Artigos

Conteúdo (article):

Ambient temperature and hospitalizations of children due to respiratory diseases in Cuiabá, Brazil.

Luiz Fernando Costa Nascimento 1, 2, #
ORCID: 0000-0001-9793-750X
e-mail: fernando.nascimento@unesp.br
Ana Cristina Gobbo César 3
ORCID: 0000-0001-8618-8654
e-mail: anacristinagobbo@gmail.com
João Andrade de Carvalho Jr 4
ORCID: 0000-0002-5599-9611
e-mail: ja.carvalho@unesp.br
1 Postgraduate Program in Environmental Sciences. University of Taubaté, Taubaté, SP, Brazil.
2 Senior Professor. Faculty of Engineering and Sciences. UNESP / Campus Guaratinguetá, Brazil.
3 Professor. Federal Institute of Education, Science and Technology of São Paulo (IFSP), Campus Bragança Paulista, SP, Brazil.
4 Professor. Department of Chemistry and Energy. Faculty of Engineering and Sciences. UNESP / Campus Guaratinguetá, Brazil.
# corresponding author.
Address: Av Ariberto Pereira da Cunha, 333.
CEP 12516-410 Tel. (+55 12) 3123-2161.
E-mail: fernando.nascimento@unesp.br

Abstract
This study evaluated the role of temperature and fine particulate matter in hospitalizations of children living in Cuiabá, MT, obtained from DATASUS, between 01/01/2016 and 12/31/2018. Daily concentrations of the pollutant fine particulate matter were estimated using the CAMS mathematical model, made available by CPTEC. Diagnoses of tracheitis and laryngitis, pneumonia, bronchitis, bronchiolitis and asthma were included. INMET provided data on maximum and minimum temperatures and relative humidity. Statistical analysis was performed with three generalized additive Poisson regression models, one of which only included the minimum temperature, another including the pollutant and the last with an interaction variable. There were 1612 hospitalizations in the period; in the multivariate model, associations were identified between minimum temperature and hospitalizations in lags 1 to 5; the effect of increasing the minimum temperature by 4°C resulted in an increase in the risk of hospitalizations by 18%; 15.2% of hospitalizations are attributed to this increase and an excess of approximately US$ 68,000.00 in expenses for the health system during the period evaluated. In addition to the known effects of exposure to pollutants on health, it was possible to identify that an increase in the minimum daily temperature can cause damage to children\'s health.

Keywords: temperature; Children\'s health; diseases of the respiratory system; climate changes

Resumo.
Este estudo avaliou o papel da temperatura e do material particulado fino (PM2.5) nas internações de crianças residentes em Cuiabá, MT, obtidas no DATASUS, entre 01/01/2016 e 31/12/2018. As concentrações diárias do PM2.5 foram estimadas através do modelo matemático CAMS, disponibilizado pelo CPTEC. Foram incluídos diagnósticos de traqueíte e laringite, pneumonia, bronquite, bronquiolite e asma. O INMET forneceu dados sobre temperaturas máximas e mínimas e umidade relativa. A análise estatística foi realizada com três modelos aditivo generalizado de regressão de Poisson, um dos quais incluiu apenas a temperatura mínima, outro incluindo o poluente e o último com uma variável de interação. Houve 1.612 internações no período; no modelo multivariado foram identificadas associações entre temperatura mínima e internações nas defasagens 1 a 5; o efeito do aumento da temperatura mínima em 4°C resultou num aumento do risco de hospitalizações em 18%; A esse aumento são atribuídos 15,2% das internações e um excesso de aproximadamente US$ 68,000.00 em gastos com o sistema de saúde no período avaliado. Além dos conhecidos efeitos da exposição aos poluentes na saúde, foi possível identificar que o aumento da temperatura mínima diária pode causar danos à saúde das crianças.

Palavras-chave: temperatura; saúde da criança; doenças do sistema respiratório; mudanças climáticas.

Introduction

In Brazil, the average temperature has been rising 25% faster than the global average since 1910, which implies that the population may be particularly exposed to the effects of rising temperatures compared to other populations. (1) These effects may include diseases of the respiratory system (chapter X of CID 10th revision) in children. Children, in 2019, corresponded to 450 thousand hospitalizations through the Unified Health System (SUS), in Brazil, with expenses of R$ 380 million. In the state of Mato Grosso alone, there were 8,500 hospitalizations that generated an expense of R$8.5 million. (2)
Recent epidemiological analyzes of the health effects of temperature have included measures of air pollution as explanatory variables in the analysis, estimates of temperature effects that are adjusted for air pollution. (3) In some cases, data and modeling approaches initially used for air pollution epidemiological investigations were later used for ambient temperature investigations. Although adjusting for air pollution in temperature studies has become common, this practice is not always appropriate. (4) Temperature has been described as a modifier of the effects of exposure to air pollutants, especially particulate matter, with behavior that is not linear, but rather something close to a U-shape. (4-6)
A study carried out in 1814 Brazilian cities identified that for every 5°C increase in the average daily temperature during the hot seasons of 2000-2015, the estimated risk of hospitalization during the interval of 0 to 7 days increased by 4.0%, being attributed 132 hospitalizations per 100,000 residents from heat exposure. The attributable rate was higher in children <5 years old. (7)
An increase of 6°C (10°F) in temperature was associated with an increase in cases of all respiratory diseases, especially pneumonia, representing an excess risk on the same day of exposure that can reach up to 15% in children between 0 and 5 years of age and up to 5% in people over 65 years of age. (8)
The temperature variation between a day and the previous day, that is, an abrupt drop in temperature, may also be associated with adverse effects on the child\'s health, such as pneumonia, being more associated than the daily temperature range, the difference between maximum temperatures and minimal. (9) In the case of asthma in children, the daily temperature range was associated with the triggering of attacks, with increases of 5°C in this range, cumulatively over nine days, being associated with a 31% increase in asthma cases. (10)
Due to physiological, metabolic and behavioral characteristics, children are more sensitive to high and low temperatures than adults. Children under one year of age are at high risk of heat-related mortality. Temperature extremes are likely to cause more morbidity among children in relation to infectious diseases and allergic diseases. (11)
The objective of this study is to identify the role of minimum temperature and its increase in hospitalizations for some respiratory diseases in children living in Cuiabá, capital of the state of Mato Grosso, in the years 2016 to 2018, and the possible interaction of exposure to fine particulate matter.

Methods

City of study:
Cuiabá has a population of approximately 600,000 inhabitants. It is located in the center of South America, at 15° 36\'S and 56° 06\'W, with a tropical climate. The rainy season runs from October to April, the climate for the rest of the year is very dry. This municipality has a Human Development Index (HDI) of 0.785 and has 14 hospitals with around 1600 hospital beds, being approximately 1100 of which dedicated to SUS care (available in http://tabnet.datasus.gov.br/cgi/tabcgi.exe?cnes/cnv/estabmt.def). According to the National Institute of Meteorology (INMET), in the hottest month, the temperature can reach 40°C with an average of 26°C and in winter, the temperature can drop below 10°C due to the arrival of polar masses that come from the south of the continent. (12) This drop in temperature is known as “chilling”.
Pollution levels in Cuiabá are mainly determined by emissions from local industries, as well as a large number of forest fires recorded per year and the large fleet of vehicles, above 400,000, in addition to the burning of urban waste and also the transport of air pollutants from the burning of biomass and industrial emissions from neighboring cities.

Type of study:
An ecological time series study was developed with hospitalization data for children up to 10 years of age and living in Cuiabá, MT, in the period between 01/01/2016 and 12/31/2018 according to exposure to air pollutants. Hospitalization data were obtained from DATASUS and refer to diagnoses, according to the ICD-10th revision, of acute laryngitis and tracheitis (code J04), pneumonia (J12 to J18), acute bronchitis and bronchiolitis (J20 and J21) and asthma ( J45).
Daily data on estimated concentrations of fine particulate matter (PM2.5) were obtained from the CPTEC-INPE portal according to the Copernicus Atmosphere Monitoring Service (CAMS) model: CAMS-Reanalysis (2019) and CAMS-N real time (2019) from the European Center for Medium-Range Weather Forecasts (ECMWF) (13). These data were downloaded with a spatial resolution of 0.125 degrees (approximately 12.5 km) for the years 2003 to 2018 and an acquired temporal resolution of 6 hours.
Information on daily minimum values for air temperature and average values for relative air humidity were obtained from the station in Cuiabá (MT) of the National Institute of Meteorology (INMET) (12).
Tables were created with mean values and respective standard deviations, maximum and minimum values for PM2.5 concentration, minimum temperature, relative air humidity and number of hospitalizations and the interquartile difference for minimum temperature values (DIQ) was estimated. Pearson correlation coefficients were obtained between these variables.
Statistical analysis used the generalized additive model (GAM) class, with Poisson regression, assuming a significance level of 5%. The GAM model was chosen, as it is not necessary to define, a priori, the relationships and structures between the health indicator and the explanatory variables. The Poisson regression model, with a logarithmic link function, was chosen because the Poisson probability distribution is the one that comes closest to the frequency of hospitalizations, as it is a count, with an excess of zeros and an asymmetric and asymptotic distribution.
Initially, daily hospitalization values were analyzed, according to minimum temperature values, with lags of 0 to 5 days (lag 0 – lag 5) as the effects of exposures on health can occur days after these exposures, relative air humidity (RH), controlled by day of the week (short-term trend - WD) and by calendar day (long-term trend - CD) as the equation below:

Log [µ] = β + β1(Tmin) + β2(RH) + β3(WD) + β4(CD) (eq 1)

Next, a model was built including daily concentrations of PM2.5 to identify the possible role of this pollutant in the effects of minimum temperature on the risk of hospitalizations.

Log [µ] = β + β1(Tmin) + β2(PM2,5) + β3(RH) + β4(WD) + β5(CD) (eq 2)

An interaction variable was constructed between PM2.5 and the minimum temperature (Tmin * PM2.5), with PM2.5 concentrations categorized into binary values according to the median PM2.5 concentrations, with the value zero being assigned for concentrations lower than the median value and 1 for values equal to or greater than the median. (14)

Log [µ] = β + β1(Tmin) + β2(Tmin* PM2,5) + β3(RH) + β4(WD) + β5(CD) (eq 3)

Under these conditions, three Poisson regression models were constructed.
The estimated effects for exposure to the minimum temperature were compared and 95% confidence intervals were obtained according to the expression: 1 - 2 ) ± 1,96√ [SÊ1)2 + (SÊ2)2] , where 1 and 2 are the estimates of the two categories and (SÊ1)2 + (SÊ2)2 are the respective variances that are provided by the models as coefficients and standard errors, respectively.
The βi coefficients provided by the model for the concentrations at minimum temperature (Tmin) that presented a p-value <0.05 were included in the final model and transformed into a percentage increase in relative risk (RR), according to the expression RR = [exp(βi* ΔTmin) – 1]*100, where ΔTmin is the interquartile difference for the minimum temperature. The proportional attributable ratio (PAR) given by the expression PAR = 1 – (1/RR) and the population attributable fraction (PAF) given by the expression PAF = PAR*N were estimated, where N is the total of the outcome studied, in this case, hospitalizations for respiratory diseases in children. The costs of these hospitalizations for the public health system were estimated. This information is also available at the same address that provides information on hospitalizations: “hospital morbidity – by place of residence – content – average cost of hospitalization” (in Portuguese: “morbidade hospitalar – por local de residência – conteúdo – valor médio da internação”).
The analysis was carried out using the Statistica™ computer program (15).
As this is a study with secondary data made publicly available by DATASUS and with information without the possibility of identifying people, submission to the Research Ethics Committee was not necessary.

Results

There were 1612 hospitalizations in the period with a daily average of 1.47 (± 1.54), varying between zero and nine hospitalizations. Hospitalizations with diagnoses of pneumonia (1451) and bronchitis and bronchiolitis (116) predominated, corresponding to 97.2% of hospitalizations, and those with diagnoses of asthma, tracheitis and laryngitis accounted for 45 hospitalizations, 2.8% of the total. The average age was 3.59 years (±2.81) and hospitalizations predominated among males - 910 cases (56.5%).
The descriptive analysis of the study variables, with average values and respective standard deviations, maximum and minimum values are in Table 1. PM2.5 concentrations reached values well above those acceptable by the WHO; the lowest minimum temperature value was 8.6ºC, possibly caused by the phenomenon called cold weather that occurs due to the penetration of cold fronts coming from the South of the country. The daily distribution of minimum temperature, relative humidity and case values are shown in Figure 1 (A – C).
The Pearson correlation matrix (Table 2) identified non-significant correlations between hospitalizations and the variables minimum temperature and humidity, but negatively correlated with the pollutant PM2.5.
The results of the coefficients and their respective standard deviations for each lag according to the models analyzed are in Table 3.
Model 1, containing the values of minimum temperature and relative humidity, in addition to the variables days of the week and long-term trends, equation 1, identified a significant association between exposure to minimum temperatures and hospitalizations in lags 1, 3 and 4. The highest association occurred in lag 3; the RR and its 95% CI corresponding to this lag are RR = 1.033 (95% CI 1.017 – 1.051). With an increase of 4ºC (DIQ) the RR values and their 95% CI were RR = 1.141 (1.068 – 1.218). In this situation, the corresponding PAR is 12.3% and PAF = 198.9, that is, an increase in the minimum temperature by 4ºC would correspond to an excess of 198.9 hospitalizations with an excess expense of around R$ 200 thousand (approximately US$ 52,000) for the SUS, considering the average value of 1 US$ = R$ 3.85, in the three years evaluated, with the average cost being R$ 1,060 (approximately US$ 275).
In the case of the results provided by model 2 (with the inclusion of PM2.5), they point to positive associations between exposure to minimum temperatures and hospitalizations for respiratory diseases, after one, two, three, four and five days after exposure, as shown in Table 3. The highest risk was identified in lag 3 of the model containing PM2.5 concentrations, with an increase of 4ºC resulting in a RR = 1.179 (95% CI; 1.105 – 1.259).
These results indicate that 15.2% of hospitalizations can be attributed to this increase, that these 244 hospitalizations corresponded to excess expenses of around R$ 260 thousand (approximately US$ 68,000) for the SUS, in the three years evaluated, being the average cost equal to R$1,060 (approximately US$275).
With the data provided by model 3, the associations occurred at the same lags as model 2; however the greatest effect was at lag 4, with relative risk after an increase of 4ºC RR = 1,173 (95% CI 1,097 – 1,254), values slightly below those found in model 2, corresponding to a RAP = 0.148 and FAP = 237 hospitalizations, with a cost of around R$252 thousand (approximately US$65,000).

Discussion

To the authors’ knowledge, this is the first study carried out in a medium-sized city in Brazil that estimates the effect of increasing temperature, in this case the minimum temperature, on increasing the risk of hospitalization for respiratory diseases in children. It is also the first national study carried out with data estimated by mathematical modeling from the CAMS platform. The highest risks are associated with the inclusion of PM2.5 in and also with the inclusion of the variable corresponding to the PM2.5 X Minimum Temperature interaction. It was possible to identify that the inclusion of PM2.5 in the analysis increases exposure to the minimum temperature. On the other hand, the interaction factor (model 3) presented an effect close to, but lower than, that found in model 2. Non-linear polynomial lag models were also not used.
The CAMS platform replaces the CCATT-BRAMS platform previously used for this type of study. (16)
A recent study, which included more than 1800 municipalities that contain around 80% of Brazil\'s population, identified an association between heat waves and the risk of hospitalizations for nine different causes. The authors identified that children up to 10 years of age and elderly people over 70 years of age were the most susceptible groups in situations when daily temperatures were higher and lasted for a long time, with the risks for hospitalization being lower in the hotter regions. (17)
Also analyzing the effects of exposure to high temperatures, Xu et al. (18), using data from Brazilian cities, identified that less developed cities exhibited stronger associations between heat exposure and hospitalizations for all causes and certain types of hospitalizations for specific causes in Brazil. An increase of 5ºC in the average temperature in the hot season was associated with 5.1% of deaths from all causes in the most vulnerable populations, those with low income, and 2.6% when they were people with higher income. It appears that this may be exacerbated by the quality of health care existing in these cities and socioeconomic inequalities under the effects of climate change.
Higher daily temperature amplitudes were also associated with mortality in a multicenter study carried out in 445 communities in 20 countries. If there is no attenuation of these amplitudes, increases between 0.4 and 1.6ºC would result in a projected excess mortality of between 0.2 and 7.4% in some regions such as the USA, Central-Southern Europe, Mexico and South Africa. (19)
A study carried out in Taiwan concluded that the risks for all-cause mortality and outpatient visits were higher with lower temperatures, while the risk of emergency room care was higher with high temperatures. Also, certain socioeconomic factors could modify significantly mortality risks related to low temperature, including the number of people with a job, elderly people living alone in low-income families, access to public services, and numbers of medical doctors. (20)
The daily temperature amplitude was associated with the triggering of asthma in children, with the effects being more evident with the range above 10ºC, in a study carried out in Australia, with more than 13 thousand hospitalizations. Increases of 5ºC in this range were responsible for a 31% increase in emergency admissions affecting children between five and nine years of age, although the reasons for this age group to be more susceptible are still unknown, and preferably in boys who contribute around 20 % of asthma cases in Australia. The reason for this male participation may be explained by differential lung growth rates and airway size, along with immunological differences; these effects were more evident with a cumulative approach between 0 and 9 days. (21)
Although adjusting for air pollution in temperature studies has become common, this practice is not always appropriate. In studies of the health effects of temperature, the rationale for adjusting for air pollution is sometimes described as concern about potential confounding factors. The study of the present article includes an interaction between temperature and pollutant, in this case PM2.5, according to the methodology used by He et al. (12)
These studies show that there is still no consensus on the effects of temperature on children\'s health. This fact shows the difficulty of comparing the findings in Cuiabá, the effects of minimum temperature on hospitalizations for respiratory diseases in children, with other studies that specifically addressed minimum temperature.
The minimum temperature was shown to be a protective factor in a study carried out in the Republic of Cameroon involving 1306 children hospitalized with respiratory diseases and aged between two months and 18 years, with RR = 0.72 (95% CI; 0.59 – 0.87) contrary to the findings in this study, RR = 1.031 (95% CI; 1.014 – 1.049). (22)
A study carried out in Japan involving 3427 children aged up to 12 years identified that daily temperature downward changes had OR = 1.033 (95% CI; 1.005-1.063) and OR = 1.027 (95% CI; 0.995-1.060) for an increase in daily temperature, concluding that greater changes in temperature values, regardless of direction, were related to a greater chance of hospitalization for asthma, but room temperature itself was not associated with this type of hospitalization. The exact mechanism of the temperature effect is not yet known, but this result suggests that fluctuation in daily temperature has a greater impact on the condition of asthma patients than the temperature itself. (23)
The findings in Cuiabá show the effect of the association of minimum temperature on these hospitalizations. Generally, the role of temperature, and also of air relative humidity in time series, is to adjust the effects of exposure to air pollutants on the risks of hospitalizations for respiratory diseases. It is common to observe that temperature values are usually negatively associated with hospitalizations, that is, the higher the temperature, the lower the risk of hospitalization. This temperature information is not typically included in these time series studies where pollutant concentrations are the focus. (3-6)
The susceptibility of very young children to respiratory illnesses during high temperatures may be due, in part, to their underdeveloped respiratory systems and poor adaptive skills. This vulnerability is consistent with the immaturity of children\'s thermoregulatory systems. Understanding how high temperatures influence respiratory diseases is an area where further research and development are clearly needed. The burden of such diseases is expected to increase in developed countries. (24)
The excess risk of hospitalization in Cuiabá was 15.2%, due to an increase of 4ºC in the minimum temperature (interquartile difference), with 245 attributable hospitalizations, which generated an additional expense of approximately R$ 260 thousand in these three years of study (approximately US$ 68,000). The estimated financial value could, if there was no change in the minimum temperature, be directed to other health-related activities. On the other hand, the social cost of these excess hospitalizations cannot be estimated, unfortunately.

Conclusion

This study may have limitations like any ecological study, as it is not possible to connect individual exposure to pollutants and changes in temperature with hospitalization. Errors in diagnoses can lead to underestimation or overestimation of the causes; failure in the residence address, which may be in another neighboring city, but reported in Cuiabá in search of better service. Possible comorbidities such as the presence of smokers in the residence or housing conditions were also not considered as information is not available in the SUS databases, which basically have accounting purposes, but are widely used in this type of study. Patients initially treated on an outpatient basis and those served by health plans, agreements and private individuals. Another possible limitation is that the pollutants analyzed were estimated using mathematical modeling, but there is the possibility of using this type of data in cities where there is no environmental monitoring.
Despite possible limitations, it was possible to identify that the increase in minimum temperature may be associated with an increased risk of hospitalization for respiratory diseases in children in Cuiabá.

AUTHOR CONTRIBUTIONS: All authors contributed equally to the investigation, analysis, writing and approval of the published manuscript.

Acknowledgement: Luiz Fernando Costa Nascimento is grateful to Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the Productivity Fellowship, process 305656/2017-1.


References
1. National Oceanic and Atmospheric Administration (NOAA). 2017. Climate at a glance: Time series. http://www.ncdc.noaa.gov [accessed 11 Feb 2021].
2. Brasil. Ministério da Saúde. Departamento de Informática do SUS – DATASUS. Available in: http://tabnet.datasus.gov.br/cgi/tabcgi.exe?sih/cnv/nruf.def Accessed Jan 18, 2021.
3. Turner LR, Barnett AG, Connell D, Tong S. Ambient temperature and cardiorespiratory morbidity: a systematic review and meta-analysis. Epidemiology. 2012; 23: 594-606
4. O\'Neill MS, Zanobetti A, Schwartz J. Modifiers of the temperature and mortality association in seven US cities. Am J Epidemiol. 2003; 157: 1074-1082
5. Rodrigues PCO, Pinheiro SL, Junger W, Ignotti E, Hacon SS. Climatic variability and morbidity and mortality associated with particulate matter. Rev. Saúde Pública 2017; 51: 91
6. Sun S, Cao P, Chan KP, Tsang H, Wong CM, Thach TQ. Temperature as a modifier of the effects of fine particulate matter on acute mortality in Hong Kong. Environ Pollut. 2015;205:357-364
7. Zhao Q, Li S, Coelho MSZS, Saldiva PHN, Hu K, Arblaster JM et al. Geographic, Demographic, and Temporal Variations in the Association between Heat Exposure and Hospitalization in Brazil: A Nationwide Study between 2000 and 2015. Environ Health Perspect 2019; 127(1): 017001-9.
8. Green RS, Basu R, Malig B, Broadwin R, Kim JJ, Ostro B. The effect of temperature on hospital admissions in nine California counties. Int J Public Health 2010; 55:113–121
9. Xu Z, Hu W, Tong S. Temperature variability and childhood pneumonia: an ecological study. Environ Health 2014; 13: 51
10. Xu Z, Etzel RA , Su H, Huang C, Guo Y, Tong S. Impact of ambient temperature on children\'s health: A systematic review, Environ Res, 2012;117: 120-131.
11. Xu Z, Hu W, Tong S. Temperature variability and childhood pneumonia: an ecological study. Environ Health 2014; 13: 51
12. He X, Zhai S, Liu X, Liang L, Song G, Song H, et al.. Interactive short-term effects of meteorological factors and air pollution on hospital admissions for cardiovascular diseases. Environ Sci Pollut Res Int. 2022; 29(45): 68103-68117.
13. Instituto Nacional de Meteorologia – INMET. Available in: https://portal.inmet.gov.br/https://portal.inmet.gov.br/ accessed in Jan 18, 2021.
14. Centro de Previsão do Tempo e Estudo Climáticos – CPTEC. Disponível em http://queimadas.dgi.inpe.br/queimadas/sisam/v2/dados/descricao. Accessed in Jan 18, 2021.
15. STATISTICA (data analysis software system), v 7. Available in: www.statsoft.com.
16. Vieira, LCPFS, Rizol, PMSR and Nascimento, LFC. Fuzzy logic and hospital admission due to respiratory diseases using estimated values by mathematical model. Ciênc. saúde coletiva, 2019; (24).1083-1090)
17. Zhao Q, Li S, Coelho MSZS, Saldiva PHN, Hu K, Huxley RR, et al. The association between heatwaves and risk of hospitalization in Brazil: A nationwide time series study between 2000 and 2015. PLoS Med 2019; 16(2): e1002753.
18. Xu R, Zhao Q, Coelho MSZS, Saldiva PHN, Abramson MJ, Li S, et al. Socioeconomic level and associations between heat exposure and all-cause and cause-specific hospitalization in 1,814 Brazilian cities: A nationwide case-crossover study. PLoS Med 2020; 17(10): e1003369.
19. Lee W, Kim Y, Sera F, Gasparrini A, Park R, Choi HM, et al. Projections of excess mortality related to diurnal temperature range under climate change scenarios: a multi-country modelling study. Lancet Planet Health. 2020; 4: e512-e521.
20. Lin Y-K, Maharani AT, Chang F-T, Wang Y-C. Mortality and morbidity associated with ambient temperatures in Taiwan. Sci. Total Environ. 2019; 651: 210-217. https://doi.org/10.1016/j.scitotenv.2018.09.161PMID: 30227291.
21. Xu Z, Crooks JL, Davies JM. Khan AF, Hu W, Tong S . The association between ambient temperature and childhood asthma: a systematic review. Int J Biometeorol 2018; 62:471–481.
22. Tchidjou HK, Vescio F, Boros S, Guemkam G, Minka E, Lobe M, et al. Seasonal pattern of hospitalization from acute respiratory infections in Yaoundé, Cameroon. J Trop Pediatr. 2010; 56 (5):317-20.
23. Ueda K, Nitta H, Odajima H. The effects of weather, air pollutants, and Asian dust on hospitalization for asthma in Fukuoka. Environ Health Prev Med. 2010; 15(6):350-7.
24. Ahdoot S, Pacheco SE and The Council On Environmental Health Global. Climate Change and Children’s Health. Pediatrics 2015; 136 (5) e1468-e1484;