Impact of allergic rhinitis on asthma: effects on bronchial hyperreactivity

Authors


G. Ciprandi
Semeiotica Medica I
Padiglione 3
A.O.U. San Martino
Largo R. Benzi 10
16132 Genoa
Italy

Abstract

Background:  Remarkable relationship exists between upper and lower airways. Bronchial hyperreactivity (BHR) is a paramount feature of asthma and may be considered a strong risk factor for the onset of asthma in patients with allergic rhinitis.

Objective:  This study is aimed at evaluating the presence of BHR in a large group of patients with moderate-severe persistent allergic rhinitis alone, and at investigating possible risk factors related to severe BHR.

Methods:  Three hundred and forty-two patients with moderate-severe persistent allergic rhinitis were prospectively and consecutively evaluated. Clinical examination, skin prick test, spirometry and bronchial methacholine (MCH) test were performed in all patients.

Results:  Twenty-two (6.4%) patients had severe BHR, 74 (21.6%) patients had mild BHR and 192 (56.2%) had borderline BHR; 54 (15.8%) patients had a negative MCH test. The logistic regression analysis evidenced that trees and house dust mites sensitization (ORAdj: 8.1), rhinitis duration > 5 years (ORAdj: 5.4) and FEV1 ≤ 86% of predicted (ORAdj: 4.0) were significantly associated with severe BHR. The discriminative ability of this model is appreciably satisfactory, being the AUC = 0.90.

Conclusion:  This study highlights the close link between upper and lower airways and the role of some risk factors, such as tree and mite sensitization, > 5-year duration, and ≤ 86% FEV1 values, as risk factors for severe BHR in patients with moderate-severe persistent allergic rhinitis alone. Therefore, BHR is frequently present in patients with chronic rhinitis and should be suspected in the presence of defined risk factors.

Allergic rhinitis is the most frequent IgE-mediated disease. Its prevalence is high in the general population and is still progressively increasing (1).

Allergen inhalation activates mast cells with consequent release of mediators, mainly histamine and leukotrienes, and cytokines capable of inducing the recruitment and activation of inflammatory cells, including eosinophils, neutrophils and Th2 lymphocytes (2). These inflammatory events lead to the onset of typical nasal symptoms, including itching, sneezing, rhinorrhea and obstruction. Clear relationship between nasal obstruction, Th2-dependent inflammation and nasal airflow limitation has been recently evidenced in patients suffering from allergic rhinitis (3). Moreover, nasal airflow limitation and inflammation are closely associated with bronchial function, mainly concerning FEV1 values, in patients with respiratory allergy (4). Indeed, numerous studies have demonstrated a close association between allergic rhinitis and asthma (5, 6). Moreover, allergic rhinitis has been demonstrated to be a strong risk factor for the onset of asthma in adults (7).

Asthma is characterized by a reversible airflow obstruction and forced expiratory volume/1 s (FEV1) is considered the gold standard to evaluate bronchial obstruction (8, 9). Nevertheless, there is increasing interest to consider the involvement of small airways in the pathogenesis of asthma (10). Even though there is no direct parameter able of assessing small airways, it has been originally considered that the forced expiratory flow at the 25% and 75% of the forced vital capacity (FEF25–75) might be more sensitive to obstruction in small airways than FEV1 (11). Moreover, it has been demonstrated that both FEV1 and mainly FEF25–75 are impaired in patients with allergic rhinitis and perceiving nasal symptoms alone (12). Moreover, FEF25–75 has been evidenced to be a reliable marker of early bronchial impairment in allergic rhinitis (13).

On the other hand, bronchial hyperreactivity (BHR) is a paramount feature of asthma. Moreover, BHR may be observed in a high proportion of rhinitics (14). In this regard, it has been hypothesized that a positive bronchial challenge to methacholine could be considered as predictive for those rhinitics would progress to develop asthma (15). BHR was reported both in patients with perennial allergic rhinitis and in subjects with seasonal allergic rhinitis (16). In addition, a seasonal variability in BHR was described in subjects sensitized to pollens (17). Very recently, we demonstrated that in a large cohort (> 2000) of patients with allergic rhinitis 70% of them had BHR (12).

Therefore, allergic rhinitis may be considered as the first step of a progression of respiratory allergy toward asthma. Indeed, it has been recently published the WHO document ‘the impact of allergic rhinitis on asthma’ (ARIA) that clearly underlined the role of allergic rhinitis as risk factor for asthma development (18). Moreover, this ARIA document revised the classification of allergic rhinitis defining two types: ‘intermittent’ and ‘persistent’; the new classification does not consider the type of allergen, but the duration (days/week and consecutive weeks) and the severity of symptoms (mild or moderate-severe).

Nevertheless the possible presence of BHR in patients with persistent allergic rhinitis has been poorly documented, and overall exploring possible associated risk factors. Therefore, the aims of the present study were: (1) to evaluate bronchial hyperreactivity in a large group of patients with moderate-severe persistent allergic rhinitis and (2) to investigate the possible role of independent predictors in the association with severe bronchial hyperreactivity.

Materials and methods

Study design

The study included patients with moderate-severe persistent allergic rhinitis. All of them were evaluated performing skin prick test, spirometry and methacholine bronchial challenge.

Subjects

Three hundred and forty-two patients with moderate-severe persistent allergic rhinitis, perceiving nasal symptoms alone and with normal FEV1 values (≥ 80% of predicted) were prospectively and consecutively evaluated. Demographic characteristics, including gender, age and duration of rhinitis (expressed in years), are reported in Table 1. All of them were Navy soldiers who had to refer to Navy Hospital for mandatory periodic visit for maintaining the qualification and an informed consent was obtained from each patient. A detailed clinical history was taken and a complete physical examination was performed. The patients were included in the study on the basis of a clinical history of persistent allergic rhinitis and presence of moderate-severe nasal symptoms according to validated criteria (18). We excluded all the subjects who met the following exclusion criteria: any prior history of asthma or presence of asthma symptoms, including cough, wheezing, breathlessness and shortness of breathing, acute or chronic upper respiratory infections, anatomical nasal disorders (i.e. nasal polyps, septum deviation, etc.), previous or current smoking, previous or current specific immunotherapy and use of nasal or oral corticosteroids, nasal or oral vasoconstrictors, antileukotrienes and antihistamines during the previous 4 weeks.

Table 1.   Demographic and clinical parameters of the patients divided in four groups according to their bronchial hyperreactivity level
 All patients N = 342PC20 < 1 mg/ml N = 22PC20 = 1–4 mg/ml = 74PC20 = 4–16 mg/ml N = 192PC20 > 16 mg/ml = 54P
  1. All percentages in round brackets are calculated over the total number of subjects reported at top of the column. All numbers in square brackets represent 1st–3rd quartiles.

  2. *Chi-square test.

  3. †Kruskal–Wallis test.

Gender – males, N (%)258 (75.4)20 (90.9)61 (82.4)141 (73.4)36 (66.7)0.059*
Age (years), median23 [21–24]24 [22–26]21 [21–23]23 [22–25]23 [21–24]< 0.0001†
Rhinitis duration (years), median2 [2–5]8.5 [3–10]5 [2–6]2 [2–3]2 [2–3]< 0.0001†
FVC% of predicted median101.0 [95–105]101.5 [86–102]105.0 [100–106]100.0 [97.5–104.5]103.0 [91–104]0.004†
FEV1% of predicted median91.0 [89–97]84.5 [83–86]94.5 [91–95]90.0 [89–95]98.0 [94–103]< 0.0001†
FEF25–75% of predicted median72.0 [71–76]63.0 [57–72]73.0 [71–73]72.0 [71–74]79.0 [72–100]< 0.0001†

All patients were treated only on demand with drugs alone. The diagnosis of persistent allergic rhinitis was made on the basis of a history of nasal symptoms and positive skin prick test according with validated criteria (18).

Skin prick test.  It was performed as stated by the European academy of allergy and clinical immunology (19). The panel consisted of: house dust mites (Dermatophagoides farinae and pteronyssinus), cat, dog, grasses mix, Compositae mix, Parietaria officinalis, birch, hazel, olive tree, Alternaria tenuis, Cladosporium, Aspergilli mix (Stallergenes, Milan, Italy).

Spirometry.  It was performed by using a computer-assisted spirometer (Pulmolab 435-spiro 235, Morgan, UK), with optoelectronic whirl flow meter. Spirometry was performed as stated by European Respiratory Society (8, 9).

Methacholine bronchial challenge.  It was performed to evaluate BHR only if basal FEV1 was equal or more than 80% of predicted. Aerosol is delivered using a dosimetric computerized supply (MEFAR MB3, Marcos, Italy). The test was performed following the American Thoracic Society guidelines for methacholine challenge (20). The threshold dose causing a 20% fall of FEV1 (PC20) was calculated.

Degree of BHR.  Three categories of BHR were considered on the basis of PC20: severe PC20 < 1 mg/ml, mild PC20 between 1 and 4 mg/ml, borderline PC20 ranging from 4 to 16 mg/ml, according with the criteria of the American Thoracic Society guidelines for methacholine challenge (20). Subjects without response to the cumulative dose of 16 mg/ml were considered having normal bronchial responsiveness.

Statistical analysis and data definitions

Descriptive statistics were firstly performed and quantitative parameters were reported as means and standard deviations (SD), or as medians with quartiles in case of skewed distribution. Qualitative data were reported as frequencies and percentages. Comparison of qualitative data among various groups of patients was made by the chi-square test (or by the Fisher’s exact test in case of expected frequencies less than five). Comparison of quantitative variables between the three groups of patients was made by means of the analysis of variance or by the nonparametric counterpart (Kruskal–Wallis test) whenever the normality assumption was not fulfilled; posthoc comparisons were performed using the Scheffé or the nonparametric Dunn’s test.

In order to evaluate the role of different independent explanatory variables in the relationship with severe BHR (PC20 < 1 mg/ml vs PC20 ≥ 1 mg/ml) a multiple logistic regression analysis was performed on patients with BHR only (N = 288). Variables that were statistically significant in the bivariate analysis or that were considered a priori important were entered in the model. Before performing logistic regression, some continuous predictors were dichotomized on the basis of the ROC curve analysis, and sensitivity, specificity and diagnostic odds ratios (DOR) were calculated (21). The following predictors were evaluated in the logistic regression model: rhinitis duration (>5years vs≤ 5years), FVC (≤104 vs >104% of predicted), FEV1 (≤86%vs≥ 86% of predicted), FEV25–75 (≤57%vs≥ 57% of predicted), type of allergen classified into three categories (trees and pollens/trees and house dust mites/trees absent). Age of the patient (>23 vs≤23years) did not enter in the model due to collinearity with rhinitis duration (>5 vs≤5years). Sensitization to trees and sensitization to house dust mites did not enter in the model due to collinearity with type of allergen (trees and pollens/trees and house dust mites/trees: absent). For the multivariate analysis, the step-down strategy was chosen; the effect was expressed in terms of adjusted odd ratio (ORAdj) and 95% confidence interval (CI) and was tested by means of the likelihood ratio test (LR test). The area under ROC curve (AUC) of the model was used as an indicator of the model’s predictive ability. All tests were two sided and a P value<0.05 was considered statistically significant. The package ‘statistica release 6’ (StatSoft Corp., Tulsa, OK, USA) was used for the bivariate analyses and the ‘stata release 7’ (Stata Corporation, College Station, TX, USA) for multivariate analyses.

Results

Three hundred and forty-two patients, 258 males (75.4%) and 84 females (24.6%) were included in the study. The mean age was 23.6years (SD: 4.5) with a minimum age of 18 and a maximum of 48years. Two hundred and twenty-eight patients (66.7%) had a diagnosis of rhinitis and 114 (33.3%) had a diagnosis of rhinoconjunctivitis. Patients were studied on the base of their MCH response: 54 patients (15.8%) turned out to be MCH negative, 192 patients were borderline (56.1%) and 96 patients (28.1%) were MCH positive. There were 22 (6.4%) patients with PC20< 1 mg/ml (severe), 74 (21.6%) with PC20 between 1 and 4 mg/ml, 192 (56.2%) with PC20 between 4 and 16 mg/ml and 54 (15.8%) with PC20 > 16 mg/ml (i.e. negative patients). There was a similar gender distribution in the 4 groups of patients (Table 1); disease duration was different in the four groups of patients: the more severe was the category of BHR, the longer was the rhinitis duration (Table 1, Fig. 1); the mild group of patients were significantly younger than the severe group of patients and were also younger with respect to the borderline group of patients (Table 1). Spirometric parameters were also significantly different in the four groups of patients; in particular FEV1 was significantly lower in the severe group of patients (Fig. 2).

Figure 1.

 Rhinitis duration in the four groups of patients according to their category of BHR.

Figure 2.

 FEV1 in the four groups of patients according to their category of BHR.

In order to evaluate possible risk factors for severe BHR, all patients presenting BHR (PC20 ≤ 16 mg/ml; N = 288) were more deeply investigated.

A ROC curve analysis was firstly performed in order to find, for the quantitative predictors, the best cut-off value discriminating between patients with severe BHR (PC20 < 1 mg/ml) from patients with BHR (PC20: 1–16 mg/ml). The results of this analysis are shown in Table 2. As shown in this table, three variables had a particularly interesting predictive ability (AUC ROC > 0.70): rhinitis duration, FEV1 and FEF25–75% of predicted. The variable FVC had a too high false positive rate (67%) and its ROC curve shape was atypical; also the false positive rate of the variable age was unsatisfactory (36%).

Table 2.   Sensitivity, specificity, positive (PPV) and negative predictive values (NPV), diagnostic odds ratios (ORdiagn), area under curve (AUC) and best cut-off values for different predictors of severe BHR (PC20 < 1 mg/ml), obtained by ROC curve analysis (N = 288)
PredictorsSensitivitySpecificityPPVNPVORdiagn (95% CI)AUC (95% CI)Best cut-off
Age (years)0.730.640.140.974.72 (1.79–12.47)0.65 (0.59–0.70)> 23 
Rhinitis duration (years)0.640.830.240.978.83 (3.49–22.31)0.79 (0.74–0.83)> 5  
FVC% of predicted0.950.330.110.9910.21 (1.35–77.13)0.59 (0.53–0.65)≤ 104
FEV1% of predicted0.820.810.260.9819.44 (6.30–59.95)0.76 (0.71–0.81)≤ 86 
FEF25–75% of predicted0.500.920.350.9612.30 (4.75–31.87)0.70 (0.64–0.75)≤ 57 

In Table 3, demographic and clinical parameters of the patients with severe BHR and with mild an borderline BHR are presented. There was no difference between the two groups of patients in terms of gender distribution, but very significant differences were found with respect to age category, rhinitis duration and spirometric parameters: patients with severe BHR were more frequently patients with age > 23 years, with a rhinitis duration > 5 years and with FVC ≤ 104, FEV1 ≤ 86 and FEF25–75 ≤ 57% of predicted.

Table 3.   Demographic and clinical parameters in patients with severe BHR (PC20 < 1 mg/ml) and in patients with mild and borderline BHR (PC20 1–16 mg/ml)
 PC20 < 1 mg/ml, N = 22PC20: 1—16 mg/ml, N = 266P
  1. Values denote N (%). All percentages in round brackets are calculated over the total number of subjects reported at top of the column.

  2. *Chi-square test.

  3. †Fisher’s exact test.

Gender – males20 (90.9)202 (75.9)0.11*
Age > 23 years,16 (72.7)96 (36.1)0.0007*
Rhinitis duration > 5 years14 (63.6)44 (16.5)< 0.0001†
FVC ≤ 104% of predicted21 (95.5)179 (67.3)0.006*
FEV1≤ 86% of predicted18 (81.2)50 (18.8)< 0.0001*
FEF25–75≤ 57% of predicted11 (50.0)20 (7.5)< 0.0001†

The description of sensitizations in the two groups of patients is shown in Table 4. Patients with severe BHR were more frequently positive to house dust mites, trees, perennial and poly sensitizations. The combination of different allergens were more deeply investigated and patients sensitized to trees and house dust mites resulted more frequently the patients with severe BHR (P < 0.0001) (Table 4).

Table 4.   Description of allergen sensitizations in patients with severe BHR (PC20 < 1 mg/ml) and in patients with mild and borderline BHR (PC20 1–16 mg/ml)
 PC20 <1 mg/ml N = 22PC20 1–16 mg/ml N = 266P
  1. Values denote N (%). All percentages in round brackets are calculated over the total number of subjects reported at top of the column.

  2. *Chi-square test.

  3. †Fisher’s exact test.

House dust mites18 (81.8)151 (56.8)0.02*
Trees18 (81.8)93 (35.0)< 0.0001*
Parietaria13 (59.1)166 (62.4)0.76*
Graminae5 (22.7)115 (43.2)0.06*
Moulds1 (4.5)61 (22.9)0.06†
Dog or cat1 (4.5)29 (10.9)0.71†
Compositae0 (0.0)10 (3.8)1.00†
Perennial: yes19 (86.4)161 (60.5)0.02*
Polysensitization: yes20 (90.9)183 (68.8)0.03*
Type of allergen
 Trees absent4 (18.2)173 (65.0)< 0.0001†
 Trees and pollens1 (4.5)43 (16.2) 
 Trees and house dust mites17 (77.3)50 (18.8) 

Multivariate logistic analysis

In order to evaluate the role of possible predictors for severe BHR (PC20 < 1 mg/ml), a logistic regression model was performed (Table 5).

Table 5.   Best fitting logistic regression models for independent predictors of the presence of severe BHR (PC20 < 1 mg/ml) (= 288)
Outcome variable: severe BHR (PC20 < 1 mg/ml)
Explanatory variablesOdds ratioAdj95% CIP value*
  1. 95% CI: 95% confidence interval.

  2. *Likelihood ratio test.

Type of allergen:   
 Trees and pollens (reference category: trees absent)1.1(0.1–11.6)0.002
 Trees and house dust mites (reference category: trees absent)8.1(2.1–30.9) 
Rhinitis duration (> 5 years)5.4(1.7–16.6)0.003
FEVI ≤86 % of predicted (reference category: > 86 % of predicted)4.0(1.0–15.5)0.037
AUC of the model: 0.90   

Three predictors turned out to be significantly associated with severe BHR: sensitization to trees and house dust mites (ORAdj: 8.1), rhinitis duration > 5 years (ORAdj: 5.4), and FEV1 ≤ 86% of predicted (ORAdj: 4.0). The model’s discriminative ability is very satisfactory, being the AUC = 0.90.

Discussion

Allergic rhinitis and asthma may be considered as a single syndrome involving two parts of the respiratory tract (17). Patients with allergic rhinitis may quite frequently present asthma symptoms and/or spirometric impairment. Indeed, it has been very recently demonstrated that patients with allergic rhinitis alone may show reduced FEF25–75 values as early marker of bronchial impairment (13). Thus, this finding remarks the link between upper and lower airways.

The present study investigated the presence of BHR in a large cohort of patients with moderate-severe persistent allergic rhinitis alone, such as perceiving only nasal symptoms. It is to note that this study is one of the largest conducted to this date on this topic.

Firstly, this study may be considered confirmatory of previous studies as a remarkable percentage of these patients show BHR (28.1%). This finding highlights the concept that a bronchial involvement is frequent in allergic rhinitis, also in absence of overt asthmatic symptoms. Therefore, the recommendation, suggested by ARIA document (18), of always considering this issue in each patient with allergic rhinitis has to be strongly supported.

Secondly, several risk factors, concerning a bronchial involvement in terms of serious BHR, may be defined. The sensitization to trees and mites, the duration of allergic rhinitis > 5 years, and FEV1 values ≤ 86% of predicted appear to be relevant for predicting severe BHR.

Thirdly, both FEV1 and FEF25–75 are more impaired in patients with BHR and this issue underlines the close relationship between bronchial function impairment and BHR as previously reported (12, 17, 20).

Fourthly, this study underlines that the subpopulation of patients with severe BHR is distinct from other ones. Patients with severe BHR are obviously more serious than others and more probably will evolve in overt asthma than others. It is to consider that in this sub-group the difference between FEV1 and FEF25–75 values is greater than in other patients: about 30. In this regard, it has been previously stated that a difference > 20 between FEV1 and FEF25–75 values is predictive for BHR in patients with respiratory allergy (20). Thus, this finding remarks that the patients with severe BHR have a more evident bronchial impairment than other patients. Moreover, this study confirms and increases the strength of the role of BHR as marker of bronchial involvement in patients suffering from isolated persistent allergic rhinitis. This issue supports the concept that airway allergic inflammation and airflow impairment are closely associated as well as upper and lower airways. Moreover, gender does not seem to represent a significant risk factor (OR 0.32; C.I. 0.07–1.39; LR test P = 0.08), whereas previous report evidenced conflicting result (22). This issue might represent a limitation of the study as most of the evaluated patients are male sailors with restrictive entry criteria. However, a very recent multicenter study conducted on 418 Italian patients with allergic rhinitis showed that > 52% of them suffered from moderate-persistent PER and that there was no difference between genders (23).

Lastly, the most clinically relevant finding of this study is represented by the suggestion of carefully considering some risk factors such as the sensitization to trees and mites, the duration of allergic rhinitis and values still normal of FEV1 (such as ≤ 86%). These findings may be considered as the most important risk factors for a possible severe BHR in allergic rhinitis.

In this regard, it is noteworthy to consider the relevant role as risk factor played by trees sensitization that is more important than other pollens and animal danders. A possible explanation might be that trees allergy in Ligurian patients is characterized by a relevantly increased prevalence and severity of symptoms (24). In addition, it has been very recently reported that patients with severe trees allergy also outside the pollen season have high IL-17 serum levels in comparison with patients sensitized to other pollens that have lower or undetectable IL-17 values (25). Moreover, another recent study pointed out that only mite allergy constitutes a very relevant risk factor for asthma development in patients with allergic rhinitis (26). Another possible explanation might be that population under study is scarcely exposed to domestic pets as Navy sailors. This issue represents therefore a limitation of this study. Thus, these findings should be confirmed in a second population more representative.

Therefore, from a clinical point of view it appears to be mandatory to evaluate in each patient with moderate-severe persistent allergic rhinitis the type of sensitization, the duration of nasal symptoms and to consider of performing spirometry. Thus, as suggested by ARIA document (18), a spirometry should be performed, also in absence of overt asthmatic symptoms, in patients with moderate-severe persistent allergic rhinitis with sensitization to trees and mites and long duration of the disease to prematurely detect the possible presence of serious BHR.

In conclusion, this study highlights the close link between upper and lower airways and the role of some risk factors, such as duration, some sensitizations and border-line FEV1 values as early prognostic marker of clinically relevant BHR in patients with moderate-severe persistent allergic rhinitis alone.

Ancillary