The Effect of Coronavirus 2019 Disease Control Measures on the Incidence of Respiratory Infectious Disease and Air Pollutant Concentrations in Yangtze River Delta Region, China

Lan Wang (  wanglan@zju.edu.cn ) Zhejiang University School of Medicine First A liated Hospital https://orcid.org/0000-0003-1369-0584 Kehan Wang Renmin University of China School of Statistics Hui Zhong Sun Yat-sen University School of Data and Computer Science Na Zhao Chinese Academy of Sciences Yunmei Yang Zhejiang University School of Medicine First A liated Hospital Yiran He Renmin University of China School of Statistics Shelan Liu Zhejiang Provincial CDC: Zhejiang Provincial Center for Disease Control and Prevention https://orcid.org/0000-0003-2053-0941 Wangli Xu Renmin University of China School of Statistics


Introduction
Coronavirus disease 2019 (COVID-19) was rst identi ed in Wuhan, China, in December 2019, and, since then, the number of COVID-19 cases has rapidly surged worldwide [1]. It was declared a pandemic by the World Health Organization (WHO) on 11 March 2020 [2]. By 7 March 2020, 116,166,652 con rmed cases and 2,582,528 deaths had been reported in at least 200 countries, areas or territories [3].
In China, the emergence of a Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) was rst con rmed in Wuhan city, SARSCoV-2 spread rapidly to many other cities of Hubei province and outside of Hubei provinces rapidly during January 2020 [4][5]. In response the threat, the Chinese government implemented numerous strict measures to control the SARS-CoV-2 transmission during the initial emergency response to COVID-2019 outbreak stage between January and March 2020 [6]. China controlled its initial COVID-19 epidemic in March 2020 and since April 29 of 2020, China have entered the routine control stages of suppression COVID-19 [7]. Although dozens of small COVID-19 outbreaks occurred in Shijiazhuang of Hebei Province, Beijing, Jilin Province, and the Xinjiang Uygur Autonomous Region since June of 2020 [8-9], they were controlled by unremitting containment and suppression [10].
However, the effect of the measures taken to contain COVID-19 on the rate of respiratory infectious disease (other than  in China remains unclear. Of the various areas affected by COVID-19 in China, the Yangtze River Delta region (YRD) is one of the top ve regions, with 1,835 and 1,322 con rmed cases having been respectively reported in Shanghai and Zhejiang Province as of 16 March 2021 by Local Health Commissions. This YRD is one of the most densely populated regions on earth and belongs to developed areas of China, with over 150 million registered residents, and it is characterized by highly sensitive disease surveillance, reporting, and detection capabilities regarding infectious disease. However, a better understanding is required of the effect of the COVID-19 control measures, implemented in 2020, on the incidence of respiratory infections in Shanghai and Zhejiang Province of China.
In 2019, the WHO listed 10 environmental threats to global health; of these, air pollution was considered to be the greatest [16]. Previous studies have demonstrated a positive association between exposure to air pollution and the rate of respiratory infectious disease [16][17]. For example, recent studies from China reported that short-term exposure to PM 2.5 , PM 10, CO, NO 2 and O 3 concentrations was signi cantly associated with con rmed COVID-19 cases [18]. Thus, a decision was made in the current study to assess the association between air pollutant concentration and the incidence of respiratory infectious disease during the COVID-19 pandemic in 2020. A comparison was performed of the incidence of eight common respiratory infectious diseases (i.e. epidemic parotitis, in uenza, measles, pulmonary tuberculosis, rubella, scarlet fever, and pertussis) and six air pollutants (i.e. PM 2.5 , PM 10 , NO 2 , SO 2 , CO and O 3 ) in Shanghai and Zhejiang Province in 2020 and in the 2017-2019 period. In addition, an attempt was made to identify potential environmental risks associated with this change of RID in Shanghai and Zhejiang Province of China.

Ethics statement
This study was conducted according to the principles and guidelines of the Declaration of Helsinki, and was approved by the Research Ethics Committee of the Zhejiang Provincial Center for Disease Control and Prevention(No.2020-24). All initial information identifying patients was anonymized in this study.

Data collection (eight RID)
Two data sources, Shanghai Municipal Health Commission and the Health Commission of Zhejiang Province of China, were used to obtain information on the incidence of the aforementioned eight common RID. The data included details (i.e. number of cases, incidence and patient data strati ed by onset date [month and year] and area) of probable, clinically diagnosed and con rmed cases of respiratory infectious disease. The population data were acquired from the National Bureau of Statistics of the People's Republic of China, updated at the end of each year.
The mean monthly and monthly data on the concentrations of the aforementioned six air pollutants in Shanghai and Zhejiang Province were obtained from an air pollution database in the east of China managed by Sun Yat-sen University. The data were obtained from 19 and 56 air quality monitoring stations in Shanghai and Zhejiang Province of China, respectively (Supplementary Fig. 1).

Case De nitions
The diagnostic criteria of all eight RID has been issued by the National Health Commission of the People's Republic of China, see Supplementary table1(1 ~ 7). However, the case de nition of seasonal in uenza was changed since 2019. The con rmed cases for seasonal in uenza were de ned as: the clinical presentation is that of any of a number of acute febrile respiratory diseases, i.e., fever, cough, coryza, di culty breathing or with a history of contact with a con rmed or suspected case and a laboratory test positive for in uenza virus, including in uenza antigen, PCR, viral isolation or a four-fold or greater increase in serum antibodies speci c for this virus isolated in paired sera.

Statistical analysis
The monthly and annual incidence of respiratory infectious disease (per 100,000) was de ned as the number of monthly and annual cases of respiratory infectious diseases divided by population size. To visually demonstrate the impact of COVID-19 on the incidence of respiratory infectious disease, the data were categorized according to different stages (i.e. the 2017-2019 period [pre-COVID-19] and 2020 [post ). Comparisons of the average annual incidence of RID and air pollutant concentrations in aforementioned two stages were conducted using a two proportional test statistics with asymptotical normal distribution. We used the seasonal ARIMA model (Box and Jenkins,1976) to predict the incidence of respiratory diseases and the concentration of air pollutants during 2020 COVID-19 outbreak, and two proportional tests were constructed to check whether the real incidence is same as the predicted incidence of RID during 2020. The widely-used Pearson correlation coe cient was to calculate the association between the incidence of respiratory diseases and air pollutants for Shanghai and Zhejiang.
A two-ratio test for checking whether the incidence is equal during the emergency stage and in a routine period. The piecewise regression model in Toms and Lesperance (2003) is to identify of changes of incidence data. A p value of < 0.05 was considered to be statistically signi cant. All analysis in this study were conducted using R statistical software. The seasonal ARIMA model was constructed by using the function "Auto. arima" in the forecast package, and the piecewise linear regression model was constructed by using the "piecewise.linear" in the Sizer package.

Results
Descriptive analysis of the difference in the incidence of RID between the 2017-2019 period and 2020 Between 2017 and 2020, 81,345 and 1,048,511 cases of eight different types of RID were reported in Shanghai and Zhejiang Province, respectively. The average annual incidence of RID (per 100,000 people) was 140.64 cases and 525.01 cases in Shanghai and Zhejiang Province, respectively. The incidence (per 100,000) for seven of the eight RID (excepting in uenza) decreased by 37.80% (95% con dence interval   Fig. 3). Generally, the average annual incidence of overall RID was considerably higher in Zhejiang Province than that in Shanghai for both periods (Supplementary Table 4).
Descriptive analysis of the difference in the incidence of overall RID during the emergency and routine responses toCOVID-19 in 2020 The monthly incidence of overall RID was considerably higher in Zhejiang Province than in Shanghai for both periods (p = < 0.050) (Fig. 4). The 2020 period under evaluation was divided into an emergency response (to COVID-19) stage (January to April 2020) and a routine response (to COVID-19) stage (May to December 2020). During the emergency response stage, the overall incidence of seven RID decreased signi cantly, by 37.76% in Shanghai (95% CI: 28.07-47.46) (p = < 0.001) ( Fig. 5 Table 6).

and Supplementary
Descriptive analysis of the difference in overall actual and predicted rates of respiratory infectious disease in 2020 A seasonal ARIMA model was used to predict the incidence of RID and to compare the difference in actual and predicted incidence. In Shanghai, actual overall eight RID incidence was 52.60% (95% CI: 51.41 to 53.78, p < 0.001) lower than the predicted rate (131.63 vs. 277.69 per 100,000 population, Fig. 7 and Supplementary Table 7). Similarly, in Zhejiang Province, true incidence was considerably lower (by 24.06%) than predicted incidence (95% CI: 23 Descriptive analysis of the difference in air pollutant concentrations between the 2017-2019 period and 2020 The monthly concentrations of overall six air pollutants (i.e. PM 2.5 , PM 10 , NO 2 , SO 2 , CO and O 3 ) decreased by 12.7% (95%CI: 4.82-20.49, p = 0.003) and 12.85% (95%CI: 3.80-21.89, p = 0.008) in Shanghai and Zhejiang in 2020 compared to the 2017-2019 period (Supplementary Table 9 Table 9). The next largest decrease was observed in PM 2.5 concentrations, which dropped by 26.30% in Zhejiang Province in Zhejiang Province when evaluated using Poisson regression (p = < 0.050) (Fig. 9).
The negative association between O 3 and in uenza incidence was both found in Shanghai (r = -0.64) and Zhejiang Province (r =-0.54).
We used the piecewise linear regression models to determine the impact of air pollutant concentration on the incidence of in uenza. The results showed that SO 2 and PM 2.5 concentrations were positively associated with in uenza incidence in Shanghai. The corresponding change points for SO 2 and PM 2.5 concentrations were 0.24 µg/m 3 and 0.05 µg/m 3 , respectively (Fig. 10a). That is, when the concentration of SO 2 was greater than 0.24µg/m 3 or the concentration of PM 2.5 was greater than 0.05 µg/m 3 , the incidence of in uenza increased with the SO 2 and PM 2.5 concentrations increased. Similarity, PM 10 , NO 2 and CO concentrations were shown to positively correlate with in uenza incidence. However, O 3 concentrations were observed to negatively correlate in uenza incidence (Fig. 10b). In Zhejiang Province, PM 2.5 and CO concentrations had a positive association with in uenza incidence; conversely, O 3 concentrations negatively correlated with in uenza incidence (Fig. 10c).

Discussion
Using representative data from the Yangtze River Delta Region, China, the current study identi ed a marked decline in the overall annual incidence of eight respiratory infectious diseases in 2020, compared to the previous three years. The overall incidence of respiratory infectious disease in 2020 was lower than that predicted in both regions in 2020. A similar decrease was demonstrated in the concentrations of six air pollutants in 2020; of these, SO 2 , PM 2.5 and PM 10 were signi cantly associated with a decrease in in uenza incidence.
Since Shanghai and Zhejiang Province are geographically proximal to Hubei Province, a large number of the population returned from Hubei Province before Wuhan went into lockdown on 23 January 2020. Both areas were the rst to be impacted by COVID-19 [5]. Zhejiang Province and Shanghai reported the rst imported cases of COVID-19 (i.e. from Hubei Province) on 20 January 2020 and 23 January 2020, respectively [19]. Thereafter, COVID-19 rapidly spread, over the next two months, to 11 cities in Zhejiang Province and to 16 Shanghai districts. In response to this novel threat, on 23 January 2020, the governor initiated a top levl emergency response to COVID-19 for implementation in Zhejiang Province and Shanghai. The application of strict precautionary measures, such as lockdowns, home stays, the closure of schools and the suspension of large-scale events, disrupted the transmission of COVID-19, which simultaneously resulted in a decrease in the incidence of other respiratory infectious diseases, including in uenza. In this regard, the ndings of the current study support those of studies conducted outside China [12][13][14]20]. Several explanations have been posited for the decrease in the incidence of respiratory infectious diseases. Firstly, these diseases have a similar mode of transmission to COVID-19 (i.e., via respiratory transmission or contact) [12][13]; therefore, it may be assumed that interventions to prevent COVID-19 have reduced the spread of respiratory infectious disease. Secondly, di culty accessing hospital services during COVID-19 owing to strict quarantine measures, together with fears of contracting COVID-19 at a hospital, have inhibited opportunities for disease transmission [11]. Lastly, people's hygiene habits have improved considerably as it is mandatory to wear masks, regularly wash hands and implement proper ventilation. Generally, it is feasible that the lockdown, in conjunction with healthseeking behavioural changes and improved personal hygiene, has reduced the risk of the transmission of respiratory infectious disease pathogens [10][11][12][13][14]20].
In the current study, the rate of in uenza slightly increased during the emergency response (to  stage in Zhejiang Province, which could have been owing to awareness of similarities in respiratory infectious disease and COVID-19 symptoms, together with enhanced surveillance, testing and diagnostic strategies for in uenza-like illnesses in early 2020. However, the incidence of in uenza remained signi cantly low during the routine response (to COVID-19) stage in both regions. This nding could be attributed to the e cacy of SARS-CoV-2 prevention strategies.
COVID-19 prompted a period of nationwide public lockdown, which provided an invaluable opportunity for an evaluation of the correlation between air pollutant concentrations and RID. In the current study, the monthly concentrations of six air pollutants were found to decrease signi cantly in 2020, compared to the previous three years, and the levels were much lower than those stipulated in the Chinese guidelines [16].The most signi cant decrease was seen in SO 2 concentrations in both regions. In 2020, this reduction can be attributed to by the reduction of the primary air pollutant emissions, for example, almost all medium and small industries except power plants and large-scale enterprises were closed [21]. In addition, o cially, the YRD cities started its full lockdown on January 23rd -25th and remained in place until the end of April, these policies have led to reduced human activities, which caused improvement of air quality as a side-product. Similar ndings were reported elsewhere in this regard [21].
It was curious to determine if low air pollutant concentrations were associated with the decrease in the incidence of RID. As with previous ndings, it was established that the decrease in short-term exposure by people to air pollutants inhibited the spread of RID in the Yangtze River Delta Region, China. Specially, the SO 2 and NO 2 concentration in Shanghai and PM 2.5 and CO in Zhejiang Province had signi cant effects on the incidence of in uenza.
The reasons were not clear, which might be driven by the difference of primary emissions, population dentistry, and energy and industrial strategy, etc. Firstly, a recent epidemiological study concluded that decreased short-term exposure to particulate matter was associated with a decline in the use of healthcare services for acute lower respiratory infections [22]. Secondly, exposure to urban airborne particulate matter has been demonstrated to alter the macrophage-mediated in ammatory response to respiratory viral infection [18,22]. Thirdly, exposure to PM 10 and PM 2.5 concentrations could signi cantly enhance RNA virus infections, such as H1N1 and H5N1, in A459 human lung epithelial cells by increasing viral replications [16,23]. Fourthly, air pollutants affect the lower respiratory tract protease-antiprotease balance and micro ora, which are associated with respiratory infections [18].
To the best of our knowledge, this is the rst study to have evaluated differences in the incidence of respiratory infectious disease and air pollutant concentrations in the Yangtze River Delta Region, China, between the period, 2017-2019 (pre-COVID- 19), and 2020 (post COVID-19). However, the study had several limitations. Firstly, the evaluation was limited to air pollutants, and it is possible that other meteorological factors might have in uenced the transmission and pathogenesis of respiratory infections. Secondly, this study might have the reporting biases because of health-care seeking behavior and laboratory capacity.Thirdly, this is an ecological study, and we can nd the decline of RID and six air pollutants. However, current evidence cannot justify any causal relationship between air pollution and RID downtrends during 2020.

Conclusion
The current study indicated that the incidence of eight RID and the concentration of six air pollutants in Shanghai and Zhejiang Province decreased signi cantly in 2020 compared to the previous three years. The most signi cant reduction pertained to SO 2 concentrations in both areas. During the most stringent emergency response period, the overall RID has been reduced by 37.76% and 22.76% in Shanghai and Zhejiang Province respectively. The in uenza reduction was positively related with the NO 2 and SO 2 in Shanghai but associated with the CO and PM 2.5 in Zhejiang Province. This study provided the additional evidences that the measures taken for COVID-19 were effective in improving the air quality greatly and decreased the other RID. In future studies, the accurate mechanisms should be explored further that utilise a large-sample design; more stringent regional jointcontrol within YRD should be pushed forward to achive a better air quality and infectious disease control.       Differences in the average monthly incidence of eight respiratory infectious diseases in the emergency response stage (January to April 2020) and the routine response stage (May to December 2020) in Shanghai, China Notes: The pink lines re ect differences in 95% con dence interval values. The pink dot indicates the 50% median point. Changes = (x2-x1)/x1×100% , x1: average monthly incidence in 2017-2019, x2: average monthly incidence in 2020. The p-values were computed using two proportional tests.

Declarations
For the emergency or routine response stages, the p-values for the emergency versus the routine response stages were computed using two ratio tests.

Figure 6
Differences in the average monthly incidence of eight respiratory infectious diseases in the emergency response stage (January to April 2020) and the routine response stage (May to December 2020) in Zhejiang Province, China Notes: The pink lines re ect differences in 95% con dence interval values. The pink dot indicates the 50% median point. Changes = (x2-x1)/x1 ×100%, x1: average monthly incidence in 2017-2019, x2: average monthly incidence in 2020. The p-values were computed using two proportional tests. For the emergency or routine response stages, the p-values for the emergency versus the routine response stages were computed using two ratio tests.

Figure 7
A comparison of the true and predicted average annual incidence of eight respiratory infectious diseases in Shanghai and Zhejiang of China in 2020 Notes: The pink lines re ect differences in 95% con dence interval values. The pink dot indicates the 50% median point. The p-values were computed through two proportional tests . Changes = (x2-x1)/x1×100%, x1: average yearly incidence in 2020 prediction; x2: average yearly incidence in 2020.  Pearson's correlation coe cients for air pollution concentrations and the incidence of seven respiratory infectious disease in Shanghai and Zhejiang Province, China, for the period 2017-2020 Notes: Showing PM2.5 particulate matter with an aerodynamic diameter of < 2.5 μm and PM10 particulate matter with an aerodynamic diameter of < 10 μm, SO2: sulphur dioxide, NO2: nitrogen dioxide, O3: ozone, CO: carbon monoxide .*: p-value 0.050 ≥ p-value of > 0.010; **: p-value of 0.01 ≥ p-value of > 0.001; **: p-value of ≤ 0.001 PTB: Pulmonary tuberculosis. On the left gure notes the data from Shanghai; On the right gure notes the data from Zhejiang Province.