- Email: [email protected]
Frequency of Sleep Apnea in Adults With the Marfan Syndrome
Frequency of Sleep Apnea in Adults With the Marfan Syndrome Meike Rybczynski, MDa, Dietmar Koschyk, MDa, Andreas Karmeier, MDd, Nele Gessler, MDa, Sara Sheikhzadeh, MDa, Alexander M. J. Bernhardt, MDa, Christian R. Habermann, MDb, Hendrik Treede, MDa, Jürgen Berger, PhDc, Peter N. Robinson, MD, MSce, Thomas Meinertz, MDa, and Yskert von Kodolitsch, MD, MBAa,* Obstructive and central sleep apneas are treatable disorders, which contribute to cardiovascular morbidity in older adults. Younger adults with Marfan syndrome may also be at risk for sleep apnea, but the relation between cardiovascular complications and sleep apnea is unknown. We used MiniScreen8 portable monitoring devices for polygraphy in 68 consecutive adults with Marfan syndrome (33 men, 35 women, 41 ⴞ 14 years old) to investigate frequency of sleep apnea and its relation to cardiovascular morbidity. The apnea– hypopnea index (AHI) was 6 to 15/hour in 14 subjects (mild sleep apnea, 21%), and AHI was >15/hour in 7 subjects (moderate or severe sleep apnea, 10%). Among established risk factors for sleep apnea, only older age (Spearman rho ⴝ 0.35, p ⴝ 0.004) and body mass index (rho ⴝ 0.26, p ⴝ 0.03) were associated with increased AHI. Of all cases of apnea, 12 ⴞ 27 were obstructive, 11 ⴞ 25 central, and 3 ⴞ 9 mixed. AHI was associated with decreased left ventricular ejection fraction (rho ⴝ ⴚ0.33, p ⴝ 0.01), increased N-terminal pro– brain natriuretic peptide levels (rho ⴝ 0.35, p ⴝ 0.004), enlarged descending aortic diameters (rho ⴝ 0.44, p ⴝ 0.001), atrial fibrillation (phi ⴝ 0.43, p ⴝ 0.002), and mitral valve surgery (phi ⴝ 0.34, p ⴝ 0.02). Of these, left ventricular ejection fraction, N-terminal pro– brain natriuretic peptide levels, atrial fibrillation, and mitral valve surgery were associated with AHI independently of age and body mass index. We found similar associations with oxygen desaturation index. In conclusion, sleep apnea exhibits increased frequency in Marfan syndrome and is not predicted by classic risk factors. Obstructive and central sleep apneas may relate to cardiovascular disease variables. © 2010 Elsevier Inc. All rights reserved. (Am J Cardiol 2010;105:1836 –1841) Epidemiologic studies have documented an association between obstructive sleep apnea (OSA) and cardiovascular diseases such as arterial hypertension, coronary artery disease, arrhythmias, heart failure, stroke, and aortic dissection, whereas central sleep apnea (CSA) has predominantly been related to heart failure.1 Continuous positive airway pressure is an effective treatment, which has been shown to decrease the risk of cardiovascular disease.1 Recently, 2 centers reported an increased frequency of OSA in Marfan syndrome.2,3 Marfan syndrome is an autosomal, dominantly inherited disease of the connective tissue with manifestations in the eyes, lungs, dura, and skeleton, which results from mutations in the gene coding for fibrillin-1, FBN1.4 Untreated patients with Marfan syndrome die prematurely due to dissection and rupture of the aorta. Despite effective treatment of aortic root disease, cardiovascular morbidity a Centre of Cardiology and Cardiovascular Surgery and bDepartments of Diagnostic and Interventional Radiology and cMedical Biometry and Epidemiology, University Hospital Eppendorf, Hamburg, Germany, and dCentre of Pulmonary Disease, Sleep Medicine, Hamburg, Germany, and eInstitute of Medical Genetics, Charité Universitätsmedizin Berlin, Berlin, Germany. Manuscript received November 7, 2009; revised manuscript received and accepted January 15, 2010. This study was supported by a grant to the Centre of Cardiology and Cardiovascular Surgery, University Hospital Eppendorf, Hamburg, Germany, from Heinen and Löwenstein, Bad Ems, Germany. *Corresponding author: Tel: 4940-7410-57328; fax: 4940-7410-54840. E-mail address: [email protected] (Y. von Kodolitsch).
0002-9149/10/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.01.369
remains high due to arrhythmia, heart failure, complications at distal sites of the aortic vessel, and heart valve dysfunction.4 We performed the present study to investigate, first, whether sleep apnea is present in a relevant subset of unselected adults with Marfan syndrome, and second, whether sleep apnea relates to cardiovascular morbidity. Methods We used portable 8-channel monitoring devices for ambulatory, unattended respiratory polygraphy. All patients were ⬎17 years of age, all presented to our clinic within a period of 1 year, all lived within the Hamburg metropolitan area, and all were evaluated according to the Ghent nosology.5,6 We did not enroll patients in whom the Marfan syndrome had been excluded,5 pregnant women, patients with blindness, and patients on continuous positive airway pressure therapy. We continued all medical regimens. We evaluated 106 consecutive outpatients. Of these, 30 patients did not fulfill criteria of Marfan syndrome, 2 were on continuous positive airway pressure treatment, 1 was blind, 1 did not complete our diagnostic program, 2 declined to participate in our study, and 2 had inadequate MiniScreen8 measurements. The remaining 68 patients with Marfan syndrome constituted our study group, with 33 men and 35 women with a mean age of 41 ⫾ 14 years (range 18 to 70; Table 1). The Hamburg research ethics committee approved our protocol, which was in accordance with principles of the Declaration of Helsinki. All subjects gave written informed consent. We performed baseline clinical examinations including basewww.AJConline.org
Miscellaneous/Apnea–Hypopnea Index in Marfan Syndrome
1837
Table 1 Relation between baseline clinical variables and apnea– hypopnea index or oxygen desaturation index in 68 patients with Marfan syndrome Variable
Male gender Age (years) Body weight (kg) Body height (m) Body mass index (kg/m2) Body surface area (m2) Neck circumference (cm) Active smoker Blood glucose (mg/dl) Total cholesterol (mg/dl) High-density lipoprotein cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) Triglyceride (mg/dl) Epworth Sleepiness Score ⬎10 Berlin questionnaire ⱖ2
All Patients (n ⫽ 68)
33 (49%) 41 ⫾ 14 79 ⫾ 20 1.85 ⫾ 0.2 22 ⫾ 4 2 ⫾ 0.2 37 ⫾ 4 5 (8%) 93 ⫾ 17 185 ⫾ 42 58 ⫾ 18 104 ⫾ 35 123 ⫾ 78 13 (19%) 17 (25%)
Range
AHI
18–70 47–127 1.52–2.08 15–32 1.49–2.65 29–50 64–155 103–290 27–104 45–191 40–474
ODI
Correlation Coefficient*
p Value†
Correlation Coefficient*
p Value†
0.29 0.35 0.24 0.01 0.26 0.17 0.23 0.16 0.18 0.23 0.05 0.11 0.16 0.27 0.16
0.06 0.004 0.05 0.9 0.03 0.2 0.06 0.45 0.2 0.08 0.7 0.4 0.2 0.09 0.42
0.26 0.46 0.36 0.09 0.38 0.31 0.36 0.11 0.24 0.35 ⫺0.03 0.26 0.25 0.15 0.27
0.11 ⬍0.001 0.003 0.5 0.002 0.01 0.003 0.71 0.06 0.007 0.8 0.05 0.06 . 46 0.08
Data are presented as mean ⫾ SD or frequency (percentage). * Spearman correlation coefficient (rho) for continuous variables. AHI and ODI were grouped 0 to 5, 6 to 15, and ⬎15 to assess the phi coefficient for binary variables. † The p value of Spearman correlation or phi coefficient.
line serum examinations, blood pressure measurement, standard 12-lead electrocardiography, transthoracic echocardiography,7 and magnetic resonance angiography of the entire aorta within 24 hours of polygraphy in all patients. An experienced technician provided individual training on how to use the MiniScreen8 device and a standardized written and illustrated device instruction to all patients. We used MiniScreen8 type 3 devices (Heinen and Löwenstein, Bad Ems, Germany) in all patients.8 During sleep the devices record nasal flow with a pressure transducer system, oxygen saturation and pulse rate by finger oximetry, body position through a magnetic sensor, snoring sounds, and thoracic and abdominal movements through belts with pneumatic cushions for pressure measurement. All patients used the device for 1 full night at home. We performed automated analysis of downloaded physiological data within 2 days of polygraphy. We also reviewed all electronically stored data manually using EasyScreen Viewer 5.09 (Heinen und Löwenstein) jointly with a board-certified physician of sleep medicine. All investigators were blinded to medical history and other medical data. In 7 subjects with technically insufficient recordings we initiated a second polygraphy yielding adequate data in 5 subjects, and we excluded 2 subjects because they declined to repeat polygraphy. We assessed body height using a wall-fixed height rule, body mass index, body surface area, and neck circumference measured at the level of the cricothyroid membrane.9 We considered active smoking with any inhaled intake of nicotine within ⱕ7 days before polygraphy and we measured fasting blood glucose levels and fasting lipid levels ⱕ24 hours of the study. We used the Epworth Sleepiness Scale questionnaire10 and the Berlin questionnaire11 to assess subjective sleepiness (Table 1). We presented heart rates and blood oxygen saturation as the mean during total recording time with documentation of lowest saturation and longest duration of desaturation. We defined an episode of apnea as the cessation of airflow
lasting ⬎10 seconds and hypopnea as a decrease in airflow of ⱖ50% lasting ⬎10 seconds, associated with a decrease in oxygen saturation of ⱖ4%.12 We quantified severity of sleep apnea as the number of apneas– hypopneas per hour (apnea– hypopnea index [AHI]) and as the number of oxygen desaturations (oxygen desaturation index [ODI]) ⱖ4% per hour of study.12 We used established thresholds for AHI and ODI to classify a sleep-related breathing disorder as absent with ⱕ5 events per hour, as mild with 6 to 15 events per hour, as moderate with ⬎15 events per hour, and as severe with ⬎30 events per hour.12 We considered obstructive apnea as ongoing ventilatory efforts documented by thoracic and abdominal movement signals for ⬎2 respiratory cycles. We defined apnea as central when apnea episodes presented without any ongoing ventilatory efforts (Figure 1), and mixed apnea when airflow continued to be absent despite resumption of respiratory efforts in an apnea episode with initial cessation of respiratory efforts.12 We assessed snoring severity as the number of snoring events per hour (Table 2). We obtained systolic and diastolic blood pressures on sphygmomanometer after 15 minutes of rest in a supine position and we documented heart rates and presence of sinus rhythm or atrial fibrillation on standard 12-lead electrocardiograms.5 We performed 2-dimensional standard echocardiography for left ventricular ejection fraction (LVEF)7 and maximum aortic diameters at the level of the aortic sinuses according to Roman et al.13 We used magnetic resonance angiography for diameters of the ascending and descending aortas at established levels. We did not consider diameters at aortic sites with an aortic prosthesis. We noted aortic dissection with a history of dissection or intramural hemorrhage at any site of the aorta reported in surgical records or on tomographic images. We considered proximal aortic surgery with a documented history of a David operation (19 patients with type 1 operation), a con-
1838
The American Journal of Cardiology (www.AJConline.org)
Figure 1. Polygraphic records of oxygen saturation (SaO2), pulse rate (Puls), oronasal airflow (Flow), thoracic ventilatory effort (Thorax), abdominal ventilatory effort (Abdomen), registration of snoring sounds (Schnarchen), and body position (Lage). (Top) A 43-year-old man with classic Marfan syndrome who had undergone an elective David operation 2 years previously. LVEF was 74%, and the N-terminal pro– brain natriuretic peptide level was 81.7 pg/ml. The patient was in a right lateral body position and he exhibited a typical pattern of obstructive episodes of apnea and snoring sounds. (Bottom) A 41-year-old man with classic Marfan syndrome who had undergone surgical reconstruction of his mitral valve at the age of 36 years and a David operation 1 year later. He was in New York Heart Association class III, his LVEF was 23%, and his N-terminal pro– brain natriuretic peptide level was 1,038 pg/ml. The patient was in a supine body position, he did not snore, and he had a typical breathing pattern of nonobstructive apnea with Cheyne–Stokes respiration.
duit operation with an artificial (15 patients) or a biological (3 patients) valve prosthesis, or with a Yacoub operation (2 patients). We documented distal aortic intervention with
presence of a tube graft in the descending (4 patients) or abdominal (3 patients) aorta. We considered mitral valve prolapse with prolapse on 2-dimensional echocardiogram
Miscellaneous/Apnea–Hypopnea Index in Marfan Syndrome
1839
Table 2 Relation between polygraphic variables and apnea– hypopnea index or oxygen desaturation index in 68 patients with Marfan syndrome Variable
Total recording time (minutes) Mean heart rate (beats/min) Oxyhemoglobin saturation Mean (%) Lowest (%) Longest desaturation (minutes) Episodes of apnea (number)‡ Total Obstructive Central Mixed Longest apnea (seconds) Episodes of hypopnea (number) Longest hypopnea (seconds) Snore index (per hour)
All Patients (n ⫽ 68)
Range
AHI
601 ⫾ 122 65 ⫾ 11
409–803 45–103
0 0.14
92 ⫾ 12 80 ⫾ 19 1.1 ⫾ 0.9
88–96 66–94 0.12–3.43
27 ⫾ 49 12 ⫾ 27 11 ⫾ 25 3⫾9 24 ⫾ 17 18 ⫾ 24 36 ⫾ 33 67 ⫾ 126
0–289 0–163 0–145 0–66 11–87 0–109 11–181 0–111
ODI
Correlation Coefficient*
p Value†
Correlation Coefficient*
p Value†
0.9 0.3
⫺0.07 0.35
0.6 0.004
⫺0.44 ⫺0.35 0.47
⬍0.001 0.004 ⬍0.001
⫺0.51 ⫺0.51 0.64
⬍0.001 ⬍0.001 ⬍0.001
0.63 0.49 0.34 0.47 0.59 0.76 0.62 0.33
⬍0.001 ⬍0.001 0.006 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.008
0.24 ⫺0.04 0.07 0.25 0.27 0.88 0.58 0.44
0.05 0.6 0.6 0.048 0.03 ⬍0.001 ⬍0.001 ⬍0.001
Data are presented as mean ⫾ SD. * Spearman correlation coefficient (rho). † The p value for Spearman correlation. ‡ Number of episodes measured during total recording time.
Table 3 Relation between cardiovascular disease variables and apnea– hypopnea index or oxygen desaturation index in 68 patients with Marfan syndrome Variables
Mean systolic blood pressure (mm Hg) Mean diastolic blood pressure (mm Hg) Baseline heart rate (beats/min) Atrial fibrillation Left ventricular ejection fraction (%) N-terminal pro–brain natriuretic peptide (pg/ml) Aortic root diameter (mm) Ascending aorta diameter (mm) Descending aorta diameter (mm) Aortic dissection Proximal aortic surgery Distal aortic intervention Mitral valve prolapse Mitral valve regurgitation Mitral valve surgery Any aortic medication
All Patients (n ⫽ 68)
Range
125 ⫾ 13 70 ⫾ 10 65 ⫾ 11 5 (8%) 55 ⫾ 11 263 ⫾ 316 38 ⫾ 5 29 ⫾ 6 25 ⫾ 7 14 (21%) 36 (53%) 7 (10%) 36 (53%) 12 (18%) 8 (12%) 38 (56%)
100–162 50–90 45–89 23–75 25–1,444 27–51 20–44 16–48
AHI
ODI
Correlation Coefficient*
p Value†
Correlation Coefficient*
p Value†
0.09 0.04 0.07 0.43 ⫺0.33 0.35 0.14 0.27 0.44 0.24 0.13 0.12 0.06 0.28 0.34 0.28
0.5 0.8 0.6 0.002 0.01 0.004 0.5 0.1 0.001 0.15 0.58 0.59 0.90 0.07 0.02 0.08
0.13 0.13 0.10 0.36 ⫺0.17 0.40 0.26 0.40 0.47 0.14 0.13 0.15 0.05 0.40 0.40 0.15
0.3 0.3 0.5 0.014 0.2 ⬍0.001 0.2 0.03 ⬍0.001 0.54 0.57 0.49 0.92 0.005 0.01 0.46
Data are presented as mean ⫾ SD or frequency (percentage). * Spearman correlation coefficient (rho) for continuous variables. AHI and ODI were grouped 0 to 5, 6 to 15, and ⬎15 to assess the phi coefficient for binary variables. † The p value of Spearman correlation or of the phi coefficient.
irrespective presence of valvular regurgitation14 and mitral regurgitation with at least moderate insufficiency diagnosed by 2-dimensional and Doppler echocardiography.15 Mitral valve surgery was present with reconstruction (5 patients) or replacement (3 patients) of the mitral valve. We considered patients on aortic medication with intake of  blockers (metoprolol or bisoprolol in 37 patients), angiotensin-converting enzyme inhibitors (ramipril in 14 patients), angiotensin II receptor blockers (valsartan or candesartan in 7
patients), or calcium antagonists (amlodipine or lercanidipine in 5 patients). In addition, we measured N-terminal pro– brain natriuretic peptide serum levels with an electrochemiluminescence sandwich immunoassay (Roche Diagnostics GmbH, Mannheim, Germany) on an Elecsys System 2010 with a detection limit ⱖ5 pg/ml (Table 3). We presented all data as mean ⫾ SD with ranges or frequencies. To evaluate the association of baseline patient characteristics (Table 1), polygraphic measurements (Table
1840
The American Journal of Cardiology (www.AJConline.org)
Table 4 Multivariate ordered logistic regression analysis (logit model) of cardiovascular disease variables in 68 patients with Marfan syndrome Variable
Atrial fibrillation Left ventricular ejection fraction (%) N-terminal pro–brain natriuretic peptide (ng/ml) Ascending aorta diameter (mm) Descending aorta diameter (mm) Mitral regurgitation Mitral valve surgery
ODI†
AHI* Regression Coefficient
95% CI
p Value
Regression Coefficient
95% CI
p Value
2.07 ⫺0.07 0.002
0.13–4.01 ⫺0.13 to ⫺0.002 0.0004–0.04
0.036 0.043 0.016
3.17
0.46–5.89
0.022
⫺0.006
⫺0.103 to 0.091
1.75
0.14–3.36
0.901 0.033
0.002 0.032 0.034 1.68 2.71
0.0001–0.004 ⫺0.191 to 0.256 ⫺0.059 to 0.127 ⫺0.07 to 3.43 0.69–4.73
0.041 0.775 0.479 0.06 0.008
* Each variable was adjusted for age and body mass index. † Each variable was adjusted for age, body mass index, neck circumference, and total cholesterol; due to a smaller number of cases, the variables ascending aorta diameter and descending aorta diameter were adjusted only for age and neck circumference but not for total cholesterol. CI ⫽ confidence interval.
2), and cardiovascular disease variables (Table 3) with AHI or ODI we used the Spearman correlation coefficient (rho) for continuous variables and the phi coefficient for nominal variables. We used these nonparametric tests because of the value distribution characteristics of AHI and ODI. To calculate the correlation coefficient rho, we used AHI and ODI as a continuous variable, and for phi coefficients with construction of contingence tables, we grouped AHI and ODI as 0 to 5, 6 to 15, and ⬎15/hour. To test, whether cardiovascular disease variables were associated with AHI or ODI independently of baseline clinical characteristics, we performed ordered logistic regression with grouped values of AHI or ODI as the dependent variable (Table 4). For ordered logistic regression, we considered only baseline characteristics that were associated with AHI or ODI at a level of significance set at a p value ⬍0.05 (2-sided). The Brant test or approximate likelihood ratio test corroborated the model assumption of proportional odds in all analyses. We used STATA 9.0 (STATA Corp., College Station, Texas) for ordered logistic regression including tests to check the model assumption and SPSS 15.0 for Windows (SPSS, Inc., Chicago, Illinois) for all other tests. Results We found an AHI ⱕ5/hour in 47 subjects (69%), an AHI 6 to 15/hour in 14 subjects (21%), and an AHI ⬎15/hour in 7 subjects (10%), including 2 subjects with an AHI ⬎30/ hour. Similarly, the ODI was ⱕ5/hour in 48 subjects (71%), 6 to 15/hour in 12 subjects (19%), and ⬎15/hour in 8 subjects (12%), including 3 subjects with an ODI ⬎30. Only older age and higher body mass index were related to AHI and ODI, whereas other epidemiologic and clinical characteristics were related only to AHI or ODI or were unrelated to AHI and ODI (Table 1). About 1/2 of all episodes of apnea were classified as obstructive (12 ⫾ 27), whereas remaining episodes of apnea were classified as central (11 ⫾ 25) or mixed (3 ⫾ 9; Figure 1, Table 2). Increased frequency of atrial fibrillation, increased N-terminal pro– brain natriuretic peptide levels, enlarged diameters of the descending aorta, and mitral valve surgery related to AHI and ODI. Conversely, a decreased LVEF related only to AHI, and enlarged diameters of the ascending aorta and increased frequency of mitral regurgitation
related only to ODI (Table 3). Ordered logistic regression showed that atrial fibrillation, LVEF, N-terminal pro– brain natriuretic peptide levels, and mitral valve surgery were associated with AHI independently of age and body mass index. Similarly, atrial fibrillation, N-terminal pro– brain natriuretic peptide levels, and mitral valve surgery were associated with ODI (Table 4). Discussion Our study documents that about 1/3 of unselected adults with Marfan syndrome exhibited sleep apnea compared to 12% reported in the general population of similar age as our study patients.3 Surprisingly, we found that about 1/2 of all episodes of apnea in patients with Marfan syndrome were central rather than obstructive. Moreover, sleep apnea related to occurrence of atrial fibrillation, decreased LVEF, increased N-terminal pro– brain natriuretic peptide serum levels, and increased frequency of mitral valve surgery. The composition of our study group was not biased for patient characteristics such as medication, presence of previous surgery, or symptoms of sleep-disordered breathing. The only other study of unselected adults with Marfan syndrome corroborated our estimate of 30% of sleep apnea in patients with Marfan syndrome.3 Conversely, the higher frequency of 64% of sleep apnea in a study of 25 patients with Marfan syndrome may reflect some selection bias.16 A questionnaire-based study of 15 patients with Marfan syndrome also supported a frequency of only 27% of sleep apnea.17 Our study demonstrated that only age and body mass index were associated with increased AHI and ODI, whereas predisposing conditions such as body weight, body surface area, enlarged neck circumference, total cholesterol levels, active smoking, or hyperglycemia were unrelated to AHI and/or ODI. Similarly, signs of subjective sleepiness did not relate to AHI or ODI in our patients.3 Thus, classic clues for sleep apnea may be of limited value for identifying increased risk of sleep apnea in Marfan syndrome. Thus, we suggest screening patients with Marfan syndrome even when sleep apnea is not clinically suspected. Cistulli et al2,18,19 suggested upper airway collapsibility, high nasal airway resistance, and craniofacial abnormalities as causal mechanisms of OSA in Marfan syndrome, whereas
Miscellaneous/Apnea–Hypopnea Index in Marfan Syndrome
Kohler et al3 did not find increased frequencies of retrognathia or abnormalities of mean neck circumference, cricomental distance, and Mallampati score in their patients. Kohler et al used ApneaLink type 4 devices (ResMed, MAP Medicine Technology, Martinsried, Germany), which permitted only flow limitations and snoring to be measured3 but which were therefore of limited use for distinguishing OSA from CSA. Conversely, our type 3 devices measured thoracic and abdominal ventilatory efforts and thus provided better means to identify obstructive and central apnea (Figure 1). We found only weak relations between aortic diameters and AHI and ODI, most likely because many patients previously underwent operation of the aorta and could not be considered for analysis. However, previous studies have found an association of AHI with aortic dilatation3 and aortic dissection,20 and 3 reports of patients with Marfan syndrome have documented cessation of aortic diameter progression on continuous positive airway pressure therapy.21,22 Patients with Marfan syndrome exhibit primary myocardial dysfunction.4 In patients without Marfan syndrome, the association of myocardial dysfunction with OSA and CSA is well established.23,24 In our patients sleep apnea was independently associated with decreased LVEF and increased N-terminal pro– brain natriuretic peptide levels. It is likely that increased N-terminal pro– brain natriuretic peptide levels were not caused by systolic dysfunction alone, but also by diastolic25 dysfunction, which usually evolves in Marfan syndrome before LVEF decreases.7 An increased risk for atrial fibrillation is established only in CSA with preserved26 and impaired27 LV function, but not in OSA.1 Interestingly, our study with a high frequency of CSA documents an association of AHI and ODI with atrial fibrillation. Studies of chronic heart failure have demonstrated decreased mitral regurgitation with continuous positive airway pressure treatment28 and decreased severity of CSA after heart valve repair.29 This is well in line with our association of mitral valve surgery and increased AHI and ODI. Our findings require corroboration in larger patient populations. Given the high prevalence of cardiovascular disease in the general population, our small study group, and comparatively small regression coefficients, the relation of OSA and CSA with cardiovascular disease needs further investigation including in-laboratory polysomnography.
8.
9.
10. 11.
12.
13.
14.
15.
16. 17.
18. 19.
20.
21.
22.
23. 24.
1. Bradley TD, Floras JS. Sleep apnea and heart failure: part II: central sleep apnea. Circulation 2003;107:1822–1826. 2. Cistulli PA, Gotsopoulos H, Sullivan CE. Relationship between craniofacial abnormalities and sleep-disordered breathing in Marfan’s syndrome. Chest 2001;120:1455–1460. 3. Kohler M, Blair E, Risby P, Nickol AH, Wordsworth P, Forfar C, Stradling JR. The prevalence of obstructive sleep apnoea and its association with aortic dilatation in Marfan’s syndrome. Thorax 2009; 64:162–166. 4. von Kodolitsch Y, Rybczynski M, Detter C, Robinson PN. Diagnosis and management of Marfan syndrome. Future Cardiol 2008;4:85–96. 5. Rybczynski M, Bernhardt AM, Rehder U, Fuisting B, Meiss L, Voss U, Habermann C, Detter C, Robinson PN, Arslan-Kirchner M, Schmidtke J, Mir TS, Berger J, Meinertz T, von Kodolitsch Y. The spectrum of syndromes and manifestations in individuals screened for suspected Marfan syndrome. Am J Med Genet A 2008;146A:3157–3166. 6. De Paepe A, Devereux RB, Dietz HC, Hennekam RC, Pyeritz RE. Revised diagnostic criteria for the Marfan syndrome. Am J Med Genet 1996;62:417– 426. 7. Rybczynski M, Koschyk DH, Aydin MA, Robinson PN, Brinken T, Franzen O, Berger J, Hofmann T, Meinertz T, von Kodolitsch Y.
25.
26.
27.
28.
29.
1841
Tissue Doppler imaging identifies myocardial dysfunction in adults with Marfan syndrome. Clin Cardiol 2007;30:19 –24. Standards of Practice Committee of the American Sleep Disorders Association. Practice parameters for the use of portable recording in the assessment of obstructive sleep apnea. Sleep 1994;17:372–377. Davies RJ, Ali NJ, Stradling JR. Neck circumference and other clinical features in the diagnosis of the obstructive sleep apnoea syndrome. Thorax 1992;47:101–105. Miletin MS, Hanly PJ. Measurement properties of the Epworth sleepiness scale. Sleep Med 2003;4:195–199. Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med 1999;131:485– 491. Feinberg I, Campbell IG. Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:7– 89. Roman MJ, Devereux RB, Kramer-Fox R, O’Ranghlin J. Two dimensional aortic root dimensions in normal children and adults. Am J Cardiol 1989;64:507–512. Freed LA, Levy D, Levine RA, Larson MG, Evans JC, Fuller DL, Lehman B, Benjamin EJ. Prevalence and clinical outcome of mitralvalve prolapse. N Engl J Med 1999;341:1–7. Zoghbi WA, Enriquez-Sarano M, Foster E, Grayburn PA, Kraft CD, Levine RA, Nihoyannopoulos P, Otto CM, Quinones MA, Rakowski H, Stewart WJ, Waggoner A, Weissman NJ. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16:777– 802. Cistulli PA, Sullivan CE. Sleep-disordered breathing in Marfan’s syndrome. Am Rev Respir Dis 1993;147:645– 648. Verbraecken J, Declerck A, Van de Heyning P, De Backer W, Wouters EF. Evaluation for sleep apnea in patients with Ehlers–Danlos syndrome and Marfan: a questionnaire study. Clin Genet 2001;60:360 –365. Cistulli PA, Sullivan CE. Sleep apnea in Marfan’s syndrome. Increased upper airway collapsibility during sleep. Chest 1995;108:631– 635. Cistulli PA, Sullivan CE. Influence of maxillary morphology on nasal airway resistance in Marfan’s syndrome. Acta Otolaryngol 2000;120: 410 – 413. Sampol G, Romero O, Salas A, Tovar JL, Lloberes P, Sagales T, Evangelista A. Obstructive sleep apnea and thoracic aorta dissection. Am J Respir Crit Care Med 2003;168:1528 –1531. Cistulli PA, Wilcox I, Jeremy R, Sullivan CE. Aortic root dilatation in Marfan’s syndrome: a contribution from obstructive sleep apnea? Chest 1997;111:1763–1766. Verbraecken J, Paelinck BP, Willemen M, Van de Heyning P, De Backer W. Aortic root diameter and nasal intermittent positive airway pressure treatment in Marfan’s syndrome. Clin Genet 2003;63:131–134. Naughton MT, Lorenzi-Filho G. Sleep in heart failure. Prog Cardiovasc Dis 2009;51:339 –349. Yumino D, Wang H, Floras JS, Newton GE, Mak S, Ruttanaumpawan P, Parker JD, Bradley TD. Prevalence and physiological predictors of sleep apnea in patients with heart failure and systolic dysfunction. J Card Fail 2009;15:279 –285. Grewal J, McKelvie R, Lonn E, Tait P, Carlsson J, Gianni M, Jarnert C, Persson H. BNP and NT-proBNP predict echocardiographic severity of diastolic dysfunction. Eur J Heart Fail 2008;10:252–259. Leung RS, Huber MA, Rogge T, Maimon N, Chiu KL, Bradley TD. Association between atrial fibrillation and central sleep apnea. Sleep 2005;28:1543–1546. Sin DD, Fitzgerald F, Parker JD, Newton G, Floras JS, Bradley TD. Risk factors for central and obstructive sleep apnea in 450 men and women with congestive heart failure. Am J Respir Crit Care Med 1999;160:1101–1106. Tkacova R, Liu PP, Naughton MT, Bradley TD. Effect of continuous positive airway pressure on mitral regurgitant fraction and atrial natriuretic peptide in patients with heart failure. J Am Coll Cardiol 1997;30:739 –745. Abe H, Takahashi M, Yaegashi H, Eda S, Kitahara H, Tsunemoto H, Kamikozawa M, Koyama J, Yamazaki K, Ikeda U. Valve repair improves central sleep apnea in heart failure patients with valvular heart diseases. Circ J 2009;73:2148 –2153.
0002-9149/10/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.01.369
remains high due to arrhythmia, heart failure, complications at distal sites of the aortic vessel, and heart valve dysfunction.4 We performed the present study to investigate, first, whether sleep apnea is present in a relevant subset of unselected adults with Marfan syndrome, and second, whether sleep apnea relates to cardiovascular morbidity. Methods We used portable 8-channel monitoring devices for ambulatory, unattended respiratory polygraphy. All patients were ⬎17 years of age, all presented to our clinic within a period of 1 year, all lived within the Hamburg metropolitan area, and all were evaluated according to the Ghent nosology.5,6 We did not enroll patients in whom the Marfan syndrome had been excluded,5 pregnant women, patients with blindness, and patients on continuous positive airway pressure therapy. We continued all medical regimens. We evaluated 106 consecutive outpatients. Of these, 30 patients did not fulfill criteria of Marfan syndrome, 2 were on continuous positive airway pressure treatment, 1 was blind, 1 did not complete our diagnostic program, 2 declined to participate in our study, and 2 had inadequate MiniScreen8 measurements. The remaining 68 patients with Marfan syndrome constituted our study group, with 33 men and 35 women with a mean age of 41 ⫾ 14 years (range 18 to 70; Table 1). The Hamburg research ethics committee approved our protocol, which was in accordance with principles of the Declaration of Helsinki. All subjects gave written informed consent. We performed baseline clinical examinations including basewww.AJConline.org
Miscellaneous/Apnea–Hypopnea Index in Marfan Syndrome
1837
Table 1 Relation between baseline clinical variables and apnea– hypopnea index or oxygen desaturation index in 68 patients with Marfan syndrome Variable
Male gender Age (years) Body weight (kg) Body height (m) Body mass index (kg/m2) Body surface area (m2) Neck circumference (cm) Active smoker Blood glucose (mg/dl) Total cholesterol (mg/dl) High-density lipoprotein cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) Triglyceride (mg/dl) Epworth Sleepiness Score ⬎10 Berlin questionnaire ⱖ2
All Patients (n ⫽ 68)
33 (49%) 41 ⫾ 14 79 ⫾ 20 1.85 ⫾ 0.2 22 ⫾ 4 2 ⫾ 0.2 37 ⫾ 4 5 (8%) 93 ⫾ 17 185 ⫾ 42 58 ⫾ 18 104 ⫾ 35 123 ⫾ 78 13 (19%) 17 (25%)
Range
AHI
18–70 47–127 1.52–2.08 15–32 1.49–2.65 29–50 64–155 103–290 27–104 45–191 40–474
ODI
Correlation Coefficient*
p Value†
Correlation Coefficient*
p Value†
0.29 0.35 0.24 0.01 0.26 0.17 0.23 0.16 0.18 0.23 0.05 0.11 0.16 0.27 0.16
0.06 0.004 0.05 0.9 0.03 0.2 0.06 0.45 0.2 0.08 0.7 0.4 0.2 0.09 0.42
0.26 0.46 0.36 0.09 0.38 0.31 0.36 0.11 0.24 0.35 ⫺0.03 0.26 0.25 0.15 0.27
0.11 ⬍0.001 0.003 0.5 0.002 0.01 0.003 0.71 0.06 0.007 0.8 0.05 0.06 . 46 0.08
Data are presented as mean ⫾ SD or frequency (percentage). * Spearman correlation coefficient (rho) for continuous variables. AHI and ODI were grouped 0 to 5, 6 to 15, and ⬎15 to assess the phi coefficient for binary variables. † The p value of Spearman correlation or phi coefficient.
line serum examinations, blood pressure measurement, standard 12-lead electrocardiography, transthoracic echocardiography,7 and magnetic resonance angiography of the entire aorta within 24 hours of polygraphy in all patients. An experienced technician provided individual training on how to use the MiniScreen8 device and a standardized written and illustrated device instruction to all patients. We used MiniScreen8 type 3 devices (Heinen and Löwenstein, Bad Ems, Germany) in all patients.8 During sleep the devices record nasal flow with a pressure transducer system, oxygen saturation and pulse rate by finger oximetry, body position through a magnetic sensor, snoring sounds, and thoracic and abdominal movements through belts with pneumatic cushions for pressure measurement. All patients used the device for 1 full night at home. We performed automated analysis of downloaded physiological data within 2 days of polygraphy. We also reviewed all electronically stored data manually using EasyScreen Viewer 5.09 (Heinen und Löwenstein) jointly with a board-certified physician of sleep medicine. All investigators were blinded to medical history and other medical data. In 7 subjects with technically insufficient recordings we initiated a second polygraphy yielding adequate data in 5 subjects, and we excluded 2 subjects because they declined to repeat polygraphy. We assessed body height using a wall-fixed height rule, body mass index, body surface area, and neck circumference measured at the level of the cricothyroid membrane.9 We considered active smoking with any inhaled intake of nicotine within ⱕ7 days before polygraphy and we measured fasting blood glucose levels and fasting lipid levels ⱕ24 hours of the study. We used the Epworth Sleepiness Scale questionnaire10 and the Berlin questionnaire11 to assess subjective sleepiness (Table 1). We presented heart rates and blood oxygen saturation as the mean during total recording time with documentation of lowest saturation and longest duration of desaturation. We defined an episode of apnea as the cessation of airflow
lasting ⬎10 seconds and hypopnea as a decrease in airflow of ⱖ50% lasting ⬎10 seconds, associated with a decrease in oxygen saturation of ⱖ4%.12 We quantified severity of sleep apnea as the number of apneas– hypopneas per hour (apnea– hypopnea index [AHI]) and as the number of oxygen desaturations (oxygen desaturation index [ODI]) ⱖ4% per hour of study.12 We used established thresholds for AHI and ODI to classify a sleep-related breathing disorder as absent with ⱕ5 events per hour, as mild with 6 to 15 events per hour, as moderate with ⬎15 events per hour, and as severe with ⬎30 events per hour.12 We considered obstructive apnea as ongoing ventilatory efforts documented by thoracic and abdominal movement signals for ⬎2 respiratory cycles. We defined apnea as central when apnea episodes presented without any ongoing ventilatory efforts (Figure 1), and mixed apnea when airflow continued to be absent despite resumption of respiratory efforts in an apnea episode with initial cessation of respiratory efforts.12 We assessed snoring severity as the number of snoring events per hour (Table 2). We obtained systolic and diastolic blood pressures on sphygmomanometer after 15 minutes of rest in a supine position and we documented heart rates and presence of sinus rhythm or atrial fibrillation on standard 12-lead electrocardiograms.5 We performed 2-dimensional standard echocardiography for left ventricular ejection fraction (LVEF)7 and maximum aortic diameters at the level of the aortic sinuses according to Roman et al.13 We used magnetic resonance angiography for diameters of the ascending and descending aortas at established levels. We did not consider diameters at aortic sites with an aortic prosthesis. We noted aortic dissection with a history of dissection or intramural hemorrhage at any site of the aorta reported in surgical records or on tomographic images. We considered proximal aortic surgery with a documented history of a David operation (19 patients with type 1 operation), a con-
1838
The American Journal of Cardiology (www.AJConline.org)
Figure 1. Polygraphic records of oxygen saturation (SaO2), pulse rate (Puls), oronasal airflow (Flow), thoracic ventilatory effort (Thorax), abdominal ventilatory effort (Abdomen), registration of snoring sounds (Schnarchen), and body position (Lage). (Top) A 43-year-old man with classic Marfan syndrome who had undergone an elective David operation 2 years previously. LVEF was 74%, and the N-terminal pro– brain natriuretic peptide level was 81.7 pg/ml. The patient was in a right lateral body position and he exhibited a typical pattern of obstructive episodes of apnea and snoring sounds. (Bottom) A 41-year-old man with classic Marfan syndrome who had undergone surgical reconstruction of his mitral valve at the age of 36 years and a David operation 1 year later. He was in New York Heart Association class III, his LVEF was 23%, and his N-terminal pro– brain natriuretic peptide level was 1,038 pg/ml. The patient was in a supine body position, he did not snore, and he had a typical breathing pattern of nonobstructive apnea with Cheyne–Stokes respiration.
duit operation with an artificial (15 patients) or a biological (3 patients) valve prosthesis, or with a Yacoub operation (2 patients). We documented distal aortic intervention with
presence of a tube graft in the descending (4 patients) or abdominal (3 patients) aorta. We considered mitral valve prolapse with prolapse on 2-dimensional echocardiogram
Miscellaneous/Apnea–Hypopnea Index in Marfan Syndrome
1839
Table 2 Relation between polygraphic variables and apnea– hypopnea index or oxygen desaturation index in 68 patients with Marfan syndrome Variable
Total recording time (minutes) Mean heart rate (beats/min) Oxyhemoglobin saturation Mean (%) Lowest (%) Longest desaturation (minutes) Episodes of apnea (number)‡ Total Obstructive Central Mixed Longest apnea (seconds) Episodes of hypopnea (number) Longest hypopnea (seconds) Snore index (per hour)
All Patients (n ⫽ 68)
Range
AHI
601 ⫾ 122 65 ⫾ 11
409–803 45–103
0 0.14
92 ⫾ 12 80 ⫾ 19 1.1 ⫾ 0.9
88–96 66–94 0.12–3.43
27 ⫾ 49 12 ⫾ 27 11 ⫾ 25 3⫾9 24 ⫾ 17 18 ⫾ 24 36 ⫾ 33 67 ⫾ 126
0–289 0–163 0–145 0–66 11–87 0–109 11–181 0–111
ODI
Correlation Coefficient*
p Value†
Correlation Coefficient*
p Value†
0.9 0.3
⫺0.07 0.35
0.6 0.004
⫺0.44 ⫺0.35 0.47
⬍0.001 0.004 ⬍0.001
⫺0.51 ⫺0.51 0.64
⬍0.001 ⬍0.001 ⬍0.001
0.63 0.49 0.34 0.47 0.59 0.76 0.62 0.33
⬍0.001 ⬍0.001 0.006 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.008
0.24 ⫺0.04 0.07 0.25 0.27 0.88 0.58 0.44
0.05 0.6 0.6 0.048 0.03 ⬍0.001 ⬍0.001 ⬍0.001
Data are presented as mean ⫾ SD. * Spearman correlation coefficient (rho). † The p value for Spearman correlation. ‡ Number of episodes measured during total recording time.
Table 3 Relation between cardiovascular disease variables and apnea– hypopnea index or oxygen desaturation index in 68 patients with Marfan syndrome Variables
Mean systolic blood pressure (mm Hg) Mean diastolic blood pressure (mm Hg) Baseline heart rate (beats/min) Atrial fibrillation Left ventricular ejection fraction (%) N-terminal pro–brain natriuretic peptide (pg/ml) Aortic root diameter (mm) Ascending aorta diameter (mm) Descending aorta diameter (mm) Aortic dissection Proximal aortic surgery Distal aortic intervention Mitral valve prolapse Mitral valve regurgitation Mitral valve surgery Any aortic medication
All Patients (n ⫽ 68)
Range
125 ⫾ 13 70 ⫾ 10 65 ⫾ 11 5 (8%) 55 ⫾ 11 263 ⫾ 316 38 ⫾ 5 29 ⫾ 6 25 ⫾ 7 14 (21%) 36 (53%) 7 (10%) 36 (53%) 12 (18%) 8 (12%) 38 (56%)
100–162 50–90 45–89 23–75 25–1,444 27–51 20–44 16–48
AHI
ODI
Correlation Coefficient*
p Value†
Correlation Coefficient*
p Value†
0.09 0.04 0.07 0.43 ⫺0.33 0.35 0.14 0.27 0.44 0.24 0.13 0.12 0.06 0.28 0.34 0.28
0.5 0.8 0.6 0.002 0.01 0.004 0.5 0.1 0.001 0.15 0.58 0.59 0.90 0.07 0.02 0.08
0.13 0.13 0.10 0.36 ⫺0.17 0.40 0.26 0.40 0.47 0.14 0.13 0.15 0.05 0.40 0.40 0.15
0.3 0.3 0.5 0.014 0.2 ⬍0.001 0.2 0.03 ⬍0.001 0.54 0.57 0.49 0.92 0.005 0.01 0.46
Data are presented as mean ⫾ SD or frequency (percentage). * Spearman correlation coefficient (rho) for continuous variables. AHI and ODI were grouped 0 to 5, 6 to 15, and ⬎15 to assess the phi coefficient for binary variables. † The p value of Spearman correlation or of the phi coefficient.
irrespective presence of valvular regurgitation14 and mitral regurgitation with at least moderate insufficiency diagnosed by 2-dimensional and Doppler echocardiography.15 Mitral valve surgery was present with reconstruction (5 patients) or replacement (3 patients) of the mitral valve. We considered patients on aortic medication with intake of  blockers (metoprolol or bisoprolol in 37 patients), angiotensin-converting enzyme inhibitors (ramipril in 14 patients), angiotensin II receptor blockers (valsartan or candesartan in 7
patients), or calcium antagonists (amlodipine or lercanidipine in 5 patients). In addition, we measured N-terminal pro– brain natriuretic peptide serum levels with an electrochemiluminescence sandwich immunoassay (Roche Diagnostics GmbH, Mannheim, Germany) on an Elecsys System 2010 with a detection limit ⱖ5 pg/ml (Table 3). We presented all data as mean ⫾ SD with ranges or frequencies. To evaluate the association of baseline patient characteristics (Table 1), polygraphic measurements (Table
1840
The American Journal of Cardiology (www.AJConline.org)
Table 4 Multivariate ordered logistic regression analysis (logit model) of cardiovascular disease variables in 68 patients with Marfan syndrome Variable
Atrial fibrillation Left ventricular ejection fraction (%) N-terminal pro–brain natriuretic peptide (ng/ml) Ascending aorta diameter (mm) Descending aorta diameter (mm) Mitral regurgitation Mitral valve surgery
ODI†
AHI* Regression Coefficient
95% CI
p Value
Regression Coefficient
95% CI
p Value
2.07 ⫺0.07 0.002
0.13–4.01 ⫺0.13 to ⫺0.002 0.0004–0.04
0.036 0.043 0.016
3.17
0.46–5.89
0.022
⫺0.006
⫺0.103 to 0.091
1.75
0.14–3.36
0.901 0.033
0.002 0.032 0.034 1.68 2.71
0.0001–0.004 ⫺0.191 to 0.256 ⫺0.059 to 0.127 ⫺0.07 to 3.43 0.69–4.73
0.041 0.775 0.479 0.06 0.008
* Each variable was adjusted for age and body mass index. † Each variable was adjusted for age, body mass index, neck circumference, and total cholesterol; due to a smaller number of cases, the variables ascending aorta diameter and descending aorta diameter were adjusted only for age and neck circumference but not for total cholesterol. CI ⫽ confidence interval.
2), and cardiovascular disease variables (Table 3) with AHI or ODI we used the Spearman correlation coefficient (rho) for continuous variables and the phi coefficient for nominal variables. We used these nonparametric tests because of the value distribution characteristics of AHI and ODI. To calculate the correlation coefficient rho, we used AHI and ODI as a continuous variable, and for phi coefficients with construction of contingence tables, we grouped AHI and ODI as 0 to 5, 6 to 15, and ⬎15/hour. To test, whether cardiovascular disease variables were associated with AHI or ODI independently of baseline clinical characteristics, we performed ordered logistic regression with grouped values of AHI or ODI as the dependent variable (Table 4). For ordered logistic regression, we considered only baseline characteristics that were associated with AHI or ODI at a level of significance set at a p value ⬍0.05 (2-sided). The Brant test or approximate likelihood ratio test corroborated the model assumption of proportional odds in all analyses. We used STATA 9.0 (STATA Corp., College Station, Texas) for ordered logistic regression including tests to check the model assumption and SPSS 15.0 for Windows (SPSS, Inc., Chicago, Illinois) for all other tests. Results We found an AHI ⱕ5/hour in 47 subjects (69%), an AHI 6 to 15/hour in 14 subjects (21%), and an AHI ⬎15/hour in 7 subjects (10%), including 2 subjects with an AHI ⬎30/ hour. Similarly, the ODI was ⱕ5/hour in 48 subjects (71%), 6 to 15/hour in 12 subjects (19%), and ⬎15/hour in 8 subjects (12%), including 3 subjects with an ODI ⬎30. Only older age and higher body mass index were related to AHI and ODI, whereas other epidemiologic and clinical characteristics were related only to AHI or ODI or were unrelated to AHI and ODI (Table 1). About 1/2 of all episodes of apnea were classified as obstructive (12 ⫾ 27), whereas remaining episodes of apnea were classified as central (11 ⫾ 25) or mixed (3 ⫾ 9; Figure 1, Table 2). Increased frequency of atrial fibrillation, increased N-terminal pro– brain natriuretic peptide levels, enlarged diameters of the descending aorta, and mitral valve surgery related to AHI and ODI. Conversely, a decreased LVEF related only to AHI, and enlarged diameters of the ascending aorta and increased frequency of mitral regurgitation
related only to ODI (Table 3). Ordered logistic regression showed that atrial fibrillation, LVEF, N-terminal pro– brain natriuretic peptide levels, and mitral valve surgery were associated with AHI independently of age and body mass index. Similarly, atrial fibrillation, N-terminal pro– brain natriuretic peptide levels, and mitral valve surgery were associated with ODI (Table 4). Discussion Our study documents that about 1/3 of unselected adults with Marfan syndrome exhibited sleep apnea compared to 12% reported in the general population of similar age as our study patients.3 Surprisingly, we found that about 1/2 of all episodes of apnea in patients with Marfan syndrome were central rather than obstructive. Moreover, sleep apnea related to occurrence of atrial fibrillation, decreased LVEF, increased N-terminal pro– brain natriuretic peptide serum levels, and increased frequency of mitral valve surgery. The composition of our study group was not biased for patient characteristics such as medication, presence of previous surgery, or symptoms of sleep-disordered breathing. The only other study of unselected adults with Marfan syndrome corroborated our estimate of 30% of sleep apnea in patients with Marfan syndrome.3 Conversely, the higher frequency of 64% of sleep apnea in a study of 25 patients with Marfan syndrome may reflect some selection bias.16 A questionnaire-based study of 15 patients with Marfan syndrome also supported a frequency of only 27% of sleep apnea.17 Our study demonstrated that only age and body mass index were associated with increased AHI and ODI, whereas predisposing conditions such as body weight, body surface area, enlarged neck circumference, total cholesterol levels, active smoking, or hyperglycemia were unrelated to AHI and/or ODI. Similarly, signs of subjective sleepiness did not relate to AHI or ODI in our patients.3 Thus, classic clues for sleep apnea may be of limited value for identifying increased risk of sleep apnea in Marfan syndrome. Thus, we suggest screening patients with Marfan syndrome even when sleep apnea is not clinically suspected. Cistulli et al2,18,19 suggested upper airway collapsibility, high nasal airway resistance, and craniofacial abnormalities as causal mechanisms of OSA in Marfan syndrome, whereas
Miscellaneous/Apnea–Hypopnea Index in Marfan Syndrome
Kohler et al3 did not find increased frequencies of retrognathia or abnormalities of mean neck circumference, cricomental distance, and Mallampati score in their patients. Kohler et al used ApneaLink type 4 devices (ResMed, MAP Medicine Technology, Martinsried, Germany), which permitted only flow limitations and snoring to be measured3 but which were therefore of limited use for distinguishing OSA from CSA. Conversely, our type 3 devices measured thoracic and abdominal ventilatory efforts and thus provided better means to identify obstructive and central apnea (Figure 1). We found only weak relations between aortic diameters and AHI and ODI, most likely because many patients previously underwent operation of the aorta and could not be considered for analysis. However, previous studies have found an association of AHI with aortic dilatation3 and aortic dissection,20 and 3 reports of patients with Marfan syndrome have documented cessation of aortic diameter progression on continuous positive airway pressure therapy.21,22 Patients with Marfan syndrome exhibit primary myocardial dysfunction.4 In patients without Marfan syndrome, the association of myocardial dysfunction with OSA and CSA is well established.23,24 In our patients sleep apnea was independently associated with decreased LVEF and increased N-terminal pro– brain natriuretic peptide levels. It is likely that increased N-terminal pro– brain natriuretic peptide levels were not caused by systolic dysfunction alone, but also by diastolic25 dysfunction, which usually evolves in Marfan syndrome before LVEF decreases.7 An increased risk for atrial fibrillation is established only in CSA with preserved26 and impaired27 LV function, but not in OSA.1 Interestingly, our study with a high frequency of CSA documents an association of AHI and ODI with atrial fibrillation. Studies of chronic heart failure have demonstrated decreased mitral regurgitation with continuous positive airway pressure treatment28 and decreased severity of CSA after heart valve repair.29 This is well in line with our association of mitral valve surgery and increased AHI and ODI. Our findings require corroboration in larger patient populations. Given the high prevalence of cardiovascular disease in the general population, our small study group, and comparatively small regression coefficients, the relation of OSA and CSA with cardiovascular disease needs further investigation including in-laboratory polysomnography.
8.
9.
10. 11.
12.
13.
14.
15.
16. 17.
18. 19.
20.
21.
22.
23. 24.
1. Bradley TD, Floras JS. Sleep apnea and heart failure: part II: central sleep apnea. Circulation 2003;107:1822–1826. 2. Cistulli PA, Gotsopoulos H, Sullivan CE. Relationship between craniofacial abnormalities and sleep-disordered breathing in Marfan’s syndrome. Chest 2001;120:1455–1460. 3. Kohler M, Blair E, Risby P, Nickol AH, Wordsworth P, Forfar C, Stradling JR. The prevalence of obstructive sleep apnoea and its association with aortic dilatation in Marfan’s syndrome. Thorax 2009; 64:162–166. 4. von Kodolitsch Y, Rybczynski M, Detter C, Robinson PN. Diagnosis and management of Marfan syndrome. Future Cardiol 2008;4:85–96. 5. Rybczynski M, Bernhardt AM, Rehder U, Fuisting B, Meiss L, Voss U, Habermann C, Detter C, Robinson PN, Arslan-Kirchner M, Schmidtke J, Mir TS, Berger J, Meinertz T, von Kodolitsch Y. The spectrum of syndromes and manifestations in individuals screened for suspected Marfan syndrome. Am J Med Genet A 2008;146A:3157–3166. 6. De Paepe A, Devereux RB, Dietz HC, Hennekam RC, Pyeritz RE. Revised diagnostic criteria for the Marfan syndrome. Am J Med Genet 1996;62:417– 426. 7. Rybczynski M, Koschyk DH, Aydin MA, Robinson PN, Brinken T, Franzen O, Berger J, Hofmann T, Meinertz T, von Kodolitsch Y.
25.
26.
27.
28.
29.
1841
Tissue Doppler imaging identifies myocardial dysfunction in adults with Marfan syndrome. Clin Cardiol 2007;30:19 –24. Standards of Practice Committee of the American Sleep Disorders Association. Practice parameters for the use of portable recording in the assessment of obstructive sleep apnea. Sleep 1994;17:372–377. Davies RJ, Ali NJ, Stradling JR. Neck circumference and other clinical features in the diagnosis of the obstructive sleep apnoea syndrome. Thorax 1992;47:101–105. Miletin MS, Hanly PJ. Measurement properties of the Epworth sleepiness scale. Sleep Med 2003;4:195–199. Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med 1999;131:485– 491. Feinberg I, Campbell IG. Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:7– 89. Roman MJ, Devereux RB, Kramer-Fox R, O’Ranghlin J. Two dimensional aortic root dimensions in normal children and adults. Am J Cardiol 1989;64:507–512. Freed LA, Levy D, Levine RA, Larson MG, Evans JC, Fuller DL, Lehman B, Benjamin EJ. Prevalence and clinical outcome of mitralvalve prolapse. N Engl J Med 1999;341:1–7. Zoghbi WA, Enriquez-Sarano M, Foster E, Grayburn PA, Kraft CD, Levine RA, Nihoyannopoulos P, Otto CM, Quinones MA, Rakowski H, Stewart WJ, Waggoner A, Weissman NJ. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16:777– 802. Cistulli PA, Sullivan CE. Sleep-disordered breathing in Marfan’s syndrome. Am Rev Respir Dis 1993;147:645– 648. Verbraecken J, Declerck A, Van de Heyning P, De Backer W, Wouters EF. Evaluation for sleep apnea in patients with Ehlers–Danlos syndrome and Marfan: a questionnaire study. Clin Genet 2001;60:360 –365. Cistulli PA, Sullivan CE. Sleep apnea in Marfan’s syndrome. Increased upper airway collapsibility during sleep. Chest 1995;108:631– 635. Cistulli PA, Sullivan CE. Influence of maxillary morphology on nasal airway resistance in Marfan’s syndrome. Acta Otolaryngol 2000;120: 410 – 413. Sampol G, Romero O, Salas A, Tovar JL, Lloberes P, Sagales T, Evangelista A. Obstructive sleep apnea and thoracic aorta dissection. Am J Respir Crit Care Med 2003;168:1528 –1531. Cistulli PA, Wilcox I, Jeremy R, Sullivan CE. Aortic root dilatation in Marfan’s syndrome: a contribution from obstructive sleep apnea? Chest 1997;111:1763–1766. Verbraecken J, Paelinck BP, Willemen M, Van de Heyning P, De Backer W. Aortic root diameter and nasal intermittent positive airway pressure treatment in Marfan’s syndrome. Clin Genet 2003;63:131–134. Naughton MT, Lorenzi-Filho G. Sleep in heart failure. Prog Cardiovasc Dis 2009;51:339 –349. Yumino D, Wang H, Floras JS, Newton GE, Mak S, Ruttanaumpawan P, Parker JD, Bradley TD. Prevalence and physiological predictors of sleep apnea in patients with heart failure and systolic dysfunction. J Card Fail 2009;15:279 –285. Grewal J, McKelvie R, Lonn E, Tait P, Carlsson J, Gianni M, Jarnert C, Persson H. BNP and NT-proBNP predict echocardiographic severity of diastolic dysfunction. Eur J Heart Fail 2008;10:252–259. Leung RS, Huber MA, Rogge T, Maimon N, Chiu KL, Bradley TD. Association between atrial fibrillation and central sleep apnea. Sleep 2005;28:1543–1546. Sin DD, Fitzgerald F, Parker JD, Newton G, Floras JS, Bradley TD. Risk factors for central and obstructive sleep apnea in 450 men and women with congestive heart failure. Am J Respir Crit Care Med 1999;160:1101–1106. Tkacova R, Liu PP, Naughton MT, Bradley TD. Effect of continuous positive airway pressure on mitral regurgitant fraction and atrial natriuretic peptide in patients with heart failure. J Am Coll Cardiol 1997;30:739 –745. Abe H, Takahashi M, Yaegashi H, Eda S, Kitahara H, Tsunemoto H, Kamikozawa M, Koyama J, Yamazaki K, Ikeda U. Valve repair improves central sleep apnea in heart failure patients with valvular heart diseases. Circ J 2009;73:2148 –2153.