Newborns with a strong immune reaction in week one had triple the eczema and food allergy rate at age 1

Taken at face value, 39 against 12 per cent is a ratio of 3.25, so the reported tripling is fair ^[1][1]. The absolute gap is 27 percentage points ^[2]. The harder question is how much of the eventual disease the first week accounts for. "What we ideally want is a weak seeding reaction," Brodin said ^[13]. With the cohort split roughly in half ^[6], about 76 per cent of the eczema and food allergy cases would have come from strong-reaction babies, and about 24 per cent from babies who already had the weak reaction he describes as the goal ^[3].

The conceptual claim is bigger than the percentages. "If you ask 100 immunologists, they would say that immune development begins at birth and continuously develops over time, without discrete phases," Brodin said ^[2]. His group argues that a response in the first week has an oversized effect on how the immune system develops afterwards ^[23]. "That's a novel concept," he said during the talk ^[4].

The team's explanation for so short a window is anatomical. "Healthy infants are born with a leaky gut that closes up after about a week," Brodin said ^[11]. While the gut is open, the researchers think, bacteria such as Bacteroides can reach the bloodstream and activate immune cells ^[10].

Strong-reaction babies carried substantially more Bacteroides in their gut ^[7]. "A high abundance of Bacteroides bacteria seems to drive seeding reactions to be very strong," Brodin said ^[8]. The data also allow the reverse reading: the bacterial levels could be a result of how severe the reaction was ^[9]. According to Brodin, babies pick up these bacteria through contact with their mother and other family members ^[21].

The feeding split, which the team describes as three immune trajectories among strong-reaction babies ^[14], is also an association with a proposed cause attached. In breastfed strong-reaction babies, high bifidobacteria went with half the rate seen at low levels ^[5]. Brodin's proposed reason is that bifidobacteria break down human milk oligosaccharides into short-chain fatty acids, anti-inflammatory molecules that can stop the immune system overreacting ^[17]. The team plans further work on why the other links emerged ^[18]. The account of the talk gives the subgroup results as percentages without the number of babies in each, and does not describe adjustment for other factors ^[22].

Every comparison in the work is between babies sorted by a reaction they already had, then followed through their first year ^[5][12]. Whether changing a baby's first week, through feeding or gut bacteria, would change that outcome is a question for a trial.

"This is a really important piece of work. Only a few studies have looked at immune development during the first week after birth," said Lucy Crouch at the University of Birmingham ^[19].