Exposure to Ambient Air Pollutions and Upper Respiratory Tract Infection in Zhengzhou City, China: A Approximate Five-Year Surveillance Study

Background: Studies indicated that air pollutions were associated with respiratory disease have with a lag exposure–response relationship, but not linear. However, few evidences in Zhengzhou, one of the most polluted cities for China. Method: Upper respiratory tract infection (URTI) outpatient visits in the hospital, meteorological parameters and air pollutions data were obtained from October 28, 2013 to May 1, 2018 and were used for evaluating the risk effects of the air pollutants with a distributed lag non-linear model (DLNM), including the stratied analysis of gender and age. Result: 475013 cases were included, with obvious seasonal uctuations,higher in cool/cold and lower in warm. Every increase of 10μg/m 3 of PM 2.5 , PM 10 , SO 2 , NO 2 and CO showed similar impacts on URTI outpatient visits in different genders and age sub-groups,within 0 to15 days of lag. PM 10 , SO 2 and NO 2 had the strongest immediately risk at lag 0 [RR PM10 : 1.0011, 95%CI (1.0002-1.0020); RR SO2 : 1.0084, 95%CI (1.0039-1.0130); RR NO2 : 1.0149, 95%CI: (1.0111-1.0188), respectively], while PM 2.5 and CO got highest risk at lag 15 days [RR PM2.5 : 1.0014, 95%CI (1.0003-1.0025); RR CO : 1.0002, 95%CI: (1.0001-1.0003), respectively]. In addition, calculating overall accumulated effects of each 10μg/m3 increase in PM 10 , SO 2 , NO 2 , and CO was greater in females than in males, as well as greater in the adolescents (aged 0-18 years) and elderly (aged ≥ 60 years) than in adults (aged 19-59 years), except CO was greater in the adolescents and adults than in the elderly. No signicant cumulative effects were found in PM 2.5 . O 3 levelwasno signicant correlation withURTI outpatient visits throughout the lag period. Conclusions: Our results indicated that PM 10 , SO 2 , NO 2 and CO had strong immediate and lag cumulative effects in the females, adolescents, and elderly. PM 2.5 has lag effects but has no signicant lag cumulative impact effects on gender and age. Spearman rank correlation analysis was performed for the detections of the relationships between air pollutions, URTI outpatients and meteorological conditions. DNA damage, and another mechanism causing and aggravating pulmonary disease 43 . Acute and long-term exposure to high-level CO in enclosed spaces can cause serious health hazards to humans, including death, while number of studies found that associations between low-level CO and benecial in pulmonary health 44,45 . of pollutants on health complex and vary with different ages, and risk factors. outpatient visits in different age groups. For PM 10 , SO 2 and NO 2 , the risk effects in the adolescents (aged ≤ 18 years old) (aged 19–59 years old), of the adults in the air than adults, CO Li short-term A rstly,


Introduction
The World Health Organization considered health was threatened by air pollution, which was the largest environment risk, with 90% of people breathing in contaminated air every day and residents of developing countries experienced severer air pollutant problems 1 . The 2018 Environmental Performance Index report indicated China's air quality was poor, ranked 3rd from the bottom 2 . The air pollutants may mainly come from industrial pollution sources, fuel combustion and ground dust 3 , which are dynamic and complex mixtures of gaseous pollutants (NOX, SO 2 , O 3 and CO) and particulate matters (PM 10 and PM 2.5 ) 4 . Numerous researches showed that air pollutions could penetrate into the system of respiratory and circulatory, exacerbating pulmonary disease, damaging extrapulmonary organs 5 . And it also caused more harmful to vulnerable populations around the world, such as females, children, elderly, and people with low socioeconomic status 6 .
URTI is a common respiratory disease, and located in the upper respiratory tract 7 , which is the gateway to the lung and is more easily to be affected by air pollutants. A study in Hong Kong showed that URTI accounted for 26.4% of outpatients 8 . Another study from Turin, a city in northwestern Italy, showed that in the emergency hospital outpatient clinics, teenagers who came to the clinic due to respiratory infections account for 81% of adolescents with respiratory diseases 9 . Numerous reaches have revealed that long-term or short-term exposed to air pollutions (PM 2.5 , PM 10 , SO 2 , NO 2 , CO and O 3 ) related to respiratory diseases 10,11 . However, the researches on the association between air pollutions and clinic patients with URTI are relatively rare in China. For all we know, in China, only Beijing 12 , Shenyang 13 , Hefei 14 , Guangdong 15 , Hongkong 16 , Taiwan 17 , Wuhan 18 have carried on study of the relationship between air pollutions and URTI, while in Henan Province has not been studied yet. Air pollutions and demographic characteristics are different in each region.
It is necessary to research local particular conditions to provide ideas for local government to prevent and control pollutants, as well as clinical treatments.

Research area
Zhengzhou, the capital of Henan Province, is located in the southern part of the North China Plain (112°42'-114°13' E, 34°16'-34°58' N). It belongs to the northern temperate continental monsoon climate, which is characterized by low rainfall, low temperatures, low humidity, and high frequencies of calm wind in autumn and winter, most of the time it belongs to static and stable weather. Zhengzhou has a special geographical location, bordering the Yellow River to the north, Mount Song to the west, and the vast Huanghuai Plain to the southeast. Moreover, the dominant wind in winter is the northwest wind, which is blocked by the surrounding Taihang Mountain and Songshan Mountain, resulting pollutants do not disperse easily. In addition, as of December 2017, the number of motor vehicles in Zhengzhou has reached 3.4711 million, there is serious pollution from automobile exhaust. Zhengzhou's central heating season, which runs from mid-November to mid-March, is dominated by coal-red heating, which is a major cause of the city's heavy pollution during winter and spring. These factors combine to result a range of health-damaging pollutants accumulate and deposit over the urban area, making Zhengzhou become one of the most worst air polluted area for China 19 The in uence of meteorological factors were taken into consideration, and those data were obtained from the website: Weather Underground 22 . In this study, daily mean temperature, relative humidity, and wind speed were used.

Analyzed statistically
Using the SPSS 22.0 software to statistically describe daily clinic visits for URTI, meteorological factors and air pollution levels, and Spearman rank correlation analysis was performed for the detections of the relationships between air pollutions, URTI outpatients and meteorological conditions.
Using Poisson distribution to analyze statistically the daily outpatient visits for URTI. Meanwhile, the relationship of air pollutions and incidence of a respiratory system diseases was non-linear, and respiratory illness changed with lagged effects of air pollutants levels 23 .
Therefore, adopting the quasi-Poisson regression with a distributed lag non-linear model (DLNM) to research the effect of air pollutants on URTI outpatients. The impact of secular trends, seasonality and weekends were controlled, and the relationship between air pollutions and URTI outpatients were matched. Take the daily mean temperature as the meteorological indicator and relative humidity was controlled. The two dimensions of predictor and lags were conducted to study air pollutants on URTI outpatient visits. Take the average of each index as the reference level, and 15d is the maximum lag period. The reference model was used in the calculations as follows 24 : In the model, E(Y i ), the number of URTI outpatients on day i; NS(time t ), the natural cubic spline of the time (date) t, df (degree of freedom) = 7/year; hum i , the mean relative humidity on day i, df = 3; wind i , the wind speed at day i, df = 3; w j , the regression coe cient of variable; strata jt , the dummy variable of the week and month (such as public holidays and weekends) effects; l, days of lag, with a maximum lag of 15 days; βT t , l1 , the temperature matrix with a lag of l-day obtained by DLNM with a coe cient of β; γPollution t , l2 , the Pollution (PM 2.5 , PM 10 , SO 2 , NO 2 , CO, and O 3 ) matrix with a lag of l-day obtained by DLNM with a coe cient of γ [25][26][27] . ξ and α represent the residual and the intercept, respectively.
In addition, the relative risk (RR) of the lag effects and the cumulative effects of 10 µg/m 3 increment of each pollutant on the URTI outpatients were calculated with the R 3.6.1 software. All data analyses were carried out using R3.6.1 software. Using the "dlnm" package to t DLNM.

Description of the general condition
Summarizing the distribution of daily URTI outpatients, concentrations of air pollutants and factors of meteorology in Table 1. A total of 475013 cases of URTI outpatients was obtained, with a female-to male-ratio of 1:1.13. The distribution of daily URTI outpatients uxed with mean air pollutant concentrations during the period were presented in Fig. 1, with obvious seasonal uctuations, and were higher in cool/cold season and lower in warm season relatively. Concentration of O 3 level showed opposite trend of uctuations in cool/cold and warm seasons.
As shown in Table 2, daily URTI outpatient visits were positively related to PM 2.5 , PM 10 , SO 2 , NO 2 and CO concentrations, especially NO 2 (r = 0.362, P < 0.01), while were negatively related to temperature, relative humidity, wind speed and O 3 . Temperature, relative humidity, wind speed and each air pollutants were signi cantly correlated, especially the correlation of PM 10 and PM 2.5 (r = 0.872, P < 0.01).  Summarizing the cumulative relative risk impacts of different air pollutants on different subgroups of outpatients with URTI within the lag period in Table S1. The in uencing effects of PM 10 , SO 2 and NO 2 was greater in female than male, and greater in the adolescents and elderly than in adults. For PM 10 , the accumulated effects of lag 0 to lag 0-12 and lag 0-14 to lag 0-15 days in females, and lag 0 to lag 0-12 days in male subgroups showed statistical signi cance, reached the peak value on lag 0-15 and lag 0-1 days, respectively (1.0150-1.0213), respectively)]. The RR-values were higher in Zhengzhou than those reported in Shenyang, but lower than Beijing. Moreover, study indicated that different pollutants in different areas had different strongest-effect-lag-time for the same disease and their corresponding RR-values, which might be considered to the human susceptibility and regional pollutants with signi cant regional characteristics 31 . More studies have been done in the north China than in the south, due to the worse air quality in the north China 32 , which was contributed by heavy industrial emission and heating season coal consuming, and also related to the local geographical environment, economic development, energy structure, air pollutions control and weather conditions 33 .
In the past few years, the levels of particulate matters (PM 10 , PM 2.5 ) in Zhengzhou were exceeding the national secondary maximum contaminant level, and NO 2 was 1.3 times higher than the national standard, while SO 2 , CO, O 3 approached the national secondary maximum contaminant level. Coal burning contributed signi cantly to PM 2.5 , PM 10 and SO 2 pollution, especially in winter 34 . NO 2 is mainly emitted into the air by burning fossil fuels, especially vehicle fuels. O 3 is formed by the interaction of nitrogen oxides (NOx) and volatile organic compounds (VOCs) with sunlight and it is also high positive correlation with vehicles exhaust emissions 35 . Using gasoline or diesel fuel cars and using carbon compounds in industrial processes are the major two responsibilities for CO being discharged to the atmosphere 36,37 .
PM 10 and PM 2.5 are mainly deposited on the trachea, especially PM 2.5 has a huge surface area and adsorbs various toxic and harmful substances, and can enter into the terminal bronchioles and alveoli 38 , leading to a series of lung injuries including destruction of the airway epithelial barrier, interfered cellular signaling pathways, destroyed lung parenchyma, cell immunity, epigenetic modi cations and autophagy 39 . Reports demonstrated that exposed to high-level SO 2 had irritative effects on the smooth muscle of the respiratory tract, causing bronchoconstriction and in ammation of the upper airway, increase airway resistance, and decrease lung function 40,41 . NO 2 is also regarded as a kind of respiratory tract irritation. Animal and human exposure studies have found that inhalation of NO 2 mainly invaded in the distal bronchioalveolar and the alveoli, causing respiratory tract infections and promoting lung in ammation 42 . O 3 , as a strong corrosive substance, can cause adverse effects on the respiratory system through in ammation, oxidative stress, airway hyperresponsiveness, DNA damage, and another mechanism causing and aggravating pulmonary disease 43 . Acute and long-term exposure to high-level CO in enclosed spaces can cause serious health hazards to humans, including death, while number of studies found that associations between low-level CO and bene cial in pulmonary health 44,45 .
The in uences of air pollutants on health are complex and vary with different genders, ages, and other risk factors.
From the gender perspective, most studies have discovered that females were more vulnerable to pollutants than males. In this study, We also found that calculating the overall cumulative risks of PM10, SO2, and NO2 concentrations increasing by 10µg/m3, lagging 15 days, females were signi cantly higher than males. It indicated that female was more susceptible and threatened potentially by pollutants. However, some studies suggested contrary view, in which males were more susceptible to particulate matter (PM 2.5 , PM 10 ) exposures than females 29,46 . There were also studies that showed that in Arak city of Iran, PM 2.5 , PM 10 , SO 2 , NO 2 and CO had signi cantly higher in uence on the hospitalization of males with respiratory diseases than that of females, and there was no gender difference in the in uence on the hospitalization of circulatory disease 47,48 . The differences in lung growth rates and function decrease between men and women may in uence the incidence of respiratory in ammation 49,50 . In addition, particulate matter was deposited in women's lungs more than in men, which making them more susceptible to respiratory disease 42,51 . Complexity was re ected in the actual exposure environment, occupation characteristics, smoking behavior, education level and ethnicity of different gender population, which impacted the health effects of air pollutant exposure 52 . Such as both active and passive smoking can lead to a decline in lung function, especially in small airway, moreover, the decline in lung function of active smoking people was more serious than passive smoking 53 . Therefore, the impact of air pollution on gender differences still needs to be further studied.
From the perspective of age, adolescents and elderly groups were more vulnerable to air pollutants than in adults aged 19-59 years old.
In this study, the total cumulative effect of each pollutant increased by 10 µg/m 3 was calculated, the incidence of URTI outpatient visits were different in different age groups. For PM 10 , SO 2 and NO 2 , the risk effects were greater in the adolescents (aged ≤ 18 years old) than in adults (aged 19-59 years old), while for the effect of CO was greater in the adolescents and adults than in the elderly. It indicated that elderly and adolescents were more susceptible to air pollutants than adults, except CO exposure. These results were consistent to previous studies. Li reported that exposed by air pollution in short-term was related to increased risk of URTI for aged 0-14 years old in Hefei 14 . A study from Beijing indicated that the elder more than 65 years old was the most sensitive to air pollution, followed by adolescents younger than 14 years old 54 . There are two reasons for the above phenomenon: rstly, compared with adults, adolescents' lungs and immune system are not mature, which were susceptible to air pollutions and adolescents tend to spend more time doing intense outdoors activities, so they may breathe a higher amount of outdoor air pollutants 55,56 ; secondly, for the elderly, both physical function and physical activity had been on a downward trend, showing a decline in the ability of the immune system to ght against exogenous hazards, as well as the cumulative effect caused by long-term exposed from air pollutants, making the elderly more susceptible to air pollutants 57 . It also implied that the health effects of air pollutant exposure in different age groups were related to the actual exposure level, the sensitivity of pollutant exposure (long-term exposure to pollutants reduces sensitivity to pollutants, people in clean air environment were more sensitive), and the physiological characteristics of aging 58,59 .
However, there were several limitations in this study: 1) due to the interaction of the pollutants, there was bias to evaluate the risk effects of each single pollutant; 2) the risk effects of indoor pollutants were not controlled; 3) the impacts of individual living habits and occupations on URTI were not considered.

Conclusions
Our results indicated that PM 10 , SO 2 , NO 2 and CO have strong immediate and lag cumulative effects in the females, adolescents, and elderly. PM 2.5 has lag effects but has no signi cant lag cumulative impact effects on gender and age.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
We declare no competing interests.  All the authors read the paper, offered interpretation, and provided comments on the nal the draft of manuscript.