3.3 Allergic and immune status of children with SARS-CoV-2 infection
As shown in Table 2, the most common allergic disease was allergic rhinitis (36, 83.7%), followed by atopic dermatitis, food allergy, asthma and urticaria. Only one child, a 13-year-old boy, was previously diagnosed with asthma. Ten (23.3%) children had self-reported allergy to drugs, all of which were penicillin. Among allergic children, 9 (20.9%) had more than one kind of allergic diseases. Other few reported allergens included dust mites, eggs and mangos, each in one child.
In comparison of COVID-19-related measurements between allergic and non-allergic children, no statistical difference was found in the demographics, clinical features and disease course (Table 1). Blood cell counts and biochemical results were mostly similar between two groups, except that eosinophils counts, and serum levels of PCT, D-dimer and AST were lower in allergic group (p < 0.05), but these measurements were all in normal ranges, without clinical relevance (Table 3).
Serum total IgE level of allergic children was higher than that of non-allergic individuals (median, 46.30 vs 28.75, unit of IU/ml, p = 0.048), and allergic children had greater percentage of increased IgE (42.4% vs 29.3%, p = 0.155) without statistical difference. In general, all of the tested immunological parameters were not different between these two groups (p > 0.05), including immunoglobulins, complements and cytokines (Table 4).
The lymphocyte subsets were not found significantly different between the allergic and non-allergic COVID-19 children (Table 4), as well as between patients with and without pneumonia (Table 6). The results of correlation analysis between immunological parameters were displayed in Figure 3. There appeared three clusters of correlations. The first one is lymphocyte, total T cells, CD4+ and CD8+ T cells as well as B cells and NK cells. This correlation was relatively strong in all patients including the allergic group and pneumonia group. The second correlation was in total IgG, IgM, IgA, IgE, C3 and C4 levels. It was relatively stronger in pneumonia group whereas the correlations within this cluster decreased in allergy group. Immunoglobulin levels and lymphocyte subsets showed a negative correlation within all patients and pneumonia group. The third cluster was between the analyzed cytokines, which showed a stronger correlation between each other in pneumonia group. In addition, the duration of hospitalization was positively correlated with the time of RT-PCR conversion, but these two temporal indices had no significant correlation with immunological and inflammatory measurements, except for the slightly negative correlation between the duration of hospitalization and the level of IFN-γ in patients with pneumonia (r = -0.276) (Figure 4).