Study	Reason for exclusion
Arshad 1992	Experimental intervention included dietary antigen avoidance in infant's diet, as well as the mother's, and also included application of acaricidal powder to reduce exposure to inhaled dust mite antigen.
Arshad 2007	Maternal antigen avoidance during lactation included, in the experimental group, a subgroup of breastfed that was combined with (and inseparable from) a subgroup of non‐breastfed infants fed hydrolyzed formula, and both subgroups also received interventions to reduce exposure to house dust mite: acaricide treatment of upholstery and carpets and plastic mattress covers.
Chandra 1986	Included in previous versions of this review. Now excluded because investigations by Chandra's university (Memorial University of Newfoundland) and by the Canadian Broadcasting Corporation strongly suggest that the published data may have been fabricated.
Chandra 1989	Included in previous versions of this review. Now excluded because investigations by Chandra's university (Memorial University of Newfoundland) and by the Canadian Broadcasting Corporation strongly suggest that the published data may have been fabricated.
Hattevig 1989	Experimental and control interventions allocated by site (2 different hospital clinics), not randomized or quasi‐randomized.
Vance 2004	Offspring atopic outcomes reported only in those with cord serum IgG levels obtained (136 of the 229 randomized), and numbers in each treatment arm not provided.
Zeiger 1989	Experimental intervention included dietary antigen avoidance in infant's diet, as well as the mother's. In addition, the trial suffered from large and grossly unequal losses to follow up due to non‐compliance with dietary restriction.