In this randomized crossover study, we found that a two-hour exposure to diesel exhaust in a controlled exposure facility resulted in a significant increase in the hematocrit immediately post-exposure without an increase in erythrocyte number. This increase appeared more pronounced in the metabolic syndrome subjects. Healthy, but not metabolic syndrome subjects demonstrated an increase in platelet count a day following DE exposure, without any change in other measured circulating blood cell fractions. Using a multiplex assay, we observed small and not significant increases in several proteins with DE compared to FA, including circulating endothelial adhesion molecules, myeloperoxidase, adiponectin, interleukins-1β, 6, and 10. While the finding would not be robust to multiple testing corrections, the increase in IL-1β immediately post-exposure achieved borderline statistical significance in standard assessment.
This is the first experimental study to examine the hematological indices, circulating peripheral cell count and a variety of inflammatory mediators in a chronological manner in humans following DE exposure. DE is a major contributor to traffic related pollution and contains higher fine particulate content. Therefore, our findings provide some support to the hypothesis that acute exposure to traffic related air pollution initiates hemoconcentration, systemic inflammatory process activating the endothelial–blood cell interface, and increases circulating platelets. All these biological processes are potential important steps in mediating adverse cardiovascular events associated with fine particulate matter exposure .
Our finding of a rapid increase in hematocrit without any changes in erythrocytes is in contrast to Seaton’s study that first showed a negative association between 3-day average personal PM10 exposures and blood parameters such as hemoglobin, packed red cell volume [16, 17]. This striking difference could be due to the different specific concentrations and the experimental setting that we employed to study the acute effects of traffic related air pollutants. Other experimental studies have reported peripheral blood counts, but not hematocrit or hemoglobin from normal healthy subjects. Those studies have also examined the systemic markers in humans for only up to 6hrs [18, 19]. It is unclear how fine particulate matter exposure would lead to hemoconcentration, but we speculate that this could be due to a combination of mechanisms such as volume status or stress [20–22] that could affect the blood viscosity. It is unlikely that our results could be influenced purely by our subjects’ fluid intake, as they go through much of the study in a fasting state, identical in each exposure day (DE or FA), and blinded to exposure status, and analyses compare the same time of day for each exposure. Since only DE exposure resulted in a significant hemoconcentration, it implies a direct effect on hemoconcentration, though it is also possible that DE exposure was associated with less desire to consume fluids.
Our recent observation of a 5mm Hg increase in the systolic blood pressure immediately after exposure to DE also indicates sympathetic nervous system mediated effects of PM2.5 and future studies will address these questions more specifically .
Few studies have addressed the direct effect of PM2.5 and other ambient air pollutants on platelets. In our study, we found that inhalation of DE in healthy subjects increased platelet count 22 hours following exposure. It is of note that we did not observe an increase in platelets in the metabolic subjects, who were also older and may have an underlying inflammatory process or impaired bone marrow function which could blunt the response seen in the normal healthy subjects (Figure 1). We previously reported little effect of DE on markers of thrombosis including D-Dimer, von-Willebrand factor in metabolic syndrome subjects,  or in healthy subjects . In contrast, another group has shown that exposure to DE results in impaired fibrinolysis and recently showed that DE inhalation increases ex vivo thrombus formation and in vivo platelet activation in healthy subjects . Our observations could be seen as at odds with these prior findings, but might also be due to differences in exposure characteristics or differences in the molecular targets of our outcome measures. Similar to the human experimental studies, animal studies showed that mice exposed to concentrated ambient particles had a significant increase in their platelet count with evidence of platelet activation . Both arterial and venous thrombosis were noted with increased platelet aggregation in hamsters treated with intra-tracheal instillation of DE particles . Although, we have only seen an increase in the platelet count, this is most likely the first process for thrombogenesis, or could be due to a direct bone marrow stimulatory effect as demonstrated in other studies  or a reactive thrombocytosis from a systemic inflammatory response.
In the multiplex luminex assay, the trend toward an increase in the endothelial adhesion molecules (E-selectin, ICAM-1, VCAM-1) and inflammatory cytokines (such as interleukin 1β, interleukin-6, interleukin-10, myeloperoxidase, and matrix metalloproteinase-9) are fairly consistent with other air pollution studies; none of these findings were statistically significant and none were robust to adjustment for multiple comparisons. Epidemiological study from children in the Mexico City observed significant increases in inflammatory mediators and vasoconstrictors, including tumor necrosis factor (TNF) alpha, prostaglandin, C-reactive protein, interleukin-1β, and endothelin-1, and down regulation of endothelial adhesion molecules compared to the children from a less polluted city . An in vitro experiment using DE particles caused an inflammatory response in macrophages characterized by increased IL-8 and TNF-α concentration. This inflammatory condition generated by macrophages in response to DEP was a potent activator of endothelial cells than direct particle treatment [28, 29].
All the results noted in our current study provide some insights into the mechanisms underlying the short-term exposure to air pollution and cardiovascular diseases. While prior experimental studies in animals and humans have shown that exposure to fine particulate matter results in atherosclerosis progression [30–32]. Less is known about the association between particulate matter exposure and atherothrombosis. Atherothrombosis occurs due to interplay between thrombosis, procoagulation, endothelial erosion, and possibly increased viscosity, typically occluding the lumen acutely leading to acute events such as unstable angina, myocardial infarctions [33–35]. We also noted an increase in markers of neutrophil activation such as myeloperoxidase and matrix metalloproteinase-9, but not increased neutrophil count suggesting that DE exposure could promote systemic inflammation by activating neutrophils than by increasing their production. These mechanisms are proposed based on previous epidemiological studies that showed an association between short-term increases in traffic related air pollution exposure and increased incidence of myocardial infarction [1, 35] and deep venous thrombosis . While there are not many human experimental studies that have shown an association between air pollution exposure and increased hematocrit, prior cardiovascular epidemiology studies have shown an association between elevated hematocrit and coronary artery disease suggesting that elevated hematocrit is associated with a greater risk of acute coronary syndromes, which is hypothesized to be from increased whole blood viscosity [37–39]. Similar to this, a panel study of elderly patients with CAD showed an increase in inflammatory cytokines (interleukin-6), oxidative biomarkers (myeloperoxidase), and platelet activation [11, 40].
Endothelial adhesion molecules have been considered as markers of endothelial injury,  and have been associated with adverse cardiovascular events . Endothelial activation results in up regulation of leukocyte adhesion proteins, attracts monocytes through release of growth factors and pro-inflammatory cytokines, and initiates vascular inflammation . However, in vivo intra-vital microscopy studies have shown that platelet-endothelium adhesion can occur in intact endothelium without activation of the coagulation cascade  through activation of platelet and endothelial adhesion molecules like ICAM-1, VCAM-1, and E-selectin. Our previous study did not show a significant increase in the pro-coagulant and pro-thrombotic factors including von Willebrand factor, plasminogen activator inhibitor-1, and D-dimer, suggesting that the activation of the coagulation cascade may not occur as an early signal in our experimental model or could be due to lack of adequate sensitivity of these assays [14, 15]. Based on our observations, we hypothesize that platelet-endothelial activation rather than procoagulation could be one of the underlying mechanisms in subjects exposed to DE that have to be confirmed in future studies.
The major strengths of our study are the use of a rigorously controlled experimental design in which each subject serves as their own control, the enrollment of both healthy subjects and those with metabolic syndrome, and observations at early (7 hours following exposure initiation) and late (22 hours) time points. Using a randomized crossover study design, we were also able to study the effect of DE directly on a broad range of cytokines and cell counts in a sequential manner.
Our study has several limitations. We would expect an experimental study like ours to have more robust abilities to detect an effect, if present, than observational designs, but our findings are not strongly positive. Several factors may explain this. Although we were able to assess the effect on multiple cytokines using the most sensitive assay,  there was wide variability in the calculated concentrations. Prior studies have validated the multiplex assay in comparison to the ELISA, using samples other than plasma [46, 47]. Other investigators who employed this assay in inflammation research  and in non-diseased healthy subjects,  reported that their findings are difficult to interpret due to the low sensitivity of this assay from multiple interactions with other plasma proteins and poor threshold of detection for the inflammatory markers such as IL-1β, TNF-α and IL-6. In addition to these, our study samples utilized for the multiplex assay also came from a subset of heterogeneous population consisting of normal healthy subjects and metabolic syndrome. Furthermore, our experimental findings cannot distinguish if the source of the inflammatory markers is pulmonary or systemic in origin. It is also unclear from our study if the increased platelet in the circulation is from bone marrow stimulation or a reactive thrombocytosis phenomenon due to increased inflammation. Hence, to address these issues, we plan to use more sensitive, specific ELISA assays for the endothelial adhesion molecules, neutrophil inflammatory markers, cytokines, and platelet activation assays in our future studies. Clinical studies such as these are logistically difficult, and the small number of participants seems an obstacle to obtaining sufficient statistical power. Nevertheless, we were well powered to detect clinically significant changes in our outcome parameters.