Final Report: Vascular Response to Traffic-Derived Inhalation in Humans

EPA Grant Number: R834796C004
Subproject: this is subproject number 004 , established and managed by the Center Director under grant R834796
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).

Center: University of Washington Center for Clean Air Research
Center Director: Vedal, Sverre
Title: Vascular Response to Traffic-Derived Inhalation in Humans
Investigators: Kaufman, Joel D. , Larson, Timothy V.
Institution: University of Washington
EPA Project Officer: Callan, Richard
Project Period: December 1, 2010 through November 30, 2015 (Extended to November 30, 2017)
RFA: Clean Air Research Centers (2009) RFA Text |  Recipients Lists
Research Category: Health Effects , Air


Project 4 examined the acute vascular effects of commute traffic exhaust exposures in human subjects in a multipollutant context. This double-blind, randomized, controlled crossover trial tested whether traffic-derived mixed pollution atmospheres of diesel exhaust and gasoline engine exhaust, experienced through travel on roadways in a passenger car, caused an increased vascular response (brachial artery vasoconstriction, increased blood pressure, reduced retinal arteriolar diameter) compared with filtered air (FA) in healthy subjects.

Summary/Accomplishments (Outputs/Outcomes):

A study vehicle was prepared and pilot-tested to confirm the effectiveness of the filter intervention in reducing in-vehicle particle counts and black carbon concentrations.

Project 4 was launched in Year 4 of the Center (2014). In this project, we used a “typical commute” study design and pertinent experience in human exposure studies to advance the Center'’s research agenda with a double-blind, exposure crossover clinical trial in 16 subjects, randomized to order. Using an innovative approach in which contrasts of in-vehicle exposure and potential participant susceptibility by genotype are nested in the experiment, we addressed several hypotheses in the study. Building on our prior work, we used this typical commute model to determine whether traffic-derived aerosols (e.g., mixed on-road environment with diesel and gasoline engine exhaust components) exert demonstrable and important acute vascular effects in human subjects, and whether traffic-derived aerosols acutely induced increased lipid peroxidation, response to oxidized phospholipids, and resulted in measurable impacts on gene expression and DNA methylation in pathways that are related not only to the triggering of acute cardiovascular events, but also to the development and progression of atherosclerosis. All of the outcomes we measured are transient and completely reversible, with exposures designed to be those experienced in a typical urban commute.

Screening and Enrollment

Subjects were screened to determine eligibility. At screening, subjects were required to be in the normal range for BMI, blood sugar, cholesterol and triglyceride levels, lung function, and blood pressure, and to have a normal ECG. Subjects also completed questionnaires describing past illness, health history, traffic and chemical exposure, smoking history, and occupation. Buccal swab samples were collected in order to achieve a balance of the TRPV1 (SNP I-585V, rs8065080) gene, which our prior work suggests modifies vascular response to traffic-related air pollution. 

Study Design

Sixteen eligible subjects completed three monitoring sessions consisting of three 2-hour commutes that traveled I-5, extending from north Seattle to roadways in south Seattle (e.g., Duwamish Valley). During each drive, subjects were accompanied by research staff responsible for collecting subject health measurements and monitoring conditions. Each drive was separated by at least 3 weeks. In two of the monitoring sessions there was no in-vehicle filtration. In the third session, cabin air was filtered using HEPA filters (see Year 5 Annual Report for additional details). The order of the sessions was randomized, and the scenario was conducted in a double-blinded manner. The cabin ventilation controls were adjusted such that air was entrained and directed to the floor vents and the temperature inside the vehicle was comfortable for the occupants. Van windows remained closed during the drive and subjects wore N95 masks while transitioning from the lab to the UW van regardless of drive condition.

Health Measurements

Subjects completed health measurements at baseline, during the drive, immediately after the drive, 3 hours later, 5 hours later and 24 hours later (i.e., 22 hours after the end of the 2-hour exposure). The health measurements included questionnaires, blood markers, Holter ECG, ambulatory blood pressure, 24-hour urine, brachial artery reactivity, retinal photography, and Finometer measurements. Subjects were required to submit samples for pregnancy and cotinine tests.

Air Monitoring

A consistent route was used for each subject exposure drive. Each day included air monitoring using the following suite of monitors in order to collect real-time measurements of the pollutants: PM2.5 (Nephelometer, Radiance Research), black carbon (microAethalometer, Aeth Labs), particle count (P-Trak, TSI Inc), particle-bound polycyclic aromatic hydrocarbons (PAHs) (PAS 2000CE, EcoChem), NO2 (CAPS, Aerodyne Research Inc), NOx (UV absorbance Model 410, 2B Technologies), ozone (chemiluminescence 3.02P, Optec), CO (CO T15n, Langan), CO2 (CO2 K-30-FS Sensor, CO2, temp/RH (Precon HS-2000, Kele Precision Mfg), location (GPS BU-353, US GlobalSat), and noise (dosimeter). The on-road concentrations of some air pollutants can be dramatically higher than concentrations of the same pollutants even a short distance from a major roadway. These pollutant gradients are one of the rationales for conducting on-road measurements inside a vehicle where the study subject can have both physiological responses and air pollutant exposures characterized. During 2014, CCAR Project 4 prepared and extensively pilot tested the UW mobile monitoring vehicle in order to test filter efficiency and particle size ranges present during on road exposures.

Important operational concerns included the ability to create a different “control,” or minimal exposure case, for baseline comparison with the pollutant exposure situation. Pilot studies were conducted that began with an assessment of pollutant exposures along major highways in the Seattle area, then continued with an investigation of methods for distinguishing the clean filtered air control case from the exposure condition inside the vehicle. Multiple test drives were conducted to refine the route to include exposure to higher concentrations, and to diesel exhaust from heavy truck traffic, specifically. More details can be found in previous Annual Reports.

Filtered air drives had nearly an order of magnitude reduction in particle number, particle concentration as estimated by nephelometry and black carbon concentrations compared with unfiltered air drives. Gaseous species were minimally affected by the filtration system, with a slight reduction in oxides of nitrogen, while CO2 and CO were slightly higher in the filtered air drives, but with considerable variation.

Health Effects Results

Noise and Perception of Exposure

All subjects had similar in-vehicle background noise levels. The condition of the car on a given day (filtered vs. unfiltered) was correctly identified only 23 percent of the time--no better than chance--—indicating that participants were successfully blinded to the exposure situation.

Blood Pressure

Blood pressure data was collected using a Finapres Finometer machine at 14 different time points corresponding to immediately before the commute began (1 time point), during the commute (9 time points) and after the commute (4 time points up to 24 hours after). There were 14 participants with usable blood pressure data. We performed an analysis comparing blood pressure between exposure groups (filtered vs. unfiltered days) using a mixed effects model with random effects for participant. Filtration status was interacted by time point, creating effect estimates that correspond to individual changes in blood pressure compared to baseline readings comparing unfiltered versus filtered days.

Diastolic blood pressure was significantly elevated relative to baseline at several time points during the commute (time points from 9 to 120 minutes after baseline) and up to 24 hours after the commute. Effect sizes were large (up to 10 mmHg during exposure and 4 mmHg 24 hours after exposure). At approximately 1 hour after drive start, diastolic blood pressure relative to pre-drive levels was on average 7.6mmHg units higher (95% CI: 3.2,12.0) in unfiltered drives compared to filtered drives. The effect of unfiltered air decreased to nearly zero at 2 hours but then increased again post-exposure. The pattern for systolic blood pressure was similar. The point at 120 minutes when average blood pressure returned to baseline is consistent with previous controlled exposure data from our group and may indicate subacute compensatory responses to blood pressure changes. This preliminary analysis indicates that exposure to some traffic-related air pollutants during in-vehicle commutes may affect blood pressure acutely and for at least 24 hours after exposure. Results additionally indicate that in-vehicle use of carbon filters may be effective at reducing biological responses to exposure. Further analyses of these data will include other outcomes such as baroreceptor sensitivity and attempt to assess the role of specific pollutants using continuous exposure metrics. We conclude that urban traffic pollutant inhalation causes substantial changes in blood pressure in realistic conditions that occur rapidly, are sustained over 24 hours, and can be reduced with effective air filtration.

Gene Expression

Changes in gene expression were investigated at 5 and 22 hours post-exposure (compared to pre-exposure). We selected genes involved in ion channel/vasculature regulation (CALCA, CALM1, TRPV1, TRPA1, ICAM1), inflammation (COX2, CCL2, CCL5, IL-1b, IL-6, MMP2, SLPI), oxidative stress (GCLc, eNOS, HIF-1a, MT3) and barrier function (OCLN, ZO-1). Blood was collected in a BD Vacutainer CPT glass molecular diagnostics tube containing density gradient polymer gel and sodium citrate. Cells were pelleted by centrifugation and resuspended in TRIzol for RNA extraction. RNA was isolated from peripheral blood mononuclear cells (PBMCs) using an RNeasy kit (Qiagen; Valencia, CA); complementary DNA was synthesized with Oligo-dT primers and reverse transcriptase (Qiagen), using the same total RNA concentration for each sample. Each sample was multiplexed using the Fluidigm-based qPCR assay on a 96.96 Dynamic Array Integrated Fluidic Circuit BioMark HD System (Fluidigm, San Francisco, CA). An internal control for each gene was spiked in at various dilutions to generate a standard curve to estimate threshold cycle C(t). Samples were run in duplicate or triplicate. Data were normalized to the reference gene GAPDH selected from four reference genes tested (GAPDH, β-actin, TATA-box protein and peptidylprolyl isomerase A). Data were collected using the Fluidigm BioMark Data Collection Software and analyzed using Fluidigm Real-Time PCR Analysis Software 4.1.2.

Compared with filtered air exposures, we did not detect any significant changes in the expression of these genes following exposure to the unfiltered drives. These results suggest that these levels of exposure may not induce a detectable response in healthy adults (at least of the measurements made in this study). Alternatively, circulating markers of exposure may not reflect effects occurring at the tissue level or time points other than those investigated. It is also possible that study power was inadequate to detect intervention-related changes.

Blood Markers

Blood was collected before the drive (baseline), and 5 hours and 22 hours after. Outcomes were complete blood count with differential, cortisol, C-reactive protein and IL-6. Within-person filtration effect on each outcome was estimated using a mixed-effects model. Filtration reduced in-vehicle particle count by 85 percent and gases by 20 percent. Compared to filtered drives, exposure to traffic pollution was associated with significant decrease in total circulating leukocytes and non-granulocyte populations: WBCs [-360/µl (95% CI:-680 to -30)], monocytes [-70/µl (95% CI:-130 to -10)], and lymphocytes [-290/µl (95% CI:-460 to -120)] at 5 hours post exposure. A decrease in lymphocyte count persisted for 22 hours post exposure [-220/µl (95% CI:-400 to -40)]. There were no significant changes in cortisol, IL-6 or C-reactive protein. We conclude that short-term exposure to on-road air pollutants results in small but significant reductions in circulating mononuclear and lymphocyte cell populations, which can be prevented with effective particle filtration.

Further analyses of these data will include other outcomes: retinal photography (CRAE and CRVE), Holter ECG (HRV and QTc), brachial artery reactivity (diameter and FMD), 24-hour urine (PAHs and 1-nitropyrene) and finometry (stroke volume, cardiac output, peripheral resistance and baroreflex sensitivity). In the future, we will attempt to identify specific pollutants associated with health effects using continuous exposure metrics.

Some of these results will be presented at the International Society of Environmental Epidemiology 2018 Annual Conference and the European Respiratory Society 2018 Congress.


Despite a change in study design, we achieved our aims, which were to test the hypothesis that traffic–-derived aerosols exert vascular effects in human subjects and to provide insight into the most toxic components and mechanisms underlying epidemiological observations of cardiovascular disease events and mortality. Our study design allowed us to test the difference between unfiltered on-road exposures and those experienced using vehicle filtration systems. We found that short-term exposure to on-road air pollutants causes changes in blood pressure and certain blood markers when compared to exposures under filtered conditions.

Journal Articles on this Report : 2 Displayed | Download in RIS Format

Other subproject views: All 2 publications 2 publications in selected types All 2 journal articles
Other center views: All 187 publications 87 publications in selected types All 86 journal articles
Type Citation Sub Project Document Sources
Journal Article Cosselman KE, Krishnan RM, Oron AP, Jansen K, Peretz A, Sullivan JH, Larson TV, Kaufman JD. Blood pressure response to controlled diesel exhaust exposure in human subjects. Hypertension 2012;59(5):943-948. R834796C004 (Final)
R827355 (Final)
R830954 (Final)
  • Full-text from PubMed
  • Abstract from PubMed
  • Associated PubMed link
  • Full-text: Hypertension-Full Text HTML
  • Abstract: Hypertension-Abstract
  • Other: Hypertension-Full Text PDF
  • Journal Article Cosselman KE, Navas-Acien A, Kaufman JD. Environmental factors in cardiovascular disease. Nature Reviews Cardiology 2015;12(11):627-642. R834796 (2015)
    R834796 (Final)
    R834796C004 (2015)
    R834796C004 (2016)
    R834796C004 (Final)
  • Abstract from PubMed
  • Abstract: Nature Reviews Cardiology-Abstract
  • Other: Johns Hopkins University-Abstract
  • Supplemental Keywords:

    Cardiovascular health, diesel exhaust, gasoline exhaust, fine particles, volatile organic compounds, blood pressure., Health, Scientific Discipline, Air, ENVIRONMENTAL MANAGEMENT, Air Quality, air toxics, Health Risk Assessment, Risk Assessments, mobile sources, Biochemistry, Risk Assessment, ambient air quality, atmospheric particulate matter, particulate matter, aerosol particles, air pollutants, motor vehicle emissions, vehicle emissions, air quality models, motor vehicle exhaust, airway disease, bioavailability, air pollution, particle exposure, atmospheric aerosols, ambient particle health effects, vascular dysfunction, cardiotoxicity, atmospheric chemistry, exposure assessment

    Relevant Websites:

  Exit Exit


    Progress and Final Reports:

    Original Abstract
  • 2011 Progress Report
  • 2012 Progress Report
  • 2013 Progress Report
  • 2014
  • 2015 Progress Report
  • 2016 Progress Report

  • Main Center Abstract and Reports:

    R834796    University of Washington Center for Clean Air Research

    Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
    R834796C001 Exposure Mapping – Characterization of Gases and Particles for ExposureAssessment in Health Effects and Laboratory Studies
    R834796C002 Simulated Roadway Exposure Atmospheres for Laboratory Animal and Human Studies
    R834796C003 Cardiovascular Consequences of Immune Modification by Traffic-Related Emissions
    R834796C004 Vascular Response to Traffic-Derived Inhalation in Humans
    R834796C005 Effects of Long-Term Exposure to Traffic-Derived Particles and Gases on Subclinical Measures of Cardiovascular Disease in a Multi-Ethnic Cohort