A sweeping analysis of nearly three million children pinpoints which early-life factors most strongly predict food allergy risk, revealing how skin barrier health, family history, and early exposures intersect to shape outcomes.
Study: Risk Factors for the Development of Food Allergy in Infants and Children. Image credit: Alona Siniehina/Shutterstock.com
A systematic review and meta-analysis of 190 studies was recently published in JAMA Pediatrics, aiming to identify the most strongly associated risk factors. Food allergy in children can affect nutritional status, growth, and development; however, this large global review focused specifically on its incidence and early-life predictors.
Early-life predictors of food allergy remain unclear
Uncertainty remains about the incidence and risk factors of food allergies in early life. This condition affects over 33 million people in the US. The most common type is mediated by high immunoglobulin (Ig) E levels and occurs early in life. It is typically lifelong and may cause life-threatening anaphylaxis reactions.
Identifying risk factors could help shape targeted preventive strategies based on modifiable factors. The current study sought to systematically assess and integrate food allergy incidence and risk factors in early life.
Meta-analysis spans 190 studies and 40 countries
The review included 190 studies with 2.8 million participants across 40 countries. Incidence estimates were derived from all studies that confirmed food allergy by food challenge. Risk factor analyses focused on various types of study, including case-control and cross-sectional studies of children below the age of six years, and included studies using both food-challenge-confirmed and other physician- or history-based definitions of food allergy.
The researchers performed random-effects meta-analyses to estimate pooled incidence rates and adjusted odds ratios, and also examined absolute risk differences when available. They also evaluated the certainty of evidence and the risk of bias in each included study.
Food allergy affects nearly 1 in 20 children
The likely pooled incidence of food allergy was 4.7%, based solely on studies that used food challenges. However, Australia had a ~10% incidence, vs ~1.8% in Africa.
The risk analysis included 176 studies, evaluating 342 risk factors for food allergy in infants and children. The level of certainty varied across factors. The following risk factors had the strongest associations and the highest certainty of evidence. These include major factors, linked to twofold or higher odds of food allergy, and minor factors, with smaller increases in odds.
Skin barrier dysfunction strongly predicts food allergy
Among the major risk factors related to allergic history, a prior history of wheezing or allergic rhinitis/conjunctivitis more than doubled or tripled the odds of food allergy, respectively, patterns consistent with the atopic march or diathesis. Atopic dermatitis in the first year of life was associated with a fourfold increase in odds. Higher transepidermal water loss, reflecting impaired skin barrier function, was associated with roughly a 3-fold increase in odds. Late introduction of solid foods, specifically peanuts, after 12 months of age was associated with more than twice the odds.
Regarding antibiotic exposure, systemic antibiotic use in the first month of life was associated with approximately fourfold higher odds of food allergy. Antibiotic exposure before birth was associated with about 30% higher odds, and exposure during the first year of life with about 40% higher odds.
Racial and migration patterns reveal complex influences
Social and genetic factors were also strongly associated. A family history of allergy doubled the odds when the mother or both parents had food allergy, increased the odds by more than twofold when siblings were affected, and raised the odds by about 70% when the father had food allergy. Parental migration prior to the child’s birth was associated with more than a 3-fold increase in odds.
In addition, children self-reported as Black had approximately fourfold higher odds compared with White children and about twofold higher odds compared with non-Hispanic White children, although the authors note that such associations likely reflect complex social and environmental influences rather than biological race alone.
Smaller but significant early-life predictors identified
Minor risk factors included gene variants in the filaggrin gene, which were associated with nearly doubled odds (OR ~1.93). Male sex was associated with 24% higher odds, being firstborn with 13% higher odds, and cesarean delivery with 16% higher odds. These increases were statistically significant but represented relatively small differences in absolute risk.
Several suspected risk factors lack strong support
Some potentially important factors lacked strong supporting evidence. These included specific eczema locations, social status, maternal medications and depression, farming background, certain metabolic and genetic markers, and infant gut microbial markers. Bias risk was high or probably high for one or more criteria in 125 studies, most commonly due to the use of non-standardized tools or methods to assess prognostic factors across participants.
Interestingly, birth weight, breastfeeding, and maternal stress during pregnancy were not associated with important differences in risk in pooled analyses.
This is the first review to systematically analyze hundreds of risk factors in relation to each other. This integrative approach clarifies associations that have not shown concordance across studies, such as male sex and family history, by applying consistent grading of evidence certainty across factors.
Importantly, instead of supporting competing food allergy hypotheses, the study supports the multifactorial origin of food allergy. For instance, the authors note that migration may alter the timing of exposure to allergens. This could aggravate atopic dermatitis and alter gut microbiome composition, potentially driving the development of food allergy, according to this model. However, the authors emphasize that these identified factors are predictive associations and may not necessarily be causal.
Its systematic, quantitative approach enabled a more uniform identification of populations at risk.
By filling current evidence gaps, the study may inform multiple stakeholders in food allergy prevention and management, helping to shape preventive strategies, clinical guidelines, and mechanistic research, while underscoring the need for well-designed birth cohorts with standardized adjustment for confounding.
Evidence supports a multifactorial allergy model
The study’s limitations primarily reflect those related to the existing literature. These include low-certainty evidence and the lack of uniform and/or comprehensive adjustment for confounders. This may limit risk inferences and the estimation of independent effects. These limitations emphasize the need to use birth cohorts with standardized adjustment for confounding factors.
Causality could not be established, and the studies included mostly those from developed countries, limiting their generalizability.
Study informs prevention without proving cause
The current study identified the major and minor risk factors that are associated with early-life food allergy. While these findings improve risk stratification and understanding of potential pathways, they do not establish causation. This could shape preventive strategies and future research.
Download your PDF copy now!
What happens to your body when you eat takeaway food too often?