Methods: Random stool samples were obtained from 4-week-old infants with eczema (n=15) and without any allergy (n=15) at 9 months. Genomic DNA extracted by PowerSoil DNA Isolation Kit (MO BIO Laboratories) was sequenced using Ion PGM Seqeuncing 200 Kit v2, Ion 318 TM Chip v2 on Ion PGM System (Ion Torrent). Reads from each patient were filtered for low quality (Phred <20). Microbial diversity was evaluated using Shannon-Weaver diversity index in Swedish (J Allergy Clin Immunol 2012;129:434-40). The taxonomic classification of the reads was assigned by BLASTn.
Results: 5 controls had insufficient DNA for sequencing. No significant association was detected between eczema and any bacteria with ≥1% relative abundance, including Bacteroides, Escherichia, Klebsiella, Bifidobacterium, Streptococcus and Lactobacillus. Among the less abundant genera (relative abundance <1%), Campylobacter was more abundant in cases (median 0.008%, IQR 0.003-0.022%) than controls (median 0.001%, IQR 0.001-0.004%) while Roseburia was less abundant in eczema (median 0%, IQR 0-0.063%) than controls (median 0.055%, IQR 0.002-0.270%). Nonetheless, Shannon-Weaver diversity index of stool microbiota at 4 weeks was similar between infants with eczema and non-allergic controls at 9 months (median [IQR]: 1.28 [0.94-1.93] versus 1.47 [1.31-1.80]; P = 0.698). Comparing microbial compositions in our newborns and Swedish, Escherichia coli was found among top 5 genera only in both our cases and controls whereas enterobacter only in Swedish newborns. Clostridium, parabacteroides and lactobacillus were found only in Chinese eczema and healthy Swedish newborns.
Conclusions: Campylobacter and Roseburia appear to be less frequently detected in stool of 4-week-old Chinese infants who subsequently develop eczema. Microbial diversity is not associated with eczema susceptibility. This study confirms ethnic-specific early-life faecal microbial compositions.
Funding: Research Committee Group Research Scheme (3110087) and Direct Grant for Research (2013.2.033), CUHK