Modelling bacterial transmission in human allergen-specific IgE sensitization

The impact of bacterial transmission from mother to child on human allergy development is poorly understood. The aim of the present work was therefore to use a temporal collected dataset of 117 mothers and their children to model the potential effect of mother-to-child bacterial transmission on allergy (IgE) sensitization.

We have recently shown a negative IgE correlation to high Escherichia coli levels until the age of 1 year, with a shift to positive correlation to high Bacteroides fragilis levels at the age of 2. In the present work, we used the previous published data to model the persistence and interaction effects of E. coli and B. fragilis with respect to IgE sensitization. Temporal modelling was made by first defining a stochastic model for sensitization state based on Markov chains and regression tree analyses. Subsequent simulations were used to determine the impact of mother-to-infant bacterial transmission. The regression tree analyses showed that E. coli colonization within 4 days was negatively correlated to sensitization, while lack of E. coli colonization at day 4 combined with B. fragilis colonization after 4 months was positively correlated. With Markov chain analyses, we found that E. coli was highly persistent in infants until the age of 4 months, while the persistence of B. fragilis increased with age.

Simulations showed that the mother's bacterial composition correlated significantly to the child's IgE sensitization state at the age of 2 years. High E. coli and low B. fragilis levels in the mother were negatively correlated, while low E. coli and high B. fragilis were positively correlated to IgE.

Our results support that allergy could partly be communicable, being transferred from mother to infant through the gut microbiota.