Epidemiology of HBoV1 infection and relationship with meteorological conditions in hospitalized pediatric patients with acute respiratory illness: a 7-year study in a subtropical region

Human bocavirus 1 (HBoV1), which belongs to family Parvoviridae, was firstly identified in respiratory secretions of children with respiratory tract disease in 2005 [1, 2]. HBoV1 has been confirmed as an important respiratory pathogen and is found in respiratory infections in children and adults worldwide. The prevalence of HBoV1 nucleic acid detection varies from 1.5 to 33% in patients with acute respiratory illness (ARI), according to different studies [3–7]. Serological and nucleic acid test results are generally consistent [8–11], showing HBoV1 infection is very common. HBoV1 can cause both upper respiratory illness (URI) and lower respiratory illness (LRI) [12–18]. Infection with HBoV1 can lead to development of a cough, rhinitis, fever and other common clinical symptoms [15, 19]. In some cases, it can cause respiratory distress, hypoxia, wheezing and other severe respiratory symptoms [18, 20]. Clinical diagnosis is mainly pneumonia, bronchitis, pneumothorax, mediastinal emphysema and otitis media and other complications [18–22]. In some cases, patients develop severe respiratory injury symptoms, which can be fatal [21, 23]. HBoV1 can be detected in fecal samples [24], blood samples [25, 26], urine [27, 28], cerebrospinal fluid [29–31], river water [32] and sewage [33, 34], indicating that HBoV1 may be associate with a variety of diseases. Current in vitro studies modeling tissue-like airway epithelial cells cultures show HBoV1 infection can lead to disruption of the tight-junction barrier, loss of cilia and epithelial cell hypertrophy [35–37], similar to lung injury tissue changes in vivo. There is currently no vaccine or specific treatment for this virus; prevention and treatment of HBoV1-related diseases still require further research. The prevalence of respiratory viruses is associated with many factors, including local climate, which may impact the survival and spread of the viruses [38]. Studying the epidemiology of HBoV1 and its relationship with meteorological conditions will improve diagnosis, treatment, control and prevention of this virus.

In this study, we investigated the epidemiology of HBoV1 infection in children (≤14 years old) hospitalized with ARI in a subtropical region in China over a 7-year period. In addition, we collected climate data to determine if there was a relationship between HBoV1 prevalence and meteorological conditions. This study will add to existing epidemiological data on HBoV1 and its relationship with climate conditions in subtropical regions and will play a positive role in HBoV1 control and prevention.

ARI is one of the most common human diseases, predominantly caused by different respiratory viruses [41, 42]. One of these viruses, HBoV1 infection, causes global epidemics, has a high public health burden and circulates with different patterns in different areas [3–7, 43]. In general, the prevalence of viruses varies because of factors such as geographical location, climatic conditions, population and social activity [38]. Epidemiology of HBoV1 in temperate regions has been described in more detail and a high incidence of infection has been observed in children under the age of 2 years in winter and spring [15, 16, 39, 44].

To describe the epidemiology of HBoV1 in Guangzhou, we collected throat swabs from 11,399 children (≤14 years old), hospitalized with ARI and monitored HBoV1 and other common respiratory pathogens over a 7-year period (Table 1).

In the current study, 86.5% (9857/11399) of patients were under the age of 5 years, with a median age of 1.75 years, indicating that infants and young children were most at risk of ARI, consistent with previous reports [45, 46]. Overall, 49.2% (5606/11399) of patients tested positive for one or more respiratory pathogens, 2.2% (248/11399) of patients were tested with HBoV1 infection (Table 1). A higher prevalence of HBoV1 was detected in male patients compared with female patients (p = 0.019), consistent with previous reports [15, 16, 39, 44].

Co-infection with HBoV1 and other pathogens is common [14, 15]. In our study, 45.2% (112/248) of HBoV1-positive patients also tested positive for other pathogens (Table 1). This may be partly caused by coinciding epidemics of HBoV1 and other pathogens. In our study, the HBoV1 seasonal distribution and total positive pathogen distribution were consistent, confirming this inference (Fig. 2). Current research shows that HBoV1 infection can lead to the collapse of the first line of defense of airway epithelium [35–37], which may lead to a higher susceptibility to other pathogens, explaining the high rate of co-infection. Whether co-infection leads to more severe disease is currently unknown and more research is needed to determine this. The characteristics of the HBoV1 infection are likely to be a good model for studying the effects of co-infections.

In this study, there was a significant difference in prevalence of HBoV1 in patients of different ages (p < 0.001). The majority of HBoV1 infections occurred in patients under 2 years old and the peak frequency of HBoV1 infection occurred in patients aged 7–12 months (Fig. 1), consistent with previous serological and epidemiological reports on the virus [8–11, 15, 16, 39, 44]. This might be because children’s immune systems are still under development and maternal antibodies gradually disappear in this age group. The distribution of HBoV1 in patients of different ages will provide important reference for future vaccines and new drug research and development, as well as providing important data for disease prevention and control.

Many factors affect the epidemiology of pathogens, such as geographical location and local climate. Guangzhou, a central city and main transport hub in southern China, is located in a subtropical region. Guangzhou is hot and has high annual rainfall, long summers, short winters and the annual precipitation and high temperature are almost in the same period (Fig. 3). In this study, two HBoV1 peaks were observed (Fig. 2). The large prevalence peaks of HBoV1 infection occurred between June and September of each year, which are the summer months in Guangzhou, with mean temperatures of higher than 25 °C (Fig. 3). Small peaks of HBoV1 infection occurred in winter, between November and December of each year. This seasonal distribution is similar to the prevalence in subtropical regions reported previously [47], but different from the HBoV1 epidemics in temperate regions, which mostly occur in winter and spring [15, 16, 39, 44], as well as from tropical regions, such as India, where no obvious epidemic season has been found [48].

To analyze the correlation between HBoV1 prevalence and meteorological conditions, multiple linear regression analysis was performed, with HBoV1 monthly prevalence as the dependent variable and mean temperature (or mean temperature in the preceding month), mean relative humidity, mean wind speed and sunshine duration as the independent variables (Table 2). Both regression models were established (p < 0.001) and the adjusted R² value (0.373) of the temperature dorp 1 month model was greater than the adjusted R² value (0.231) of the current monthly temperature model, indicating that the temperature dorp 1 month model had better explanatory power than the current monthly temperature model. Both of the models showed that the prevalence of HBoV1 was significantly correlated with temperature and relative humidity (Table 2). In detail, HBoV1 prevalence was positively correlated with temperature, that is consistent with previous reports [47, 49]. Conversely, HBoV1 prevalence was negatively correlated with relative humidity, this was different from a previous report in Suzhou [47], which may be related to Guangzhou high humidity (mean monthly relative humidity was 77.2 ± 7.3%) (Fig. 3). It is common for pathogen prevalence to fluctuate over time because of a variety factors. In this study, HBoV1 prevalence was relatively low in 2013 to 2014. It might be partly related to the relatively higher mean relative humidity during this period (Fig. 3). Climate conditions may impact the survival and spread of respiratory viruses, however no significant linear relationship between HBoV1 infection and wind speed or sunshine duration were found in this study (p > 0.05) (Table 2).

Some limitations of this study should be noted. First, because our study mainly focused on HBoV1 circulation in hospitalized patients with ARI, HBoV1 in outpatients and the asymptomatic population were not included. Second, many factors can affect virus epidemics, meteorological data analysis alone may not serve as a final conclusive interpretation. Third, the study was only conducted in three hospitals and may not be representative of the overall population.

Our study has provided a better understanding of the epidemiology of HBoV1 in subtropical regions, specifically correlations with climate data; these data will be helpful for future control and prevention of HBoV1 infections.