High Number of Early Respiratory Infections in Association with Allergic Sensitization to Mold promotes Childhood Asthma

To the Editor

The relationship between early life respiratory infections, aeroallergen sensitization and subsequent development of childhood asthma is unclear. To examine the effects of respiratory infections and allergic sensitization on asthma development, data from the Cincinnati Childhood Allergy and Air Pollution Study (CCAAPS) birth cohort were evaluated.

Details of the methods can be found at www.jacionline.org.¹. At annual visits at ages 1–4 and again at 7 years, CCAAPS children underwent skin prick testing (SPT) to 15 aeroallergens¹. High number of respiratory infections (HNRI) was defined as >6 infections during the 1^st year of life. At each age, sensitization was defined as ≥1 positive SPT; early sensitization was defined as ≥1 positive SPT at ages 1–3. Asthma was defined at age 7, as previously described². All children completed at least one visit between ages 1–4, and 81% were evaluated for asthma at age 7 (Table E1). Baseline characteristics were compared between the groups (See Table E2).

Children with a HNRI had a higher proportion of sensitization to ragweed (p=0.04) and mold (p=0.02) when evaluated at age 3, compared to children with a low number of respiratory infections (LNRI, Figure 1B–C). By 7 years, mold sensitization did not differ by respiratory infection number (Figure 1C). Even after adjustment, HNRI during the 1^st year was associated with mold (aOR 1.81, 95% CI 1.1–3.2) and ragweed (aOR 2.12, 95%CI 1.03–4.4) sensitization at age 3.

FIG 1.

Prevalence of aeroallergen sensitization according to the age and number of respiratory infections during the 1st year. A. There were no differences in any aeroallergen prevalence between children having HNRI versus low number of respiratory infections (LNRI). B. At three years, children with HNRI had a higher prevalence of ragweed and mold sensitization compared to children with LNRI.

Asthma was associated with HNRI (aOR 2.87, 95%CI 1.6–5.0, data not shown), and any aeroallergen sensitization (aOR 2.64, 95%CI 1.5–4.7, Table E3). Asthma was also associated with early mold (aOR 2.49, 95%CI 1.4–4.3), dust mite, dog and cat sensitization (Table E3).

Since mold was the only aeroallergen found to be associated with both HNRI and asthma (Figure 1C and Table E3), the combination of early mold sensitization and HNRI was evaluated with asthma risk. Children presenting with both HNRI and early mold sensitization had a markedly higher risk of asthma at age 7 (35% versus 4%, aOR 12.8, 95%CI 4.4–43.6, Figure 2A) compared to children with neither HNRI nor aeroallergen sensitization.

FIG 2.

A. Children with sensitization to mold (without HNRI) and children with HNRI (without aeroallergen sensitization) had an increased risk for asthma. Early mold sensitization and HNRI conferred highest asthma risk. B. Early mold sensitization with HNRI confers a higher asthma risk than children sensitized to other aeroallergen(s) without mold. Adjusted for pets, daycare, breastfeeding, birth season, family income, race and antibiotic use.

To determine that it was specifically mold, and not just aeroallergen sensitization, that combined with HNRI to increase asthma risk, we compared mold sensitized children (with or without other aeroallergen sensitization) to children sensitized to other aeroallergens but negative to mold. Indeed, the risk of asthma was highest in the group sensitized to mold with HNRI (35% versus 13%; aOR 3.8, 95%CI 1.5–10.1, Figure 2B) compared to children sensitized to other aeroallergen(s) but not mold and LNRI. While aeroallergen sensitization other than mold still combined with HNRI to increase asthma risk (23%), the effect size was 1.5-fold lower than those with mold sensitization (aOR 2.5, 95%CI 1.0–6.7, Figure 2B). We further evaluated children that were mono-sensitized to mold with HNRI and observed the same significant trend (data not shown).

To ensure that HNRI was not just a proxy for early wheeze, we also adjusted our final model for wheezing in the first year of life. We found that while wheeze was associated with asthma (OR 4.4, 95%CI 2.1–9.5), the HNRI+mold sensitization combination was still the strongest predictor (aOR 7.16, 95%CI 2.3–25.5, data not shown).

Discussion

Our study demonstrates that early sensitization to mold combined with HNRI in the first year of life increases asthma risk over 12 times those that are not sensitized without HNRI, suggesting a synergistic effect of mold sensitization and HNRI. These findings may represent an enhanced Th2 response associated with repeated respiratory infections in the setting of prior or concurrent mold exposure; consistent with our previous report that mold are potent immunomodulators and have powerful effects on asthma independent of their potential to act as antigens³.

Our findings that HNRI is associated with an increased risk for asthma at age 7 and aeroallergen sensitization are consistent with previous studies. Specifically, HNRI during the 1^st year of-life is consistently associated with an increased risk for asthma at age 7^4–6. Thus, the load and timing of viral exposure likely play a role in an individual’s progression to asthma. The association between early respiratory infections and aeroallergen sensitization has been previously reported⁷, however, mold sensitization was not specifically evaluated.

Interestingly, most of the children that were sensitized to mold lost the sensitization at 7 years of age. It may be possible that recurrent respiratory infections promotes transient mold sensitization. However, based on our previous published work, mold exposure has potent and long lasting immunomodulatory effects independent of sensitization³. Thus, the waning in mold SPT positivity does not mean that there are not ongoing immune responses to mold exposure. Although asthma was more strongly associated with having cat or dog sensitization, these allergens were not associated with HNRI. Among children sensitized to cat, 33% had a cat and among children sensitized to dog, 43% had a dog, but it is possible that public exposures may play a role.

Our study does have limitations. History of respiratory infections was collected by questionnaire, which could lead to recall bias and inability to evaluate the infectious etiology. Specific IgE to aeroallergens were not collected, which could have supplemented our data. Performing multiple testing for the association between HNRI and the individual aeroallergen sensitization may have introduced bias. Findings cannot be extrapolated to minority groups (other than of African American race) and infants who are not at risk for atopy. Lastly, the timing of viral infection and sensitization cannot be established, since SPTs were not performed prior to age 1.

In conclusion, both early sensitization to mold and HNRI confer individual risk for asthma at age 7, but in combination result in a higher risk suggesting a potentiation effect. Having a better understanding of this interaction may help to prevent the development of asthma later in life. Future mechanistic studies are needed to determine the exact mechanism by which respiratory infections and mold sensitization synergize to promote asthma development.

Abbreviations

CCAAPS

Cincinnati Childhood Allergy and Air Pollution Study

SPT

Skin prick test

HNRI

High number of respiratory infections

LNRI

Low number of respiratory infections

OR

Odds Ratio