Statistics from Altmetric.com
Concern over the health effects of air pollution has existed for many decades, arguably peaking in the UK with the infamous London smog of 1952, which caused thousands of deaths and led to the passing of the Clean Air Act in 1956 and similar legislation elsewhere. There have been very large reductions in ambient pollution levels in most high income countries in the decades since. However, even at the relatively low pollutant levels seen more recently in these settings, the epidemiological evidence for adverse effects on human health is clear.
Early time-series studies in the 1990s demonstrated an effect of short-term changes in the levels of pollutants, in particular particulate matter, on overall mortality.1 Subsequent studies of cardiovascular mortality and morbidity suggested that both day-to-day changes in pollutant levels and longer term exposure affect risk. Our recent systematic review found compelling evidence for an effect of air pollution on myocardial infarction specifically.2 A major review of the epidemiological evidence on air pollution and cardiovascular disease more generally, conducted for the UK Department of Health, concluded that ‘a large number of time-series studies show very clearly that, with few exceptions, all of the commonly measured pollutants (particles, ozone, sulphur dioxide, nitrogen dioxide and carbon monoxide) are positively associated with increased mortality and hospital admissions for cardiovascular disease’.3
These consistently observed effects clearly require some mechanistic explanation. To date, the main candidate hypotheses have held that pollutant exposure increases the risk of cardiovascular events via disruption of the autonomic nervous system, an inflammatory response, or both.
Out of control?
One of the leading theories postulates that pollution exposure leads to disturbances in the autonomic control of the heart, reflecting altered balance of the sympathetic and parasympathetic nervous systems. Inhalation of pollutants is conjectured to initiate such a response via cytokine release and a subsequent autonomic stress response, or by direct stimulation of receptors in the upper airway.3
Testing this theory in observational studies presents a challenge: how can autonomic control be evaluated? Measurement of heart rate variability (HRV; changes in the time interval between heart beats) has become a preferred way to quantify cardiac autonomic control, and a number of observational studies examining this parameter suggest that increasing levels of particulate air pollution are associated with decreases in HRV,4 5 although the evidence is somewhat mixed.6 Reduced HRV is itself associated with an increased risk of cardiac events.7 A few epidemiological studies have taken advantage of the detailed data available from implantable cardioverter defibrillator devices. Such analyses are naturally limited to selected subgroups of patients who have received an implantable cardioverter defibrillator and are therefore at high risk of developing serious cardiac arrhythmias. Nevertheless, among this group, arrhythmias recorded by the devices appeared to be more frequent following increases in ambient exposure to fine particles, NO2, CO and black carbon in Boston,8 though the findings were not convincingly replicated in a similar London-based study.9
However, there are common limitations to these observational studies: exposure data are usually collected from central monitoring sites and therefore do not reflect personal exposure accurately; only a handful of specific pollutants are routinely measured; there may be questions of residual confounding (eg, by ambient temperature or activity); and, given the large number of collinear pollutant exposures, disentangling pollutant-specific associations can be problematic.
Experimental evidence is therefore needed. In their paper published in this issue of Heart, Mills et al (see page 544) report on a timely double-blind randomised crossover study in which 52 men (32 healthy and 20 with stable coronary artery disease) were exposed for 1 h to dilute diesel exhaust or filtered air.10 HRV was assessed 1 h following exposure. Not surprisingly those with prior coronary heart disease had significantly impaired autonomic function compared with healthy volunteers, but the comparison of interest revealed no change in heart rhythm or HRV on exposure to diesel exhaust versus filtered air for either group. The authors report that their results are in keeping with the negative findings in the only previous controlled exposure study investigating the effects of diesel exposure on autonomic function.11
Is this a fatal blow for the autonomic dysfunction hypothesis? Mills et al conclude that it does not appear to be a dominant mechanism explaining air pollution effects on cardiovascular disease risk; their study indeed provides some of the best quality evidence to date on the question, and strong evidence against the hypothesis. However, it is not entirely definitive. The authors acknowledge some important limitations: given the experimental design, very fast-acting (<1 h) effects could not be excluded; 75% of the higher risk group of volunteers were on β-blockers which could have mitigated any pollution effect; and the particular choice of dilute diesel exposure is unlikely to reflect the full range of pollutants to which people are exposed in everyday life. Since several observational studies have suggested pollutant effects on HRV (usefully summarised in the paper), an explanation is needed for the discrepancy between these epidemiological data and negative results in controlled conditions. Further studies examining exposure to different pollutant mixes in different populations will be needed to definitively exclude autonomic dysfunction as an important mechanism.
An inflammatory issue?
A second key hypothesised mechanism for air pollution effects on cardiovascular disease is that deposition of fine particulate matter, through increased oxidative stress, causes an alveolar inflammatory response, with release of mediators to the systemic circulation which could increase the risk of cardiovascular events (primarily heart attacks and strokes) through effects on blood coagulability and thrombotic risk.12
Direct evidence for inflammation associated with pollution increases is mixed; again, there is some disagreement between observational and experimental evidence. Increases in levels of C-reactive protein5 and other inflammatory markers13 at times of higher ambient pollution have been observed, yet a number of experimental studies have reported no clear systemic inflammatory response on controlled exposure to fine and ultrafine particles,14 dilute diesel exhaust15 and concentrated ambient particles.16
The prediction of increased coagulability and decreased fibrinolysis is supported more consistently by the data. Plasma viscosity increased among individuals exposed to a severe air pollution episode in Germany in 1985.17 Controlled exposure experimental studies have demonstrated exposure to concentrated environmental particles leading to an increase in plasma fibrinogen levels in healthy volunteers,18 and exposure to dilute diesel leading to a decrease in fibrinolytic capacity, specifically through reduced release of tissue plasminogen activator.15 On the other hand, findings are not entirely consistent: a further controlled study found that 2 h of concentrated ambient particulate exposure had no effect on fibrinolytic function among either healthy middle-aged volunteers or patients with prior coronary heart disease, despite delivery at 3–5 times US Environmental Protection Agency National Ambient Air Quality Standards.16 This result highlights that choice of pollutant composition in controlled exposure studies might be crucial.
Conclusions and research challenges
Alternative mechanisms have been proposed to compete with (or complement) the two leading theories described above. Perhaps the simplest of all invokes the ability of inhaled ultrafine particles to pass directly into the systemic circulation, postulating direct physical effects on the heart and vascular system19; another possibility is suggested by a study in which controlled exposure to a mixture of concentrated ambient particles and ozone in humans led to arterial vasoconstriction.20 It is possible that exposure to air pollution may affect the risk of acute cardiac events through more than one mechanism. Yet promising mechanistic hypotheses appear in some cases to be undermined by experimental evidence. Neither the observational nor the experimental data alone can be conclusive; both study types have their limitations, and the variation in research findings doubtless in part reflects the methodological complexity of studying mechanistic questions.
Experimental studies arguably allow for a more finely tuned examination of the various mechanistic hypotheses that have been put forward but, given the many permutations of pollutant mixtures, participant demographics and proposed mechanistic pathways, several more rigorous controlled exposure studies along with large population-based analyses with better exposure data are likely to be needed before the picture is clear.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.