Sleep apnea and arterial stiffness
J Sleep Res. (2014) 23, 700–708
Obstructive sleep apnea is associated with increased arterial stiffness in severe obesity IAN W. SEETHO1, ROBERT J. PARKER2, SONYA CRAIG2, NICK DUFFY2, K E V I N J . H A R D Y 3 and J O H N P . H . W I L D I N G 1 1
Department of Obesity and Endocrinology, University of Liverpool, Clinical Sciences Centre, University Hospital Aintree, Liverpool, UK, Department of Respiratory Medicine, University Hospital Aintree, Liverpool, UK and 3Department of Diabetes and Endocrinology, St Helens and Knowsley Teaching Hospitals NHS Trust, St Helens, UK 2
Keywords arterial stiffness, obstructive sleep apnea, severe obesity Correspondence Ian Seetho, Department of Obesity and Endocrinology, University of Liverpool, Clinical Sciences Centre, University Hospital Aintree, Longmoor Lane, Liverpool L9 7AL, UK. Tel.: 0151 5295885; fax: 0151 5295888; e-mail: [email protected] Accepted in revised form 8 March 2014; received 21 November 2013 DOI: 10.1111/jsr.12156
SUMMARY
Obstructive sleep apnea is associated with obesity and metabolic syndrome, leading to greater cardiovascular risk. Severely obese patients with obstructive sleep apnea may still be at risk of adverse health outcomes, even without previous cardiovascular disease. Pulse wave analysis non-invasively measures peripheral pulse waveforms and derives measures of haemodynamic status, including arterial stiffness, augmentation pressure and subendocardial viability ratio. We hypothesized that the presence of obstructive sleep apnea in severe obesity, even in the absence of an antecedent history of cardiovascular disease, would affect measurements derived from pulse wave analysis. Seventy-two severely obese adult subjects [obstructive sleep apnea 47 (body mass index 42 7 kg m 2), without obstructive sleep apnea (non-OSA) 25 (body mass index 40 5 kg m 2)] were characterised using anthropometric, respiratory and cardio-metabolic parameters. Groups were similar in age, body mass index and gender. More subjects with obstructive sleep apnea had metabolic syndrome [obstructive sleep apnea 60%, without obstructive sleep apnea (non-OSA) 12%]. Those with obstructive sleep apnea had greater arterial stiffness, augmentation pressure and decreased subendocardial viability ratio (all P < 0.001), with significantly higher systolic (P = 0.003), diastolic (P = 0.04) and mean arterial pressures (P = 0.004) than patients without obstructive sleep apnea (nonOSA). Arterial stiffness correlated with mean arterial blood pressure (P = 0.003) and obstructive sleep apnea severity (apnea–hypopnea index; P < 0.001). apnea–hypopnea index significantly predicted arterial stiffness in multiple regression analysis, but components of the metabolic syndrome did not. Thus, patients with obstructive sleep apnea with severe obesity have increased arterial stiffness that may potentially influence cardiovascular risk independently of metabolic abnormalities. The presence of obstructive sleep apnea in severe obesity identifies a group at high cardiovascular risk; clinicians should ensure that risk factors are managed appropriately in this group whether or not treatment of obstructive sleep apnea is offered or accepted by patients.
INTRODUCTION Obstructive sleep apnea (OSA) is a common condition, with prevalence in middle-aged men and women at 4% and 2%, respectively (Young et al., 1993). Obesity is an important risk factor for OSA, with prevalence estimates in severe obesity
700
between 40% and 90% (Schwartz et al., 2008). OSA is characterised by repetitive intermittent hypoxia, periodic arousals and sleep fragmentation, with cardiovascular and metabolic sequelae. Randomised trials have shown that OSA treatment with continuous positive airway pressure (CPAP) can improve hypertension and atherosclerosis (Drager et al., ª 2014 European Sleep Research Society
Arterial stiffness in OSA and severe obesity 2007a; Pepperell et al., 2002). Furthermore, fewer long-term adverse cardiovascular events have been observed when severe OSA is treated with CPAP (Marin et al., 2005). The proposed mechanisms underlying the cardiovascular associations of OSA include increased sympathetic drive, hypothalamic–pituitary–adrenal axis activation, insulin resistance and associated metabolic disturbances, increased systemic inflammation, oxidative stress and a prothrombotic state (Kohler and Stradling, 2010). Obstructive sleep apnea is associated with increased arterial stiffness that may contribute to the associated cardiovascular risk (Doonan et al., 2011). It has been reported that increased arterial stiffness (measured as augmentation index) derived from the radial artery is predictive of cardiovascular events (Weber et al., 2005). Arterial stiffness is associated with OSA severity (Doonan et al., 2011), and is elevated in OSA (Kohler et al., 2008). This is attenuated following CPAP therapy (Drager et al., 2007a). Such changes in arterial stiffness indicate that it may be linked with an increased cardiovascular risk in patients with OSA (Buchner et al., 2012). Pulse wave analysis (PWA) is a non-invasive measurement technique that allows the study of peripheral pulse waveforms with measurements using a micromanometertipped probe. Intra-arterial pulse pressure is transmitted through the arterial wall to the sensor, and peripheral pressure waveforms are calibrated using peripheral blood pressure (BP) readings, with subsequent derivation of aortic pressure waveforms based on a transfer function (O’Rourke et al., 2001). Previous work has shown that PWA demonstrates high levels of reproducibility (Wilkinson et al., 1998). The use of PWA can provide information concerning haemodynamic status, enabling derivation of arterial stiffness, augmentation pressure (AP) and subendocardial viability ratio (SEVR) measurements that may allow assessment of cardiovascular risk. The augmentation index (Aix) is a composite measure of central arterial stiffness and peripheral wave reflection. The reflected waves cause changes in pulse wave morphology that affects haemodynamic performance and vascular compliance. Aix is the difference between the pulse height of the primary systolic and reflected peak pressure waves divided by the central pulse pressure expressed as a percentage (Siebenhofer et al., 1999). Augmentation pressure is defined as the difference between the primary and reflected systolic peak pressures. It is a measure of central wave reflection in the vasculature and is a marker of cardiovascular risk in OSA (Noda et al., 2008). Subendocardial viability ratio is expressed as a ratio of diastolic pressure time interval to systolic pressure time interval. It is a marker of myocardial oxygen supply and demand, and subendocardial ischaemia (Siebenhofer et al., 1999). A higher SEVR is better in terms of cardiovascular health, as lower values (at ~50%) indicate decreased diastolic perfusion times and reduced coronary perfusion (Ferro et al., 1995; Hoffman and Buckberg, 1978). ª 2014 European Sleep Research Society
701
Although PWA has been used to investigate the effects of OSA on arterial stiffness (Phillips et al., 2013), to date, few studies have investigated these effects in severely obese cohorts [defined as a body mass index (BMI) ≥ 35 kg m 2]. Jelic et al. (2002) studied changes in arterial stiffness during sleep cycles with OSA in an obese cohort with a mean BMI of 35 kg m 2 (Jelic et al., 2002). In another study, the differential effects of OSA severity on arterial properties in subjects with a mean BMI of 38 kg m 2 were investigated, but a control group was not used (Protogerou et al., 2008). YimYeh et al. (2010) investigated the effects of age and OSA on vascular function in a cohort with a mean BMI of 37 kg m 2 (Yim-Yeh et al., 2010). Bakker et al. (2011) investigated the effects of continuous and auto-adjusted positive airway pressure on 12 severely obese patients, but did not study patients of similar BMI without OSA (Bakker et al., 2011). Severely obese patients with OSA may still be at risk of adverse health outcomes, even without a previous history of cardiovascular disease. Therefore, this knowledge will be useful when patients are assessed in clinical practice, especially in the presence of OSA, enabling risk factors to be modified and the potential to improve health outcomes with early treatment. In this study, we hypothesized that the presence of OSA in severe obesity, even in the absence of an antecedent history of cardiovascular disease, would affect measurements derived from PWA. In order to test this, we measured the Aix, AP and SEVR derived from PWA in patients recruited from weight management clinics. In order to determine the independence of the observed associations relating to the Aix, a regression analysis adjusted for obesity and other variables such as age, BP, lipid levels, fasting glucose, body measurements (neck and waist-to-hip ratio), body fat percentage and the apnea –hypopnea index (AHI) was performed. MATERIALS AND METHODS Ethics statement The study was approved by the local research ethics committee (NRES 12/NW/123). The study was performed in accordance with the Declaration of Helsinki. All patients gave written informed consent. Participants We studied severely obese patients from multidisciplinary weight management clinics at University Hospital Aintree. Patients were recruited from March 2011 to January 2012. Inclusion and exclusion criteria were assessed according to the clinical history, physical examination and analysis of the medical notes. Patients were eligible if they were ≥21 years old and had a BMI ≥ 35 kg m ². Exclusion criteria were patients who were being treated or had prior treatment for OSA and those with known cardio-respiratory disease; hypertension (defined
702
I. W. Seetho et al.
as BP >140/90 or on BP-lowering medications; Chobanian et al., 2003); current smokers or those with more than 10pack years smoking history; kidney and liver disease; acute illnesses and pregnancy. Protocol All patients attended a study visit day between 08:00 and 10:00 h, and underwent a detailed history and physical examination. Body composition measurements, anthropometry, fasting blood and urine biochemical tests, PWA, venous blood gases (to assess for hypercarbia), and spirometry testing were performed. An overnight sleep study was performed 2–3 weeks later, and patients then grouped according to their sleep status. This ensured that the recruiter was blinded to the OSA status of the patient, especially in relation to the PWA measurements taken on the study visit day. Given the high prevalence of OSA in obesity, it was anticipated that there would be more OSA than non-OSA patients. We estimated it would be necessary to screen approximately 90 patients to ensure adequate recruitment of non-OSA subjects. Blood pressure Blood pressure was measured at the arm in a sitting position after a rest for at least 5 min at 1-min intervals between each measurement using an oscillometric digital BP monitor (HEM-705CP, Omron, Kyoto, Japan). The mean of three measurements was calculated. Body measurements All measurements were done in triplicate. Weight and height were measured without shoes and with light clothing. Body composition measurements used bioimpedance scales (TBF521, TANITA, Tokyo, Japan). This method has been previously validated (Jebb et al., 2000). Other measurements included neck circumference at the level of the laryngeal prominence; waist circumference midway between the lower rib and iliac crest; and hip circumference was measured horizontally over the widest part of the gluteal region. The tape measure was ensured to be snug and not compressing the skin, parallel to the floor with measurement at the end of a normal expiration. PWA Pulse wave analysis was performed at the same time each day (~09:30 h) for all patients in order to ensure uniformity in measurements. All measurements were made in the fasting state, with abstinence from alcohol and caffeine for at least 8 h, in the upright position, in a temperature-controlled room (24 °C). PWA measurements were taken by a single operator (IWS) by applying a hand-held tonometer attached to the SphygmoCor system (AtCor Medical, Sydney, Australia). The tonometer comprised a high-fidelity micromanometer-tipped probe that was gently applied to the skin surface at the radial
artery of the non-dominant arm in a non-occlusive manner, with the wrist slightly flexed and palm facing upwards. By this means, radial artery waveforms were acquired and digitised through the SphygmoCor software linked to a computer. Radial artery waveforms were calibrated from brachial pressures that were measured using an oscillometric digital BP monitor (HEM-705CP, Omron, Kyoto, Japan). A minimum of 10 radial waveforms for each patient was required to generate a corresponding ascending aortic pressure waveform by a validated mathematical transfer function within the SphygmoCor software (AtCor Medical, Sydney, Australia). To ensure quality control, only measurements with a Quality Index ≥80% and a signal strength ≥500 units were accepted. As heart rate (HR) influences the Aix, all values were normalised to a HR of 75 bpm. The tonometry data acquired were utilised by the SphygmoCor software to derive values for Aix, AP and SEVR. Spirometry assessment Spirometry was performed with a Spiro Air LT system (Medisoft, Sorinnes, Belgium), supervised by an experienced technician. Sleep diagnostic assessment Daytime somnolence was assessed using an Epworth Sleepiness Scale (ESS) questionnaire, where a score >10 indicated increased sleepiness (Johns, 1993). Diagnosis was confirmed by overnight multichannel respiratory-limited polysomnography (Somnoscreen Digital PSG & EEG acquisition system, Version 2.0, SomnoMedics, Randersacker, Germany). Sleep studies were independently assessed by experienced sleep physiologists using software [Domino PSG analysis software (version 2.5.0), SomnoMedics, Randersacker, Germany]. apnea was defined as a cessation of airflow for >10 s. Hypopnoea was defined as a 50% reduction in airflow accompanied by a >4% desaturation and a reduction in chest wall movement. Sleep apnea was diagnosed if the AHI was ≥5 (Flemons et al., 1999). Biochemical measurements Serum, urea, creatinine and electrolytes, fasting lipids, thyroid function, and fasting glucose were measured using standard laboratory assays (Roche, UK). Blood gases were analysed with a Cobas Blood Gas Analyser (Roche, UK). Metabolic syndrome Subjects were assessed for metabolic syndrome according to the National Cholesterol Education Program (NCEP) guidelines (Cleeman et al., 2001). Patients had metabolic syndrome if three or more risk factors were present: waist circumference (males >102 cm, females >88 cm); triglycerides ≥1.7 mM; high-density lipoprotein cholesterol ª 2014 European Sleep Research Society
Arterial stiffness in OSA and severe obesity (males
J Sleep Res. (2014) 23, 700–708
Obstructive sleep apnea is associated with increased arterial stiffness in severe obesity IAN W. SEETHO1, ROBERT J. PARKER2, SONYA CRAIG2, NICK DUFFY2, K E V I N J . H A R D Y 3 and J O H N P . H . W I L D I N G 1 1
Department of Obesity and Endocrinology, University of Liverpool, Clinical Sciences Centre, University Hospital Aintree, Liverpool, UK, Department of Respiratory Medicine, University Hospital Aintree, Liverpool, UK and 3Department of Diabetes and Endocrinology, St Helens and Knowsley Teaching Hospitals NHS Trust, St Helens, UK 2
Keywords arterial stiffness, obstructive sleep apnea, severe obesity Correspondence Ian Seetho, Department of Obesity and Endocrinology, University of Liverpool, Clinical Sciences Centre, University Hospital Aintree, Longmoor Lane, Liverpool L9 7AL, UK. Tel.: 0151 5295885; fax: 0151 5295888; e-mail: [email protected] Accepted in revised form 8 March 2014; received 21 November 2013 DOI: 10.1111/jsr.12156
SUMMARY
Obstructive sleep apnea is associated with obesity and metabolic syndrome, leading to greater cardiovascular risk. Severely obese patients with obstructive sleep apnea may still be at risk of adverse health outcomes, even without previous cardiovascular disease. Pulse wave analysis non-invasively measures peripheral pulse waveforms and derives measures of haemodynamic status, including arterial stiffness, augmentation pressure and subendocardial viability ratio. We hypothesized that the presence of obstructive sleep apnea in severe obesity, even in the absence of an antecedent history of cardiovascular disease, would affect measurements derived from pulse wave analysis. Seventy-two severely obese adult subjects [obstructive sleep apnea 47 (body mass index 42 7 kg m 2), without obstructive sleep apnea (non-OSA) 25 (body mass index 40 5 kg m 2)] were characterised using anthropometric, respiratory and cardio-metabolic parameters. Groups were similar in age, body mass index and gender. More subjects with obstructive sleep apnea had metabolic syndrome [obstructive sleep apnea 60%, without obstructive sleep apnea (non-OSA) 12%]. Those with obstructive sleep apnea had greater arterial stiffness, augmentation pressure and decreased subendocardial viability ratio (all P < 0.001), with significantly higher systolic (P = 0.003), diastolic (P = 0.04) and mean arterial pressures (P = 0.004) than patients without obstructive sleep apnea (nonOSA). Arterial stiffness correlated with mean arterial blood pressure (P = 0.003) and obstructive sleep apnea severity (apnea–hypopnea index; P < 0.001). apnea–hypopnea index significantly predicted arterial stiffness in multiple regression analysis, but components of the metabolic syndrome did not. Thus, patients with obstructive sleep apnea with severe obesity have increased arterial stiffness that may potentially influence cardiovascular risk independently of metabolic abnormalities. The presence of obstructive sleep apnea in severe obesity identifies a group at high cardiovascular risk; clinicians should ensure that risk factors are managed appropriately in this group whether or not treatment of obstructive sleep apnea is offered or accepted by patients.
INTRODUCTION Obstructive sleep apnea (OSA) is a common condition, with prevalence in middle-aged men and women at 4% and 2%, respectively (Young et al., 1993). Obesity is an important risk factor for OSA, with prevalence estimates in severe obesity
700
between 40% and 90% (Schwartz et al., 2008). OSA is characterised by repetitive intermittent hypoxia, periodic arousals and sleep fragmentation, with cardiovascular and metabolic sequelae. Randomised trials have shown that OSA treatment with continuous positive airway pressure (CPAP) can improve hypertension and atherosclerosis (Drager et al., ª 2014 European Sleep Research Society
Arterial stiffness in OSA and severe obesity 2007a; Pepperell et al., 2002). Furthermore, fewer long-term adverse cardiovascular events have been observed when severe OSA is treated with CPAP (Marin et al., 2005). The proposed mechanisms underlying the cardiovascular associations of OSA include increased sympathetic drive, hypothalamic–pituitary–adrenal axis activation, insulin resistance and associated metabolic disturbances, increased systemic inflammation, oxidative stress and a prothrombotic state (Kohler and Stradling, 2010). Obstructive sleep apnea is associated with increased arterial stiffness that may contribute to the associated cardiovascular risk (Doonan et al., 2011). It has been reported that increased arterial stiffness (measured as augmentation index) derived from the radial artery is predictive of cardiovascular events (Weber et al., 2005). Arterial stiffness is associated with OSA severity (Doonan et al., 2011), and is elevated in OSA (Kohler et al., 2008). This is attenuated following CPAP therapy (Drager et al., 2007a). Such changes in arterial stiffness indicate that it may be linked with an increased cardiovascular risk in patients with OSA (Buchner et al., 2012). Pulse wave analysis (PWA) is a non-invasive measurement technique that allows the study of peripheral pulse waveforms with measurements using a micromanometertipped probe. Intra-arterial pulse pressure is transmitted through the arterial wall to the sensor, and peripheral pressure waveforms are calibrated using peripheral blood pressure (BP) readings, with subsequent derivation of aortic pressure waveforms based on a transfer function (O’Rourke et al., 2001). Previous work has shown that PWA demonstrates high levels of reproducibility (Wilkinson et al., 1998). The use of PWA can provide information concerning haemodynamic status, enabling derivation of arterial stiffness, augmentation pressure (AP) and subendocardial viability ratio (SEVR) measurements that may allow assessment of cardiovascular risk. The augmentation index (Aix) is a composite measure of central arterial stiffness and peripheral wave reflection. The reflected waves cause changes in pulse wave morphology that affects haemodynamic performance and vascular compliance. Aix is the difference between the pulse height of the primary systolic and reflected peak pressure waves divided by the central pulse pressure expressed as a percentage (Siebenhofer et al., 1999). Augmentation pressure is defined as the difference between the primary and reflected systolic peak pressures. It is a measure of central wave reflection in the vasculature and is a marker of cardiovascular risk in OSA (Noda et al., 2008). Subendocardial viability ratio is expressed as a ratio of diastolic pressure time interval to systolic pressure time interval. It is a marker of myocardial oxygen supply and demand, and subendocardial ischaemia (Siebenhofer et al., 1999). A higher SEVR is better in terms of cardiovascular health, as lower values (at ~50%) indicate decreased diastolic perfusion times and reduced coronary perfusion (Ferro et al., 1995; Hoffman and Buckberg, 1978). ª 2014 European Sleep Research Society
701
Although PWA has been used to investigate the effects of OSA on arterial stiffness (Phillips et al., 2013), to date, few studies have investigated these effects in severely obese cohorts [defined as a body mass index (BMI) ≥ 35 kg m 2]. Jelic et al. (2002) studied changes in arterial stiffness during sleep cycles with OSA in an obese cohort with a mean BMI of 35 kg m 2 (Jelic et al., 2002). In another study, the differential effects of OSA severity on arterial properties in subjects with a mean BMI of 38 kg m 2 were investigated, but a control group was not used (Protogerou et al., 2008). YimYeh et al. (2010) investigated the effects of age and OSA on vascular function in a cohort with a mean BMI of 37 kg m 2 (Yim-Yeh et al., 2010). Bakker et al. (2011) investigated the effects of continuous and auto-adjusted positive airway pressure on 12 severely obese patients, but did not study patients of similar BMI without OSA (Bakker et al., 2011). Severely obese patients with OSA may still be at risk of adverse health outcomes, even without a previous history of cardiovascular disease. Therefore, this knowledge will be useful when patients are assessed in clinical practice, especially in the presence of OSA, enabling risk factors to be modified and the potential to improve health outcomes with early treatment. In this study, we hypothesized that the presence of OSA in severe obesity, even in the absence of an antecedent history of cardiovascular disease, would affect measurements derived from PWA. In order to test this, we measured the Aix, AP and SEVR derived from PWA in patients recruited from weight management clinics. In order to determine the independence of the observed associations relating to the Aix, a regression analysis adjusted for obesity and other variables such as age, BP, lipid levels, fasting glucose, body measurements (neck and waist-to-hip ratio), body fat percentage and the apnea –hypopnea index (AHI) was performed. MATERIALS AND METHODS Ethics statement The study was approved by the local research ethics committee (NRES 12/NW/123). The study was performed in accordance with the Declaration of Helsinki. All patients gave written informed consent. Participants We studied severely obese patients from multidisciplinary weight management clinics at University Hospital Aintree. Patients were recruited from March 2011 to January 2012. Inclusion and exclusion criteria were assessed according to the clinical history, physical examination and analysis of the medical notes. Patients were eligible if they were ≥21 years old and had a BMI ≥ 35 kg m ². Exclusion criteria were patients who were being treated or had prior treatment for OSA and those with known cardio-respiratory disease; hypertension (defined
702
I. W. Seetho et al.
as BP >140/90 or on BP-lowering medications; Chobanian et al., 2003); current smokers or those with more than 10pack years smoking history; kidney and liver disease; acute illnesses and pregnancy. Protocol All patients attended a study visit day between 08:00 and 10:00 h, and underwent a detailed history and physical examination. Body composition measurements, anthropometry, fasting blood and urine biochemical tests, PWA, venous blood gases (to assess for hypercarbia), and spirometry testing were performed. An overnight sleep study was performed 2–3 weeks later, and patients then grouped according to their sleep status. This ensured that the recruiter was blinded to the OSA status of the patient, especially in relation to the PWA measurements taken on the study visit day. Given the high prevalence of OSA in obesity, it was anticipated that there would be more OSA than non-OSA patients. We estimated it would be necessary to screen approximately 90 patients to ensure adequate recruitment of non-OSA subjects. Blood pressure Blood pressure was measured at the arm in a sitting position after a rest for at least 5 min at 1-min intervals between each measurement using an oscillometric digital BP monitor (HEM-705CP, Omron, Kyoto, Japan). The mean of three measurements was calculated. Body measurements All measurements were done in triplicate. Weight and height were measured without shoes and with light clothing. Body composition measurements used bioimpedance scales (TBF521, TANITA, Tokyo, Japan). This method has been previously validated (Jebb et al., 2000). Other measurements included neck circumference at the level of the laryngeal prominence; waist circumference midway between the lower rib and iliac crest; and hip circumference was measured horizontally over the widest part of the gluteal region. The tape measure was ensured to be snug and not compressing the skin, parallel to the floor with measurement at the end of a normal expiration. PWA Pulse wave analysis was performed at the same time each day (~09:30 h) for all patients in order to ensure uniformity in measurements. All measurements were made in the fasting state, with abstinence from alcohol and caffeine for at least 8 h, in the upright position, in a temperature-controlled room (24 °C). PWA measurements were taken by a single operator (IWS) by applying a hand-held tonometer attached to the SphygmoCor system (AtCor Medical, Sydney, Australia). The tonometer comprised a high-fidelity micromanometer-tipped probe that was gently applied to the skin surface at the radial
artery of the non-dominant arm in a non-occlusive manner, with the wrist slightly flexed and palm facing upwards. By this means, radial artery waveforms were acquired and digitised through the SphygmoCor software linked to a computer. Radial artery waveforms were calibrated from brachial pressures that were measured using an oscillometric digital BP monitor (HEM-705CP, Omron, Kyoto, Japan). A minimum of 10 radial waveforms for each patient was required to generate a corresponding ascending aortic pressure waveform by a validated mathematical transfer function within the SphygmoCor software (AtCor Medical, Sydney, Australia). To ensure quality control, only measurements with a Quality Index ≥80% and a signal strength ≥500 units were accepted. As heart rate (HR) influences the Aix, all values were normalised to a HR of 75 bpm. The tonometry data acquired were utilised by the SphygmoCor software to derive values for Aix, AP and SEVR. Spirometry assessment Spirometry was performed with a Spiro Air LT system (Medisoft, Sorinnes, Belgium), supervised by an experienced technician. Sleep diagnostic assessment Daytime somnolence was assessed using an Epworth Sleepiness Scale (ESS) questionnaire, where a score >10 indicated increased sleepiness (Johns, 1993). Diagnosis was confirmed by overnight multichannel respiratory-limited polysomnography (Somnoscreen Digital PSG & EEG acquisition system, Version 2.0, SomnoMedics, Randersacker, Germany). Sleep studies were independently assessed by experienced sleep physiologists using software [Domino PSG analysis software (version 2.5.0), SomnoMedics, Randersacker, Germany]. apnea was defined as a cessation of airflow for >10 s. Hypopnoea was defined as a 50% reduction in airflow accompanied by a >4% desaturation and a reduction in chest wall movement. Sleep apnea was diagnosed if the AHI was ≥5 (Flemons et al., 1999). Biochemical measurements Serum, urea, creatinine and electrolytes, fasting lipids, thyroid function, and fasting glucose were measured using standard laboratory assays (Roche, UK). Blood gases were analysed with a Cobas Blood Gas Analyser (Roche, UK). Metabolic syndrome Subjects were assessed for metabolic syndrome according to the National Cholesterol Education Program (NCEP) guidelines (Cleeman et al., 2001). Patients had metabolic syndrome if three or more risk factors were present: waist circumference (males >102 cm, females >88 cm); triglycerides ≥1.7 mM; high-density lipoprotein cholesterol ª 2014 European Sleep Research Society
Arterial stiffness in OSA and severe obesity (males
