Article Text
Abstract
Objectives: Although associations have been found between levels of ambient airborne particles and cardiovascular disease (CVD) in the general population, little is known about possible cardiovascular effects from high exposure to particles in underground railway systems. This study investigates risk markers for CVD in employees exposed to particles in the Stockholm underground system.
Methods: 79 workers (54 men and 25 women) in the Stockholm underground were investigated between November 2004 and March 2005. All were non-smokers aged 25–50 years. Three exposure groups were delineated: 29 platform workers with high exposure to particles, 29 train drivers with medium exposure and 21 ticket sellers with low exposure (control group). A baseline blood sample was taken after 2 non-working days, and a second sample after 2 working days, for analysis of levels of plasminogen activator inhibitor-1 (PAI-1), high-sensitivity C-reactive protein (hs-CRP), interleukin-6, fibrinogen, von Willebrand factor and factor VII. The study investigated changes in plasma concentrations between sample 1 and sample 2, and differences in average concentrations between the groups.
Results: No changes between sample 1 and 2 were found that could be attributed to particle exposure. However, the highly exposed platform workers were found to have higher plasma concentrations of PAI-1 and hs-CRP than the ticket sellers and train drivers. This suggests that particle exposure could have a long-term inflammatory effect. These differences remained for PAI-1 in the comparison between platform workers and ticket sellers after adjusting for body mass index.
Conclusions: Employees who were highly exposed to airborne particles in the Stockholm underground tended to have elevated levels of risk markers for CVD relative to employees with low exposure. However, the differences observed cannot definitely be linked to particle exposure as such.
Statistics from Altmetric.com
Levels of particles with an aerodynamic diameter of less than 10 μm (PM10) measured on an underground platform in Stockholm were about five times higher than on one of the busiest city streets.1 High levels of airborne particles in underground systems have also been reported in London,2 3 New York4 and Rome.5 Both transportation workers and commuters are exposed, especially while on underground platforms. The particles are derived chiefly from wheels, rails and brakes, and contain a high proportion of iron. They are mainly in the 1–10 μm range and are thus larger than particles generated by combustion engines.3 6 7
There is a well-established association between particle levels in urban air and cardiovascular morbidity and mortality in the general population.8–10 A common hypothesis is that inhaled particles are deposited in the small airways and cause an inflammatory response that increases the risk of cardiovascular disease (CVD) by increasing blood coagulability, influencing the atherosclerotic process or altering the autonomic nervous control of the heart.11–13 Several haematological factors, including plasminogen activator inhibitor type 1 (PAI-1), C-reactive protein (CRP), interleukin-6 (IL-6), fibrinogen and factor VII, are known to be predictive of CVD and are also markers of inflammatory reactions.11 13 The above hypothesis has gained support from several empirical studies. Air pollution from traffic was found to be associated with high concentrations of plasma fibrinogen in office workers in London.9 Exposure to airborne particles inside patrol cars in North Carolina was found to be associated with increased numbers of neutrophils and elevated levels of CRP and von Willebrand factor (vWF), as well as changed heart rate variability.14 IL-6 and fibrinogen levels were found to be increased in Norwegian tunnel construction workers after exposure to dust.15 The hypothesis that air pollution has an effect on fibrinolytic function is also supported by the recent observations of Mills and colleagues, who found that exposure to diesel exhaust impairs fibrinolytic capacity in men with coronary heart disease by reducing the acute release of endothelial tissue plasminogen activator (t-PA).16
In contrast to the well-documented detrimental effects on the cardiovascular system of particles in city air, most of which are derived from traffic, very little is known about the possible cardiovascular effects of exposure to particles in underground railway systems. Experimental data have shown that particulate matter from undergrounds has marked inflammatory effects on cultured human lung cells3 17 18 and induces more oxidative stress than particles from street level.17 19 The high exposure to airborne particles in the underground system, the experimental data indicating that these particles have inflammatory properties and the lack of studies in humans led us to investigate risk markers for CVD in people working in the Stockholm underground system who are exposed to such particles.
METHODS
Study group
The study group comprised 79 employees in the Stockholm underground, who were investigated between November 2004 and March 2005. The participants were non-smokers of both sexes aged 25–50 years. There were no criteria regarding freedom from disease, although subjects on medication that could influence the haemostatic balance (such as anti-coagulation medication) were not included. The participants were divided into three exposure groups according to level of exposure to underground particles (see below). There were 29 highly exposed platform workers (cleaners and ticket collectors mainly working on platforms), 29 moderately exposed train drivers and 21 ticket sellers with low exposure (control group).
Information on particle exposure
For 44 of the participants (11 cleaners, 12 ticket collectors, 13 train drivers and eight ticket sellers), exposure to particles (measured as PM2.5 and by DataRAM) was investigated by personal sampling during two work shifts.20 PM2.5 levels for the cleaners were about seven times higher than for the control group (ticket sellers), about five times higher for ticket collectors, and about twice as high for train drivers, with an average PM2.5 of 79 μg/m3 (standard deviation (SD) 17) for cleaners, 50 μg/m3 (SD 8) for ticket collectors, 19 μg/m3 (SD 3) for train drivers and 10 μg/m3 (SD 3) for the control group. The corresponding DataRAM (a nefelometric instrument scanning particles 1–10 μm in size) levels were 256 μg/m3 (SD 97), 108 μg/m3 (SD 26), 33 μg/m3 (SD 12) and 13 μg/m3 (SD 3), respectively.
Background characteristics
The participants were examined and interviewed about the state of their health and were asked if they were on medication. Blood samples were taken for the analysis of markers of inflammation and coagulation (see below) and lipoprotein parameters, and blood pressure, height and weight were recorded. We calculated body mass index (BMI) from information on weight and height (BMI = weight in kg/(height in m)2). We coded individuals with a history of drug- or diet-treated diabetes as being diabetic. We assessed the presence of hypertension from information on the use of antihypertensive drugs, or a systolic blood pressure exceeding 160 or a diastolic pressure exceeding 90 at the examination. Individuals who had smoked on a regular basis any time before inclusion were coded as ex-smokers. Eight people had asthma and three of these were on medication with inhaled corticosteroids on a regular basis (one in each exposure group). One person (a train driver) had mild ulcerative colitis and was on medication with a topical anti-inflammatory agent. Table 1 shows the background characteristics of the exposure groups.
Blood sampling
The first blood sample was taken before the start of a work shift after at least 2 non-working days. The second sample was taken 48 h later after 2 days of work. Both blood samples were taken after 1 night of fasting and at the same time of day, with a maximum difference of 2 h. The blood samples were collected for analysis of the inflammatory markers PAI-1, high-sensitivity CRP (hs-CRP) and IL-6 and the coagulation parameters fibrinogen, vWF and factor VII. The coagulation parameter fibrinogen is also a marker of inflammation. The first blood sample was also tested for lipoprotein parameters. Blood samples for the analysis of coagulation parameters and inflammatory markers were centrifuged immediately at 2200 g for 15 min at 15°C. The blood plasma was then immediately frozen with solid carbon dioxide and stored at −70°C. Blood samples for the analysis of cholesterol and triglycerides were kept at room temperature for 30 min, centrifuged at 2200 g for 15 min and then stored at room temperature. All blood samples were collected and prepared in a room adjacent to the workplace and were transported to the laboratory at Karolinska University Hospital on the same day.
Statistical methods
To investigate the short-term effects of particle exposure, we analysed whether there was a significant increase in plasma concentrations from the first to the second sample. Since the data were not normally distributed, we used Wilcoxon signed ranks test in these analyses. We also compared plasma concentrations in sample 2 between groups (cross-sectional comparison). We used the second blood sample to examine acute and long-term effects. In the comparison between groups we used the Mann-Whitney test for non-adjusted data. We investigated any correlation between log-transformed blood parameters and BMI by linear regression. The blood parameters PAI-1, hs-CRP, IL-6 and fibrinogen were correlated to BMI (Pearson correlation factor of 0.547, 0.191, 0.293 and 0.366, respectively). The correlation was significant at the 0.01 level (two-tailed) for all these blood parameters except hs-CRP (p = 0.09). We adjusted the values of PAI-1, hs-CRP, IL-6 and fibrinogen for BMI in the comparison between groups and used t test of logarithmically transformed BMI-adjusted data.
The study was evaluated and approved by the Regional Ethics Committee in Stockholm, Sweden. All study subjects gave their informed consent before inclusion.
RESULTS
Table 2 shows the median plasma concentrations of PAI-1, vWF, factor VII, IL-6, hs-CRP and fibrinogen for samples 1 and 2 in the various exposure groups.
There was a significant increase from the first to the second sample in the plasma concentration of PAI-1 in ticket sellers (p = 0.03) and fibrinogen in train drivers (p = 0.01), but in platform workers, who were highly exposed, there was no significant increase in any of the blood markers. Thus, we found no acute changes that could be attributed to particle exposure. However, when comparing the plasma concentrations between the groups, we found that platform workers had higher levels of PAI-1 and hs-CRP than the two groups with lower exposure, both in samples 1 and 2. The plasma concentration of PAI-1 in sample 2 was significantly higher in platform workers compared to ticket sellers (p = 0.02) and train drivers (p = 0.02), and hs-CRP in sample 2 was significantly higher in platform workers compared to train drivers (p = 0.04). After adjusting for BMI, the difference remained significant for PAI-1 in the comparison between platform workers and ticket sellers (p = 0.02) (table 3).
Additional analyses of sample 2 were restricted to individuals without diabetes or high blood pressure, with simultaneous adjustment for BMI. The findings were similar to those for all individuals included (in the comparison between platform workers and ticket sellers, p = 0.04 for PAI-1, p = 0.28 for hs-CRP, p = 0.17 for IL-6 and p = 0.17 for fibrinogen).
DISCUSSION
After 2 days working in the Stockholm underground there was no increase in the plasma concentrations of markers of inflammation and coagulation in platform workers who were highly exposed to airborne particles. It is possible that particles in the underground system do not have a short-term effect on the risk markers that we studied. It is also possible that the work-free period of 2 days before the first blood sample was insufficient to reach baseline or that 2 days of exposure is insufficient to cause an effect. However, the lack of effect could also be due to the relatively small sizes of the exposure groups. The reason for the increase in PAI-1 in ticket sellers and fibrinogen in train drivers is unclear, but may be due to some parameter that differs between the situation at work and leisure time. Although physical activity may cause a short-term inflammatory response, it seems unlikely that this can explain the findings since ticket sellers and train drivers are sedentary workers.
The highly exposed platform workers showed elevated levels of two blood markers of inflammation compared to the control group, both before and after working 2 days in the underground. After adjusting for BMI, the difference remained significant for PAI-1 and there was still a tendency for higher levels of hs-CRP. These findings indicate a possible long-term inflammatory effect of exposure to airborne particles in the underground system. However, the study was primarily designed to make inter-individual comparisons of risk markers before and after 2 days of work. The observed difference in plasma levels between the groups cannot be definitely linked to particle exposure. Other possible explanations include diurnal variation, physical activity and metabolic factors, which are discussed below. The three exposure groups are also relatively small for these kinds of comparisons.
PAI-1 is subject to diurnal variation, with the highest levels occurring in the early morning. However, the platform workers were sampled at the same time of day as the control group. Platform workers, especially cleaners, have a more physically demanding job than ticket sellers and train drivers, but regular physical activity has been shown to lower CRP levels and lessen inflammation, probably because it reduces the production of IL-6, which plays a part in the synthesis of CRP.21 22 An increased inflammatory response is only a short-term effect of exercise.22 23 Existing data on exercise and the effects on fibrinogen are more conflicting.24 25 Differences in job strain would probably not explain the results, although job strain may be associated with plasma fibrinogen concentrations.26 However, several of the inflammatory markers, especially PAI-1, are associated with parameters related to the metabolic syndrome.27 28 Although the findings were adjusted for BMI, it cannot be excluded that some of the differences in the plasma levels of inflammatory markers between the groups may have been caused by differences in parameters other than BMI. We investigated this possibility by restricting the analysis to individuals without diabetes or high blood pressure, with simultaneous adjustment for BMI, but this did not change the results to any marked degree.
Main messages
Employees exposed to high levels of particles in the Stockholm underground tended to have elevated levels of inflammatory markers, relative to employees with low exposure.
Policy implications
The present findings encourage continued efforts to reduce the levels of airborne particles in the underground system for the benefit of highly exposed employees.
The elevated markers of inflammation in highly exposed workers may thus actually reflect activation of low-grade inflammation in blood vessels. Such activation of the inflammatory response would theoretically increase the risk of CVD,11 13 29 30 especially myocardial infarction, and exposure to particles in the underground system is a possible explanation for this activation. The hypothesis that there is an inflammatory response to particle exposure has support from several empirical studies,9 14 15 and experimental data show that airborne particles in underground systems have an inflammatory effect.3 17–19
Our classification into exposure groups was confirmed by personal exposure measurements for about half of the participants. Thus, the risk of misclassification of exposure was low. The contrast between highly exposed and mildly exposed workers was high; in terms of PM2.5 the exposure levels were about six times higher for platform workers than for the control group, and for DataRAM the levels were about 14 times higher.20
In conclusion, we found no increase in markers of inflammation and coagulation after 2 working days in employees who were highly exposed to airborne particles in the Stockholm underground. However, we found elevated plasma concentrations of the inflammatory marker PAI-1 and a tendency to elevated plasma concentrations of hs-CRP in the highly exposed group of platform workers, relative to employees with low exposure. Our findings indicate a possible association between exposure to particles in the underground and long-term effects on inflammatory markers, which suggests that there may be an increased risk of future CVD in individuals who are highly exposed to such particles. Since the study was primarily designed to investigate short-term effects in individuals, the observed elevation cannot be linked to particle exposure with certainty, even though the results cannot be explained by the prevalence of overweight, metabolic syndrome, physical activity or smoking habits.
Acknowledgments
We thank Stina Gustavsson of the Department of Occupational and Environmental Health in Stockholm, who collected and prepared the blood samples and participated in the examinations. We also thank the staff at the clinical chemistry laboratory of Karolinska University Hospital, who provided us with equipment for sampling and preparation of the blood samples for field use. We thank the managers and employees of Connex Sverige AB in Stockholm for their contributions to the study.
REFERENCES
Footnotes
-
Funding: The study was supported financially by the Swedish Council for Working Life and Social Research (grant number 2004–0276).
-
Competing interests: None.