European Respiratory Society


There is increasing evidence that inflammation plays an important role in the development of cardiovascular complications in patients with obstructive sleep apnoea (OSA).

No previous works have studied levels of soluble tumour necrosis factor-α receptor (sTNFR)-1 in patients with OSA. The aims of the present study were to examine serum levels of sTNFR-1 and the effect of nasal continuous positive airway pressure (CPAP) in patients with OSA.

A prospective, randomised, placebo-controlled crossover study was performed. In total, 30 consecutive newly diagnosed OSA patients (apnoea/hypopnoea index 43.8±27.0 events·h−1) and 15 healthy obese patients were selected. Urinary levels of norepinephrine and epinephrine, as well as plasma sTNFR-1, tumour necrosis factor (TNF)-α, interleukin (IL)-6 and leukotriene (LT)B4 levels were obtained at baseline and after 3 months of CPAP or sham CPAP.

Nocturnal urinary levels of norepinephrine, epinephrine and sTNFR-1 (1,053±269 versus 820±166 pg·mL−1) were significantly higher in OSA patients. There were no significant differences in plasma levels of IL-6, LTB4, or TNF-α between the two study groups. There were no significant differences in blood pressure, urinary catecholamine levels, or plasma IL-6, LTB4 and TNF-α levels after both treatment modalities. However, after 3 months of effective CPAP usage, sTNFR-1 levels were significantly reduced (1,053±269 versus 899±254 pg·mL−1).

Obstructive sleep apnoea patients have higher levels of soluble tumour necrosis factor-α receptor 1 than individuals without OSA; soluble tumour necrosis factor-α receptor 1 levels are lowered by continuous positive airway pressure therapy. These findings further corroborate a potential role of inflammation in the natural history of obstructive sleep apnoea.

Obstructive sleep apnoea (OSA) is the most common form of sleep-disordered breathing, affecting 4% of middle-aged (30–60 yrs) males and 2% of middle-aged (30–60 yrs) females. OSA is characterised by repeated partial or complete closure of the upper airway during sleep leading to chronic intermittent hypoxia, sleep fragmentation and daytime sleepiness. This medical condition is associated with a great variety of cardiovascular disturbances and stroke, resulting in severe morbidity and mortality 1. OSA has been associated with an increased risk of systemic hypertension, left and right ventricular dysfunction, ventricular hypertrophy, heart failure, metabolic dysregulation, pulmonary hypertension, coronary artery disease and cardiac arrhythmias. There are many risk factors for OSA, including sex (male), advanced age and obesity, which are the same for cardiovascular disease, making it difficult to accurately determine the role of OSA as an independent risk factor.

Systemic inflammation is crucial in the pathogenesis of all stages of atherosclerosis. There is also increasing evidence that inflammation plays an important role in the development of cardiovascular complications in patients with OSA 2.

Conversely, several studies have demonstrated that tumour necrosis factor (TNF)-α and its soluble receptors, especially soluble TNF-α receptor (sTNFR)-1, are frequently elevated in patients with heart failure, providing useful prognostic information 35. In addition, higher inflammatory marker levels may predict survival and development of overt heart failure in patients with myocardial infarction 6, 7 and may antedate new-onset heart failure even in patients with asymptomatic left ventricular systolic dysfunction without prior myocardial infarction 8. TNF-α soluble receptors have also been linked to disorders in glucose metabolism 9, and sTNFR-1 is probably an important regulator of sleep–wake cycles 10.

There are no data about the possible presence of elevated plasma sTNFR-1 levels in individuals with OSA. Moreover, previous studies evaluating the effects of continuous positive airway pressure (CPAP) therapy, the treatment of choice for OSA, on inflammatory marker levels were not examined with a randomised, placebo-controlled design 1115. The aims of the current study were: 1) to evaluate whether levels of sTNFR-1, as well as other pro-inflammatory cytokines, including TNF-α, interleukin (IL)-6, and leukotriene (LT)B4, are elevated in OSA and otherwise healthy obese patients compared with obese subjects without OSA; and 2) to determine whether treatment with CPAP decreases cytokine levels in a randomised, placebo-controlled crossover study design.


Study population

In total, 33 consecutive newly diagnosed OSA patients (apnoea/hypopnoea index (AHI) ≥10 events·h−1 and excessive daytime sleepiness (Epworth sleepiness scale >11)) and 15 obese control subjects (AHI <5 events·h−1 and Epworth sleepiness scale <10) were selected to be studied from January 2004 to June 2005. None of the patients had ever received treatment for OSA. Control subjects were recruited from a list of healthy subjects from the present authors’ Madrid metropolitan area of healthcare assigned to the Hospital La Paz (Madrid, Spain), who had undergone a routine health examination in the previous 3 months. A control subject similar in sex, age (±2 yrs), weight (±2 kg) and height (±5 cm) with regard to the two preceding patients was randomly selected to be included in the study. Exclusion criteria for both study groups were: 1) unwillingness or inability to perform the testing procedure; 2) obstructive or restrictive lung disease demonstrated on pulmonary function testing; 3) current use of cardioactive drugs; 4) cardiac rhythm disturbances, including sinus bradycardia and sinus tachycardia; 5) known hypertension, or 24-h mean blood pressure of 135 systolic and/or 85 mmHg diastolic or more; 6) left ventricular ejection fraction <50%; ischaemic or valvular heart disease; hypertrophic, restrictive or infiltrative cardiomyopathy; pericardial disease or stroke by history or physical examination, ECG, chest radiography, conventional stress testing and echocardiography; 7) diabetes mellitus, by history or two random blood glucose levels ≥126 mg·dL−1; 8) morbid obesity (body mass index (BMI) >40 kg·m−2); and 9) daytime hypoxaemia (arterial oxygen tension <70 mmHg (9.3 kPa)) or hypercapnia (arterial carbon dioxide tension >45 mmHg (6.0 kPa)). Withdrawal criteria were: 1) clinical exacerbation leading to a change in medication; 2) hospital admission for ≥10 days; and 3) average nightly CPAP usage <3.5 h. Three patients were deemed ineligible for inclusion after initial assessments (one had left ventricular systolic dysfunction secondary to ischaemic heart disease and two had unknown mitral stenosis); therefore, 30 OSA patients were finally included in the study.

The study was approved by the Institutional Ethics Committee at the Hospital Universitario La Paz (Madrid, Spain). All subjects gave their written informed consent prior to enrolment.

Study protocol

A single-centre, prospective, randomised, double-blind, placebo-controlled, crossover study was performed, in which OSA patients received effective CPAP (Aria LX CPAP device; Respironics, Murrysville, PA, USA) and sham CPAP therapy for two 12-week periods (fig. 1). The sham CPAP device consisted of a conventional CPAP device, in which the area of the exhalation port was amplified, thereby nearly cancelling nasal pressure (1–2 cmH2O (0.1–0.2 kPa)); an orifice resistor was connected between the tubing and the CPAP unit that loads the blower with the same airflow resistance as in effective CPAP.

Fig. 1—

Protocol of the study. CPAP: continuous positive airway pressure.

OSA patients underwent a full-night CPAP titration study with an automated pressure setting device (Auto Set; ResMed, Sydney, Australia). Patients were given detailed instructions on using CPAP equipment, but they were not informed of the type of therapy they were receiving. Compliance data were measured with a run-time counter.

All subjects underwent baseline sleep study, 24-h ambulatory blood pressure monitoring (ABPM) and a transthoracic echocardiogram. Plasma levels of sTNFR-1, TNF-α, LTB4 and IL-6 were measured. Urine specimens were also collected. OSA patients were randomised to receive either effective or sham CPAP therapy for 12 weeks. They were then readmitted to the hospital and the CPAP device was switched to the alternative mode of therapy for another 12 weeks. Echocardiogram, ABPM, plasma cytokines and catecholamines were measured just after each period with either effective or sham CPAP treatment.


LTB4 concentrations were measured by a forward sequential competitive enzyme immunoassay (R&D Systems, Minneapolis, MN, USA). The inter- and intra-assay variations between the samples were 10.5 and 9.3%, respectively, and the detection limit of the assay was 27.6 pg·mL−1.

Quantitative sandwich enzyme immunoassay kits (R&D Systems) were used to measure IL-6, TNF-α and sTNFR-1 concentrations in plasma. The inter- and intra-assay variations between the samples were as follows. IL-6: 3.3 and 1.6%; TNF-α: 4.6 and 4.2%; and sTNFR-1: 5.0 and 5.2%, respectively, i.e. the inter- and intra-assay variabilities were <10%. The detection limits of the assays were 0.7, 0.5 and 0.43 pg·mL−1, respectively.

During the same night that urine specimens were collected, respiratory polygraphy was performed in OSA patients and control subjects. A previously validated portable respiratory recording device was used (Sibel Home-300; Sibel S.A., Barcelona, Spain) 16, which records oronasal airflow (using a thermistor and nasal cannula prongs), chest wall impedance, oxygen saturation, snoring and body position. Oximetry data were averaged over 2-s sampling intervals. Artefacts were removed by eliminating all changes of oxygen saturation between consecutive sampling intervals of >4%·s−1, and any oxygen saturation <20%.

Respiratory events were classified as either obstructive or central on the basis of presence or absence of respiratory effort. Respiratory events were scored as apnoea when there was a cessation of oronasal airflow lasting ≥10 s. Hypopnoea was defined as a decrease of 50% in oronasal airflow lasting >10 s, associated with a fall in arterial oxygen saturation (Sa,O2) of >4% of the preceding baseline level. Mean night-time Sa,O2, minimum Sa,O2(lowest values recorded during sleep), oxygen desaturation index and percentage of time with Sa,O2<90% on nocturnal oximetry were computed as indices of nocturnal oxygen saturation.

Spirometry, echocardiogram and 24-h ABPM were performed on each patient as previously described 17, 18. Urinary excretion of norepinephrine and epinephrine were also determined as previously described 17.

Statistical analyses

Values are expressed as mean±sd or percentage. All statistical tests were two-sided. The comparisons between the groups were performed using the Wilcoxon rank-sum test, while the Chi-squared test was used for evaluating frequencies. Relationships between variables were determined by the Pearson's correlation analysis. Comparisons of effects of the treatments over time were made with repeated-measures ANOVA, using treatment as a within-subject factor and order as a between-subject factor. When ANOVA results showed significant differences between treatment conditions, post hoc multiple comparisons were performed with the Bonferroni test. A value of p<0.05 was considered statistically significant.


All individuals in both groups were male. Characteristics of the studied population are shown in table 1. There were no differences in either demographic, blood pressure or lung function data between the two study groups. Nocturnal levels of norepinephrine and epinephrine were higher in patients with OSA. Nocturnal epinephrine dip was higher in control subjects. Although the values were within the normal limits, OSA patients had higher values of left ventricular posterior wall, interventricular septum thickness, left ventricular mass, and left ventricular mass index.

View this table:
Table 1—

Baseline characteristics in obstructive sleep apnoea(OSA) patients and control subjects

Baseline levels of sTNFR-1 were higher in OSA patients compared with control obese subjects (fig. 2). There was no significant difference in TNF-α, IL-6 or LTB4 levels between the two study groups (table 2).

Fig. 2—

Serum soluble tumour necrosis factor-α receptor (sTNFR)-1 levels in obstructive sleep apnoea (OSA) patients and control subjects. Box and whisker plots represent median, interquartile range and range. #: p = 0.032 by Wilcoxon test.

View this table:
Table 2—

Plasma levels of pro-inflammatory cytokines in obstructive sleep apnoea (OSA) patients and control subjects

Several associations between baseline variables and cytokine levels were revealed. There were no significant relationships between plasma levels of TNF-α or LTB4 and any studied variable. Significant positive relationships were found between weight and sTNFR-1 (r = 0.446, p = 0.033) and IL-6 (r = 0.736, p = 0.003) levels. Also, BMI was associated with high sTNRF-1 (r = 0.414, p = 0.049) and IL-6 (r = 0.693, p = 0.006) levels. Low epinephrine nocturnal dip was associated with higher sTNFR-1 (r =  -0.484, p = 0.019) levels. In addition, levels of IL-6 were positively correlated with diurnal norepinephrine levels (r = 0.667, p = 0.009) and negatively correlated with forced vital capacity % predicted (r =  -0.636, p = 0.015) and forced expiratory volume in one second % pred (r =  -0.583, p = 0.029).

Pearson's correlation coefficients between levels of sTNFR-1, TNF-α, IL-6 and LTB4 and sleep parameters in patients with OSA are shown in table 3. There was no significant relationship between any of the sleep parameters and circulating cytokine levels.

View this table:
Table 3—

Correlation between plasma cytokine levels and sleep parameters in obstructive sleep apnoea patients

Five OSA patients failed to complete the study, two because of an average nightly usage of CPAP of <3.5 h, and three were lost from follow-up. Complete measurements were available in 25 patients. Mean CPAP pressure was 10±2 cmH2O (0.98±0.20 kPa), and the average nightly usage was similar on sham CPAP and effective CPAP (6.3±1.6 versus 6.2±1.1 h; p = 0.6636). There were no differences in heart rate, weight, blood pressure recordings or urinary catecholamines at baseline and after sham or effective CPAP in OSA patients (table 4). There was no randomisation of patients in the cross-over design. After 12 weeks on effective CPAP, a significant reduction in sTNFR-1 levels was observed (table 5 and fig. 3). After both modes of treatment, no significant differences were observed in either systolic function or ventricular structure parameters.

Fig. 3—

Individual serum soluble tumour necrosis factor-α receptor (sTNFR)-1 levels at baseline, after sham continuous positive airway pressure (CPAP) and CPAP therapy in 25 obstructive sleep aponea patients. *: p<0.05. •: mean values.

View this table:
Table 4—

Weight, blood pressure data, and urinary catecholamines in 25 obstructive sleep apnoea patients at baseline and after both sham continuous positive airway pressure(CPAP) and effective CPAP

View this table:
Table 5—

Plasma levels of pro-inflammatory cytokines in 25 obstructive sleep apnoea patients at baseline and after both sham continuous positive airway pressure (CPAP) and effective CPAP


In the present study, levels of sTNFR-1 were significantly higher in obese OSA and otherwise healthy patients compared with obese control subjects. In addition, circulating TNF-α, IL-6 and LTB4 levels were not different between the two study groups. Interestingly, after 3 months of CPAP treatment, circulating sTNFR-1 levels were lower in OSA patients.

It has been shown that elevated sTNFR-1 levels predict the incidence of fatal and nonfatal cardiovascular events in several conditions, such as heart failure and acute coronary syndrome 37. In that sense, in patients with acute myocardial infarction sTNFR-1 levels have been shown to be a major predictor of mortality and heart failure 6, 7. Increasing circulating levels of sTNFR-1 are also associated with increased mortality in a wide range of heart failure patients 35. Reported levels of sTNFR-1 in both heart failure and ischaemic patients were higher than those observed in the present OSA patient population, probably reflecting the fact that the present OSA patients were free of any overt cardiovascular disease. Whether higher levels of sTNFR-1 are present in a nonselected population of OSA subjects needs to be determined in future studies. Heart failure and coronary artery disease share multiple cardiovascular pathophysiological mechanisms with OSA, including endothelial dysfunction, systemic inflammation, impaired baroreflex and sympathetic overactivity.

Elevated sTNFR-1 levels are also reported to be associated with type-2 diabetes mellitus, regardless of obesity 9. It is conceivable that the association of OSA and high sTNFR-1 levels may raise the rate of cardiovascular events in this patient population. Increased plasma levels of pro-inflammatory cytokines may be linked not only to poor cardiovascular outcomes in OSA but also to excessive daytime sleepiness, which is a typical feature of OSA 19. Moreover, elevated circulating cytokine levels may be the result of the enhanced cellular release of such cytokines in response to intermittent apnoea-related hypoxia, supporting a direct role for inflammation as a potential mechanism of atherogenesis in OSA 2, 20, 21. In this regard, OSA and chronic intermittent hypoxaemia are associated with the activation of nuclear factor κB, which is involved in the transcription of numerous genes involved in the inflammatory cascade; its downstream products (including the TNF-α system) may be key factors in the link between OSA and cardiovascular diseases. However, further studies are needed in order to evaluate whether increased sTNFR-1 levels can play a role in predicting subsequent prognosis during long-term follow-up in the entire spectrum of individuals with OSA due to initial inflammatory responses, or whether, on the contrary, they are an epiphenomenon secondary to sleep deprivation.

Previous studies have reported that patients with OSA develop systemic inflammation with increased levels of circulating pro-inflammatory cytokines, such as TNF-α, IL-6 and others, as compared with healthy subjects, and CPAP therapy reduces the levels of OSA-induced inflammatory cytokines 1113, 15, 22. Importantly, the increased levels of cytokines correlated with OSA severity, regardless of obesity 23. Conversely, some authors have reported no increased levels of inflammatory mediators in individuals with OSA when compared with subjects without OSA 14, 24, 25. In the present study, only sTNFR-1 levels were higher in OSA patients than in control subjects, but no differences were found for the other pro-inflammatory markers studied. These opposing results may be due to differences in the duration of the disease in the study samples, variable presence of comorbid conditions or the measurement techniques used. Cellular actions of TNF-α are mediated by TNF receptors. Circulating levels of TNF-α are frequently undetectable or highly variable in humans and sTNFR-1 levels may be a better marker for the biological effects and activation status of the TNF-α system.

In the present study, two groups of obese individuals were studied. A positive correlation was found between sTNFR-1, IL-6 levels and BMI in OSA patients, but there were no significant relationships between sleep parameters and plasma cytokine levels. However, BMI did not differ between the two study groups. Obesity is very common among OSA patients and it may also be associated with inflammation. In view of the current results, the present authors speculate that at the initial stages of the disease, when clinical cardiovascular disturbances have not yet developed, the enhanced sympathetic activity secondary to repetitive nocturnal obstructive respiratory events may be the predominant mechanism responsible for the rise in sTNFR-1 levels. An ongoing prospective randomised clinical trial is studying whether the increased risk of cardiovascular complications in individuals with OSA is caused primarily by the inflammatory effects of OSA or obesity 26.

The present study has several limitations. All individuals were males. Therefore, the findings may not be readily extrapolated to female patients. In addition, the present study included a limited number of patients. That is explained by the fact that the process of including only newly diagnosed middle-aged OSA and otherwise healthy patients to achieve the purpose of the present study was very difficult. As is well known, in the global population of patients with OSA the presence of coexistent comorbid conditions is very common. The present data do not allow an analysis of whether the arousal index is related to TNF-α and sTNFR-1, since the patients were diagnosed by means of polygraphy. Since the principal target of the present study was to value the effect of the CPAP on serum levels of sTNFR-1 in OSA patients, neither the design nor the sample size are suitable for the evaluation of the possible determinants of the levels of sTNFR-1 in these patients. However, the present authors speculate that the impact of sleep fragmentation on sTNFR-1 levels in OSA patients might be considered to be discrete. Total sleep deprivation increases the plasma levels of sTNFR-1 and IL-6, but there have not been any identified changes associated with a partial sleep deprivation, which corresponds better to the situation of the OSA patients. Finally, the present study did not use polysomnography, but a portable respiratory recording device, to diagnose OSA. Although conventional polysomnography is considered the gold standard for the diagnosis of OSA, it has been previously demonstrated that an accurate diagnosis of OSA can be made without using neurological variables, even in an unattended setting, suggesting that the portable respiratory recording device is therefore suitable for use in epidemiological studies 27.

In conclusion, high circulating soluble tumour necrosis factor-α receptor-1 levels may be observed in obese obstructive sleep apnoea patients with no associated comorbid conditions. Further studies will determine whether it could serve as an early marker of ongoing systemic inflammation in obstructive sleep apnoea, increasing the risk for eventual future cardiovascular adverse events. Chronic continuous positive airway pressure application might be useful in reducing the potential cardiovascular risks of elevated circulating plasma soluble tumour necrosis factor-α receptor-1 and other pro-inflammatory cytokines in patients with obstructive sleep apnoea.

Support statement

This research was partly supported by a grant from the Fondo de Investigación Sanitaria (FIS; Spain; grant code exp. 01/0278) and Neumomadrid (Madrid, Spain; 2000).

Statement of interest

None declared.

  • Received January 15, 2008.
  • Accepted May 2, 2008.


View Abstract