Environmental conditions and childhood asthma in Lisbon: an exploratory analysis for autumn thunderstorm

ENViroNMENTaL CoNDiTioNS aND CHiLDHooD aSTHMa iN LiSBoN. aN EXpLoraTorY aNaLYSiS For aUTUMN THUNDErSTorMS paulo canário1 Marcelo fragoSo2 carla Mora3 helena nogueira4 abstract – asthma attacks triggered by thunderstorms have been described in seve ral countries. the rises in pollen concentrations, attributed to atmospheric electrical discharges, were considered the leading cause of these outbreaks. the presence of variou s pollutants as well as certain weather conditions may contribute synergistically to the onse t of exacerbations in patients with a previous diagnosis of asthma. this study aims to investigate whether atmospheric electrical discharges exert any influence on the number of emergency admissions due to childhood asthma, in autumn, in Lisbon. the investigation will examine whether there are isolated or combined effects of meteorological variables, pollu tant and pollen concentrations that favour an increase in the number of exacerbations of asthma symptoms. for this purpose we compared the number of asthma admissions in period s with and without thunderstorms in a Lisbon paediatric emergency service. increa sing pollen concentrations attributed to thunderstorms reported in the literature were not found in Lisbon in the days analysed. there was also an absence of relationship between admissions for asthma and air pollution. associations between hospitalisations due to asthm a crisis and atmospheric variables were documented. in Lisbon, received: May 2014; accepted: september 2014. 1 PhD student, researcher at CeG - iGOt – UL, research Group ZePHyrUs – alterações Climáticas e sistemas ambientais. edifício da faculdade de Letras, alameda da Universidade, 1600-214 Lisboa, Portugal. email:[email protected] 2 assistant Professor, researcher at CeG - iGOt – UL, research Group ZePHyrUs – alterações Climáticas e sistemas ambientais. edifício da faculdade de Letras, alameda da Universidade, 1600-214 Lisboa, Portugal. email: [email protected] 3 invited assistant Professor, researcher at CeG - iGOt – UL, research Group ZePHyrUs – alterações Climáticas e sistemas ambientais. edifício da faculdade de Letras, alameda da Universidade, 1600-214 Lisboa, Portugal. email: [email protected] 4 assistant Professor, researcher at Cias – UC, research Group Cias - research Centre for anthropology and Health. faculdade de Ciências e tecnologia, Universidade de Coimbra apartado 3046 3001-401 Coimbra. email: [email protected] Finisterra, XLiX, 98, 2014, pp. 179-196 180 P. Canário, M. Fragoso, C. Mora and H. Nogueira no influence of atmos pheric electrical discharges in the variation of the number of emergency admissions due to childhood asthma was verified in the studied period (autumn). On the contrary, it was verified that a decrease in minimum temperature positively reinforces these admissions. Keywords: asthma, thunderstorms, atmospheric electrical discharges, minimum temperature, Lisbon resumo – condiçõeS aMBientaiS e aSMa infantil eM liSBoa. análiSe exploratória eM diaS de troVoada no outono. Crises de asma desencadeadas por trovoadas têm sido descritas em vários países, como consequência do aumento das concen trações de pólenes. Vários poluentes e certas condições meteorológicas podem contribuir sinergicamente para a agudização dos sintomas em indivíduos com asma. investiga-se se as descargas eléctricas atmosféricas (Dea) aumentam as admissões hospitalares à urgência infan til, no Outono, em Lisboa. analisaram-se as entradas numa urgência pediátrica em Lisboa, comparando o número de admissões por asma em períodos com e sem trovoadas. O aumento das concentrações de pólenes no ar, atribuído às trovoadas não se verificou, em Lisboa, no período analisado. Confirmou-se também a ausência de relação entre admissões por asma e poluen tes e documentaram-se associações entre aquelas e as variáveis atmosféricas. não se observou associação entre descargas eléctricas atmosféricas e admissões ao serviço de Urgên cia pediátrico por asma. no entanto, verificou-se que a diminuição da temperatura mínima provocou o aumento de admissões hospitalares por asma no período estudado. Palavras-chave: asma, trovoadas, descargas eléctricas atmosféricas, temperatura mínima, Lisboa. résumé – conditionS atMoSphériqueS et aSthMe infantile à liSBonne: une analySe préliMinaire portant Sur leS JourS orageux. On a décrit en divers pays une augmentation des crises d’asthme pendant les jours d’orage, surtout à cause d’un accroissement des concentrations en pollens, liées aux décharges électriques. Mais d’autres causes peuvent être la présence de divers polluants dans l’air ou encore certaines situations météorologiques. On a comparé le nombre de décharges électriques au nombre d’admissions en urgence infantile causées par l’asthme, pendant un automne à Lisbonne. On n’a pas observé d’augmentation de la concentration de l’air en pollens, attribuée aux orages dans la bibliographie pendant cette période, ni de rapport entre les admissions pour asthme et la concentration en polluants ou la fréquence de décharges électriques. il semble, par contre, que la baisse de la température minimale provoque une augmentation de ces admissions. Mots-clés: asthme, orage, décharges électriques atmosphériques, température minima le, Lisbonne. 181 i. intrODUCtiOn as part of the “raiDen” project, financed by fCt, the Portuguese foundation for science and technology, research was conducted on the spatial and temporal variability of thunderstorms over mainland Portugal based on the first seven years (2003-2009) of atmospheric electric discharge (aeD) records, provided by the aeD detection network setup by Instituto Português do Mar e da Atmosfera (iPMa). for more on these records please see santos et al. (2012), sousa et al. (2013) and Correia (2013). as part of the aforementioned project, Professor Henrique andrade, in collaboration with physicians of Dona Estefânia Hospital in Lisbon, initiated a study on the possible links between thunderstorms and the increase in the number of asthma attacks in children living in Lisbon. the increase in number of asthma attacks linked to thunderstorms has been repor ted around the world. the first studies were made in the beginning of the 1980s in the UK (D’amato, 2012; anderson, 2001). Other countries to report surges in the number of asthma attacks linked to the occurrence of thunderstorms are australia (Bellomo, 1992; Girgis, 2000; Marks, 2001), the Usa (Grundstein, 2008), Canada (Villeneuve, 2005) and italy (D’amato, 2005; Losappio, 2011). several mechanisms were put forward as the triggers of asthma crisis, the most widely accepted being an increase in the concentrations of airborne pollen during and immediately after the o ccurrence of thunderstorms (Marks, 2007). the increa se in pollen concentrations is due to the rupture of pollen grains through osmotic shock, which occurs during thunderstorms on the smaller sized particles that, when inhaled, reach the lower air passages thus increasing the allergic reaction (suphioglu, 1992; Vaidyanathan, 2006; taylor, 2007; Grudstein, 2009). Other factors can work in synergy to increase asthma attacks during thunderstorms, for example the increase in pollutants such as ozone, nitrogen oxides and particulate matter (PM) (fauroux, 2000; Johnston, 2006; ruffoni, 2013). alcoforado (1991) found a link between increases in the occurrence of asthma attacks and clear, calm anticyclone conditions in the city of Lisbon during spring. the records for thunderstorms in Portugal show a relatively low frequency of these events, even for the country’s centre, interior and southern regions where the occurren ce is higher, with a yearly average aeD for 2003-2009 of 15 days to 20 days with registered electrical activity. in the Lisbon area the average was of 10 days a year. autumn and more specifically the september-november trimester is the time of year with the highest occurrence of aeD in mainland Portugal (Correia, 2013). the month of september registers the highest number of incidences with an ave rage of 23.5% of the yearly total of days with thunderstorms. During this season the ave rage density of aeD are nevertheless lower than during the summer, especially in the interior regions of Portugal. as for the spatial distribution, aeD are more frequent in the centre, interior and south of Portugal. During both autumn and the spring-summer seasons, there is a well-defined daily cycle in the temporal distribution of aeD, which occur more frequently during the afternoon, with peaks between 4p.m. and 5p.m. Environmental conditions and childhood asthma in Lisbon 182 P. Canário, M. Fragoso, C. Mora and H. Nogueira ii. aiM the aim of this study is to investigate whether thunderstorms, namely, the record s of atmospheric electric discharges, exert any influence on the number of asthma admissions to the paediatric emergencies (in autumn, in Lisbon). in paralle l, the study aims at understanding which atmospheric variables are associated with the increase of paediatric asthma admissions. Lastly, it is also important to identify the environmental conditions (isolated or a combination of meteorological, pollutant and pollen concentration effects) that are consistent with asthma symptoms. iii. MetHODs anD Data 1. atmospheric Electric Discharge Data aeD are one of the most important manifestations in the development cycle of vertically developed clouds, such as Cumulonimbus or Cumulus congestus, which occur durin g unstable atmospheric conditions known as thunderstorms. in 2002 iPMa installed and has since maintained a network for the detection of aeD in mainland Portugal, made-up of four sensors that capture two types of aeD, intra-cloud and cloud-ground. in order to build a sample of thunderstorm days in autumn in the Lisbon area, aeD data from the period of 2003-2009 were made available by iPMa through the raiDen project collaboration protocol. as stated above, the information was obtai ned from the detection network implemented over mainland Portugal, through sensors that function combining two detection methods (known as Df, Direction finder s and tOa, time of arrival), which are efficient and precise at capturing the electromagnetic signal produced by the aeD. in this study, only the cloud-ground aeD were considered since the intra-cloud discharges do not have a significant impact on atmospheric conditions that might influence allergenic symptoms. in this manner, the sample for “thunderstorm days in the Lisbon area” during autumn (2003-2009), are days where aeD cloud-ground were registered within a quadrangular sector (fig. 1) with around 40 km each side, centred about the Dona Estefânia Hospital. Despite the lack of control over whethe r patients arriving at that hospital with asthma symptoms are doing so from within the defined area, the defined spatial limits (fig. 1) allow for the hypothesis that the Hospital’s admissions for asthmatic patients are related to the thunderstorms (aeD) detected within the area. the parameters considered for the aeD were the total amount of discharges within the area of study. Other parameters, such as pola rity, current intensity or multiplicity, were not considered relevant. 183 fig. 1 – aeD occurrence area considered in order to define sample days during autumn in Lisbon (2003-2009). the dots indicate the locations of eaD observed in the sample (except for the maritime area of the rectangle). Fig. 1 – Área de ocorrência de DEA considerada na definição da amostra de dias de trovoada de Outono, na área de Lisboa (2003-2009). Os pontos assinalam as localizações das DEA registadas nesta amostra, que não incluiu ocorrências na área marítima do rectângulo. the sample contained a total of 20 days with registered thunderstorms betwee n the 1st of september and the 15th of november, a period that is considered to be representative of the warmest phase in autumn. table i presents a general description of the atmospheric electric discharge during those 20 days. Out of the sample total, 45% of days with registered thunderstorms occurred in september and were characterised by the severity of the storms with a higher number of registered aeD – 433 – making up a total of 68% of all the registered aeD. Of note is a single day (9th of september 2009) with an exceptionally high number of aeD (178), an extreme event in the context of atmospheric electric discharges in the Lisbon area for the period of 2003-2009 (Carraça et al., 2012). in October and november the days with registered thunderstorms had similar numbers of daily aeD, with a daily maximum of 23 and 25 aeD. Environmental conditions and childhood asthma in Lisbon 184 P. Canário, M. Fragoso, C. Mora and H. Nogueira table i – Description of thunderstorm days of the sample, in autumn, based on the occurrence of aeD in the Lisbon area. Quadro I – Caracterização sumária dos dias de trovoada da amostra estudada, no Outono, com base na ocorrência de DEA na área de Lisboa. source: iPMa. thunderstorm days aeD counts Minimum, average and maximum daily aeD counts september 9 433 9 / 48.1 / 178 October 8 140 8 / 17.5 / 23 novemberer 3 61 11 / 20.3 / 25 Total 20 634 8 / 31.7 / 178 in addition to this sample, a control sample was created made-up of a random selection of 40 days without registered aeD, during the same time-period in autumn and for the same region. 2. Meteorological variables the meteorological data were obtained from the national Climatic Data Center (nCDC). Values of temperature, humidity, precipitation, wind, cloud cover and atmos pheric pressure for the Gago Coutinho weather station in Lisbon were used. this station (located within the Dona Estefânia Hospital’s area of influence) is not influenced by terrain or urban construction, as it is situated next to the airport, which is why it is used as a reference station representative of regional climate (andrade, 2003; alcoforado e andrade, 2006). 3. air quality When thunderstorms occur, electrical discharges can increase pollen concentrations and can also contribute towards the production of nitrogen oxides, and thus towards an increase in ozone concentrations (Dales et al., 2003; Grundstein and sarnat 2009; ruffoni et al., 2013). the increase in pollutant concentrations and the fragmentation of airborne particles, making them more easily inhaled, are phenomena that can contribute to increase the number of asthma attacks (Burch and Levetin, 2002; Moreira et al, 2008; D’amato, Cecchi and annesi-Maesano, 2012). the concentrations of nitrogen dioxide (nO2), ozone (O3), PM less than 2.5 µg/m3 and PM less than 10 µg/m3 were analysed for the 40 days of the control sample and the sample’s 20 days. Hourly data from the stations closest to the Dona Estefânia Hospital in Avenida da Liberdade, Beato, Olivais, Odivelas and Entrecampos (Qualar – Portuguese environmental agency - table ii), were analysed. air quality was determined by both the time-limits of pollutant concentration set by decree-law no. 102/2010, of the 23rd of september, and by two levels: alert levels (level 185 above which a short period of exposure represents risk to human health and imme diate measures must be adopted) and information levels (level above which a short period of exposure represents risk to human health for that part of the population that is parti cularly sensitive and above which the need for immediate public advice in needed). table ii – Brief statistics of pollutants. Quadro II – Caracterização sumária dos poluentes. source: Qualar - agência Portuguesa do ambiente. Pollutants stations average ( µg /m3) standard deviation ( µg /m3 ) amplitude ( µg /m3) exceedances (average daily hours) nitrogen dioxide nO2 entrecampos 5.8 21.1 71.5 0.2 Olivais 36.5 17 58.2 0 av Liberdade 63.3 22.2 78.6 2.4 Odivelas 40.4 20.5 72.9 0.4 Beato 32.2 16.7 57 0 Ozone O3 entrecampos 34.8 18.7 59.3 0 Olivais 44.4 20.7 63.8 0 Odivelas 52.2 20.4 67 0 Beato 45.6 19.6 64.3 0 PM < 2.5 µg/m3entrecampos 19.2 5.7 20.1 0 Olivais 13.9 4.7 15.6 0 PM < 10 µg/m3 entrecampos 42.6 11.5 39.5 5.9 Olivais 26.8 9.8 33.3 2.1 av Liberdade 48.2 13.9 48.9 10 Odivelas 33.8 9.3 33.9 2.6 PM – Particulate matter Values of PM 2.5 and PM 10 greater than 50 µg/m3 were considered above the limit for Human Health safety (Decree-law no. 102/2010); values lower than 30 µg/m3 are considered to be good air quality levels and values between 30 and 50 are considered medium air quality levels. Hourly mean nitrogen dioxide levels were fixed between 120 µg/m3 and 100 µg /m3 as the upper and lower limits, respectively. as for ozone concentrations, value s should not be above 180 µg /m3 for the information level, or above 240 ug/m3 for the alert level, also for a reference period of 1 hour. for the analysed period, ozone concentrations were below the above-referenced levels. as for nO2, there were several days where the upper limits were exceeded. this gas is a product of the oxidisation of nO, which is originated by vehicle emissions, Environmental conditions and childhood asthma in Lisbon 186 P. Canário, M. Fragoso, C. Mora and H. Nogueira making it a good indicator for the amount of traffic. the Avenida da Liberdade weather station illustrates this point as it is the station closest to an area of high traffi c intensity and was the station that registered the highest nO2 concentrations (table ii). the Entrecampos weather station registered similar values, however Olivais station values did not pass the legal limits and Beato station only exceeded these limits during a two-hour period. the latter two stations are located farthest from the main road arteries and essentially reflect the regional signal. Critical values were registered during two days in Entrecampos (during 2 hours on 12/9/2007 and 1 hour on 19/9/2007) and during a 7 day period in Avenida da Liberdade (between 1 to 13 hours, daily). the month of september 2007 registered the highest number of days above the critical values, which were also the warmest days in the sample. Only two weather stations (Entrecampos and Olivais) had data for the smaller dimension PM (<2,5 µg/m3), and never registered values above the legal limit of 50 µg/m3 (table ii). Moderate quality data were registered for 8 hours in Entre campos and during 6 hours in Olivais, with both stations registering the highest values for the same five days. PM 10 µg/m3 values were registered in 4 of the chosen stations, with Avenida da Liberdade station showing the highest values (16 days), followed by Entre campos (15 days), Olivais (10 days) and Odivelas (7 days). During the warmest days in 2007 (19/9, 20/9, 21/9 and 16/10) values were above critical levels during 24 hours in Avenida da Liberdade. 4. Emergency admission records During the 60 days analysed, clinicians consulted 14,464 children resulting in the same number of admissions in the hospital’s emergency, and 1023 were asthma admissions (table iV). from each of the patient’s admission charts information relating to time of admission, date of admission, age, sex and patient postal code were collected. 5. pollen information Pollen-concentration data were obtained from the Portuguese network of aerobiology (http://www.rpaerobiologia.com/). Data were retrieved for the analysed days and for the five days preceding and succeeding those dates. the predominant pollens were from the Parietaria and Chenopodium grain types, registering values that were, on average, higher during the control days than the days with thunderstorms. 6. Statistical analysis an initial comparison between numbers of admission in both samples (thunderstorm and control) was made using an analysis of variance ANOVA. in all analyses the total daily number of asthma cases was considered, as well as the percentage of 193 different meteorological and air quality variables in order to determine the mode l capa ble of explaining the most (the variables used were the same as those used for the discriminant analysis). the Poisson regression confirmed the relative importance of minimum (maximum) temperatures and its inverse (direct) relationship with the increa se in asthma cases (table X). furthermore, it was verified that the increase in asthma admissions was higher by winds blowing from the south and West. table X – Poisson regression - analysis of parameter estimates (see text for details). Quadro X – Regressão de Poisson – Análise de parâmetros (detalhes no texto). Parameters Coef. standard deviation Wald 95% iC Chi -square p-value Maximum temperature 0.0046 0.0306 -0.0553 0.0646 0.02 0.8795 Minimum temperature -0.086 0.0496 -0.1832 0.0113 3.00 0.0831 relative humidity 0.004 0.0064 -0.0086 0.0165 0.39 0.5328 Wind speed south wind West wind 0.0021 0.4107 -0.4875 0.0162 0.3072 0.3833 -0.0297 -0.1915 -1.2387 0.0338 1.0129 0.2638 0.02 1.79 1.62 0.8988 0.1813 0.2035 V. DisCUssiOn anD COnCLUsiOns this study found no relationship between the occurrence of thunderstorms and emergency paediatric asthma admissions in Lisbon during autumn. Despite the fact that, on average, days with thunderstorms showed a higher number of emergency asthma admissions, no statistical evidence was found to link the two phenomena. in the same manner, no statistical evidence was found linking the number of daily electric discharges with the number of asthma emergency admissions. the increase in airborne pollen concentrations that several authors have asso ciated with the occurrence of thunderstorms did not occur in Lisbon d uring the time of the study. the smallness of the sample of days with thunderstorms could have been a contributing factor. On the other hand, it is possible that air flux dynamics, described as causing an increase in the number of asthma cases, might not have occurred, or if they did, they might not have affected the risk population (taylor et al., 2004; nasser, Pulimood, 2009; D’amato et al., 2012). a utumn was chosen for the analysis as the period of the year with the highest number of aeD in Portugal, together with the fact that, as it is not the period with the highest pollen concentration (concentrations are low and stable), it would allow for an easy identification of any peaks that could more easily be associated with the occurrence of thunderstorms. it is possible, however, that the relationship might occur for months with higher pollen concentrations in Lisbon, such as the months of april and May (Caeiro et al., 2007). Environmental conditions and childhood asthma in Lisbon 194 P. Canário, M. Fragoso, C. Mora and H. Nogueira the absence of a relationship between asthma admission and the pollutants is due to the relatively low concentration values and the relatively low number of days where the risk levels were exceeded, despite the relationships previously esta blished when concentration values were below the risk levels determined by law (fauroux, 2000). the inversely proportional relationship between hospital admissions and mini mum temperatures and the directly proportional relationship between hospital admis sions and maximum temperatures was established using discriminant analysis, and confirmed by the Poisson regression, albeit with low statistically s ignificant values (with the maximum temperatures established as less significant). H owever, there was no statistical relationship between the daily temperature range for the sampled days and asthma admissions, contrary to what was established d uring spring in Lisbon (alcoforado, 1991) where days with a large range in tempe rature were associated to an increase in the number of asthma cases. the relationships here established only demonstrate the relative importance of the coldest and warmest days. the inversely proportional relationship with minimum temperatures confirms the conclusions also reached in other studies (rossi, 1993; Hervás, 2013), although other authors have shown the existence of directly proportional relationships (Hashimoto, 2004; Mireku et al., 2009) such as the one found here. in all the analyses performed, the wind speed variable is directly proportional to the number of asthma attacks (despite having a low importance). as for wind d irection, Poisson regression analysis found opposite signs for the different d irections considered. the importance of wind direction is a matter deserving a more in depth analysis, where a larger sample of consecutive days registering higher pollen counts (spring) should be considered. the relationship between the asthma attack variables and humidity was contradictory for the two analyses p erformed, which did not allow for a conclusion. as a final note: this study did not find that atmospheric electric discharges had any influence on the number of paediatric emergency asthma admissions. at the same time, no influence between aeD and the increase in pollen concentrations was found for Lisbon. these results might be due to the fact that the period of thunderstorms occurrences analysed was autumn when pollen concentrations are low, due to the limited number of sample days and due to the type of analysis performed. a drop in the minimum temperature does seem to increase the amount of asthma emergencies. aCKnOWLeDGeMents the study presented here is the result of an initiative brought about by Professor Henrique andrade who pioneered this theme in Portugal, which has interested a diversity of specialists, mainly in europe and north america. With the aim of continuing his research, the team he collaborated with at the time of his death, presents this paper as the first results of an ongoing research. the authors would like to thank the imuno-allergic services of the Dona Estefânia Hospital for all the support and 195 avai lability throughout this first phase of the research. We also thank the anonymous reviewers for all comments and suggestions that helped improve this work. the research was undertaken within the project “Lightning activity in Portugal: variability p atterns and socioeconomic impacts (raiDen)”, funded by fundação para a Ciência e a tecnologia (fCt), under the contract number PtDC/Cte-atM/101931/2008. Environmental conditions and childhood asthma in Lisbon BiBLiOGrafia alcoforado MJ (1991) influência do tempo no desencadeamento de crises de dispneia em doentes respiratórios. Finisterra - Revista Portuguesa de Geografia, XXVi (51): 105-115. alcoforado MJ, andrade H (2006) nocturnal urban heat island in Lisbon (Portugal): main features and modelling attempts. Theor Appl Climatol, 84: 151-159. anderson W, Prescott GJ, Packham s, Mullins J, Brookes M, seaton a (2001) asthma admissions and thunderstorms: a study of pollen, fungal spores, rainfall and ozone. Q J Med, 94: 429-433. andrade H (2003) Bioclima humano e temperatura do ar em Lisboa. PhD thesis. University of Lisbon, 435 p. Bellomo r, Gigliotti P, treloar a, Holmes P, suphioglu C, singh MB, Knox B (1992) two consecutive thunderstorm associated epidemic of asthma in Melbourne. Med J Australia, 156: 834-837. Brito ff, Gimeno PM, Carnés J, fernández-Caldas e, Lara P, alonso aM, García r, Guerra f (2010). Grass Pollen, aeroallergens, and Clinical symptoms in Ciudad real, spain. J Investig Allergol Clin Immunol, 20(4): 295-302. Burch M. and Levetin e (2002) effects of meteorological conditions on spore plumes. Int J Biometeorol, 46: 107-117. Caeiro e, Brandão r, Carmo s, Lopes L, almeida MM, Gaspar a, Oliveira Jf, todo-Bom a, Leitão t, nunes C (2007) rede Portuguesa de aerobiologia: resultados da monitorização do pólen atmosférico (2002-2006). Rev Port Imunoalergologia, 15(3): 235-250. Carraça MGD, fragoso M, alcoforado MJ, Lopes a, Mora C, Correia s, andrade H (2012) effects of surface features on thunderstorm activity – a preliminary study for Lisbon region, Conference Proceedings of the ICUC8-8th International Conference on Urban Climates, 6th10th august 2012, UCD, Dublin (ireland). Correia s (2013) Análise de padrões temporais e espa ciais de descargas elétricas atmosféricas em Portugal Continental. Centro de estudos Geográficos, núcleo Clima e Mudanças ambien tais, 5, 124pp. Dales re, Cakmak s, Judek s, Dann t, Coates f, Brook Jr, Burnett rt (2003) the role of fungal spores in thunderstorm asthma. Chest, 123: 745-750. D’amato G, Liccardi G, silvana JG, Viegi BG (2005) thunderstorm-associated asthma in pollinosis patients. http://www.bmj.com/rapid-response/2011/10/30/thunderstorm-associated-asthma-pollinosis-patients D’amato G, Cecchi L, annesi-Maesano i (2012) a trans-disciplinary overview of case reports of thunderstorm-related asthma outbreaks and relapse. Eur Respir Rev, 21(124): 82-87. fauroux B, sampil M, Quénel P, Lemoullec y (2000) Ozone: a trigger for Hospital Pediatri c asthma emergency room Visits. Pediatr Pulm, 30: 41-46. Girgis st, Marks GB, Downs sH, Kolbe a, Car Gn, Paton r (2000) thunderstorm-associated asthma in an inland town in southeastern australia. Who is at risk? Eur Respir J, 16: 3-8. Grundstein a, sarnat se, Klein M, shepherd M, naeher L, Mote t, tolbert P (2008) thunderstorm associated asthma in atlanta, Georgia. Thorax, 63: 659-660. Grundstein a, sarnat se (2009) Meteorological Mecha nisms explaining thunderstorm-related asthma. Geography Compass, 3/1: 45-63. doi: 10.1111/j.1749-8198.2008.00195.x Hand DJ ( 2006) Classifier technology and the illusion of progress. Stat sci, 21(1): 1-15. Hashimoto M, fukuda t, shimizu t, Watanabe s, Watanuki s, eto y, Urashima M (2004) influen ce of climate factors on emergency visit s for childhood asthma attack. Pediatr Int, 46: 48-52. Hervás D, Utrera Jf, Hervás-Masip J, Hervás Ja, García-Marcos L (2013) Can meteorological factors forecast asthma exacerbation in a 196 P. Canário, M. Fragoso, C. Mora and H. Nogueira pedia tric population? Allergol Imuno, http://dx.doi.org/10.1016/j.aller.2013.07.004 Johnston nW, sears Mr (2006) asthma exacerbations 1: epidemiology. Thorax, 61: 722-728. doi: 10.1136/thx.2005.045161 Kiotseridis H, Cilio CM, Bjermer L, tunsäter a, Jacobsson H, Dahl aslög (2013) Grass pollen allergy in children and adolescentssymptoms, health related quality of life and the value of pollen prognosis. Clin Transl Immunology, 3: 19 http://www.ctajournal.com/content/3/1/19 Losappio L, Heffler e, Contento f, Cannito C, rolla G (2011) thunderstorm-related asthma epidemic owing to Olea europea pollen sensitization. Allergy, 66: 1510-1511. Marks GB, Colquhoun Jr, Girgis st, Koski MH, Han sen P, Downs sH, Car nG (2001) thunderstorm outflows preceding epidemics of asthma during spring and summer. Thorax, 56: 468-471. Marks GB, Bush rK (2007) it’s blowing in the wind: new insights into thunderstorm-related asthma. J Allergy Clin Immunol, 120: 530-532. Mireku n, Wang y, ager J, reddy rC, Baptist aP (2009) Changes in weather and the effects on pediatric asthma exacerbations. Ann Allergy Asthma Immunol, 103: 220-4. Moreira s, santos Cs, tente H, nogueira L, ferreira f, neto a (2008) Morbilidade respiratória e exposição a partículas inaláveis na cidade de Lisboa. Acta Pediatr Port, 0873-9781/08/ 39-6/223. nasser sM, Pulimood tB (2009) allergens and thunderstorm asthma. Curr Allergy Asthm Rep, 9: 384-390. rapiejko P, stanlaewicz W, szczygielski K, Jurkiewicz D (2007) threshold pollen count necessary to evoke allergic symptoms. Otolaryngol Pol, LXi(4): 591-594 rossi OVJ, Kinnula VL, tienari J, Huhti e (1993) asso ciation of severe asthma attacks with weather, pollen, and air pollutants. Thorax, 48: 244-248. ruffoni s, Passalacqua G, ricciardolo f, furgani a, negrini aC, De amici M, Ciprandi G(2013) a 10-year survey on asthma exacerbations: relationships among emergency medicine calls, pollens, weather, and air pollu tion. Rev Fr Allergol, 53: 569-575. santos Ja, reis M, sousa J, Leite s, Correia s, Janeira M, fragoso M (2012) Cloud-to-ground lightning in Portugal: Patterns and dynamical forcing, Natural Hazards and Earth System Sciences, 12: 3, 639-649. sousa Jf, fragoso M, Mendes s, Corte-real J, santos Ja (2013) statistical-dynamical modeling of the cloud-to-ground lightning activity in Portugal, Atmos Res, 132-133: 46-64. suphioglu C, singh MB, taylor P, Bellomo r, Holmes P, Puy r, Knox rB (1992) Mechanism of grasspollen-induced asthma. Lancet, 339: 569-572. taylor P e, Haflidi J (2004) thunderstorm asthma. Curr Allergy Asthma Rep, 4: 409-413. taylor Pe, Jacobson KW, House JM, Glovsky MM (2007). Links between pollen, atopy and the asthma epidemic. Int Arch Allergy Immunol, 144: 162-170. Vaidyanathan V, Miguel aG, taylor Pe, flagan rC, Glovsky MM (2006) effects of electric fields on pollen rupture. J Allergy Clin Immunol, 117(2): s157. Villeneuve PJ, Leech J, Bourque D (2005) frequency of emergency room visits for childhood asthma in Ottawa, Canada: the role of weather. Int J Biometeorol, 50: 48-56.