Figure 3.

Reduced steroid levels were observed in restricted analysis of ICS and OCS non-users. ICS/OCS use between birth to age 6 was converted to a binary variable according to ever use (TRUE) or never use (FALSE). Steroid metabolite associations with infection proneness were assessed within never users of these medications to eliminate potential confounding by their use. Models were adjusted for the same covariates as in the original models. Metabolite names marked with a superscript hash sign (#) denote metabolite identification at tier 2 level (matching Metabolon database, details in Methods).