Diarrhea caused by viruses is a global problem among young children. We investigated two of the most important agents, rotavirus and adenovirus, to provide epidemiological evidence for a better understanding of their role among children with acute diarrhea. A total of 3147 hospitalized children were enrolled in the study during 2010 ~ 2014. Antigen testing for rotavirus and DNA testing for adenovirus were performed on stool specimens collected from participants. There were 1985 cases of community-acquired diarrhea (CAD) and 1162 cases of hospital-acquired diarrhea (HAD). A total of 692 cases (22.0 %) were positive for rotavirus. Rotavirus was detected in more children with HAD than in those with CAD (24.6 %; 286/1162 vs. 20.5 %; 406/1985). A total of 324 cases (10.3 %) were adenovirus positive. There was a significant difference between the CAD group and HAD group (9.5 %; 188/1985 vs. 11.7 %; 136/1162: χ 2 = 3.957, p = 0.047). Co-infection was found in only 35 children (1.11 %), and the co-infection rate was similar between the CAD and HAD groups (χ 2 = 1.174, p = 0.279). There was no association between sex and the detection rate of these viruses. The positive rate was significantly different for rotavirus among CAD cases (χ 2 = 27.979, p < 0.001) and for adenovirus (χ 2 = 34.362, p < 0.001) in the five age groups. Compared with the other four age groups (15.8–19.8 %), the prevalence of rotaviruses was highest among children aged 12–24 months (28.6 %). Adenovirus was detected in 3.6 % of neonates compared with 5.8 % of infants from 1 to 6 months old; this increased to 12.0–13.8 % in children over 6 months of age. In HAD cases, age differences were not found for rotavirus and adenovirus. Seasonal variation of rotavirus was observed, with peaks in November and December and with through in July and August; however, no clear seasonal pattern was found for adenovirus. Detection rates for rotavirus and adenovirus were significantly higher in children with HAD than those with CAD, but co-infection was very low. A high prevalence of rotavirus was identified in neonates with diarrhea. Vaccination for rotavirus gastroenteritis should be considered in neonates.

Human adenovirus serotype 31 (HAdV-31) is closely associated with gastroenteritis in children and can cause fatal systemic disseminated diseases in immunocompromised patients. The lack of genomic data for HAdV-31, especially in China, will greatly limit research on its prevention and control. Sequencing and bioinformatics analyses were performed for HAdV-31 strains from diarrheal children in Beijing, China, during 2010–2022. Three capsid protein genes (hexon, penton, and fiber) were obtained in 37 cases, including one in which the whole genome was sequenced. HAdV-31 strains clustered into three distinct clades (I–III) in a phylogenetic tree constructed based on concatenated genes and the whole genome; the endemic strains only gathered into clade II, and most of the reference strains clustered into clade I. Compared with penton and hexon, fiber had a faster evolutionary rate (1.32 × 10⁻⁴ substitutions/site/year), an earlier divergence time (1697), lower homology (98.32–100% at the amino acid level), and greater genetic variation (0.0032). Four out of the six predicted positive selection pressure codons were also in the knob of fiber. These results reveal the molecular evolution characteristics and variations of HAdV-31 in Beijing, and fiber may be one of the main evolution driving forces.

It has been reported that chronic infection of human norovirus (HuNoV) may potentially serve as a reservoir for viral variants with the possibility to evade population immunity or alter the binding sites of HBGA receptors. In this study, a child diagnosed with Burkitt lymphoma and positive for HuNoV determined by real-time PCR (qPCR) firstly in 15 August 2016, was followed up until 20 March 2018, and 26 fecal specimens and one vomitus were collected to trace the evolutionary characteristics of HuNoV by phylogenetic analysis, meta-genomics next-generation sequencing (mNGS), and temporal evolutionary analysis of VP1 among 23 specimens positive for HuNoV. There were 15 specimens with partial RdRp gene sequences forming an independent cluster with sequences of GII.P31, 14 with the region C sequences and 11 with P domain sequences of VP1 gene clustered together with HuNoV GII.4 Sydney_2012. All these sequences showed that mutations accumulated nearly in a time order, and more mutations were shown in the key epitopes A–E or near the binding sites for HBGA in subdomain P2 with higher evolutionary rates. Analysis of NGS data identified intra-host viral quasi-species, and two genome sequences of the same length from mNGS were assembled from N705, with mutations located in the region of subdomain P2 (1171 nt–1202 nt) which led to five amino acid mutations. In conclusion, the accumulated mutations of HuNoV, especially in subdomain P2, were explored in a child with Burkitt lymphoma, and the sequencing of HuNoV from immunocompromised individuals was proven critical for monitoring intra-host quasi-species evolution and potential variant emergence, providing basic data for clinical infection control.