Association of Staphylococcus aureus colonization with food allergy occurs independently of eczema severity

Abstract

Background: Staphylococcus aureus has been implicated in the pathophysiology of eczema, allergic rhinitis, asthma, and food allergy. S aureus is a marker of more severe eczema, which is a risk factor for food sensitization/allergy. Therefore it might be that the association between S aureus and food allergy in eczematous patients is related to eczema severity.

Objective: We sought to investigate the association of S aureus colonization with specific IgE (sIgE) production to common food allergens and allergies in early childhood independent of eczema severity. We additionally determined the association of S aureus colonization with eczema severity and persistence.

Methods: In Learning Early About Peanut Allergy (LEAP) study participants eczema severity was assessed, and skin/nasal swabs were cultured for S aureus. Sensitization was identified by measuring sIgE levels. Peanut allergy was primarily determined by means of oral food challenge, and persistent egg allergy was primarily determined by using skin prick tests.

Results: Skin S aureus colonization was significantly associated with eczema severity across the LEAP study, whereas at 12 and 60 months of age, it was related to subsequent eczema deterioration. Skin S aureus colonization at any time point was associated with increased levels of hen's egg white and peanut sIgE independent of eczema severity. Participants with S aureus were more likely to have persistent egg allergy and peanut allergy at 60 and 72 months of age independent of eczema severity. All but one of the 9 LEAP study consumers with peanut allergy (9/312) were colonized at least once with S aureus.

Conclusion: S aureus, independent of eczema severity, is associated with food sensitization and allergy and can impair tolerance to foods. This could be an important consideration in future interventions aimed at inducing and maintaining tolerance to food allergens in eczematous infants.

Keywords: Food sensitization; Learning Early About Peanut Allergy; Staphylococcus aureus; atopic dermatitis; eczema; egg allergy; food allergy; microbiome; peanut allergy; prevention.

FIG 1.

Eczema severity by skin S aureus colonization at the preceding visit. Data are presented for all participants who were in the LEAP and LEAP-On studies with available SCORAD data for each study assessment time point divided into groups based on whether subjects had skin S aureus at the previous visit (in red) or did not have skin S aureus at the previous visit (in blue). Black diamonds represent model predicted means, boxes represent 25th and 75th percentiles, error bars represent 2.5th and 97.5th percentiles, and the middle line of the box represents the median. The total number of subjects contributing to the analysis at each time point, P values, mean differences, and 95% CIs around that difference directly above each assessment time point refer to the least squares mean difference (S aureus – no S aureus) and P value comparison between those who had skin S aureus at the previous visit and those who did not have skin S aureus at the previous visit by using a longitudinal repeated-measures model adjusted for SCORAD score at the previous visit, time, S aureus status at the previous visit, and the interaction between S aureus status at the previous visit and time.

FIG 2.

Peanut sIgE over time by skin S aureus colonization status. Data are presented for all participants who were in the LEAP and LEAP-On studies with available peanut sIgE data for each study assessment time point divided into groups based on whether subjects ever had skin S aureus from baseline to 60 months (in red) or never had skin S aureus from baseline to 60 months (in blue). Black diamonds represent model predicted means, boxes represent 25th and 75th percentiles, error bars represent 2.5th and 97.5th percentiles, and the middle line of the box represents the median. The total number of subjects contributing to the analysis at each time point, P values, mean differences, and 95% CIs around that mean difference directly above each assessment time point refer to the comparison between those who never have S aureus and those who ever have S aureus groups by using a longitudinal repeated-measures model adjusted for SCORAD, time, S aureus status, and the interaction between S aureus status and time. Average SCORAD scores at each time point are annotated directly below the box plots for those who ever had skin S aureus (red) and those who never had skin S aureus (blue).

FIG 3.

Relative distribution of hen’s egg white and peanut sIgE over time by skin S aureus colonization status. These figures show the relative distribution of hen’s egg white sIgE and peanut sIgE levels between those who ever have skin S aureus (shown in red) from 4 to 11 months to 60 months and those who never have skin S aureus (shown in blue). Vertical reference lines indicate where the distribution begins to significantly differ (P < .05) between the 2 groups by using bootstrap sampling of 1000 replicates of the upper percentiles, indicating that those with S aureus colonization are overrepresented at the higher end of the sIgE distribution (which is more indicative of allergy). A reference panel is included to illustrate the 67.8% of the trial participants who never had skin S aureus and the 32.2% who ever had skin S aureus and what a pattern with no association of skin S aureus with sIgE levels would look like.

FIG 4.

Peanut allergy in relation to skin S aureus colonization and treatment assignment percentages (from raw data), ORs, and 95% CIs from multiple multivariate logistic regression models by using the Firth penalized likelihood method are displayed. One model was fit for the 60-month data (outcome of interest being peanut allergy, as assessed by means of oral food challenge at 60 months), and another model was fit for the 72-month data (outcome of interest being peanut allergy as assessed by means of oral food challenge or the relevant diagnostic algorithm at 72 months). Predictors of interest included skin S aureus colonization status adjusted for SCORAD scores (at 60 and 72 months, respectively), LEAP study treatment assignment, and interaction between skin S aureus status and treatment assignment. A, For the plot, summary of the relationship between peanut allergy and skin S aureus colonization status (overall, within consumers, and within avoiders). In the Percent panel numerators refer to the number of subjects with peanut allergy, whereas the denominator refers to the number of subjects with skin S aureus (in red) and those without skin S aureus (blue). B, For the plot, summary of the relationship between peanut allergy and peanut consumption (overall, within those with skin S aureus, and within those without skin S aureus). In the Percent panel numerators refer to the number of subjects with peanut allergy, whereas the denominator refers to the number of subjects in the avoidance group (in gray) and those in the consumption group (green). Interpret results with caution because a small number of patients with peanut allergy (especially in the peanut consumption arm) contribute to these analyses.