Independent and interactive effects of viral species on early-life lower respiratory tract illness

Abstract

Objectives: To determine the association between viral species and odds of severe lower respiratory tract illnesses (sLRI) versus upper respiratory illness (URI) among children under 2 years of age.

Methods: Infants (n=2061) enrolled in the Puerto Rican Infant Metagenomic and Epidemiologic Study of Respiratory Outcomes were surveilled for respiratory illnesses until age 2 years (March 2020 to April 2024). Nasal swabs from 1363 illnesses (774 participants) were screened for 21 pathogens.

Results: RSV infections occurred in 23% of sLRIs and increased odds of sLRI vs URI (OR=9.28; 95% CI, 5.43-15.85). Metapneumovirus, parainfluenza, and non-SARS-CoV-2 coronavirus also increased odds of sLRIs, while SARS-CoV-2 was associated with lower risk of sLRIs. Rhinovirus (43%) and bocavirus (16.1%) were commonly detected, but were not associated with sLRI risk. Co-infection with multiple viral species was associated with 2.92-fold greater odds of sLRI (95% CI, 2.05-4.16) compared to single viral species infections. Rhinovirus-bocavirus was the most common co-infection, and interaction between these viruses was associated with increased odds of sLRI.

Conclusions: Diverse viral pathogens drive early-life sLRIs. Some (e.g. RSV and metapneumovirus) have an intrinsic propensity to cause sLRIs, while other viruses' lower airway pathogenicity depends on other factors, including co-infection.

Keywords: Coinfection; Infants; Pediatrics; Respiratory syncytial virus infections; Respiratory tract infections; Rhinovirus.

Fig. 1.

Respiratory illness surveillance from March 3, 2020, to April 29, 2024. A) 30-day rolling incidence of upper respiratory illnesses (blue), mild lower respiratory illnesses (orange), and severe lower respiratory illnesses (red) in the PRIMERO cohort from March 3, 2020, to April 29, 2024. The number of illness events per 1000 person months was calculated as the sum of the number of illness events divided by the average number of participants surveilled in a window of 15 days before to 15 days after the date times 1000. The grey dashed line indicates the number of participants under active surveillance. The red shaded area indicates the first year of the COVID-19 pandemic (March 3, 2020, to March 1, 2021). B) Kaplan-Meier estimates and 95% confidence intervals for the probability of experiencing one or more URIs (blue) and LRIs (purple) from birth to 24 months of age for children born during the first year of the COVID-19 pandemic (solid lines) and children born after March 1, 2021 (dashed lines). P-values comparing curves between those born before and after March 1, 2021, are from the log-rank test.

Fig. 2.

Viral detections in early life respiratory illnesses. A) Rates of viral detection among URIs (blue), mild LRIs (orange) and severe LRIs (red). B) Detection of viral species per month among illnesses assessed in-person. The viral detection rate is calculated as the number of viral positives divided by the number of illnesses successfully tested for all viruses during that month.

Fig. 3.

Association of co-infection with illness severity. A) Frequency of negative, single and multiple viral detections in URIs (blue), mild LRIs (orange) and severe LRIs (red). Each bar is annotated with N (%). B) Forest plot of adjusted odds ratios comparing the odds of severe LRI vs URI between illnesses that test positive and negative for specific viral infections and co-infections. Blue dots represent the estimated aOR, and whiskers indicate the 95% confidence interval.