Introduction

Hantaviruses (HVs) are a genus of single-stranded, enveloped, negative-sense RNA viruses that usually cause chronic asymptomatic infection in rodents1. The genome of HV includes large (L), medium (M) and small (S) segments, which encode viral RNA polymerase, envelope glycoproteins Gn and Gc, and nuclear protein (NP), respectively2. The L segment is the most conserved among the three segments; the M segment expresses Gn and Gc proteins, which have neutralizing antigen sites and haemagglutination active sites that can induce the production of neutralizing antibodies and haemagglutination-inhibiting antibodies; and the NP encoded by the S segment can stimulate the humoral and cellular immune responses of the body3. Regarding the relationship between HV and natural hosts, the ecological environment and evolution, scholars have found that the epidemic of HV and the evolution of the epidemic area are the result of the long-term interaction between the virus and the host4. The geographical distribution and natural history of natural hosts play a key role in the prevalence range and intensity of HV-related diseases5.

When humans are incidentally infected with some strains of HV through inhalation of virus-containing aerosols or being bitten by rodents, two human syndromes can result: HV cardiopulmonary syndrome (HCPS) in the Americas and haemorrhagic fever with renal syndrome (HFRS) in Asia6. HFRS is a group of rodent infectious diseases with typical clinical manifestations of fever, haemorrhage, headache, abdominal pain and acute kidney damage that varies from mild to severe and is related to different pathogens7. In Europe, the primary HFRS pathogens are Puumala virus (PUUV) and Dobrava-Belgrade virus (DOBV), while in Asia, the primary pathogens are Hantan virus (HTNV), Amur virus (AMRV) and Seoul virus (SEOV)8.

Suspected HFRS records might be traced back to the 1930s, but it did not attract substantial attention until more than 3000 United Nations soldiers were infected from 1950 to 1953 in the Korean War9. In 1978, Lee discovered that its causative agent was HTNV, and the natural reservoir was the striped field mouse, Apodemus agrarius10. Although the understanding of HV infection has increased considerably worldwide over the past few decades, both the amplitude and the magnitude of HV outbreaks have been increasing, and several novel HVs with unknown pathogenicity have been identified in some insectivore hosts11.

It is estimated that more than 20000 HFRS cases occur every year globally, with the majority occurring in Asia. Since the first clinical HFRS case was detected in mainland China in 1931, China has gradually become the most prevalent country in the world, with 209209 HFRS cases and 1855 deaths reported during 2004-2019, accounting for more than 90% of the cases worldwideData analysis

Classification data are expressed as rates, and continuous data are expressed as the mean ± standard deviation or interquartile range (IQR) to reflect the concentration discrete trend. The temporal periodicity and spatial clustering of HFRS morbidity in Jiangxi from 2005 to 2021 were explored by periodogram based on Fourier transformation and spatial autocorrelation analysis determined by Moran’s I indicator using GeoDa software (Version 1.16), respectively20,21. Joinpoint regression was conducted by Joinpoint software (Version 4.9.1.0) to describe the time trend changing pattern of HFRS morbidity22. The join points were identified by the grid search method when linear trends changed significantly in direction or magnitude, and each trend change in the final model was described by an annual percentage change (APC) and average annual percentage change (AAPC) with 95% CI23. The multiple sequence alignment and amino acid variation analysis of the sequences included in this research were applied by MEGA X software (Version 11.0), and the neighbour-joining (NJ) method was used to construct the phylogenetic tree with a bootstrap test of 1000 replicates. Then, BioEdit software (Version 7.0) was used to conduct homology analysis. RDP4 (Version 4) and SimPlot software (Version 2.5) were used to conduct gene recombination signal analysis for the whole variant strain sequence to understand the variation in epidemic strains and vaccine strains.

Ethics approval

All experimental protocols were approved by the Jiangxi Center for Disease Control and Prevention Ethics Committee. The data obtained from the CISDCP and GenBank were anonymized so that subjects could not be identified. The ethics committee of Jiangxi Center for Disease Control and Prevention waived the need for informed consent from patients due to the retrospective nature of the study. All methods in our study were used in accordance with the relevant guidelines and regulations. All animal experiments were reported in accordance with ARRIVE guidelines (Supplementary Material 2).

Results

Epidemic of human HFRS in Jiangxi

Distribution of HFRS cases

A total of 8981 HFRS cases were reported in Jiangxi between 2005 and 2021, with a case fatality rate (CFR) of 1.16% (104/8981). The median age was 47 years (Q1 = 34, Q3 = 58) with a sex ratio of 2.11:1 (male vs. female: 6093:2888). The occupation of patients was mainly farmers (5666), accounting for 63%, followed by students (720, 8%) and houseworkers (714, 8%).

Spatiotemporal distribution

Figure 2 was drawn based on the monthly morbidity time series, which showed that the prevalence of HFRS had distinct peaks in winter and summer every year (Fig. 2A). Furthermore, the periodogram showed two significantly higher spikes at approximate frequencies of 0.166 and 0.083 Hz, which means that HFRS morbidity existed at 6 months (1/0.166) and 12 months (1/0.083) on the time-series scale (Fig. 2B). Spatial autocorrelation analysis found that the average annual morbidity of HFRS was not randomly distributed in Jiangxi from 2005 to 2021 (Moran’s I = 0.327, P = 0.002), with a “High-High” clustering area around Gaoan County and a main “Low-Low” clustering area in southern Jiangxi (Fig. 3).

Figure 2
figure 2

Monthly distribution and the periodogram of HFRS morbidity in Jiangxi from 2005 to 2021.

Full size image
Figure 3
figure 3

The local indicators of spatial association (LISA) cluster map of HFRS morbidity in Jiangxi Province from 2005 to 2021. The map was generated by ArcGIS software (Version 10.4 ESRI, Redlands, CA, USA, https://www.esri.com/sofware/arcgis/arcgis-for-desktop).

Full size image

Trend in HFRS morbidity

The trend of HFRS morbidity reported in Jiangxi demonstrated four stages, but the APC in each stage was not significant, and the annual incidence rate was stable at 0.8~1.6/100000 (AAPC = −1.4, 95% CI = −9.1 to −6.9, P =0.729) (Table 1, Fig. 1 in Supplementary Materials 1). The changes in HFRS morbidity trends among three age groups (0~15 yrs, 16~60 yrs and ~61 yrs and older) in high- and low-prevalence areas were further explored (an average annual morbidity greater than 5/100000 was defined as a high-prevalence area). As Table 1 and Fig. 4 show, the HFRS morbidity in the 0~15 yrs group significantly increased by an average of 6% per year from 2005 to 2021 (APC=6.0, P <0.05) in the low prevalence area and by 12% per year from 2007 to 2018 (APC=12.0, P <0.05) in the high prevalence area. The HFRS morbidity at 16~60 yrs (HFRS vaccine-targeted population) showed a significant increase of 11% per year from 2008 to 2015 (APC=11.0, P <0.05) in the low prevalence area, while no significant change was found in the high prevalence area. For ~61 yrs and older, the HFRS morbidity markedly increased by 11.8% per year from 2005 to 2016 in the low prevalence area (APC=11.8, P <0.05) and by 18.3% per year from 2005 to 2013 in the high prevalence area (APC=18.3, P <0.05).

Table 1 Morbidity trend by age according to the join points identified by the analysis.
Full size table
Figure 4
figure 4

The trends in HFRS morbidity among the three age groups in high- and low-prevalence areas. A represents low prevalence areas and B represents high prevalence areas; 1, 2 and 3 represent 0 ~ 15 yrs, 16 ~ 60 yrs and ~ 61 yrs and older, respectively.

Full size image

Relationship between HV prevalence in rodents and the HFRS epidemic in humans

Rodent HV prevalence

A total of 1382200 mousetraps were set out in Gaoan County, one of the monitoring sites in China, and 3933 rats were captured, with an average rat density of 2.80% and an average annual antigen-positive rate of 2.01%. The average HV antigen-positive rate of rodents was 2.24% in fields, and the main species of rodents were Rattus losea (915/1957, 47%) and Apodemus agrarius (873/1957, 45%). The average HV antigen-positive rate of rodents was 1.79% in residential areas, and the main species of rodents were Rattus norvegicus (1679/1976, 85%) and Mus musculus (255/1957, 13%). Based on the rat density and antigen-positive rate, we calculated the index of rat with virus (IRV) to reflect the HV prevalence intensity among rodents. In general, the IRV showed a significant upwards trend from 2005 to 2013 and then remained relatively stable at approximately 2.5 in the next 8 years (Tables 1 and 2 in Supplementary Material 1 and Fig. 2 in Supplementary Material 1).

Figure 5
figure 5

Correlation between human and rodent hantavirus from 2005 to 2021 in the counties surrounding Gaoan County. (A) shows the overall counties surrounding Gaoan County. (B, D, E and F) show the overall, 0 ~ 15, 16 ~ 60 and ~ 61 yrs and older subgroups in the counties that started HFRS vaccination in 2009, respectively. C, G, H and I show the overall, 0 ~ 15, 16 ~ 60 and ~ 61 yrs and older subgroups in the counties that started HFRS vaccination in 2013. The blue line is the HFRS morbidity fitted curve by loess regression, and the yellow line is the HFRS morbidity fitted curve by loess regression; the dots indicated the HFRS morbidity.

Full size image
Table 2 Epidemiological correlation between human and rodent hantavirus from 2005 to 2019 in the surrounding counties of Gaoan.
Full size table

Relationship between IRV and HFRS morbidity

Due to the strict nonpharmaceutical interventions for coronavirus disease 2019 (COVID-19), the morbidity of many infectious diseases has declined from 2020 to 2021, including HFRS38. The reasons for this finding could be attributed to three potential factors: first, due to the tremendous number of young people leaving villages for better work, elderly people and children must engage in more agricultural activities that increase their exposure risk39; second, accelerated urbanization changed the rodent living environment, which promoted contact between humans and rodents and accelerated the spread of HV40; and finally, HFRS immunization was only carried out among people aged 16–60 years in Jiangxi41. As this paper found, although the prevalence of HV in rodents increased or remained stable from 2009 to 2019, the morbidity among vaccination-targeted people remained at a certain level or decreased, which reflects the effectiveness of the vaccine from the perspective of ecological epidemiology.

Jiangxi is a mixed epidemic area containing two dominant specific hosts of A. agrarius and R. norvegicus with two coprevalent kinds of HV (HTNV and SEOV), which is a possible reason for a bimodal distribution of HFRS morbidity40. HTNV in Jiangxi was highly homologous and formed a relatively independent branch that was not closely related to HTNV viruses at home and abroad, suggesting that HTNV in Jiangxi may be a new subtype and may have geographical limitations. Unlike HTNV, there were multiple branches of SEOV in Jiangxi, and these genetic relationships were closely related to Guangdong, Hubei and Zhejiang provinces around Jiangxi, indicating the diversity of HV and the dynamics of virus variation in Jiangxi. This phenomenon may be related to the fact that the main host of SEOV, R. norvegicus, a domesticated rodent, is more likely to migrate with human activities42.

The results of this paper indicated that the Immunization Expanded Program played a certain role in curbing the incidence of HFRS in Jiangxi from the perspective of ecological epidemiology; however, the vaccine strains Z10 (HTNV) and Z37 (SEOV) were isolated in the 1980s from Zhejiang Province43, and their protective effect against current strains in Jiangxi still needs further monitoring and analysis. The HTNV strains prevalent in Jiangxi were in an independent branch, and the genetic distance was relatively far from Z10 with approximately 3% amino acid heterogeneity. Thus, the protective effect of the vaccine strain Z10 used in Jiangxi needs further evaluation from serological and epidemiological perspectives.

The current study has several limitations. First, due to the lack of a specific research design, this paper did not find direct evidence that vaccines could reduce the prevalence of HFRS. Second, this study did not distinguish HTNV strain-type and SEOV strain-type in human cases, so the direct relationship between the HV epidemic and HFRS morbidity needs further research with laboratory evidence. Last, although we observed that the S and M sequences of HTNV in Jiangxi were in an independent branch compared to other regions, we could not determine that this distinction between the sequences of Jiangxi and other regions was caused by temporal or spatial factors because the collection time of many of the reference sequences was unknown and the compared fragments of HV sequences collected after 2010 were not matched in this study.

Conclusion

The prevalence of HFRS in Jiangxi has obvious temporal periodicity and spatial clustering. Morbidity among children and elderly individuals had a rapid upwards trend with the increasing prevalence of HV in rodents, indicating that it is urgent to accelerate the implementation of expanded immunization in this population. HTNV isolated in humans or rodents in Jiangxi is relatively genetically independent, while SEOV in Jiangxi presents much more diversity. Although the amino acid homology of both HTNV and SEOV prevalent in Jiangxi with the vaccine strains remained relatively high, HV variation still needs to be continuously monitored in the future to determine vaccine protective effectivity due to the dozens of amino acid site variations.