Reproducibility of exhaled nitric oxide in smokers and non-smokers: relevance for longitudinal studies
© Bohadana et al; licensee BioMed Central Ltd. 2008
Received: 12 October 2007
Accepted: 28 February 2008
Published: 28 February 2008
Currently, there is much interest in measuring fractional exhaled nitric oxide (FE NO ) in populations. We evaluated the reproducibility of FE NO in healthy subjects and determined the number of subjects necessary to carry out a longitudinal survey of FE NO in a population containing smokers and non-smokers, based on the assessed reproducibility.
The reproducibility of FE NO was examined in 18 healthy smokers and 21 non-smokers. FE NO was assessed once at 9 AM on five consecutive days; in the last day this measurement was repeated at 2 PM. Respiratory symptoms and medical history were assessed by questionnaire. The within- and between-session repeatability of FE NO and log-transformed FE NO was described. The power of a longitudinal study based on a relative increase in FE NO was estimated using a bilateral t-test of the log-transformed FE NO using the between-session variance of the assay.
FE NO measurements were highly reproducible throughout the study. FE NO was significantly higher in males than females regardless of smoking status. FE NO was positively associated with height (p < 0.001), gender (p < 0.034), smoking (p < 0.0001) and percent FEV1/FVC (p < 0.001) but not with age (p = 0.987). The between-session standard deviation was roughly constant on the log scale. Assuming the between-session standard deviation is equal to its longitudinal equivalent, either 111 or 29 subjects would be necessary to achieve an 80% power in detecting a 3% or a 10% increase in FE NO respectively.
The good reproducibility of FE NO is not influenced by gender or smoking habits. In a well controlled, longitudinal study it should allow detecting even small increases in FE NO with a reasonable population size.
Fractional exhaled nitric oxide (FE NO ) is now widely used as a surrogate marker for eosinophilic airway inflammation . Yet, several personal and environmental factors including active or passive smoking may influence FE NO thus acting as causes of bias .
Good reproducibility is an important precondition for using a physiological test both for clinical and epidemiological purposes . While the variability and reproducibility of FE NO have been evaluated in patients with asthma, studies in healthy subjects usually excluded smokers [4–7]. Since exposure to cigarette smoke decreases FE NO [8–13] a given instrumental or biological variation will yield higher coefficients of variation if the denominator is lower.
Currently, there is much interest in assessing longitudinal changes in FE NO levels in populations. Ensuring a low variability and good reproducibility in this setting is critical in distinguishing true biological changes from other sources of variation. In this study, we examined the influence of smoking status and gender on the reproducibility of FENO measurements in a group of healthy individuals. Then, based on the reproducibility, we determined the sample size necessary to carry out a longitudinal study on the evolution of FE NO , a matter on which there is virtually no information in the literature.
Subjects and Protocol
Thirty-nine young subjects were recruited among clerks, technicians, engineers, postgraduate students and medical residents from the University of Nancy. There were 21 non-smokers (10 males and 11 females) and 18 smokers (9 males and 9 females). The selection criteria were a negative history of respiratory symptoms and allergic diseases (e.g. rhinitis, eczema), and normal spirometry. Subjects with acute respiratory infection in the last 4 weeks were excluded. None of the volunteers was taking any drug or medication. All subjects gave informed consent and the research protocol was approved by our local ethics in medical research committee.
Subjects were approached a first day when a brief respiratory symptoms questionnaire was administered through a face-to-face interview and pulmonary function tests were performed. The questionnaire covered personal and demographic information, past and present chest diseases, respiratory symptoms and allergies, and past and present smoking habits. After inclusion, six sessions of three measurements were planned over one week. All subjects had their FE NO measured once on five consecutive days, from Monday through Friday, between 09:00–10:00 h during visits 1, 2, 3 and 4; during the last visit FE NO was measured twice between 09:00–10:00 h and 14:00–15:00 h.
Spirometry was carried out in the sitting position, using an electronic spirometer (Oscilolink, Datalink, Montepellier, France). Forced vital capacity (FVC) and forced expiratory volume in one second (FEV1) were obtained by having the subject expire forcefully after a maximal inspiratory maneuver. At least three forced expiratory maneuvers, satisfactory according to recommended criteria  were recorded. The curve with the highest sum of FVC+ FEV1 was used for analysis. Curves with an FEV1/FVC ratio of 80% or higher and a predicted  FEV1 ≥ 75% were considered as normal.
FE NO was measured according to the ATS/ERS recommendations . Measurements were done using a chemiluminescence analyzer (NIOX® 2.0 system; Aerocrine AB, Solna, Sweden). The subject was in the sitting position and exhaled against an oral pressure of 5 cm H20 – sufficient to close the velum – at a flow rate of 50 mL/s. At each session, three correctly performed exhalations were recorded. Any exhalation which did not meet the ATS/ERS requirements was rejected by the NIOX system and the subject was asked to perform a new exhalation maneuver. Calibrations were performed every 13 days according to the standards of the manufacturer. Subjects avoided eating for 1 hour and smoking for 8 hours before testing.
Statistical analysis was performed using the Stata package . The Shapiro-Wilks test was used to check the normality and log-normality of the measurements. Reproducibility of FE NO measurements was assessed by the Bland Altman analysis . Within-session repeatability was assessed by plotting the difference between each of the three measurements of a session and their session mean against the session mean. This procedure generated three data points by subject by session, thus yielding 702 data points (39 subjects × 3 measures by session × 6 sessions = 702). Between-session repeatability was calculated by plotting the difference between a mean session FENO and the overall individual mean against the overall individual mean both on the natural and log-transformed (on base 10) basis. This procedure generated 1 data point by subject by session thus yielding 234 data points (39 subjects × 6 sessions = 234). The possible learning effect was assessed by comparing the 6 successive sessions and the morning sessions on day 1 and day 5 using a repeated measure analysis of variance. In turn, the diurnal variation was assessed by comparing, using a matched t-test, the morning session vs. the afternoon session on day 5.
A multiple linear regression model was used to assess the association between log-transformed FE NO , taken as the dependent variable, and other, independent variables including age, height, sex, smoking habits and spirometric variables.
A sample size determination was calculated to achieve an 80% power in a study in which FE NO would be measured twice e.g. before and after exposure to some risk factor. The statistical test is based on the individual difference of log-transformed values at time 2 and time 1 tested to be equal to zero by a standard two-sided Student test. In other words, it tests an increase in percent of the initial measurement. The expected standard deviation of this difference of (log-transformed) measurements is estimated using the within-subject between-session standard deviation of the log-transformed mean of the three measurements which is assumed to be equal to the individual long-term variability.
Baseline characteristics of participants. Values are mean (SD).
Nb of subjects
BMI, kg/m 2
Geometrical mean (ppb) and geometrical SD of FENO measurement on six occasions in smokers and non-smokers stratified by gender.
Men (n = 9)
Women (n = 9)
Men (n = 10)
Women (n = 10)
Multiple linear regression model with log-FENO as the dependent variable in relation to age, height, gender, smoking status and spirometry among 39 healthy subjects.
[95% confidence interval]
Gender: Men vs. Women
Smoker vs. Non-smoker
FEV 1 /FVC (% predicted)
In this study, we demonstrate that FE NO measurements are reproducible, a finding consistent with results reported previously in healthy subjects and patients with asthma [4–7]. Yet, we provide new knowledge by showing that the reproducibility of FE NO is not influenced by gender nor adversely affected by smoking status in smokers refraining from smoking before testing. In addition, we show that significant changes in FE NO are potentially detectable with fairly small numbers of subjects, an aspect of practical importance when planning epidemiological studies. However, it should be pointed out that our calculations were based on the assumption that the variance of the mean FE NO measurement within the week in which the essay was conducted represents the intrinsic long-term variability of FE NO in absence of any external factor influencing it. Although this hypothesis seems reasonable – as supported by the remarkable seasonal stability of FE NO  – only a much longer follow up of this measurement in a similar population could validate it. Finally, it should be stressed that this reasoning does not apply to managing individual asthmatics in the clinical setting.
Our study differed from those quoted above [4–7] in three ways. First, we examined males and females separately, while some authors preferred to examine males only  or merge males and females together [6, 7]; one study did not mention the subjects' gender . Second, we examined smokers and non-smokers of both sexes separately, while the above mentioned studies [4–6] did not take smoking status into account. Finally, our series included only healthy subjects while other series included asthmatics as well [5, 6].
For the sake of comparability, we used a protocol similar to that of Kharitonov and colleagues . The within-session standard deviation was close to that noted by these authors (1.8 ppb vs. 2.1 ppb) although it increased with the mean FE NO . However, in contrast to Kharitonov et al, we did not exclude any a priori outliers as it would lead to removing almost all measurements of subjects with high FE NO values. If such procedure were applied in any clinical or epidemiological study it would imply removing the most interesting subjects.
By plotting the distance between the session means and the overall subject specific mean vs. the latter we showed the within-subject between-session variance increased with the subject mean; this feature disappeared when using log-transformed measurements, which we would therefore recommend as the relevant variable. This means simply that any evolution of a subject's FE NO is better expressed in percent of the initial value rather than as an increase or decrease in the absolute value.
The relationship between gender and FE NO has been examined previously. Olin and colleagues  measured FE NO in a large random adult general population sample (n = 2,200) and found that when both height and gender were included in multiple regression model the contribution of gender was not significant. They postulated that the alleged association FE NO /male gender was probably due to a height-dependent increase in the total airway mucosal surface area that produces NO. However, conflicting results were published by others. Olivieri and colleagues  measured FE NO in 204 healthy, non-smoking, non-atopic individuals and documented significantly higher FE NO values in men compared with women. Travers and colleagues  examined subjects aged between 25 and 75 years and found that sex, atopy and smoking status affected FE NO levels; they presented reference ranges adjusted for these factors. More recently, Robin Taylor and colleagues  examined 895 adults aged 32 and found that FE NO levels were nearly 25% less in females and suggested that reference values should be stratified by sex. In the present study we also found that women had significantly lower FE NO than men and that the association between FE NO and gender remained significant when height was included in the regression model.
There is evidence that genetic factors might play a role on the genesis of the reported gender-related differences in FE NO . Graserman and colleagues  studied NO and genetic variants in NO synthases in 105 healthy non-smoking and smoking subjects. They found FE NO to be significantly higher in males than females among both non-smokers and smokers. More importantly, they noted that healthy non-smoking females with greater number of repeats in neuronal NO synthase had significantly lower NO levels than did females with fewer numbers of repeats. They concluded that variants in the neuronal NO synthase gene contribute to the variability of airway NO concentrations in healthy females. More recently, Lund and colleagues  examined 377 adult twins identified through the Norwegian Twin Registry and showed that genetic effects accounted for 60% of the variation in FE NO . The influence of genetic factors on our results is not known and clearly more studies are necessary to clarify this issue.
Our study confirms the well-established association between cigarette smoking and FE NO levels [8–13]. The mechanisms by which smoking reduces FE NO are not completely understood. Several possibilities have been proposed including (i) a reduction of the endogenous NO synthesis by feedback inhibition due to the high concentrations of NO contained in the cigarette smoke (ii) an inadequate supply of cofactors necessary for NO production  and (iii) an increase in the breakdown of NO . Recently, Malinovschi and colleagues  examined the effect of past, current and passive smoking on FE NO in a general population sample using flow-independent NO exchange parameters. They found lower FE NO levels in current smokers and attributed this to reduced NO levels in both the airways and alveoli. Conversely, Pietropaoli and colleagues  measured FE NO at different expiratory flow rates in asymptomatic smokers and age-matched non-smokers and concluded that the diminished NO expired by smokers resulted from diminished NO production by the tissues of the conducting airways and not by the alveoli. Whatever the mechanism, there appears that FE NO levels decrease with increasing years smoked and with recent smoking . Had it been present in our study, the latter factor could have produced variations in FE NO levels and adversely affected our reproducibility. However, to prevent such possibility and achieve stability we asked our subject to avoid smoking for 8 hours before FE NO measurements.
Atopy might have influenced baseline FE NO levels of our subjects [19, 28, 29] although we are not aware of such an effect in absence of any airway inflammation. However, while the NO level of atopic subjects can be expected to increase specifically under the influence of a risk factor , in absence of any such factor, there is no reason it should vary considerably. Excluding atopic subjects in a longitudinal study would thus bias the results of that study towards the absence of effect. By consequence, we decided not to screen our subjects for asymptomatic atopy (e.g. by performing skin prick tests to common allergens) nor eliminate outliers.
A final word must be said about our protocol. As with other tests, FE NO measurement may be influenced by technical and biological sources of variation. To deal with the former, all measurements were carried out by the same operator, using the same instrument – which met the ATS criteria for precision and accuracy  – according to a standardized procedure. Further, care was taken to ensure the subjects' comprehension, comfort and cooperation during the procedure. As for biological sources, we took care to record carefully the subjects' recent exposures and activities. In addition, we dealt with the possible role of circadian rhythms by asking each subject to come to the laboratory every day at the same hour. Incidentally, our average FE NO values were within the limits of reported reference values established using NO analyzers of the same manufacturer [21, 31]. For smokers, our values are below both the upper limit of 36.4 ppb for females and the upper limit of 45.9 ppb for males proposed by Travers and colleagues . For non-smokers, our values are similar to the 31.2 ppb value predicted for men and below the 29.7 ppb value predicted for women by Olin and colleagues  and far below the upper limit of 44.6 and 56.5 ppb proposed respectively for women and men by Travers and colleagues .
In summary, this study improves our understanding of the reproducibility of FE NO in two ways. First, it shows that this reproducibility is not influenced by gender or smoking status in subjects refraining from smoking before FE NO assessment. Second, it provides evidence that FE NO is potentially useful in the survey of populations, fairly small samples being necessary to assess significant changes in FE NO levels. This finding could be of practical importance for the survey of populations at risk of asthma since traditional means of monitoring such as tests of lung function and bronchial provocation tests are not directly related to airway inflammation. Further studies, carried out over longer periods of time, would be necessary to confirm the assumption that the short-term variance of FE NO represents its intrinsic long-term variability.
- FENO :
Fractional exhaled nitric oxide
- FEV1 :
Forced expiratory volume in one second
Forced vital capacity.
The authors thank the volunteers for their kind participation.
- Taylor DR, Pijnenburg NW, Smith AD, Jongste DC: Exhaled nitric oxide measurements: clinical application and interpretation. Thorax. 2006, 61: 817-27. 10.1136/thx.2005.056093.View ArticlePubMedPubMed CentralGoogle Scholar
- Kharitonov S, Barnes PJ: Clinical aspects of exhaled nitric oxide. Eur Respir J. 2000, 16: 781-792. 10.1183/09031936.00.16478100.View ArticlePubMedGoogle Scholar
- Butler J: The pulmonary function test – Cautious overinterpretation. Chest. 1981, 79: 498-450. 10.1378/chest.79.5.498.View ArticlePubMedGoogle Scholar
- Purokivi M, Randell J, Hirvonen MR, Tukiainen H: Reproducibility of measurements of exhaled NO, cell count and cytokine concentrations in induced sputum. Eur Respir J. 2000, 16: 242-246. 10.1034/j.1399-3003.2000.16b10.x.View ArticlePubMedGoogle Scholar
- Ekroos H, Karjalainen J, Sarna S, Laitinen LA, Sovijarvi ARA: Short-term variability of exhaled nitric oxide in young male patients with mild asthma and healthy subjects. Respir Med. 2002, 96: 895-900. 10.1053/rmed.2002.1378.View ArticlePubMedGoogle Scholar
- Kharitonov SA, Gonio F, Kelly C, Meah S, Barnes PJ: Reproducibility of exhaled nitric oxide measurements in healthy and asthmatic adults and children. Eur Respir J. 2003, 21: 433-438.View ArticlePubMedGoogle Scholar
- Stark H, Purokivi M, Kiviranta J, Randell J, Tukiainen H: Short-term and seasonal variations of exhaled and nasal NO in healthy subjects. Respir Med. 2007, 101: 265-271. 10.1016/j.rmed.2006.05.009.View ArticlePubMedGoogle Scholar
- Schilling J, Holzer P, Guggenbach M, Gyurech D, Marathia K, Geroulanos S: Reduced endogenous nitric oxide in the exhaled air of smokers and hypertensives. Eur respir J. 1994, 7: 467-471. 10.1183/09031936.94.07030467.View ArticlePubMedGoogle Scholar
- Kharitonov SA, Robbins RA, Yates D, Keatings V, Barnes PJ: Acute and chronic effects of cigarette smoking on exhaled nitric oxide. Am J Respir Crit Care Med. 1995, 152: 609-612.View ArticlePubMedGoogle Scholar
- Malinovschi A, Janson C, Holmkvist T, Norback D, Merilainen P, Hogman M: Effect of smoking on exhaled nitric oxide and flow-independent nitric oxide exchange parameters. Eur Respir J. 2006, 28: 339-345. 10.1183/09031936.06.00113705.View ArticlePubMedGoogle Scholar
- Persson MG, Zetterstrom O, Agrenius V, Ihre E, Gustafsson LE: Single-breath nitric oxide measurements in asthmatic patients and smokers. Lancet. 1994, 343: 146-147. 10.1016/S0140-6736(94)90935-0.View ArticlePubMedGoogle Scholar
- Verleden GM, Dupont LJ, Verpeut AC, Demets MG: The effect of cigarette smoking on exhaled nitric oxide in mild steroid-naïve asthmatics. Chest. 1999, 116 (1): 59-64. 10.1378/chest.116.1.59.View ArticlePubMedGoogle Scholar
- McSharry CP, McKay IC, Cahudhuri R, Livingston E, Fraser I, Thomson NC: Short and long-term effects of cigarette smoking independently influence exhaled nitric oxide concentration in asthma. J Allergy Clin Immunol. 2005, 116: 88-93. 10.1016/j.jaci.2005.03.025.View ArticlePubMedGoogle Scholar
- Standardization of Spirometry, 1994 Update. American Thoracic Society. Am J Respir Crit Care Med. 1995, 152: 1107-1136.
- Quanjer Ph H, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault J-C: Lung volumes and forced respiratory flows. Report working party. Standardization of Lung Function Tests. European Community for Coal and Steel. Eur Respir J. 1993, 6 (Suppl 16): 5-40.View ArticlePubMedGoogle Scholar
- ATS/ERS: Recommendations for Standardized Procedures for the Online and Offline Measurement of Exhaled Lower Respiratory Nitric Oxide and Nasal Nitric Oxide. Am J Respir Crit Care Med. 2005, 171: 912-930. 10.1164/rccm.200406-710ST.View ArticleGoogle Scholar
- Statistical software, release 9.0. 2005, Stata Corporation, 702 University Drive, College Station, TX 77840
- Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986, i: 307-310.View ArticleGoogle Scholar
- Olin AC, Rosengren A, Thelle DS, Lissner L, Bake B, Toren K: Height, age and atopy are associated with fraction of exhaled nitric oxide in a large adult population sample. Chest. 2006, 130: 1319-1325. 10.1378/chest.130.5.1319.View ArticlePubMedGoogle Scholar
- Olivieri M, Talamini G, Corradi M, et al: Reference values for exhaled nitric oxide (reveno) study. Resp Res. 2006, 30: 94-99. 10.1186/1465-9921-7-94.View ArticleGoogle Scholar
- Travers J, Marsh S, Aldington S, et al: Reference ranges for exhaled nitric oxide derived from a random community survey of adults. Am J Respir Crit Care Med. 2007, 176: 238-242. 10.1164/rccm.200609-1346OC.View ArticlePubMedGoogle Scholar
- Robin Taylor D, Mandhane P, Greene JM, Hancox RJ, Filsell S, McLachlan CR, Willimason AJ, Cowan JO, Smith AD, Sears MR: Factors affecting exhaled nitric oxide measurements: the effect of sex. Resp Res. 2007, 8: 82-10.1186/1465-9921-8-82.View ArticleGoogle Scholar
- Graserman H, van's Gravesande KS, Büscher R, Drazen JM, Ratjen F: Effects of sex and of gene variants in constitutive nitric oxide sysnthase on exhaled nitric oxide. Am J Respir Crit Care Med. 2003, 167: 1113-1116. 10.1164/rccm.200211-1342OC.View ArticleGoogle Scholar
- Lund MB, Kongerud J, Nystad W, Boe J, Harris JR: Genetic and environmental effects on exhaled nitric oxide and airway responsiveness in a population-based sample of twins. Eur Respir J. 2007, 29: 292-298. 10.1183/09031936.00044805.View ArticlePubMedGoogle Scholar
- Higman DJ, Strachan AM, Buttery L, Hicks RC, Springall DR, Greenhalgh RM, Powell JT: Smoking impairs the activity of endothelial nitric oxide synthase in saphenous vein. Artherioscler Thromb Vasc Biol. 1996, 16: 546-552.View ArticleGoogle Scholar
- Balint B, Donnely LE, Hanazawa T, Kharitonov SA, Barnes PJ: Increased nitric metabolites in exhaled breath condensate after exposure to tobacco smoke. Thorax. 2001, 56: 456-461. 10.1136/thorax.56.6.456.View ArticlePubMedPubMed CentralGoogle Scholar
- Pietropaoli AP, Perillo IB, Perkins PT, Frasier LM, Speers DM, Frampton MW, Utell MJ, Hyde RW: Smokers have reduced nitric oxide production by conducting airways but normal levels in the alveoli. Inhal Toxicol. 2007, 19: 533-41. 10.1080/08958370701260673.View ArticlePubMedGoogle Scholar
- Gratziou C, Dassiou M, Lignos M, Roussos C: Exhaled nitric oxide and bronchial hyperreactivity in atopic and non-atopic patients. Am J Respir Crit Care Med. 1998, 157: A611-Google Scholar
- Ho LP, Wood FT, Robson A, Innes JA, Greening AP: Atopy influences nitric oxide levels in adults. Chest. 2000, 118: 1327-1331. 10.1378/chest.118.5.1327.View ArticlePubMedGoogle Scholar
- Olin AC, Alving K, Toren K: Exhaled nitric oxide: relation to sensitization and respiratory symptoms. Clin Exp Allergy. 2004, 34: 221-226. 10.1111/j.1365-2222.2004.01888.x.View ArticlePubMedGoogle Scholar
- Olin AC, Bake B, Toren K: Fraction of exhaled nitric oxide at 50 mL/s. Reference values for adult lifelong nevers-smokers. Chest. 2007, 131: 1852-1856. 10.1378/chest.06-2928.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2466/8/4/prepub