Bacterial Signatures of Paediatric Respiratory Disease: An Individual Participant Data Meta-Analysis

Area of Science:

• Microbiology
• Pediatric Medicine
• Bioinformatics

Background:

• The airway microbiota's role in pediatric respiratory diseases is under investigation, with limited understanding of specific bacterial associations versus general disease links.
• Previous studies on respiratory microbiota are often small, hindering definitive conclusions about disease-specific or non-specific bacterial patterns.

Purpose of the Study:

• To investigate overarching patterns of bacterial association with acute and chronic pediatric respiratory diseases.
• To conduct an individual participant data (IPD) meta-analysis of respiratory microbiota using 16S rRNA gene sequences.
• To determine if specific bacteria are linked to particular diseases or if a general microbiota association exists.

Main Methods:

• An IPD meta-analysis was performed on raw microbiota data from published studies.
• Included cross-sectional analyses of pediatric (<18 years) microbiota in acute and chronic respiratory conditions (>10 cases).
• Utilized a uniform bioinformatics pipeline for sequence processing and employed diversity approaches, machine learning, and biomarker analyses.

Main Results:

• The analysis included 20 studies with data from 2624 children.
• Disease was associated with lower bacterial diversity in nasal and lower airway samples.
• Increased relative abundance of nasal taxa, including Streptococcus and Haemophilus, was observed in diseased children.

Conclusions:

• IPD meta-analysis identified a non-specific disease association in the respiratory microbiota across multiple pediatric conditions.
• Studying single diseases may obscure these broader, non-specific microbiota-disease links.
• Machine learning shows potential as a supplementary tool for clinical diagnosis in respiratory conditions, despite current limitations.