Several epidemiological studies suggest that diabetes confers increased risk for health effects from air pollution exposure [,,]. We previously reported that UFP exposure transiently increased blood platelet activation and levels of circulating von-Willebrand factor [], consistent with effects on vascular endothelium that promote coagulation. Diabetes also causes autonomic dysfunction, associated with reductions in both time and frequency domain variables of HRV []. Impaired parasympathetic autonomic function reduces vagal tone, which can blunt the high-frequency HRV response. As shown in Table 1, ECG parameters at baseline in our subjects with diabetes differed from healthy, younger, nondiabetic subjects in our prior studies. These findings indicate that cardiac physiology and autonomic control of the heart are altered in people with diabetes, potentially increasing susceptibility to air pollution.
In otherwise healthy people with type 2 diabetes, we found that inhalation of elemental carbon UFP while resting caused a delayed increase in heart rate and a reduction in HF HRV, in comparison with clean air exposure. There were also small non-significant reductions in pNN50, rMSSD, and SDNN associated with UFP exposure. We found no convincing effects on parameters of cardiac repolarization or cardiac rhythm. In the prolonged recordings, there were small ST depressions in all 3 of the precordial leads that were analyzed, but none of the changes were statistically significant. These findings provide further evidence of an effect of UFP inhalation on the autonomic nervous system in people with diabetes, with reductions in vagal relative to sympathetic influences.
The UFP influence on heart rate was delayed in our study, and appears to be an accentuation of the increase in heart rate that occurred as the subjects left the Clinical Research Center and returned to their home environment. As shown in Figures 1A and 2A, the mean heart rate increased 24 and 48 hours after both air and UFP exposure, probably reflecting increased physical activity and stress associated with the subjects’ return home. The increase in heart rate was greater following UFP exposure then air. The decreases in HF HRV (Figure 1D), and the trends toward decreases in the other markers of HRV associated with UFP exposure (Figures 1B,1C, and 2B,2C,2E), were most pronounced during the night following the exposure, when subjects were presumably sleeping. Parasympathetic influences predominate during sleep, slowing the heart rate and increasing heart rate variability []. In our study, this increase in parasympathetic tone appeared to be blunted following UFP exposure. We postulate that, in diabetics who have abnormal autonomic function, UFP exposure further shifts autonomic balance toward sympathetic and away from parasympathetic influences. The delay in the effects on heart rate suggests this is not a direct effect of particles on nerve endings within the respiratory epithelium, but a response to a cascade of events, the nature of which is incompletely understood at this time.
One could speculate regarding a number of pathways for delayed effects on heart rate. Some of our subjects were obese and could have had sleep apnea. It is possible that the brief UFP exposure altered sleep dynamics or worsened sleep-disordered breathing in the nights immediately following exposure, thereby blunting the normal increase in parasympathetic tone that occurs during sleep []. Increased systemic inflammation as a result of UFP exposure could alter autonomic tone, although we found no increases in the circulating inflammatory marker C reactive protein in these subjects []. The small, nonsignificant decreases in the ST segment (Figure 3) could be consistent with an effect on the myocardium, as a result of subtle changes in cardiac perfusion, afterload, or preload. Changes in atrial repolarization could also affect the ST segment, but this is unlikely in the absence of effects on ventricular repolarization (T wave and QTc). The mean increases in heart rate were small (approximately 3 beats per minute) and not clinically important for people without heart disease. However, even small effects on heart rate may be clinically important. Most episodes of cardiac ventricular arrhythmia, or myocardial ischemia, are preceded by an increase in heart rate []. Increases in heart rate require greater myocardial oxygen consumption, leading to myocardial ischemia in the presence of obstructing coronary artery disease. A faster heart rate shortens left ventricular filling times, which may increase left atrial filling pressures, leading to pulmonary edema in patients with systolic or diastolic left ventricular failure. Resting heart rate is a predictor of all-cause and cardiovascular mortality [], with an increase in mean heart rate of 10 beats/min associated with a 9.8% increase in all-cause mortality among patients with coronary artery disease []. In middle-aged subjects without clinical heart disease, increased night-time heart rate is predictive of increased mortality and cardiovascular risk after adjustment for other cardiovascular risk factors []. Our previous studies in healthy subjects [-] suggest that UFP inhalation subtly alters both systemic and pulmonary vascular function. Such effects in people with critical cardiac disease could contribute to the observed associations between PM exposure and cardiovascular morbidity and mortality.
Most of the ECG measures in this study did not differ significantly between air and UFP exposure, including time-domain indicators of HRV (rMSSD, SDNN, pNN50), cardiac repolarization (QTc, T wave amplitude), ischemia (ST segment), or arrhythmia. The autonomic dysfunction associated with diabetes may have reduced the ability of these subjects to respond with further reductions in HRV. Also, subjects with known cardiovascular disease or coronary artery disease were excluded.
Clinical exposure studies have shown variable effects of PM on ECG parameters. Some epidemiology and panel studies have shown small effects of PM on heart rate [-], but most have not. We previously studied ECG changes after exposure of healthy subjects to laboratory generated ultrafine carbon particles and clean air []. We found no effects on heart rate, and generally non-significant effects on HRV, cardiac repolarization, and the ST segment, with trends suggesting increased parasympathetic tone with UFP exposure. Samet et al. [] studied 19 young, healthy subjects exposed to concentrated ambient ultrafine particles and clean air, with intermittent exercise. They found increased frequency-domain markers of HRV, indicating elevated, rather than reduced, vagal (parasympathetic) input to the heart, consistent with our studies with laboratory generated particles in healthy subjects. In contrast, Gong et al. [] found no convincing effects of concentrated ambient UFP, with intermittent exercise, on heart rate or HRV in healthy and asthmatic subjects. Controlled clinical studies in young, healthy subjects of concentrated ambient fine particles, and diesel exhaust [,], have shown no consistent effects on HRV. However, studies in elderly subjects have shown reductions in HRV in response to concentrated ambient fine particles []. Most clinical studies have not looked beyond 24 hours after exposure.
The bulk of evidence that PM exposure reduces HRV comes from panel studies of exposure to ambient air pollution, involving older subjects and those with heart disease. For example, Henneberger and colleagues [] found air pollution effects on cardiac repolarization duration, morphology, and variability in 56 male patients with ischemic heart disease. Pope et al. [] found reductions in HRV associated with increases in concentrations of particulate matter less than or equal to 2.5 μm (PM2.5) in elderly subjects. Rich et al. [] studied 76 patients in a cardiac rehabilitation program who had a recent myocardial infarction or unstable angina. Exposures to fine particles and UFP were associated with decreases in parasympathetic modulation, prolongation of late repolarization duration, increased blood pressure, and systemic inflammation. In 28 elderly subjects in Boston, ambient PM2.5 exposure was associated with decreased rMSSD and pNN50 []. There were stronger effects of black carbon on SDNN in subjects with previous myocardial infarction. Associations with heart rate were not reported. Exposures to ambient air pollution while riding in a car reduced HF HRV in 21 people with type 2 diabetes [], although associations of changes in HF HRV with individual pollutant concentrations in the vehicle were not significant.
To the best of our knowledge, this is the first clinical study of the cardiac effects of exposure to UFP in people with diabetes. The strengths of our study include use of a potentially susceptible, under-studied population; the double-blind, randomized, crossover design; housing of the subjects in a controlled environment overnight prior to exposure; and controlled exposures to well-characterized elemental carbon UFP as a surrogate for ambient UFP. In addition, continuous ECG recordings for 48 hours allowed assessment of delayed effects. The limitations include those inherent in human clinical studies [], with a relatively limited number of study subjects, short term exposures, and possible confounding effects of ambient pollutant exposures prior to the experimental exposures. Exposures were performed at rest, and effects may differ with exercise. In addition, our study utilized laboratory generated pure elemental carbon UFP, which differ from ambient UFP. Our elemental carbon particles may underestimate the potential effects of ambient UFP, which represent a complex mixture of chemical species, including organics. However, we feel that studies such as ours using surrogates for ambient UFP have a role in helping to understand UFP health effects. Ultrafine particle concentrators do not efficiently concentrate the smallest of the ambient UFP size fraction, and probably cause alterations in surface chemistry of the particles that do get concentrated. Soluble particles do not get concentrated, and volatile surface components are lost. Thus, concentrated particles must also be considered imperfect surrogates of ambient UFP. The other alternative is studying people undergoing real-world exposures, as in panel studies. Here the exposures involve complex mixtures of pollutants that include larger particles and gases that vary continuously in concentration and composition, and it is impossible to completely sort out the role of UFP in causing any observed effects. Understanding of health effects of UFP exposure requires synthesis of data from various approaches, including clinical studies of laboratory generated and concentrated ambient UFP, panel studies of ambient exposures, epidemiology, and animal exposure studies.
While exposure mass concentrations of ~50 μg/m3 used in this study were relevant to ambient PM exposures, the exposure number concentrations were substantially higher than those in most ambient settings. Nevertheless, UFP number concentrations on busy highways can reach peaks within an order of magnitude of those used in this study [].