Exhaled NO reference limits in a large population-based sample using the Lambda-Mu-Sigma method

摘要

Absolute values are used in the interpretation of the fraction of exhaled nitric oxide (FeNO), but it has been suggested that equations to calculate reference values may be a practical and clinically useful approach. We hypothesize that the application of the Lambda-Mu-Sigma (LMS) method may improve FeNO reference equations and their interpretation. Our aims were to develop FeNO reference equations with the LMS method and to describe the difference between this method and the absolute fixed cut-offs of the current recommendations. We utilized the United States National Health and Nutrition Examination Surveys 2007-2012 and included healthy individuals with no respiratory diseases and blood eosinophils 300/mm(3) (n = 8,340). Natural log-transformed FeNO was modeled using the LMS method, imbedded in the generalized additive models for location, scale, and shape models. A set of FeNO reference equations was developed. The explanatory variables were sex, age, height, smoking habits, and race/ethnicity. A significant proportion of individuals with normal FeNO given by the equations were classified as having intermediate levels by the current recommendations. Further lower predicted FeNO compared with previous linear models was seen. In conclusion, we suggest a novel model for the prediction of reference FeNO values that can contribute to the interpretation of FeNO in clinical practice. This approach should be further validated in large samples with an objective measurement of atopy and a medical diagnosis of asthma and rhinitis.