Measurement of fractional exhaled nitric oxide concentration (FeNO) has been proposed as a useful biomarker for monitoring and management of airway diseases. Limited information is available regarding reference levels of FeNO levels in Chinese adults. This study aimed to investigate the reference equation and determinants of FeNO in Chinese adults.
1093 (577 males) healthy nonsmoking subjects aged 18–90 years were recruited. FeNO was measured online using a chemiluminescence analyser. Other assessments included spirometry, skin prick tests, total serum IgE levels and eosinophil count in peripheral blood.
The geometric mean FeNO was 32.6 (95% reference interval (RI) 31.4–33.7) ppb for all subjects. FeNO values were higher in males than females (geometric mean (95% RI) 38.3 (36.5–40.2) ppb versus 27.1 (25.8–28.5) ppb, p<0.0001), and in atopic than nonatopic subjects (34.6 (33.0–36.3) ppb versus 29.8 (28.3–31.4) ppb, p<0.0001). FeNO correlated with age (r2 = 0.23), height (r2 = 0.20), IgE level (r2 = 0.18) and percentage eosinophil count (r2 = 0.18) (all p<0.0001), but not with spirometric parameters. Based on multiple regression modelling, the reference equation of FeNO value was:
log(FeNO) = 0.781 + 0.104(sex) + 0.004(age) + 0.084(atopy) + 0.003(height in cm), where for sex 1 = male and 0 = female, age is measured in years, for atopy 1 = atopic and 0 = nonatopic, and height is measured in cm.
The FeNO of Chinese adults is higher than that of the Caucasian population, and is affected by age, sex, height and atopic status. This study provides useful references for the interpretation of FeNO.
Exhaled NO fraction in Chinese adults is higher than in Caucasians, and is affected by age, sex, height and atopy http://ow.ly/l5mNR
Noninvasive assessment of airway inflammation has been a popular research topic in recent years. Assessment of biomarkers in exhaled breath, such as fractional exhaled nitric oxide concentration (FeNO) or biomolecules in the exhaled breath condensate, is entirely noninvasive and has great potential for serial monitoring of airway inflammation . There is a growing body of literature on FeNO measurement. As FeNO measurement becomes more important and popular, the American Thoracic Society (ATS) and the European Respiratory Society (ERS) have published guidelines for the measurement of FeNO [2, 3]. FeNO was first demonstrated to be significantly elevated in subjects with asthma two decades ago . Treatment of asthma with inhaled corticosteroids could reduce FeNO in asthmatics . FeNO correlated with other markers of inflammation such as peripheral blood eosinophil count and eosinophil cationic protein level [6, 7]. Some studies have demonstrated that FeNO measurement might be useful for the diagnosis of asthma in adults [8, 9]. FeNO measurement has also been investigated in the treatment algorithm for asthma. However, randomised controlled algorithm asthma control trials revealed equivocal benefits when adding FeNO measurement to the routine guideline management including spirometry [1, 10].
Correct interpretation of FeNO value is important when applying this tool for assessment of airway inflammation. Previous studies on reference values of FeNO in the general population were mainly limited to children [11–13]. There are limited data on the reference value of FeNO in the adult population [14–18], and there is very limited information on FeNO in a healthy, nonsmoking Chinese population. FeNO is a promising noninvasive marker for the assessment of airway inflammation, especially for asthma. Understanding FeNO in the normal population and factors that may affect its level can improve our understanding of the role of FeNO in the clinical management of respiratory diseases. The aim of this study was to establish a reference equation of FeNO in a large healthy Chinese adult population.
This was a cross-sectional observational study in which FeNO was measured in healthy Chinese adults in Hong Kong. Subjects were recruited by advertisements posted at the Chinese University of Hong Kong, Prince of Wales Hospital (Hong Kong, China) and newspapers. Subjects aged between 18 and 90 years who had expressed interest in this study were stratified into different age groups and randomly selected to participate in the study so that each age group of both sexes would have a similar number of subjects. All subjects were not current smokers. Previous smokers must have stopped smoking for at least 1 year with a smoking history of <10 pack-years. Subjects with a history of chronic respiratory diseases like asthma, chronic obstructive pulmonary disease, bronchiectasis, previous lung surgeries, chronic cough and sputum and wheeze over the past year were excluded. The study was approved by the ethics committee of the Chinese University of Hong Kong (approval number: CRE-2006.438) and informed consent was obtained from each subject.
FeNO was measured in the morning (between 09:00 h and 11:00 h) before spirometry. FeNO was measured online using a chemiluminescence analyser (NOA 280i; Sievers Instruments, Boulder, CO, USA) according to ATS/ERS recommendations . Subjects were in the sitting position (with no nose clip); they exhaled to residual volume, inserted a mouth piece, inhaled to total lung capacity, and then exhaled for 10 s at a constant flow rate of 50 mL·s−1. The measurement was repeated until three FeNO values varied <10% or two values varied <5%. The mean FeNO (in ppb) was then recorded. All subjects had to refrain from strenuous physical activity or exercise for ≥30 min prior to FeNO measurement. In addition, subjects avoided eating for 1 h and caffeine ingestion for 6 h before the test. Subjects were not tested within 4 weeks of an upper or lower respiratory tract infection.
Spirometry (pre- and post-bronchodilator) was performed using the MicroLab 3300 spirometer (CareFusion, Basingstoke, UK), according to the ATS standards .
Skin prick test
Allergen skin prick test was performed on the volar aspect of forearm using a panel of eight common aeroallergens according to standard methods . A wheal size 3 mm larger than negative control 15 min after skin prick was considered a positive test. Allergens including Dermatophagoides pteronyssinus and D. farinae, house dust, cockroach, Cladosporidium, Bermuda grass, cat fur and dog hair (ALK-Albelló, Hørsholm, Denmark) were used.
Peripheral blood was taken for measurement of eosinophil count and total IgE level.
Data were categorised and analysed using the Statistical Package for Social Sciences (SPSS) for Windows release 17.0 (SPSS Inc., Chicago, IL, USA). FeNO values were log-transformed before analysis. FeNO is presented as geometric mean and 95% reference interval (RI) (RI was calculated by exponentiating the reference limit obtained from the log-transformed data) and median with interquartile range, as appropriate. The associations between FeNO and anthropometric measurements, spirometric variables and atopic status were assessed by multivariate linear regression and Spearman correlations. The reference equation was computed by multiple linear regression modelling. From the predictors selected a priori, the variables for the final regression model were chosen using backward stepwise regression analysis. The upper 95% cut-off limits were calculated from the regression model based on the whole population, taking into account age, sex, height and atopic status of the subjects using mid-class values in the group. All comparisons were made two-sided, and p-values <0.05 were considered significant.
In total 1113 subjects were recruited for this study. Among these subjects, 1093 (98.2%) were able to perform satisfactory FeNO measurement (i.e. able to maintain the flow rate or produce reproducible results) and were included in the final analyses. Among those 20 subjects (six males and 14 females) who could not perform satisfactory FeNO measurement, nine (45%) and four (20%) were aged >60 and >70 years, respectively. In the 1093 subjects with satisfactory FeNO measurement, 24 (2.2%) had bronchodilator reversibility in the spirometry examination, and only two subjects (0.2%) had post-bronchodilator forced expiratory volume in 1 s <80% predicted normal. The demographic characteristics of the subjects are shown in table 1.
The geometric mean FeNO was 32.6 (95% RI 31.4–33.7) ppb. The range of FeNO was 4.2–315 ppb. FeNO value was higher in males than females (geometric mean (95% RI) 38.3 (36.5–40.2) ppb versus 27.1 (25.8–28.5) ppb, p<0.0001) and higher among atopic when compared with nonatopic subjects (34.6 (33.0–36.3) ppb versus 29.8 (28.3–31.4) ppb, p<0.0001). FeNO correlated with age (r2 = 0.23, p<0.0001), height (r2 = 0.20, p<0.0001), IgE level (r2 = 0.18, p<0.0001) and percentage eosinophil count (r2 = 0.18, p<0.0001). However, FeNO had no correlation with spirometric parameters. We entered parameters that had significant correlations with FeNO into a multivariate linear regression model, and found that sex, age, height, atopic status, serum IgE level and blood percentage eosinophil count were independently associated with FeNO value (table 2).
By putting the demographic parameters, including age, sex and height, together with atopic status by skin prick test, into a multiple linear regression modelling, the reference equation of FeNO in our Chinese adults is as follows:
log(FeNO) = 0.781 + 0.104(sex) + 0.004(age) + 0.084(atopy) + 0.003(height)
where for sex 1 = male and 0 = female, age is measured in years, for atopy 1 = atopic and 0 = nonatopic and height is measured in cm. The intercept value was 0.78, r2 was 0.144 and residual sd was 0.24.
Using this reference equation, we calculated the upper 95% cut-off limits for FeNO, according to height and age (using the mid-class values). The data are presented in table 3 and the corresponding graphic presentation is shown in figure 1 for both atopic and nonatopic subjects.
This has been the first large scale study assessing the FeNO value in the Chinese adult population. We have developed a reference equation for prediction of FeNO value in this population. We have shown that sex, age, height and atopic status by skin prick tests were determinants of FeNO.
Reference ranges for FeNO measured in accordance with the current ATS/ERS standards have been reported previously in children [11–13, 17]. There were studies that assessed the reference equations of FeNO in healthy adults and, among these studies [14–18], only one involved healthy nonsmoking adults as in our study . Many studies involved subjects who were current smokers or ex-smokers, with airway diseases or respiratory tract infections [15, 16, 18, 21, 22]. For example, a study from Germany included current smokers, asthma subjects and subjects with respiratory tract infections (in 24.3%, 3.8% and 20.2%, respectively, among the 897 subjects) . Other large-scale studies from Sweden  and Germany  measured FeNO values in >2000 and >1000 subjects, respectively. In these studies, subjects with physician-diagnosed asthma or those on inhaled steroids , or smokers or patients with respiratory infections were included . A very recent study from the USA reported the FeNO value in >13 000 subjects aged 6–80 years . This study involved children, adults, smokers, subjects with asthma and subjects on asthma medications. Subjects with self-reported asthma or wheezing, or who had prescriptions for asthma/wheezing in the past 12 months, were excluded when normal values and thresholds were calculated. Since the prediction model equations in this study were derived for age groups of 6–11 years and 12–80 years, teenagers <18 years of age were also included in the older age group .
When compared with the Caucasian population, adult Chinese subjects had a higher geometric mean FeNO value. table 4 summarises the comparison of FeNO of normal adults without known lung diseases between different populations. Compared with the study performed in Sweden by Olin et al.  with 1131 subjects, our geometric mean FeNO was much higher (32.6 ppb versus 16.6 ppb). This ethnic difference was consistent with our previous observation in paediatric subjects involving 258 local and 33 Caucasian students with a mean age of 14 years in Hong Kong. We found that their mean FeNO was 25.3, 15.8, 14.9 and 10.1 ppb for Chinese boys, Chinese girls, Caucasian boys and Caucasian girls, respectively . Another large-scale paediatric study in Canada involving 656 school children aged 9–12 years also found that Asian-Canadian subjects had a higher FeNO value than white subjects and the African-Canadian subjects . Furthermore, a study involving 62 children in the UK found that FeNO was significantly higher, after correcting for atopic status, by an average of 36% in South Asians when compared to the white subjects . A recent study from Taiwan involving 681 Asian children aged 5–18 years found that the geometric mean FeNO and the upper 95% CI were 13.7 and 29.7 ppb, respectively, and this was also higher when compared with the Caucasian population . In fact, other smaller-scale studies involving Chinese and Korean adults also showed a similar level of FeNO value to our study (table 4) [25–27]. A study involving 895 African-Americans noted that the mean FeNO for males and females were 27±26 ppb and 18±18 ppb, respectively . The FeNO value of this group of African-American subjects appeared to be greater than the Caucasian population but less than that of the Asian population. As this study involved asthma subjects and subjects with respiratory tract infection (current/past week), it is thus difficult to compare their results directly with “healthy” subjects of other populations or ethnicity . Another large US study involving Hispanic, white, black and subjects of other ethnicities also found that using white subjects as reference, other ethnic groups had higher FeNO value. However, the composition of the other ethnic groups was unspecified in this study . There is also some smaller-scale study of FeNO value in normal healthy subjects of other populations .
Our study found that male subjects had a higher FeNO value than female subjects and atopy was associated with higher FeNO value. In addition, height and age were both positively associated with the FeNO level. Furthermore, these factors all had independent association with the FeNO value as multivariate analysis with adjustment of the other factors found a statistically significant association. Age and height were important factors for FeNO in children [12, 13]. The effect of age and height affecting FeNO in adults is less consistent, with studies showing conflicting results [14, 15]. Concerning the effect of sex on FeNO, our study noted a major difference between sexes, similar to the results reported by Travers et al.  and Taylor et al. . However, this was not observed in the study by Olin et al. . There is also a controversial relationship between FeNO and atopy. Some studies have suggested that atopy, such as the number of positive skin prick tests and total IgE level, may affect FeNO value [34, 35], whereas other studies suggested that FeNO value was not influenced by atopy . Concerning the upper limit of normal of the FeNO value, our study found that the upper limits of FeNO value ranged from 19 to 62 ppb, depending on age, sex, height and atopic status. There was only one study that assessed the FeNO value in a large population of healthy adults, and that found that the upper limits of FeNO ranged from 24.0 to 54.0 ppb, depending on age, height and atopy . When comparing this Swedish study with our study, apart from the difference in ethnicity, their subjects were older, taller and had less atopy. Difference in body build and height can affect lung volume and this may also account for the variations of FeNO value among different populations.
Apart from the demographic factors of the subjects (such as age and ethnicity), other factors may also affect FeNO level. A previous study found that exhaled NO measurements in healthy subjects and patients with airways disease differed according to the type of analyser used . Their study compared three brands of machines (Ecomedics (Dürnten, Switzerland), NIOX (Aerocrine AB, Solna, Sweden) and Logan Research Ltd (Rochester, UK)). Sievers, the brand of machine used in this study, was not assessed. It was thus not certain whether the difference in the FeNO level in our population when compared to other studies would be attributable to machine difference. This factor might not be very significant as all current machines should follow the ATS/ERS guidelines [2, 3] for calibration and maintenance. Conversely, a study from Singapore using the NIOX machine in testing 45 first-year Asian medical students and another study from Korea involving 166 adults aged 20–80 years using the same Sievers machine as in this study , found similar FeNO values to our study .
We speculated that differences in environmental exposures or genetic polymorphisms of high-producing nitric oxide synthase genotypes would be related to the FeNO level in different populations. The production of endogenous nitric oxide from l-arginine is dependent on the enzyme NO synthase (NOS). All NOS isoenzymes convert l-arginine to l-citrulline, with the generation of NO. Three isoforms of NOS are known. NOS1 and NOS3 are both constitutively expressed in the human airway, whereas NOS2 is inducible by inflammation . It is possible that differences in genetic background may affect the activity of NOS enzymes resulting in differences in NO production. A population-based study of young adult twins revealed that variation in FeNO was explained by genetic and nonshared environmental effects . A previous study noted that the NOS3 missense sequence variant in the endothelial NOS gene (G894T) was associated with FeNO in an American cohort of subjects with asthma. The TT genotype had a significantly higher FeNO than the GT genotype . An inhibitor of iNOS was noted to produce marked inhibition of FeNO in both normal and asthmatic subjects . Further studies are needed to assess the genetic contributions of FeNO value in different ethnicities. Apart from genetic factors, environmental factors such as air pollution, allergens and diet may also affect the FeNO level, but we have not assessed these factors in this study [42–45].
This study had some limitations. FeNO was measured on just one occasion for the subjects and repeatability was not tested. In addition, we recruited volunteers to join the study and the subjects were not randomly recruited from a large population or community. The percentage of atopic subjects in this study was on the high side and this might be due to selection bias.
In summary, we have shown that sex, age, atopic status and height were determinants of FeNO value in a large population of Chinese subjects. FeNO value in the Chinese subjects appeared to be higher than that of the Caucasian population. In this population, the upper limits of FeNO ranged from 19 to 62 ppb, depending on their age, sex, height and atopic status. Further studies are needed to explore the genetic and other determinants of FeNO. This study provides useful population-based reference values for the accurate interpretation of FeNO levels in the Chinese population.
Support statement: This study is supported by a Hong Kong Lung Foundation Research Grant and the Respiratory Research Fund of the Chinese University of Hong Kong.
Conflict of interest: None declared.
- Received August 20, 2012.
- Accepted November 6, 2012.
- ©ERS 2013