Meta-analysis identifies key risk factors for food allergy in young children

IMPORTANCE: The incidence and risk (predictive) factors for early life food allergy development remain uncertain. OBJECTIVE: To estimate the incidence and quantify risk factors for food allergy development. DATA SOURCES: MEDLINE and Embase were systematically searched to January 1, 2025. Data were analyzed from June 1, 2025, to November 25, 2025. STUDY SELECTION: Incidence estimates included studies confirming food allergy via food challenge. Risk factor analyses included cohort, case-control, and cross-sectional studies in any language assessing children younger than 6 years using multivariable analyses. DATA EXTRACTION AND SYNTHESIS: Paired reviewers independently extracted data. Random-effects meta-analyses pooled incidence and adjusted odds ratios (ORs). Risk of bias was assessed using the QUIPS tool, and certainty of evidence assessed using GRADE. MAIN OUTCOME AND MEASURE: The primary outcome was food allergy to age 6 years. RESULTS: A total of 190 studies involving 2.8 million participants across 40 countries were analyzed. Among studies using food challenge, overall food allergy incidence was likely 4.7% (moderate certainty). Among 176 studies identifying 342 risk factors with varying certainty, the strongest and most certain factors included prior allergic conditions (eg, atopic dermatitis [eczema] within the first year of life [OR, 3.88; risk difference [RD], 12.0%; 95% CI, 8.8%-15.7%], allergic rhinitis [OR, 3.39; RD, 10.1%; 95% CI, 6.7%-14.4%], and wheeze [OR, 2.11; RD, 5.0%; 95% CI, 2.1%-8.8%]), severity of atopic dermatitis (OR, 1.22; RD, 1.0%; 95% CI, 0.6%-1.6%), increased skin transepidermal water loss (OR, 3.36; RD, 10.0%; 95% CI, 6.3%-14.8%), filaggrin gene sequence variations (OR, 1.93; RD, 4.2%; 95% CI, 2.4%-6.4%), delayed solid food introduction (eg, peanut after age 12 months [OR, 2.55; RD, 6.8%; 95% CI, 1.9%-14.6%]), infant antibiotic use (first month [OR, 4.11; RD, 12.8%; 95% CI, 0.4%-40%], first year [OR, 1.39; RD, 1.8%; 95% CI, 0.8%-3.1%], during pregnancy [OR, 1.32; RD, 1.5%; 95% CI, 0.6%-2.5%]), male sex (OR, 1.24; RD, 1.1%; 95% CI, 0.7%-1.6%), firstborn child (OR, 1.13; RD, 0.6%; 95% CI, 0.3%-1.0%), family history of food allergy (eg, mother [OR, 1.98; RD, 4.4%; 95% CI, 2.5%-6.8%], father [OR, 1.69; RD, 3.2%; 95% CI, 1.3%-5.5%], both parents [OR, 2.07; RD, 4.8%; 95% CI, 1.3%-5.5%], siblings [OR, 2.36; RD, 6.0%; 95% CI, 4.4%-8.0%]), parental migration (OR, 3.28; RD, 9.7%; 95% CI, 4.9%-16.3%), self-identification as Black (vs White [OR, 3.93; RD, 12.1%; 95% CI, 5.2%-22.5%], vs non-Hispanic White [OR, 2.23; RD, 5.5%; 95% CI, 3.0%-8.7%]), and cesarean delivery (OR, 1.16; RD, 1.0%; 95% CI, 0.3%-1.2%). Factors like low birth weight, postterm birth, maternal diet, and stress during pregnancy showed no significant risk difference. CONCLUSIONS AND RELEVANCE: In this meta-analysis, the most credible risk factors associated with development of childhood food allergy are a combination of major and minor risk factors, including early allergic conditions (atopic march/diathesis), delayed allergen introduction, genetics, antibiotic exposure, demographic factors, and birth-related variables.