Is the prevalence of peanut allergy increasing? A 5-year follow-up study in children in Montreal.

Background: Studies suggest that peanut allergy prevalence might be increasing, but these results have not yet been substantiated.
Objective: We conducted a follow-up study to determine whether peanut allergy prevalence in Montreal is increasing.
Methods: Questionnaires regarding peanut ingestion were administered to parents of children in randomly selected kindergarten through grade 3 classrooms between December 2000 and September 2002 and between October 2005 and December 2007. Respondents were stratified as (1) peanut tolerant, (2) never/rarely ingest peanut, (3) convincing history of peanut allergy, or (4) uncertain history of peanut allergy. Children in group 3 with positive skin prick test responses were considered to have peanut allergy. Children in groups 2 and 4 with positive skin prick test responses had peanut-specific IgE levels measured, and if the value was less than 15 kU/L, an oral peanut challenge was performed. Multiple imputation was used to generate prevalence estimates that incorporated respondents providing incomplete data and nonrespondents.
Results: Of 8,039 children surveyed in 2005-2007, 64.2% of parents responded. Among those providing complete data, the prevalence was 1.63% (95% CI, 1.30% to 2.02%) in 2005-2007 versus 1.50% (95% CI, 1.16% to 1.92%) in 2000-2002. After adjustment for missing data, the prevalence was 1.62% (95% credible interval, 1.31% to 1.98%) versus 1.34% (95% credible interval, 1.08% to 1.64%), respectively. The differences between the prevalences in 2005-2007 and 2000-2002 were 0.13% (95% credible interval, -0.38% to 0.63%) among those providing complete data and 0.28% (95% credible interval, -0.15% to 0.70%) after adjustment for missing data.
Conclusions: This is the first North American study to document temporal trends in peanut allergy prevalence by corroborating history with confirmatory tests. The results suggest a stable prevalence, but wide CIs preclude definitive conclusions.