In children, FE_NO is age dependent [19, 45, 48] with an increase of about 5 % or 1 ppb/year [19, 48]. Besides age, height, body mass index (BMI), gender, and race can affect FE_NO. FE_NO has been shown to have a positive correlation with height [45, 46, 48], and a negative relationship with BMI [45]. However, no significant correlation with height, weight, BMI, or body surface area is noted after adjusting for gender [50] and another study using BMI z score, which is age-independent and sex-independent, also shows no correlation between BMI and FE_NO [53]. Males tend to have higher FE_NO than females [8, 45, 50]. Non-white population, particularly Asians, have higher FE_NO [45, 48, 50].

FE_NO can be increased by ingestion of nitrate-rich l-arginine containing foods, such as lettuce, spinach, ham, cucumber, potato, and tomato [54, 55]. FE_NO increases steadily after nitrate or nitrate-rich food ingestion with a maximum at 120 min, and then decline. However, FE_NO at 3 h after ingestion still remains higher than the baseline prior to ingestion [54, 55]. Therefore, nitrate-rich diet should be avoided in 3 h before the measurement. Dietary intake of fats such as butter has a positive correlation with FE_NO, while consumption of salad, one of the major sources of antioxidants, is found to be negatively associated with FE_NO level [56]. FE_NO decreases with drinking water 5–20 s before exhalation maneuver, and water temperature does not affect this effect [57]. FE_NO decreases significantly in the first hour after coffee or caffeine consumption, and this drop remains consistently for 4 h after consumption compared to the placebo [58]. Alcohol also decreases FE_NO in asthmatic subjects but not in normal individuals [59]. Previously refraining from eating and drinking for 1 h before FE_NO measurement is recommended by ATS/ERS [2], but longer hours of refraining might be needed to avoid the dietary effect.

Many medications can affect FE_NO level. Oral or intravenous l-arginine has been shown to increase FE_NO in healthy adults [60, 61], but a recent study did not show a significant increase in FE_NO after oral arginine administration regardless of history of allergy [62], at the same dose studied in the previous study [60]. The nitrite-reducing conditions in the oral cavity by using antibacterial mouthwash such as chlorhexidine acetate or sodium bicarbonate reduce FE_NO [55, 63]. Inhaled and oral corticosteroid reduce FE_NO in asthma [32, 64–66] and leukotriene receptor antagonist such as montelukast also decreases FE_NO [67]. However, nedocromil does not reduce FE_NO in asthmatic children [66]. Omalizumab decreases FE_NO and the degree of inhibition of FE_NO is similar to that seen for inhaled steroid alone [68]. Bronchodilators like beta-agonists have been shown to increase FE_NO in asthmatics [67, 69–71] while low-dose theophylline does not affect FE_NO in mild asthma despite reducing eosinophilic inflammation [72]. Bronchoconstriction by methacholine decreases FE_NO in healthy volunteers [70]. However, no significant fall in FE_NO is observed after allergen or isocapnic cold air challenge in atopic asthma [73]. The reason for these changes is not clear but is believed to be due to neural mechanisms leading to increase in NO release from lower airways or mechanical effect on airway caliber.

Cigarette smoking has been shown to reduce FE_NO both acutely and on long-term basis [74, 75]. FE_NO is also reduced with passive smoke exposure both in adults [45, 76] and in early infancy [34, 77, 78] and school age children [45]. Infants exposed pre- and postnatally to smoke show lower FE_NO than infants exposed only after birth and never-exposed infants [77]. However, this is not the case in infants of mother with atopy or asthma, and FE_NO increase [78]. Another study done in young children with mean age 51.3 weeks [79] has a different result, showing higher FE_NO with exposure to parental smoking. A dose–response relationship between FE_NO and the number of smoking parents is also noted [79]. This discrepancy in result could be related to differences in age of subjects, duration of passive smoking exposure, and prenatal environmental factors such as maternal smoking and atopy.

Spirometric maneuvers have been shown to transiently reduce FE_NO [69, 71, 73, 80, 81], as early as 1 min after spirometry and FE_NO returned to baseline over 1 h. It is important to obtain FE_NO consistently prior to spirometry. However, FE_NO maneuver itself and body plethysmography do not appear to affect plateau FE_NO level [69, 81]. Reduction in FE_NO with spirometry is blunted by bronchoprovocation with isocapnic cold air hyperventilation or allergen [73]. FE_NO is reduced after sputum induction [82, 83]. However, this change occurs with sputum induction by hypertonic saline, not by isotonic saline, and the decreased FE_NO is observed over 4 h and returns to baseline after 24 h [83]. A change in FE_NO occurs with exercise, and the largest drop in FE_NO is noted at 5 min after exercise [80]. Therefore, ATS/ERS recommends avoiding strenuous exercise for 1 h before the measurement [2].

The effect of circadian rhythm on FE_NO is not conclusive. No variation is shown in healthy and asthmatic adults and children as well as infants [52, 84–86] while other studies report the morning levels higher than the evening levels in asthmatic and healthy children [20, 87]. The opposite finding in circadian variation is also noted with FE_NO at 4 pm higher than FE_NO at 4 am in nocturnal asthma [86]. If possible, it is ideal to perform serial measurements of FE_NO in the same period of the day to avoid the effects of circadian rhythm. Seasonal variation in FE_NO is reported due to natural pollen exposure in children with seasonal asthma and pollen allergy [88].

FE_NO does not alter significantly with gestation during pregnancy although amniotic nitrite concentration decreases after 37 weeks of gestation [89]. Menstrual cycle may affect FE_NO results. FE_NO is higher before menstruation than after, in women with complaint of premenstrual asthma [90], but no effect of the menstrual cycle on FE_NO is noted in other studies [91, 92].

The amount of NO in ambient air can affect FE_NO [73, 93]. Ambient NO at the time of each test recorded, and breathing NO free or scrubbed air during the study is recommended. Underlying disease condition has shown to affect FE_NO. FE_NO levels in adults with hypertension, particularly male, and in patients undergoing major surgery are significantly lower than healthy volunteers [75]. Renal failure and dialysis do not appear to have a significant impact on FE_NO [75]. The application of positive end-expiratory pressure has been shown to increase FE_NO in animals [94–96]. High FE_NO is also detected in mechanically ventilated patients with septic syndrome [75]. Hypoxia causes a dose-dependent decrease in FE_NO in both animal research [94] and human studies [97, 98] while carbon dioxide also causes a dose-dependent reduction in FE_NO [94, 95]. Change in pulmonary blood flow did not affect FE_NO in humans but the effect of hemodynamic change on FE_NO is noted in animal study [94]. Many studies report elevated FE_NO in atopy and allergy [30, 31, 33, 99]. Other conditions affecting FE_NO are discussed in the clinical practice section.

Many studies have assessed the roles of FE_NO in asthma. FE_NO can be used to support the diagnosis of asthma, evaluate eosinophilic airway inflammation, assess potential response to anti-inflammatory agents, guide dose titration of anti-inflammatory medications, predict asthma exacerbation, predict asthma relapse, and evaluate adherence to anti-inflammatory medications [1, 4–6].

The official ATS Clinical Practice Guidelines [5] reported a strong recommendation for using FE_NO to assess eosinophilic airway inflammation and steroid responsiveness. Other strong recommendations supported by ATS are age as a factor affecting FE_NO in children <12 years of age; measures of low FE_NO of <20 ppb in children (<25 ppb in adults) indicating less likelihood of eosinophilic inflammation and responsiveness to corticosteroids; cautious interpretation of FE_NO values between 20 and 35 ppb in children (25–50 ppb in adults); persistent and/or high allergen exposure as a factor associated with higher levels of FE_NO; and the use of FE_NO to monitor airway inflammation in asthma. On the other hand, ATS Guidelines provided only weak recommendations for the use of FE_NO as supporting the diagnosis of asthma. Recommendations are also weak for identifying proper cut points in interpretation and in quantitating a significant increase or decrease in values between visits.

FE_NO is considered as an indirect marker of airway eosinophilic inflammation, and many studies describe the correlation between FE_NO and eosinophils measured in sputum, bronchoalveolar lavage, bronchial biopsy, and blood. This relationship was also studied in children [8–14]. In summary, eosinophilic inflammation is unlikely present when FE_NO is low.

Airway inflammation in asthma is heterogeneous and is not always associated with eosinophilic inflammation. The use of FE_NO for diagnostic purpose has been evaluated. In epidemiologic studies, FE_NO can be used as an indicator for allergic airway inflammation in children [14, 49, 99, 100], and performs better than respiratory function measured and bronchodilator responsiveness in identifying preschool children with probable asthma [101], and in predicting subsequent wheezing treated with systemic steroid in infants and toddlers [29]. Sensitivity, specificity, and positive and negative predictive values as diagnostic of asthma are high [101, 102]. However, a cross-sectional survey in adolescent children shows high negative predictive values of FE_NO for asthma but positive predictive value is low for the diagnosis of asthma [14]. High FE_NO can be also noted in atopic children without asthma [47, 100]. The combination of FE_NO and methacholine provocation test may have more diagnostic power in epidemiological studies than FE_NO alone for allergic asthma [103].

In children with respiratory symptoms, FE_NO in the diagnosis of asthma can be a predictor of asthma [101, 102, 104]. Significant association between FE_NO and reported asthma symptoms is also shown [14, 99, 101, 105], while the correlations between symptoms and spirometry are poor [105]. In nonallergic patients, normal FE_NO does not exclude the diagnosis of asthma, and in patients who have already been treated with inhaled steroids, is significantly reduced from previously elevated FE_NO in inflammatory airway diseases. Overall, the results of studies examining the use of FE_NO in the diagnosis of asthma have also been inconsistent in the adult literature. Therefore, FE_NO reflects only one aspect of the asthma phenotype, and should be used as a supportive method to other diagnostic tests.

It is well known that not all patients with asthma respond to corticosteroids. High FE_NO (>35 ppb) in adults has been shown as a positive indicator that the patient would likely respond to inhaled corticosteroids (likelihood ratio of a positive response, 4.9; 95 % confidence interval, 2.2–10.9) [106]. In children, the response to anti-inflammatory treatment was also found to be correlated with the FE_NO [107–109]. High FE_NO (median 17.4 ppb) is associated with a good FEV₁ response (>15 % increase) while lower FE_NO (median 11.1 ppb) is associated with a poor response (<5 % increase) [107]. FE_NO is predictive of steroid responsiveness more consistently than spirometry, bronchodilator response, peak flow variation, or airway hyperreactivity to methacholine [107, 110–112], even when no sputum eosinophilia is demonstrated [113]. The reduction response of FE_NO to corticosteroid treatment is both rapid (within 1 week, potentially as early as 48 h) and dose dependent [112, 114, 115]. Anti-leukotrienes also reduce FE_NO in asthma, but to a lesser extent [109]. FE_NO is also helpful in predicting asthma relapse after clinical remission [116] and in anticipating eligibility for inhaled corticosteroid dose reduction [117, 118]. However, tailoring anti-asthma therapy according to FE_NO in comparison to usual care treatment strategies shows no added value of FE_NO for clinical symptoms, asthma exacerbations, pulmonary function, β-agonist use, and overall daily dose of inhaled corticosteroid [119–123]. A recent systematic review and meta-analysis [124] also conclude no significant benefit of adding FE_NO to traditional treatment algorithms with respect to asthma exacerbations, asthma symptom scores, or forced expiratory flow in 1 s (FEV₁). In children, FE_NO-based treatment group received significantly higher dose of inhaled corticosteroid [123], and this finding is opposite to that adult studies [124]. However, pregnant women with asthma [125] may benefit the most from FE_NO monitoring, resulting in significant reduction in asthma exacerbation during pregnancy.

A strong positive correlation between the reduction of FE_NO and the adherence to anti-inflammatory treatment is noted [126]. Therefore, besides reviewing adequate doses of anti-inflammatory treatment, checking adherence and inhaler technique is needed for children with high exhaled NO who are already being treated with anti-inflammatory treatment.

FE_NO may be able to predict long-term outcome. In adults with difficult-to-treat asthma, FE_NO predicts accelerated decline in lung function [127]. In infants and toddlers with recurrent wheeze, high FE_NO also predicts deterioration in z-scores of forced expiratory volumes and flows, as measured 6 months later [29].

It is important to choose the appropriate cut point in relation to the clinical setting (Table 14.2). Clinical practice strategies for use of FE_NO in patients with asthma have been explained in detail in the text and tables of the ATS Clinical Practice Guidelines [5].

Since NO functions in host defense against viral infections, elevation of FE_NO is likely beneficial to the host by inhibiting viral replication. In both healthy and asthmatic adults, higher FE_NO was noted during viral respiratory tract infections [128–130]. In contrast, infants with rhinorrhea [131] or acute virus-associated wheezy bronchitis [85] have significantly lower FE_NO than healthy infants. With resolution of symptoms, a significant increase in FE_NO is noted in some of the infants with rhinorrhea [131]. These findings suggest possible downregulation of NO production, impaired NO diffusion into the airway due to epithelial damage and increased airway secretions, and an inflammatory reaction mostly related to neutrophils. However, a more recent study in children with respiratory syncytial virus bronchiolitis shows FE_NO higher (but without statistical significance) than healthy control during the acute phase of illness, and a positive correlation between FE_NO and the clinical score (Downes) of bronchiolitis is noted with higher FE_NO in children with more distress [132]. The inconsistency of the effect of viral infection on FE_NO level is not fully explained but the differences in method, sample size, gender, atopy, allergy, asthma of study subjects, maternal smoking as well as possible virus-specific disease process can be speculated. The infants with future wheezing episodes in 2 years [132], and young children <4 years of age with clinical index for predicting asthma at school age [133] already present elevated level of FE_NO. FE_NO in young children with history of recurrent wheezing is also higher than healthy control and children with other pulmonary diseases [134], and atopic wheezers show higher FE_NO than nonatopic wheezers [134].

Data on FE_NO values for children with CLD of prematurity are inconsistent possibly due to the differences in the age of study subjects, gestational age, definitions and severity of CLD, or measurement techniques. Compared to controls, equal [134–136] or higher [137–139] FE_NO was noted in infants and younger children with CLD. FE_NO is particularly elevated in infants with moderate or severe CLD [138]. Although FE_NO is not elevated in infants with CLD, calculated NO output V_NO (nl/min), by multiplying online flow rate (L/min) and FE_NO (ppb), is reduced in infants with lower gestational age, higher clinical risk index for babies score, longer duration of oxygen therapy, postnatal treatment with corticosteroids, and more severe CLD [135]. Even at school age, no significant differences are observed in FE_NO among children with CLD of prematurity, healthy control, and preterm-born children without CLD, unless atopy is not present [140]. FE_NO in prematurely born atopic children is significantly higher than nonatopic children without prematurity [140]. On the other hand, Baraldi et al. [141] reports lower FE_NO in school-age children with CLD of prematurity than healthy matched term-born control and preterm-born children without CLD. FE_NO in children with CLD is four times lower than children with asthma despite a comparable airflow limitation in FEV₁ in both CLD and asthma groups [141]. The low FE_NO in children with CLD is not due to effects of medications such as corticosteroid or leukotriene receptor antagonist, based on study protocol. However, other factors affecting FE_NO such as atopy, allergy, smoking exposure, and severity of CLD are not described for the study subjects.

NO in airway gases obtained by bronchoscopy, NO reaction products in exhaled breath condensate or bronchoalveolar lavage fluid, FE_NO, and V_NO are low in adults with PHN, and are negatively correlated with pulmonary artery (PA) pressures and with years since diagnosis of PHN [142–144]. Therapeutic interventions for PHN are associated with increased levels of FE_NO [144, 145]. Therefore, FE_NO may have a role in monitoring disease severity and response to therapy in children with PHN but further investigation is required.

FE_NO is low in both children [40, 63, 146, 147] and infants [134, 148] with CF despite chronic airway inflammation, although FE_NO levels similar to or higher than control subjects are also reported in children and infants with CF [93, 139, 149]. FE_NO is negatively correlated with lung clearance index in CF [40] and FE_NO is lower in patients with severe CF lung disease than in those with mild disease [147, 150]. However, this association has not been shown in other trial [93]. FE_NO also has an inverse relationship with age, regardless of CF genotypes, and gradually decline throughout young childhood [149]. Several mechanisms for reduced FE_NO in CF [151] have been proposed including reduced iNOS or iNOS expression in CF [152–155], presence of NO reductase in Pseudomonas aeruginosa [156], trapping of NO metabolites in thick secretions [157], deficiency of l-arginine by increased sputum or systemic arginase activities [150, 158], and metabolism and consumptions of NO by reactive oxygen species present in the inflamed environment resulting in elevated nitrate and nitrotyrosine [159, 160].

After therapeutic interventions, the relationship between FE_NO and pulmonary function is not consistent. A significant increase in FE_NO following intravenous antibiotic treatment is noted in children with CF, but does not correlate with lung function [161]. On the other hand, inhaled l-arginine improves both FE_NO and pulmonary function [162]. FE_NO is also not associated with any marker of airway inflammation measured in bronchoalveolar lavage in infants with CF [149].

Decreased iNOS in bronchial epithelium and consequently decreased NO are associated with reduced bactericidal activity in the lung [152, 155], and with increased susceptibility of airway colonization with P. aeruginosa [163]. Interestingly, children with chronic P. aeruginosa infection demonstrated no significant increase in FE_NO after intravenous antibiotics [161, 164].

A study including both children and adult shows lower nasal NO concentration in CF than in controls and asthmatics [93]. However, no relationship between nasal NO and pulmonary function is noted in patients with CF.

C _alv and J _NO have been evaluated in children with CF [40, 43, 44]. J _NO can be similar to or lower than healthy children without CF but variable results were noted for C _alv. C _alv correlates negatively with other systemic inflammatory markers in CF [40]. Children with chronic P. aeruginosa colonization have higher levels of systemic inflammatory markers than those not colonized and also lower levels of C _alv [40].

Although FE_NO may have some utility as a biomarker of CF severity in children, roles of FE_NO in CF-related lung disease need further study and the use of FE_NO is currently very limited in CF.

Children with PCD have lower FE_NO than healthy children. However, some overlap in FE_NO has been shown [165]; thereby, FE_NO cannot be used to differentiate patients with PCD from healthy controls. On the other hand, extremely low levels of nasal NO occur in children with PCD, and can effectively discriminate those with PCD from children with other causes of bronchiectasis, CF, asthma, and healthy controls [104, 166–169]. A detailed description of the utility of nasal NO as a screening tool in PCD is explained further in Chap. 4.

Increased FE_NO has been noted with acute rejection [170], pulmonary infection [171], and development of bronchiolitis obliterans syndrome (BOS) [172, 173] in adult lung transplant recipients. However, no elevated FE_NO is also reported in pulmonary infection or BOS in human lung transplantation [170]. The data on pediatric lung and cardiac transplant recipients demonstrated that C _alv was significantly elevated in cardiac transplant recipients when compared to controls; lung transplant recipients had higher levels of C _alv than controls; however this difference did not reach statistical significance [174]. Nasal NO is significantly lower in pediatric lung transplant recipients when compared to other solid-organ recipients or healthy controls, and was negatively correlated with tacrolimus levels [175]. Higher FE_NO was also observed in children with pulmonary complications after hematopoietic stem cell transplantation [176].