Secondary Logo

Journal Logo

Original Article

Air Pollution and Heart Rate Variability Among the Elderly in Mexico City

Holguín, Fernando*†; Téllez-Rojo, Marta M.; Hernández, Mauricio; Cortez, Marlene†‡; Chow, Judith C.§; Watson, John G.§; Mannino, David; Romieu, Isabelle

Author Information
doi: 10.1097/
  • Free


An association of increased cardiopulmonary morbidity and mortality with particulate matter (PM) pollution has been shown in epidemiologic studies from various countries.1 Suspended particle exposure is also associated with increased rates of hospital admissions and emergency room visits for respiratory and cardiovascular causes.2-4 However, the mechanisms responsible for the increased cardiovascular morbidity and mortality are poorly understood. Environmental exposure to suspended particles has been associated with an increase in mean heart rate and cardiac rhythm abnormalities independent of oxygenation levels in both animals and humans.5,6 The link between particulate matter pollution and cardiac conduction abnormalities has led to research focusing on changes in autonomic cardiac regulation assessed by heart rate variability.7-9,13 Recent studies suggest that PM levels are associated with a reduced heart rate variability, which in turn is known as an independent risk factor for cardiovascular mortality.10-12 However, these studies were conducted with a small number of participants over a limited period of time, and the effects of PM2.5 (particulate matter less than 2.5 μm in diameter) and other specific pollutants on heart rate variability was either not reported7,9,14 or reported only partially.8 In the Mexico City metropolitan area air pollutant levels (in particular, fine particulates and ozone) frequently exceed the Environmental Protection Agency standards.15 Therefore, to explore the autonomic cardiac response in relation to air pollutant exposure, we conducted a longitudinal study among elderly residents of a nursing home located in the northeast of the metropolitan area, an area frequently exposed to high levels of suspended particles and ozone.16


Study Population

The population under study consisted of volunteers who were the permanent residents of a nursing home in the northeast area of the metropolitan area. We followed participants over a period of 3 months, from February 8 to April 30, 2000, obtaining daily local measures of fine particulate matter (PM2.5) and studies every alternate day of heart rate variability. We excluded current smokers, participants with pacemakers or underlying cardiac arrhythmias, and persons who were unable to sign a consent form or attend the study site located within the nursing home. The ethics committee of the Mexican National Institute of Public Health approved this protocol.

Exposure Assessment

Daily 24-hour measurements of PM2.5 were determined by gravimetric analysis using Mini-Vol portable air samplers (version 4.2, Eugene, OR) with 47-mm Teflon filters (Pall Gelman, Vancouver, WA) and flows set at 4 L/min. Outdoor monitors were located on the roof of the nursing home, approximately 150 feet from the entrance and the closest road. Indoor monitors were placed in common living areas. We performed gravimetric analysis of Teflon filters at the air laboratory of the National Center for Environmental Research and Training (CENICA) in Mexico City.

Ambient levels of O3 (ozone), nitrogen dioxide (NO2), sulfur dioxide (SO2), and carbon monoxide (CO), as well as climatic variables, were obtained from an automated monitoring station (Tacuba) located 3 km (NE) upwind from the study site. All pollutant levels were measured by Thermoelectron devices (Franklin, MA) We determined NO2 by chemiluminescence (Thermoelectron 42 MCA), SO2 by an ultraviolet fluorescent analyzer (Thermoelectron 42 MCA), ozone by ultraviolet photometry (Thermoelectron 49 MCA), and ambient CO levels by infrared light absorption (Thermoelectron 48 MCA). Filter gravimetric analysis was performed under controlled climatic conditions at a room temperature of 22°C (±°) with a relative humidity of 40% (±5%); each filter was exposed for 24 hours before gravimetric analysis. Filter weights were obtained by a micrometric scale (Cahn C; Thermo Cahn, Madison, WI) under laminal flow.

Heart Rate Variability Analysis

Participants were scheduled to have a heart rate variability analysis every other day between 8:00 AM and 1:00 PM using an LPPac Q, Predictor 3.0 (Arrhythmia Research Technology, Houston, TX), which meets standards of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology.17,18 Before measuring heart rate variability, we had participants rest in a supine position for 5 minutes. Each recording lasted 5 minutes while the participants continued in the supine position. We identified a QRS template by an automated algorithm that identifies the waveform most representative of the patient’s dominant rhythm. This template was compared with subsequent R waves using a correlation coefficient of 0.75 as previously described in the study by Liao et al.7 We discarded any test having more than 15% of abnormal QRS processes. A total of 132 heart rate variability studies (12.8%) were excluded through the editing process. All patients had at least 1 study excluded with an average of 3.81-15 excluded per participant. We estimated the power spectral density using a Fast Fourier Transformation and a Hanning window with a smoothing weight of 10. Three frequencies were obtained from the power spectral density: a high-frequency component (HRV–HF; 0.15-0.40 Hz), a low-frequency component (HRV–LF 0.04-0.15 Hz), and a very-low-frequency component (HRV–VLF 0.033-0.04 Hz)19

Other Variables

All participants answered a questionnaire, which included demographic information, smoking habits, and a daily diary to record their hourly activities. We extracted the medical history, diagnosis of hypertension, and medication use from the participants’ medical files.

Statistical Analysis

To analyze the effects of PM2.5 and other pollutants on the population average HRV, we transformed the response variables HRV–HF and HRV–LF (msec2) using a log10. We calculated estimates of total PM2.5 exposure based on the time activity patterns and PM2.5 microenvironmental concentrations (indoors and outdoors). Our index of personal exposure to PM2.5 was calculated using the following formula: (PM2.5 indoors × hours spent indoors/24) + (PM2.5 outdoors × hours spent outdoors/24).

We initially studied the association of PM2.5 and ozone with HRV–HF, HRV—LF, and HRV–LF/HF ratio using generalized estimating equations. These models are specially developed to account for autocorrelation and allow for the use of both fixed and time-dependent covariates.20 We adjusted models for potential confounding factors, including participants’ age and the average heart rate during each HRV measurement. Hypertension, sex, minimum temperature, and relative humidity did not confound the association between HRV and PM2.5 and were not included in the regression models. Given the fact that patients with hypertension are more likely to have deregulation of the autonomic nervous system and reduced heart rate variability,21 we stratified our analyses separately for subjects with and without hypertension. All analyses were conducted using air pollutant concentrations on the same day and with a 1-day lag for PM2.5, ozone, as well as SO2, NO2, and CO ambient concentrations; multipollutant models, including PM2.5 and ozone, were also examined. All statistical analyses were conducted using Stata 6.0 (College Station, TX).


Study Population

Of the initial 42 screened participants, 34 (81%) were followed up during the study period. The mean age was 79 years, and 44% were men. Thirteen participants had a diagnosis of hypertension (38%) and 6 had diabetes mellitus, other diagnoses included 4 patients with chronic bronchitis, 4 patients with Parkinson’s disease and 1 each with transient ischemic attacks, hypertension and ischemic cardiomyopathy. Participants spent 86% of the time indoors during the study period as determined by the results of an activity questionnaire completed daily. The mean number of hours spent indoors for the whole group was 21 hours (range: 12–24) and hours spent outdoors were 0.6 hours (0-6) while at the nursing home. The rate of reexamination was high, with a mean of 18 studies per participant and a total of 595 monitoring sessions. Subjects with a medical diagnosis of hypertension were more likely to be women and to spend more time indoors. The mean heart rate variability indices, averaged over the 3-month follow up, were higher for subjects without hypertension than for those with hypertension (Table 1).

Characteristics and Heart Rate Variability Measurements of the Study Participants, Mexico City, 2000

Exposure Data

Results from the air pollutants measured during the study are shown in Table 2. The 24-hour NAAQS (National Ambient Air Quality Standards) for PM2.5 were exceeded on 6 days, and the average indoor and outdoor levels where higher than previously reported in similar heart rate variability studies.7-9 Levels of ozone were also high during the study period, with more than 75% of days being above the NAAQS for the ozone 1-hour maximum. There was a moderate correlation (r = 0.52) between PM2.5 indoors and PM2.5 outdoors, and the correlations of PM2.5 estimates of total exposure with PM2.5 indoors and PM2.5 outdoors were r = 0.92 and r = 0.55, respectively. The correlation of the ozone one-hour maximum with PM2.5 indoors, PM2.5 outdoors, and PM2.5 total exposure were r = 0.27, r = 0.47 and r = 0.18, respectively.

Ambient Pollution and Temperature Levels Measured at the Automated Network of Environmental Monitoring in Tacuba (North Mexico City During HRV Monitoring

Measurements of Heart Rate Variability

The observed associations between the daily variation of PM2.5 total exposure and the HRV indices are presented in Table 3. After adjusting for age, heart rate, and hypertension, we observed an inverse relation between the high-frequency component of heart rate variability and total PM2.5. A 10-μg/m3 increase in PM2.5 was associated with a 5.0% change in HRV–HF. The exposure to indoor levels of PM2.5 was associated with a substantial reduction in the HRV–HF component (β = −0.049; confidence interval = −0.090–0.007 for a 10-μg/m3 increase in PM2.5) and exposure to outdoor PM2.5 was associated with a more modest reduction in high frequency (β = −0.023; −0.058-0.010 for a 10-μg/m3 increase in PM2.5). When restricting the analysis to participants who had a diagnosis of hypertension, a larger reduction in the HRV–HF component was observed, corresponding to a 7.1% change in HRV–HF for a 10-μg/m3 increase in PM2.5. We used associations with other metric parameters of heart rate variability to assess the impact of air pollutants on cardiac autonomic dysfunction; low frequency (HRV–LF) showed a moderate inverse association with PM2.5, whereas the LH/FH ratio showed a modest positive association with levels of PM2.5. In addition, changes in heart rate variability were strongly related to ambient ozone concentrations among subjects with hypertension for HRV–HF and HRV–LF (Table 4), corresponding to a 2.0% change in heart rate variability per 10 ppb increase in ozone 1-hour maximum. Control for diabetes did not change the observed associations. Of our patients with hypertension, 8 were taking angiotensin-converting enzyme inhibitors, 2 were taking beta-blockers, 4 were taking a calcium channel blocker, and 2 were taking a thiazide diuretic. Because of the small number of participants, we were unable to stratify the effect of PM2.5 on HRV by each group of antihypertensive medications.

Adjusted β-coefficients and 95% Confidence Intervals of the Change in Heart Rate Variability (HRV) for a 10-μg/m3 Increase in Same-day PM2.5
Adjusted β-coefficients and 95% Confidence Intervals of the Change in Heart Rate Variability for a 10-ppb Increase in Same-day Ozone 1-hour Maximum

Other pollutants (NO2, CO, and SO2) were not related to heart rate variability (Table 5). All of these results are for air pollution parameter on the same day as the heart rate variability assessments; results were similar when we used 1-day time-lag exposure for the pollutants.

Adjusted β-coefficients and 95% Confidence Intervals of the Association Between Heart Rate Variability and Gaseous Pollutants

In multipollutant models, including both PM2.5 and ozone, the magnitude of the relation of PM2.5 with the HRV–HF component decreased slightly (Table 6), corresponding to the following percent changes: 5.0% in our total sample, 4.0% among participants without hypertension, and 6.4% among participants with hypertension for a 10-μg/m3 increase in PM2.5. Ozone was no longer associated with HRV–HF (Table 6).

Adjusted β-coefficients and 95% Confidence Intervals of the Association of HRV with PM2.5 and Ozone 1-hour Maximum


This is one of the largest studies to date on the effects of PM2.5 and ozone on heart rate variability, and it is the first of its kind in Mexico. Our results suggest that exposure to ambient air concentrations of PM2.5 and ozone is associated with autonomic cardiac dysfunction in the elderly, characterized by a reduction in the high frequency or vagal component of heart rate variability. Furthermore, in our study, participants with a diagnosis of hypertension appeared to be more susceptible with larger reductions in the HRV–HF associated with PM2.5 exposure. Although individuals diagnosed with hypertension are known to have lower heart rate variability,21 an increased susceptibility to particulate air pollution in this group of patients has not been previously described. In our study, exposure to ambient levels of ozone was also associated with reductions in the HRV–HF component, although this effect was not as strong in the multipollutant models that adjusted for PM2.5.

Numerous mechanistic hypotheses have been proposed to explain how exposure to ambient levels of PM2.5 might produce changes in the autonomic cardiac regulation. Translocation of inhaled ultrafine particles into the systemic circulation,22 the activation of pulmonary irritant receptors that mediate the stimulation of parasympathetic pathways,23 an increase in plasma viscosity,24 and the elevation of inflammatory mediators25 are some of the potential biologic explanations. It is possible that the stimulation of irritant receptors from ozone-induced airway inflammation and oxidation23 can either independently lead to changes in heart rate variability or potentiate the effect of inhaled particulates. In general, it is conjectured that interactions among inflammation, abnormal hemostatic function, and altered cardiac rhythm might play an important role in the pathogenesis of cardiopulmonary diseases related to air pollution.26

Several limitations have to be considered in interpreting our results. We used only short-term heart rate variability recordings (5 min), which could have hindered the possibility of performing an adequate assessment of lower frequencies. However, there is a high correlation between short recordings of HRV–HF using spectral analysis and using longer time-domain analysis.27 Because we obtained short-term study results that were performed during the morning, we were unable to account for any changes related to circadian variation in heart rate variability parameters. Maneuvers to elicit vagal responses, such as metronomic breathing, were not performed, but controlling respiratory rate does not seem to confound the association between PM2.5 and changes in heart rate variability.8 The same trained technician performed the entire heart rate variability recording, thus reducing intertest variability. PM2.5 exposure was limited to 384 observations from 34 subjects who had complete time activity data; these coefficients were not substantially different from coefficients calculated from the 595 heart rate variability monitoring sessions using PM2.5 indoor exposure.

Restricting analysis to outdoor PM2.5 was associated with a modest reduction in the HRV–HF component, suggesting that some indoor sources might play a role. We were unable to identify significant indoor exposure sources and the nursing home was a smoke-free environment. However, idling diesel buses were parked for a few hours close to the nursing home indoor living areas at least 3 times a week. We speculate that such emissions could have contributed to the larger effects seen in association with indoor PM2.5.

Information about ozone concentration and other pollutants was available through a nearby automated monitoring station, which was situated 3 km upwind from the study site, allowing a reasonable estimation of the nursing home ozone air pollution. Measurements for NO2 and CO are more susceptible to local sources of emission and traffic flow, and therefore might not be representative of the nursing home environment. Our results are from a single nursing home located in the northeast metropolitan area and might not be applicable to a more general population; still our findings are consistent with a similar study of elderly nursing home residents.7 A reduction in heart rate variability has been positively correlated with increased rates of cardiovascular morbidity and mortality in both the high-risk and the general population.11,12 However, further studies are needed to determine whether the daily fluctuations in autonomic cardiac balance associated with exposure to ambient particulate matter and ozone will provide meaningful clinical and epidemiologic data. Results from this study complement previous reports by identifying subjects with hypertension as a population potentially at risk of experiencing the adverse cardiovascular health effects associated with ambient ozone and particulate air pollution.


We thank the CENICA (Centro Nacional de Investigación y Capacitación Ambientales in Mexico City) group and Victor Gutierrez for their contribution in performing the gravimetric analyses in our study; the DIF (Departamento de Integración Familiar, México) for allowing them to perform the study in their nursing home, and Isabel García and Rafael Santibañnez for their technical support.


1.Schwartz J. Air pollution and daily mortality: a review and meta analysis. Environ Res. 1994;64:36–52.
2.Burnett RT, Smith-Doiron M, Stieb D, et al. Effects of particulate matter and gaseous pollution on cardiorespiratory hospitalizations. Arch Environ Health. 1999;54:130–139.
3.Wordley J, Walters S, Ayres JG. Short term variations in hospital admissions and mortality and particulate pollution. Occup Environ Med. 1997;54:108–116.
4.Tolbert P, Klein M, Metzger KB, et al. Interim results of the study of particulates and health in Atlanta (SOPHIA). Journal of Exposure and Environmental Epidemiology. 2000;10:446–460.
5.Pope CAI, Dockery DW, Kanner RE, et al. Oxygen saturation, pulse rate, and particulate air pollution: a daily time-series panel study. Am J Respir Crit Care Med. 1999;159:365–372.
6.Godleski JJ, Verrier RL, Koutrakis P, et al. Mechanisms of morbidity and mortality from exposure to ambient air particles. Health Effects Institute Research Reports, 2000. Available at:
7.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.
8.Gold DR, Litonjua A, Schwartz J, et al. Ambient pollution and heart rate variability. Circulation. 2000;101:1267–1273.
9.Pope CAI, Verrier RL, Lovett EG, et al. Heart rate variability associated with particulate air pollution. Am Heart J. 1999;138:890–899.
10.Tsuji H, Venditti FJ, Manders ES, et al. Reduced heart rate variability and mortality risk in an elderly cohort. Circulation. 1994;90:878–883.
11.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. Am J Epidemiol. 1997;145:899–908.
12.Tsuji H, Larson MG, Venditti FJ, et al. Impact of reduced heart rate variability on risk for cardiac events. Circulation. 1996;94:2850–2855.
13.Magari SR, Schwartz J, Paige L, et al. The associations between personal measurements of environmental exposure to particulates and heart rate variability. Epidemiology. 2002;13:305–310.
14.Magari RS, Hauser R, Schwartz J, et al. Association of heart rate variability with occupational and environmental exposure to particulate air pollution. Circulation. 2001;104:986–991.
15.Borja-Aburto VH, Castillejos M, Gold DR, et al. Mortality and ambient fine particles in southwest Mexico City. Environ Health Perspect. 1998;106:849–856.
16.Carabias LJ, Provencio ED, Fernández AB, et al. Almanaque de datos y tendencias de la calidad del aire en ciudades Mexicanas. Dirección general de gestión e Información Ambiental del instituto Nacional de Ecología. CENICA; 2000.
17.The Task Force of The European 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.
18.Kennedy HL. Heart rate variability instruments from commercial manufacturers. In: Malik M, Camm AJ, eds. Heart Rate Variability. Armonk, NY: Futura Publishing; 1st ed. 1995:127–132.
19.Stein PK, Bosner MS, Kleiger RE, et al. Heart rate variability: a measure of cardiac autonomic tone. Am Heart J. 2001;127:1376–1381.
20.Diggle PJ, Liang KY, Zeger SL. Analysis of Longitudinal Data. New York: Oxford University Press; 2001.
21.Singh JP, Larson MG, Tsuji H, et al. Reduced heart rate variability and new-onset hypertension: insights in to the pathogenesis of hypertension: the Framingham Heart Study. Hypertension. 1998;32:293–297.
22.Nemmar A, Hoet PHM, Vanquickenborne B, et al. Passage of inhaled particles into the blood circulation in humans. Circulation. 2002;105:411–414.
23.Watkinson WP, Campen CJ, Nolan JP, et al. Cardiovascular and systemic responses to inhaled pollutants in rodents: effects of ozone and particulate matter. Environ Health Perspect. 2001;109(suppl 4):539–546.
24.Koening WSMLH, Döring A, Ernst E. Association between plasma viscosity and all-cause mortality: results from the MONICA-Ausburg cohort study. Br J Haematol. 2000;109:453–458.
25.Schwartz J. Air pollution and blood markers of cardiovascular risk. Environ Health Perspect. 2001;109(suppl 3):405–409.
26.Arden Pope C III. 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.
27.Sinnreich RK, Friedlander Y, Sapoznikov D, et al. Five minute recordings of heart rate variability for population studies: repeatability and age-sex characteristics. Heart. 1998;80:156–162.

A Stein is a Stein is a Stein?

Readers of our July issue may wonder how Epidemiology could have mistaken a physician who left her home country to become a renowned epidemiologist, for a physician who left her home country to become a renowned poet.

We’re wondering too.

The Editors

Gertrude Stein by Pablo PicassoZena Stein by Bachrach

particulate air pollution; ozone; heart rate variability; Mexico; elderly

© 2003 Lippincott Williams & Wilkins, Inc.