Temporal trends in RSV-specific antibody seroprevalence among hosptalized children during the COVID-19 pandemic in Japan: a retrospective cross-sectional study

Abstract

Respiratory syncytial virus (RSV) is a major cause of acute respiratory illness in young children. Given the transient nature of natural immunity, understanding age-specific seroprevalence is critical for optimising vaccine development and strategies. This study investigated the impact of the COVID-19 pandemic on RSV immunity in children under five (0–60 months), specifically aiming to determine the age with the lowest antibody prevalence and assess overall RSV seroprevalence during the pandemic years (2019–2021). Analysis of serum IgG concentrations revealed a significant decline in overall RSV IgG levels and seropositivity rates. Median IgG concentrations dropped from 912.94 mIU/mL in 2019 to 527.74 mIU/mL in 2021 (p < 0.001). A one-way ANOVA confirmed a significant effect of year on IgG levels (p < 0001), with post-hoc Tukey HSD tests showing significant differences between all annual pairs (2019 vs 2020: p < 0.0001; 2020 vs 2021: p = 0.0015). Age-stratified analysis consistently showed a U-shaped pattern in mean log IgG levels across all years: high in young infants (maternal antibodies), declining through early childhood, and increasing in older children. The IgG lowest point was delayed in 2021 (9–12 months) versus 2019 (5–8 months). Consistently lower IgG levels and seropositivity rates were observed in 2020 and 2021. Peak second quarter seropositivity dropped from 50.0–50.3% (2019–2020) to 25.0% (2021), and older children’s peak seropositivity in 2021 was 47.0% vs. 85.0% in 2019. These findings suggest that reduced RSV circulation during the COVID-19 pandemic may have yielded lower RSV-specific antibody levels among hospitalized children. While not representative of the general pediatric population, these results highlight the value of hospital-based serological monitoring to detect shifts in viral exposure and inform immunization strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-13100-4.

Introduction

Respiratory syncytial virus (RSV) is a leading cause of lower respiratory infections in aediatrics worldwide. It is responsible for an estimated 3 million hospitalisations and over 60,000 deaths annually, with nearly half of fatalities occurring in infants under six months of age [1]. In Japan, RSV accounted for 9711 reported cases in 2017 and 8509 in 2018, with approximately 25% of cases resulting in hospitalisation. This corresponds to a hospitalisation rate of 23.2 per 1000 person-years, rising to 35.4 per 1000 among infants aged 6 months [2]. Children under five years, especially infants, are particularly vulnerable due to their immature immune systems and lack of prior exposure. Severe RSV infections can result in bronchiolitis and pneumonia, often necessitating hospitalisation or intensive care [3–5]. Understanding the immune responses of young children to RSV is crucial for developing effective prevention and management strategies.

Immunoglobulin G (IgG) play a critical role in RSV-specific immune defence by neutralizing RSV and preventing severe disease [6]. Maternal IgG antibodies are passively transferred across the placenta during pregnancy, providing transient protection in early infancy. However, these antibodies wane over time, leaving infants increasingly susceptible to infection [7]. Measuring serum RSV-specific IgG levels therefore provides valuable insights into antibody dynamics during infancy and early childhood, reflecting both passive and active immune protection [8].

The emergence of the COVID-19 pandemic introduced unprecedented changes in the epidemiology of respiratory infections. Non-pharmaceutical interventions such as social distancing, mask mandates, school closures, and lockdowns led to a global reduction in the circulation of respiratory viruses, including RSV. Several countries reported suppressed RSV activity and altered seasonality during the pandemic period, followed by a resurgence of cases after restrictions were lifted, often with unusual timing and increased severity [9–12]. These shifts have raised concern regarding the interruption of typical viral exposure patterns and the resulting impact on population immunity especially in cohorts of young children with limited prior exposure [13].

Recent studies have emphasized the need to examine RSV-specific antibody trends before and during the pandemic to understand the immunological consequences of reduced viral circulation [10, 14–17]. Understanding these trends is essential for anticipating future outbreak risk and informing immunization strategies. Before the pandemic, RSV outbreaks followed predictable seasonal patterns, but these were disrupted by pandemic-related mitigation measures [15]. As social restrictions eased, the lack of prior exposure likely contributed to more severe and widespread RSV epidemics in several regions [18].

Notably, Japan offers a unique setting for seroepidemiological studies. Its consistently low infant mortality rate reflects high-quality pediatric care, a robust public health system, and minimal confounding factors like widespread infections or malnutrition. This context allows for a relatively pure assessment of RSV-specific humoral immunity, minimally influenced by background mortality or co-infections. Such data are invaluable for understanding the natural dynamics of immune waning and susceptibility during early childhood.

In this study, we conducted a retrospective cross-sectional seroepidemiological study to assess the impact of changing viral epidemiology during the COVID-19 pandemic on RSV immunity in hospitalized children aged 0–60 months admitted to Mie National Hospital, Japan, between 2019 and 2021. Our primary objectives were to characterize the temporal trends in RSV-specific IgG antibody levels across the pre-pandemic and pandemic periods, and to identify the age range with the lowest antibody concentrations in this vulnerable population. RSV-specific IgG antibodies were quantified using enzyme-linked immunosorbent assay (ELISA).

Material and methods

Study design

This retrospective cross-sectional study examined trends in RSV IgG antibody concentrations and seropositivity rates among hospitalized children aged 0–60 months. Serum samples were obtained from children admitted to Mie National Hospital (Tsu, Mie, Japan) between January 2019 and December 2021. The inclusion criteria were residency in Mie Prefecture, admission to Mie National Hospital during the study period, and availability of a stored serum sample collected at the time of admission. To avoid duplication, re-admissions were excluded, ensuring that each sample corresponded to a unique participant. This cohort reflects hospitalized children rather than the general pediatric population. However, because Mie National Hospital is the primary facility providing pediatric inpatient care for Tsu and its surrounding communities, these data offer valuable insight into antibody trends among children requiring medical attention, despite potential selection bias.

The study protocol was approved by the Institutional Review Board (IRB) of Mie National Hospital (Approval No: 2023.35). Informed consent was obtained from parents or legal guardians by using the opt-out approach by displaying study details on the hospital notice board and website, giving patients and their guardians the opportunity to refuse participation prior to sample collection. All methods were performed in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki and the Japanese Ethical Guidelines for Medical and Biological Research involving Human Subjects.

Data collection and sample processing

Demographic and birth-related characteristics, including sex, birth weight, gestational age at birth, and number of siblings, were extracted from electronic health records. Cross-sectional sampling was conducted over three years to minimise bias from seasonal RSV transmission and to provide an accurate average seroprevalence estimate. Each year was divided into four quarters, with 60 samples collected per quarter. These 60 samples were evenly distributed among 10 age groups, with 6 samples in each group. This design resulted in a total of 240 samples per year and 720 archived, randomly selected samples over the three years. Across the entire dataset, each age group was represented by 72 samples (24 collected per year). Within each quarter, samples were randomly selected from the available stored sera of eligible children to ensure even temporal representation. The sample size was determined using standard methods to achieve a seroprevalence estimate of 50% with a precision of ±14%, which was appropriate for exploratory analysis of relative temporal trends [19].

All hospital admissions were categorised by age: every two months up to 12 months, then in groups of 13 to <18 months, 19 to <24 months, 25 to <36 months, and 37 to <60 months. Patients were linked to serum samples taken at admission, and RSV antibody measurements were performed on these samples.

Expression and purification of trimerized secretory RSV pre-F (DS-Cav1)

The trimerized secretory RSV pre-F (DS-Cav1) protein was expressed and purified as follows;

The construct was derived from codon-optimized human RSV F protein (A2 strain) cDNA (Sino Biological, Inc., Cat. VG40042-UT). Key pre-fusion stabilizing mutations (DS-Cav1: S155C/S190F/V207L/S290C) [20] and three naturally occurring substitutions (P102A/I379V/M447V) [21] were introduced via PCR-based mutagenesis. To enable trimerization and purification, a Foldon domain (from T4 bacteriophage fibritin) [22] and a 6x histidine tag were added to the ectodomain (residues 1–513) of the F protein, yielding DS-Cav1-Foldon-6xHis. This final construct was subcloned into the pWPI-IRES-Puro Ak vector (Addgene, Watertown, MA, USA) using SmaI and SpeI sites. Lentiviruses expressing DS-Cav1-Foldon-6xHis were produced by transfecting 293T cells (ATCC) with pCMV-VSV-G and psPAX2 plasmids (Addgene #8454, #12260, respectively). Lentiviral transduction was followed by selection with 20 μg/ml puromycin in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and 100 U/ml penicillin G, and 10 μg/ml streptomycin (P/S) for one week. Secreted DS-Cav1-Foldon-6xHis protein was harvested from Opti-MEM (Thermofisher) containing 10% FBS and P/S, centrifuged, and stored at −80 °C.

Protein purification involved dialysis of the supernatant against 20 mM imidazole in buffer A (0.3 M NaCl, 10 mM Tris-HCl, pH 7.4) at 4 °C. The dialysate was loaded onto a Ni2+ chelating sepharose column (Cytiva). After washing with 20 mM imidazole, the protein was eluted with 500 mM imidazole in buffer A. Further purification was performed by gel filtration chromatography using a Superdex 200 10/30 column (Cytiva) equilibrated with 1 × PBS (-). Protein concentration was determined using the BCA method (Thermo Fisher Scientific). Protein purity, confirmed to be over 98% by 10% SDS-PAGE, is presented in Supplementary Fig. 2.

RSV IgG quantification in human serum

Serum samples were processed at the Mie National Hospital Clinical Research Laboratory, a collaborative research center. Detection of human anti-RSV pre-fusion (Pre-F) protein IgG antibody was performed using an in-house enzyme-linked immunosorbent assay (ELISA). This assay was developed by adapting the well-established Abcam traditional ELISA testing protocol.

For the ELISA, plates (ThermoFisher Scientific® - F96 MaxiSorp™ Nunc-immuno™ Plate) were coated with 100 ng/well of 1 µg/ml RSV Pre-F protein (dissolved in phosphate-buffered saline (PBS)). The coated plates were incubated overnight at 4 °C, followed by a 1.5-hour blocking step with 1% BlockAce (KAC, cat. #UKB80). Optimal dilutions for serum samples and the standard curve were established using the First International Standard for Antiserum to RSV (NIBSC code 16/284). The standard was serially diluted two-fold from 1:1000 to 1:128,000. Serum samples were diluted 1:4000 in 0.4% BlockAce. Both standards and samples were added in duplicate to ELISA plate wells and incubated at room temperature for 2 hours.

Plates were then washed eight times with 260 µl/well of 0.05% Tween 20 in PBS (PBS-T). A secondary antibody, polyclonal goat anti-human IgG H&L (HRP) (Abcam, cat. #ab97165), diluted 1:10000 in 0.4% BlockAce, was added at 100 µl/well to each well and incubated for 1 hour at room temperature. After eight washes with PBS-T, plates were incubated with a 1:1 mixture of KPL 3,3′,5,5′-tetramethylbenzidine (TMB) Peroxidase substrate and KPL Peroxidase Substrate Solution B (Seracare, cat. # 5120–0047) for 10 minutes in the dark at room temperature (RT). The colorimetric reaction was stopped with 2 N H₂SO₄. Optical densities (OD) were measured at 450 nm on a Tecan Spark Plate reader. IgG titers were quantified by interpolating the serum dilutions at OD 0.2 via 4-parameter logistic regression.

Statistical analysis

Statistical analyses were performed using R software (Version 2022.12.0 + 353). Continuous variables were summarised as mean ± SD or median (IQR), while categorical variables were presented as frequencies and percentages. Chi-squared tests were used to assess demographic differences across 2019, 2020, and 2021 for categorical variables, and one-way ANOVA and the Kruskal-Walli’s test were used for continuous variables with p-value of <0.05 considered statistically significant. Trends of log-transformed RSV IgG concentrations were analysed with one-way ANOVA and pairwise comparisons using Tukey’s HSD test.

To define data-driven seropositivity, a Gaussian Mixture Model (GMM) was applied to the 2019 log10-transformed IgG data using the Mclust function, selecting the optimal model components to be two based on the Bayesian Information Criterion (BIC) [23]. The 2019 baseline data was used to define an empirical cut-off, recognizing that antibody distributions and relatively normal cut-off may have shifted over time for other years in which COVID-19 was active. Subsequently, a two-component GMM was specifically fitted using mixtools, and its parameters were used to calculate the cutoff point as the intersection of the probability density functions of the two identified seronegative and seropositive components, establishing a seropositivity threshold of 3.07 Log m mIU/mL. Mean IgG concentrations and seropositivity rates were calculated by age group in quarterly analyses [24–26].

Results

Regional RSV epidemiological context

To contextualize the serological findings, we first examined weekly reported RSV cases in Mie Prefecture from 2019 to 2021 (Fig. 1). In 2019, RSV exhibited a typical seasonal pattern. In contrast, RSV activity was markedly reduced in 2020. In 2021, an out-of-season resurgence was observed, with case counts peaking during the second and third quarters. These temporal changes in community-level RSV circulation were considered in relation to subsequent analyses of pediatric antibody concentrations.

Fig. 1.

Weekly reported cases of RSV infection per sentinel site in Mie Prefecture, Japan, by year (2019–2021). The y-axis represents the number of reported cases per sentinel site, including hospital outpatient departments and community clinics, and the x-axis represents the epidemiological week of the year. The sharp resurgence in the second and third quarters of 2021 highlights the altered seasonality of RSV circulation following the implementation and subsequent relaxation of COVID-19 non-pharmaceutical interventions (NPIs). (Source: Mie Prefecture Infectious Disease Surveillance Centre, Japan. Data retrieved from https://www.kenkou.pref.mie.jp/disease_rs_virus_infection.html)

Study population characteristics

A total of 720 children were selected from the admission register and stratified into 10 age groups, each comprising 48 children. Participant demographics and birth characteristics for the years 2019 to 2021 are summarized in Table 1. Categorical variables were compared using Chi-squared test, and continuous variables were analysed using the Kruskal-Wallis test. A p-value of <0.05 was considered statistically significant.

Table 1.

Comparison of the characteristics of study participants among the three years

n	Group	2019	2020	2021	P value
		240	240	240
Sex (%)	Female	109 (45.4)	97 (40.4)	104 (43.3)	0.539
	Male	131 (54.6)	143 (59.6)	136 (56.7)
Weight at Birth (%)	<2500 g	13 (6.3)	30 (14.0)	24 (11.6)	0.063
	2500 g−4000 g	192 (93.2)	183 (85.1)	179 (86.5)
	≥4000 g	1 (0.5)	2 (0.9)	4 (1.9)
Gestational Age at Birth (%)	Post-term	0 (0.0)	2 (1.1)	0 (0.0)	0.278
	Preterm	11 (8.2)	14 (7.9)	21 (11.4)
	Term	123 (91.8)	162 (91.0)	164 (88.6)
Number of Siblings (%)	Zero	54 (37.0)	68 (36.8)	67 (36.6)	0.505
	One	60 (41.1)	75 (40.5)	72 (39.3)
	Two	25 (17.1)	32 (17.3)	28 (15.3)
	Three	7 (4.8)	6 (3.2)	8 (4.4)
	Four	0 (0.0)	4 (2.2)	4 (2.2)
	Five	0 (0.0)	0 (0.0)	3 (1.6)
	Six	0 (0.0)	0 (0.0)	1 (0.5)
Concentrations [median(min-max)]		912.94 [91.86, 153164.80]	764.70 [65.31, 31201.60]	527.74 [62.97, 26319.60]	<0.001

The sex distribution was relatively consistent across the three years, with a slightly higher proportion of males in each cohort (p = 0.539). The proportion of low-birth-weight infants (<2500 g) showed a non-significant upward trend in 2020 (14.0%) compared to 2019 (6.3%) and 2021 (11.6%) (p = 0.063). Gestational age at birth was comparable across years, with the majority of participants born at term (p = 0.278). The number of siblings also showed no significant variation (p = 0.505), with most children having one or no siblings.

Annual trends in RSV IgG levels

RSV-specific IgG levels declined progressively over the three-year period. Median concentrations were highest in 2019 (912.94 mIU/mL), followed by a decline in 2020 (764.70 mIU/mL), and reaching their lowest in 2021 (527.74 mIU/mL) (Fig. 2A). One-way ANOVA revealed a highly significant effect of year on median IgG concentrations (p = 5.3 × 10⁻⁹). Post-hoc Tukey HSD tests confirmed significant pairwise differences between 2019 and 2020 (p = 1.2 × 10⁻⁷), 2019 and 2021 (p = 0.014), and 2020 and 2021 (p = 0.0015). Although inter-individual variability was observed, the overall trend indicates a substantial reduction in RSV-specific IgG levels, with potential implications for population immunity.

Fig. 2.

(A) Distribution of RSV IgG levels across three years (2019–2021) this box plot illustrates the distribution of log-transformed RSV IgG concentrations across 2019, 2020, and 2021. Median IgG levels declined over time, with the highest levels in 2019, followed by 2020, and the lowest in 2021. An ANOVA test revealed a significant difference in IgG levels across the years (p < 0.0001). Pairwise comparisons were also significant: 2019 vs. 2020 (p < 0001), 2019 vs. 2021 (p = 0.014), and 2020 vs. 2021 (p = 0.0015). Outliers are represented as individual points. (B) IgG levels by age group across three Years. The line graph illustrates the mean log-transformed RSV IgG concentrations across different age groups (x-axis) for three years: 2019 (red), 2020 (green), and 2021 (blue). The y-axis represents the mean log IgG concentration (log10 mIU/mL). IgG levels tend to be highest in the youngest age group (0–2 months) and decline steadily until around 7–10 months. A gradual increase was observed beyond 12 months, with the highest levels occurring in older age groups (37–60 months), particularly in 2019. The 2020 and 2021 trends showed lower IgG levels in comparison to 2019, suggesting a decreasing trend in RSV IgG levels over time. (C) Seasonal (Quarters of the Year) variation in mean RSV IgG concentrations by year. Line graph showing the mean log-transformed RSV IgG concentrations (log₁₀ mIu/ml) of participants across the four quarters of the year for 2019 (red), 2020 (green), and 2021 (blue). Error bars represent the standard error of the mean. In 2019, RSV IgG levels peaked in the second quarter and dipped in the third, whereas 2020 and 2021 showed relatively stable but lower concentrations throughout the year. These trends suggest potential seasonal and inter-annual differences in RSV transmission dynamics

Age- and year-specific trends in mean RSV IgG concentrations

Figure 2B shows the mean log-transformed RSV IgG concentrations (log10 mIU/mL) across age groups for 2019–2021. In all years, infants aged 0–10 months exhibited a decline in mean IgG levels, consistent with the waning of maternally acquired antibodies. The lowest concentrations were observed between 5 and 6 and 9–10 months. A rise in IgG levels typically occurred after 11–12 months, with 2019 showing an earlier and more robust increase than 2020 or 2021.

Beyond 12 months, mean log IgG concentrations steadily increased with age, reflecting cumulative exposure and active immunity development. However, clear inter-annual differences were evident: 2019 consistently exhibited the highest mean IgG levels, peaking at approximately 3.70 log10 mIU/mL in the 37–60-month age group. Levels in 2020 were moderately lower, while 2021 showed the lowest concentrations across nearly all age groups, especially beyond 5 months of age.

Quarterly trends in mean log RSV IgG concentrations

Temporal changes in mean log-transformed RSV-specific IgG levels by calendar quarter are presented in Fig. 2C. For this study, first quarter (Q1) to fourth quarter (Q4) refer to the calendar quarters: January–March, April–June, July–September, and October–December, respectively. In 2019, mean log RSV IgG concentrations peaked in the second quarter (~3.2 Log mIU/mL), coinciding with a rise in RSV case reports in Mie Prefecture (Fig. 1). In 2020, IgG levels declined steadily from Q1 to Q4, ranging between 2.85 and 3.0 Log mIU/mL, paralleling the suppressed RSV activity that year. In contrast, 2021 displayed the lowest mean RSV IgG concentrations across all quarters. The nadir was observed in second quarter (Q2) (~2.62 Log mIU/mL), followed by a gradual increase through Q4 (~2.85 Log mIU/mL), aligned with a resurgence in RSV cases (Fig. 1).

Gaussian mixture model of IgG concentration

Gaussian Mixture Model (GMM) analysis identified a two-component model as the optimal fit for the Log10-transformed IgG data, based on Bayesian Information Criterion (BIC) (see Supplementary Fig. 1). This model suggests the existence of two distinct subpopulations within the dataset.

As shown in Fig. 3A, overall distribution (black curve) deviated from unimodality, and the GMM identified two Gaussian components: a lower-concentration population (blue curve) with a mean of ~2.8 Log10 mIU/mL (SD), and a higher-concentration population (red curve) with a mean of ~3.5 Log10 mIU/mL (SD). The intersection of these two components, marked at 3.07 Log10 mIU/mL, was selected as the cutoff threshold for seropositivity classification. Dataset from 2019 was used for threshold determination due to its assumed minimal influence of the COVID pandemic on RSV cases and serology.

Fig. 3.

(A) Gaussian Mixture Model of Log10 concentration data. This figure shows how a Gaussian Mixture Model (GMM) decomposes the Log10 concentration data into two Gaussian components. The black curve represents the overall observed distribution, while the blue and red curves represent the first and second Gaussian components, respectively. The vertical dotted line at 3.07 Log10 concentration indicates a potential boundary between the two subpopulations. The Y-axis represents probability density, and the X-axis represents Log10 concentration values. (B) Age-dependent trends in mean log IgG concentrations (log10 mIU/mL) from 2019 to 2021. This figure illustrates the mean log IgG concentrations (log10 mIu/ml) across various paediatric age groups (0–60 months) for the years 2019, 2020, and 2021. The graph demonstrates a U-shaped pattern in IgG levels across all three years, indicating high levels in young infants (0–2 months), a subsequent decline during infancy and early childhood (3–4 to 11–12 months), and a gradual increase in older age groups (13–18 to 37–60 months). (C) Quarterly seropositivity rates from 2019 to 2021. This figure illustrates the quarterly seropositivity rates for 2019, 2020, and 2021, demonstrating the temporal dynamics of seroprevalence. Seropositivity rates fluctuated across the three years, with a general trend of higher rates in 2019 and 2020, particularly in the second quarter. In contrast, 2021 exhibited consistently lower seropositivity rates across all quarters

Age-specific RSV seropositivity rates (2019–2021)

Seropositivity rates by age groups (0–60 months) for each year are depicted in Fig. 3B. All years followed a U-shaped curve: higher seropositivity in neonates, a decline in infancy, and a gradual rise with age.

In 2019, the seropositivity rate began at ~46% in the 0–2 months group, dropped to ~15% at 5–8 months, and increased steadily to ~85% in the 37–60 months group. Similar patterns were observed in 2020, starting at ~60%, dipping to ~15% and rising to ~82%. In contrast, 2021 exhibited lower rates across all age groups, starting at ~47%, dipping to ~8% at 9–10 months, and peaking at only ~47% in the oldest age group.

Quarterly seropositivity rates

Figure 3C illustrates quarterly RSV seropositivity rates across the three years. In 2019, seropositivity started increased from 38.3% in Q1 to 50% in Q2, declined to 40% in third quarter (Q3), and rose slightly to 45% in the Q4. In 2020, similar fluctuations were observed: 40.0% in Q1, peaking at 50.3% in Q2, followed by 38.3% and 28.3% in Q3 and Q4, respectively. In contrast, 2021 demonstrated uniformly lower rates, starting at 16.7% in Q1, and increasing gradually to 31.7% in Q4. The second quarter represented the peak for 2019 and 2020, while 2021 remained significantly lower across all quarters.

Discussion

This study aimed to determine the age at which RSV-specific antibodies were least prevalent and to assess trends in RSV-specific IgG concentrations and seropositivity among hospitalized children under five years of age before and during the COVID-19 pandemic (2019–2021). Our analysis of serum samples revealed a significant decline in both RSV-specific IgG concentrations and seropositivity rates, particularly in 2021. We also observed a delayed nadir in IgG levels during early childhood and attenuated seasonal peaks in seropositivity, findings that are consistent with changes in viral circulation during the pandemic.

Demographic characteristics, including sex, birth weight, gestational age, and number of siblings, remained largely consistent across the three years. While there was a slight increase in proportion of low-birth-weight infants in 2020, this difference was not statistically significant. The overall stability in baseline characteristics suggests that the observed changes in RSV IgG levels were unlikely to result from shifts in the underlying population structure.

A major finding of this study is the marked and statistically significant decline in overall RSV IgG antibody concentrations and seropositivity between 2019 and 2020. Median IgG concentrations fell from 912.94 mIU/mL in 2019 to 527.74 mIU/mL in 2021 (p < 0.001), supported by highly significant results from ANOVA and post hoc comparisons. These findings are consistent with global reports of reduced RSV activity during the pandemic [27]. Non-pharmaceutical interventions (NPIs) implemented to control SARS-CoV-2, such as social distancing, school closures, and improved hygiene practices likely suppressed the circulation of other respiratory viruses, including RSV, leading to decreased natural exposure and weakened immune stimulation among young children [15]. The temporal patterns of IgG concentrations in our study closely aligned with the prefectural surveillance data in Fig. 1, which shows a sharp resurgence in RSV cases in Mie Prefecture during 2021 following the near-total suppression of viral activity in 2020. The consistently lower IgG levels and seropositivity in 2021 reflect prolonged reductions in RSV exposure and a potential buildup of immunologically naïve individuals. This phenomenon, described as “immunity debt”, has been proposed as a contributing factor to the resurgence of severe RSV outbreaks in 2021 and beyond [28, 29]. Recent international evidence supports this trend. For example, a study in around ten European countries found a significant drop in RSV-specific antibody levels among children aged 3–4 to years when comparing pre-pandemic data to the post-pandemic period [30]. Although this study cannot establish causality, the temporal patterns of IgG concentrations observed in our study closely aligned with RSV surveillance data from Mie Prefecture, reinforcing the relationship between viral circulation in the community and the immunity of the clinical population.

In age-stratified analysis, a consistent U-shaped pattern in IgG levels was observed across all years. High concentrations in infants aged 0–2 months likely reflect passively acquired maternal antibodies [27, 31]. These antibody levels declined sharply through infancy, reaching a nadir before gradually increasing with age as a result of active immune acquisition through natural infection. Notably, the nadir in 2021 appeared later (9–12 months) compared to 2019 (5–8 months), suggesting a delayed onset of primary RSV infections due to reduced exposure during the pandemic. Even among older children (19–60 months), IgG concentrations in 2020 and 2021 remained consistently lower than in 2019, indicating persistent immunity gap across all age groups.

With respect to seasonal variation, IgG levels peaked in the second quarter of 2019, corresponding with increased RSV activity and typical seasonal circulation. In contrast, both 2020 and 2021 demonstrated lower and delayed IgG levels throughout the year. The nadir observed in the second quarter of 2021, before a surge in RSV cases later that year, suggests a link between reduced population immunity and heightened susceptibility to out-of-season outbreaks. These findings reflect a broader shift in RSV epidemiology during the pandemic, characterized by altered seasonality and reduced seroprevalence, a trend supported by other studies [16, 17, 32].

To classify seropositivity with greater precision, we employed a Gaussian Mixture Model (GMM) on log-transformed IgG concentration data. Our primary reliance on the Bayesian Information Criterion (BIC) with Mclust for model selection was due to its statistical consistency in identifying the ‘true’ number of underlying components. This aligned with our goal of distinguishing seronegative and seropositive populations, as BIC’s stronger penalty for complexity encourages more parsimonious and biologically interpretable models. This analysis identified two distinct subpopulations-one with lower mean antibody concentrations (~2.8 log mIU/mL) and another with higher mean levels (~3.5 log mIU/mL). The intersection point (~3.07 log mIU/mL) was used as an objective cutoff for seropositivity. This data-driven approach avoids reliance on arbitrary thresholds and provides a statistically robust framework, particularly relevant in populations with waning immunity or heterogeneous exposure histories.

Several limitations should be acknowledged. The study was limited to a single hospital in one geographic region of Japan and may not represent community-level antibody patterns. Its cross-sectional design prevents longitudinal tracking of individual immune trajectories. Although all serum samples were obtained from children admitted to Mie National Hospital, a precise diagnostic breakdown for each individual donor (for example, trauma or infectious disease) was not available in this retrospective dataset. As a result, we could not assess whether RSV antibody profiles differed by underlying reason for admission, which may limit the generalizability of our findings to specific clinical subgroups. Potential confounders such as use of RSV prophylaxis (palivizumab or nirsevimab) and maternal antibody status were not accounted for and could influence antibody levels. In addition, the sample size was designed for exploratory analysis and descriptive trend identification, rather than precise population-level estimation.

Despite these limitations, this study provides valuable insight into temporal changes in RSV antibody dynamics among hospitalized children during the pandemic. The findings underscore the importance of integrating serological monitoring with clinical and virological surveillance to understand the interplay between viral transmission and immunity. Longitudinal, multi-center studies are needed to track maternal antibody waning, timing of primary infection, and evolving transmission dynamics. Continued seroepidemiological monitoring during and after the pandemic will be essential to inform RSV immunization strategies and anticipate future outbreak patterns.

Conclusion

This study observed a decline in RSV-specific IgG levels and seropositivity rates among hospitalized children in Japan during the COVID-19 pandemic, suggesting reduced viral exposure in this period. The findings indicate shifts in RSV immunity patterns, including age-specific and seasonal variations, likely influenced by changes in viral transmission dynamics. These results underscore the importance of ongoing serological and epidemiological surveillance to better understand evolving RSV circulation and to guide future public health and immunisation strategies as global patterns continue to change.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

RSV

Respiratory syncytial virus

IgG

Immunoglobulin G

COVID-19

Coronavirus disease 2019

ELISA

Enzyme-linked immunosorbent assay

GMM

Gaussian Mixture Model

ANOVA

Analysis of variance

NPI

Non-pharmaceutical intervention

mIU/mL

Milli-international units per milliliter

IRB

Institutional Review Board

SD

Standard deviation

IQR

Interquartile range

Author contributions

E.D.K., M.N., J.X.A., K.S., and K.T. conceived the study. Data acquisition and analysis were performed by E.D.K., M.N., J.X.A., P.B.T., and A.S. Data curation was carried out by E.D.K. and N.O. The original draft was prepared by E.D.K. All authors contributed to the review and editing of the manuscript. J.O., M.F., and T.N. made DS-Cav1-Foldon-6xHis construct. S.Y. and N.O. lentivirally expressed and purified the trimerized secretory preF protein. Project administration was managed by M.N., K.S., H.K., and K.T. All authors have read and approved the final version of the manuscript.

Funding

This work was supported by the program on research and development (R&D) of new generation vaccine including new modality application from the SCARDA (Strategic Center of Biomedical Advanced Vaccine Research and Development for Preparedness and Response) of the Japan Agency for Medical Research and Development (AMED) (Grant No. JP233fa827011).

Data availability

The datasets generated during the current study are not publicly available due to data protection regulations, but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Institutional Review Board (IRB) of Mie National Hospital (Approval No: 2023.35). Informed consent was obtained from parents or legal guardians prior to sample collection.

Consent for publication

Not applicable.

Competing interests

J.O., N.O., M.F., and T.N. hold shares of stock in BioComo, Inc. E.D.K., M.N., A.S., J.O., S.Y., N.O., M.F., T.N., and K.T. are patent applicants for RSV preF ELISA. The other authors declare no competing interests. None of the funders influenced the study’s design, data collection, analysis, publication decision, or manuscript preparation.