Background: Severe asthma presents a major challenge to health care and negatively affects the quality of life of patients. Understanding the factors predicting the development of severe asthma is limited.

Objective: To characterize patients with severe asthma and establish risk factors for the development of severe asthma in a Finnish sample with a nationwide coverage of population, health care, and drug register data.

Methods: We used data between January 1, 2014 and December 31, 2020. Pooled data over the years were used to identify characteristics of patients with severe asthma. Annual data were used in machine learning methods and logistic regression to identify factors predicting the development of severe asthma.

Results: Analysis of pooled data including 242,164 individuals revealed that patients with severe asthma were more often women, slightly older, multimorbid, and had higher body mass index values compared with patients with nonsevere asthma. They also had higher use of nonasthma-related medications, manifesting as polypharmacy. Annual data from 6908 patients revealed that the most significant predictors of the development of severe asthma were being aged 51 to 60 years (odds ratio [OR] 3.90 [95% CI: 3.42-4.47]), chronic sinusitis (OR 2.48 [95% CI: 2.12-2.89]), and higher blood eosinophil counts (≥600 cells/μL, OR 2.10 [95% CI: 1.56-2.28]). Increases in all medications (nonasthma and asthma medications) were observed in the year before the onset of severe asthma.

Conclusion: The results provide a clinically relevant risk factor profile for early identification of the patients at risk of developing severe asthma.