag 0.03 (0.04, 0.11) 7.02 (2.59, 17.00) 0.43 (0.02, 0.81)
CO 1-day lag 0.15 (0.05, 0.24) 13.68 (0.37, 27.72) 0.33 (0.12, 0.98)
2-day lag 0.13 (0.05, 0.20) 12.57 (1.11, 23.65) 0.36 (0.06, 0.88)
3-day lag 0.12 (0.06, 0.19) 10.35 (1.11, 19.59) 0.33 (0.05, 0.78)

The statistics were calculated by generalized linear mixed-effect models adjusting for age, sex, BMI, smoking, history of hypertension and CAD, weekday for health exami-
nation, and daily-average temperature and relative humidity before the examination.

Table 4
Percent changes and 95% CI in heart rate variability indices for an interquartile range increase of air pollution at 1- to 3-day lags.

Pollutant Day lags SDNN rMSSD VLF LF HF LF:HF ratio

% (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI)

PM2.5 1-day lag 2.2 (8.0, 3.9) 7.8 (15.1, 0.1) 12.9 (23.4, 0.8) 19.8 (32.7, 4.4) 10.0 (3.5, 25.3) 17.6 (28.2, 5.4)
2-day lag 0.9 (5.3, 7.6) 3.3 (11.5, 5.6) 5.1 (17.5, 9.2) 11.5 (26.6, 6.7) 8.3 (5.5, 24.1) 10.2 (22.6, 4.3)
3-day lag 4.0 (5.1, 13.9) 1.3 (13.1, 12.1) 0.2 (18.4, 22.2) 9.4 (30.7, 18.6) 10.3 (9.2, 34.0) 6.9 (24.9, 15.5)
OC 1-day lag 2.6 (6.5, 12.6) 3.4 (15.2, 10.1) 4.3 (15, 28.0) 13.7 (34.6, 13.9) 23.3 (0.2, 51.6) 8.5 (26.7, 14.1)
2-day lag 8.3 (0.7, 18.2) 3.3 (8.7, 16.9) 20.8 (0.5, 46.6) 2.9 (21.2, 34.4) 16.8 (3.9, 41.8) 7.8 (12.9, 33.4)
3-day lag 6.2 (0.2, 13.0) 0.0 (8.4, 9.2) 8.7 (5.5, 25.1) 4.4 (20.9, 15.6) 13.2 (1.2, 29.6) 0.4 (13.8, 17.0)
EC 1-day lag 1.2 (5.5, 8.3) 0.2 (9.0, 10.3) 5.3 (9.1, 22.0) 3.5 (20.9, 17.8) 10.2 (5.1, 28.1) 0.5 (15.0, 16.5)
2-day lag 4.8 (0.6, 10.5) 3.3 (4.2, 11.4) 11.9 (0.4, 25.7) 5 (10.5, 23.2) 5.7 (6.1, 19.0) 6.8 (5.9, 21.2)
3-day lag 5.1 (0.7, 11.3) 0.0 (7.8, 8.4) 6.3 (6.4, 20.8) 3.4 (18.6, 14.7) 9.0 (3.6, 23.4) 0.8 (12.2, 15.7)
SO2 1-day lag 2.8 (9.5, 4.4) 3.7 (13.0, 6.5) 8.2 (21.3, 7.1) 5.1 (23.1, 17.2) 0.6 (15.2, 16.6) 7.2 (21.4, 9.6)
2-day lag 1.8 (6.2, 2.8) 1.3 (7.4, 5.3) 1.8 (11.2, 8.6) 0.4 (12.3, 14.8) 0.6 (10.1, 10.0) 1.5 (11.5, 9.7)
3-day lag 1.1 (5.3, 3.4) 0.0 (6.0, 6.4) 0.6 (9.8, 9.5) 2.6 (9.8, 16.8) 1.5 (10.6, 8.6) 0.7 (9.2, 11.6)
NO2 1-day lag 3.1 (11.7, 6.4) 8.1 (19.2, 4.7) 7.6 (24.9, 13.7) 9.7 (32.1, 20.1) 6.9 (14, 33.0) 6.7 (25.7, 17.1)
2-day lag 0.7 (6.3, 5.2) 4.3 (11.8, 3.8) 0.9 (13, 12.9) 1.3 (17.3, 17.9) 5.3 (8.0, 20.4) 0.2 (13.0, 15.4)
3-day lag 0.3 (5.3, 4.9) 3.9 (10.6, 3.2) 1.1 (11.8, 10.9) 0.2 (14.2, 17.1) 4.0 (7.6, 17.0) 1.3 (10.5, 14.7)
CO 1-day lag 6.4 (13.1, 0.8) 7.6 (16.9, 2.8) 11.4 (24.9, 4.5) 8.8 (27.5, 14.7) 0.3 (16.5, 20.4) 8.9 (24, 9.2)
2-day lag 3.7 (9.1, 2.0) 4.4 (12, 3.8) 4.0 (15.6, 9.2) 0.6 (16.6, 18.6) 0.6 (13.6, 14.3) 1.9 (14.7, 12.8)
3-day lag 3.3 (7.9, 1.5) 3.4 (9.8, 3.6) 2.9 (12.8, 8.2) 1.7 (12.1, 17.6) 1.7 (12.4, 10.2) 0.0 (10.9, 12.3)

The statistics are calculated by mixed models adjusting for age, sex, BMI, smoking, history of hypertension and CAD, weekday for health examination, and daily-average
temperature and relative humidity before the examination.

reported the positive associations between PM and CRP (Chuang
et al., 2007; Delfino et al., 2008; Dubowsky et al., 2006; Huttunen
et al., 2012; Kelishadi et al., 2009; Zhao et al., 2013); whereas
other studies failed to demonstrate such relationships (Diez-Roux
et al., 2006; Liu et al., 2007; Pope et al., 2004; Ruckerl et al.,
2007; Steinvil et al., 2008; Sullivan et al., 2007; Tsai et al., 2012;
Wu et al., 2012; Zeka et al., 2006). A systemic review suggested
that elevated CRP levels were only found among children and
healthy adults with exposures to higher levels of PM; however, PM-
induced CRP responses were not consistently found in adults with
chronic inflammatory conditions (Li et al., 2012). Further studies
are needed to better quantify the changes in CRP level, as well as
other pro-inflammatory markers, in response to short-term air
pollution between different study populations.
The thromboembolic reaction has been postulated the final
pathway to trigger acute cardiovascular events (Brook et al., 2010;
Franchini and Mannucci, 2011; Nelin et al., 2012). Previous
studies demonstrated that short-term exposures to air pollution
are associated with increased in fibrinogen (Chuang et al., 2007;
Ruckerl et al., 2007; Zeka et al., 2006). However, D-dimer, a sur-
rogate marker of fibrin turnover and predictor of cardiovascular
morbidity and mortality (Alehagen et al., 2004; Sadanaga et al.,
2010), was rarely addressed in relationships between air pollution
and cardiovascular effects. Two panel studies reported insignificant
associations between D-dimer and PM constituents in elderly and
subjects with chronic obstructive pulmonary disease (Delfino et al.,
2008; Sullivan et al., 2007). Our study results demonstrated that
short-term CO exposures increased fibrinogen concentrations. OC,
EC, SO2, and NO2 were found to increase concentrations of D-dimer
among the subjects with multiple CVD risk factors, and the esti-
mates in D-dimer were higher than fibrinogen with exposures to air
pollution.
Likewise, reduced HRV has been demonstrated to associate with
an increased risk of sudden cardiac events (Camm et al., 2004; La
Rovere et al., 1998; Nolan et al., 1998). In this study, PM2.5 was
observed to primarily decrease a variety of frequency-domain HRV
indices, including VLF, LF and LF/HF. Our findings agreed with those
investigations of decreases in vago-sympathetic balance in humans

Please cite this article in press as: Chen, S.-Y., et al., Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance
in subjects at risk for cardiovascular disease, Environmental Pollution (2017), http://dx.doi.org/10.1016/j.envpol.2017.01.037
 S.-Y. Chen et al. / Environmental Pollution xxx (2017) 1e6
with exposures to particulate air pollution (Gold et al., 2000; Liao
et al., 2004; Luttmann-Gibson et al., 2010; Park et al., 2005; Pope
et al., 2004; Schwartz et al., 2005). Additionally, we first introduced
the VLF to evaluate the effects of short-term exposures to air
pollution on autonomic dysfunction and the results showed a
decrease in VLF with exposures to PM2.5 at 1-day lag. VLF has been
recognized as a strong predictor for cardiac events independently
of LF (Hadase et al., 2004; Nolan et al., 1998). Regarding time-
domain HRV indices, only rMSSD was found to borderline
decrease with exposures to PM2.5. Our study suggest that
frequency-domain HRV indices, which are capable of describing the
autonomic contribution to cardiac oscillation more accurately than
time-domain measures obtained from recordings of different du-
rations (Task force of the European Society of Cardiology and the
North American Society of Pacing and Electrophysiology. 1996),
may be an ideal tool in evaluation of the effects of air pollution on
autonomic nervous system among susceptible populations.
We found that PM2.5 resulted in autonomic imbalance after 1-
day exposure, while PM2.5 chemical compositions, OC and EC, and
gaseous pollutants triggered thrombogenicity and the effects lasted
to 1- to 3-days after exposures. The results may implicate that the
physical and chemical properties of PM2.5 may adversely affect
cardiovascular effects with different biological pathway and time
durations. PM2.5 mass directly deposit in pulmonary alveoli that
directly stimuli the extra-pulmonary reflex and alter HRV imme-
diately. But PM chemical compositions and gaseous pollutants can
further dissolve into the blood and cause a prolonged hypercoag-
ulable and hypofibrinolytic status (Franchini and Mannucci, 2011).
OC and EC are two major urban air pollutants that are emitted
primarily from vehicles and account for significant portions of PM
in the Taipei metropolitan area (Chen et al., 2001). OC and EC also
contributed a large amount of black carbon components of PM2.5
absorbance which had been significantly associated with a higher
risk of subclinical atherosclerosis in middle-aged residents of
Taiwan (Su et al., 2015) and increased diastolic BP in elderly resi-
dents of Taipei city (Chen et al., 2015). SO2 and CO are secondary
gaseous pollutants that mostly coming from the major hazardous
component of traffic-related air pollutants in Taipei metropolis
(Chuang et al., 2007). Our findings imply that traffic-emitted air
pollution may play an important role in increasing the thrombo-
genicity for patients at risk of CVD.
4.2. Study limitations
Several limitations should be addressed. First, we used a single
monitoring station to estimate personal exposures to air pollutants,
which may result in a potential measurement error. The estimates
of the health effects maybe toward to null via a nonsystematic
exposure misclassification (Zeger et al., 2000). Therefore, the
magnitude of effects upon exposures to urban air pollution ob-
tained from regression modeling may be smaller than the actual
impact. Second, lack of each participant's time-activity information
during study period, hence to what extent of outdoor and indoor air
pollution exposures for study subjects was not available in this
study, which may also results in exposure misclassification. How-
ever, Meier et al. (2015) demonstrated that outdoor air pollution
can be an important determinant of indoor air pollution in certain
architectural style. Third, the possible confounding effects of
medication for CVD were not considered because of lack of infor-
mation of medication in this study. Some classic anti-hypertensive
drugs, such as b-blockers or angiotensin receptor blockers were
reported to associate with the attenuation of HRV (Izzo et al., 2015).
Fourth, because lacking a group of healthy subjects for comparison,
we cannot attribute the changes observed to the CVD.
Please cite this article in press as: Chen, S.-Y., et al., Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance
in subjects at risk for cardiovascular disease, Environmental Pollution (2017), http://dx.doi.org/10.1016/j.envpol.2017.01.037
5
4.3. Conclusion
In summary, our findings demonstrate that short-term exposure
to urban particulate and gaseous pollution jointly exert multiple
biological pathways, including systemic inflammation, thrombotic
factors activation, and autonomic imbalance, among high risk CVD
subjects.
Contributions
Conceived and designed the study: SYC, TCS, and CCC. Estab-
lished and provided the cohort: TCS. Analyzed and interpreted the
data: SYC, TCS, and CCC. Wrote the paper: SYC and TCS.
Acknowledgement
This study was supported by grants from National Science
Council, Taiwan (NSC97-2923-1-002-001-MY4, NSC 100-2314-B-
002-151-MY3, and NSC 101-2314-B-002-184-NY3), and the grant
from National Taiwan University (NTU-CESRP-102R7620) also
partially supported this study.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.envpol.2017.01.037.
References
Adar, S.D., Gold, D.R., Coull, B.A., et al., 2007. Focused exposures to airborne traffic
particles and heart rate variability in the elderly. Epidemiology 18, 95e103.
Alehagen, U., Dahlstrom, U., Lindahl, T.L., 2004. Elevated D-dimer level is an inde-
pendent risk factor for cardiovascular death in out-patients with symptoms
compatible with heart failure. Thromb. Haemostat. 92, 1250e1258.
Brook, R.D., Rajagopalan, S., Pope 3rd, C.A., et al., 2010. Particulate matter air
pollution and cardiovascular disease: an update to the scientific statement from
the american heart association. Circulation 121, 2331e2378.
Camm, A.J., Pratt, C.M., Schwartz, P.J., et al., 2004. Mortality in patients after a recent
myocardial infarction: a randomized, placebo-controlled trial of azimilide using
heart rate variability for risk stratification. Circulation 109, 990e996.
Chen, J.L., Lee, C.T., Chang, S.Y., et al., 2001. The elemental contents and fractal di-
mensions of PM2.5 in Taipei City. Aerosol. Air Qual. Res. 1, 9e20.
Chen, S.Y., Wu, C.F., Lee, J.H., et al., 2015. Associations between long-term air
pollutant exposures and blood pressure in elderly residents of Taipei city: a
cross-sectional study. Environ. Health Perspect. 123, 779e784.
Chuang, K.J., Chan, C.C., Chen, N.T., et al., 2005. Effects of particle size fractions on
reducing heart rate variability in cardiac and hypertensive patients. Environ.
Health Perspect. 113, 1693e1697.
Chuang, K.J., Chan, C.C., Su, T.C., et al., 2007. The effect of urban air pollution on
inflammation, oxidative stress, coagulation, and autonomic dysfunction in
young adults. Am. J. Respir. Crit. Care Med. 176, 370e376.
Dabass, A., Talbott, E.O., Venkat, A., et al., 2015. Association of exposure to partic-
ulate matter (PM) air pollution and biomarkers of cardiovascular disease risk in
adult NHANES participants (2001-2008). Int. J. Hyg. Environ. Health. http://
dx.doi.org/10.1016/j.ijheh.2015.12.002.
Dales, R., 2004. Ambient carbon monoxide may influence heart rate variability in
subjects with coronary artery disease. J. Occup. Environ. Med. 46, 1217e1221.
de Hartog, J.J., Lanki, T., Timonen, K.L., et al., 2009. Associations between PM2.5 and
heart rate variability are modified by particle composition and beta-blocker use
in patients with coronary heart disease. Environ. Health Perspect. 117, 105e111.
Delfino, R.J., Staimer, N., Tjoa, T., et al., 2008. Circulating biomarkers of inflamma-
tion, antioxidant activity, and platelet activation are associated with primary
combustion aerosols in subjects with coronary artery disease. Environ. Health
Perspect. 116, 898e906.
Diez-Roux, A.V., Auchincloss, A.H., Astor, B., et al., 2006. Recent exposure to par-
ticulate matter and c-reactive protein concentration in the multi-ethnic study
of atherosclerosis. Am. J. Epidemiol. 164, 437e448.
Dubowsky, S.D., Suh, H., Schwartz, J., et al., 2006. Diabetes, obesity, and hyperten-
sion may enhance associations between air pollution and markers of systemic
inflammation. Environ. Health Perspect. 114, 992e998.
Ebelt, S.T., Wilson, W.E., Brauer, M., 2005. Exposure to ambient and nonambient
components of particulate matter: a comparison of health effects. Epidemiology
16, 396e405.
Folino, A.F., Scapellato, M.L., Canova, C., et al., 2009. Individual exposure to partic-
ulate matter and the short-term arrhythmic and autonomic profiles in patients
with myocardial infarction. Eur. Heart J. 30, 1614e1620.
6 S.-Y. Chen et al. / Environmental Pollution xxx (2017) 1e6
Franchini, M., Mannucci, P.M., 2011. Thrombogenicity and cardiovascular effects of
ambient air pollution. Blood 118, 2405e2412.
Franchini, M., Guida, A., Tufano, A., et al., 2012. Air pollution, vascular disease and
thrombosis: linking clinical data and pathogenic mechanisms. J. Thromb.
Haemostat. 10, 2438e2451.
Gold, D.R., Litonjua, A., Schwartz, J., et al., 2000. Ambient pollution and heart rate
variability. Circulation 101, 1267e1273.
Hadase, M., Azuma, A., Zen, K., et al., 2004. Very low frequency power of heart rate
variability is a powerful predictor of clinical prognosis in patients with
congestive heart failure. Circ. J. 68, 343e347.
Huttunen, K., Siponen, T., Salonen, I., et al., 2012. Low-level exposure to ambient
particulate matter is associated with systemic inflammation in ischemic heart
disease patients. Environ. Res. 116, 44e51.
Izzo, J.L., Khan, S.U., Saleem, O., et al., 2015. Ambulatory 24-hour cardiac oxygen
consumption and blood pressure-heart rate variability: effects of nebivolol and
valsartan alone and in combination. J. Am. Soc. Hypertens. 9, 526e535.
Kelishadi, R., Mirghaffari, N., Poursafa, P., et al., 2009. Lifestyle and environmental
factors associated with inflammation, oxidative stress and insulin resistance in
children. Atherosclerosis 203, 311e319.
Khafaie, M.A., Salvi, S.S., Ojha, A., et al., 2013. Systemic inflammation (c-reactive
protein) in type 2 diabetic patients is associated with ambient air pollution in
pune city, India. Diabetes Care 36, 625e630.
La Rovere, M.T., Bigger Jr., J.T., Marcus, F.I., et al., 1998. Baroreflex sensitivity and
heart-rate variability in prediction of total cardiac mortality after myocardial
infarction. ATRAMI (autonomic tone and reflexes after myocardial infarction)
investigators. Lancet 351, 478e484.
Li, Y., Rittenhouse-Olson, K., Scheider, W.L., et al., 2012. Effect of particulate matter
air pollution on c-reactive protein: a review of epidemiologic studies. Rev.
Environ. Health 27, 133e149.
Liao, D., Duan, Y., Whitsel, E.A., et al., 2004. Association of higher levels of ambient
criteria pollutants with impaired cardiac autonomic control: a population-
based study. Am. J. Epidemiol. 159, 768e777.
Liu, L., Ruddy, T.D., Dalipaj, M., et al., 2007. Influence of personal exposure to par-
ticulate air pollution on cardiovascular physiology and biomarkers of inflam-
mation and oxidative stress in subjects with diabetes. J. Occup. Environ. Med.
49, 258e265.
Luttmann-Gibson, H., Suh, H.H., Coull, B.A., et al., 2010. Systemic inflammation,
heart rate variability and air pollution in a cohort of senior adults. Occup. En-
viron. Med. 67, 625e630.
Meier, R., Schindler, C., Eeftens, M., et al., 2015. Modeling indoor air pollution of
outdoor origin in homes of SAPALDIA subjects in Switzerland. Environ. Int. 82,
85e91.
Nelin, T.D., Joseph, A.M., Gorr, M.W., et al., 2012. Direct and indirect effects of
particulate matter on the cardiovascular system. Toxicol. Lett. 208, 293e299.
Nolan, J., Batin, P.D., Andrew, s R., et al., 1998. Prospective study of heart rate
variability and mortality in chronic heart failure: results of the United Kingdom
heart failure evaluation and assessment of risk trial. Circulation 98, 1510e1516.
Please cite this article in press as: Chen, S.-Y., et al., Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance
in subjects at risk for cardiovascular disease, Environmental Pollution (2017), http://dx.doi.org/10.1016/j.envpol.2017.01.037
Park, S.K., O'Neill, M.S., Vokonas, P.S., et al., 2005. Effects of air pollution on heart
rate variability: the VA normative aging study. Environ. Health Perspect. 113,
304e309.
Pope, C.A., Hansen, M.L., Long, R.W., et al., 2004. Ambient particulate air pollution,
heart rate variability, and blood markers of inflammation in a panel of elderly
subjects. Environ. Health Perspect. 112, 339e345.
Riediker, M., Cascio, W.E., Griggs, T.R., et al., 2004. Particulate matter exposure in
cars is associated with cardiovascular effects in healthy young men. Am. J.
Respir. Crit. Care Med. 169, 934e940.
Ruckerl, R., Greven, S., Ljungman, P., et al., 2007. Air pollution and inflammation
(interleukin-6, c-reactive protein, fibrinogen) in myocardial infarction survi-
vors. Environ. Health Perspect. 115, 1072e1080.
Sadanaga, T., Sadanaga, M., Ogawa, S., 2010. Evidence that D-dimer levels predict
subsequent thromboembolic and cardiovascular events in patients with atrial
fibrillation during oral anticoagulant therapy. J. Am. Coll. Cardiol. 55,
2225e2231.
Schwartz, J., Litonjua, A., Suh, H., et al., 2005. Traffic related pollution and heart rate
variability in a panel of elderly subjects. Thorax 60, 455e461.
Steinvil, A., Kordova-Biezuner, L., Shapira, I., et al., 2008. Short-term exposure to air
pollution and inflammation-sensitive biomarkers. Environ. Res. 106, 51e61.
Su, T.C., Hwang, J.J., Shen, Y.C., et al., 2015. Carotid intima-media thickness and long-
term exposure to traffic-related air pollution in middle-aged residents of
Taiwan: a cross-sectional study. Environ. Health Perspect. 123, 773e778.
Sullivan, J.H., Hubbard, R., Liu, S.L., et al., 2007. A community study of the effect of
particulate matter on blood measures of inflammation and thrombosis in an
elderly population. Environ. Health 6, 3.
Task force of the European Society of Cardiology and the North American Society of
Pacing and Electrophysiology, 1996. Heart rate variability: Standards of mea-
surement, physiological interpretation and clinical use. Circulation 93,
1043e1065.
Tsai, D.H., Amyai, N., Marques-Vidal, P., et al., 2012. Effects of particulate matter on
inflammatory markers in the general adult population. Part. Fibre. Toxicol. 9, 24.
Wheeler, A., Zanobetti, A., Gold, D.R., et al., 2006. The relationship between ambient
air pollution and heart rate variability differs for individuals with heart and
pulmonary disease. Environ. Health Perspect. 114, 560e566.
Wu, S., Deng, F., Wei, H., et al., 2012. Chemical constituents of ambient particulate
air pollution and biomarkers of inflammation, coagulation and homocysteine in
healthy adults: a prospective panel study. Part. Fibre. Toxicol. 9, 49.
Zeger, S.L., Thomas, D., Dominici, F., et al., 2000. Exposure measurement error in
time-series studies of air pollution: concepts and consequences. Environ.
Health Perspect. 108, 419e426.
Zeka, A., Sullivan, J.R., Vokonas, P.S., et al., 2006. Inflammatory markers and par-
ticulate air pollution: characterizing the pathway to disease. Int. J. Epidemiol.
35, 1347e1354.
Zhao, J., Gao, Z., Tian, Z., et al., 2013. The biological effects of individual-level PM2.5
exposure on systemic immunity and inflammatory response in traffic police-
men. Occup. Environ. Med. 70, 426e431.