Umeå universitets logga

The relationship between lung function decline and airway inflammation among asthmatics has important therapeutic implications, but has rarely been studied in large samples or in population-based asthma cohorts.

A population-based adult asthma cohort (n=2055) was recruited during 1986-2001 and clinically examined including spirometry. In 2012-2014, all still eligible subjects (n=1425) were invited to a clinical follow-up including spirometry, blood sampling, and a structured interview, and n=1006 participated (55% women, mean age 59y, 32-92y). Linear regression was performed with age, sex, smoking habits, year of first examination, family history of asthma, socioeconomic status, eosinophils (EOS)>=0.3x109/L, and neutrophils (NEUT)>=5.0x109/L as independent variables and pre-bronchodilator FEV1 decline/year (ml and % of predicted [pp], respectively) as dependent. In secondary models, both ICS use at baseline and ICS use at follow-up were also included.

The mean annual FEV1 decline in ml (pp) among asthmatics with EOS<0.3, 0.4>EOS>=0.3 and EOS>=0.4x109/L, respectively, was 26ml (0.03pp), 29ml (0.10pp) and 34ml (0.27pp) (p<0.001). In adjusted analyses, EOS>=0.3 was significantly associated with FEV1 decline, both in terms of ml (4ml excess annual decline vs EOS<0.3) and pp. The association between EOS and FEV1 decline in pp, but not ml, remained when additionally adjusted for ICS use. The association with NEUT>=5.0x109/L was less clear.

On group level, adult asthmatics with higher levels of eosinophils in blood have a history of excess FEV1 decline compared to asthmatics with lower levels of eosinophil inflammation, independent of other factors such as ICS use.