Clinical evidence linking allergy to increased risk of middle ear inflammation
From a clinical point of view, it is extremely important to distinguish among different phenotypes of otitis media, AOM, RAOM and OME. Clinical evidence from the literature seems to be in favor of an association between allergic rhinitis/atopic status and OME; whereas results about a hypothetical correlation with AOM and RAOM are contradictory and inconclusive because one of the major critical points is that some authors fail to distinguish among the different phenotypes of middle ear inflammation in children (articles included in the qualitative analyses are reported in table 2). Salah et al. did not find evidence linking atopy or allergy to AOM/RAOM26, whereas other authors10, 27, 28 demonstrated that children with AR had a significantly higher risk of otitis, even if they did not distinguish among different phenotypes of OM. Accordingly, authors29,30 demonstrated that atopic children may have a significantly increased risk of recurrent or persistent OM needing repeated tympanostomy tube insertions without distinguishing between AOM and OME. Finally, in a recent study authors demonstrated that RAOM was significantly associated with turbinate hypertrophy in children suffering from allergic rhinitis.31
On the other hands most authors seem to be in agreement with experimental data confirming also a significant clinical association between allergy and OME. Several manuscripts32-35suggested that AR may specifically influence the onset of OME. Authors demonstrated that AR prevalence was notably higher among OME patients than in the controls, with remarkable differences in eosinophil counts and serum IgE concentrations in the allergic rhinitis patients, compared to other OME patients.36 Cheng et al. in a metanalysis including cross-sectional and case-controlled studies confirmed that OME and AR are prevalent in pre-school and school-aged children and that allergy is a risk factor for OME.32Roditi et al. interestingly demonstrated that allergic rhinitis was significantly associated to OME even age was demonstrated as an effect modifier of the association; indeed a significant relationship was observed in children 6 years of age and older, whereas there was no significant association in younger children.37Furthermore, studies have documented thatAR patients have a higher risk of ET dysfunction, particularly during childhood, than non-allergic ones.27, 38Therefore, AR seems to be a common risk factor in children with OME and there is significant hearing loss compared to non-AR children with OME; in addition, the hearing levelimproved significantly with medical therapy.10
Recently, Torretta et al.39 demonstrated that allergy or atopy were significantly higher in otitis-prone children with OME than in those with simple RAOM: atopy was found in 73% of the children with both RAOM and OME versus 39.5% of those with simple RAOM, while allergy was found in 60% and 36%, respectively. The authors39 suggested that children with both RAOM and OMEare clinically different from children with simple RAOM, as they have a more complex clinical presentation that may be sustained by underlying allergy or atopy. Interestingly, in the same manuscript, children with both RAOM and OME were more frequently affected by adenoidal hypertrophy (AH) (p = 0.016). This is the second critical point that makes it difficult to analyze data in this regard, since ET dysfunction and AH are frequently associated with OME. Quaranta et al.40analyzed children with AH showing that 60% had OME and that AR was rarely present in this group. AH may constitute a concomitant clinical entity associated with allergy that exponentially promotes the development of OME and may be a confounding factor in the analysis. Accordingly, some authors demonstrated that adenoid tissue and middle ear mucosa share the same inflammatory pattern in allergic children affected by OME.33
Most relevant clinical articles33-37,39-42investigating specifically association between OME and allergy and included in the qualitative analyses are summarized in table 3.