Alveolar concentration and bronchial flux of nitric oxide: two linear modeling methods evaluated in children and adolescents with allergic rhinitis and atopic asthma

Objective: Alveolar concentration (C(A)NO) and bronchial flux (J(aw)NO) of nitric oxide (NO) characterize the contributions of peripheral and proximal airways to exhaled NO. Both parameters can be estimated using a two-compartment model if the fraction of NO in orally exhaled air (FE(NO)) is measured at multiple constant expiratory flow rates (V). The aim of this study was to evaluate how departures from linearity influence the estimates of C(A)NO and J(aw)NO obtained with the help of linear regression analysis of the relationships between FE(NO) and 1/V (method P), and between the NO output (V(NO) = FE(NO) × V) and V (method T). Furthermore, differences between patients with atopic asthma (AA) and allergic rhinitis (AR) and between methods P and T were assessed.

Design: Measurements of FE(NO) were performed with a chemiluminiscence analyzer at five levels of V ranging from 50 to 250 ml/sec in school children and adolescents with mild to moderate-severe AA treated by inhaled corticosteroids (N = 42) and AR (N = 20).

Results: Violation of the linearity condition at V ≤ 100 ml/sec caused shifts between methods with regard to the partition of exhaled NO into alveolar (C(A)NO: P > T) and bronchial (J(aw)NO: T > P) components. Both methods gave similar results in the linear range of 150-250 ml/sec: The mean ratios P/T and limits of agreement calculated in AA and AR patients were 1.03 (0.49-1.56) and 1.07 (0.55-1.59) for C(A)NO and 1.03 (0.73-1.33) and 0.99 (0.90-1.10) for J(aw)NO, respectively. No significant differences between AA and AR were found in C(A)NO and J(aw)NO calculated in the linear range by the T method {medians (inter-quartile ranges): 1.7 ppb (0.9-3.9) vs. 2.3 ppb (0.8-3.7), P = 0.91; 1,800 pl/sec (950-3,560) vs. 1,180 pl/sec (639-1,950), P = 0.061}. However, the flow-dependency of the estimates was markedly higher in AA than in AR patients: C(A) NO was decreased 2.8-fold vs. 1.5-fold and J(aw) NO was increased 1.5-fold vs. 1.2-fold in the linear range as compared to the range of 50-250 ml/sec. In both groups, the median standard errors (SE) of the J(aw) NO estimates were similar for the metods P and T and small (<15%) regardless of the range for expiratory flows. The precision of C(A) NO estimates was less in all ranges. For both methods, the SE of the estimates obtained in the range of 150-250 ml/sec exceeded 50% in asthmatics and 30% in AR patients, respectively. The results show that FE(NO) has to be measured at several expiratory flows ≥100 ml/sec for the accurate estimation of C(A) NO and J(aw) NO using linear methods P and T in children and adolescents with AA and AR. A stepwise procedure for detecting nonlinearity and evaluating the quality of FE(NO) measurements is suggested.