METHODS

Study design and participants

In the international multi-centre (Austria, France, and Italy) cross-sectional AIS Life + study, conducted between 2013 and 2014, we enrolled participants suffering from nasal allergy. A convenient sample of individuals with an active condition of pollen-induced AR was selected from pre-existing epidemiological study databases or through web advertisement (Pisa, Italy), clinics of general practitioners (Paris, France) or public health database and pulmonary clinics (Vienna, Austria) and invited to participate in this epidemiological survey. All potential participants were administered a screening questionnaire through a telephone interview to check whether they were eligible for the study. We included individuals who: 1) were adults (≥ 18 years of either sex); 2) reported allergic rhinitis diagnosis/symptoms or positive clinical tests to pollen in the last 12 months; 3) spent most of the week (at least 5 days/week) living, studying, or working in the areas where this study was conducted; and 4) were not treated with allergen immunotherapy over the previous 6 years.
The study was approved by the ethics committees of the participating centres in Italy (Ethics Committee of University-Hospital of Pisa; Protocol No. 14248) and in Austria (Berlin Charite University Ethics: EA1/119/12), and signed informed consents were obtained from all the participants before recruitment. In France, the approval by an external ethic committee was declared as not applicable at that time: instead, the study was approved by the Hospital ethic committee, by the National Committee for Information Management on Medical Research (Comite´ Consultatif sur le Traitement de l’Information en Matière de Recherche dans le domaine de la sante´ ) and by the National Commission on Informatics and Health (CNIL, Commission Nationale Informatique et Liberte´ ). In France, the CNIL approved in 2016 that all data acquired prior to 2016, without the previous need of an authorization of an Ethic Committee, could still be utilized. In any case, patients were seen in the frame of routine care (“soins courants ”). The AIS study was conducted according to the Declaration of Helsinki and reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines18.

Instruments and variables

A standardized health questionnaire was administered to all eligible individuals in order to obtain information on their demographic characteristics (age, gender, body mass index [BMI], level of education), and potential risk factors (smoking status, exposure to second-hand smoke, and drug consumption), on symptoms/diagnosis of AR and asthma, disease severity level, as well as on previous clinical tests (spirometry, skin prick test and serum level of immunoglobulin E [IgE]).
The health-related quality of life (HRQoL) of the participants was assessed using a modified version of the validated RHINASTHMA questionnaire (the higher the score, the lower the HRQoL), i.e., the only available instrument that allows evaluating the concomitant impact of AR and asthma on HRQoL19. This 30-item questionnaire provides individual scores for upper airways, lower airways, respiratory allergy impact, and a composite score (the Global Summary [GS] score, which indicates the overall impact of the disease). The details of the instrument and the scoring system can be found elsewhere19. Patients used a five-point Likert scale (’not at all’, ’a little’, ’fairly’, ’much’, ’very much’) to indicate the extent to which they were bothered by each AR and asthma during the year preceding the completion of the questionnaire. These responses are then converted into scores, from 0 to 100, with larger scores corresponding to worse HRQoL. A RHINASTHMA-GS score from 0 to 20, indicating minimal or absent disease impact on patient life, was considered reflective of optimal HRQoL.
The control of AR and asthma was evaluated by a modified version of the Control of Allergic Rhinitis/Asthma Test (CARAT) (the higher the score, the higher the disease control)20,21. CARAT is a 10-item questionnaire containing information about the frequency of symptoms, sleep impairment, activities limitation, and need for more medication: the response options for all the questions follow a 4-point Likert scale (range 0-3). The range of CARAT score is 0–30, 0 being the complete absence of control: the minimal clinically important difference (MCID) is 3.522. The Global Initiative for Asthma (GINA) classification 201723 and ARIA (2008)6 were used to classify asthma according to its severity.

Statistical analyses

Data were described as frequency (%), mean (standard deviation [SD]), or median (interquartile range [IQR]) for categorical, continuous, and ordinal variables, respectively. To test the association among quality of life and control (RHINASTHMA and CARAT – total and subdomains) scores, and AR-asthma (independent variable), we first used a bivariate analysis using Wilcoxon rank-sum test. Then, we constructed univariable (unadjusted) and multivariable (adjusted) regression models among the independent variable and HRQoL and control scores using a mixed effect Poisson regression model. As potential confounders, we tested fixed factors (age, sex, BMI, smoking status, exposure to smoke, education, ARIA grade, sensitivity to allergens, and drugs taken in the last 12 months) and a random factor (the country). To include confounders in the regression models, we used a priori evidence criteria, i.e., covariates were considered as confounders if were found consistent in previous literature. However, confounders were retained in the model if they modified the estimates of the remaining variables by more than 10%. We checked the collinearity of the confounders using the variance inflation factor (VIF). The parsimony of the models was confirmed by Akaike’s information criteria (AIC).
We also performed two secondary analyses. Firstly, we tested if there was any effect modification by obesity on the association between AR-asthma, and the HRQoL and control scores. Secondly, we performed meta-analyses to determine if there were any heterogeneity in the HRQoL and control (total) scores between the participating countries. All analyses were conducted using a complete case approach in Stata V.16 (StataCorp, College Station, TX, USA), and a p-value < 0.05 was considered statistically significant.