Defining the window of opportunity and target populations to prevent peanut allergy

Abstract

Background: Peanut allergy affects 1% to 2% of European children. Early introduction of peanut into the diet reduces allergy in high-risk infants.

Objective: We aimed to determine the optimal target populations and timing of introduction of peanut products to prevent peanut allergy in the general population.

Methods: Data from the Enquiring About Tolerance (EAT; n = 1303; normal risk; 3-year follow-up; ISRCTN14254740) and Learning Early About Peanut Allergy study (LEAP; n = 640; high risk; 5-year follow-up; NCT00329784) randomized controlled trials plus the Peanut Allergy Sensitization (PAS; n = 194; low and very high risk; 5-year follow-up) observational study were used to model the intervention in a general population. Peanut allergy was defined by blinded peanut challenge or diagnostic skin prick test result.

Results: Targeting only the highest-risk infants with severe eczema reduced the population disease burden by only 4.6%. Greatest reductions in peanut allergy were seen when the intervention was targeted only to the larger but lower-risk groups. A 77% reduction in peanut allergy was estimated when peanut was introduced to the diet of all infants, at 4 months with eczema, and at 6 months without eczema. The estimated reduction in peanut allergy diminished with every month of delayed introduction. If introduction was delayed to 12 months, peanut allergy was only reduced by 33%.

Conclusions: The preventive benefit of early introduction of peanut products into the diet decreases as age at introduction increases. In countries where peanut allergy is a public health concern, health care professionals should help parents introduce peanut products into their infants' diet at 4 to 6 months of life.

Keywords: Peanut allergy; diet; early introduction; population; prevention.

Graphical abstract

Fig 1

Relationship between age at baseline and reported duration and eczema severity on the likelihood of peanut allergy at baseline in the first year of life. Bars represent prevalence of peanut allergy at baseline (raw data), defined by baseline oral food challenge or SPT wheal of >4 mm at screening, for participants in the LEAP screening cohort (7 LEAP RCT and 76 PAS group IV participants). Participants aged 4 to 11 months were assessed in the study at baseline and defined as low risk (all group I subjects, assumed to be tolerant), high risk and high risk sensitized (groups II and III from early introduction group, assessed by baseline peanut challenge), and likely allergy (group IV, assumed to be peanut allergic, with a peanut wheal of >4 mm) (Table E1). Those randomized to peanut avoidance (groups II and III) were omitted because they were not assessed for peanut allergy by oral food challenge at baseline. Proportion of infant peanut allergy by (A) tertile of age at screening (months) and (B) tertile of duration of eczema at screening (months); duration was the more important risk factor (Fig E7). The number with baseline peanut allergy is annotated above each bar, and the sample size is below each bar.

Fig 2

Trajectory of peanut wheal sizes of avoidance group participants allergic to peanut at the final assessment (n = 53; 36 months for EAT and 60 months for LEAP and PAS participants). Each line represents an allergic participant’s SPT values over the course of the study, starting with their age (in months) at baseline. SPT was not collected in the EAT avoidance group at 3 months; therefore, a distribution was imputed on the basis of the EAT early introduction group SPT distribution at baseline. Because 99% of SPT distribution at 3 months in the EAT early introduction group was between 0 and 1 mm, points were jittered within this interval so that lines could be connected between the 3-, 12-, and 36-month assessments. Participants with a >4 mm wheal at screening are identified with red lines (PAS group IV) and only had SPT data available at the screening visit and the 60-month visit. Orange lines represent EAT and LEAP allergic, avoidance group participants whose wheal sizes were >4 mm by their 12-month visit. Black lines represent allergic participants from the avoidance group whose wheal sizes were <4 mm by their 12-month visit. Assuming that participants with an SPT wheal of >4 mm are allergic to peanut,, , , approximately 60% of participants with peanut allergy at the end of the study were allergic at or before their 12-month visit based on wheal sizes of >4 mm. PA, Peanut allergy.

Fig 3

Relative reduction in burden of peanut allergy in a normalized population by age at introduction for (A) raw data from each study, (B) EAT-modeled effect plus whole-population model, and (C) whole-population model by eczema severity. All relative reductions estimate the treatment effect between early peanut introduction and avoidance. (A) EAT ITT and per-protocol (PP; restricted to only those exposed to the intervention) point estimates are displayed as red squares and are calculated as relative reductions between the standard introduction and early introduction arms. The blue points and blue smoothed regression line using a spline term for age shows relative reduction estimates from the raw high-risk LEAP screening population data (that is, LEAP + PAS, with imputed treatment effect among the PAS cohort, where the imputed benefit in PAS group IV was 0). (B) Red dashed line shows EAT-modeled estimates using the LEAP ITT treatment effect (Fig E4) applied at 3 months and 12 months. The whole-population (EAT + LEAP + PAS)-modeled ITT effect with bootstrapped 95% confidence intervals is shown in black and gray (see Fig E14 in the Online Repository at www.jacionline.org for sensitivity analyses). (C) The whole-population–modeled ITT effect is shown by eczema severity. Additional sensitivity analyses and modeling details relevant to these analyses are shown in the Online Repository (see Figs E12 and E13, and Tables E4, E5, and E6).