Serum eosinophil cationic protein: distribution and reproducibility in a randomly selected sample of men living in rural Norfolk, UK

Background It has been proposed that serum levels of eosinophil cationic pr otein (ECP) may be a clinically useful measure in allergic illness. The aim of this report is to describe the distribution and reproducibility of seru m ECP levels in a population sample and to examine its relationship with ot her markers of disease. Methods The study was conducted in rural areas of Norfolk, UK in a random s ample of men aged 20-44 years enriched with subjects drawn from general pra ctice 'asthma registers'. Asthma symptoms were assessed using the EC Respir atory Health Survey questionnaire. Atopy was measured by skin prick tests a nd serum IgE. Airway hyper-responsiveness (AHR) was tested by methacholine challenge test. Serum IgE and ECP was measured by fluoroimmunoassay using t he Pharmacia CAP system. Reference equations were derived in subjects from the random sample who did not have symptoms of asthma, abnormal lung functi on or AHR. The relation of serum ECP with various clinical characteristics was examined in the whole study population. Reproducibility of serum ECP me asurement was assessed in 57 subjects 4 weeks after the initial test. Results The study population comprised 311 from the random sample and 58 fr om the asthma sample. The reference equation for serum ECP for healthy men was log(10) ECP = 1.3966 - [(age - 20) x 0.0057]. The estimated mean serum ECP for a 20-yearold man was 25 mu g/L. Current smokers have higher serum E CP levels that non-smokers (P=0.014). ECP levels were not related to the sk in prick test reactivity, serum IgE, a questionnaire-based diagnosis of ast hma, or impaired lung function (all P > 0.05). Levels were higher in subjec ts with AHR (P = 0.003) and those who reported wheeze (P = 0.017) but there was no clinically useful separation in ECP levels between subjects classif ied by these criteria. The test was moderately reproducible over a 4-week p eriod (intraclass correlation coefficient = 0.62). Discussion Serum ECP levels were higher in this rural English population th an reported in a comparable population in Sweden. Serum ECP is a reproducib le test but cross-sectionally does not relate in any clinically useful way to markers of asthma. The meaning of between-subject differences in ECP lev els requires further exploration.