If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Keys findings: High concentration of NO2, measured near patients' homes, increased the risk for emergency department visit for asthma.
•
What is known: A relationship between short-term exposure to air pollution and emergency room visits for asthma has been demonstrated.
•
What does this add: Our study confirms the excess risk of asthma emergency visits linked to air pollution close to where the patients live.
•
What is the implication: Emissions of NO2 are linked to road traffic. Preventative strategies should be implemented to reduce and vehicles emissions.
Abstract
Introduction
Particulate matter, nitrogen dioxide (NO2) and ozone are recognized as the three pollutants that most significantly affect human health. Asthma is a multifactorial disease. However, the place of residence has rarely been investigated. We compared the impact of air pollution, measured near patients' homes, on emergency department (ED) visits for asthma or trauma (controls) within the Provence-Alpes-Côte-d’Azur region.
Methods
Variables were selected using classification and regression trees on asthmatic and control population, 3–99 years, visiting ED from January 1 to December 31, 2013. Then in a nested case control study, randomization was based on the day of ED visit and on defined age groups. Pollution, meteorological, pollens and viral data measured that day were linked to the patient's ZIP code.
Results
A total of 794,884 visits were reported including 6250 for asthma and 278,192 for trauma. Factors associated with an excess risk of emergency visit for asthma included short-term exposure to NO2, female gender, high viral load and a combination of low temperature and high humidity.
Conclusion
Short-term exposures to high NO2 concentrations, as assessed close to the homes of the patients, were significantly associated with asthma-related ED visits in children and adults.
Commission Nationale de l’Informatique et des Libertés
ED
Emergency Department
EEDA
Electronic Emergency Department Abstracts
ICD
International Classification of Disease
ICD-10
International Classification of Disease, Revision 10
INSEE
National Institute of Statistics and Economic Surveys
NO2
Nitrogen dioxide
O3
Ozone
ORa
adjusted Odds Ratio
ORUPACA
Observatory of Provence-Alpes-Côte d’Azur region
PACA
Provence-Alpes-Cote-d’Azur
PM
Particulate matter
PM10
Particulate matter with a diameter less than 10 μm
PM2,5
Particulate matter with a diameter less than 2,5 μm
P
Probability values
WHO
World Health Organization
1. Introduction
Air pollution harms human health and the environment. Particulate matter (PM), nitrogen dioxide (NO2) and ground-level ozone (O3) are now generally recognized as the three pollutants that most significantly affect human health. In Europe, emissions of many air pollutants have decreased substantially in the past decade. However, a significant proportion of the population, especially in cities, still lives in areas where air quality standards set for maximum allowable pollutants are exceeded [
Air pollution — European Environment Agency. (late updated 3 June 2016). Available online: http://www.eea.europa.eu/themes/air/intro. (Late accessed 16 October 2016).
]. NO2, O3 and PM induce airway inflammation and airway hyper-responsiveness (oxidative stress, immunology response and remodeling). According to the World Health Organization (WHO), in 2012, almost 3.7 million persons prematurely died as a result of air pollution in the world [
In France, the French Institute of Public Health Surveillance reported that PM2.5 were estimated to cause more than 48,000 premature deaths per year, more than half occurring in cities larger than 100,000 inhabitants [
Impacts de l’exposition chronique aux particules fines sur la mortalité en France continentale et analyse des gains en santé de plusieurs scénarios de réduction de la pollution atmosphérique. 2016
In recent decades, several epidemiological studies reported an association between short-term exposure to air pollution and adverse health effects, showing an increase in emergency room visits or hospital admissions for asthma as a result of increased pollution levels [
Short-term effects of the particulate pollutants contained in saharan dust on the visits of children to the emergency department due to asthmatic conditions in Guadeloupe (French Archipelago of the Caribbean).
]. However, only a few epidemiological studies took into account confounding factors such as meteorological conditions, viral infections and pollen counts according to place of residence [
Short-term effects of the particulate pollutants contained in saharan dust on the visits of children to the emergency department due to asthmatic conditions in Guadeloupe (French Archipelago of the Caribbean).
Associations between ozone, PM2.5, and four pollen types on emergency department pediatric asthma events during the warm season in New Jersey: a case-crossover study.
]. Moreover, no French study has evaluated the impact of short-term pollution exposure on asthmatic disease in the past decade, despite the lowering of the WHO-recommended threshold values.
The mediterranean area, including Provence-Alpes-Côte-d’Azur (PACA; south-eastern France) region, has a dry and sunny weather that favors the generation of secondary pollutants. Its spatial heterogeneity is interesting as it includes coastal highly urbanized zones, major road traffic and high-density industrial places but also rural and mountainous ones with forest fires. Moreover, the air quality-monitoring network AIR PACA reported a total of 122 days when PM has exceeded the daily limit, while the national average is 35 days per year [
The aim of our study was to measure the impact of air pollution, assessed close to the homes of the patients, on asthma-related hospital visits to the emergency department (ED) within the PACA region in 2013, and to estimate the risks from pollution, meteorological conditions, pollen exposure and viral load.
2. Methods
2.1 Study area
The study was conducted in the PACA region of southeastern France (31,400 km2; 4,937,445 inhabitants or 7.5% of the French population: National Institute of Statistics and Economic Surveys (INSEE 2013)).
This region is an intensively urbanized area with 9 out of 10 inhabitants residing in large urban and metropolitan areas. In 2013, it was the third most populated region of France.
Large urban areas, dense road, motorway networks and industrial zones make it the most significant area in France for air pollutant emissions.
2.2 Population under study
Electronic Emergency Department Abstracts (EEDA), recommended for every patient admitted to an ED in France, are directly available from the patients' computerized medical files. EEDAs are anonymously transmitted daily to “Santé Publique France” (a public health network) and included in the French syndromic surveillance system [
]. EEDAs report the date of the emergency room visit, age of patient, ZIP code of residence and final diagnosis using a national standardized thesaurus based on International Classification of Disease (ICD) codes validated by the French Society of Emergency Medicine. Since 2008, the ED Observatory of PACA region (ORUPACA) has collected these data from EDs in the region.
From January 1 to December 31, 2013, EEDAs transmitted by EDs located in PACA were included in this study if they concerned a 3- to 99-year-old patient living in PACA.
We defined 3 age groups: 3- to 17-year-old, 18- to 49-year-old and 50- to 99-year-old. We excluded children under 3 years of age because it is more difficult to be sure about the diagnosis of asthma.
Each factor was recorded on the same day than the ED visit.
The controls were defined as patients consulting for trauma in ED. When the patient was discharge from the ED, the emergency physician had to code the final diagnosis using a national thesaurus of International Classification of Disease, Revision 10 (ICD-10) codes. Visits for an exacerbation of asthma (J45-J46) and for trauma (S00-T98) were defined according to this thesaurus.
This control group was chosen because they are a homogenous group whose attendance at the ED is less likely to be related to air pollution.
Patients' ZIP codes were used to link EEDA to the nearest (“as the crow flies”) monitoring station for pollution, pollen and meteorological data.
The study was approved by the “Commission Nationale de l’Informatique et des Libertés” (CNIL), the French data protection authority (n°1,887,366).
This study was purely observational and consent of participants was not required because the research involved no intervention or contact with the patient. Only anonymous data were used.
2.3 Exposure data
2.3.1 Air pollution data
For each ZIP code, respectively, two primary air pollutants and one secondary pollutant were considered and measured in μg/m3: daily average of particulate matter with a diameter less than 10 μm (PM10), daily 1-hour maximum of NO2, and daily 1-hour maximum of O3.
The air quality-monitoring network AIR PACA, with 80 stations throughout the PACA region, took these measurements.
A spatial pollution surface model was created from the data for these three pollutants. Data were processed via the deterministic CHIMERE chemistry-transport model over a 4-km grid. Then a communal aggregation mesh was performed, using an average weight by population residing in each mesh.
2.3.2 Meteorological data
For each ZIP code, 4 daily weather indicators were recorded: average temperature (°C); air humidity (%); average wind speed (m/s) and rainfall (mm). These data were obtained from Météo-France (the French national meteorological company) and taken from a total of 237 monitoring stations. Each ZIP code was assigned to the nearest “as the crow flies” measuring station.
2.3.3 Pollen data
The choice of pollens was based upon two criteria: they had to be the most important pollen types causing sensitization and allergic symptoms in this geographic area and they had to be present in relatively high and significant concentrations.
For each ZIP code, the main taxa (cupressaceae, birch, ash, poaceae and urticaceae) were collected from 7 stationary pollen traps in PACA by the national network for aerobiological monitoring. Each ZIP code was linked to the nearest monitoring station.
2.3.4 Virology data
For each day in 2013, viral load was defined by the ratio of the daily number of visits in each ED of PACA for viral diseases and of the daily number of total visits for each ED.
Viral diseases encoded on the EEDA used to define the viral load were: bronchiolitis (ICD-10: J21), bronchitis (ICD-10: J20), laryngitis (ICD-10: J04), viral pneumonia (ICD-10: J12), influenza (ICD-10: J11), nasopharyngitis (ICD-10: J00) and angina (ICD-10: J02). The viral load was assigned to patients in terms of ED and date of visit.
The grouping of all the codes relating to a seasonal viral pathology is not validated in the literature. However, in France, Public Health France for the epidemiological surveillance of these seasonal viral diseases use the diagnosis of Electronic Emergency Department Abstracts [
]. Syndromic surveillance is the collection, analysis and interpretation of health-related data in near real time. Emergency data using discharge diagnoses are often included in these surveillances. These methods have been applied in EDs for monitoring the three main viral epidemic diseases: influenza, bronchiolitis and gastroenteritis.
2.4 Statistical analysis
First, variables were selected using classification and regression trees (CART) on asthmatic and control population by limiting itself to 3 levels for each category (i.e. 2 splits from the “root” node, the start node of tree). CART is a non-parametric regression approach. The CART method on this population allows having more staff and therefore having more bearing in pollution, pollens and meteorological profiles. That allows taking into account combined effects of several pollutants, pollens or meteorological data, as describe by Gass et al. [
Associations between ambient air pollutant mixtures and pediatric asthma emergency department visits in three cities: a classification and regression tree approach.
Environ. Health Glob. Access Sci. Source.2015; 14: 58
]. This analysis provides 3 qualitative variables: pollutant profile, pollen profile and meteorological profile.
For each variable (pollution, meteorological and pollen data), the CART method was carried out for selected multiple categories for the purpose of evaluating the joint effects of exposure mixtures. The referent group contained the lowest concentrations for pollutants, meteorological factors or taxa for pollens. The right-hand branch of each split always corresponded to the highest concentration. By definition, terminal nodes were located at the end of the tree and so cannot have had off springs. Each terminal node was able to be viewed as a unique mixture. The determination of a cut-off is carried out by a statistical method of binarization (CART). That is, by recursively considering each values of the variable as a potential threshold. The chosen threshold will be that which will optimize a statistical criteria. This method is equivalent to that of clusters in determining the thresholds but adding an explained variable.
Using a nested case control design, a random sampling was performed to improve quality and power of analysis by comparing each asthmatic with 15 controls. Each control was randomly matched on the emergency room visit day, regardless of hospital, and in order by age groups.
The impact of pollution, weather, pollens and virus risk on asthma-related emergency room visits was assessed by conditional logistic regression. For each factor, ORs were estimated regarding their relationship to asthma attacks. Interactions between meteorological factors and pollution or pollen data were also taken into account.
The analysis design used univariate analyses to select variables (p < 0.20) via conditional logistic regressions, and to assess interactions between covariates. Multivariate analysis was performed as a stepwise procedure using conditional logistic regression, to accurately estimate the adjusted Odds Ratio (ORa). Probability values (p) of less than 0.05 were considered as significant. This design allowed identical risk exposure to be considered for this large cohort.
Statistical analyses were performed using R3.1.3 software ((C) 2015 The R Foundation for Statistical Computing, Vienna, Austria).
3. Results
Of the 55 EDs in the PACA region, 37 participated in the study and transmitted 1,048,575 EEDAs in 2013. A total of 794,884 visits concerned patients aged from 3 to 99 including 6250 (0.79% of total visits) visits for asthma and 278,192 (35% of total visits) for trauma. Fig. 1 shows the flow chart of this study.
Fig. 1Flow chart of the study population: nested case control comparing asthma and trauma ED visits in Provence-Alpes-Cote-d’Azur from 1/1/2013–12/31/2013.
Patients visiting for asthma attacks were younger than the total population (average age of 22.67 vs. 41.49 years), but the gender's repartition was the same (50% vs. 51.7% of boys). In total, 75,825 controls were matched to 5055 cases. Table 1 describes the distribution of ED visits in PACA region in terms of age and gender.
Table 1Distribution of study patients: nested case control comparing asthma and trauma ED visits in Provence-Alpes-Cote-d’Azur from 1/1/2013–12/31/2013.
Asthmatics
Controls
Total population 3- to 99-years
Number n=
6250 (0.79%)
278,192 (35%)
794,884
Number of patients in age group
3- to 17-years
3565 (2%)
80,024 (45.69%)
175,156
18- to 49-years
1772 (0.5%)
117,915 (36.12%)
326,487
50- to 99-years
913 (0.3%)
80,253 (27.37%)
293,241
Age (average ± SD) years
22.67 ± 0.8
36.57 ± 11.51
41.49 ± 15.32
Age (median) years
13
31
38
Age (IQR)
30
38
42
Gender
Male
3130 (0.76%)
157,205 (38.27%)
410,746
Female
3120 (0.81%)
120,987 (31.50%)
384,138
SD: standard deviation.
IQR: interquartile range (from 25th to 75th percentile).
For pollution data, only NO2 and O3 were selected by this method. PM10 was not found a significant variable with CART analysis except for CART analysis with different characteristics: by increasing the number of levels in trees, we found PM10 was only associated with asthma in the youngest patients (3–17 years old). For meteorological data, temperature, humidity and wind speed were selected by CART. As for pollen data, urticaceae, poaceae, ash and cupressaceae, were selected by this method. Fig. 2, Fig. 3, Fig. 4 show CART analysis, respectively, for pollutants, meteorological data and pollens.
Fig. 2Tree resulting from CART analysis illustrating the joint effects of pollutants treated as ordinal variables for asthma and trauma ED visits in Provence-Alpes-Cote-d’Azur from 1/1/2013–12/31/2013. (75,825 controls were matched to 5,055 cases)
In diagram, are stated in boldface the number of asthmatics on the number of total population (asthmatics and traumatology) aged 3- to 99-year-old in each split (expressed in decimal number 0 to 1).
Fig. 3Tree resulting from CART analysis illustrating the joint effects of meteorological factors treated as ordinal variables for asthma and trauma ED visits in Provence-Alpes-Cote-d’Azur from 1/1/2013–12/31/2013. (75,825 controls were matched to 5,055 cases)
In diagram, are stated in boldface the number of asthmatics on the number of total population (asthmatics and traumatology) aged 3- to 99-year-old in each split (expressed in decimal number 0 to 1).
Fig. 4Tree resulting from CART analysis illustrating the joint effects of pollen treated as ordinal variables for asthma and trauma ED visits in Provence-Alpes-Cote-d’Azur from 1/1/2013–12/31/2013. (75,825 controls were matched to 5,055 cases)
In diagram, are stated in boldface the number of asthmatics on the number of total population (asthmatics and traumatology) aged 3- to 99-year-old in each split (expressed in decimal number 0 to 1).
Table 2 describes the average daily concentrations in pollutants and average measurements of meteorological and pollen factors.
Table 2Descriptive statistics of pollutants (daily averages PM10 and daily 1-hour maximum of NO2 and O3), meteorological, pollens and viral variables for asthmatics cases and traumatology controls during the study period. PACA, France. 2013: n = 5055 cases versus 75,825 controls.
Variables
Pop
Mean
Min
p 25
Median
p 75
Max
IQR
NO2 (μm/m3)
T
12.47
0.002
5.67
10.36
18.39
63.67
12.72
A
13.83
0.002
6.71
12.38
20.22
57.14
13.51
Tot
13.15
0.002
6.19
11.37
19.31
60.41
13.12
O3 (μm/m3)
T
24.14
1.61
19.06
23.93
28.51
58.25
9.45
A
23.82
2.01
18.79
23.68
28.27
51.13
9.48
Tot
23.98
1.81
18.93
23.81
28.39
54.69
9.46
PM10 (μm/m3)
T
24.19
2.63
17.07
22.65
29.55
88.95
12.48
A
24.89
3.51
17.59
23.26
30.4
88.95
12.81
Tot
24.54
3.07
17.33
22.96
29.98
88.95
12.65
Rainfall (mm)
T
2.16
0
0
0
0.2
187.7
0.2
A
2.15
0
0
0
0.2
98.4
0.2
Tot
2.16
0
0
0
0.2
143.05
0.2
Humidity (%)
T
68.08
14
59
69
78
100
19
A
68.32
14
60
69
78
100
18
Tot
68.56
14
59.5
69
78
100
18.5
Temperature (°C)
T
13.9
- 13.1
8.9
13.9
19
31.5
10.1
A
14.17
−7.1
9.3
14.1
19.2
31.5
9.9
Tot
14.04
−10.1
9.1
14
19.1
31.5
10
Wind speed (m/s)
T
3.45
0
1.9
2.8
4.3
22.5
2.4
A
3.45
0
1.9
2.9
4.3
21.6
2.4
Tot
3.45
0
1.9
2.85
4.3
22.05
2.4
Cupressaceae (grains/m3)
T
37.15
0
0
2.8
15.9
897.3
15.9
A
39.14
0
0
2.1
15.5
897.3
15.5
Tot
38.15
0
0
2.45
15.7
897.3
15.7
Urticaceae (grains/m3)
T
7.92
0
0
0
7
166
7
A
7.66
0
0
0
6
166
6
Tot
7.79
0
0
0
6.5
166
6.5
Poaceae (grains/m3)
T
5.63
0
0
0
3
218
3
A
5.16
0
0
0
3
147
3
Tot
5.4
0
0
0
3
182.5
3
Ash (grains/m3)
T
5.94
0
0
0
1
487
1
A
6.2
0
0
0
1
456
1
Tot
6.07
0
0
0
1
471.5
1
Birch (grains/m3)
T
0.93
0
0
0
0
382
0
A
0.86
0
0
0
0
382
0
Tot
0.9
0
0
0
0
382
0
Viral load (%)
T
6.86
0
2.08
4.62
9.8
54.55
7.72
A
8.39
0
2.78
6.14
12.5
45.57
9.72
Tot
7.63
0
2.43
5.38
11.15
50.06
8.72
Pop: population.
→ T: traumatology patients – A: asthmatics patients – Tot: Total population (study cohort).
Compared to trauma control, women were at higher risk of asthma-related ED visits with Ora = 1.52 (95% CI: 1.43–1.61; p < 0.01).
Pollution analysis showed that when ground-level O3 concentration was ≤25.63μm/m3 and NO2 concentration >12.37μm/m3, there was a statistically significant risk for asthma-related ED visits with an ORa of 1.35 (95% CI: 1.17–1.57; p = <0.01) (Fig. 2 and Table 3).
Table 3Multivariate analysis, of a case-control nested in a cohort, of different selected variables on the emergency room visits for asthma in Provence-Alpes-Cote-d’Azur region (n = 5055 cases versus 75,825 controls).
Variables
Terminal Node
ORa
95% IC
p
Female
1.52
1.43–1.61
<0.01**
Pollution
3: referent
4
1.35
1.17–1.57
<0.01**
6
0.87
0.72–1.05
0.15
7
1.28
1.04–1.58
0.02*
Meteorology
4: referent
5
1.11
0.95–1.29
0.20
6
1.28
1.05–1.57
0.02*
7
1.28
0.83–1.99
0.26
Pollen
3: referent
4
0.84
0.66–1.07
0.16
7
0.87
0.73–1.03
0.11
8
0.92
0.51–1.65
0.78
9
0.95
0.79–1.16
0.63
Viral load
1.04
1.04–1.05
<0.01**
Terminal nodes are located at the end of the trees about pollution, pollen and meteorological data. Each terminal node can be viewed as a unique mixture.
The referent group contains the lowest concentrations for pollutants, meteorological factors or taxa for pollens.
Similarly, when ground-level O3 was higher (>25.63μm/m3) and concentration NO2 remained beyond the limit (>12.37μm/m3), there was still a risk for asthma-related ED visits with Ora = 1.28 (95% CI: 1.04–1.58; p = 0.02) (Fig. 2 and Table 3).
These results showed the NO2 effect to be independent of ground-level ozone concentration.
Meteorological analysis revealed that low temperatures (<21.8 °C) and high humidity (>83%) were statistically linked to asthma-related ED visits, with Ora = 1.28 (95% CI: 1.05–1.57; p = 0.02) (Fig. 3 and Table 3).
The association between viral load and asthma-attacks was also statistically significant with ORa = 1.04 (95% CI: 1.04–1.05; p < 0.01) (Table 3).
No association between the risk for asthma-related ED visits and pollen was found. Asthma-attacks were not found to increase with meteorological factors and pollution or pollen data either (Table 3).
4. Discussion
Our study shows a significant association between short-term exposure to high concentration of NO2 measured near patients' home, independent of ground-level ozone, and asthma-related ED visits in 3- to 99-year-old patients in the PACA region in 2013. The study, which took into account viral data, pollens and meteorological conditions, also identifies other risk factors: the female gender, a combination of low temperature and high humidity, and a high viral load.
Our study is original because we evaluated patients' exposure to atmospheric pollution based on their home ZIP code, rather than on the receiving hospital, and because we used the CART method. Our assessment is more accurate because, according to INSEE, the average commuting distance between home and work in Provence-Alpes-Côte-d'Azur is 10.6 km [
]. Few previous studies have taken into account the patients' place of residence. Shmool and al. reported that, after incorporating residential-level spatial exposure, an excess risk per 10 ppb ozone exposure was significant on lag days 1 through 5, ranging from 7.3 to 9.4% for asthma outpatient in emergency departments in New York City [
Associations between ambient air pollutant mixtures and pediatric asthma emergency department visits in three cities: a classification and regression tree approach.
Environ. Health Glob. Access Sci. Source.2015; 14: 58
]. Of course, asthma diagnosis may be overrepresented because the ICD-10 only permits to code asthma attack regardless of the age of the patient and the number of previous similar episodes. So, in pre-school children, some wheezy episodes may have been wrongly coded as asthma attacks. Similarly, older patients coming to ED for a first wheezy attack may also have been coded as asthma attack. An advantage of our study is to include, for the first time, a specific control group in order to limit the effect of pollution on control case. In our study, we chose patients with trauma because we believe they are less likely to be affected by pollution than other medical conditions. Control group selection bias due to diagnostic coding is more likely. Hence OR in asthmatics are higher in our study than in the literature.
Our study has some limitations because we have not used continuous measurement of personal exposure to air pollutants, which would be considered as the gold standard for exposure assessment [
]. It would be very expensive to measure personal exposure in a large number of participants necessary to satisfy statistical power requirements. Our estimates of pollution exposure are thus based on a series of measurement of pollution from stationary positions. While we cannot rule out coding errors on the ED software, asthma and trauma have easily recognizable clinical presentation, so the potential for errors should be fairly low in this cohort.
Our finding of an association between asthma attacks and elevated concentrations of NO2 is consistent with the literature. Indeed, NO2 is the principal pollutant reported to be associated with hospital admissions or consultation for asthma. Outdoors air pollution (NO2, but also sulfur dioxide SO2) was associated with increased risk of asthma hospitalization in Shanghai [
]. Another study found a significant association between short-term increase in NO2 (for 6-day cumulative mean pollutant concentration) and timing of asthma diagnosis, solely for May to October. This study reported that a 10-parts per billion increase in NO2 was significantly associated with an increased rate of asthma diagnosis from 2.7% to 7% [
Association of short-term increases in ambient air pollution and timing of initial asthma diagnosis among medicaid-enrolled children in a metropolitan area.
]. A systematic review has reported an association between ambient exposure to NO2 and a higher prevalence of diagnosis of childhood asthma or increased asthma risk at 3 years [
]. Simons et al. demonstrated that an NO2 reduction of 10% could potentially decrease respiratory drugs use with an associated direct cost saving of 107.845€ (95%IC: €71.483- €143.823) [
We did not find any statistically significant correlation between PM10 level and asthma-related emergency room visits. PM10 has been reported to be a risk factor for asthma attacks but with an interquartile range for PM10 concentration generally higher than that observed in out study [
]. We used a CART methodology to select pollutant in order to take into account their combined effect. To our knowledge, no previous study has evaluated the impact of PM10 in asthma with this method. It was only used for PM10 forecast [
]. Using the CART method, PM10 was not selected. Upon increasing the number of level in trees, PM10 was only associated with asthma in the youngest patients (3–17 years old), but not in older patients. We have not analyzed the pollution according to each age group as patients were matched according to age groups. It is possible that other confounding factors could explain the absence of any effect of PM10 e.g. environmental cofactors, or behavioral and socio-economic factors.
We found a link between asthma attacks and viral load, consistent with previous reports. We chose to focus on viral diseases globally rather than on each individual virus. Bonnelyke et al. showed a relationship between asthma attacks and number of respiratory infections, but not the virus itself in their study of 313 children [
Regarding meteorological factors, we found that a combination of low temperature and high humidity was a risk factor for asthma-related emergency room visits. Zhang reported that a daily mean temperature below the median of 18.7 °C, was linked with an increase in asthma hospital admissions [
]. In Shanghai, outdoor air pollution was associated with an increased risk of asthma hospitalization, particularly in the cold season (from October to March) [
Our study differs from the literature because we found no link between pollen exposure and asthma-related emergency room visits. This may be because we only had 7 pollen-monitoring stations distributed throughout the region. This meant that exposure levels for cases and controls were very close. It is possible that the low number of stations adversely affected the multivariate analysis due to the coarse nature of the pollen exposure estimates. In other studies, emergency calls for asthma exacerbation were significantly correlated with pollen exposure, but only between April and August [
]. However, associations have also been reported between mid-spring pollen types (maple, birch, beech, ash, sycamore/London plane tree and oak) and both over-the-counter allergy medication sales and ED visits for asthma in children aged (5–17) [
The associations between daily spring pollen counts, over-the-counter allergy medication sales, and asthma syndrome emergency department visits in New York City, 2002-2012.
Environ. Health Glob. Access Sci. Source.2015; 14: 71
Our study did not evaluate the impact of socio-economic factor on asthma-related visits and air pollution. A recent systematic review shed light on the role of socioeconomic status as an effect-modifier of the association between air pollutants (mainly O3, followed by NO2 and PM10) and important asthma-related outcomes, with a strong negative influence on children (3–6 and 12–18 years) living in low socio-economic conditions [
The role of socioeconomic position as an effect-modifier of the association between outdoor air pollution and children's asthma exacerbations: an equity-focused systematic review.
It would be interesting to set up an alert system for patients with asthma, using a lower threshold than the current WHO recommendations. A recent quasi-experimental study reported a doubling of emergency admissions for all relevant conditions (asthma, chronic obstructive pulmonary disease or coronary heart disease) and a fourfold increase in admissions for respiratory conditions after implementation for 2 years an alert system automatically triggered by high pollution levels [
Effects of an air pollution personal alert system on health service usage in a high-risk general population: a quasi-experimental study using linked data.
J. Epidemiol. Community Health.2016; 70: 1184-1190
]. We might envisage a personalized alert system for each asthmatic patient to trigger an action (for example: to avoid strenuous physical activity, individual adaptation of treatment …) when daily 1-hour maximum of NO2 exceeds 12 μm/m3, threshold found in CART method, in the PACA region.
5. Conclusion
Our study shows that short-term local exposure to high concentration of NO2, as assessed close to the home of the patients, is associated with an increased risk of emergency department visits for asthma exacerbation in patients aged 3–99-years in the PACA region. These findings require further confirmation from other studies, using larger patients' cohorts. Emissions of NO2 is primarily linked to road traffic. If our data are confirmed, preventative strategies should be implemented to reduce traffic pollutions and vehicles emissions.
Funding sources
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Financial disclosure
The authors have no financial relationships relevant to this article to disclose.
Conflicts of interest
None.
Clinical trial registration
“Commission Nationale de l’Informatique et des Libertés” (CNIL) (The French data protection authority): n° 1887366.
Contributor's statement
Julie Mazenq: She conceptualized and designed the study, coordinated and supervised data collection, and drafted the initial manuscript.
Jean-Christophe Dubus: Pr Dubus conceptualized and designed the study, reviewed and revised the manuscript.
Jean Gaudart: Dr Gaudart carried out the statistics analyses, critically reviewed the manuscript.
Denis Charpin: Pr Charpin designed the pollen data collection instruments and critically reviewed the manuscript.
Gilles Viudes: Dr Viudes is the promoter of the study. He conceptualized and designed the study, critically reviewed the manuscript.
Guilhem Noel: Dr Noel conceptualized and designed the study, coordinated and supervised data collection, reviewed and revised the manuscript.
All the authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Acknowledgments
We thank Mr. Robin Dominique and his staff at the air quality-monitoring network (AirPACA) for collecting the pollution data.
We thank Mrs Boiron who has reviewed the English version of the manuscript.
References
Air pollution — European Environment Agency. (late updated 3 June 2016). Available online: http://www.eea.europa.eu/themes/air/intro. (Late accessed 16 October 2016).
Impacts de l’exposition chronique aux particules fines sur la mortalité en France continentale et analyse des gains en santé de plusieurs scénarios de réduction de la pollution atmosphérique. 2016 (Environnement et santé/Rapports et synthèses/Publications et outils/Accueil. Available online:)
Quelle est la part des pics de pollution dans les effets à court terme de la pollution de l’air sur la santé dans les villes de France?2016 (Environnement et santé/Rapports et synthèses/Publications et outils/Accueil. Available online:)
Short-term effects of the particulate pollutants contained in saharan dust on the visits of children to the emergency department due to asthmatic conditions in Guadeloupe (French Archipelago of the Caribbean).
Associations between ozone, PM2.5, and four pollen types on emergency department pediatric asthma events during the warm season in New Jersey: a case-crossover study.
Associations between ambient air pollutant mixtures and pediatric asthma emergency department visits in three cities: a classification and regression tree approach.
Environ. Health Glob. Access Sci. Source.2015; 14: 58
Association of short-term increases in ambient air pollution and timing of initial asthma diagnosis among medicaid-enrolled children in a metropolitan area.
The associations between daily spring pollen counts, over-the-counter allergy medication sales, and asthma syndrome emergency department visits in New York City, 2002-2012.
Environ. Health Glob. Access Sci. Source.2015; 14: 71
The role of socioeconomic position as an effect-modifier of the association between outdoor air pollution and children's asthma exacerbations: an equity-focused systematic review.
Effects of an air pollution personal alert system on health service usage in a high-risk general population: a quasi-experimental study using linked data.
J. Epidemiol. Community Health.2016; 70: 1184-1190
30316-5&id=gr1.jpg){kind=link}
30316-5&id=gr2.jpg){kind=link}
30316-5&id=gr3.jpg){kind=link}
30316-5&id=gr4.jpg){kind=link}