Airways inflammation may precede pulmonary dysfunction in wheezing individuals. The fraction of exhaled nitric oxide (FE_NO) has been described as a useful method for wheezing diagnosis in children, however, its application requires evidence. This study aimed to determine the accuracy of FE_NO in identifying wheezing in preschoolers.

A cross-sectional study was carried out with children from 3 to 5 years old, from Brazilian day care centers and public schools. They were evaluated by FE_NO measurement through the single breath method, and by ATS-DLD-78-C questionnaire that is used as a gold standard to phenotype wheezing patterns.

The sample consisted of 243 non-wheezing children, 118 non-recurrent wheezing and 62 recurrent wheezing. The means of FE_NO and confidence intervals of 95%, were 5.4 (CI 95%, 5.2–5.6); 7.5 (CI 95%, 6.9–8.2) and 11.2 (CI 95%, 9.6–12.7), respectively. The sensitivity, specificity, positive and negative predictive FE_NO values in the 6 parts per billion (ppb) cut-off point that best diagnosed wheezing of non-wheezing children, were: 65.5%, 84.3%, 75.6% and 76.7%, respectively, with an area under the curve (AUC) = 0.77. At 10 ppb, the best cut-off points for differentiating recurrent wheezing of non-recurrent wheezing were: 56.4%, 81.3%, 61.4%, 78.0%, respectively, with an AUC = 0.69. The post-test probability for each FE_NO cut-off points was increased by 33% for wheezing and 20% for recurrent wheezing diagnosis when associated with clinical examination.

FE_NO can provide a reliable and accurate method to discriminate the presence and type of wheezing in preschoolers with 92% of acceptable in this study population.