Objective Exposure to air pollution is associated with increases in cardiovascular morbidity and mortality. This study was undertaken to determine the effect of diesel exhaust inhalation on heart rhythm and heart rate variability in healthy volunteers and patients with coronary heart disease.
Design and setting Double-blind randomised crossover studies in a university teaching hospital.
Patients 32 healthy non-smoking volunteers and 20 patients with prior myocardial infarction.
Interventions All 52 subjects were exposed for 1 h to dilute diesel exhaust (particle concentration 300 μg/m3) or filtered air.
Main outcome measures Heart rhythm and heart rate variability were monitored during and for 24 h after the exposure using continuous ambulatory electrocardiography and assessed using standard time and frequency domain analysis.
Results No significant arrhythmias occurred during or following exposures. Patients with coronary heart disease had reduced autonomic function in comparison to healthy volunteers, with reduced standard deviations of the NN interval (SDNN, p<0.001) and triangular index (p<0.001). Diesel exhaust did not affect heart rate variability compared with filtered air (p>0.05 for all) in healthy volunteers (SDNN 101±6 vs 91±6, triangular index 20±1 vs 21±1) or patients with coronary heart disease (SDNN 47±5 vs 38±4, triangular index 8±1 vs 7±1).
Conclusions Brief exposure to dilute diesel exhaust does not alter heart rhythm or heart rate variability in healthy volunteers or well-treated patients with stable coronary heart disease. Autonomic dysfunction does not appear to be a dominant mechanism that can explain the observed excess in cardiovascular events following exposure to combustion-derived air pollution.
- Air pollution
- diesel exhaust
- heart rate variability
- autonomic function
- autonomic regulation
- heart rate variability
- Holter ECG
Statistics from Altmetric.com
- Air pollution
- diesel exhaust
- heart rate variability
- autonomic function
- autonomic regulation
- heart rate variability
- Holter ECG
Observational and epidemiological studies consistently report that exposure to air pollutants is associated with excess cardiovascular morbidity and mortality,1 and may be an important modifiable risk factor for cardiovascular disease.2 If public health measures are to be implemented to reduce this risk, a better understanding of the mechanisms and components of urban air pollution responsible for these observations is urgently required.
Short-term increases in air pollution exacerbate cardiorespiratory disease leading to hospitalisation for conditions including acute myocardial infarction3 and deaths from coronary heart disease, heart failure and arrhythmia.2 Furthermore, in patients with implanted cardiac defibrillators there appears to be a relationship between ambient particulate matter (PM) and the incidence of ventricular tachycardia and fibrillation.4–6 These associations are strongest for fine particulate air pollutants, with emissions from road traffic implicated both directly and indirectly in these reports.
There is an important relationship between autonomic regulation of the cardiac cycle and cardiovascular mortality.7 Variation in the interval between consecutive heart beats, or heart rate variability, is controlled by the contrasting effects of the sympathetic and parasympathetic nervous systems. Reduction in heart rate variability reflects either an increase in sympathetic drive or a decrease in vagal parasympathetic tone. Reduced heart rate variability increases the risk of cardiovascular morbidity and mortality in both healthy individuals8 and patients following myocardial infarction.9
Several panel studies have reported associations between measures of heart rate variability and high ambient PM.10–13 There is marked heterogeneity in the nature, magnitude, direction and duration of these associations between studies with no clear and consistent effect apparent. Differences may in part be due to imprecision in the measurement of pollution exposure and the effect of potential confounding environmental and social factors. Controlled exposures of air pollutants can help to address these shortcomings by providing a precisely defined exposure in a regulated environment. Using a carefully characterised exposure system, we have previously shown in healthy volunteers that exposure to dilute diesel exhaust causes lung inflammation,14 depletion of airway antioxidant defences15 and impairment of vascular and fibrinolytic function.16 Furthermore, we have described ischaemic and prothrombotic effects in patients with coronary heart disease.17 To date, there have been no controlled exposure studies to assess the direct effects of diesel exhaust on heart rate variability in patients with coronary heart disease.
We aimed to assess the effect of dilute diesel exhaust inhalation on heart rhythm and heart rate variability in healthy volunteers and an ‘at-risk’ population of patients with stable coronary heart disease.
Fifty-two men participated in this study, which was performed with the approval of the local research ethics committee, in accordance with the Declaration of Helsinki and with the written informed consent of all volunteers. Thirty-two healthy non-smoking volunteers and 20 patients with stable coronary artery disease were recruited.
Healthy non-smoking volunteers (age 20–38 years) on no regular medications and patients with coronary heart disease were recruited from the University Hospital, Umeå, Sweden. All patients (age 51–67 years) had proven coronary heart disease with a previous myocardial infarction (>6 months previously) treated by primary angioplasty and stenting, and were receiving standard secondary preventive therapy. Patients with angina pectoris (Canadian Cardiovascular Society grade ≥2), diabetes mellitus, uncontrolled hypertension, renal or hepatic failure, or those with unstable coronary disease (acute coronary syndrome or unstable symptoms within 3 months) were excluded. All volunteers were invited to a prestudy screening visit for exercise stress testing and patients unable to achieve stage 2 of the Bruce protocol were excluded. Current smokers and those with asthma, significant occupational exposure to air pollution or an intercurrent illness were excluded from the study. In patients, regular medications were continued throughout the study, with the exception of ACE inhibitor therapy that was withdrawn 7 days before each exposure.
All subjects underwent an identical randomised double-blind crossover study design, with healthy volunteers and patients attending at 08.00 h on two occasions at least 2 weeks apart for controlled exposure to dilute diesel exhaust or filtered air. Each subject was exposed for 1 h in a specially built diesel exposure chamber according to a previously described standard protocol.14 During each exposure, subjects performed two 15 min periods of exercise on a bicycle ergometer separated by two 15 min periods of rest. The ergometer workload required to achieve a minute ventilation of 25 l/min/m2 for each healthy volunteer and 15 l/min/m2 for each patient with coronary heart disease was determined to ensure that subjects within each group received a similar exposure.
All subjects were fitted with Holter electrocardiographic monitors (Reynolds Medical Lifecard, Delmar Reynolds, Hertford, UK) prior to each exposure. ECG monitoring was continued for 24 h following the start of the exposure. All subjects remained indoors for the 24 h period following exposure to minimise additional exposure to ambient air pollution. Subjects were asked to abstain from alcohol for 24 h and from food, tobacco and caffeine-containing drinks for at least 4 h before each exposure.
To assess the acute effects of exposure on heart rate variability, 15 healthy volunteers and all patients were asked to rest supine in a quiet temperature-controlled room maintained at 22–24°C for 20 min immediately before and 2 and 6 h after the start of each exposure. All volunteers also underwent blood pressure measurements and a vascular assessment after the final rest period as detailed in our previous publications.16 17
Diesel exhaust exposure
The diesel exhaust was generated from an idling Volvo diesel engine (Volvo TD45, 4.5L, 4 cylinders, 680 rpm) from Swedish Low Sulphur Gasoil E10 (Preem, Gothenburg, Sweden) as described previously.14 16 Over 90% of the exhaust was shunted away and the remaining part was diluted with ambient filtered air heated to 20°C (humidity ∼50%) before being fed into a whole body exposure chamber at a steady state concentration. Air in the exposure chamber was continuously monitored with exposures standardised using online measurements of nitrogen oxide concentrations (NOx) to deliver approximately 1.2×106 suspended particles/cm3 at a particulate concentration of 300 μg/m3. There was little variation in NOx (4.45±0.02 parts per million, ppm), NO2 (0.92±0.02 ppm), NO (3.35±0.02 ppm), CO (2.97±0.08 ppm) and total hydrocarbon (2.50±0.09 ppm) concentrations between exposures.
ECG recordings were analysed using the Reynolds Medical Pathfinder Digital 700 Series Analysis System (Delmar Reynolds). An experienced single operator, blinded to both subject characteristics and exposure, verified any abnormal rhythms and performed manual editing of aberrant beats and electrical interference prior to generating RR data tables. Where less than 95% of the RR data was valid, the subject was excluded and the recording was not analysed further. RR data were analysed using the HRV Tools software package (Delmar Reynolds) to determine time and frequency components of heart rate modulation over the entire 24 h period and during the final 5 min of each rest period prior to exposure and 2 h and 6 h following the start of each exposure.
Standard time domain measures were calculated including the mean NN interval (time interval between consecutive sinus beats), SD of NN interval values (SDNN, an index that expresses overall variability), percentage successive NN interval differences >50 ms (PNN50), root mean square of successive NN interval differences (RMSSD) and the triangular index (an estimate of overall heart rate variability). SDNN, PNN50 and RMSSD are measures of high frequency variation mediated primarily by the vagus nerve. Frequency domain analysis determined the low frequency (LF; 0.1 Hz) and high frequency (HF; 0.25 Hz) components of the power spectrum in absolute values of power (ms2). LF and HF were also expressed in normalised units (LFn and HFn) to account for variation in the total power and very low frequency components, as well as the HF/LF ratio.
Continuous variables are reported as mean±SEM. Statistical analyses were performed with GraphPad Prism (Graph Pad Software, USA) using analysis of variance (ANOVA) with repeated measures and the two-tailed Student t test where appropriate. Statistical significance was taken at p<0.05.
Although one healthy volunteer and one patient were excluded from the analysis due to interference on the ECG recording, 98% and 99% of the data were valid in the remaining healthy volunteers (n=31) and patients (n=19) respectively. Patients were middle aged (60±1 years) men who were on typical cardiac medication (table 1). Patients and healthy controls did not experience any symptoms or serious arrhythmias during either exposure or during the 24 h study period.
Heart rate and heart rate variability over the 24 h study period were reduced in patients with coronary heart disease compared with young healthy controls (table 2). Inhalation of dilute diesel exhaust for 1 h did not affect time or frequency domain measures of heart rate variability over the 24 h period in either healthy volunteers or patients with coronary heart disease (table 2).
Baseline measures of heart rate and heart rate variability averaged over a 5 min interval immediately before both exposures were similar in each group (tables 3 and 4). Blood pressure was not affected by exposure (data not shown). There were no differences in either time or frequency domain measures of heart rate variability at 2 or 6 h in healthy volunteers. In patients with coronary artery disease there was an increase in SDNN at 2 and 6 h compared with baseline (p=0.029), but this was not affected by exposure to diesel exhaust (p=0.483).
Exposure to dilute diesel exhaust for 1 h did not affect heart rhythm or heart rate variability in healthy volunteers or patients with coronary heart disease. We suggest that the induction of autonomic dysfunction and arrhythmia is unlikely to explain the association between combustion-derived air pollution and adverse cardiovascular events. Our findings contrast with several observational studies that report associations between ambient air pollution and heart rate variability, and this apparent discrepancy requires further discussion.
Controlled exposure studies
Controlled exposures of air pollutants provide a precisely defined exposure in a regulated environment and can overcome many of the potential biases and confounders inherent to observational studies. Our findings are consistent with the only previous study of diesel exhaust exposure on autonomic function by Peretz et al who also did not identify any reproducible effects on heart rate variability in 16 young healthy volunteers.18 Taken together with our findings, we believe this clearly demonstrates that diesel exhaust inhalation does not impact on heart rate variability or induce autonomic dysfunction.
We believe that our model of exposure is relevant both in composition and magnitude of exposure for the assessment of the acute effects of combustion-derived air pollution on autonomic function and heart rhythm in man. This is the largest controlled exposure study to date, and we have studied the effects of exposure in a cohort of patients who may be particularly susceptible to the effects of air pollution. The crossover study design employed minimises the potential for confounding by other environmental factors and within-subject comparisons increase the power to detect small changes in autonomic function. Despite this, we were unable to demonstrate an acute or persistent effect of exposure to dilute diesel exhaust on heart rhythm or heart rate variability, whether this was assessed using time or frequency domain analysis in either healthy subjects or patients with coronary heart disease.
While more than 30 epidemiological studies report associations between air pollution and changes in heart rate variability, there is considerable heterogeneity between these studies in the nature, magnitude, direction and duration of the associations between PM and heart rate variability (table 5). Liao et al were the first to report an association between fine particulate air pollution (PM2.5) and heart rate variability in a panel of elderly subjects.11 The authors considered their finding somewhat exploratory, but the analysis revealed an inverse correlation between same day PM2.5 concentrations and both the HF and LF power components of heart rate variability. Reduced heart rate variability is associated with an increase in sudden cardiac death and all-cause mortality in survivors of acute myocardial infarction.9 Liao et al hypothesised that an effect of PM exposure on the autonomic control of heart rate and rhythm may explain the association between PM and adverse cardiovascular outcomes. Subsequently, numerous panel studies have explored this mechanistic hypothesis by examining the associations between levels of different air pollutants and changes in heart rate variability or incidence of cardiac arrhythmia.
Most observational studies report negative associations between fine particulate air pollution and either time or frequency domain parameters of heart rate variability, but there is considerable disagreement between these studies. Differences may be due to variation in the composition of PM or the effect of confounding environmental and social factors. For example, in a multicentre study, Timonen et al found the effects of PM on heart rate variability in patients with coronary heart disease were dependent on local sources of PM.19 Increases in PM2.5 were associated with a reduction in HF power in Finland, but a similar increase in PM2.5 was associated with an increase in HF power in Germany. In the study by Wheeler et al, exposure to PM2.5 was associated with reductions in SDNN in patients with previous myocardial infarction.20 However, the direction of this effect was reversed in patients with chronic obstructive pulmonary disease where SDNN increased following exposure to ambient PM2.5.20
Differences in statistical analyses make comparison of the effect size between observational studies difficult. In the largest study to date, Liao et al report weak associations between particulate air pollution and heart rate variability in a population of 6784 healthy adults, suggesting a small role for PM in modulating autonomic function.21 An increase in 1SD of PM10 (11.5 μg/m3) in the 3-day period prior to assessment was associated with an increase in heart rate of 0.32 beats/min and a decrease in SDNN of 1.03 ms. Previous studies showing that heart rate variability is an important prognostic marker have assessed the relationship between heart rate variability and long-term outcome.9 Whether transient hourly or daily fluctuations in heart rate and heart rate variability of this magnitude impact on cardiovascular outcome is unknown.
The implication from these observational studies is that reduced heart rate variability following exposure to air pollution may predispose to serious tachyarrhythmias resulting in hospitalisation or sudden cardiac death. Direct evidence that PM may be a trigger for arrhythmia is derived from studies of high-risk patients with implantable cardioverter defibrillators (ICD). In a pilot study, fine particulate and other traffic-derived air pollutants were associated with an increase in the number of defibrillator interventions among 100 patients with ICDs.4 However, in a larger more complete analysis with longer follow-up, there was no increase in the risk of ventricular arrhythmia unless the analysis was restricted only to those patients requiring frequent defibrillator interventions.5 Indeed, in a much larger study, Anderson et al recently reported a fixed stratum case crossover analysis in 705 patients who experienced 5462 activation days over an average of 1200 days observation in London. Overall they concluded that there was little evidence of an association between air pollution and activation of ICDs.22
While controlled exposure to dilute diesel exhaust is a good model for studying the effects of combustion-derived air pollution and diesel exhaust is a major source of ambient ultrafine particles, it is important to appreciate that ambient air pollution contains a range of particulate pollutants from a variety of atmospheric sources. Notably, diesel exhaust does not contain appreciable quantities of atmospheric metals and there is some experimental evidence to suggest that metals may modify the effect of PM on the autonomic nervous system.23 In a cohort of boiler construction workers exposed to high concentrations of PM and metals, cardiac autonomic function was associated with six common atmospheric metals including vanadium and lead.12 Furthermore, circulating concentrations of common atmospheric metals have been found to correlate with measures of heart rate variability in healthy persons without occupational exposure.24 Consistent with this hypothesis, exposure to pure ultrafine carbon particles does not affect heart rate variability in either healthy volunteers25 or patients with coronary heart disease,26 whereas exposure to concentrated ambient particles with a range of organic and inorganic components may influence autonomic function.27 The effects following exposure to concentrated ambient particles are more pronounced with fine particles28 and in elderly individuals,29 but no study to date has attributed these differences to the metal content of PM. It appears that, even when controlling for potential environmental confounders and individual subject differences through controlled exposures, particle composition is a major determinant of the health effects of PM.
While we do not believe limitations in our study design can explain our negative findings, we acknowledge that a number of relevant factors may have influenced the outcome of our studies. Given potential safety concerns, we recruited patients who had stable and symptomatically well-controlled coronary heart disease on optimal medical therapy, which included maintenance beta-blocker therapy in the majority of patients. The beneficial effects of beta-blockade in patients with coronary artery disease and heart failure are well established and are thought to be mediated in part through enhanced cardiac vagal control,30–32 with treatment increasing HF power by as much as 50%.33 It is possible that an adverse effect of diesel exhaust on autonomic function could have been masked by beta-blockers or other pharmacological therapies in our patients with coronary heart disease. Indeed, a recent reanalysis of the study by Timonen19 suggests that the use of beta-blockers in patients with coronary heart disease may have modified the association between PM2.5 and heart rate variability, partly explaining the inconsistencies identified in the original analysis.34 However, exposure to diesel exhaust did not affect time or frequency domain measures of heart rate variability in our young healthy volunteers, none of whom were on regular medication, suggesting that the use of cardiovascular therapies is not the primary explanation for the absence of an effect of diesel exhaust on autonomic function in our patients with coronary heart disease.
We chose to assess heart rate variability under controlled conditions 1 h following the end of the exposure to minimise any effect of exercise during the exposure on heart rate and heart rate variability. We cannot discount the possibility that exposure to diesel exhaust may have caused an immediate effect on autonomic function during the exposure itself or within the first 1 h following the exposure. While it is possible that our study was insufficiently powered to detect small effects on autonomic function, the study size was larger than any previous human exposure study addressing the effect of particle exposure on autonomic function.
Brief exposure to dilute diesel exhaust does not alter heart rhythm or heart rate variability in healthy volunteers or patients with coronary heart disease on optimal cardioprotective medication. While we cannot exclude an immediate effect during exposure, diesel exhaust did not impair autonomic function in the hours following exposure or over the 24 h period. Our findings from a carefully controlled exposure to an important air pollutant suggest that autonomic dysfunction may not be the central mechanism with which to explain the observed excess in cardiovascular events following exposure to road traffic or combustion-derived air pollution.
The authors thank Frida Holmström, Annika Johansson, Margot Johansson, Veronica Sjögren, Maj-Cari Ledin, Chris Llewellyn, JoanHenderson and the staff in the Department of Respiratory Medicine and Allergy, Umeå and Wellcome Trust Clinical Research Facility, Edinburgh for their assistance with these studies.
See Editorial, p 519
Linked articles 212183.
Funding NLM is supported by an Intermediate Clinical Research Fellowship from the British Heart Foundation (FS/10/024/28266); British Heart Foundation Programme Grant (RG/10/9/28286); National Health Service Research and Development Fund (SPG2005/27); Swedish Heart Lung Foundation; Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS); Swedish National Air Pollution Programme; Swedish Emission Research Programme; Heart and Lung Associations in Sollefteå and Örnsköldsvik, and County Council of Västerbotten, Sweden.
Competing interests None.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the Regional Ethical Review Board, Sweden.
Provenance and peer review Not commissioned; externally peer reviewed.
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.