Online public concern about allergic rhinitis and its association with COVID-19 and air quality in China: an informative epidemiological study using Baidu index

The identification of keywords in this study can be shown in the flowchart in Figure S1 (see Additional file 1). Keywords specific retrieval process is detailed in Additional file 1. A total of 33 keywords were generated for the subsequent analysis. According to allergic rhinitis guidelines and medical practitioners' recommendations [1]. The 33 search terms were further grouped into the four categories of “disease”, “etiology”, “symptoms/complications”, and “disease treatment/management”, which are listed in Table S1 (see Additional file 1). The daily BSIs of each term were extracted and used for subsequent analysis.

The Chinese government announced the closure of Wuhan on January 23, 2020, so we chose this time point to separate the two-year periods before and after the COVID-19 pandemic [25]. Following that, we defined the period from January 23, 2017 to January 22, 2020 as the “pre-COVID-19” segment, while the period from January 23, 2020 to June 23, 2022 as the “post-COVID-19” segment. According to the time when the WHO named the major prevalent variants and they spread in China, “post-COVID-19” has been divided into four periods, named the original-variant (from January 2020 to November 2020), Alpha (from December 2020 to April 2021), Delta (from May 2021 to November 2021) and Omicron (from December 2021 to June 2022) period respectively [26‐28]. Original-variant referred to all mutant variants before Alpha. Additionally, in order to understand the relationship between air quality and the public concern about allergic rhinitis or mask-wearing, we choose Beijing as a proxy city, the detailed description is shown in Additional file 2. We also retrieved daily BSIs of 33 allergic rhinitis-related keywords from January 23, 2019 to June 23, 2022 and the “mask wearing” term for the period from January 23, 2020 to June 23, 2022 for consolidation and analysis. The air quality index (AQI) data were obtained from the website of the China Air Quality Online Monitoring and Analysis Platform [29]. From this platform, the daily AQI of Beijing from 23 January 2019 to 23 June 2022 was obtained. AQI is a composite index reflecting the level of air pollution. The daily AQI values usually indicates real-time air pollution and warns people of potentially poor air quality. The higher the AQI value is, the more severe the air pollution is.

To analyze search trends for allergic rhinitis, we summarized the monthly BSI (sum of daily BSIs from the first day of a month to the last day of a month) for all search terms related to allergic rhinitis from January 23, 2017 to June 23, 2022, and plotted curves of search volume for all keywords to describe changes in public attention. Meanwhile, the time series data were seasonally decomposed using the ‘decompose’ function of R software (version 4.1.2) into trend factors, seasonal components and random variables, representing both irregular and periodical changes over time.

To take into account the impact of the number of internet users during a certain period, we extracted the number of Internet users (million) in 2019 and 2020 from the China Statistical Yearbook respectively, and calculated the adjusted daily BSI values by the formula of daily BSI/number of Internet users (million) [30, 31]. The Mann–Whitney U test was implemented to test the null hypothesis of this study, that is, there was no statistically significant difference in the adjusted allergic rhinitis BSIs before and after COVID-19. The median adjusted BSI was used to describe the search volume for each theme of allergic rhinitis. Spearman’s “Rho” correlation was used to measure the association between the values of daily AQI and the daily BSIs related to allergic rhinitis. Statistical analysis was performed using IBM SPSS Version 23.0 (IBM Corporation), and p-values < 0.05 were considered to be statistically significant.

We created monthly time-series curves for the total BSI of all allergic rhinitis-related keywords from January 23, 2017 to June 23, 2022 (Fig. 1). Overall, we found there is a discernible seasonal fluctuation in search interest of allergic rhinitis-related keywords. This seasonal fluctuation could be described as two separate peaks from March to May and from August to October. Following the COVID-19 outbreak, the search interest for allergic rhinitis dropped significantly during the first peak period from March to May 2020 compared to the same period in 2017, 2018 and 2019.

After the decomposition, the trend component of the total BSI of allergic rhinitis began to rise rapidly since January 2017, and reached its peak in April 2019. After that, it showed a downward trend in the following months, and hit the bottom in May 2020. Meanwhile, a regular seasonal component with two peaks from March to April and in August was also observed, indicating the seasonality of the public’s search interest for allergic rhinitis (Fig. 2a). We further separately investigated the four themes of allergic rhinitis: “disease”, “etiology”, “symptoms/complications” and “disease treatment/management” (Fig. 2b), and found that the “disease treatment/management” theme revealed similar pattern to the overall searching trends. More details could be found in Additional file 3.

Comparison of changes in overall search interest for allergic rhinitis between the pre-COVID-19 and post-COVID-19 periods found that the adjusted daily BSI showed a significant decrease in search volume after the pandemic of COVID-19 (25.65 vs 23.48; p < 0.001), as shown in Table 1. Among the four allergic rhinitis-related themes, the search interest of the “symptoms/complications” theme increased significantly (1.34 vs 1.66; p < 0.001) after COVID-19 whereas the research interest of the “disease” (5.16 vs 3.76; p < 0.001), “etiology” (2.27 vs 2.20; p = 0.022), and “disease treatment/management” (16.96 vs 15.85; p < 0.001) themes all decreased significantly after COVID-19. In detail, the adjusted BSI for the term “symptoms of allergic rhinitis” in the “symptoms/complications” theme has the most significant increase after the occurrence of COVID-19 (p < 0.001) (see Additional file 4, Table S2). There are also representative changes in some other keywords as detailed in Additional file 4.

Our study also investigated the effect of air quality on public concern about allergic rhinitis. The higher the AQI index, the worse the air quality. The trend of Beijing allergic rhinitis BSI was summarized in Figure S2 (see Additional file 5), where we found that similar to the national-wide results, at the beginning of the COVID-19 outbreak, Beijing’s public concern about allergic rhinitis was significantly lower than that in the same period of 2019. We first analyzed the correlations between each BSI of allergic rhinitis-related keywords and AQI during the one-year period before COVID-19 outbreak, and found 8 keywords with statistically significant correlations with the AQI (p < 0.05) (see Additional file 6, Table S3).

Subsequently, we analyzed the association between BSIs of allergic rhinitis search behavior (both the total and the 4 themes) and AQI after COVID-19 outbreak and compared them with one year before COVID-19 outbreak. Notably, as shown in Fig. 3, there was a positive and significant correlation (r = 0.108, p < 0.05) between AQI and the BSIs of the “etiology” theme of allergic rhinitis one year before COVID-19 outbreak, but negative and significant correlations between each pair of AQI and four themes of allergic rhinitis after the occurrence of COVID-19 (p < 0.05). We also found that AQI was significantly negatively correlated with all allergic rhinitis themes in the Original-variant period (total: r = -0.195, p < 0.05), and was partially (not significantly) negatively associated with allergic rhinitis in the Alpha and Delta-variant periods. In the Omicron period, a higher level of AQI was associated with increased search interest of allergic rhinitis, showing a significant positive correlation (total: r = 0.377, p < 0.05). More specific descriptions can be seen in Additional file 7.

We also explored the link between the search behavior of wearing masks and AQI during the four epidemic periods (Fig. 4). The result showed positive correlations between mask-wearing BSI and AQI during the Original-variant, Delta and Omicron variant dominance periods, implying that the AQI in Beijing positively affected the mask wearing BSI in these periods. However, the positive correlation coefficient was small and not significant (r = 0.069, p = 0.224) between AQI and mask wearing BSI in Original-variant period, whereas such correlation increased to a significant level in the Delta (r = 0.198, p = 0.004) and Omicron (r = 0.206, p = 0.003) variant period.

In addition, during the Omicron-variant period, the correlations between BSIs of allergic rhinitis and mask wearing showed stronger and significant positive associations (r = 0.585, p < 0.05) (see Additional file 8, Table S4). More details could be seen in Additional file 8.

Our findings suggested that public concern about allergic rhinitis has generally declined during the COVID-19 pandemic, especially in the first half of 2020. Meanwhile, we also detected a regular seasonal component with two peaks in spring and summer for the allergic rhinitis-related internet searches in China. In addition, we examined the relationship between air quality and allergic rhinitis-related searches in the city of Beijing from 2019 to 2022. One year before COVID-19, people’s search behavior for allergic rhinitis increased as air quality became worse. However, during the COVID-19 pandemic, allergic rhinitis-related search behavior decreased with deteriorating air quality. This may be related to limited outdoor activities under the quarantine policy and widely-adopted protective behaviors such as wearing masks during the pandemic.

Our study showed that allergic rhinitis BSI had a marked seasonal variation, which was consistent with the two previously reported peaks of airborne pollen [32]. The seasonality of allergic rhinitis search behavior may be related to both pollen and climate-changing-induced allergic rhinitis. Firstly, March to May is generally recognized as the pollen season, and pollen can act as an allergen to induce the allergic rhinitis onset [33, 34]. Secondly, from July to September, the climate is relatively hot and the frequent use of air conditioners can easily cause indoor air drying, resulting in impaired blood circulation in the nasal mucosa, which in turn triggers the onset of allergic rhinitis [35]. Therefore, along with the increased incidence of allergic rhinitis, the public's attention to allergic rhinitis also reaches its peaks.

When assessing the trend component after removing the seasonal effects, our results showed that public attention to allergic rhinitis on the Internet in China gradually increased from 2017 onwards, reaching a peak in April 2019. This was similar to previous epidemiological studies elucidated that the prevalence of allergic rhinitis maintained at a high level from 2017 to 2019 [3, 36]. Subsequently, public interest in allergic rhinitis declined in the first half of 2020, but began to rise again in the second half of that year. In summary, public concern about allergic rhinitis was significantly downward in the early stages of the outbreak, compared with before the COVID-19 outbreak in China. This might be attributed to the actions taken by the Chinese government at all levels to contain COVID-19 in the early stages of the outbreak, as preventive measures against COVID-19 virus also played an integral role in reducing the onset of allergic rhinitis caused by outdoor allergens, as confirmed in previous studies [37, 38].

We carefully comparing Internet searches for the four themes of allergic rhinitis before and after the COVID-19 pandemic, the initial increase in searches of “symptoms/complications” may have been due to some overlap between early symptoms of the COVID-19 infection and respiratory symptoms in patients with allergic rhinitis [39]. Therefore, people may utilize Internet searches for rhinitis symptoms to make a preliminary differential diagnosis of the two conditions. It was also worth noting that, the search volume of the theme “symptoms/complications” had another rapid growth peak after May 2021, which was exactly when the State Council accelerated the promotion of COVID-19 vaccination nationwide. At the same time, the CDC of China released the COVID-19 Vaccination Adverse Reaction Surveillance Report. Therefore, increasing internet searches of “symptoms/complications” of allergic rhinitis may also be driven by those potential adverse reactions that people experienced or were concerned about after the COVID-19 vaccination [40], including allergic rhinitis and other allergies.

On the other hand, for “disease”, “etiology” and “disease treatment/management” themes, the implementation of the quarantine policy has restricted people’s outdoor activities and reduced their opportunities for strenuous exercise, which may subsequently reduce the incidence of allergic rhinitis triggered by excessive physical activity [41], leading to decreased search interest of allergic rhinitis and its treatment after the COVID-19 outbreak. Moreover, our results showed that pollen-related keywords “hay fever” and “pollen allergy” dropped significantly after the COVID-19 outbreak. One explanation is that in order to prevent the spread of the COVID-19 virus, people consciously maintained social distance and wore masks in public places [10, 37, 38, 42], and these measures have blocked airborne pollen, allergen exposure and cold air irritation to a certain extent, therefore effectively avoiding the onset of the disease in patients with allergic rhinitis.

In this study, the context of the COVID-19 pandemic provided us an opportunity to explore the relationship between daily air quality and public concern about allergic rhinitis, as well as their correlations with mask-wearing behavior. The correlation analysis focusing on Beijing area provided following main findings.

On the one hand, we noted that is, as air quality became worse, allergic rhinitis BSI increased accordingly before 2020, but decreased after the COVID-19 pandemic. Previous studies have shown that air pollutants can aggravate the symptoms of people with allergic rhinitis, and thus, air pollution was associated with allergic rhinitis episodes [20, 43]. Meanwhile, the implementation of national and local quarantine policies during the COVID-19 pandemic led to a reduction in human-related activities, such as industrial production and transportation, indirectly diminishing levels of urban air pollution [23, 24]. Therefore, we speculated that the effect of poor air quality on triggering allergic rhinitis may be attenuated by COVID-19 preventive and controlling measures. In the context of the COVID-19 epidemic, wearing masks can not only reduce the spread of the virus through the respiratory tract, but also simultaneously assist people with allergic rhinitis in reducing their exposure to air pollutants to cause allergic rhinitis onset [15, 44]. Therefore, the simultaneous improvements in air quality during quarantines and the adoption of mask-wearing were accompanied by a decline in search behavior related to allergic rhinitis.

On the other hand, we found that the association between air quality and allergic rhinitis BSI differed during the four different variant epidemics. Such dissimilarity was particularly pronounced during the epidemics of the Original and Omicron variants, with AQI and allergic rhinitis BSI being significantly negatively correlated during the Original-variant period but significantly positively correlation during the Omicron period. This may be explained by the changes in quarantine policies and people’s mask-wearing behavior due to the differences in transmissibility and toxicity of various variants of the COVID-19 virus [45, 46]. One possible explanation is that, at the beginning of the epidemic of the Original-variant, the mass media in China broadcasted lots of personal protection measures and urged people to wear masks, and thus most people strictly enforced the mask-wearing behavior regardless of air quality during the Original-variant period [47]. In addition, the implementation of the quarantine policy restricted also greatly increased the exposure to indoor allergic substances, leading to the incidence of allergic rhinitis [48, 49]. So even though the air quality became better in the period, people’s search behavior for allergic rhinitis increased. However, during the Omicron period, most of the mass media calls for wearing masks were “whether or not to wear a mask or not is determined based on the risk of infection” [50]. The masks were only required in public areas where crowds gathered, and the enforcement of mask wearing was not as strict as it was during the Original-variant period, while society gradually resumed production and work, and air pollution generated by human activities gradually increased. Therefore, during the Omicron period, the worse the air quality was, the more people were interested in searches of masks, and the deterioration of air quality also led to a simultaneous increase in the onset of allergic rhinitis and related internet searches.

In summary, in the context of the COVID-19 pandemic, the Baidu index reveals the potential demand of the public for allergic rhinitis-related information and mask-wearing behavior under different air quality levels. In fact, along with the liberalization of the COVID-19-related policies, personal protective behaviors will be decreased compared with the early stage of the pandemic. In this case, the health administrative department should encourage patients with allergic rhinitis not only to be aware of daily air quality, but also to wear masks accordingly to reduce exposure to allergens, especially during seasonal peak period of allergic rhinitis.