*co-last authors

Biological and clinical determinants of persistent asthma have been poorly investigated in non-severe asthma from population-based studies.

Adults from the French Constances cohort with current asthma (CA) at inclusion were followed-up. At inclusion and follow-up, CA was defined by reports in the last 12 months of asthma attacks, symptoms or asthma medication. Having CA at both time-points defined persistent asthma, otherwise asthma was labelled remittent. Blood eosinophil (0.25x10⁹/L) and neutrophil (5x10⁹/L) cut-offs defined paucigranulocytic (P), neutrophilic (N), eosinophilic (E) and mixed (M) phenotypes. Associations of asthma evolution (persistent/remittent) with asthma characteristics or inflammatory phenotypes at inclusion were studied using logistic models adjusted for age, sex, smoking status, education level, body mass index, and French deprivation index at inclusion.

Among 16000 participants with CA at inclusion, 3060 had data available 5 years later (65% with persistent asthma; at inclusion: mean age 47 years, 46% men; P, N, E and M phenotypes: 58%, 6%, 33% and 3%). Persistent asthma was significantly associated with asthma attacks, nocturnal symptoms, chronic bronchitis, dyspnoea, and asthma medication at inclusion (adjusted aOR from 1.29 to 3.45). The likelihood of persistent asthma increased with increasing asthma symptom score at inclusion (aOR from 0.77 to 4.80, p-trend<0.001). The probability of persistent asthma was greater in phenotypes E (aOR=1.40[1.18-1.66]) and M (aOR=1.74[1.09-2.79]) compared to P.

This study highlights the importance of blood inflammatory phenotypes and clinical burden as determinants of non-severe asthma.