Ozone Air Pollution Is Associated With Acute Myocardial Infarction
http://www.100md.com
循环学杂志 2005年第2期
INSERM U558, Faculte de Medecine (J.-B.R., M.C., M.G., J.F.)
Department of Epidemiology, CHU (J.-B.R., J.F.); InVS (S.C.)
ORAMIP (M.M.), Toulouse, France.
Abstract
Background— Despite the diversity of the studied health outcomes, types and levels of pollution, and various environmental settings, there is substantial evidence for a positive link between urban air pollution and cardiovascular diseases. The objective of this study was to test the associations between air pollutants and the occurrence of acute myocardial infarction (AMI).
Methods and Results— Pollutant concentrations (SO2, NO2, and O3) were measured hourly as part of the automated air quality network. Since 1985, an AMI registry (the Toulouse MONICA Project) has been collecting data in the southwest of France. All cases of AMI and sudden and probable cardiac deaths are recorded for subjects 35 to 64 years of age. We studied the short-term exposure effect of pollution on the risk of AMI (from January 1, 1997, to June 30, 1999) using a case-crossover design method. We performed a conditional logistic regression analysis to calculate relative risks (RRs) and their 95% CIs. After adjustment for temperature, relative humidity, and influenza epidemics, the RRs (for an increase of 5 μg/m3 of O3 concentration) for AMI occurrence were significant for the current-day and 1-day-lag measurements (RR, 1.05; 95% CI, 1.01 to 1.08; P=0.009; and RR, 1.05; 95% CI, 1.01 to 1.09; P=0.007, respectively). Subjects 55 to 64 years of age with no personal history of ischemic heart disease were the most susceptible to develop an AMI (RR, 1.14; 95% CI, 1.06 to 1.23). NO2 and SO2 exposures were not significantly associated with the occurrence of AMI.
Conclusions— Observational data confirm that short-term O3 exposure within a period of 1 to 2 days is related to acute coronary events in middle-aged adults without heart disease, whereas NO2 and SO2 are not.
Key Words: morbidity ; myocardial infarction ; air pollution ; registries ; vasoconstriction
Introduction
Despite the diversity of the studied health outcomes, types and levels of pollution, various environmental settings, and different methodological approaches, substantial evidence for a positive link between urban air pollution and cardiovascular diseases exists. Increasing levels of air pollutant concentration were found to be associated with increased cardiovascular or ischemic heart disease (IHD) mortality and a higher rate of hospital admissions for vascular diseases1–13; however, few reports14 have investigated jointly the relationships between ambient air pollution and cardiovascular morbidity and mortality at the population level. Few studies have investigated mortality on the basis of only the underlying causes stated in death certificates or investigated morbidity on the basis of cardiovascular hospital admissions with a potential selection bias with regard to acute myocardial infarction (AMI) cases. A systematic recording of AMI at the population level over a long period enabled us to study the relationships between air pollutants and mortality and morbidity of IHD simultaneously. We could investigate these relationships thanks to the existence of an acute myocardial infarction registry (Monitoring Trends and Determinants in Cardiovascular Disease [MONICA]) and an air quality network (ORAMIP) covering the same population in the Toulouse area. The objective of the present study was to test the associations between various air pollutants and the daily occurrence of AMI.
Methods
Air Pollution and Meteorological Parameter Measurements
Gas pollutant concentrations were measured hourly as part of the automated French air quality network (ORAMIP) in southwestern France. Sulfur dioxide (SO2) concentrations were measured by UV fluorescence; nitrogen dioxide (NO2) concentrations, by chemiluminescence; and ozone (O3) concentrations, by UV photometry. We primarily used French Environment SA (eg, AC31M for NO2 and O341M for O3) or SERES (SF 2000 for SO2) equipment for the ORAMIP measurements.
At the beginning of the study (January 1, 1997), 14 monitoring stations were located mainly in the metropolitan area of Toulouse (southwestern France), including 4 lead measurement stations. Three types of stations were identified according to the European classification15: stations measuring urban background concentrations of pollutants (n=5) to which the population is exposed, including suburban (n=1) and observation (n=1) stations; stations directly exposed to traffic pollution (n=3); or stations exposed to industrial pollution (n=6), including stations close to a chemical plant site (n=2) and those close to a lead factory (n=4). Only data provided by urban monitoring stations were taken into account to test the relationships between air pollutants and the occurrence of IHD because the goal of the study was to estimate the average daily exposure-response relationship, not to assess the health impact of exposure to air pollution from specific sources (eg, road traffic). For each pollutant (NO2 and SO2), the 24-hour mean concentration was calculated from midnight to midnight and took into account the results from all the urban monitoring stations. For O3, the moving average of the 8 consecutive hours with the highest values of the day was taken into account in the statistical analysis.
Mean daily temperature (degrees Celsius) for the minimum and the maximum values and relative humidity (percent) were available from the weather station located in the metropolitan area of Toulouse. Data on influenza epidemics, expressed as the number of occurrences per week, were obtained from the general practitioner sentinel network of the metropolitan area of Toulouse.
A standardized procedure was used to complete days with missing data.16 A missing value () on day i in year k in monitoring station j was replaced by a weighted average value of the measurements from the rest of the monitoring stations; ie, ijk = i.k x (.jk/..k). Globally, the proportion of missing values in relation to the total values of measured pollutants averaged 2%.
Registry Data
Since 1985, an acute myocardial infarction registry (the Toulouse MONICA Project) has been carried out in southwestern France.17–19 All cases of AMI, sudden cardiac deaths, and probable cardiac deaths are recorded for subjects 35 to 64 years of age. The studied population in 1999 (national population census) was 395 744 inhabitants in the whole area covered by the registry for the eligible age range. The population living in the area covered by both the registry and the air quality network made up the study population. The population exposed during the study period (January 1, 1997, to June 30, 1999) was 685 895 inhabitants. All private and public hospitals within the geographic area of the registry were screened by the medical staff for suspected AMI, and death certificates (suspected coronary deaths, sudden deaths, and fatal cases with insufficient data) were analyzed. An event was considered new if its apparent date of onset was >28 days after any previous AMI event. An event was considered fatal if death occurred before midnight on the 27th day after the initial event. Four definitions were used in this study: definition 1, definite AMI plus possible coronary death plus sudden death plus death with insufficient data; definition 2, definite AMI plus possible coronary death plus sudden death; definition 3, definite AMI plus possible coronary death; and definition 4, definite AMI.20,21
Briefly, definite AMI corresponded to subjects for whom clinical, ECG, and enzymatic data were available. All deaths classified as coronary deaths, sudden deaths, or deaths with insufficient data were validated from all available medical and forensic reports and interviews of general practitioners. Possible coronary deaths corresponded to subjects who died rapidly after suffering from chest pain (typical or atypical symptoms) or who had a known history of CHD without clear evidence supporting another cause of death. Sudden deaths corresponded to subjects who died rapidly (<1 hour) without symptoms, who had no personal history of CHD, or who had evidence supporting another cause of death. Deaths with insufficient data corresponded to deaths with no symptoms (or inadequately described), no history of CHD, and no other diagnosis.21
Only the population living in the Toulouse area covered by urban monitoring stations was studied, and only acute episodes occurring during the studied period were included in the statistical analysis. During the study period, 635, 564, 475, and 399 acute episodes met the required criteria for the definitions 1, 2, 3, and 4, respectively. The proportions of the various events for the most complete definition (definition 1, n=635) were 62.8%, 12.0%, 14.0%, and 11.2% for definite AMI, possible coronary death, sudden death, and death with insufficient data, respectively.
Statistical Analysis
To predict the likely occurrence of an acute event, we used fixed-effect methods for nonrepeated events using conditional logistic regression.22 The case-crossover design was first developed by Maclure23 to study the short-term exposure effect on transient risks. In our study, when 1 subject in the registry population experienced at least 1 event during the studied period, the day of occurrence was defined as the "case day." Two methods were used to select the referent period (control period) on the basis of knowledge of the cyclic and time variability in pollution in the area studied for the case-crossover analysis: the ambidirectional sampling 7-day referent lag (control days were the same day of the previous week and the same day of the week after the case day) and the time-stratified (by months) sampling selection.24 In the ambidirectional sampling selection, the 2 control days were the same day of the week as the case day. For example, for an acute event that occurred on Monday, the 2 control days were the Monday of the previous week and the Monday of the next week. When 2 events occurred on the same day, the number of control days was established accordingly. For each event occurring on a specific day, a matched pair was established. In the time-stratified sampling selection, stratification into separate months, with a retrospective fixed-interval referent selection, was done. The control days were selected 7, 14, 21, and 28 days before the case day. The models were systematically adjusted for relative humidity (in percentage), temperature (minimum or maximum values in degrees Celsius), and influenza epidemics (number of incidents per week). All adjusted variables (relative humidity, temperature, and influenza epidemics) were entered into each model as a single quantitative term.
The relationships between exposure and AMI occurrence were tested considering the exposure to pollutants during the day of AMI occurrence (concurrent day, or D0), the day before AMI occurrence (day lag 1, or D-1), 2 days before (day lag 2, or D-2), and 3 days before (day lag 3, or D-3), as well as the average values of pollutant concentration over 3 days (D0 to D-2) and (D-1 to D-3).
A conditional logistic regression analysis (fixed-effect logistic regression) was performed to calculate relative risks (RRs) and their 95% CIs (PROC PHREG procedure with SAS software). First, to ascertain the correct scale in the logit association between exposure and the occurrence of AMI, the estimated coefficients of the corresponding variables were assessed by calculation of adjusted RRs for tertiles of each pollutant, specifically for O3. The linear trend of the logit was tested with the ordinal values of the tertiles in the conditional logistic regression. These relationships were tested for the 4 definitions of AMI given in this study and for every single-day lag or for the average of 3 consecutive days. The results have shown that the shape of the relationship can be modeled as a logit linear relationship. Finally, the linear association between the pollutant and the occurrence of AMI was analyzed, with the pollutant entered as a single quantitative term in each model. For each pollutant, the RRs of AMI occurrence were given for a continuous variation with an increment of 5 μg/m3 of this pollutant.
The analyses, performed in the whole population living in the area covered by the monitoring stations of the ORAMIP network, were carried out by age groups (35 to 54 and 55 to 64 years), sex, personal history of IHD, and survival status. Finally, RRs were estimated in the population with the potentially highest risk level, ie, the oldest subjects (55 to 64 years), subjects with personal history of IHD, and those who died within 28 days after the occurrence of AMI. Because the 2 methods of referent period (time-stratified or ambidirectional) selection gave similar results, only the results obtained with the time-stratified sampling were taken into consideration.
Results
In Table 7, RRs between O3 exposures and occurrence of acute events were estimated in the oldest subjects with no personal history of IHD. For an increase of 5 μg/m3 (5 units) of O3 concentration, the RRs were similar for a given day lag regardless of the definition of AMI. They were significant for the current day for AMI definitions 1 and 2 and were significant for all AMI definitions for day lag 1 and for the cumulative concentration of O3 (from day lag 0 to 2). The highest RRs and the most significant associations were observed for day lag 1. The RRs were 1.11 (95% CI, 1.04 to 1.19), 1.14 (95% CI, 1.06 to 1.23), 1.15 (95% CI, 1.06 to 1.25), and 1.13 (95% CI, 1.04 to 1.23) for definitions 1, 2, 3, and 4, respectively.
Discussion
The main finding of this study is a positive relationship between acute IHD occurrence and O3 exposure. Moreover, analyses performed from the current day of exposure to day lag 3 suggest a short time effect of ozone on acute IHD occurrence during the first 24 hours after exposure. Indeed, the highest RRs were observed on day lag 1 and then decreased until day lag 3. The cumulative effect of O3 exposure over 3 days was not observed in this study; the RRs were lower than or of a similar magnitude as those of day lag 1. Finally, the sensitivity analysis done with different AMI definitions showed a consistent association between ozone air concentration and AMI occurrences. For day lag 1, the attributable risk for the population 35 to 64 years of age ranged from 4% to 5% according to the definition of AMI used. The attributable risk reached 12% when the oldest population (55 to 64 years of age) without history of IHD was considered.
Analysis of the relationships by individual level of age and personal history of IHD revealed that the oldest subjects (in a middle-aged population of 35 to 64 years) without a personal history of IHD were more susceptible to an acute event when ozone concentrations increased. The associations according to survival status were identical, however. In this case, the relationship was of a similar strength among subjects who died after an acute IHD event and those who had a 28-day survival. The relationships were statistically significant in the latter group because of higher statistical power (the number of subjects was double).
The short-term effect of O3 on various health outcomes has been studied, with a focus on mortality and hospital admissions,13,25,26 but few studies have assessed a positive association between O3 exposure and AMI morbidity/mortality. A large multicenter study carried out in the United States, NMMAPS (National Morbidity, Mortality, and Air Pollution Study), reported a significant impact of O3 during the summer and an increased mortality of 0.41% associated with an increase of 20 μg/m3 in daily O3 concentrations at day lag 0.25
In our study, the lack of association between NO2 or SO2 and AMI occurrence could originate from the moderate levels of air concentration pollutants in the Toulouse area, resulting in a limited range variability of concentration values. Generally speaking, comparisons with cities where associations between gaseous pollutants and cardiovascular hospitalization rates were significant showed median values of air pollutant concentrations that were 2- to 3-fold higher than in the Toulouse area (NO2=29.3 μg/m3, SO2=7.4 μg/m3) (Rome: NO2=86.0 μg/m3; Hong Kong: NO2=53.5 μg/m3, SO2=14.5 μg/m3; London: NO2=61.2 μg/m3, SO2=20.6 μg/m3; Denver: SO2=15.2 μg/m3).6–8 On the other hand, O3 concentration was positively associated with the risk of AMI occurrence in Toulouse (median value of O3=74.8 μg/m3). In Denver, which has a similar median level of O3 concentration (50.3 μg/m3), a 25th to 75th percentile change of ozone increased the risk of hospitalization for AMI and coronary atherosclerosis when the current day of O3 exposure and AMI admissions was considered.7 In Barcelona (median value of O3=70.8 μg/m3),2 an increase of 100 μg/m3 of ozone air concentration was significantly associated (RR, 1.09) with increased cardiovascular mortality in the elderly during the summer. In Mexico City (median O3=87.0 μg/m3), the mean concentration of O3 over a 2-day period was associated with a 1.8% increase in cardiovascular mortality.3 In Hong Kong and London,6 with half the O3 concentration recorded in Toulouse (median O3=28.3 and 32.0 μg/m3, respectively), associations between O3 exposure and cardiac admissions were less consistent and remained uncertain. In Montreal,5 with a median concentration of ozone that was one third the value recorded in Toulouse, cardiovascular mortality increased by 2.5% (95% CI, 0.2 to 5.0) with an increase of 21.3 μg/m3 of O3. Globally, the conflicting results could be due to the multiplicity of the studied health outcomes, the various types and levels of pollution, and investigation methodologies.
The age-related increasing risk observed in the present study was an expected result because of a higher prevalence of deleterious pathologies among the elderly (congestive heart failure or arrhythmia) and a growing sensitivity to the effects of ambient air pollutants.27 On the other hand, the higher risk among subjects without a personal history of IHD was unexpected. Some pathophysiological hypotheses can be suggested to explain the association between the short-term effect of ozone and AMI occurrence.
One possibility is that alterations in arterial tone could be a relevant mechanism explaining an increased incidence of acute cardiac events in populations exposed to air pollution. Short-term exposure to a mix of fine particulate air pollution (PM 2.5) and O3 at levels observed in urban environment induces artery vasoconstriction in healthy adults28; however, the effect of O3 alone on endothelial function and arterial tone remains to be investigated. In rabbits,29 exposure to ozone for 4 hours increased baseline values of total vascular resistance in pulmonary vessels that were directly exposed, but systemic vessels were not investigated. Thus, one could hypothesize that O3 may alter systemic vascular tone like it alters pulmonary tone in direct contact, but this hypothesis needs to be investigated. The lack of a significant association between ozone exposure and AMI occurrence in subjects with a history of IHD could be explained by the preventive action of standard treatments against relapses that can be taken by these patients: vasodilator drugs (nitrate), statin with its pleiotropic effects (improvement of endothelial dysfunction, increased nitric oxide bioavailability, antioxidant properties, inhibition of inflammatory responses, and stabilization of atherosclerotic plaques),30 and ACE inhibitors (improvement of endothelial function, contribution to increased NO bioactivity, antiproliferation activity, modulation of sympathetic activity, and attenuation of atherosclerosis).31
This study has some limitations. This relationship was established at the population level, and no specific information system had been developed for this investigation. We used an information system (AMI MONICA registry) existing since 1985 in the Toulouse area. The methodology of the MONICA project has been validated and can be considered a reliable device for the AMI registration.17,20 The recording of AMI at the population level does not completely eliminate all the risks of potential biases but enables the reduction of selection biases of AMI cases.
The age of the population covered by the registry ranged from 35 to 64 years, thus limiting the generalization of the conclusions because subjects >64 years of age are more vulnerable to pollution exposure than younger subjects. Nevertheless, our results underline the deleterious influence of air pollution on the vascular system even at young ages.
Pollution risk factors were measured as ecological variables with measurements provided by the urban monitoring stations, assuming that exposure was homogeneous all over the studied area, thus limiting the strength of the relationship because of probable exposure misclassification for some individuals. Indeed, exposure misclassification would only bias to the null, and the association could be underestimated.32
The case-crossover design is a reliable method to study short-term exposure effects on transient risks. It eliminates the potential biases resulting from seasonality and long-term trends of air pollutants insofar as the referent periods are meticulously selected.24,33 In the present study, the results obtained with the 2 methods of referent-period selections (ambidirectional sampling and time-stratified sampling selection) were similar.
In conclusion, these observational data confirm that short-term O3 exposure within a period of 1 to 2 days is related to acute coronary events in middle-aged adults without heart disease, whereas NO2 and SO2 are not.
Acknowledgments
We thank the Federation Fran;aise de Cardiologie, the Institut National de la Sante et de la Recherche Medicale, and the Institut de Veille Sanitaire for their support that enabled this work. We thank Thibaud Bouillie for preparing the database, all the investigators of the Toulouse MONICA registry for their invaluable contributions to the careful collection and validation of the data, and the physicians and cardiologists of private and public institutions who helped in this process.
References
Schwartz J, Morris R. Air pollution and hospital admissions for cardiovascular disease in Detroit, Michigan. Am J Epidemiol. 1995; 142: 23–35.
Sunyer J, Castellsague J, Saez M, Tobias A, Anto JM. Air pollution and mortality in Barcelona. J Epidemiol Community Health. 1996; 50 (suppl 1): s76–s80.
Borja-Aburto VH, Castillejos M, Gold DR, Bierzwinski S, Loomis D. Mortality and ambient fine particles in southwest Mexico City, 1993–1995. Environ Health Perspect. 1998; 106: 849–855.
Ballester F, Tenias JM, Perez-Hoyos S. Air pollution and emergency hospital admissions for cardiovascular diseases in Valencia, Spain. J Epidemiol Community Health. 2001; 55: 57–65.
Goldberg MS, Burnett RT, Brook J, Bailar JC 3rd, Valois MF, Vincent R. Associations between daily cause-specific mortality and concentrations of ground-level ozone in Montreal, Quebec. Am J Epidemiol. 2001; 154: 817–826.
Wong CM, Atkinson RW, Anderson HR, Hedley AJ, Ma S, Chau PY, Lam TH. A tale of two cities: effects of air pollution on hospital admissions in Hong Kong and London compared. Environ Health Perspect. 2002; 110: 67–77.
Koken PJ, Piver WT, Ye F, Elixhauser A, Olsen LM, Portier CJ. Temperature, air pollution, and hospitalization for cardiovascular diseases among elderly people in Denver. Environ Health Perspect. 2003; 111: 1312–1317.
D’Ippoliti D, Forastiere F, Ancona C, Agabiti N, Fusco D, Michelozzi P, Perucci CA. Air pollution and myocardial infarction in Rome: a case-crossover analysis. Epidemiology. 2003; 14: 528–535.
Zmirou D, Barumandzadeh T, Balducci F, Ritter P, Laham G, Ghilardi JP. Short term effects of air pollution on mortality in the city of Lyon, France, 1985–90. J Epidemiol Community Health. 1996; 50 (suppl 1): S30–S35.
Nafstad P, Haheim LL, Wisloff T, Gram F, Oftedal B, Holme I, Hjermann I, Leren P. Urban air pollution and mortality in a cohort of Norwegian men. Environ Health Perspect. 2004; 112: 610–615.
Peters A, Dockery DW, Muller JE, Mittleman MA. Increased particulate air pollution and the triggering of myocardial infarction. Circulation. 2001; 103: 2810–2815.
Le Tertre A, Medina S, Samoli E, Forsberg B, Michelozzi P, Boumghar A, Vonk JM, Bellini A, Atkinson R, Ayres JG, Sunyer J, Schwartz J, Katsouyanni K. Short-term effects of particulate air pollution on cardiovascular diseases in eight European cities. J Epidemiol Community Health. 2002; 56: 773–779.
Brook RD, Franklin B, Cascio W, Hong Y, Howard G, Lipsett M, Luepker R, Mittleman M, Samet J, Smith SC Jr, Tager I, for the Expert Panel on Population and Prevention Science of the American Heart Association. Air pollution and cardiovascular disease: a statement for healthcare professionals from the Expert Panel on Population and Prevention Science of the American Heart Association. Circulation. 2004; 109: 2655–2671.
Eilstein D, Quenel P, Hedelin G, Kleinpeter J, Arveiler D, Schaffer P. Air pollution and myocardial infarction: Strasbourg France, 1984–89. Rev Epidemiol Sante Publique. 2001; 49: 13–25.
Classification and Criteria for Setting Up Air Quality Monitoring Stations. Paris, France: ANDEME Editions; 2002.
Katsouyanni K, Schwartz J, Spix C, Touloumi G, Zmirou D, Zanobetti A, Wojtyniak B, Vonk JM, Tobias A, Ponka A, Medina S, Bacharova L, Anderson HR. Short term effects of air pollution on health: a European approach using epidemiologic time series data: the APHEA protocol. J Epidemiol Community Health. 1996; 50 (suppl 1): S12–S18.
Marques-Vidal P, Ruidavets JB, Cambou JP, Ferrières J. Incidence, recurrence, and case fatality rates for myocardial infarction in southwestern France, 1985 to 1993. Heart. 2000; 84: 171–175.
Marrugat J, Ferrières J, Masia R, Ruidavets J, Sala J. Differences in use of coronary angiography and outcome of myocardial infarction in Toulouse (France) and Gerona (Spain): the MONICA-Toulouse and REGICOR investigators. Eur Heart J. 2000; 21: 740–746.
Ferrières J, Cambou JP, Ruidavets JB, Pous J. Trends in acute myocardial infarction prognosis and treatment in southwestern France between 1985 and 1990 (the MONICA Project-Toulouse). Am J Cardiol. 1995; 75: 1202–1205.
Tuomilehto J, Kuulasmaa K. WHO MONICA Project: assessing CHD mortality and morbidity. Int J Epidemiol. 1989; 18 (suppl 1): S38–S45.
Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project: registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation. 1994; 90: 583–612.
Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 1989.
Maclure M. The case-crossover design: a method for studying transient effects on the risk of acute events. Am J Epidemiol. 1991; 133: 144–153.
Levy D, Lumley T, Sheppard L, Kaufman J, Checkoway H. Referent selection in case-crossover analyses of acute health effects of air pollution. Epidemiology. 2001; 12: 186–192.
Samet JM, Dominici F, Curriero FC, Coursac I, Zeger SL. Fine particulate air pollution and mortality in 20 U.S. cities, 1987–1994. N Engl J Med. 2000; 343: 1742–1749.
Le Tertre A, Quenel P, Eilstein D, Medina S, Prouvost H, Pascal L, Boumghar A, Saviuc P, Zeghnoun A, Filleul L, Declercq C, Cassadou S, Le Goaster C. Short-term effects of air pollution on mortality in nine French cities: a quantitative summary. Arch Environ Health. 2002; 57: 311–319.
Mann JK, Tager IB, Lurmann F, Segal M, Quesenberry CP Jr, Lugg MM, Shan J, Van Den Eeden SK. Air pollution and hospital admissions for ischemic heart disease in persons with congestive heart failure or arrhythmia. Environ Health Perspect. 2002; 110: 1247–1252.
Brook RD, Brook JR, Urch B, Vincent R, Rajagopalan S, Silverman F. Inhalation of fine particulate air pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation. 2002; 105: 1534–1536.
Delaunois A, Segura P, Montano LM, Vargas MH, Ansay M, Gustin P. Comparison of ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol Appl Pharmacol. 1998; 150: 58–67.
Davignon J. Beneficial cardiovascular pleiotropic effects of statins. Circulation. 2004 15; 109 (suppl III): III-39–III-43.
Ferrari R. Effect of ACE inhibition on myocardial ischaemia. Eur Heart J. 1998; 19 (suppl J): J30–J35.
Armstrong BG. Effect of measurement error on epidemiological studies of environmental and occupational exposures. Occup Environ Med. 1998; 55: 651–656.
Robins J, Pike M. The validity of case-control studies with non random selection of controls. Epidemiology. 1990; 1: 273–284.(Jean-Bernard Ruidavets, M)
Department of Epidemiology, CHU (J.-B.R., J.F.); InVS (S.C.)
ORAMIP (M.M.), Toulouse, France.
Abstract
Background— Despite the diversity of the studied health outcomes, types and levels of pollution, and various environmental settings, there is substantial evidence for a positive link between urban air pollution and cardiovascular diseases. The objective of this study was to test the associations between air pollutants and the occurrence of acute myocardial infarction (AMI).
Methods and Results— Pollutant concentrations (SO2, NO2, and O3) were measured hourly as part of the automated air quality network. Since 1985, an AMI registry (the Toulouse MONICA Project) has been collecting data in the southwest of France. All cases of AMI and sudden and probable cardiac deaths are recorded for subjects 35 to 64 years of age. We studied the short-term exposure effect of pollution on the risk of AMI (from January 1, 1997, to June 30, 1999) using a case-crossover design method. We performed a conditional logistic regression analysis to calculate relative risks (RRs) and their 95% CIs. After adjustment for temperature, relative humidity, and influenza epidemics, the RRs (for an increase of 5 μg/m3 of O3 concentration) for AMI occurrence were significant for the current-day and 1-day-lag measurements (RR, 1.05; 95% CI, 1.01 to 1.08; P=0.009; and RR, 1.05; 95% CI, 1.01 to 1.09; P=0.007, respectively). Subjects 55 to 64 years of age with no personal history of ischemic heart disease were the most susceptible to develop an AMI (RR, 1.14; 95% CI, 1.06 to 1.23). NO2 and SO2 exposures were not significantly associated with the occurrence of AMI.
Conclusions— Observational data confirm that short-term O3 exposure within a period of 1 to 2 days is related to acute coronary events in middle-aged adults without heart disease, whereas NO2 and SO2 are not.
Key Words: morbidity ; myocardial infarction ; air pollution ; registries ; vasoconstriction
Introduction
Despite the diversity of the studied health outcomes, types and levels of pollution, various environmental settings, and different methodological approaches, substantial evidence for a positive link between urban air pollution and cardiovascular diseases exists. Increasing levels of air pollutant concentration were found to be associated with increased cardiovascular or ischemic heart disease (IHD) mortality and a higher rate of hospital admissions for vascular diseases1–13; however, few reports14 have investigated jointly the relationships between ambient air pollution and cardiovascular morbidity and mortality at the population level. Few studies have investigated mortality on the basis of only the underlying causes stated in death certificates or investigated morbidity on the basis of cardiovascular hospital admissions with a potential selection bias with regard to acute myocardial infarction (AMI) cases. A systematic recording of AMI at the population level over a long period enabled us to study the relationships between air pollutants and mortality and morbidity of IHD simultaneously. We could investigate these relationships thanks to the existence of an acute myocardial infarction registry (Monitoring Trends and Determinants in Cardiovascular Disease [MONICA]) and an air quality network (ORAMIP) covering the same population in the Toulouse area. The objective of the present study was to test the associations between various air pollutants and the daily occurrence of AMI.
Methods
Air Pollution and Meteorological Parameter Measurements
Gas pollutant concentrations were measured hourly as part of the automated French air quality network (ORAMIP) in southwestern France. Sulfur dioxide (SO2) concentrations were measured by UV fluorescence; nitrogen dioxide (NO2) concentrations, by chemiluminescence; and ozone (O3) concentrations, by UV photometry. We primarily used French Environment SA (eg, AC31M for NO2 and O341M for O3) or SERES (SF 2000 for SO2) equipment for the ORAMIP measurements.
At the beginning of the study (January 1, 1997), 14 monitoring stations were located mainly in the metropolitan area of Toulouse (southwestern France), including 4 lead measurement stations. Three types of stations were identified according to the European classification15: stations measuring urban background concentrations of pollutants (n=5) to which the population is exposed, including suburban (n=1) and observation (n=1) stations; stations directly exposed to traffic pollution (n=3); or stations exposed to industrial pollution (n=6), including stations close to a chemical plant site (n=2) and those close to a lead factory (n=4). Only data provided by urban monitoring stations were taken into account to test the relationships between air pollutants and the occurrence of IHD because the goal of the study was to estimate the average daily exposure-response relationship, not to assess the health impact of exposure to air pollution from specific sources (eg, road traffic). For each pollutant (NO2 and SO2), the 24-hour mean concentration was calculated from midnight to midnight and took into account the results from all the urban monitoring stations. For O3, the moving average of the 8 consecutive hours with the highest values of the day was taken into account in the statistical analysis.
Mean daily temperature (degrees Celsius) for the minimum and the maximum values and relative humidity (percent) were available from the weather station located in the metropolitan area of Toulouse. Data on influenza epidemics, expressed as the number of occurrences per week, were obtained from the general practitioner sentinel network of the metropolitan area of Toulouse.
A standardized procedure was used to complete days with missing data.16 A missing value () on day i in year k in monitoring station j was replaced by a weighted average value of the measurements from the rest of the monitoring stations; ie, ijk = i.k x (.jk/..k). Globally, the proportion of missing values in relation to the total values of measured pollutants averaged 2%.
Registry Data
Since 1985, an acute myocardial infarction registry (the Toulouse MONICA Project) has been carried out in southwestern France.17–19 All cases of AMI, sudden cardiac deaths, and probable cardiac deaths are recorded for subjects 35 to 64 years of age. The studied population in 1999 (national population census) was 395 744 inhabitants in the whole area covered by the registry for the eligible age range. The population living in the area covered by both the registry and the air quality network made up the study population. The population exposed during the study period (January 1, 1997, to June 30, 1999) was 685 895 inhabitants. All private and public hospitals within the geographic area of the registry were screened by the medical staff for suspected AMI, and death certificates (suspected coronary deaths, sudden deaths, and fatal cases with insufficient data) were analyzed. An event was considered new if its apparent date of onset was >28 days after any previous AMI event. An event was considered fatal if death occurred before midnight on the 27th day after the initial event. Four definitions were used in this study: definition 1, definite AMI plus possible coronary death plus sudden death plus death with insufficient data; definition 2, definite AMI plus possible coronary death plus sudden death; definition 3, definite AMI plus possible coronary death; and definition 4, definite AMI.20,21
Briefly, definite AMI corresponded to subjects for whom clinical, ECG, and enzymatic data were available. All deaths classified as coronary deaths, sudden deaths, or deaths with insufficient data were validated from all available medical and forensic reports and interviews of general practitioners. Possible coronary deaths corresponded to subjects who died rapidly after suffering from chest pain (typical or atypical symptoms) or who had a known history of CHD without clear evidence supporting another cause of death. Sudden deaths corresponded to subjects who died rapidly (<1 hour) without symptoms, who had no personal history of CHD, or who had evidence supporting another cause of death. Deaths with insufficient data corresponded to deaths with no symptoms (or inadequately described), no history of CHD, and no other diagnosis.21
Only the population living in the Toulouse area covered by urban monitoring stations was studied, and only acute episodes occurring during the studied period were included in the statistical analysis. During the study period, 635, 564, 475, and 399 acute episodes met the required criteria for the definitions 1, 2, 3, and 4, respectively. The proportions of the various events for the most complete definition (definition 1, n=635) were 62.8%, 12.0%, 14.0%, and 11.2% for definite AMI, possible coronary death, sudden death, and death with insufficient data, respectively.
Statistical Analysis
To predict the likely occurrence of an acute event, we used fixed-effect methods for nonrepeated events using conditional logistic regression.22 The case-crossover design was first developed by Maclure23 to study the short-term exposure effect on transient risks. In our study, when 1 subject in the registry population experienced at least 1 event during the studied period, the day of occurrence was defined as the "case day." Two methods were used to select the referent period (control period) on the basis of knowledge of the cyclic and time variability in pollution in the area studied for the case-crossover analysis: the ambidirectional sampling 7-day referent lag (control days were the same day of the previous week and the same day of the week after the case day) and the time-stratified (by months) sampling selection.24 In the ambidirectional sampling selection, the 2 control days were the same day of the week as the case day. For example, for an acute event that occurred on Monday, the 2 control days were the Monday of the previous week and the Monday of the next week. When 2 events occurred on the same day, the number of control days was established accordingly. For each event occurring on a specific day, a matched pair was established. In the time-stratified sampling selection, stratification into separate months, with a retrospective fixed-interval referent selection, was done. The control days were selected 7, 14, 21, and 28 days before the case day. The models were systematically adjusted for relative humidity (in percentage), temperature (minimum or maximum values in degrees Celsius), and influenza epidemics (number of incidents per week). All adjusted variables (relative humidity, temperature, and influenza epidemics) were entered into each model as a single quantitative term.
The relationships between exposure and AMI occurrence were tested considering the exposure to pollutants during the day of AMI occurrence (concurrent day, or D0), the day before AMI occurrence (day lag 1, or D-1), 2 days before (day lag 2, or D-2), and 3 days before (day lag 3, or D-3), as well as the average values of pollutant concentration over 3 days (D0 to D-2) and (D-1 to D-3).
A conditional logistic regression analysis (fixed-effect logistic regression) was performed to calculate relative risks (RRs) and their 95% CIs (PROC PHREG procedure with SAS software). First, to ascertain the correct scale in the logit association between exposure and the occurrence of AMI, the estimated coefficients of the corresponding variables were assessed by calculation of adjusted RRs for tertiles of each pollutant, specifically for O3. The linear trend of the logit was tested with the ordinal values of the tertiles in the conditional logistic regression. These relationships were tested for the 4 definitions of AMI given in this study and for every single-day lag or for the average of 3 consecutive days. The results have shown that the shape of the relationship can be modeled as a logit linear relationship. Finally, the linear association between the pollutant and the occurrence of AMI was analyzed, with the pollutant entered as a single quantitative term in each model. For each pollutant, the RRs of AMI occurrence were given for a continuous variation with an increment of 5 μg/m3 of this pollutant.
The analyses, performed in the whole population living in the area covered by the monitoring stations of the ORAMIP network, were carried out by age groups (35 to 54 and 55 to 64 years), sex, personal history of IHD, and survival status. Finally, RRs were estimated in the population with the potentially highest risk level, ie, the oldest subjects (55 to 64 years), subjects with personal history of IHD, and those who died within 28 days after the occurrence of AMI. Because the 2 methods of referent period (time-stratified or ambidirectional) selection gave similar results, only the results obtained with the time-stratified sampling were taken into consideration.
Results
In Table 7, RRs between O3 exposures and occurrence of acute events were estimated in the oldest subjects with no personal history of IHD. For an increase of 5 μg/m3 (5 units) of O3 concentration, the RRs were similar for a given day lag regardless of the definition of AMI. They were significant for the current day for AMI definitions 1 and 2 and were significant for all AMI definitions for day lag 1 and for the cumulative concentration of O3 (from day lag 0 to 2). The highest RRs and the most significant associations were observed for day lag 1. The RRs were 1.11 (95% CI, 1.04 to 1.19), 1.14 (95% CI, 1.06 to 1.23), 1.15 (95% CI, 1.06 to 1.25), and 1.13 (95% CI, 1.04 to 1.23) for definitions 1, 2, 3, and 4, respectively.
Discussion
The main finding of this study is a positive relationship between acute IHD occurrence and O3 exposure. Moreover, analyses performed from the current day of exposure to day lag 3 suggest a short time effect of ozone on acute IHD occurrence during the first 24 hours after exposure. Indeed, the highest RRs were observed on day lag 1 and then decreased until day lag 3. The cumulative effect of O3 exposure over 3 days was not observed in this study; the RRs were lower than or of a similar magnitude as those of day lag 1. Finally, the sensitivity analysis done with different AMI definitions showed a consistent association between ozone air concentration and AMI occurrences. For day lag 1, the attributable risk for the population 35 to 64 years of age ranged from 4% to 5% according to the definition of AMI used. The attributable risk reached 12% when the oldest population (55 to 64 years of age) without history of IHD was considered.
Analysis of the relationships by individual level of age and personal history of IHD revealed that the oldest subjects (in a middle-aged population of 35 to 64 years) without a personal history of IHD were more susceptible to an acute event when ozone concentrations increased. The associations according to survival status were identical, however. In this case, the relationship was of a similar strength among subjects who died after an acute IHD event and those who had a 28-day survival. The relationships were statistically significant in the latter group because of higher statistical power (the number of subjects was double).
The short-term effect of O3 on various health outcomes has been studied, with a focus on mortality and hospital admissions,13,25,26 but few studies have assessed a positive association between O3 exposure and AMI morbidity/mortality. A large multicenter study carried out in the United States, NMMAPS (National Morbidity, Mortality, and Air Pollution Study), reported a significant impact of O3 during the summer and an increased mortality of 0.41% associated with an increase of 20 μg/m3 in daily O3 concentrations at day lag 0.25
In our study, the lack of association between NO2 or SO2 and AMI occurrence could originate from the moderate levels of air concentration pollutants in the Toulouse area, resulting in a limited range variability of concentration values. Generally speaking, comparisons with cities where associations between gaseous pollutants and cardiovascular hospitalization rates were significant showed median values of air pollutant concentrations that were 2- to 3-fold higher than in the Toulouse area (NO2=29.3 μg/m3, SO2=7.4 μg/m3) (Rome: NO2=86.0 μg/m3; Hong Kong: NO2=53.5 μg/m3, SO2=14.5 μg/m3; London: NO2=61.2 μg/m3, SO2=20.6 μg/m3; Denver: SO2=15.2 μg/m3).6–8 On the other hand, O3 concentration was positively associated with the risk of AMI occurrence in Toulouse (median value of O3=74.8 μg/m3). In Denver, which has a similar median level of O3 concentration (50.3 μg/m3), a 25th to 75th percentile change of ozone increased the risk of hospitalization for AMI and coronary atherosclerosis when the current day of O3 exposure and AMI admissions was considered.7 In Barcelona (median value of O3=70.8 μg/m3),2 an increase of 100 μg/m3 of ozone air concentration was significantly associated (RR, 1.09) with increased cardiovascular mortality in the elderly during the summer. In Mexico City (median O3=87.0 μg/m3), the mean concentration of O3 over a 2-day period was associated with a 1.8% increase in cardiovascular mortality.3 In Hong Kong and London,6 with half the O3 concentration recorded in Toulouse (median O3=28.3 and 32.0 μg/m3, respectively), associations between O3 exposure and cardiac admissions were less consistent and remained uncertain. In Montreal,5 with a median concentration of ozone that was one third the value recorded in Toulouse, cardiovascular mortality increased by 2.5% (95% CI, 0.2 to 5.0) with an increase of 21.3 μg/m3 of O3. Globally, the conflicting results could be due to the multiplicity of the studied health outcomes, the various types and levels of pollution, and investigation methodologies.
The age-related increasing risk observed in the present study was an expected result because of a higher prevalence of deleterious pathologies among the elderly (congestive heart failure or arrhythmia) and a growing sensitivity to the effects of ambient air pollutants.27 On the other hand, the higher risk among subjects without a personal history of IHD was unexpected. Some pathophysiological hypotheses can be suggested to explain the association between the short-term effect of ozone and AMI occurrence.
One possibility is that alterations in arterial tone could be a relevant mechanism explaining an increased incidence of acute cardiac events in populations exposed to air pollution. Short-term exposure to a mix of fine particulate air pollution (PM 2.5) and O3 at levels observed in urban environment induces artery vasoconstriction in healthy adults28; however, the effect of O3 alone on endothelial function and arterial tone remains to be investigated. In rabbits,29 exposure to ozone for 4 hours increased baseline values of total vascular resistance in pulmonary vessels that were directly exposed, but systemic vessels were not investigated. Thus, one could hypothesize that O3 may alter systemic vascular tone like it alters pulmonary tone in direct contact, but this hypothesis needs to be investigated. The lack of a significant association between ozone exposure and AMI occurrence in subjects with a history of IHD could be explained by the preventive action of standard treatments against relapses that can be taken by these patients: vasodilator drugs (nitrate), statin with its pleiotropic effects (improvement of endothelial dysfunction, increased nitric oxide bioavailability, antioxidant properties, inhibition of inflammatory responses, and stabilization of atherosclerotic plaques),30 and ACE inhibitors (improvement of endothelial function, contribution to increased NO bioactivity, antiproliferation activity, modulation of sympathetic activity, and attenuation of atherosclerosis).31
This study has some limitations. This relationship was established at the population level, and no specific information system had been developed for this investigation. We used an information system (AMI MONICA registry) existing since 1985 in the Toulouse area. The methodology of the MONICA project has been validated and can be considered a reliable device for the AMI registration.17,20 The recording of AMI at the population level does not completely eliminate all the risks of potential biases but enables the reduction of selection biases of AMI cases.
The age of the population covered by the registry ranged from 35 to 64 years, thus limiting the generalization of the conclusions because subjects >64 years of age are more vulnerable to pollution exposure than younger subjects. Nevertheless, our results underline the deleterious influence of air pollution on the vascular system even at young ages.
Pollution risk factors were measured as ecological variables with measurements provided by the urban monitoring stations, assuming that exposure was homogeneous all over the studied area, thus limiting the strength of the relationship because of probable exposure misclassification for some individuals. Indeed, exposure misclassification would only bias to the null, and the association could be underestimated.32
The case-crossover design is a reliable method to study short-term exposure effects on transient risks. It eliminates the potential biases resulting from seasonality and long-term trends of air pollutants insofar as the referent periods are meticulously selected.24,33 In the present study, the results obtained with the 2 methods of referent-period selections (ambidirectional sampling and time-stratified sampling selection) were similar.
In conclusion, these observational data confirm that short-term O3 exposure within a period of 1 to 2 days is related to acute coronary events in middle-aged adults without heart disease, whereas NO2 and SO2 are not.
Acknowledgments
We thank the Federation Fran;aise de Cardiologie, the Institut National de la Sante et de la Recherche Medicale, and the Institut de Veille Sanitaire for their support that enabled this work. We thank Thibaud Bouillie for preparing the database, all the investigators of the Toulouse MONICA registry for their invaluable contributions to the careful collection and validation of the data, and the physicians and cardiologists of private and public institutions who helped in this process.
References
Schwartz J, Morris R. Air pollution and hospital admissions for cardiovascular disease in Detroit, Michigan. Am J Epidemiol. 1995; 142: 23–35.
Sunyer J, Castellsague J, Saez M, Tobias A, Anto JM. Air pollution and mortality in Barcelona. J Epidemiol Community Health. 1996; 50 (suppl 1): s76–s80.
Borja-Aburto VH, Castillejos M, Gold DR, Bierzwinski S, Loomis D. Mortality and ambient fine particles in southwest Mexico City, 1993–1995. Environ Health Perspect. 1998; 106: 849–855.
Ballester F, Tenias JM, Perez-Hoyos S. Air pollution and emergency hospital admissions for cardiovascular diseases in Valencia, Spain. J Epidemiol Community Health. 2001; 55: 57–65.
Goldberg MS, Burnett RT, Brook J, Bailar JC 3rd, Valois MF, Vincent R. Associations between daily cause-specific mortality and concentrations of ground-level ozone in Montreal, Quebec. Am J Epidemiol. 2001; 154: 817–826.
Wong CM, Atkinson RW, Anderson HR, Hedley AJ, Ma S, Chau PY, Lam TH. A tale of two cities: effects of air pollution on hospital admissions in Hong Kong and London compared. Environ Health Perspect. 2002; 110: 67–77.
Koken PJ, Piver WT, Ye F, Elixhauser A, Olsen LM, Portier CJ. Temperature, air pollution, and hospitalization for cardiovascular diseases among elderly people in Denver. Environ Health Perspect. 2003; 111: 1312–1317.
D’Ippoliti D, Forastiere F, Ancona C, Agabiti N, Fusco D, Michelozzi P, Perucci CA. Air pollution and myocardial infarction in Rome: a case-crossover analysis. Epidemiology. 2003; 14: 528–535.
Zmirou D, Barumandzadeh T, Balducci F, Ritter P, Laham G, Ghilardi JP. Short term effects of air pollution on mortality in the city of Lyon, France, 1985–90. J Epidemiol Community Health. 1996; 50 (suppl 1): S30–S35.
Nafstad P, Haheim LL, Wisloff T, Gram F, Oftedal B, Holme I, Hjermann I, Leren P. Urban air pollution and mortality in a cohort of Norwegian men. Environ Health Perspect. 2004; 112: 610–615.
Peters A, Dockery DW, Muller JE, Mittleman MA. Increased particulate air pollution and the triggering of myocardial infarction. Circulation. 2001; 103: 2810–2815.
Le Tertre A, Medina S, Samoli E, Forsberg B, Michelozzi P, Boumghar A, Vonk JM, Bellini A, Atkinson R, Ayres JG, Sunyer J, Schwartz J, Katsouyanni K. Short-term effects of particulate air pollution on cardiovascular diseases in eight European cities. J Epidemiol Community Health. 2002; 56: 773–779.
Brook RD, Franklin B, Cascio W, Hong Y, Howard G, Lipsett M, Luepker R, Mittleman M, Samet J, Smith SC Jr, Tager I, for the Expert Panel on Population and Prevention Science of the American Heart Association. Air pollution and cardiovascular disease: a statement for healthcare professionals from the Expert Panel on Population and Prevention Science of the American Heart Association. Circulation. 2004; 109: 2655–2671.
Eilstein D, Quenel P, Hedelin G, Kleinpeter J, Arveiler D, Schaffer P. Air pollution and myocardial infarction: Strasbourg France, 1984–89. Rev Epidemiol Sante Publique. 2001; 49: 13–25.
Classification and Criteria for Setting Up Air Quality Monitoring Stations. Paris, France: ANDEME Editions; 2002.
Katsouyanni K, Schwartz J, Spix C, Touloumi G, Zmirou D, Zanobetti A, Wojtyniak B, Vonk JM, Tobias A, Ponka A, Medina S, Bacharova L, Anderson HR. Short term effects of air pollution on health: a European approach using epidemiologic time series data: the APHEA protocol. J Epidemiol Community Health. 1996; 50 (suppl 1): S12–S18.
Marques-Vidal P, Ruidavets JB, Cambou JP, Ferrières J. Incidence, recurrence, and case fatality rates for myocardial infarction in southwestern France, 1985 to 1993. Heart. 2000; 84: 171–175.
Marrugat J, Ferrières J, Masia R, Ruidavets J, Sala J. Differences in use of coronary angiography and outcome of myocardial infarction in Toulouse (France) and Gerona (Spain): the MONICA-Toulouse and REGICOR investigators. Eur Heart J. 2000; 21: 740–746.
Ferrières J, Cambou JP, Ruidavets JB, Pous J. Trends in acute myocardial infarction prognosis and treatment in southwestern France between 1985 and 1990 (the MONICA Project-Toulouse). Am J Cardiol. 1995; 75: 1202–1205.
Tuomilehto J, Kuulasmaa K. WHO MONICA Project: assessing CHD mortality and morbidity. Int J Epidemiol. 1989; 18 (suppl 1): S38–S45.
Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project: registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation. 1994; 90: 583–612.
Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 1989.
Maclure M. The case-crossover design: a method for studying transient effects on the risk of acute events. Am J Epidemiol. 1991; 133: 144–153.
Levy D, Lumley T, Sheppard L, Kaufman J, Checkoway H. Referent selection in case-crossover analyses of acute health effects of air pollution. Epidemiology. 2001; 12: 186–192.
Samet JM, Dominici F, Curriero FC, Coursac I, Zeger SL. Fine particulate air pollution and mortality in 20 U.S. cities, 1987–1994. N Engl J Med. 2000; 343: 1742–1749.
Le Tertre A, Quenel P, Eilstein D, Medina S, Prouvost H, Pascal L, Boumghar A, Saviuc P, Zeghnoun A, Filleul L, Declercq C, Cassadou S, Le Goaster C. Short-term effects of air pollution on mortality in nine French cities: a quantitative summary. Arch Environ Health. 2002; 57: 311–319.
Mann JK, Tager IB, Lurmann F, Segal M, Quesenberry CP Jr, Lugg MM, Shan J, Van Den Eeden SK. Air pollution and hospital admissions for ischemic heart disease in persons with congestive heart failure or arrhythmia. Environ Health Perspect. 2002; 110: 1247–1252.
Brook RD, Brook JR, Urch B, Vincent R, Rajagopalan S, Silverman F. Inhalation of fine particulate air pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation. 2002; 105: 1534–1536.
Delaunois A, Segura P, Montano LM, Vargas MH, Ansay M, Gustin P. Comparison of ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol Appl Pharmacol. 1998; 150: 58–67.
Davignon J. Beneficial cardiovascular pleiotropic effects of statins. Circulation. 2004 15; 109 (suppl III): III-39–III-43.
Ferrari R. Effect of ACE inhibition on myocardial ischaemia. Eur Heart J. 1998; 19 (suppl J): J30–J35.
Armstrong BG. Effect of measurement error on epidemiological studies of environmental and occupational exposures. Occup Environ Med. 1998; 55: 651–656.
Robins J, Pike M. The validity of case-control studies with non random selection of controls. Epidemiology. 1990; 1: 273–284.(Jean-Bernard Ruidavets, M)