Variability of eosinophilia in the LEAD study: Pulmonary and extrapulmonary traits

<bold>Background:</bold> Blood eosinophil counts (BEC) show moderate variability in disease types over time, but evidence for this and their association with pulmonary and extrapulmonary traits in a general population is lacking. We hypothesise a temporal variation of BEC over two phases (P1 and P2; 4.3±0.6 years) in the population-based Austrian LEAD study. <bold>Methods:</bold> Qualified subjects (n=6932, 18-82 years) were classified into four BEC groups with 210 (upper) and 128 (lower) cells/µL thresholds (Hartl et al., 2020): consistently high (BEC_CH, ≥210 cells/µL at P1 and P2), consistently low (BEC_CL, <128 cells/µL at P1 and P2), variably increasing (BEC_VI,<128 or 128-<210 cells/µL at P1 and ≥210 cells/µL at P2) and variably decreasing (BEC_VD, ≥210 cells/µL at P1 and 128-<210 or <128 cells/µL at P2). <bold>Results:</bold> <italic>1)</italic> 67.7% and 67.3% subjects from <128 and ≥210 cells/µL categories remained stable over time: 91.6% with low BEC-P1 remaining below the high threshold (<210 cells/µL) at P2. 13.4% and 7.9% showed BEC_VI and BEC_VD respectively; <italic>2)</italic> the variability in BEC had significant effects on lung function changes (post-bronchodilator ΔFEV₁ (ml/year): BEC_CL, -17.3±55.1; BEC_CH, -21.0±59.7; BEC_VD, -11.5±50.1; and BEC_VI, -24.2±59.3); and <italic>3)</italic> comorbidities (particularly asthma (odds ratio: BEC_VD, 1.7 and BEC_VI, 1.6) and skin atopy (odds ratio: BEC_VD, 1.6, and BEC_VI, 1.5)) were the largest contributing factors to variable BEC patterns in multivariate logistic regression models. Associations with body morphometry and composition, and smoking habits were notable only in univariate analyses. <bold>Conclusion:</bold> In a population-based cohort, variable BEC patterns over time are associated with changes in lung function and extrapulmonary traits.