Abstract: 【Abstract】 Objective To investigate the effect of different delivery and feeding modes on intestinal microflora in infants with atopic dermatitis （AD）. Methods A total of 33 infants with AD were enrolled from Department of Dermatology, Wuhan NO.1 Hospital from July 2019 to December 2020, and 30 healthy infants were selected as control group. Then, all infants were grouped according to different delivery and feeding modes: cesarean-delivery AD group （22 cases）, cesarean-delivery control group （19 cases）, spontaneous-delivery AD group （11 cases）, and spontaneous-delivery control group （11 cases）; mixed-feeding AD group （13 cases）, mixed-feeding control group （11 cases）, formula milk powder-feeding AD group （12 cases）, formula milk powder-feeding control group （11 cases）, breastfeeding AD group （8 cases）, and breastfeeding control group （12 cases）. The total DNA was extracted from the infant feces, PCR was performed to amplify the V1 - V9 regions of bacterial 16S rRNA gene, and PacBio Sequel sequencer was used for high-throughput sequencing. Wilcoxon rank sum test was used to compare the bacterial community composition at genus and species levels, and correlations of relative abundance of differentially abundant bacterial taxa with eosinophil counts and SCORing Atopic Dermatitis （SCORAD） scores were analyzed. Results In the spontaneous-delivery control group, cesarean-delivery control group, spontaneous-delivery AD group, and cesarean-delivery AD group, the top 5 bacterial genera with high relative abundance were Bifidobacterium, Bacteroides, Veillonella, Streptococcus, and Escherichia. In the formula milk powder-feeding control group, breastfeeding control group, mixed-feeding control group, formula milk powder-feeding AD group, breastfeeding AD group, and mixed-feeding AD group, the top 5 abundant bacterial genera were Bifidobacterium, Bacteroides, Clostridium, Veillonella, and Escherichia. Linear discriminant analysis of effect size （LEfSe） showed no significant difference in the relative abundance of bacterial taxa among different delivery mode groups; among different feeding mode groups, Akkermansia and Akkermansiamuciniphila were the most differentially abundant microbes in the formula milk powder-feeding AD group at genus （LDA = 4.78） and species （LDA = 4.91） levels, respectively. The relative abundance of Akkermansia and Akkermansiamuciniphila （both 9.6% ± 0.72%） was significantly higher in the formula milk powder-feeding AD group than in the formula milk powder-feeding control group （both 2.50% ± 0.83%, Z = 1.66, P = 0.048）, the mixed-feeding AD group （both 0, Z = 2.26, P = 0.012） and the breastfeeding AD group （both 0, Z = 1.85, P = 0.032）. Additionally, the relative abundance of Akkermansia and Akkermansia-muciniphila was positively correlated with SCORAD scores in AD patients （ρ = 0.384, 0.387, respectively, both P < 0.05）. Conclusion Different delivery modes did not significantly affect the intestinal flora of AD or healthy infants, and the relative abundance of Akkermansia and Akkermansiamuciniphila increased in the formula milk powder-feeding infants with AD, which may be involved in the occurrence of AD.

Key words: Dermatitis, atopic, Infant, Intestinal flora, Delivery mode, Feeding way