Methods

Study design

The AVN is a registry that has been collecting cases of anaphylaxis since 2002, reported by allergists from French-speaking countries (specifically France, Belgium, and Luxembourg). Cases are recorded using a structured online questionnaire comprising more than 70 variables, including demographics data, relevant clinical information, results of allergy workup (skin prick tests, specific IgE and/or component-resolved diagnostics [CRD]). All submitted cases are reviewed and validated by an expert AVN panel prior to inclusion in the database.

The study complies with the European General Data Protection Regulation and the French national data protection framework. It was approved by the National Committee of Data and Patient Protection. All data were collected anonymously in accordance with French data protection regulations, and informed consent was obtained from all participants or their relatives.

Data extraction, variables and age-group definition

Demographic, clinical variables extracted include age, sex, causative allergen, food type, quantity consumed, cofactors involved, atopic comorbidities, and CRD, when available. Only anaphylaxis cases classified as ≥ grade 2 according to Ring modified by Behrendt classification were included.28 Eliciting doses (ED) are expressed in kitchen measurements as described by Faust et al.29 Year of declaration was collected for time-trend analyses for FIA cases declared to AVN. Analyses focused on FIA caused by plant proteins relevant to PBDs, including legumes, tree nuts, cereals, seeds and pseudocereals.23 These proteins are referred to as “plant allergens (PAs)”. Age at reaction was categorized as adolescents (13–19 years) and adults (>19 years), adapted from the World Health Organization definition and Grabenhenrich et al.30,31 Clinical phenotype analyses were limited to the most frequent plant allergen (PA) elicitors of FIA in adolescents and adults. As secondary analysis, the distribution of PA-related FIA was also collected for infants (<1 year) and children (1 to <13 years).

Statistical analysis

All statistical analyses were performed using R software (version 4.3.1). Qualitative data are presented as counts and percentages. Non-normally distributed quantitative data are expressed as median and interquartile range (IQR). Time trends in annual case counts were assessed using Poisson regression models to estimate the annual percent change (APC) and its 95% confidence intervals (CIs). To test whether the trend was constant over the study period, a segmented Poisson model was fitted with the breakpoint estimated from the data, and a change in slope was evaluated using the Davies test. All descriptive analyses are based on available data without imputation. Missing values were excluded from percentage calculations. Fisher's exact or chi-square test was used to compare categorical variables between groups and the Mann-Whitney U test was used to evaluate differences in continuous variables. Clinical and demographic characteristics were compared as separate, pre-specified analyses without adjustment. The Benjamini-Hochberg correction was applied to all per-allergen analyses (age group distribution and APC) and to CRD comparisons involving 3 or more categories. P value < 0.05 was considered statistically significant.