Magari, Shannon R.1; Schwartz, Joel2; Williams, Paige L.3; Hauser, Russ1; Smith, Thomas J.4; Christiani, David C.1,5
Anumber of health outcomes have been associated with particulate air pollution. 1–4 Cardiopulmonary disease has been identified as one of the main sources of morbidity and mortality associated with exposure to particulates. 4–7 One potential mechanism for those effects involves alterations in cardiac autonomic function. To date, those investigations have used elderly subjects and have not included personal exposure measurements. This study was designed to examine the magnitude of the effect of continuous personal measurements of environmental particulate matter with a mean aerodynamic diameter ≤2.5 μm (PM2.5) on cardiac autonomic function, using standard heart rate variability (HRV) parameters.
The mechanisms responsible for the cardiac morbidity and mortality associated with particulate air pollution have not been elucidated fully. One potential mechanism is through the disruption of the autonomic nervous system. 8 The alteration in the autonomic nervous system may be induced directly through a sympathetic stress response 9 or indirectly by inflammatory cytokines produced in the lungs and released into the circulation. 10,11 These proposed mechanisms have been developed in part through the investigation of compromised animal models. Watkinson et al.12 have induced arrhythmias in rat models of pulmonary hypertension after exposures to particles.
Cardiac autonomic function can be characterized using several standard heart rate variability measures in the time domain. Standard 5-minute HRV measurements can be collected noninvasively and afford the researcher insights into the function of the autonomic nervous system. 13 The 5-minute standard deviation (SD) of normal-to-normal intervals (SDNN) is used in this study as the main HRV indicator, representing all variation occurring within the 5-minute interval. It provides an overall measure of the parasympathetic and sympathetic components of the cardiac autonomic nervous system.
The clinical value of HRV measures has been studied in survivors of myocardial infarction. Low heart rate variability in the months after a myocardial infarction has been associated with increased mortality. 14,15 In general population cohorts, low HRV has also been associated with both increased risk of cardiac events and all-cause mortality. 16–18
We used personal PM2.5 measurements obtained with a continuous real-time monitor to investigate the effects of particulate matter on cardiac autonomic function in a group of relatively young boilermaker construction workers on a day when they were not working. Moving PM2.5 averages served as the exposure metrics in this study and were generated using the continuous 5-minute PM2.5 data. We regressed both the 5-minute SDNN and 5-minute heart rate on these exposure metrics and on the function of the cardiac autonomic nervous system.
The Institutional Review Board of the Harvard School of Public Health approved this study, and written informed consent was obtained. The study population consisted of 20 male (apprentice/journeyman) boilermaker construction workers. These 20 subjects, a self-selected subset from a larger occupational cohort, consented to monitoring on a day when they were not working. We used a self-administered questionnaire to collect information on medical history, including respiratory and cardiac systems, and on subjects’ use of prescription and nonprescription medications.
Exposures included ambient air pollution and multiple indoor sources of pollutants, including tobacco smoke, cooking, and cleaning. Monitoring for 16 of the 20 subjects took place between June and August 1999; an additional 3 subjects were monitored in October 1999 and 1 subject was monitored in December 1999. Each subject was monitored for one 24-hour period. Continuous PM2.5 monitoring was conducted throughout the 24-hour monitoring period with a TSI Inc. DustTrak (Shoreview, MN), which measures airborne particles using light-scattering technology. It has a range of 0.001–100 mg/m3 and can measure particles from 0.1 to 10 μm in size. The response of the DustTrak is linear over the size range of particles investigated in this study. This is apparently the first health-effects study to make use of the DustTrak. The monitor was placed in a padded pouch and the inlet tubing secured in the participant’s breathing zone. The monitor ran throughout the 24-hour observation period. Subjects were instructed to wear the monitor while they were awake and to place the monitor on a nightstand next to their bed while they were sleeping, keeping the inlet as close as possible to their breathing zone. The monitor was programmed to record measurements every 10 seconds and to report 5-minute averages. Moving averages from 15 minutes to 9 hours were generated using these 5-minute averages.
Continuous Holter Monitoring and Tape Processing
We performed continuous heart rate monitoring using a five-lead Holter monitor, Model Dynacord 3 Channel Model 423 (Raytel Cardiac Services, Windsor, CT). Each participant’s skin was shaved if necessary, cleansed, and slightly abraded to ensure proper lead contact. Electrodes were placed in a modified V1 and V5 position. Each subject was given a daily diary to record any symptoms such as chest pain and shortness of breath, as well as activities such as eating and sleeping.
Each 24-hour tape was sent to Raytel Cardiac Services (Windsor, CT) and analyzed using a Delmar Avionics (Irvine, CA) Model Strata Scan 563. Only beats with a rate ratio (ie, normal-to-normal) interval of 0.6–1.5 seconds and a rate ratio of 0.8–1.2 were included in the analysis. Trained personnel performed all analyses, and all normal and abnormal findings could either be accepted or rejected on the basis of standard criteria to ensure quality control. Tapes were analyzed in the time domain, and reports summarizing heart rhythm, rate analysis, and ST changes were generated. The mean heart rate and the SDNN were calculated in 5-minute segments for the entire recording.
We used mixed-effects regression models (SAS Version 8 19) to investigate the effects of PM2.5 on heart rate variability parameters. Previous studies have reported that heart rate variability is associated with PM2.5 exposure in the preceding few hours. 5 Limited reports in elderly subjects found that a 4-hour average has the strongest correlation, although a recent study has reported even shorter averaging times (2 hours). 20 To address this issue, we examined averaging times ranging from 15 minutes to 9 hours, although the primary hypothesis of the present study was focused on averaging times of 2–4 hours. The log10-transformed response variables, 5-minute heart rate, and 5-minute SDNN were regressed on these moving averages. The response variables were log10 transformed to improve normality and stabilize the variance. A random effect for each study subject, as well as fixed covariates such as time of day, smoking status, age, and heart rate, were included in the models. The modifying effects of individual characteristics such as hypertension were also investigated. Analyses examining the effects of outliers on regression results were also performed.
The study population consisted of 20 white men, 9 of whom were current smokers (Table 1). Their average age was 43 years (SD = 12.7 years), and they had spent an average of 19 years (SD = 14) as boilermakers. Their mean heart rate was 80 beats per minute (SD = 9 beats per minute).
The arithmetic mean PM2.5 concentration over the 24-hour period was 150 μg/m3 (SD = 292 μg/m3) (Table 2). As expected, smokers had a higher mean PM2.5 exposure than nonsmokers, with an arithmetic mean of 216 μg/m3 (SD = 404 μg/m3) compared with 96 μg/m3 (SD = 158 μg/m3) in the nonsmokers. All nonsmoking subjects lived with other nonsmokers except for one nonsmoker who lived with four current smokers. After excluding this particular subject, the mean PM2.5 exposure measured in nonsmokers was reduced to 49 μg/m3 (SD = 35 μg/m3).
The effects of the various moving averages of exposure on heart rate variability (presented as a percentage decrease or increase in the 5-minute SDNN) are plotted in Figure 1. There are small associations between the 5-minute SDNN and the 15-minute averaging interval. A maximum decrease in the 5-minute SDNN is observed using the 3-hour averaging interval. The association weakens using averaging intervals from 4 to 7 hours. Using 8-hour and 9-hour moving average intervals, the observed association becomes positive, indicating increases in heart rate variability with increases in PM2.5 exposure.
The 3-hour PM2.5 moving average was chosen as the main exposure metric, as its effects were the largest in magnitude and most comparable with recently published investigations. 5,21 On the basis of the mixed-effects regression model, an average 1.4% decrease (95% confidence limits [CL] = −2.1, −0.6) in heart rate variability for every 100-μg/m3 increase in the 3-hour PM2.5 moving average was estimated (Table 3). The decreases are similar in direction and magnitude when stratified by current smoking status. However, after excluding the one nonsmoker living with smokers, the mean decrease is only 0.7% (95% CL = −3.1, 1.8).
The association between the 3-hour PM2.5 moving average and the 5-minute heart rate is also summarized in Table 3. Small effects were observed. For the 20 subjects as a whole, there was no mean increase (0.0%; 95% CL = −0.2, 0.2) in heart rate associated with an increase of 100 μg/m3 in the 3-hour PM2.5 moving average after adjusting for time of day, age, and current smoking status as a dichotomous variable. The nonsmoking subgroup (excluding the one individual who lived with other smokers) experienced an average 1.9% increase (95% CL = 1.2, 2.7) in heart rate for every 100-μg/m3 increase in the 3-hour moving average after adjustment for confounding variables.
The same models investigating the effects of the 3-hour moving PM2.5 average on the log10 SDNN and the log10 heart rate were rerun with the top 5% and bottom 5% of the 3-hour moving PM2.5 averages removed to examine the effects of outlying exposures (Table 4). The association became stronger for all subjects combined, with a mean 2.3% decrease (95% CL = −3.7, −0.9) in the heart rate variability for a 100-μg/m3 increase in the 3-hour PM2.5 moving average. The association in nonsmokers increases, and the greatest association is now observed in the nonsmokers’ subgroup (excluding the one nonsmoker living with smokers). This subgroup showed a mean 9.5% decrease (95% CL = −14.5, −4.5) in the 5-min SDNN with every 100 μg/m3 increase in the 3-hour PM2.5 moving average.
When the outlying exposures were removed, the association between the PM2.5 moving average and heart rate was also stronger. For all subjects there was a 1.2% increase (95% CL = 0.7, 1.6) associated with a 100-μg/m3 increase in the 3-hour PM2.5 moving average. The largest increase in heart rate, 4.7% (95% CL = 3.1, 6.2) for every 100 μg/m3 increase in the 3-hour moving average, was observed among the nonsmoking subgroup (excluding the one nonsmoker living with smokers).
The modifying effect of hypertension on the observed association between PM2.5 and both heart rate and SDNN was investigated. Five of the 20 men reported having been diagnosed by a physician with hypertension, although none was currently taking medication for this condition. There was little difference between the normotensive smokers and nonsmokers, with both groups experiencing decreases in the 5-minute SDNN similar to those reported in Table 3. The three hypertensive nonsmokers had a median SDNN of 38 ms, whereas the two hypertensive smokers had a median SDNN of 33 ms. For comparison, the normotensive nonsmokers had a median SDNN of 57 ms whereas the normotensive smokers had a median SDNN of 54 ms. There was little difference in median heart rate between hypertensive smokers and nonsmokers and between normotensive smokers and nonsmokers.
This study used continuous personal monitors to study the effects of ambient PM2.5 on heart rate variability among a relatively young cohort. Previous studies have used broad-area PM2.5 monitoring and have focused specifically on elderly, compromised populations. 5,6
Personal PM2.5 measurements observed in this study are higher than ambient-area measurements in the suburbs of Boston, but are similar to personal PM2.5 measurements reported by Lebret et al.22 and Spengler et al. 23 These researchers reported average PM2.5 levels ranging from 20 to 30 μg/m3 in nonsmoking homes and levels ranging from 30 to 60 μg/m3 in smoking homes. In our study, smoking subjects had the highest geometric mean PM2.5 averaged over the 24-hour monitoring period at 61 μg/m3, with environmental tobacco smoke presumably being the main indoor source. The nonsmoking subjects (excluding the one nonsmoker living with smokers) had a geometric mean PM2.5 averaged over the 24-hour monitoring period of 37 μg/m3, approximately two to three times the daily ambient average in Boston. Indoor sources of PM2.5 other than environmental tobacco smoke, such as cooking and cleaning, may have also contributed to these levels. 24,25
The changes in HRV associated with the 3-hour moving average in this group of 20 individuals are greater than those observed in our previous analysis of the association between occupational and environmental exposure to particulates and HRV. 21 The 20 subjects in the present paper are a self-selected subset of an occupational cohort of 40 subjects. These 20 subjects do not differ from the larger occupational cohort in mean age, mean number of years worked as a boilermaker, or smoking history. The figure illustrating the effects of the various-length moving averages on the 5-minute SDNN is similar to that observed in the analysis of the entire cohort, up to 4 hours. The effects of PM2.5 on HRV using longer moving averages (more than 4 hours) are smaller in this study, whereas the association continued to increase using longer averaging intervals in the larger cohort. Our previous study, 21 which examined the entire cohort, considering both non-work-place exposures and comparatively large work-place exposures (welding fume and residual oil fuel ash) to PM2.5, revealed a mean decrease of 0.5% (95% CL = −0.7, −0.2) in heart rate variability for every 100 μg/m3 increase in the 3-hour moving PM2.5 average. When the exposures are limited to ambient and indoor sources in this study, the observed mean decrease in heart rate variability is 1.4% (95% CL = −2.1, −0.6). This is nearly a three-fold increase in the magnitude of the association when the environmental exposures are considered alone in this group of men. The differences between the two analyses may reflect the overall differences in the time course of action, the mechanism of action, and the overall toxicity of particles derived from the workplace compared with those found in ambient and indoor air pollution.
These results can be compared with those obtained by Gold et al.5 in their study of the effects of ambient PM2.5 on heart rate variability in 21 active Boston residents age 53–87 years. They reported an overall 0.26-ms decrease in the 6-minute SDNN for a 14.4-μg/m3 increase in the 4-hour average PM2.5 after controlling for heart rate. This is a 1.8-ms decrease for every-100 μg/m3 increase in the 4-hour PM2.5 average. To facilitate comparison with our study, we reanalyzed our data using the linear 5-minute SDNN as the outcome measure and regressed it on both the 3-hour and 4-hour moving PM2.5 averages. Using the linear 5-minute SDNN, our data revealed a 1.9-ms decrease for every 100-μg/m3 increase in the 4-hour PM2.5 moving average and a 2.6-ms decrease for every 100-μg/m3 increase in the 3-hour PM2.5 moving average. The decreases associated with the 4-hour moving average in both studies are remarkably similar considering the overall profile of the two age groups studied.
Recognizing that the strata are small, these results suggest little modification by smoking status until outlying exposures have been removed. The magnitude of most effects on heart rate and rate variability increase with exclusion of these outlying exposures, especially among the nonsmoking subgroup (excluding one nonsmoker living with smokers). These results could imply that nonsmokers who are not consistently exposed to environmental tobacco smoke are particularly susceptible to low levels of environmental PM2.5. Furthermore, these results could also suggest that smokers are, in some way, less susceptible to environmental PM2.5 than nonsmokers. These results need to be confirmed in larger studies.
Again, although the strata are small, there is a suggestion that the associations outlined above are also modified by hypertension. Susceptible populations, including those with preexisting cardiopulmonary disease, have been identified in both animal and human populations, 6,9,12 and the effects of hypertension should be given further consideration in future studies.
The observed response to non-work-place environmental particle exposures are different from that reported in our earlier study. 21 This finding may be due to several factors that were not quantified in this study. The particle composition between the two environments is not identical. Work-place particles were dominated by residual fuel oil ash and metal fume derived from welding, whereas the exposures away from work were derived from multiple sources and are likely to represent a more varied composition. Although metals and residual oil fly ash are constituents of ambient PM2.5, they do not constitute its entirety. 26
In addition, personal measurements of other copollutants were not gathered in this study. Confounding by other environmental copollutants such as carbon monoxide, sulfur dioxide, ozone, and nitrogen dioxide is a concern. This concern is moderated by the report of Sarnat et al.27 regarding their assessment of the association between personal particulate and gaseous exposures. They report that personal PM2.5 exposures are not well correlated with personal exposures to ozone, nitrogen dioxide, and sulfur dioxide in a group of nonsmokers. With little association between the personal PM2.5 measurements and personal gaseous copollutants, the potential for confounding is minimized. They did not measure carbon monoxide, however, and whether a similar copollutant pattern for carbon monoxide exists in a mixed group of smokers and nonsmokers is unclear.
Also, the observed PM2.5 exposure patterns in our first study, 21 including both work and nonwork exposures, and this study, which only includes exposures away from work, are somewhat different. The pattern of particle exposure at work included large spikes in PM2.5 due to welding and other work activities, whereas the exposures away from work contained few spikes mainly due to smoking. An individual’s physiologic response to these varied profiles may account for some of the difference shown between our two studies.
This study reveals alterations in cardiac autonomic control associated with environmental PM2.5 exposure. Theses changes in young, otherwise healthy adult males have not been fully explored, although evidence from general population cohorts does indicate an increased risk of cardiac events and all-cause mortality associated with these changes.
We thank Jee Young Kim, Ema Rodrigues, and Jeanne Jackson for their assistance in data collection and laboratory analyses; Salvatore Mucci for his assistance in data entry; and David Miller and Lucille Pothier for their assistance with database management and statistical programming.
1. Samet JM, Dominici F, Curriero FC, et al
. Fine particulate air pollution and mortality in 20 U.S. cities, 1987–1994. N Engl J Med 2000; 343: 1742–1749.
2. Dockery DW, Pope CA 3rd, Xu X, et al
. An association between air pollution and mortality in six U.S. cities. N Engl J Med 1993; 329: 1753–1759.
3. Peters A, Liu E, Verrier RL, et al
. Air pollution and incidence of cardiac arrhythmia. Epidemiology 2000; 11: 11–17.
4. Schwartz J. Air pollution and hospital admissions for cardiovascular disease in Tucson. Epidemiology 1997; 8: 371–377.
5. Gold DR, Litonjua A, Schwartz J, et al
. Ambient pollution and heart rate variability. Circulation 2000; 101: 1267–1273.
6. Liao D, Creason J, Shy C, et al
. Daily variation of particulate air pollution and poor cardiac autonomic control in the elderly. Environ Health Perspect 1999; 107: 521–525.
7. Pope CA 3rd, Verrier RL, Lovett EG, et al
. Heart rate variability associated with particulate air pollution. Am Heart J 1999; 138 (5 pt 1):890–899.
8. Pope CA 3rd. Epidemiology of fine particulate air pollution and human health: biologic mechanisms and who’s at risk? Environ Health Perspect 2000; 108 (suppl 4): 713–723.
9. Godleski JJ, Verrier RL, Koutrakis P, et al
. Mechanisms of morbidity and mortality from exposure to ambient air particles. Res Rep Health Eff Inst 2000; 91: 5–103.
10. Monn C, Becker S. Cytotoxicity and induction of proinflammatory cytokines from human monocytes exposed to fine (PM2.5
) and coarse particles (PM10–2.5
) in outdoor and indoor air. Toxicol Appl Pharmacol 1999; 155: 245–252.
11. Quay JL, Reed W, Samet J, et al
. Air pollution particles induce IL-6 gene expression in human airway epithelial cells via NF-kappaB activation. Am J Respir Cell Mol Biol 1998; 19: 98–106.
12. Watkinson WP, Campen MJ, Costa DL. Cardiac arrhythmia induction after exposure to residual oil fly ash particles in a rodent model of pulmonary hypertension. Toxicol Sci 1998; 41: 209–216.
13. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 1996; 93: 1043–1065.
14. La Rovere MT, Bigger JT Jr, Marcus FI, et al
. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet 1998; 351: 478–484.
15. Bigger JT Jr, Fleiss JL, Rolnitzky LM, et al
. Frequency domain measures of heart period variability to assess risk late after myocardial infarction. J Am Coll Cardiol 1993; 21: 729–736.
16. Dekker JM, Schouten EG, Klootwijk P, et al
. Heart rate variability from short electrocardiographic recordings predicts mortality from all causes in middle-aged and elderly men. The Zutphen Study. Am J Epidemiol 1997; 145: 899–908.
17. Tsuji H, Larson MG, Venditti FJ Jr, et al
. Impact of reduced heart rate variability on risk for cardiac events. The Framingham Heart Study. Circulation 1996; 94: 2850–2855.
18. de Bruyne MC, Kors JA, Hoes AW, et al
. Both decreased and increased heart rate variability on the standard 10-second electrocardiogram predict cardiac mortality in the elderly: the Rotterdam Study. Am J Epidemiol 1999; 150: 1282–1288.
19. SAS/STAT8 Software. 8th ed. Cary, NC: SAS Institute Inc, 1997.
20. Schwartz J, Suh H, Verrier M, et al
. Fine combustion particles and heart rate variability in an elderly panel. Presented at: the 13th Conference of the International Society for Environmental Epidemiology; September 2–5, 2001; Garmisch-Partenkirchen, Germany.
21. Magari SR, Hauser R, Schwartz J, et al
. The association of heart rate variability with occupational and environmental exposure to particulate air pollution. Circulation 2001; 104: 986–991.
22. Lebret E, McCarthy JF, Spengler J, et al
. Elemental composition of indoor fine particles. Indoor Air 1987; 1: 569–573.
23. Spengler J, Ware JH, Speizer FE, et al
. Harvard’s indoor air quality respiratory health study. Indoor Air 1987; 2: 218–223.
24. Abt E, Suh HH, Allen G, et al
. Characterization of indoor particle sources: a study conducted in the metropolitan Boston area. Environ Health Perspect 2000; 108: 35–44.
25. Long CM, Suh HH, Koutrakis P. Characterization of indoor particle sources using continuous mass and size monitors. J Air Waste Manag Assoc 2000; 50: 1236–1250.
26. Wilson WE, Suh HH. Fine particles and coarse particles: concentration relationships relevant to epidemiologic studies. J Air Waste Manag Assoc 1997; 47: 1238–1249.
27. Sarnat JA, Koutrakis P, Suh HH. Assessing the relationship between personal particulate and gaseous exposures of senior citizens living in Baltimore, MD. J Air Waste Manag Assoc 2000; 50: 1184–1198.
© 2002 Lippincott Williams & Wilkins, Inc.