Cohort studies suggest that long-term exposure to higher levels of outdoor air pollution, particularly particulate matter, increases risk of cardiovascular death1–5 and cardiovascular disease.6 Whether this is because sustained exposure to higher levels of air pollution causes long-term structural changes to the cardiovascular system or because episodes of high air pollution precipitate cardiovascular events (eg, arrhythmias or plaque rupture) is not known.4,5,7 Two studies suggested that air pollution could have long-term effects on the structure of blood vessels,8,9 supporting the idea that the increased risk of cardiovascular disease is due to effects of long-term exposure, rather than the summation of the effects of acute episodes of air pollution.
The mechanism by which outdoor air pollutants could affect short- or long-term cardiovascular risk is not known,4 but 1 hypothesis is that inhalation of particles, particularly those with a diameter of less than 10 μm (PM10), provokes inflammation in the lung, causing the release of inflammatory mediators into the bloodstream that may influence hemostasis or accelerate atherosclerosis, for example. This is made more plausible by the observation that chronic low-grade inflammation is a risk factor for cardiovascular disease.10,11
In support of this idea, several epidemiologic studies have found acute effects (over hours or days) of air pollution episodes on levels of circulating markers of systemic inflammation fibrinogen12 and C-reactive protein,13–15 although other studies have found no effects16 or effects in the opposite direction.13 Experimental studies in small numbers of human volunteers have shown systemic inflammatory responses to inhalation of particles.17 Although it is possible that outdoor air pollution episodes cause acute elevations of systemic inflammatory markers, in the same way as acute illnesses, it is not clear whether these, or long-term average exposure to air pollution, have any role in determining long-term levels of systemic inflammation. We have found only 1 study examining whether long-term exposure to higher levels of air pollution could lead to chronic systemic inflammation; it reported a positive association between exposure to PM10 and white cell count that was more marked in people with metabolic syndrome.18
We aimed to study whether long-term average exposure to higher levels of air pollution was associated with higher levels of systemic inflammation, by comparing levels of inflammatory markers over a geographic area with large variation in average levels of exposure to pollutants. We examined whether long-term exposure to outdoor air pollutant levels (PM10, nitrogen dioxide [NO2], sulfur dioxide [SO2], and ozone [O3]) was related to concentrations of the markers of inflammation fibrinogen and C-reactive protein collected during 3 representative cross-sectional studies of the English population conducted in 1994, 1998, and 2003.
Participants and Outcome Measures
The Health Survey for England is a program of annual surveys of representative samples of people living in private households in England (population about 50 million). We used data on adults participating in 1994, 1998, and 2003, years in which the Health Survey for England concentrated on cardiovascular disease and its risk factors.
In each survey year, adults were recruited using a multistage sampling strategy designed to generate a representative sample of the English population (eSupplement I, http://links.lww.com/A724). Briefly, the sampling units were postcode sectors, which have a median area of about 4 km2 in England. For each survey year, 720 English postcode sectors (∼10% of the total) were selected with a probability proportional to the number of addresses within the sector. From each selected postcode sector, 19 addresses were drawn (18 in 1994). All adults (up to a maximum of 10) living in each household at those addresses were invited to complete a health questionnaire. Those who completed the questionnaire were offered a nurse visit to take a nonfasting blood sample unless they were pregnant, had a history of epilepsy, had a clotting disorder, or were taking anticoagulants.
Blood samples were analyzed for fibrinogen in 1994, 1998, and 2003 and high sensitivity C-reactive protein in 1998 and 2003. In 1998 and 2003, the Royal Victoria Infirmary, Newcastle-upon-Tyne analyzed fibrinogen with the MDA 180 analyzer (Organon Teknika, Cambridge, UK), using a modification of the Clauss thrombin clotting method.19 In 1994, the West Middlesex University Hospital analyzed fibrinogen by the nephelometric method. The Royal Victoria Infirmary analyzed C-reactive protein in both 1998 and 2003 using the N Latex CRP mono Immunoassay on the Behring Nephelometer II Analyzer (Dade Behring, Germany). Median coefficients of variation for measures of within-assay variability by survey year were 5.5% to 6.6% for fibrinogen and 3.7% to 4.1% for C-reactive protein. Median differences between assay and external sample results by survey year were for fibrinogen: 1994: 11.0%; 1998: 1.0%; and 2003: −2.9%; and for C-reactive protein: 1998: 9.0% and 2003: −3.2%. eSupplement II (http://links.lww.com/A725) provides more detail of methods of collection of blood samples, and quality control methods and results.
We analyzed data from people aged 16 and older in the 1994 and 2003 surveys and people aged 18 and older in the 1998 survey (in which a blood sample was not routinely requested in people aged 16 and 17). We analyzed data only on participants belonging to white ethnic groups, because the numbers belonging to other ethnic groups were small and ethnically heterogeneous.
We assumed that annual average pollutant exposure for people living in each postcode sector was that of the 1 km2 where its centroid (based on medians of x and y coordinates of each address within a postcode sector) was located. We estimated annual average background exposure to PM10, NO2, SO2, and O3 for each 1 km2 of England from an emission inventory by using air dispersion models including the effect of weather conditions.20 Further details of the emission inventory, the models, and their validation are provided in eSupplement III (http://links.lww.com/A726). Briefly, the models for PM10,20,21 NO2,22–24 and SO220 were constructed by identifying all known emission sources of particulates, oxides of nitrogen and SO2 (for example, roads, railways, industrial sites, power stations), and estimating quantities of emissions, using either reported emissions data from regulated industries or emission factors derived from measurements from a number of sources assumed to be representative of a particular source sector (for instance, the power station sector or the rail sector).25 To obtain NO2 maps, maps for oxides of nitrogen were adjusted for the effect of the availability of oxidant in different environments (rural, urban, city center). We used monitoring station data to calibrate modeled concentrations for PM10 and NO2. O3 maps were constructed by interpolating data from rural monitoring stations and adjusting for effects of altitude and NO2 emissions in urban areas by using the concentrations calculated with the air dispersion models.26
Because data collection for each Health Survey is conducted over the whole year and part of the following year, we averaged the exposure estimate for each participant for each pollutant for the data collection year and the previous year. Because we did not have pollution data for 1993, for analyses of Health Survey for England 1994, we used annual averages for 1994 only.
Measures of each pollutant exposure were at the postcode sector level. The median number of adults taking part in the Health Survey for England living in each postcode sector was 22 in 1994, 21 in 1998, and 21 in 2003. Therefore, we used multilevel linear regression analyses to take account of the hierarchical structure of the data. Fibrinogen and C-reactive protein concentrations were log transformed to normalize the distributions of the residuals before performing the analyses.
We performed all regression analyses for each pollutant singly for each year, controlling for the following individual-level potential confounders: age (in 10-year age groups: to protect the identity of the participants, we were not permitted to analyze age as a continuous variable); sex; body mass index (in quartiles; the relationship between body mass index and the inflammatory markers was not linear); social class of head of household (defined by UK Registrar General's classification of occupations); cigarette smoking (never, ex-, current); and region (8 regions). We included interaction terms between age, sex, and body mass index in all models because preliminary analyses showed that the associations with age and body mass index differed between men and women. We also controlled for factors that may have contributed to measurement error: indoor temperature at time of nurse visit in quintiles (not in 1994 as temperature data were not available); and month of nurse visit, because both ambient temperature and season may influence levels of inflammatory markers.27,28
We excluded people with a recorded fibrinogen concentration of less than 0.05 g/L from the analysis. The multivariable regression models excluded people with missing data on any of the potential confounders and those with a recorded body mass index of less than 10 kg/m2 or a recorded indoor temperature of less than 10°C, having assumed that these values represented errors in data collection or processing.
To examine whether our estimates were biased by including people with known risk factors for raised levels of C-reactive protein or fibrinogen, we repeated the analyses after excluding people taking β-blockers or lipid-lowering drugs; people with known cardiovascular disease, diabetes, or rheumatologic disease; and people who reported recent acute illness.
To examine whether any effect was only seen in people with higher body mass index, we repeated the analysis after stratifying for body mass index in 3 categories: <25, 25 to 30, and >30 kg/m2.
To examine whether our estimates were influenced by the effect of clustering within household of cardiovascular disease risk factors because of shared environment, we repeated the analyses with an additional level for household, as well as postcode sector.
We conducted fixed effects meta-analysis of the year-specific estimates for each air pollutant by using inverse variance weighting because there was no evidence of statistical heterogeneity across the years. We reported estimates of effect as percentage change in fibrinogen and C-reactive protein per 1-μg/m3 increase in pollutant level. We performed all analyses using STATA version 9.2 (STATA Corp., TX).
In 1994, 1998, and 2003, the National Centre for Social Research obtained ethical approval from all relevant Local and Multi-Centre Research Ethics Committees. Participants provided explicit consent to take part in each part of the study, including written consent for blood sampling. We obtained approval to link air pollution data to Health Survey for England by postcode sector from the National Centre for Social Research, having undertaken to follow additional procedures to protect the identity of the participants.
Seventy five percent of households approached by the Health Survey for England in 1994, 1998, and 2003 participated. In total, in these 3 years of surveys, 91% of adults living in participating households completed a health questionnaire. Of these, 93% were white. Table 1 shows response to completing a questionnaire, availability of fibrinogen and C-reactive protein results, and characteristics of white adults with blood test results for each of the survey years. Overall, we had fibrinogen or C-reactive protein results for 62% of white adults living in participating households who completed a questionnaire, that is, those who were eligible to have blood tests. People aged 35 to 74 were more likely than people in younger and older age groups to have fibrinogen or C-reactive protein measurements, as were men and people in nonmanual social classes. Of people with fibrinogen and C-reactive protein measurements, we had data on air pollution exposure for 93% in 1994, 96% in 1998, and 99% in 2003. Exposure to air pollutants in white adults with and without fibrinogen and C-reactive protein measurements was very similar (eSupplement I, http://links.lww.com/A724).
Table 2 shows geometric means of fibrinogen and C-reactive protein concentration for white adults (16+ in 1994 and 2003 and 18+ in 1998) participating in each of the survey years, by key covariates. Fibrinogen and C-reactive protein were higher in women, older people, smokers, people of lower social class, people with higher body mass index, people with recent acute illness, and people with cardiovascular disease.
Table 3 shows summary statistics of outdoor air pollution exposure estimates by postcode sector for white adults participating in each year.
Table 4 provides the crude and adjusted results of multilevel models estimating the associations between each of the 4 pollutants and fibrinogen concentration for each year, and the results of the meta-analysis. Figure 1 shows these results as forest plots.
The combined estimates of effect size were close to zero, suggesting a change of up to 0.1% in fibrinogen concentration with a change of 1 μg/m3 in pollutant concentration, and were generally negative, suggesting that increasing concentrations of pollutants were associated with a lower fibrinogen concentration. Where the confidence interval did not include zero for NO2 and PM10 in 1998, the effects were not in the expected direction.
No individual variable included in the model had any consistent effect on the size or direction of the estimates.
Table 5 provides the crude and adjusted results of multilevel models estimating the associations between each of the 4 pollutants and C-reactive protein concentration for each year and the results of the meta-analysis. Figure 2 shows these results as forest plots. Again, all the combined estimates of effect size were very close to zero, suggesting a change of up to 0.3% in C-reactive protein concentration with a change of 1 μg/m3 in pollutant concentration. The estimates for PM10, NO2, and SO2 were in the expected direction, suggesting that increasing concentrations of pollutants were associated with a higher C-reactive protein concentration. The combined estimate of effect size for O3 was in the opposite direction.
As for fibrinogen, no individual variable included in the model had any consistent effect on the size or direction of the estimates.
Repeating the analyses for both fibrinogen and C-reactive protein excluding people taking β-blockers and lipid-lowering drugs; people with known cardiovascular disease, diabetes, or rheumatologic disease; or people with reported recent acute illness made very little difference to the estimates (data not shown). Repeating the analyses for both fibrinogen and C-reactive protein separately in people of normal, overweight, and obese body mass index showed no evidence of effect modification or trend.
Repeating the analyses with an additional level for household and postcode sector led to negligible effects on the estimates of effect (data not shown).
Our study found no evidence of positive associations between concentrations of 2 markers of inflammation and 4 measures of chronic outdoor air pollution exposure in white adults living in England. Contrary to our expectations, many of the associations were negative. This raises the possibility that the health effects of chronic outdoor air pollution are not mediated by systemic inflammation.
We found 1 previous study examining the long-term effects of outdoor air pollution on systemic inflammation by using data from the Third National Health and Nutrition Examination Survey. This study found that annual average PM10 exposure was associated with a higher white cell count: an adjusted difference between the lowest quartile of exposure and the 3 highest quartiles of exposure of 138 × 106/L.18 Controlled experiments in human volunteers suggest that inhaled particulate matter induces systemic inflammation17 but whether this has any public health relevance is not yet clear. A number of studies have examined the short-term effects over a few days and weeks of air pollution episodes on fibrinogen and C-reactive protein concentrations, but the association is not yet conclusive. The results have not been consistent and some positive findings may have been confounded by weather conditions.12–16,29 However, even if it were established that air pollution episodes led to raised levels of inflammatory markers over the short term, this would not necessarily mean that longer term exposure to high levels of air pollution causes chronically elevated levels of systemic inflammation.
Keys strengths of our study were that it was large and nationally representative with a high level of participation and allowed us to examine effects over a wide range of exposures and to incorporate rich data on potential confounders. Using multilevel linear regression models allowed for the possibility that concentrations of fibrinogen and C-reactive protein in people living in 1 postcode sector may be more similar to each other than of people living elsewhere, because they share other risk factors due to their shared geography.
We believe that our method of estimating individual exposure to outdoor air pollution has some advantages. All techniques will be subject to some exposure misclassification. Most commonly, researchers have used single monitoring station data to assign exposure to all individuals living in large areas, which are likely to reflect individual exposure poorly because of the distance from the monitoring station and individual mobility.30 Other researchers have used population density31 or estimated traffic exposure32 in an attempt to reduce exposure misclassification. The disadvantage of this type of approach (based on land use) is that the variables used to improve on fixed site monitoring may not always be the best markers for air pollutant emission densities.
To maximize the accuracy of exposure estimates in our study, we estimated exposure for small geographic areas (postcode sectors) by using comprehensive emission inventories, combined with air dispersion models taking into account meteorologic information.20 The emission inventories were constructed using information on many different influences on air pollution levels (including population density, emissions from roads from a combination of traffic activity data [daily flows for different vehicle types on each major road link], vehicle fleet characteristics [age, relevant emission standard] and emission factors [emissions per kilometer per vehicle], railways, airports, industry, and domestic heating),25 and enabled us to account directly for dispersion and processes in the atmosphere. This meant that estimates of air pollution exposure for every postcode sector took into account a large number of possible influences on air pollution, including its experience of road traffic and other vehicles, factories, housing and population, and the weather, providing added value over monitoring station data or simple land-use methods. An independent comparison of the maps for PM10 and NO2 with data collected by local authorities showed good agreement33 (further information on model verification is provided in eSupplement III, http://links.lww.com/A726). We used a more empirical approach to model ozone concentrations.26 Comparisons with modeling studies suggest that this approach is robust.34
We acknowledge that for people living very close to main roads, proximity to the road may be a better determinant of exposure than modeled levels. People living less than 50 m from a main road have been found to have an increased prevalence of coronary artery calcification, a predictor of coronary events.9 We did not have data on distance from main road of residence. However, our study aimed not to examine the effect of living close to a main road on levels of inflammatory markers, but instead aimed at estimating the effect of average levels of air pollution on population levels of inflammatory markers. People living close to main roads and exposed to very high levels of particulates make up only a very small proportion of the population; in the UK, most of the fall in levels in particulate matter occurs within the first 10 m of the highway.35 If we had been able to analyze distance from main road, most of the participants would have lived significantly further away than 10 m; any effect of C-reactive protein or fibrinogen would have been diluted.
It may be argued that rural postcode sectors are too large for postcode sector level exposure to reflect individual exposure accurately. However, our previous analyses have suggested that more than 97% of the variance in annual mean PM10, NO2, SO2, and summer mean O3 estimates is among postcode sectors, and less than 3% of the variance is within postcode sectors (eSupplement III, http://links.lww.com/A726).
We did not have measures in our study of fine particulates (particulate matter <2.5 μm in diameter [PM2.5]), which are thought to be the most toxic.4 There are very few PM2.5 measurements in the UK (and almost none during our study period), on which to base an exposure assessment. Studies that have compared PM10 and PM2.5 in the UK have generally found strong correlations. Harrison et al36 found strong correlations between colocated hourly PM10 and PM2.5 measurements at 5 monitoring sites in the UK with r2 of between 0.59 and 0.94. The PM2.5/PM10 ratio varied from 0.63 to 0.73. Moreover, some of the most influential studies have found relationships between cardiovascular mortality and PM10 that are similar to those with PM2.5.30
Overall, we believe that our estimates of exposure by postcode sector are more likely to be better measures of true individual exposure than those that have been used to demonstrate associations between air pollution and cardiovascular mortality. In addition, postcode sectors are generally smaller than the geographic units used to assign exposure in many other studies.
Our approach to quantifying chronic exposure to outdoor air pollution assumes that participants had been living at their address for some time, and also that people do not spend significant periods of time in other postcode sectors with different patterns of exposure. We did not have information on duration of time at current address so we cannot take account of this, although it is a reasonable assumption that most participants had not moved from their home in the last year. In Western societies, people spend 16 hours a day on average in their homes,37 so it is likely that the pattern of exposure in the residential, rather than workplace, postcode sector has more influence on health outcomes.
Fibrinogen and C-reactive protein are only 2 of the many possible markers of systemic inflammation. We did not have data on levels of any other markers. However, the evidence linking C-reactive protein and fibrinogen with cardiovascular disease is more extensive and consistent than for any other markers. We acknowledge the possibility that air pollution leads to chronic inflammation by a pathway that does not include C-reactive protein or fibrinogen, but we believe that a significant degree of systemic inflammation usually results in a detectable influence on C-reactive protein, in particular.
Fibrinogen and C-reactive protein levels are highly variable over time, exhibiting seasonal and diurnal variations.27 This variability will be true of any sensitive marker of inflammation and could have increased the size of our standard errors. We did not have repeated measures of fibrinogen or C-reactive protein, so we were unable to assess how well our single measurements reflected chronic averages. A UK cohort study, however, found a considerable degree of within-subject consistency over time in fibrinogen and C-reactive protein levels.38 In our study, we found that C-reactive protein and fibrinogen levels were substantially higher in women, older people, people of lower social class, people with higher body mass index, smokers, people with a history of cardiovascular disease, and people with a history of recent acute illness, which suggests that our measures were good indicators of chronic levels. In addition, C-reactive protein and fibrinogen were correlated in both 1998 and 2003 populations (in both, Pearson's correlation coefficient 0.6).
We do not believe that the inclusion of people with cardiovascular disease or recent acute illness biased our results because our results were similar after excluding these groups.
Among those who provided questionnaire data, we found some differences between people who provided fibrinogen and C-reactive protein measurements and those who did not. The differences were small and so are unlikely to have led to spurious findings of no association. Moreover, the association between air pollution and C-reactive protein or fibrinogen would have had to be very large in those without data for their exclusion to have obscured a real association. Response to the individual questionnaire was very high in participating households, so it is unlikely that this contributed significant bias to the results. Our models did not include a number of variables that were associated with levels of inflammatory markers, in particular prevalence of cardiovascular disease, diabetes mellitus, and rheumatologic disease; and use of lipid-lowering drugs and β-blockers. However, it is unlikely that this could explain the null findings because excluding people with these conditions or taking these medications from the analyses made no difference to the estimates. Other variables that we did not include were occupation, and environmental tobacco smoke, on which we had limited information; we also thought that it was unlikely to be a cause of negative confounding, given the lack of consistent effect of active smoking on the direction and magnitude of the estimates of effect.
In our study, we did not have data on acute variations in air pollution in relation to acute levels of circulating inflammatory mediators, so we were unable to establish whether these acute variations influenced levels of C-reactive protein or fibrinogen. In any case, our analysis at the population (small-area) level was designed to assess the “average” (long-term) degree of elevation, which is what should influence the average (long-term) risk of cardiovascular disease. Short-term elevations in inflammatory markers, even if causally related to cardiovascular disease (which some dispute) would increase cardiovascular event risk only for short periods. Our study, on the other hand, explored whether chronic low-grade systemic inflammation is involved in the mechanism by which chronic higher air pollution exposure could increase risk of long-term, possibly irreversible, effects on the cardiovascular system such as structural changes to blood vessels, long-term changes to neural cardiovascular responses, or long-term coagulability, rather than simply precipitating cardiovascular events. When we designed this study, we thought that it was plausible that chronic air pollution exposure might cause chronic low-grade inflammation leading to higher risk of cardiovascular disease. There is a body of evidence suggesting that inflammation is involved in the development of atheromatous plaques,39 and cohort studies demonstrating associations between fibrinogen10 and C-reactive protein11 and later cardiovascular disease in previously healthy people. However, our study does not support this hypothesis. If low-grade systemic inflammation did link air pollution and the development of the pathology of chronic cardiovascular disease, we would have expected to have found at least a modest geographic relationship between air pollution exposure in the previous months to years and markers of inflammation. Our findings provide some evidence that the geographic and temporal association between outdoor air pollution exposure and later cardiovascular disease demonstrated in cohort studies may not be mediated by chronic systemic inflammation.
1. Dockery DW, Pope CA III, Xu X, et al. An association between air pollution and mortality in six US cities. N Engl J Med
2. Nafstad P, Haheim LL, Wisloff T, et al. Urban air pollution and mortality in a cohort of Norwegian men. Environ Health Perspect
3. Krewski D, Burnett RT, Goldberg MS, et al. Overview of the reanalysis of the Harvard Six Cities Study and American Cancer Society Study of Particulate Air Pollution and Mortality. J Toxicol Environ Health A
4. Committee on the Medical Effects of Air Pollutants. Cardiovascular disease and air pollution-a report by the Committee on the Medical Effects of Air Pollutants
. London: Department of Health; 2006.
5. Brook RD, Franklin B, Cascio W, et al. Air Pollution and Cardiovascular Disease: a Statement for Healthcare Professionals From the Expert Panel on Population and Prevention Science of the American Heart Association. Circulation
6. Miller KA, Siscovick DS, Sheppard L, et al. Long-term exposure to air pollution and incidence of cardiovascular events in women. N Engl J Med
7. Pope CA III, Burnett RT, Thurston GD, et al. Cardiovascular mortality and long-term exposure to particulate air pollution: epidemiological evidence of general pathophysiological pathways of disease. Circulation
8. Kunzli N, Jerrett M, Mack WJ, et al. Ambient air pollution and atherosclerosis in Los Angeles. Environ Health Perspect
9. Hoffmann B, Moebus S, Mohlenkamp S, et al. Residential exposure to traffic is associated with coronary atherosclerosis. Circulation
10. Danesh J, Lewington S, Thompson SG, et al. Plasma fibrinogen level and the risk of major cardiovascular diseases and nonvascular mortality: an individual participant meta-analysis. JAMA
11. Danesh J, Wheeler JG, Hirschfield GM, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med
12. Schwartz J. Air pollution and blood markers of cardiovascular risk. Environ Health Perspect
. 2001;109(suppl 9):405–409.
13. Seaton A, Soutar A, Crawford V, et al. Particulate air pollution and the blood. Thorax
14. Pope CA III, Hansen ML, Long RW, et al. Ambient particulate air pollution, heart rate variability, and blood markers of inflammation in a panel of elderly subjects. Environ Health Perspect
15. Peters A, Frohlich M, Doring A, et al. Particulate air pollution is associated with an acute phase response in men. Results from the MONICA-Augsburg Study. Eur Heart J
16. Diez Roux AV, Auchincloss AH, Astor B, et al. Recent exposure to particulate matter and C-reactive protein concentration in the multi-ethnic study of atherosclerosis. Am J Epidemiol
17. Salvi S, Blomberg A, Rudell B, et al. Acute inflammatory responses in the airways and peripheral blood after short-term exposure to diesel exhaust in healthy human volunteers. Am J Respir Crit Care Med
18. Chen JC, Schwartz J. Metabolic syndrome and inflammatory responses to long-term particulate air pollutants. Environ Health Perspect
19. von Clauss A. Rapid physiological coagulation method in determination of fibrinogen. Acta Haematol.
20. Stedman JR, Bush TJ, Vincent KJ, et al. UK air quality modelling for annual reporting 2003 on ambient air quality assessment under Council Directives 96/62/EC, 1999/30/EC and 2000/69/EC
. Didcot, Oxfordshire: AEA Technology, National Environmental Technology Centre; 2005. Report AEAT/ENV/R/1790.
21. Stedman JR, Kent AJ, Grice S, et al. A consistent method for modelling PM10 and PM2.5 concentrations across the United Kingdom in 2004 for air quality assessment. Atmos Environ
22. Stedman JR, Vincent KJ, Campbell GW, et al. New high resolution maps of estimated background ambient NOx and NO2
concentrations in the UK. Atmos Environ
23. Stedman JR, Goodwin JWL, King K, et al. An empirical model for predicting urban roadside nitrogen dioxide concentrations in the UK. Atmos Environ
24. Jenkin ME. Analysis of sources and partitioning of oxidant in the UK–Part 1: the NOx-dependence of annual mean concentrations of nitrogen dioxide and ozone. Atmos Environ
25. Dore CJ, Watterson JD, Murrells TP, et al. UK emissions of air pollutants 1970 to 1994
. National Atmospheric Emissions Inventory, AEA Technology, National Environmental Technology Centre: Didcot, Oxfordshire; 2007. AEAT/ENV/R/2359.
26. Coyle M, Smith RI, Stedman JR, et al. Quantifying the spatial distribution of surface ozone concentration in the UK. Atmos Environ
27. Rudnicka AR, Rumley A, Lowe GD, et al. Diurnal, seasonal, and blood-processing patterns in levels of circulating fibrinogen, fibrin D-dimer, C-reactive protein, tissue plasminogen activator, and von Willebrand factor in a 45-year-old population. Circulation
28. Stout RW, Crawford VL, McDermott MJ, et al. Seasonal changes in haemostatic factors in young and elderly subjects. Age Ageing
29. Pekkanen J, Brunner EJ, Anderson HR, et al. Daily concentrations of air pollution and plasma fibrinogen in London. Occup Environ Med
30. Pope CA III, Burnett RT, Thun MJ, et al. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA
31. Hoek G, Brunekreef B, Goldbohm S, et al. Association between mortality and indicators of traffic-related air pollution in the Netherlands: a cohort study. Lancet
32. Jerrett M, Burnett RT, Ma R, et al. Spatial analysis of air pollution and mortality in Los Angeles. Epidemiology
33. Stedman JR, Handley C. A comparison of national maps of NO2 and PM10 concentrations with data from the NETCEN ‘Calibration Club.’
Didcot, Oxfordshire: AEA Technology; 2001. AEAT/ENV/R/0725.
34. Hayman G, Thomson C, Abbott J, et al. Ozone Modelling for the Review of the Air Quality Strategy. A report produced for the Department for Environment, Food and Rural Affairs, the Scottish Executive, Welsh Assembly Government and the Department of the Environment Northern Ireland
. Didcot, Oxfordshire: AEA Technology plc; 2006. AEAT/ENV/R/2092 Issue 1.
35. Air Quality Expert Group. Particulate Matter in the United Kingdom
. London: Department of Environment, Food and Rural Affairs; 2005.
36. Harrison RM, Yin J, Mark D, et al. Studies of the coarse particle (2.5–10 μm) component in UK urban atmospheres. Atmos Environ
37. Fuhlbrigge A, Weiss S. Domestic gas appliances and lung disease. Thorax
. 1997;52(suppl 3):S58–S62.
38. Emberson JR, Whincup PH, Morris RW, et al. Extent of regression dilution for established and novel coronary risk factors: results from the British Regional Heart Study. Eur J Cardiovas Prev Rehabil
39. Willerson JT, Ridker PM. Inflammation as a cardiovascular risk factor. Circulation
. 2004;109(suppl 1):II2–10.
Supplemental Digital Content
© 2009 Lippincott Williams & Wilkins, Inc.