Effects of exposure to air pollution on blood coagulation

Authors

  • A. BACCARELLI,

    1. Department of Environmental Health, Harvard School of Public Health, Boston, MA, USA
    2. Department of Preventive Medicine and EPOCA Epidemiology Research Center, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan
    Search for more papers by this author
  • A. ZANOBETTI,

    1. Department of Environmental Health, Harvard School of Public Health, Boston, MA, USA
    Search for more papers by this author
  • I. MARTINELLI,

    1. A. Bianchi Bonomi Haemophilia and Thrombosis Center, Department of Internal Medicine and Medical Specialties, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan, Italy
    Search for more papers by this author
  • P. GRILLO,

    1. Department of Preventive Medicine and EPOCA Epidemiology Research Center, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan
    Search for more papers by this author
  • L. HOU,

    1. Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD, USA
    Search for more papers by this author
  • S. GIACOMINI,

    1. Department of Preventive Medicine and EPOCA Epidemiology Research Center, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan
    Search for more papers by this author
  • M. BONZINI,

    1. Department of Preventive Medicine and EPOCA Epidemiology Research Center, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan
    Search for more papers by this author
  • G. LANZANI,

    1. Air Quality Unit, Regional Environmental Protection Agency ARPA Lombardia, Milan, Italy
    Search for more papers by this author
  • P. M. MANNUCCI,

    1. A. Bianchi Bonomi Haemophilia and Thrombosis Center, Department of Internal Medicine and Medical Specialties, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan, Italy
    Search for more papers by this author
  • P. A. BERTAZZI,

    1. Department of Preventive Medicine and EPOCA Epidemiology Research Center, University of Milan and IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milan
    Search for more papers by this author
  • J. SCHWARTZ

    1. Department of Environmental Health, Harvard School of Public Health, Boston, MA, USA
    Search for more papers by this author

Andrea Baccarelli, Exposure, Epidemiology and Risk Program, Harvard School of Public Health, 401 Park Drive, Landmark Center, Suite 412F West, P.O. Box 15698, Boston, MA 02215, USA.
Tel.: +1 617 384 8838; fax: +1 617 384 8745; e-mail: abaccare@hsph.harvard.edu

Abstract

Summary. Background: Consistent evidence has indicated that air pollution increases the risk of cardiovascular diseases. The underlying mechanisms linking air pollutants to increased cardiovascular risk are unclear. Objectives: We investigated the association between the pollution levels and changes in such global coagulation tests as the prothrombin time (PT) and the activated partial thromboplastin time (APTT) in 1218 normal subjects from the Lombardia Region, Italy. Plasma fibrinogen and naturally occurring anticoagulant proteins were also evaluated. Methods: Hourly concentrations of particulate (PM10) and gaseous pollutants (CO, NO2, SO2, and O3) were obtained from 53 monitoring sites covering the study area. Generalized additive models were applied to compute standardized regression coefficients controlled for age, gender, body mass index, smoking, alcohol, hormone use, temperature, day of the year, and long-term trends. Results: The PT became shorter with higher ambient air concentrations at the time of the study of PM10 (coefficient = −0.06; P < 0.05), CO (coefficient = −0.11; P < 0.001) and NO2 (coefficient =−0.06; P < 0.05). In the 30 days before blood sampling, the PT was also negatively associated with the average PM10 (coefficient = −0.08; P < 0.05) and NO2 (coefficient = −0.08; P < 0.05). No association was found between the APTT and air pollutant levels. In addition, no consistent relations with air pollution were found for fibrinogen, antithrombin, protein C and protein S. Conclusions: This investigation shows that air pollution is associated with changes in the global coagulation function, suggesting a tendency towards hypercoagulability after short-term exposure to air pollution. Whether these changes contribute to trigger cardiovascular events remains to be established.

Introduction

Over the past two decades, a growing body of evidence has led to a heightened concern about the potential deleterious health effects of ambient air pollution and its relation to cardiovascular disease [1,2]. Several air pollutants, including particulate matter (PM), carbon monoxide (CO), nitrogen oxides (NOx), sulfur dioxide (SO2) and ozone (O3) have been associated with increased hospitalization and mortality as a result of cardiovascular disease and stroke [1–9]. The underlying mechanisms linking air pollutants and increased cardiovascular risk remain unclear, although prior studies have associated exposure to air pollution with activation of inflammatory pathways, production of reactive oxygen species, endothelial injury and dysfunction, arterial vasoconstriction and alterations in blood coagulation factors [2,10–12]. Seaton et al. [13] proposed that inhaled pollutants induce alveolar inflammation with release of mediators capable of increasing blood coagulability. The contribution of inflammation to blood coagulation is de novo synthesis of tissue factor on leukocytes and endothelial cells stimulated by inflammatory cytokines, acute phase C-reactive protein and reactive oxygen species [14,15]. Exposure of the blood to tissue factor triggers the extrinsic coagulation pathway whose function is monitored by the prothrombin time (PT) through interaction of tissue factor with activated coagulation factor VII (FVII) [16]. The plasma levels of several coagulation factors have been investigated in epidemiological studies as potential mediators of air pollution-related hypercoagulability [9,17–20]. Coagulation factors such as FVII and fibrinogen, which are part of the acute-phase responses mediated by cytokines released during inflammatory reactions, increase after short-term exposure to particles [9,19,21]. However, whether or not air pollution exposure is associated with hypercoagulability as measured with global coagulation tests has never been determined. In this study, we examined in 1218 healthy subjects from the Lombardia Region in Italy the effects of air pollution on such global coagulation tests as the PT and the activated partial thromboplastin time (APTT) assuming that, if shortened, these tests would reflect hypercoagulability. In addition, we evaluated in the same subjects the concentration and activity of an acute phase reactant such as fibrinogen and of naturally occurring anticoagulants such as antithrombin (AT), protein C and protein S.

Material and methods

Study population and laboratory methods

From January 1995 to August 2005, 1218 healthy individuals, who were partners or friends of patients with thrombosis, attended the Thrombosis Center of the University of Milan, Italy, and did agree to undergo thrombophilia screening on a voluntary basis. Only individuals resident in the Lombardia Region were chosen. Previous thrombosis and recent use of drugs affecting coagulation was excluded with a validated structured questionnaire. All participants gave written informed consent and approval for the study was obtained from the Departmental Institutional Review Board. On the day of the visit, the participants attended the Thrombosis Center at 09:00 hours when a fasting blood sample was taken. A standardized questionnaire was administered including demographic data and questions on education, occupation, smoking, alcohol consumption, diet, reproductive history and hormone use (oral contraceptives or hormone replacement therapy). Blood was collected into vacuum tubes (Becton & Dickinson, Meylan, France) containing 3.8% trisodium citrate (ratio of blood to anticoagulant: 9:1). The PT was measured on freshly collected platelet-poor plasma with human relipidated recombinant thromboplastin (Recombiplastin; Instrumentation Laboratory, Orangeburg, NY, USA) in combination with a fully automated photo-optical coagulometer (ACL, Instrumentation Laboratory). Results were expressed as ratios of test to reference plasma coagulation times (INR, international normalized ratio). The APTT was measured with the automated APTT reagent from Organon Teknika (now bioMerieux, Durham, NC, USA) using an automated photo-optical coagulometer (ACL; Instrumentation Laboratory, Lexington, MA, USA). Results for the APTT were expressed as ratios of test to reference plasma coagulation times. Normal plasma used as reference for both the PT and APTT was prepared by pooling equal portions of platelet-poor plasmas from the blood of 30 donors (15 males and 15 females). Immunoreactive fibrinogen was measured on frozen plasma (−70 °C) by enzyme immunoassay (Dako, Glostrup, Denmark). AT and protein C were measured on frozen plasma as functional activities (Electrachrome Antithrombin, Instrumentation Laboratory; and Proclot Protein C, Instrumentation Laboratory). Immunoreactive protein C and total protein S antigen were measured by enzyme-linked immunosorbent assay (ELISA) with polyclonal antibodies (Dako). Free protein S, i.e. the fraction endowed with anticoagulant activity, was measured using the aforementioned method after PEG precipitation of the C4b-binding protein–protein S complex or directly by ELISA with specific monoclonal antibody (Asserachrom Free Protein S, Stago). During the study, the quality of laboratory results was monitored through participation in the National Interlaboratory Coagulation Survey. The laboratory participates on a regular basis to the external quality control scheme organized by the Italian Committee for Standardization of Laboratory Methods.

Air pollution and weather data

We obtained from the Regional Environmental Protection Agency (ARPA Lombardia) recordings of hourly air pollution data measured from January 1994 to September 2005 by monitors located at 53 different sites throughout Lombardia (Fig. 1A). The 53 stations included in this study were selected by the Regional Environmental Protection Agency (ARPA Lombardia) from the approximately 200 monitors of the Regional Air Monitoring Network on the basis of their reliability determined by standardized quality control procedures and by correlation with in situ measurements, of continuity of recording and of the ability to represent local background air pollution. We identified nine different study areas in the region (Fig. 1A) characterized by homogenous within-area air pollution concentrations and temporal variations. Within each study area, levels of air pollutants measured by different monitors were highly correlated. In addition, mobile monitoring in each of the study areas during the study period showed high concordance with measurements taken by the permanent monitors in the same area [22]. For each study area, mean hourly concentrations of PM with an aerodiameter of equal to or less than 10 μm (PM10) of CO, nitrogen dioxide (NO2), sulphur dioxide (SO2) and ozone (O3) were averaged using an algorithm that combined levels reported by multiple monitoring locations [23]. The Southern part of Pavia province (Fig. 1A) was excluded, because this area had no local monitoring stations and showed pollution patterns in repeated point mobile recordings that differed from those measured by stationary monitors located in neighboring areas. In addition, we obtained data on mean daily linear visibility recorded at the three major airports (Milano Malpensa, Milano Linate and Bergamo Orio al Serio), and at one meteorological station (Brescia-Ghedi) available online from the U.S. National Climatic Data Center (http://www.ncdc.noaa.gov). Linear visibility data were used to calculate the extinction coefficient, which was shown to be a good predictor of fine particle concentrations [24]. In most of the areas, total suspended particles (TSPs) rather than PM10 were measured in the earlier years of the study period (year 1995 in area 4; years 1995–1996 in area 3; years 1995–1997 in area 1 and 2; and years 1995–1998 in areas 5, 6, 7, and 9), amounting to 29% of the days during the study period. In the study areas, TSP measurements were also continued after PM10 recording was introduced. For the periods in which only TSP measurements were available, PM10 was estimated as the predicted value from a linear regression model that included PM10 as the dependent variable and, as independent variables, day of the week, wind direction (divided in eight sectors) and penalized splines of TSP (d.f. = 4), temperature (d.f. = 4), barometric pressure (d.f. = 4), relative humidity (d.f. = 4), wind intensity (d.f. = 4), extinction coefficient (d.f. = 4), hour of the day (d.f. = 8) and date (d.f. = 4 per each year in the study period). The penalized splines were used to allow for non-linear associations with PM10 concentrations. The correlation (r) between the predicted and the measured data for the time periods in which both TSP and PM10 were available ranged from 0.71 (Area 9) to 0.85 (Area 6). The mean correlation for the nine areas was equal to 0.77 (SD = 0.05). The analyses performed throughout this study were performed including the predicted data. When predicted data were excluded from the analyses, the point estimates obtained were similar to those including predicted data, but had wider confidence intervals.

Figure 1.

 Lombardy region maps showing: (A) the 53 air pollution monitors located in the nine air pollution homogenous areas identified for the study; and (B) the residence of the study subjects.

Statistical analysis

In the analysis of the association of air pollutants with blood coagulation measurements, the following variables were chosen before the analysis as relevant predictors and included in the linear regression: age, gender, body mass index (BMI), current cigarette smoking (yes/no), current alcohol consumption (yes/no), hormone use (yes/no), day of the year, long-term time trend and temperature. Penalized splines were used to account for potential non-linearity in the relationship of day on the year (d.f. = 4), long-term trend (d.f. = 3) and temperature (d.f. =4). Values of functional protein C, protein C antigen, free protein S, functional protein S, functional AT and fibrinogen were log-transformed to improve normality and stabilize the variance. Regression analyses were performed in R-version 2.2.1 (R Foundation for Statistical Computing, Vienna, Austria) using generalized additive models to evaluate the relation of blood coagulation measurements with each air pollutant. To enable comparability across outcomes, effects are expressed as standardized coefficients, expressing the fraction of an SD change in outcome associated with an SD change in exposure.

Results

Characteristics of the study population and air pollution profile in the study area

The study included 490 males and 728 females between 11 and 84 years of age (mean age = 43.6 years) (Table 1). The majority of the participants were from the Milan (Area 1, 43.2%) or suburban Milan (Area 2, 22.5%) areas (Table 1 and Fig. 1B). All the participants had normal coagulation times, with the PT INR between 0.70 and 1.31 (median = 0.98, p25–p75 = 0.95–1.03) and APTT ratios (R) between 0.54 and 1.45 (median = 1.00, p25–p75 = 0.94–1.06). Similarly, levels of the other coagulation measurements were within the normal limits (Table 1). Forty-four (3.9%) and 30 (2.6%) individuals were carriers of factor V Leiden or G20210A prothrombin mutations, respectively.

Table 1.   Characteristics of the study subjects and mean levels of the coagulation measurement investigated
VariableAll subjects, (n = 1218)
  1. *International normalized ratio. Ratio to reference plasma. Free protein S was measured only on a subset of 639 subjects.

Age, years [mean, (min–max)]43.6 (11.2–84.4)
Gender, n (%)
 Male490 (40.2%)
 Female728 (59.8%)
Body mass index, kg m−2 [mean, (min–max)]23.9 (15.6–45.2)
Smokers, n (%)344 (28.2%)
Cigarettes day−1, [mean, (min–max)]10.0 (1–60)
Alcohol, n (%)647 (53.5%)
Hormone users, n (%)168 (23.1%)
Area of residence, n (%)
 Area 1, Milan, urban area526 (43.2%)
 Area 2, Milan, suburban area274 (22.5%)
 Area 3, Bergamo and Brescia18 (1.5%)
 Area 4, Po river valley (towns > 15,000 population)74 (6.1%)
 Area 5, Po river valley (remaining territory)154 (12.6%)
 Area 6, major northern cities (Varese, Como, Lecco)24 (2.0%)
 Area 7, lower Valtellina valley29 (2.4%)
 Area 8, Alps16 (1.3%)
 Area 9, Pre-Alpine territory103 (8.5%)
Altitude at the subjects’ town of residence, meters [mean, (min–max)]202.3 (19–2580)
Coagulation measurements [median, (p25–p75)]
 Prothrombin time, (INR*)0.98 (0.95–1.03)
 Activated partial thromboplastin time, (R)1.00 (0.94–1.06)
 Fibrinogen, mg dL−1282 (247–331)
 Functional antithrombin (%)100 (94–106)
 Functional protein C (%)101 (89–116)
 Protein C antigen (%)108 (96–126)
 Functional protein S (%)106 (86–127)
 Free protein S (%)103 (88–124)
Factor V Leiden mutation carriers, n (%)30 (2.6%)
G20210A prothrombin mutation carriers, n (%)44 (3.9%)

Air pollution exposure was estimated on the basis of ambient measurements taken during the study period. Air pollution levels and weather variables at the time of blood sampling are summarized in Table 2.

Table 2.   Air pollution profile and weather variables at the time of blood sampling, by season*
Air pollutantsSeasonNumber of subjectsPercentileMax
p25Medianp75
  1. *Blood sampling was performed on all subjects at 09:00 hours between January 1st, 1995 and September 1st, 2005. Average of hourly measurements from multiple monitors located in each of the areas. Concentrations were missing for the earlier periods of the study in some of the study areas. p.p.m., parts per million; p.p.b., parts per billion.

PM10 (μg m−3)Sep.–Nov.36433.151.276.5148.9
Dec.–Feb.26947.968.595.3238.3
Mar.–May30930.064.164.8158.5
Jun.–Aug.21028.044.361.394.7
CO (p.p.m.)Sep.–Nov.4041.362.183.527.73
Dec.–Feb.2802.003.114.3111.43
Mar.–May3131.031.582.145.52
Jun.–Aug.2140.731.141.585.35
NO2 (p.p.b.)Sep.–Nov.40528.136.146.796.3
Dec.–Feb.28035.643.352.2101.8
Mar.–May31230.137.346.7116.9
Jun.–Aug.20924.232.442.592.3
SO2 (μg m−3)Sep.–Nov.4046.010.317.982.3
Dec.–Feb.2809.620.832.691.1
Mar.–May3054.78.617.567.9
Jun.–Aug.2083.65.39.226.5
O3 (p.p.b.)Sep.–Nov.4023.96.09.053.0
Dec.–Feb.2793.24.87.146.1
Mar.–May3105.39.917.856.8
Jun.–Aug.21018.223.930.957.7
Weather variables Minp25Medianp75Max
Temperature (°C)Jan.–Dec.−9.25.311.117.429.8
Barometric pressure (mmHg)Jan.–Dec.865.4994.91001.11008.01037.2
Relative humidity (%)Jan.–Dec.14.663.582.494.6100.0

Air pollutant concentrations and coagulation times

Table 3 presents the estimated mean changes of PT and APTT for various lags of air pollutants, adjusted for potential confounders. PM10, CO, and NO2 showed significant negative associations with the PT. PT values became shorter with increasing concentrations at the time of blood sampling of PM10 (coefficient = −0.06, 95% CI = −0.12 to 0.00, P < 0.05), CO (coefficient = −0.11, 95% CI = −0.18 to 0.05, P < 0.001) and NO2 (coefficient = −0.06, 95% CI =−0.12 to 0.01, P < 0.05). The association between PM10 and PT remained similar when we used as a measure of air pollution exposure the average PM10 concentrations (moving averages, MAs) between 1 and 6 h before blood sampling (Fig. 2). The effect of CO on PT was larger at the time of blood sampling and gradually decreased with longer MAs. The association between NO2 and PT was statistically significant for MAs between 1 and 5 h and decreased thereafter for longer MAs.

Table 3.   Estimated changes in prothrombin time (PT) and activated partial thromboplastin time (APTT) associated with air pollutant levels at the time of blood sampling, and in the 7 and 30 days prior to the study
 Correlation with air pollutant levels at the time of blood sampling, Coefficient (95% CI)Correlation with average air pollutant levels in the 7 days prior to the study, Coefficient (95% CI)Correlation with average air pollutant levels in the 30 days prior to the study, Coefficient (95% CI)
  1. Standardized correlation coefficients, expressing the fraction of an SD change in PT or APTT associated with an SD change in exposure, adjusted for age, gender, body mass index, cigarette smoking, alcohol consumption, oral contraceptives, and penalized smoothing splines for day of the year (d.f. = 4), long-term time trend (d.f. = 3), and temperature (d.f. = 4). *P < 0.05; **P < 0.001.

PT
 PM10−0.06 (−0.12, 0.00)*−0.03 (−0.10, 0.04)−0.08 (−0.14, −0.01)*
 CO−0.11 (−0.18, −0.05)**−0.07 (−0.14, 0.01)−0.05 (−0.13, 0.02)
 NO2−0.06 (−0.12, −0.01)*−0.07 (−0.14, 0.00)−0.08 (−0.15, 0.00)*
 SO2−0.02 (−0.08, 0.05)−0.02 (−0.09, 0.06)0.00 (−0.08, 0.07)
 O30.02 (−0.06, 0.09)0.01 (−0.11, 0.12)−0.03 (−0.15, 0.09)
Activated partial thromboplastin time (APTT)
 PM100.02 (−0.04, 0.08)0.00 (−0.07, 0.06)0.01 (−0.06, 0.08)
 CO0.03 (−0.04, 0.10)0.04 (−0.04, 0.11)0.06 (−0.01, 0.14)
 NO20.06 (0.00, 0.11)0.01 (−0.06, 0.08)0.03 (−0.04, 0.11)
 SO2−0.03 (−0.09, 0.04)−0.06 (−0.14, 0.01)−0.05 (−0.13, 0.02)
 O30.03 (−0.05, 0.11)0.00 (−0.12, 0.12)−0.03 (−0.15, 0.09)
Figure 2.

 Estimated mean changes and 95% CI of the prothrombin time (PT) associated with average PM10 concentrations recorded between zero and 6 h before blood sampling (moving averages). Standardized correlation coefficients, expressing the fraction of a SD change in PT associated with an SD change in exposure are reported. Hour zero corresponds to the estimated PT change associated with the exposure at the time of blood sampling. For the following hours, hour n corresponds to the estimated PT change associated with the average PM10 concentrations in the n hours before blood sampling.

No association between air pollutant concentrations at the time of the study and the APTT was found (Table 3). Neither the PT nor the APTT were significantly associated with air pollution levels when the effects of the average concentrations of air pollutants in the 7 days before the study were evaluated (Table 3), although we still found a non-significant negative association of the PT with PM10 and NO2. PT values were negatively and significantly associated with the average concentrations of air pollutants in the 30 days before the study of PM10 (coefficient = −0.08, 95% CI −0.14 to −0.01, P < 0.05) and NO2 (coefficient = −0.08, 95% CI −0.15 to 0.00, P < 0.05). We also evaluated whether the associations of PM10, CO, or NO2 with PT were modified by carrying any of the mutations in the Leiden V or prothrombin genes, but no significant changes were found (P > 0.05 for the tests for interaction between any of the pollutant variables and carrying either or both mutations).

Associations of air pollutant concentrations with fibrinogen and natural anticoagulant proteins

Table 4 shows that the levels of the air pollutants at the time of blood sampling were not associated with plasma fibrinogen, functional AT, functional protein C, protein C antigen, functional protein S and free protein S. Similarly, no consistent associations with the average concentrations of air pollutants in the 7 or 30 days prior to the study were found. Among the results based on the 30-day averages, a decrease in functional protein S was found in association with an interquartile change in PM10 (coefficient = −0.14, 95% CI −0.23 to −0.05; P < 0.01).

Table 4.   Estimated changes in anticoagulant proteins and fibrinogen associated with air pollutant levels at the time of blood sampling, and in the 7 and 30 days prior to the study
 Correlation with air pollutant levels at the time of blood sampling, Coefficient (95% CI)Correlation with average air pollutant levels in the 7 days prior to the study, Coefficient (95% CI)Correlation with average air pollutant levels in the 30 days prior to the study, Coefficient (95% CI)
  1. Standardized correlation coefficients expressing the fraction of an SD change in outcome associated with an SD change in exposure, adjusting for age, gender, body mass index, cigarette smoking, alcohol consumption, oral contraceptives, and penalized smoothing splines for day of the year (d.f. = 4), long-term time trend (d.f. = 3), and temperature (d.f. = 4). *P < 0.05; **P < 0.01.

Fibrinogen
 PM100.01 (−0.05, 0.07)−0.03 (−0.09, 0.04)−0.02 (−0.09, 0.05)
 CO−0.04 (−0.01, 0.03)0.02 (−0.05, 0.09)−0.02 (−0.10, 0.05)
 NO2−0.03 (−0.08, 0.03)0.00 (−0.06, 0.07)−0.01 (−0.08, 0.05)
 SO20.00 (−0.06, 0.06)0.00 (−0.07, 0.07)0.00 (−0.07, 0.07)
 O3−0.02 (−0.09, 0.06)−0.13 (−0.25, −0.02)*−0.12 (−0.24, −0.01)*
Functional antithrombin
 PM10−0.02 (−0.09, 0.04)−0.06 (−0.13, 0.01)−0.06 (−0.13, 0.02)
 CO−0.01 (−0.08, 0.06)−0.06 (−0.14, 0.02)−0.04 (−0.11, 0.05)
 NO20.00 (−0.06, 0.06)−0.04 (−0.11, 0.04)−0.01 (−0.09, 0.07)
 SO20.01 (−0.06, 0.08)0.01 (−0.07, 0.09)0.02 (−0.06, 0.10)
 O30.01 (−0.07, 0.09)0.07 (−0.05, 0.20)0.15 (0.03, 0.28)
Functional protein C
 PM100.00 (−0.06, 6.1)−0.06 (−0.12, 0.01)−0.06 (−0.14, 0.01)
 CO−0.06 (−0.12, 1.1)−0.03 (−0.10, 0.05)−0.05 (−0.12, 0.03)
 NO2−0.02 (−0.08, 3.6)−0.02 (−0.09, 0.05)−0.04 (−0.12, 0.03)
 SO20.06 (0.00, 12.5)0.10 (0.03, 0.17)**0.07 (0.00, 0.15)
 O3−0.01 (−0.09, 6.7)−0.01 (−0.13, 0.11)−0.03 (−0.16, 0.09)
Protein C, antigen
 PM100.00 (−0.06, 6.0)−0.04 (−0.10, 0.03)−0.06 (−0.14, 0.01)
 CO0.02 (−0.5, 8.5)0.01 (−0.07, 0.09)0.01 (−0.04, 0.09)
 NO20.00 (−0.05, 6.2)−0.02 (−0.09, 0.05)−0.03 (−0.11, 0.04)
 SO20.05 (−0.01, 11.4)0.03 (−0.04, 0.11)0.01 (−0.07, 0.08)
 O30.01 (−0.07, 8.6)0.06 (−0.06, 0.18)0.04 (−0.08, 0.16)
Functional protein S
 PM100.04 (−0.03, 0.10)−0.03 (−0.11, 0.06)−0.14 (−0.23, −0.05)**
 CO0.04 (−0.05, 0.13)−0.02 (−0.14, 0.011)−0.08 (−0.20, 0.05)
 NO2−0.02 (−0.09, 0.06)0.02 (−0.07, 0.012)−0.08 (−0.14, 0.06)
 SO20.02 (−0.07, 0.10)0.04 (−0.05, 0.014)0.01 (−0.09, 0.10)
 O30.06 (−0.03, 0.14)0.07 (−0.07, 0.022)0.16 (0.01, 0.32)
Free protein S
 PM100.05 (−0.01, 0.10)0.01 (−0.05, 0.07)−0.01 (−0.08, 0.06)
 CO0.00 (−0.06, 0.06)0.01 (−0.06, 0.08)0.02 (−0.06, 0.09)
 NO20.01 (−0.04, 0.06)0.06 (−0.01, 0.12)0.04 (−0.03, 0.11)
 SO2−0.01 (−0.07, 0.05)0.04 (−0.03, 0.10)0.03 (−0.04, 0.10)
 O3−0.05 (−0.12, 0.02)−0.05 (−0.16, 0.06)−0.04 (−0.15, 0.08)

Discussion

In this study conducted on a large sample of healthy individuals from the Lombardia Region in Italy, we found that air pollution levels measured in the hours preceding blood sampling were associated with shortened PT (i.e. smaller INR values). In particular, increasing concentrations of PM10, CO, and NO2 were associated with shorter PT. In addition, a similar degree of shortening of the PT was related with 30-day average PM10 and NO2 levels. No important relations with air pollution were found for the APTT, fibrinogen and the natural anticoagulant proteins.

PT measures the formation of the fibrin clot through the activity of the extrinsic and common coagulation pathways, which involve the interaction of tissue factor and activated FVII, in addition to FX and FV, prothrombin and fibrinogen [25]. The test is based on plasma recalcification in the presence of tissue factor [26]. Our finding of a mildly shortened PT in association with high concentrations of PM10, CO, and NO2 apparently reflects air pollution-related changes in blood coagulation. The PT depends on the concentrations of factors in the extrinsic (FVII) and common pathways (FX, FV, FII and fibrinogen) and is shortened in the presence of traces of thrombin or other activated factors that may be produced in hypercoagulable states [27]. Tissue factor is thought to be the primary initiator of in vivo coagulation [15,28]. In the absence of tissue factor expression, endothelial cells maintain thromboresistance. Endothelial and blood cells may be induced to express tissue factor by such a variety of substances such as interleukin-1 (IL-1), tumor necrosis factor, C-reactive protein and reactive oxygen species [14,15]. Heightened tissue factor production in inflammatory conditions may account for the shortened PT observed in healthy subjects with higher leukocyte counts [29]. Gilmour et al. [30] demonstrated that tissue factor expression is enhanced in macrophages exposed to PM10. Because air pollutants are known to elicit pulmonary and systemic inflammatory responses [1,13], perhaps pollution exposure increases the levels of mediators capable of inducing tissue factor expression, thereby generating a tendency to hypercoagulability. In contrast, the APTT was not associated with air pollution exposure in this study. APTT differs from PT in that coagulation is initiated by contact activation rather than by tissue factor. Therefore, the absence of air pollution-related changes in APTT is consistent with the hypothesis that air pollutants alter blood coagulation through the induction of tissue factor. On the other hand it cannot be excluded that PT shortening was as a result of increased levels of coagulation factors, particularly activated FVII. Such hypotheses could have been tested by using specific assays, such as activated FVII (FVIIa), thrombin generation testing, or other coagulation markers that may give evidence of increased coagulation in the circulation (e.g. TAT, F1 + F2, D-dimer). Unfortunately, because the present investigation was based on a preexisting study population, we were not able to assess the correlation of air pollution with FVIIa or those markers of coagulation activation.

This study was based on healthy individuals who had normal coagulation times, with INRs ranging from 0.70 to 1.31. We estimated that an increase by one SD in air pollution at the time of the study was associated with an average downward change in the PT from 6% to 11%, scaled to the PT SD. A similar degree of shortening was observed in women taking oral contraceptives, which are an established risk factor for thrombosis [31–33]. In our study, we found significant associations of PT with air pollutant levels measured at the time of blood sampling, as well as with the average concentrations in the 30 days before the study. Although early studies on air pollution and cardiovascular health evaluated mainly short-term effects, a number of more recent investigations have tested for the association of average exposures over longer time periods. In particular, Zanobetti et al. [34] found increased cardiovascular mortality in association with particle exposures occurring more than 1 month before the event. Recently, studies conducted on a Boston population [35] and on the Multi-Ethnic Study of Atherosclerosis [36] showed that increased inflammatory markers were associated with the average exposures to particles over the 28–60 days prior to the study. Such results suggest that, in addition to acute outcomes, air pollutants have delayed effects on inflammatory pathways and cardiovascular endpoints that are consistent with the biphasic PT changes we observed in our study.

The association of PT with PM10, CO, and NO2, but not with SO2 or O3, may indicate that traffic pollution is responsible for the changes in blood coagulability. Ambient CO and NO2 have been shown to be correlated with particles from traffic, while ambient O3 is considered a surrogate for secondary particle exposures [37]. Previous reports have shown that particles originating from traffic are more associated than other pollutants with mortality effects [4].

Experimental and epidemiological studies that evaluated plasma concentrations of coagulation factors in association with air pollution exposure have produced mixed results. Some of them showed increased levels of FVII [12], fibrinogen [9,19,21] and von Willebrand factor [18,20]. On the other hand, other studies that measured the same factors found decreased levels or no association with air pollution exposure [17,20,38–41]. Consistent with the negative results of some of the previous studies, we found no association between fibrinogen and the levels of the air pollutants. In addition, we examined for the first time the concentration and activity of the natural anticoagulants proteins, but no consistent association with air pollution levels was found.

A limitation of this study is that ambient air pollution was used as a surrogate for personal exposure, which may have resulted in a measurement error. Such a measurement error would generally tend to bias estimates toward the null [42] and may have contributed to the lack of association of air pollution found for fibrinogen and natural anticoagulants. However, the consequence of using ambient measures to estimate exposure is likely to be only a modest underestimation of pollution effects [43]. Our study was based on readings of hourly air pollution data from 53 different monitoring sites throughout the Lombardia Region that were selected on the basis of their capability to represent local background air pollution, as determined by correlation with random in situ measurements in the adjacent territory. The analysis was based on nine different pollution areas that showed spatially homogenous pollution patterns, as determined by the high longitudinal correlation of the measures from the monitoring stations in the same area and of random measurements at different within-area locations. We adjusted analyses for several potential confounders that may have influenced blood clotting (age, gender, BMI, cigarette smoking, alcohol consumption, hormone use, day of the study, and temperature). Therefore, chances that the observed associations reflected bias as a result of confounders are minimized.

In conclusion, our study found shorter PT in association with higher concentrations of ambient PM10, CO, and NO2. Air pollution levels showed no consistent association with the APTT, natural anticoagulants and fibrinogen. The observation of shortened PT may indicate that air pollution determines hypercoagulability, potentially contributing to the increase in cardiovascular events observed after exposure to air pollution.

Addendum

A. Baccarelli developed the design of the study, performed the statistical analyses, and wrote the present manuscript. A. Zanobetti participated in the definition of the study objectives, statistical analysis and interpretation of the results. I. Martinelli participated in the definition of the study objectives, supervised the recruitment of the study subjects, contributed to the interpretation of the results, and participated in the writing of the manuscript. P. Grillo managed the air pollution data and contributed to the definition of the homogenous air pollution zones. L. Hou contributed to the development of the study design and interpretation of the results. S. Giacomini contributed to the development of the study hypothesis and exposure framework. M. Bonzini contributed to the analysis of the exposure data. G. Lanzani developed the exposure categorization and supervised the selection of the monitoring stations from the Regional air pollution network. P. M. Mannucci supervised the recruitment of the study subjects, contributed to the interpretation of the results, and participated in the writing of the manuscript. P. A. Bertazzi participated in the definition of the study hypotheses and of the study design and contributed to the interpretation of the results. J. Schwartz participated in the development of the study hypotheses, contributed to the definition of study design, supervised the statistical analysis, and participated in the writing of the manuscript.

Acknowledgements

We are grateful to N. Carfagno and A. Di Leo for their invaluable support in the selection of the air monitoring stations and data extraction. Many thanks to S. Melly for his assistance in the creation of the geographic maps of the Lombardia Region.

Disclosure of Conflict of Interests

This work was funded by the Regional Government of Lombardy (Health Directorate Contract 8956/RCC), CARIPLO Foundation (Health Effects of Airborne Pollutants Project), Italian Ministry for University and Research (University System Internationalization Program, 99C/2005), and EPA PM center grant R827353.

Ancillary