Effects of maternal late-pregnancy and early-infancy exposures to respiratory syncytial virus circulating intensity on the risk for respiratory syncytial virus hospitalisation in the first two years of life: a comparative retrospective cohort study between Scotland and Singapore

Background
Although it is recognised that young infants could be protected against respiratory syncytial virus (RSV) infections by maternal RSV antibodies, such protective effects are not well assessed at the population level and are confounded by varied exposures to RSV circulation in the community during early infancy. In this study, we aimed to disentangle the role of late-pregnancy and early-infancy exposures in determining RSV hospitalisation risk in offspring by explicitly accounting for the background exposure to RSV circulation.
Methods
In this comparative retrospective cohort study, we leveraged RSV surveillance and hospitalisation data from Scotland and Singapore and constructed retrospective cohorts by birth months, with each birth month cohort as the observation unit. For each birth cohort, the outcome of interest was laboratory-confirmed RSV hospitalisation until the age of two years; the exposures of interest were RSV community incidence rates during the late-pregnancy period (defined as three months before birth) and during the early-infancy period (defined as three months after birth). In Scotland, we included hospital records from children born between Jan 1, 2017 and Nov 30, 2022 from the national health database available with Public Health Scotland. A quasi-Poisson model was used to assess the role of late-pregnancy exposure and early-infancy exposure in determining the risk for RSV hospitalisation for the age of 0–<6 months, 6–<12 months and 12–<24 months, separately, while accounting for the background RSV incidence rate in the community obtained from surveillance dashboard concerning viral respiratory diseases by Public Health Scotland. The analysis above was subsequently repeated in Singapore with medical data from children born between April 1, 2005 and June 30, 2015 at KK Women's and Children's Hospital. A random-effect meta-analysis was conducted to synthesise the incidence rate ratio (IRR) derived from the regression model above across the two countries.
Findings
A total of 13,838 hospitalisations from 71 birth month cohorts and 8,475 hospitalisations from 123 birth month cohorts were included in Scotland and Singapore, respectively. Increase in early-infancy exposure to RSV circulation consistently increased the risk for RSV hospitalisation in the first six months in Scotland and Singapore (pooled IRR: 2.47, per 10/100,000 increase in the RSV incidence rate in the community; 95% CI: 2.23–2.74). In the second year of life, there was borderline significance for the early-infancy RSV exposure in Scotland and Singapore (pooled IRR: 1.15, 95% CI: 1.01–1.31). While increase in late-pregnancy exposure to RSV circulation was consistently associated with decreased risk for RSV hospitalisation for the first three months of life in Scotland and Singapore (pooled IRR: 0.56, 95% CI: 0.46–0.69), for the subsequent three months (i.e., 3–<6 months), the decreased risk persisted only in Scotland (IRR: 0.19, 95% CI: 0.10–0.36) and not in Singapore (IRR: 1.03, 95% CI: 0.83–1.28). For both countries, there was no association between late-pregnancy exposure to RSV circulation and RSV hospitalisation beyond six months of life.
Interpretation
By differentiating the effect of maternal late-pregnancy RSV exposure from early-infancy exposure, this study evaluates the protective effect of maternal RSV exposure against RSV severe disease at the population level. While this study used population-level maternal late-pregnancy exposure as a proxy for maternal RSV antibody levels, the observed shorter duration of protection associated with late-pregnancy RSV exposure in tropical (Singapore) versus temperate (Scotland) regions indicates that the potential role of local RSV circulation patterns on this protective effect warrants further consideration. Our findings should be interpreted as hypothesis-generating and more serological studies based on antibody responses are needed to evaluate the effect of late-pregnancy RSV exposure on RSV hospitalisation in regions with different circulation patterns.