Secondary Logo

Journal Logo

Air Pollution: Original Article

Impact of Fine and Ultrafine Particles on Emergency Hospital Admissions for Cardiac and Respiratory Diseases

Belleudi, Valeriaa; Faustini, Annunziataa; Stafoggia, Massimoa; Cattani, Giorgiob; Marconi, Achillec; Perucci, Carlo A.a; Forastiere, Francescoa

Author Information
doi: 10.1097/EDE.0b013e3181d5c021


The short-term effects of airborne particulate matter (with an aerodynamic diameter ≤10 μm [PM10]) on daily mortality have been studied in several countries. Meta-analyses and multicity studies have estimated effects ranging from 0.3% to 1.5% increase in mortality rates per 10 μg/m3 PM10.1,2 The impact of airborne particulates on acute cardiovascular and respiratory morbidity has also been reported for a range of outcomes, including subclinical effects, lung function, respiratory or cardiac symptoms, physician office visits, emergency room visits, and hospital admissions.2,3

There are, nonetheless, several issues regarding the health effects of airborne particles that have not yet been clarified. The size and chemical composition of the particulate matter eliciting the specific health effects have not been fully investigated. Fine (aerodynamic diameter ≤2.5 μm [PM2.5]) and ultrafine particles (diameter ≤0.1 μm) have been hypothesized as the effective toxic fractions of particulate matter, because they promote and maintain oxidative stress both at the respiratory level (the entry system) and at the systemic level where oxidative stress causes inflammation.4 Although extensive research has been conducted (mainly in the United States) on the effects of PM2.5 on mortality and hospital admissions,5,6 there are only a few investigations on the role of different particle sizes that include ultrafine particles. In Europe, despite the large problem of traffic-related air pollution in urban areas (especially from diesel vehicles), particulate matter has not been continuously monitored, limiting the number of investigations on the effect of PM2.5.7,8 In addition, the technical difficulties in measuring ultrafine particle concentration and estimating the population exposure greatly limit the overall availability of studies addressing the impact of ultrafine particles on human health.

In 1995, Seaton et al9 first proposed that ultrafine particles are the effective damaging fraction of PM10. Since then, several investigations have been conducted on the human toxicology and the pathologic mechanisms of ultrafine particles,10 with important contributions from studies investigating changes in cardiovascular11 or respiratory functions.12 However, the epidemiologic evidence linking ultrafine particles with cardiovascular or respiratory health effects is still limited and controversial. Stolzel et al,13 in an extension of the study by Wichmann et al,14 reported associations of ultrafine particles with daily mortality in Erfurt (Germany), whereas Kettunen et al15 found a positive relationship of ultrafine particles with stroke mortality during summer in Helsinki. Subsequent investigations in Helsinki failed to show a clear effect of ultrafine particles either on natural or on cardiovascular mortality.16 Andersen et al17 found that the adverse health effects of particulate matter on cardiovascular and respiratory hospital admissions in the elderly in Copenhagen were mainly attributable to PM10 and accumulation mode particles (0.1–0.5 μm) but not to particles <0.1 μm. In the 2 reports from Helsinki dealing with adults and the elderly patients, cardiovascular and respiratory hospital admissions16 and respiratory emergency room visits18 were related mainly to PM2.5 and accumulation mode particles, rather than to particles <0.1 μm.

The aim of this study was to evaluate the association of PM10, PM2.5, and ultrafine particles with emergency hospital admissions for cardiac and respiratory diseases in the city of Rome. We explored several lag times between exposures and health effects and evaluated potential effect modification by age, previous chronic obstructive pulmonary disease (COPD), and season.


Health Events

All the hospitalizations of residents of Rome (approximately 2.7 million inhabitants) discharged with a diagnosis of cardiac or respiratory disease between 10 April 2001 and 31 December 2005 were identified by the regional hospital discharge registry. Events were further restricted to admissions in hospitals within the city boundaries to increase the likelihood that the exposure corresponded with the air pollution being measured. We limited the health series to patients aged ≥35 years, and we excluded all rehospitalizations immediately after a hospital discharge (up to 2 days) to limit exposure misclassifications. To identify acute events, scheduled admissions were excluded.18,19,20

We used detailed disease definitions to increase the specificity of the impact estimate and to contribute to understanding the underlying biologic mechanisms.6,19 We selected primary diagnoses of the following cardiac and respiratory diseases: acute coronary syndrome (myocardial infarction and unstable angina; International Classification of Diseases [ICD]-9 code 410 and 411), heart failure (ICD-9 code 428), lower respiratory tract infections, including acute bronchitis and pneumonia (ICD-9 code 466 and 480–486), and chronic-obstructive pulmonary disease (COPD) (ICD-9 code 491, 492, 494, and 496). Cases of acute coronary syndrome were also identified by complications reported in the primary diagnosis (ICD-9 code 414.10, 423.0, 427.1, 427.41, 427.42, 427.5, 428.1, 429.5, 429.6, 429.71, 429.79, 429.81, 518.4, 780.2, and 785.51), together with myocardial infarction (ICD-9 code 410) recorded as the secondary diagnosis. Cases of COPD were also identified by respiratory failure (ICD-9 code 518.8) listed as a principal diagnosis and COPD as the secondary diagnosis. The diagnostic accuracy of the regional information system has been validated for myocardial infarction (ICD-9 code 410) (96%) and pneumonia (80%). For each hospitalized subject, we had information on any hospitalizations in the 5 preceding years for COPD that were recorded as a primary diagnosis or as a secondary condition, using an individual linkage procedure with the regional database.

Environmental Data

There is an extensive air pollution network in Rome, managed by the regional Environmental Protection Agency, which routinely monitors several pollutants, including PM10.21 However, PM2.5 and ultrafine particles were not regularly monitored. We used instead a monitoring station 2 km east of the city center on the grounds of the Italian National Institute of Health (NIH) in operation since 1980 and specifically set up to measure ultrafine particles for the European Union-funded Health Effects of Air Pollution on Susceptible Subpopulations (HEAPSS) project.22 The monitoring station provided data for PM10, PM2.5 and ultrafine particles for the period 10 April 2001–31 December 2005. Gravimetric measurement methods were used to determine the PM10 mass fraction of suspended particulate matter,23 and a provisional method was used for PM2.5 that follows the standards set in 2005.24 Results were reported in mean daily mass concentrations (μg/m3). Particle number concentration was measured using optical condensation particle counters (CPC, model 3022A; TSI Inc, Shoreview, MN) with 50% counting efficiency at 0.007 μm. The Health Effects of Air Pollution on Susceptible Subpopulations (HEAPSS) standard operating procedures regarding the measurement protocol of total particle number concentration in ambient air were strictly followed.25 Because ultrafine particles typically dominate ambient total particle number concentrations, we used particle number concentration as a proxy for ultrafine particles.

The NIH monitoring station was missing data at various times for at least 1 of the 3 PM fractions. A total of 45% of the PM10 data were missing. To impute missing PM10 data, we used daily concentrations from 4 stations of the regional monitoring network. These stations are located approximately at the cardinal points of the city and are far from each other. They operated over the same time period as the central NIH station and used a PM10 measurement method (based on beta array attenuation) equivalent to the reference method used at the NIH site. The PM10 aggregated daily mean from the 4 stations was preferred to that from each individual station because this showed the highest correlation (r = 0.81) with PM10 measures from the NIH station. With this method, 1704 (940 original + 764 imputed) daily data points for PM10 were available.

A total of 29% of the PM2.5 data were missing. Missing data for PM2.5 were imputed using season-specific PM2.5-PM10 regression values (measured at the NIH station), because alternative direct PM2.5 measurements were unavailable. In this way, 1448 (1228 original + 220 imputed) daily data points for PM2.5 were available. Finally, data for 19% of the particle number concentration were missing. Some missing data were imputed from a particle number concentration station located elsewhere in Rome. Although the station had worked for only 398 days during 2002–2004, the correlation between the measures from the 2 sites (r = 0.79) was high enough to allow their use to predict the missing particle number concentration values after appropriate scaling. In this way, a total of 1433 (1401 + 32 imputed) daily data points were available.

Daily information on temperature, humidity, and barometric pressure was provided by the Italian Air Force Meteorological Service from the Ciampino Airport (southeast of Rome). Apparent temperature26 (a composite index that takes into account both air temperature and humidity) was calculated. Weekly estimates of influenza incidence were reported by the Italian National Health Service from the national surveillance system.

Data Analysis

A case-crossover design was used to study the association between particles and hospitalization for cardiac or respiratory diseases.27 We selected control days using a time-stratified approach28 that divides the study period into monthly strata, selecting control days for each case on the same days of the week in the stratum. The concentration differences (for PM10, PM2.5 and particle number concentration) between the event day and the control days were the relevant exposure metric29 in the case-crossover approach. This approach controls for season by matching month and year. It also partly controls for other variables such as weather, because all comparisons are made within the same month. Finally, all time-invariant or slowly varying risk factors are controlled for by design.19

We performed a conditional logistic regression analysis to study the association of PM10, PM2.5, and particle number concentration and hospitalizations for acute coronary syndrome, heart failure, lower respiratory tract infections, and COPD. The regression model controlled for the confounding effects of apparent temperature, barometric pressure, temporary population decreases in the summer and during holidays, and influenza epidemics. Cubic splines of apparent temperature (lag 0) were used, with knots located at the 25th and the 75th percentiles of the distribution. Barometric pressure (lag 0) was adjusted using a linear term. Population decrease was categorized as follows: code 2 indicated the 16 days of August before or after 15 August (national holiday), when most of the people have vacation outside the area and a decrease in hospital services is observed; also the few days around Christmas and Easter had a code of 2. Summer days from July 16 to August 31 and other national or local holidays in the year had a code of 1. All other days were coded 0. A dummy variable was used to indicate days in which an influenza epidemic was suggested (59 days) on the basis of the surveillance system in Rome.

We explored several lags of effect. The individual lags from day 0 to day 6 were examined, and then, short-term effects were explored in the interval 0–1 day (also to allow comparisons with previous literature results) and 0–2 days, and extended effects were explored over an interval of 0–5 days (the results for 0–6 days were very similar). We expressed the results as the percent change (95% confidence interval [CI]) for a daily increase of 14 μg/m3 PM10, 10 μg/m3 PM2.5 and 9392 particles/cm3 for particle number concentration. As interquartile ranges (IQRs), these values allowed us to compare the results for different PM fractions, while allowing us to use 10 μg/m3 PM2.5 as reference. These values were derived as follows: a percentile distribution of the differences between the event day (all outcomes combined) and the control days for PM2.5 was calculated, and the IQR was divided by 10. This ratio was applied to the IQRs of the differences between the event day and the control days for PM10 and particle number concentration, and the corresponding values were used as an exposure metric comparable with that of PM2.5.

We considered age, previous COPD, and season as potential effect modifiers. The analysis was stratified into 3 age groups defined as 35–64, 65–74, and ≥75 years.6 Previous COPD as effect modifier was considered only for diagnoses of acute coronary syndrome and heart failure. Finally, to explore effect modification by season, data were analyzed stratifying for winter (December–March), summer (June–September), and transition periods (spring/fall, including April, May, October, and November).30 Separate models were run for each of the 4 outcomes.

We ran 2-pollutant models with PM10 (or PM2.5) and particle number concentration. In addition, several sensitivity analyses were conducted to evaluate the robustness of the results. In particular, we ran models without adjustment for influenza epidemics, with the actually measured PM10 (excluding the estimated data) and considering the PM10 values from the average of the city monitors. A sensitivity analysis was also conducted with respect to the spatial heterogeneity of exposure, considering only subjects living within the so called “green area,” a large portion of the city (encompassing about 56% of the population) that extends from the historical city center through the main urban area of Rome.

All analyses were conducted using SAS version 8.2 (SAS Institute Inc, Cary, NC) and STATA version 10.0 (Stata Corp, College Station, TX). There was no need for an ethical committee approval of the study because the database of hospital admissions is collected for administrative purposes and can be used for research by public health institutions.


A total of 90,056 emergency hospital admissions for cardiac diseases were observed in the eligible population during the period under study (Table 1).Acute coronary syndrome and heart failure were the most frequent causes of hospitalization (25% and 20% of the total cardiac admissions, respectively). A total of 38,735 respiratory emergency admissions were recorded, of which 29% was lower respiratory tract infections and 39% was COPD. Daily hospitalization rates for these conditions increased with age, especially, for heart failure and COPD. There was a greater tendency for very elderly patients to have emergency hospitalizations. The percentage of acute coronary syndrome and heart failure patients with previous hospitalizations for COPD was 19% and 45%, respectively.

Emergency Hospital Admissions for Cardiac and Respiratory Diseases (by Age Groups) in Rome Residents Hospitalized in Rome Between 10 April 2001 and 31 December 2005

The distribution of the environmental variables and their correlations are reported in Table 2. PM2.5 concentration was about two-thirds that of PM10. PM2.5 and PM10 were strongly correlated (r = 0.84), whereas particle number concentration was only moderately correlated with PM2.5 (r = 0.55) and PM10 (r = 0.57). The concentration of pollutants was highest in winter and lowest in summer, with the exception of PM2.5, which showed similar levels in the summer and in the transition periods. The annual mean levels of PM2.5 and PM10 varied little throughout the study period, but a decreasing trend was evident for particle number concentration. The eTable ( shows the distributions and the correlation coefficients of the differences between the event days and control days, which are the actual exposure metrics in the study.

Environmental Variables in Rome (Daily Mean, SD, Percentiles, IQR) and Pearson's Correlation Coefficients Among Environmental Variables

Table 3 reports the associations of the 3 particle fractions with the 4 health conditions, examining the individual 0–6 day lags and the cumulative lags. Admissions for acute coronary syndrome increased in relation to PM10 and PM2.5, and lag 0, lag 0–1, and lag 0–2 had the strongest effect. For an immediate lag (0), the effect estimates were 1.1% for a 14 μg/m3 increase in PM10 and 2.3% for a 10 μg/m3 increase in PM2.5 levels. No association was found for particle number concentration. Hospitalizations for heart failure increased in association with daily variations of all particulate air pollutants. Again, the effect was immediate (lag 0), with the largest effect found for PM2.5 (lag 0, 2.4% [95% CI = 0.3% to 4.5%]), although PM10 (lag 0, 1.8% [0.1% to 3.5%]) and particle number concentration (lag 0, 1.8% [0.4% to 3.2%]) were also strongly associated. Of note, an effect was also evident even after an extended lag, which decreases only at lag 6. The effect size for the cumulative 0–5 lag was 2.8% for PM10 and 2.4% for particle number concentration.

Associations (Percentage Changes) Between PM10, PM2.5 and Particle Number Concentration and Admissions for Acute Coronary Syndrome, Heart Failure, Lower Respiratory Tract Infections, and COPD in Patients Aged ≥35 Years in Rome During the Study Period

When respiratory hospitalizations were considered (Table 3), daily admissions for lower respiratory tract infections increased in association with PM10 and PM2.5 but not with particle number concentration. The largest effect was for PM2.5 at lag 2 (2.8% [0.5% to 5.2%]) and at lag 3 (3.0% [0.8% to 5.3%]). In contrast, exacerbations of COPD were positively related to particle number concentration at lag 0 (1.6% [0.0% to 3.2%]), whereas the association for PM10 and PM2.5 at lag 0 were positive but with wide confidence intervals. Negative associations were found at lag 3 for PM10 and lag 1 for PM2.5.

Table 4 examines the associations between the 3 pollutants and acute coronary syndrome for the potential effect modifiers. Only the results for lag 0 are reported. Although no clear effect modification by age was seen for PM10, the effect of PM2.5 on acute ischemic conditions was higher in people aged 65–74 years. When we examined the data for previous COPD and age, there was a strong effect of PM10 on younger subjects (35–64 years) with previous COPD. In contrast, there was a strong effect of PM2.5 among 65–74-year-old subjects with no previous COPD. The largest PM10 and PM2.5 effect on acute coronary syndrome was detected during the winter, whereas no association was seen in the summer or during spring and fall.

Associations (Percentage Changes) Between PM10, PM2.5 and Particle Number Concentration and Admissions for Acute Coronary Syndrome (Lag 0) in Patients Aged ≥35 Years, by Age, Previous COPD, and Season in Rome During the Study Period

Table 5 reports the associations for heart failure (lag 0) stratified by potential effect modifiers. When considering PM10, there was an increasing gradient of the effect by age, with the strongest association found among those aged ≥75 years. For both PM2.5 and particle number concentration, there was a clear increased risk for both ages 65–74 and ≥75 years. Previous COPD was not an obvious effect modifier, although subjects with a negative history tended to have stronger effect estimates, especially among subjects aged ≥75 years. Finally, the strongest effects for the particulate fractions were again seen in winter, although a positive effect was found also during the transitional periods.

Associations (Percentage Changes) Between PM10, PM2.5, and Particle Number Concentration and Admissions for Heart Failure (Lag 0) in Patients Aged ≥35 Years, by Age, Previous COPD, and Season in Rome During the Study Period

The detailed results for respiratory infections (lag 2) and COPD (lag 0) are reported in Table 6. The effects of PM10 and of PM2.5 on respiratory infections were strongest among those aged ≥75 years. The association of particulate number concentration with COPD admissions was strongest in subjects aged ≥65 years and was more evident during spring/fall.

Associations (Percentage Changes) Between PM10, PM2.5, and Particle Number Concentration and Admissions for Lower Respiratory Tract Infections (Lag 2) and COPD (Lag 0) in Patients Aged ≥35 Years, by Age, Previous COPD, and Season in Rome During the Study Period

When 2-pollutant models were considered, on combining PM2.5 with particle number concentration for acute coronary syndrome admissions (lag 0), the effect of PM2.5 increased (2.8% [0.5% to 5.1%]), whereas the association with particle number concentration became negative (−1.6%). For heart failure, when both PM10 and particle number concentration were considered at lag 0, the PM10 effect decreased to 0.5% (−1.6% to 2.6%), whereas the particle number concentration effect remained (1.7% [0.0% to 3.2%]). When both PM2.5 and particle number concentration were considered, the effect of both weakened, with a stronger effect for PM2.5 (1.51% [−1.0% to 4.1%]) than for particle number concentration (0.4% [−1.6% to 2.4%]). For respiratory infections (lag 2), only small decreases in the PM10 and PM2.5 effects were seen when controlling for particle number concentration. In contrast, adjustment for PM10 or for PM2.5 did not alter the association of COPD (lag 0) with particle number concentration (1.9% [0.1% to 3.8%] and 1.3% [−0.8% to 3.5%], respectively).

Because air pollution may affect human viral response, we were concerned that adjustment for influenza might mask the effect of air pollution. We then performed models without control for influenza epidemic. Similar results were obtained for all the outcomes, although the effect estimates for PM2.5 and respiratory infections, without influenza, actually increased (lag 2, 3.0% [0.7% to 5.4%]). There was no statistical interaction between influenza periods and PM2.5 on respiratory infections. Again, similar estimates for PM10 were obtained when only the actual measured values were used (rather than the measured and the estimated values) and when the PM10 average of the city monitors was used as the exposure variable. When the data set was restricted to subjects living in the main urban area of Rome (55% of acute coronary syndrome cases, 57% of heart failure cases, 48% of lower respiratory tract infections, and 42% of COPD cases), results were confirmed.


We found associations between airborne particles of various sizes and hospital admissions for cardiac and respiratory diseases. The strength of these associations, the specific disease entities, and the latency between exposure and outcome (lag) differed with particle size. The most important increases in hospital admissions were in association with PM2.5 for both acute ischemic conditions and heart failure, especially in the elderly patients and at immediate lag. The impact of fine particles was also high for respiratory tract infections, with a longer lag. Ultrafine particles played a role in hospitalizations for heart failure and COPD, although, the evidence of an association with heart failure was less than that for PM2.5. A difference emerged regarding the latency of the particle number concentration effect: the effect was immediate for COPD admissions and delayed for heart failure admissions. Previous COPD was not an obvious effect modifier, although there was the suggestion of a stronger association of PM10 with acute ischemic events for young adults with COPD.

The association of PM2.5 with acute ischemic conditions is consistent with most previous studies6,8,19 both in the strength of the association and in the short latency. A greater increase (4.5%) in coronary events was reported in a study of patients who had undergone cardiac catheterization.31 Although Forastiere et al,22 Sunyer et al,32 and De Leon et al33 have reported higher risks of circulatory deaths in subjects with COPD, 2 earlier studies20,34 found no evidence of effect modification by COPD on the association between air pollution and hospitalizations for ischemic heart disease. In our investigation, the stronger association of PM10 with acute coronary syndrome among adults aged 35–64 years with COPD may be only a chance finding, or it may indicate a specific vulnerability related to active smoking in the younger subgroup.

Studies on the effect of PM2.5 on heart failure admissions are rare, and results have been inconsistent: an effect size similar to the one estimated has been reported in the large US study,6 but negative results have also been reported.35,36 We were able to confirm that the effect was present mainly in the elderly patients, although the effect was not particularly stronger among people with previous COPD. Heart failure complication in the natural history of COPD may be independent of air pollution exposure.

A specific role of ultrafine particles in increasing cardiac hospital admissions was previously reported for first acute myocardial infarctions, with the highest impact on fatal cases37 and cardiac readmissions in myocardial infarction survivors.38 However, 2 more recent studies16,17 did not find an association of ultrafine particles on hospital admissions for cardiovascular diseases. Contrary to expectations based on toxicologic evidence,11,39 our results do not support an association of ultrafine particles with acute ischemic events. No previous studies have analyzed the association between ultrafine particles and heart failure admissions, and thus, our results are original in this respect, especially with regards to the prolonged lag.

The effects of airborne particles on hospitalizations for lower respiratory tract infections and COPD in this study are less impressive than those reported in previous literature. Host et al8 found a 2.5% excess risk of lower respiratory tract infections after short-term PM2.5 exposure in 6 French cities; Dominici et al6 found a 1.4% and 1.6% increase in hospitalization for lower respiratory tract infections and COPD, respectively, in the elderly patients; Zanobetti and Schwartz19 found a 6.5% increase in hospitalizations for pneumonia in the elderly; Peel et al40 found a 3% increase in risk of emergency department visits for pneumonia. For COPD only, excess risks of 3% (emergency room visits) in the elderly18 and of 3% (hospitalizations) for 10 μg/m3 PM2.5 were repor ted.41 Among the few articles that studied ultrafine particles and respiratory diseases,17,18,42–44 only one18 found an association between ultrafine particles and emergency room visits for asthma in children, and for combined asthma-COPD in the elderly patients. The Aitken mode particles (aerodynamic diameter of 0.03–0.1 μm) showed a positive though modest effect on asthma-COPD visits, whereas a stronger increase was observed for accumulation mode particles (aerodynamic diameter of 0.1–0.29 μm); both effects were immediate. We found a delayed PM2.5 effect for lower respiratory tract infections, especially in the elderly patients, and the results were somewhat weaker than previously reported. This might be explained by our choice of studying only adult and elderly populations (in fact, we included only 61% of the total emergency lower respiratory tract infections occurring at all ages), and there is a possibility that we missed a potentially important effect among children.

Although different measures of the smallest particle concentrations were used in the previous study in Helsinki,16 both PM2.5 and accumulation mode particles (0.1–0.29 μm) caused an increase in hospitalizations for asthma and COPD at short lag (0 and 1). Our results for COPD admissions are difficult to interpret. We observed an association of particle number concentration on COPD admissions at lag 0. However, at least in 2 instances (lag 3 for PM10 and lag 2 for PM2.5), increased pollutant levels were related to lower COPD admissions. Although chance can be an explanation for these findings, a strong positive association between PM10 and respiratory mortality has been estimated for the same period in Rome using the same case-crossover approach (lag 0–5, for 10 μg/m3 PM10, 4.9% [1.8% to 8.0%]). The effect for cardiovascular mortality was lower (lag 0–1, for 10 μg/m3 PM10, 1.2% [0.1% to 2.2%]). The PM10 effect on respiratory mortality was similar for those dying outside the hospital and inhospital. A possible suggestion is that the pool of COPD cases at risk for PM-related admissions is depleted from PM-related fatal exacerbation of COPD; as a consequence, a decreased probability of hospitalization is observed. Of course, this hypothesis of harvesting the at-risk pool needs to be confirmed in other urban locations with high PM-related respiratory mortality.

In interpreting the findings, consideration needs to be given to the inherent limitations of the data. The use of hospitalizations as the only outcome measure may underestimate the impact of the exposure when compared with emergency room visits. Although hospitalization discharge reports are collected for administrative purposes, and some degree of misclassification of the diagnoses could have occurred, we used the principal diagnoses to control this problem. By including only unplanned emergency hospital admissions, we included only well-defined disease entities with acute onset. We used the modern definition of acute coronary syndrome that also includes patients that were hospitalized with a myocardial infarction complication. Also, our definition of COPD was more elaborate, as it also included patients admitted with respiratory failure. Finally, our definition of respiratory diseases excluded conditions of the upper respiratory tract, which can be surgery-related and not influenced by air pollution.40

Problems in exposure assessment may have influenced our results. As in most time-series studies of daily variations in air pollutants, fixed monitoring stations are assumed to represent daily variation in the overall population exposure. Differential measurement error is an issue when different size fractions are compared. Vehicular traffic constitutes the main source of ultrafine particles near roads, and the variability in exposure can be much higher than in the background areas; this is in contrast to PM10 and PM2.5 mass concentrations, which have a more uniform urban distribution. A recent study45 has shown a stronger correlation between fixed monitor and indoor values for PM2.5 than for ultrafine particles. From this perspective, the comparison of the effects of various PM sizes in this study, although done using the same fixed monitor, was more favorable for PM10 and PM2.5 than for ultrafine particles. It is for this reason that the positive effects that we found between particle number concentration and heart failure and COPD are noteworthy. The limitations in the exposure assessment given the quantity of missing data also need to be considered.

In sum, our results confirm an immediate impact of PM2.5 on hospitalizations for acute ischemic events and heart failure, and a delayed effect on respiratory tract infections. Ultrafine particles showed an association with admissions for heart failure (extended lag) and COPD (immediate lag). All the effects were generally found in the elderly patients and during winter.


We thank Margaret Becker for her revision of the manuscript.


1. WHO. WHO Air Quality Guidelines. Global Update 2005. Geneva: WHO; 2006.
2. Pope CA III, Dockery DW. Health effects of fine particulate air pollution: lines that connect. J Air Waste Manag Assoc. 2006 ;56 :709–742.
3. Samet J, Krewski D. Health effects associated with exposure to ambient air pollution. J Toxicol Environ Health A. 2007 ;70 :227–242.
4. MacNee W, Donaldson K. Mechanism of lung injury caused by PM10 and ultrafine particles with special reference to COPD. Eur Respir J. 2003;21(suppl 40):47s–51s.
5. Schwartz J, Dockery DW, Neas LM. Is daily mortality associated specifically with fine particles ? J Air Waste Manag Assoc. 1996 ;46 :927–939.
6. Dominici F, Peng RD, Bell ML, et al. Fine particulate air pollution and hospital admission for cardiovascular and respiratory diseases. JAMA. 2006 ;295 :1127–1134.
7. Anderson HR, Bremner SA, Atkinson RW, Harrison RM, Walters S. Particulate matter and daily mortality and hospital admissions in the west midlands conurbation of the United Kingdom: associations with fine and coarse particles, black smoke and sulphate. Occup Environ Med. 2001 ;58 :504–510.
8. Host S, Larrieu S, Pascal L, et al. Short-term associations between fine and coarse particles and hospital admissions for cardiorespiratory diseases in six French cities. Occup Environ Med. 2008 ;65 :544–551.
9. Seaton A, MacNee W, Donaldson K, Godden D. Particulate air pollution and acute health effect. Lancet. 1995 ;345 :176–178.
10. Oberdörster G, Oberdörster E, Oberdörster J. Nanotoxicology: an emerging discipline evolving from studies of ultrafine particles. Environ Health Perspect. 2005 ;113 :823–839.
11. Delfino RJ, Sioutas C, Malik S. Potential role of ultrafine particles in associations between airborne particle mass and cardiovascular health. Environ Health Perspect. 2005 ;113 :934–946.
12. Oberdörster G. Pulmonary effects of inhaled ultrafine particles. Int Arch Occup Environ Health. 2001 ;74 :1–8.
13. Stolzel M, Breitner S, Cyrys J, et al. Daily mortality and particulate matter in different size classes in Erfurt, Germany. J Expo Sci Environ Epidemiol. 2007 ;17 :458–467.
14. Wichmann HE, Spix C, Tuch T, et al. Daily mortality and fine and ultrafine particles in Erfurt, Germany part I: role of particle number and particle mass. Res Rep Health Eff Inst. 2000 ;98 :5–86.
15. Kettunen J, Lanki T, Tiittanen P, et al. Associations of fine and ultrafine particulate air pollution with stroke mortality in an area of low air pollution levels. Stroke. 2007 ;38 :918–922.
16. Halonen JI, Lanki T, Yli-Tuomi T, Tittanen P, Kulmala M, Pekkanen J. Associations of different measures of particulate air pollution with acute cardio-respiratory hospital admissions and mortality among the elderly. Epidemiology. 2009 ;20 :143–153.
17. Andersen ZJ, Wahlin P, Raaschou-Nielsen O, Ketzel M, Scheike T, Loft S. Size distribution and total number concentration of ultrafine and accumulation mode particles and hospital admissions in children and the elderly in Copenhagen, Denmark. Occup Environ Med. 2008 ;65 :458–466.
18. Halonen JI, Lanki T, Yli-Tuomi T, Kulmala M, Tittanen P, Pekkanen J. Urban air pollution, and asthma and COPD hospital emergency room visits. Thorax. 2008 ;63 :635–641.
19. Zanobetti A, Schwartz J. Air pollution and emergency admissions in Boston, MA. J Epidemiol Community Health. 2006 ;60 :890–895.
20. Peel JL, Metzger KB, Klein M, Flanders WD, Mulholland JA, Tolbert PE. Ambient air pollution and cardiovascular emergency department visits in potentially sensitive groups. Am J Epidemiol. 2006 ;165 :625–633.
21. Forastiere F, Stafoggia M, Berti G, et al. Particulate matter and daily mortality: a case-crossover analysis of individual effect modifiers. Epidemiology. 2008 ;19 :571–580.
22. Forastiere F, Stafoggia M, Picciotto S, et al. A case-crossover analysis of out-of-hospital coronary deaths and air pollution in Rome, Italy. Am J Respir Crit Care Med. 2005 ;172 :1549–1555.
23. Comité Européen de Normalisation (CEN). Air quality—determination of the PM10 fraction of suspended particulate matter—reference method and field test procedure to demonstrate reference equivalence of measurement methods. Brussels, Belgium: CEN; 1999. Report No. EN 12341.
24. Comité Européen de Normalisation (CEN). Standard gravimetric measurement method for the determination of the PM2.5 mass fraction of suspended particulate matter. Brussels, Belgium: CEN; 2005. Report No. EN 14907.
25. Aalto P, Hämeri K, Paatero P, et al. Aerosol particle number concentration measurements in five European cities using TSI-3022 condensation particle counter over a three-year period during health effects of air pollution on susceptible sub-populations. J Air Waste Manag Assoc. 2005 ;55 :1064–1076.
26. Steadman RG. The assessment of sultriness. Part I: a temperature-humidity index based on human physiology and clothing science. J Appl Meteorol. 1979 ;18 :861–873.
27. 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.
28. Levy D, Lumley T, Sheppard L, et al. Referent selection in case-crossover analyses of acute health effects of air pollution. Epidemiology. 2001 ;12 :186–192.
29. Kunzli N, Schindler C. A call for reporting the relevant exposure term in air pollution case-crossover studies. J Epidemiol Community Health. 2005 ;59 :527–530.
30. Zeka A, Zanobetti A, Schwartz J. Individual-level modifiers of the effects of particulate matter on daily mortality. Am J Epidemiol. 2006 ;163 :849–859.
31. Pope CA III, Muhlestein JB, May HT, Renlund DG, Anderson JL, Horne BD. Ischemic heart disease events triggered by short-term exposure to fine particulate air pollution. Circulation. 2006 ;114 :2443–2448.
32. Sunyer J, Schwartz J, Tobias A, Macfarlane D, Garcia J, Anto JM. Patients with chronic obstructive pulmonary disease are at increased risk of death associated with urban particle air pollution: a case-crossover analysis. Am J Epidemiol. 2000 ;151 :50–56.
33. De Leon SF, Thurston GD, Ito K. Contribution of respiratory disease to nonrespiratory mortality associations with air pollution. Am J Respir Crit Care Med. 2003 ;167 :1117–1123.
34. Lee IM, Tsai SS, Ho CK, Chiu HF, Wu TN, Yang CY. Air pollution and hospital admissions for congestive heart failure: are there potentially sensitive groups ? Environ Res. 2008 ;108 :348–353.
35. Symmons JM, Wang L, Guallar E, et al. A case crossover study of fine particulate matter air pollution and onset of congestive heart failure symptom exacerbation leading to hospitalization. Am J Epidemiol. 2006 ;164 :421–433.
36. Wellenius GA, Yeh GY, Coull BA, Suh HH, Phillips RS, Mittleman MA. Effects of ambient air pollution on functional status in patients with chronic congestive heart failure: a repeated-measures study. Environ Health. 2007;6:
37. Lanki T, Pekkanen J, Aalto P, et al. Associations of traffic related air pollutants with hospitalization for first acute myocardial infarction: the HEAPSS study. Occup Environ Med. 2006 ;63 :844–851.
38. Von Klot S, Peters A, Aalto P, et al. Ambient air pollution is associated with increased risk of hospital cardiac readmissions of myocardial infarction survivors in five European cities. Circulation. 2005 ;112 :3073–3079.
39. Inoue K-I, Takano H, Sakurai M, et al. Pulmonary exposure to diesel exhaust particles enhances coagulatory disturbance with endothelial damage and systemic inflammation related to lung inflammation. Exp Biol Med. 2006 ;231 :1626–1632.
40. Peel JL, Tolbert PE, Klein M, et al. Ambient air pollution and respiratory emergency department visits. Epidemiology. 2005 ;16 :164–174.
41. Ko FWS, Tam W, Wong TW, et al. Temporal relationship between air pollutants and hospital admissions for chronic obstructive pulmonary disease in Hong Kong. Thorax. 2007 ;62 :779–784.
42. Penttinen P, Timonen KL, Tittanen P, Mirme A, Ruuskanen J, Pekkanen J. Ultrafine particles in urban air and respiratory health among adults asthmatics. Eur Respir J. 2001 ;17 :428–435.
43. Tittanen P, Timonen KL, Ruskanen JJ, Mirme A, Pekkanen J. Fine particulate air pollution, resuspended road dust and respiratory health among symptomatic children. Eur Respir J. 1999 ;13 :266–273.
44. Pekkanen J, Timonen KL, Ruuskanen J, Reponen A, Mirme A. Effects of ultrafine and fine particles in urban air on peak expiratory flow among children with asthmatic symptoms. Environ Res. 1997 ;74 :24–33.
45. Hoek G, Kos G, Harrison R, et al. Indoor-outdoor relationships of particle number and mass in four European cities. Atmos Environ. 2008 ;42 :156–169.

Supplemental Digital Content

© 2010 Lippincott Williams & Wilkins, Inc.