Annual Variations in Asthma Attacks in the Italian Adult Population

Abstract

Introduction: Seasonal variation plays a crucial role in the occurrence and exacerbation of allergic diseases, yet real-world, large-scale epidemiological evidence remains limited. Asthma attacks can be triggered by pollen and often worsen during the pollen season. However, seasonal patterns may vary across different climatic areas.

Objectives: The primary aim of the present study was to assess the seasonality of asthma attacks in Italian adults.

Methods: We analyzed GEIRD (Gene Environment Interaction in Respiratory Diseases) data, collected from 2008 to 2014 in seven Italian centers, three (Verona, Pavia, Turin) located in the Po Valley (sub-continental area) characterized by sub-continental climate, and four (Sassari, Ancona, Terni, Palermo) in the Mediterranean area.

Seasonality was assessed using a Poisson regression framework incorporating Fourier terms (extended Stolwijk model). Both single-peak (unimodal) and double-peak (bimodal) cyclic trends were investigated. The model selection was based on both pseudo R-squared values (R2; explanatory power) and the Goodness-of-Fit (GoF) test. The model providing the best overall fit was retained.

Results: The study included 204 subjects (137 and 67 from sub-continental and Mediterranean areas, respectively) who reported having suffered from asthma attacks for a median of 4 months in the previous year. Both unimodal and bimodal Poisson regression models were fitted to this cyclic pattern of asthma attacks. All unimodal models had acceptable GoF. However, the bimodal models had higher pseudo R-squared values with acceptable GoF. Overall, the seasonal pattern of asthma attacks was characterized by a main peak in April and a secondary peak in September (p<0.001; R2=0.503; GoF=0.95).

In the sub-continental area, a primary peak was detected in May and a secondary one in September (p<0.001; R2=0.42; GoF=0.89), with relative increases of 34% and 19%, respectively. In contrast, the seasonality of asthma attacks in the Mediterranean area was characterized by the first peak in May, whereas the second peak occurred in October (p=0.004; R2=0.36; GoF=0.99) with relative increases of 37% and 22%, respectively.

Conclusion: Asthma attacks in Italy mainly occur during spring. The bimodal cyclic pattern explained 50% of the monthly variability in asthma attacks. Nevertheless, a weaker secondary peak of asthma attacks was observed in autumn. The sub-continental (Po Valley) and Mediterranean areas had slightly different autumn peaks, which may reflect differences in ragweed and Urticaceae exposure, and warrant further investigation. In addition, the modeling approach identified significant seasonal variation.