Exhaled breath markers in patients with obstructive sleep apnoea
- First Online:
- Cite this article as:
- Petrosyan, M., Perraki, E., Simoes, D. et al. Sleep Breath (2008) 12: 207. doi:10.1007/s11325-007-0160-8
- 245 Views
The objectives of the present study were to assess the level of exhaled breath markers indicating airway inflammation and oxidative stress in patients with obstructive sleep apnoea syndrome (OSAS) in comparison with non-apnoeic (obese and non-obese) subjects and investigate whether therapy with continuous positive airway pressure (CPAP) can modify them. The design was a retrospective observational study, set in Evgeneidio Hospital. Twenty-six OSAS patients and nine obese and 10 non-obese non-apnoeic subjects participated in this study. We measured nasal nitric oxide (nNO), exhaled nitric oxide (eNO), exhaled carbon monoxide (eCO) in exhaled breath, and 8-isoprostane, leukotriene B4 (LTB4), nitrates, hydrogen peroxide (H2O2), and pH in exhaled breath condensate (EBC) before and after 1 month of CPAP therapy. The levels of eNO and eCO were higher in OSAS patients than in control subjects (p < 0.05). Nasal NO was higher in OSAS patients than in obese controls (p < 0.01). The level of H2O2, 8-isoprostane, LTB4, and nitrates were elevated in OSAS patients in comparison with obese subjects (p < 0.01). Conversely, pH was lower in OSAS patients than in non-apnoeic controls (p < 0.01). One month of CPAP therapy increased pH (p < 0.05) and reduced eNO (p < 0.001) and nNO (p < 0.05). Apnea/hypopnoea index was positively correlated with 8-isoprostane (r = 0.42; p < 0.05), LTB4 (r = 0.35; p < 0.05), nitrates (r = 0.54; p < 0.001), and H2O2 (r = 0.42; p < 0.05). Airway inflammation and oxidative stress are present in the airway of OSAS patients in contrast to non-apnoeic subjects. Exhaled breath markers are positively correlated with the severity of OSAS. One-month administration of CPAP improved airway inflammation and oxidative stress.
KeywordsObstructive sleep apnoeaExhaled breathExhaled breath condensateCPAP
Obstructive sleep apnoea syndrome (OSAS) occurs in ∼9% of men and 4% of women in the middle-aged American population, and it is characterised by recurrent episodes of upper airway collapse that lead to significant hypoxemia and disturbed sleep architecture . OSAS is considered to be an important risk factor for cardiovascular morbidity, although the exact mechanisms involved have not been fully elucidated [2, 3]. Thus, hypoxia/reoxygenation phenomenon deriving from the repeated episodes of cessation of breathing has been proposed to represent a form of oxidative stress leading to increased generation of reactive oxygen species that could injure the vascular endothelium . Furthermore, it has been suggested that the up-regulation of inflammatory mediators induced by recurrent obstructive apnoeas may lead to cardiovascular disease .
Nonetheless, confounders such as obesity, comorbidities, and/or medications hinder the adequate assessment of the presence of oxidative stress and inflammation in OSAS . Prior studies have provided inconsistent results regarding the presence of increased oxidative stress in OSAS [7–10] or have not used control subjects matched closely for body mass index and obesity [11–13], and consequently, any conclusion drawn remains controversial.
Oxidative stress and inflammation of the respiratory tract can be non-invasively monitored by measuring exhaled breath markers . Such techniques are simple to perform, may be repeated frequently, and can be applied in patients in whom invasive procedures are not possible because the manoeuver does not affect the airway function. Thus, abnormalities in exhaled breath markers reflect intrinsic fluctuations of the airway lining fluid caused by inflammation and oxidative stress and may assist in the monitoring of disease severity and the response to treatment, and accordingly in a better design of therapeutic approaches .
Therefore, the purpose of this study was to assess the “exhaled breath profile” using combined measurements of exhaled nitric oxide, carbon monoxide and breath condensate markers of airway inflammation and oxidative stress in patients with OSAS in comparison with obese and non-obese, non-apnoeic subjects. Furthermore, the study aimed to document whether these markers can be useful in monitoring the severity of OSAS before and after continuous positive airway pressure therapy.
Materials and methods
Thirty-five consecutive subjects referred to the Center of Sleep Disorders of “Evgeneidio” Hospital (Athens, Greece) during a period of 32 months participated in the study. The enrollment criteria were (1) no smoking for the previous 12 months, (2) no upper or lower respiratory tract disease (e.g. upper respiratory tract infection, rhinitis, sinusitis, bronchitis, bronchial asthma, tuberculosis, lung tumours, sarcoidosis, cystic fibrosis) including a history of nasal allergy, (3) no recent surgery involving upper airway, (4) no systemic infectious disease, (5) no use of medications known to influence exhaled breath markers (as nitrous substances, oral or nasal or inhaled steroids, anti-inflammatory drugs, l-arginine, oxymethasoline, vaccination for influenza, statins, vitamin C or E, etc.), and (6) no history of cardiovascular or renal disease. The subjects were divided into two groups according to the apnoea/hypopnoea index (AHI). Twenty-six subjects had AHI greater than 20 events per hour and created the OSAS group and nine subjects had AHI less than 5 events per hour and created the non-apnoeic obese control group. Another 10 lean (body mass index less than 24 kg·m−2) individuals who were hospital employees, fulfiled the aforementioned enrollment criteria and gave a history of normal sleep habits without snoring formed the non-apnoeic non-obese control group. The study protocol was approved by the hospital ethics committee, and all subjects gave their written informed consent prior to enrollment in the study.
In all subjects, at their first visit, exhaled nitric oxide (eNO), nasal nitric oxide (nNO) and exhaled carbon monoxide (eCO) were measured and exhaled breath condensate was collected. Then, each subject underwent pulmonary function testing and arterial blood gas analysis. A full-night diagnostic polysomnography was subsequently performed.
Patients with OSAS (AHI >20 events per hour) underwent a second overnight polysomnography with nasal continuous positive airway pressure (CPAP) titration. Optimal CPAP pressure was defined as the pressure required to eliminate snoring and flow limitation. Following titration, patients were discharged home for 4 weeks of continued CPAP treatment on the titrated pressure. Members of the research staff were in frequent contact with the participants to answer questions about mask placement and to encourage compliance with the therapy. After 4 weeks of CPAP administration, during which compliance was measured nightly by an internal clock counter in the CPAP unit and averaged over treatment period (mean daily duration of nCPAP use, 4.3 ± 1.7 h), patients underwent a third overnight polysomnography. Exhaled breath markers were then measured again.
A full-night diagnostic polysomnography (EMBLA S7000, Medcare Flaga, Reykjavik, Iceland) was performed in each subject. To determine the stages of sleep, electro-encephalogram (C4–A1, C3–A2, O2–A1, O1–A2), electro-oculogram and electro-myogram of the submentalis muscle were obtained. Arterial blood oxyhaemoglobin was recorded with the use of a finger pulse oximeter. Thoracoabdominal excursions were measured qualitatively by respiratory effort sensors [XactTrace belts featuring Respiratory Inductive Plethysmography (RIP), Medcare Flaga] placed over the rib cage and abdomen. Snoring was detected with a vibration snore sensor and body posture with a body position sensor. Airflow was monitored using an oral thermistor placed in front of the mouth and a nasal cannula/pressure transducer inserted in the opening of the nostrils.
Sleep stage was scored manually in 30-s epochs . Respiratory events were scored using standard criteria by an experienced technician . Thus, an apnoea was defined as the absence of airflow for more than 10 s in the presence of continued respiratory effort . Hypopnoeas were defined as a reduction in the chest wall movement to an amplitude that was smaller than approximately 70% of the baseline level and lasting more than 10 s, leading to a 4% decrease of haemoglobin saturation or an arousal . The number of episodes of apnoeas and hypopnoeas per hour of sleep is referred to as the apnoea/hypopnoea index (AHI).
Pulmonary function testing and arterial blood gases
Pulmonary function test was performed on the same day of the measurement of the breath markers. FEV1, FVC, FEV1/FVC ratio and DLCO were measured using a spirometer (Masterscreen, Jaeger, Wurzburg, Germany). Arterial blood gases (PaO2, PaCO2), pH and haematocrite were measured by an automatic acid–base analyser (ABL555, Radiometer, Copenhagen, Denmark).
Measurement of breath markers
Exhaled and nasal nitric monoxide and eCO were measured by a LR2000 chemiluminenscence analyser (Logan Research, Rochester, UK) on the same day at a specific time (between 11 a.m. and 13 p.m.) to minimise diurnal variation. The analyser had nitric monoxide sensitivity ranging from 1 to 5,000 parts per billion (p.p.b.) and was calibrated daily with NO/N2 calibration gas containing 116 p.p.b. In addition to NO, this analyser measures carbon dioxide concentration, pressure and volume in real time. The subjects exhaled slowly from total lung capacity over 30 ± 40 s through a mouthpiece against a mild resistance to prevent nasal contamination. All subjects maintained an exhalation flow rate of 250 mL/s. NO was sampled from a side arm attached to the mouthpiece. The mean value was taken from the point corresponding to the plateau of end-exhaled CO2 reading (5 ± 6% CO2) as representative of low respiratory tract sample. The results of the analysis were computed and graphically displayed on plots of NO and CO2 concentrations, pressure and flow rate against time.
Nasal nitric monoxide (nNO) was measured with a teflon tube inserted into one of the nostrils, whilst the subject held his/her breath (i.e. with no active exhalation), after having inspired to total lung capacity. This allowed analysis of the local NO concentration with free flow of ambient air from one nostril to the other and subsequent direction into the analyser. The NO level was measured when a plateau level was achieved and in the absence of a CO2 signal, indicating satisfactory breath holding. Three readings each of exhaled and nasal air were taken (within 10%), with the results being expressed as the mean value.
Exhaled carbon monoxide (eCO) sensitivity of the analyser employed was 1 to 500 p.p.b. Measurement of CO was carried out using an external resistance of 0.40 ± 0.05 kPa (3 ± 0.4 mm Hg) and an exhalation flow of 5–6 L/min. The subjects exhaled slowly from total lung capacity for 10–15 s, maintaining a constant flow. The mean of two reproducible measurements with less than 5% variation was recorded. Ambient CO was recorded before each measurement and subtracted from the mean value obtained during the procedures.
Exhaled breath condensate
Exhaled breath condensate (EBC) was obtained between 11 a.m. and 13 p.m. with a specific condenser (EcoScreen; Jaeger) which allowed the non-invasive collection of the non-gaseous components of the expiratory air, and recommendations of the ATS/ERS task force were followed . Subjects were breathing through a mouthpiece and a two-way non-rebreathing valve, which also served as a saliva trap. They were wearing a noseclip and breathed in a relaxed manner (tidal breathing) for 15 min. This collection time provided an adequate and constant volume for assay and was well tolerated. If subjects felt saliva in their mouths, they were instructed to swallow it. Thus, at least 2 ml of breath condensate were collected as ice at −20°C and divided in two samples in 2-ml sterile plastic tubes. The first sample was used for the immediate measurement of pH and the second was immediately stored in −80°C for the measurements of hydrogen peroxide (H2O2), 8-isoprostane, leukotriene B4 (LTB4) and nitrates. In each group, measurements were performed on the same day. The samples reported herein were collected over a 4-month period and then assayed.
Right after the collection of condensate, pH was measured, as previously described . Stable pH was achieved in all cases after deaeration of the condensate with an inert gas (argon, 350 ml/min for 10 min) and was measured using a pH meter (CONSORT P-903).
Hydrogen peroxide measurements
EBC H2O2 was measured colorimetrically by means of horseradish peroxidase-catalysed oxidation of tetramethylbenzidine according to the method previously described . Briefly, 100 mL 3,39,5,59-tetramethylbenzidine [dissolved in 0.42 M citrate buffer (pH 3.8)] and 10 mL horseradish peroxidase (Sigma Chemicals, St Louis, MO, USA; 52.5 U/mL) were reacted with 100 mL EBC for 10 min at room temperature. Subsequently, the mixture was acidified to pH 1 with 10 mL 36 M sulphuric acid. The reaction product was measured spectrophotometrically at 450 nm. A separate standard curve for H2O2 was constructed for each assay. Detection limit was 0.31 μM. Intra-assay variation was calculated to be 3.0%.
A specific enzyme immunoassay kit (Cayman Chemical, Ann Arbor, MI, USA) was used to measure 8-isoprostane concentrations in breath condensates. The intra-assay and inter-assay variabilities were ±5 and ±6%, respectively. The detection limit of the assay was 4 pg/mL.
Leukotriene B4 measurements
A specific enzyme immunoassay kit (Cayman Chemical) was used to measure LTB4 concentrations in breath condensate. The intra-assay and inter-assay variabilities were <10%. The specificity was 100% and the detection limit of the assay was 3 pg/mL.
Samples were assayed by ion chromatography with a conductivity measurement analyser (Dionex DX 100; Dionex, Sunnyvale, CA, USA). Briefly, 0.1 ml of EBC was diluted to 2 ml with purified water (Dieffe Medical, Sondrio, Italy). The analyser utilises an exchange anion column (Dionex AS 4a-SC) which separates anions, which are then detected using a conductivity analyser (expressed in Siemens). Nitrate concentration is calculated as the area of the nitrate peak of the chromatogram. The method is able to detect nitrates levels as low as 2 μM, and its intra-assay variability was 8 ± 1.3%.
Quantitative data are reported as mean±SD. All analyses were conducted using statistical software (version 12.0; SPSS, Chicago, IL, USA). Non-parametric statistics were applied. Differences in group data were evaluated by Kruskal–Wallis analysis of variance by ranks; when significant differences were found, post hoc comparisons were performed by using the Mann–Whitney U test. To evaluate the relationship between apnoea–hypopnoea index and various variables, Pearson’s correlation test was applied. A p value of <0.05 was considered to indicate statistical significance.
Anthropometric, clinical and polysomnographic characteristics of the study population
OSAS patients (n = 26)
Obese non-apnoeic subjects (n = 9)
Non-obese non-apnoeic subjects (n = 10)
55.4 ± 14.1
52.0 ± 7.4
37.7 ± 5.3*
37.5 ± 10.0
33.3 ± 2.5
22.2 ± 1.7*
63.7 ± 29.5
2.7 ± 1.7*
1.5 ± 1.2*
FEV1 % predicted
94.9 ± 13.1
98.4 ± 20.2
98.6 ± 21.2
FVC % predicted
94.4 ± 13.5
94.7 ± 15.5
95.6 ± 16.4
82.6 ± 5.7
84.6 ± 7.0
88.7 ± 6.2
DLCO % reference
92.1 ± 14.5
97.5 ± 11.1
97.8 ± 13.6
PO2, mm Hg
75.8 ± 11.1
76.9 ± 8.7
80.5 ± 9.3
PCO2, mm Hg
39.3 ± 4.0
40.2 ± 2.8
35.7 ± 3.2
7.43 ± 0.02
7.42 ± 0.01
7.43 ± 0.01
42.4 ± 5.4
41.8 ± 4.8
40.3 ± 4.6
Nocturnal average oxygen saturation, %
84.2 ± 5.1
94.9 ± 2.3*
95.3 ± 1.9*
Nocturnal lowest oxygen saturation, %
67.7 ± 10.2
89.3 ± 2.1*
90.6 ± 2.2*
Systolic blood pressure, mm Hg
126.6 ± 18.9
117.2 ± 12.0
120.2 ± 13.5
Diastolic blood pressure, mm Hg
82.0 ± 9.7
76.6 ± 7.0
74.5 ± 6.8
Heart rate beats
77.8 ± 7.6
74.4 ± 5.5
76.6 ± 5.4
Exhaled air analysis
Nasal NO was higher in OSAS patients (610.3 ± 222.5 p.p.b.) than in obese controls (366.1 ± 168.7 p.p.b.; p < 0.01; Fig. 1b). There was no difference between the level of nNO of non-obese (539.1 ± 264.4 p.p.b.) and obese subjects (366.1 ± 168.7 p.p.b.) or OSAS patients (610.3 ± 222.5 p.p.b.). After CPAP therapy, concentration of nNO decreased (516.9 ± 135.7 p.p.b.; p < 0.05) but was still significantly higher than in obese subjects (p < 0.05).
Exhaled CO concentration was higher in OSAS patients (6.4 ± 2.9 p.p.b.) compared to obese (4.8 ± 1.0 p.p.b.; p < 0.05; Fig. 1c) and non-obese subjects (4.7 ± 1.2 p.p.b.; p < 0.05; Fig. 1c). There was no difference between the level of eCO in obese (4.8 ± 1.0 p.p.b.) and non-obese subjects (4.7 ± 1.2 p.p.b.). After CPAP, therapy concentration of eCO did not change.
Breath condensate analysis
Hydrogen peroxide was found increased in patients with OSAS (5.8 ± 8.9 uM) compared with non-obese (0.34 ± 0.4 uM; p < 0.01) and obese controls (1.2 ± 0.9 uM; p < 0.05). The level of H2O2 in obese controls was higher than in non-obese controls (p < 0.05; Fig. 2b).
Concentration of 8-isoprostane was higher in OSAS patients (12.8 ± 6.8 pg/ml) than in non-obese (5.5 ± 1.9 pg/ml; p < 0.001) and obese controls (4.0 ± 0.2 pg/ml; p < 0.001; Fig. 2c). Higher levels of LTB4 were found in patients with OSAS (8.7 ± 6.2 pg/ml) in comparison to obese and non-obese controls (p < 0.01; Fig. 2d). Additionally, the levels of nitrates were significantly higher in OSAS patients (23.3 ± 9.3 uM) than in non-obese (14.4 ± 5.9 uM; p < 0.01) and obese controls (10.5 ± 5.1 uM; p < 0.001; Fig. 2e). No differences in the level of LTB4 and nitrates were detected between obese and non-obese control subjects (Fig. 2d and e).
Influence of 1 month of CPAP therapy on EBC markers
Before CPAP (n = 20)
After CPAP (n = 20)
11.6 ± 5.7
11.8 ± 8.7
10.0 ± 6.7
9.9 ± 6.8
24.5 ± 9.6
20.8 ± 11.6
Hydrogen peroxide, μM
3.7 ± 6.1
4.5 ± 6.6
7.15 ± 0.80
7.73 ± 0.17*
This study demonstrates that patients with OSAS have increased levels of eNO, nNO and eCO compared to obese subjects. Obese non-apnoeic and non-obese non-apnoeic subjects did not differ in these parameters. In addition, in EBC of OSAS patients hydrogen peroxide, 8-isoprostane, LTB4 and nitrates are increased, whereas pH is decreased.
Exhaled NO, a well-known marker of inflammation and oxidative stress, is produced by various cells within the respiratory tract and is elevated in conditions such as asthma and bronchoectasis [20, 21]. Furthermore, nasal NO is constitutively produced by steroid-resistant nitric oxide synthase mainly in the mucosa of the paranasal sinuses and seems to be elevated in diseases implicating inflammation of the upper airways such as allergic rhinitis . In addition, exhaled CO is produced ubiquitously in the body by heme oxygenase as a breakdown product of heme and is also considered as a marker of inflammation and oxidative stress .
The increased levels of eNO, nNO and eCO suggest that inflammation and oxidative stress are present in the airways of patients with OSAS. Our data are consonant with the results of Olopade et al. that showed that eNO level was significantly increased after sleep in patients with moderate-severe OSAS . Interestingly, nNO of apnoeics did not differ from nNO of non-obese non-apnoeic subjects. Although the age difference between these two groups might hinder some interpretation, it is plausible to suggest that obesity and apnoeas exert an anti-thetical influence on nNO level. Thus, obesity may reduce the level of nNO, whereas apnoeas may increase it, resulting in normal level of nNO in obese apnoeics.
According to our findings, CPAP therapy significantly reduced eNO and nNO without altering the levels of eCO. These results suggest that 1 month of CPAP therapy may neutralise airway inflammation and oxidative stress, but this effect is not reflected by eCO levels. A longer period of CPAP use might be needed for the normalisation of eCO level as well.
Isoprostanes are reliable biomarkers of lipid peroxidation and oxidative stress that are formed in vivo by free radical-catalysed peroxidation of arachidonic acid . 8-Isoprostane is widely studied as a marker of oxidative stress in various respiratory diseases including bronchial asthma, chronic obstructive pulmonary disease, cystic fibrosis and inter-stitial lung diseases . An increased level of 8-isoprostane has also been reported in patients with OSAS [7, 26]. In full accordance with the latter, our data documented that 8-isoprostane is higher in patients with OSAS than in control non-obese and obese subjects. Thus, oxidative stress, reflected by 8-isoprostane, is present in the airways of patients with OSAS independent of the level of obesity .
Leukotrienes are lipid mediators derived from arachidonic acid and released from most inflammatory cells. LTB4 is directly synthesised by monocytes, alveolar macrophages and neutrophiles and is mainly an inflammatory mediator . According to our data, the level of LTB4 in patients with OSAS is significantly higher when compared with control non-obese and obese subjects. Hence, airway inflammation, detected by LTB4, is present in OSAS patients.
Nitrates in EBC are significantly higher in patients with OSAS in comparison with non-obese and obese controls. Nitrates are increased in patients with severe asthma, COPD and cystic fibrosis . These data also confirm the presence of airway inflammation in OSAS patients.
Hydrogen peroxide is higher in patients with OSAS than in healthy and obese controls. Hydrogen peroxide of EBC is a non-invasive marker of oxidative stress, and it is increased in healthy smokers and patients with asthma, COPD, bronchoectasis and acute respiratory distress syndrome . Therefore, we can conclude that elevation of H2O2 in OSAS patients indicates oxidative stress in OSAS patients.
The pH of EBC was lower in patients with OSAS than in control healthy and obese subjects. Kostikas et al. documented that pH values of EBC reflected the amounts of endogenous acidification on the respiratory tract . The same authors also confirmed that the endogenous airway acidification is strongly related to the inflammatory process. Consequently, airway inflammation, as estimated by pH of EBC, is present in the airway of OSAS patients.
We observed a positive correlation between AHI and the level of 8-isoprostane, LTB4, nitrates and H2O2 (higher level of all parameters in patients with increased AHI). This indicates that the number of episodes of airway obstruction is related to inflammation and oxidative stress. Interestingly, this correlation did not persist when the association of oxygen saturation (mean and lowest) with the aforementioned parameters was examined. Thus, it is plausible to speculate that, in regard of the causal pathway of inflammation and oxidative stress, the number of desaturation events is more important than their severity.
We have not detected any difference between the level of eNO, eCO, nNO, LTB4, nitrates and pH in obese and non-obese non-apnoeic controls. Conversely, 8-isoprostane was reduced in obese compared to non-obese controls. These data suggest that obesity without obstructive apnoea events does not cause inflammation and oxidative stress in the airways. Our results are in agreement with the findings of Carpagnano et al.  that documented that inflammation and oxidative stress are present in the airway of OSAS patients but not in obese subjects. Similarly, other studies confirmed that eNO and LTB4 are identical in obese and non-obese children with asthma, suggesting that obesity does not influence airway inflammation .
Interestingly, 1 month of CPAP therapy significantly reduced the level of eNO and nNO and increased pH of EBC without altering the level of 8-isoprostane, LTB4, nitrates and H2O2. Conversely, Carpagnano et al. have shown that the concentration of 8-isoprostane is reduced after two nights of CPAP therapy . In the latter study, in contrast to our protocol, EBC was collected immediately after sleep. We suppose that longer periods of CPAP therapy might be needed to affect oxidative stress markers.
Some weaknesses of the current study must be acknowledged and deserve consideration. Firstly, the relatively small number of subjects studied may have precluded the detection of other correlations between examined variables. However, previous trials [7, 12, 13, 26] included similar (or lower) numbers of subjects and yet were able to provide equivalent results. Secondly, it is noteworthy that the design of the study was not randomised, and additionally, non-obese non-apnoeic subjects were younger than participants of the two other study groups. Accordingly, a degree of bias should be acknowledged in the interpretation of our results. Lastly, it is plausible to suggest that an earlier timeframe for the collection of the examined specimens would have provided a more accurate assessment of the overnight physiologic effects of sleep apnoea.
In conclusion, the present study illustrates that airway inflammation and oxidative stress, as reflected by exhaled breath markers, are present in the airway of OSAS patients in contrast to obese non-apnoeic subjects. Additionally, exhaled breath markers are positively correlated with the severity of OSAS. Lastly, 1 month of CPAP therapy reduces the degree of airway inflammation and oxidative stress.