PURPOSE OF THE STUDY.
To investigate the prevalence and demographic risk factors of food allergy (FA) and its association with other atopic diseases in a population sample.
Data were collected from 10 348 adult and children older than 1 year, who represented the national population from 30 sites across the continental United States. Blood was collected and specific immunoglobulin E (IgE) panels were run for 79.3% of the subjects.
Specific IgE levels to peanut, cow's milk, egg white, and shrimp were collected from subjects aged 6 years and older. Shrimp-specific IgE was not tested for subjects younger than 6 years. Food sensitization was defined as having at least 1 food-specific serum IgE level at ≥0.35 kU/L. FA risk categories included unlikely FA (between ≥0.35 and 2 kU/L), likely FA (egg white: ≥7 kU/L, or ≥2 kU/L if ≤2 years old; milk: ≥15 kU/L, or ≥5 kU/L if ≤2 years old; peanut: ≥14 kU/L; and shrimp: ≥5 kU/L), and possible FA (between 2 kU/L and the likely FA threshold level for each food). Clinical FA rates were based on the sum of 50% of possible FA and 95% of likely FA.
Overall food sensitization was 16.8%. Milk and egg sensitization were highest (22% and 13.9%, respectively) in children aged 1 to 5 years. Peanut sensitization was highest in older children aged 6 to 19 years (10.7%) and young adults aged 20 to 39 years (8.7%). Shrimp sensitization did not vary with age. Overall prevalence of multiple sensitizations was 4.7%. The overall estimated clinical FA rate was 2.5% ([3.1% possible FA × 0.5] + 1.0% likely FA). The highest prevalence of clinical FA was in children aged 1 to 5 years (4.2%) and lowest in adults aged 60 years or older (1.3%). Clinical FA was 1.8% in children aged 1 to 5 years for milk, egg, and peanut. Peanut (2.7%) was the most common clinical FA in older children aged 6 to 19 years. Peanut and shrimp (range: 0.9%–1.2%) were the most common clinical FA in adults aged 20 to 59, and shrimp (0.7%) was the most common clinical FA in adults aged 60 years or older. Overall prevalence of multiple clinical FA was 1.3%. Clinical FA was more prevalent in younger subjects (P < .001), male subjects (P < .001), and non-Hispanic black subjects (P < .001). Household income and education level were not significantly associated with clinical FA. Subjects with doctor-diagnosed asthma were at a higher risk for likely FA; this risk increased with increased asthma persistence and severity and with an emergency department visit for asthma in the previous year. The odds of doctor-diagnosed hay fever were increased for those with possible FA. Eczema was not significantly increased for any FA risk group.
The estimated population prevalence of clinical FA was 2.5% and was associated with childhood, male gender, and non-Hispanic black race/ethnicity. Asthma and emergency department visits for asthma were associated with likely FA.
This is an important study that investigated the prevalence of clinical FA in children and adults in the same large population sample; it confirmed early observations of association of FA with childhood, non-Hispanic black race, and asthma. Use of objective data eliminated some limitations of previous survey studies. However, these data most likely underestimate clinical FA, because the study only accounted for 4 common allergenic foods, no clinical history is included (a small but significant percentage of patients with undetectable specific IgE might have clinical FA), and confirmation with oral food challenges was not performed. The next step will be to expand the number of foods investigated and develop protocols for confirming clinical FA in a large sample.
- Copyright © 2011 by the American Academy of Pediatrics