There is a growing body of evidence for the utility of eosinophil-derived neurotoxin (EDN) as a biomarker in asthma, including association with eosinophilic airway inflammation, assessment of disease severity and potential for predicting pathogenic risks, including exacerbations. However, to interpret any biomarker data with confidence, it is first important to understand the preanalytical factors and biological variation that may affect its reliable measurement and results interpretation. In this study we defined the healthy serum EDN reference range for men and women as 1.98 to 26.10 ng/mL, with no significant gender differences. Smoking did not impact the mean EDN levels and no circadian rhythm was identified for EDN, unlike blood eosinophils (EOS) where levels peaked at 00:00h. EDN expression in different cell types was investigated and shown to occur primarily in eosinophils, indicating they are likely to be the main cellular repository for EDN. We also confirm that the quantification of serum EDN is not influenced by the type of storage tube used, and it is stable at ambient temperature or when refrigerated for at least 7 days and for up to one year when frozen at -20°C or -80°C. In summary, EDN is a stable biomarker that may prove useful in precision medicine approaches by enabling the identification of a subpopulation of asthma patients with activated eosinophils and a more severe form of the disease.

Data was collected using a commercially available ELISA. Data here is that obtained from the ELISA assay. In the publication statistical evaluations (tests, confidence intervals, correlations) were based on log transformed EDN measurements, or log transformed eosinophil counts, due to the log-normal characteristic of this type of data. Presented group means and confidence intervals are back-transformed to linear scale. The reference interval was determined according to the Clinical Laboratory Standards Institutes guidance. The reference interval is set at 95% of the  upper and lower values measured from a population, meaning that the lower and upper reference limits are estimated as the 2.5^th and 97.5^th percentile. Temporal changes in EDN and eosinophils were analysed using a mixed model for repeated measurements, with log of EDN and log of eosinophils as response variables, timepoint and cohort (asthma, healthy) as factors with interaction, and subject as a random effect to account for within-subject correlation.

Where an EDN ELISA value could not be obtained, due to a sample not being aquired, this is highlighted as ND in the dataset ie not determined.