severe obstructive sleep apnea elicits concentric left ventricular geometry

10
CE: Namrta; HJH/201793; Total nos of Pages: 9; HJH 201793 Severe obstructive sleep apnea elicits concentric left ventricular geometry Giovanni Cioffi a , Tiziano Edoardo Russo a , Carlo Stefenelli a , Alessandro Selmi a , Francesco Furlanello a , Dana Cramariuc b , Eva Gerdts b and Giovanni de Simone c Background Obstructive sleep apnea (OSA) has several negative effects on the heart including increase in myocardial end-systolic stress, venous return and sympathetic activity, all potential stimuli of left ventricular (LV) hypertrophy. The impact of the severity of OSA on LV geometry is unknown. We hypothesized that OSA is related to concentric LV geometry. Methods One hundred and fifty-seven patients with suspected OSA underwent echocardiography, ambulatory 24-h blood pressure and ECG monitoring. On the basis of the severity of OSA, patients were divided into controls, mild OSA and moderate/severe OSA (apnea–hypopnea index <5, 5–15 and >15/h, respectively). Patients with LV hypertrophy were defined as LV mass at least 49.2 g/m 2.7 for men and at least 46.7 for women. Relative wall thickness of at least 0.43 identified patients with concentric LV geometry. Results Patients with moderate/severe OSA (n U 86) had a higher body mass index and a higher prevalence of paroxysmal atrial fibrillation than those (n U 51) with mild OSA and controls (n U 20). Prevalence of hypertension, diabetes, obesity, LV mass and blood pressure did not differ between the groups. Relative wall thickness was positively related to apnea–hypopnea index (r U 0.30; P U 0.003) and the prevalence of concentric LV geometry was 20% in controls, 12% in mild OSA and 58% in moderate/severe OSA (P < 0.001). In logistic regression analysis concentric LV geometry was associated with moderate/severe OSA [odds ratio (OR) 7.6, P < 0.001], low stress-corrected midwal shortening (OR 3.38, P U 0.004), and higher body mass index (OR 1.09, P U 0.03). Conclusions Moderate/severe OSA is associated with high prevalence of concentric LV geometry. This increased prevalence may in part explain the increased rate of cardiovascular events in these patients. J Hypertens 28:000–000 Q 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins. Journal of Hypertension 2010, 28:000–000 Keywords: apnea–hypopnea index, concentric hypertrophy, desaturations, left ventricular geometry, obstructive sleep apnea Abbreviations: AHI, apnea–hypopnea index; ECG, electrocardiogram; LV, left ventricular; LVH, left ventricular hypertrophy; OSA, obstructive sleep apnea a Department of Cardiology, Villa Bianca Hospital, Trento, Italy, b Department of Heart Disease, Haukeland University, Bergen, Norway and c Department of Clinical and Experimental Medicine, Federico II, University Hospital, School of Medicine, Naples, Italy Correspondence to Giovanni Cioffi, MD, Echocardiography Laboratory, Villa Bianca Hospital, via Piave 78, 38100 Trento, Italy Tel: +39 0 461 916000; fax: +39 0 461 916874; e-mail: gcioffi@villabiancatrento.it Received 12 August 2009 Revised 8 November 2009 Accepted 18 December 2009 Introduction Obstructive sleep apnea (OSA) is a common under- diagnosed pathological condition implicated in the initiation and progression of cardiovascular diseases [1,2]. OSA has several negative effects on the heart including increase in left ventricular (LV) afterload, venous return (influencing preload) and sympathetic activity, all potential stimuli of LV hypertrophy (LVH) [1,3,4]. The magnitude of these effects parallels the severity of OSA. Patients with OSA have a high preva- lence of arterial hypertension and, more importantly, obesity, which represents the strongest predisposing factor for the sleep abnormalities [5,6]. The relation between hypertension and obesity and its impact on LV geometry are known [7,8]. Less is known about the impact of OSA on LV geometry. OSA should hypothetically produce changes in LV geometry lead- ing to a concentric remodeling through sympathetic activation, vasoconstriction and elevation in blood pressure. Accordingly, we studied a group of patients with OSA to test the hypothesis that higher degree of sleep disturb- ance is related to concentric LV geometry. Methods Study population The study group included participants above 18 years of age with ascertained OSA by a formal sleep study, who were addressed by their General Practitioners to our center for a more precise stratification of their global cardiovascular risk. All participants were free of symp- toms and clinical signs of cardiac disease (New York Heart Association functional class I). Recruitment period Original article 1 0263-6352 ß 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins DOI:10.1097/HJH.0b013e328336c90a

Upload: unina

Post on 17-Jan-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

Severe obstructive sleep apnea elicits concentric leftventricular geometryGiovanni Cioffia, Tiziano Edoardo Russoa, Carlo Stefenellia, Alessandro Selmia,Francesco Furlanelloa, Dana Cramariucb, Eva Gerdtsb andGiovanni de Simonec

Background Obstructive sleep apnea (OSA) has several

negative effects on the heart including increase in

myocardial end-systolic stress, venous return and

sympathetic activity, all potential stimuli of left ventricular

(LV) hypertrophy. The impact of the severity of OSA on

LV geometry is unknown. We hypothesized that OSA is

related to concentric LV geometry.

Methods One hundred and fifty-seven patients with

suspected OSA underwent echocardiography, ambulatory

24-h blood pressure and ECG monitoring. On the basis of

the severity of OSA, patients were divided into controls, mild

OSA and moderate/severe OSA (apnea–hypopnea index

<5, 5–15 and >15/h, respectively). Patients with LV

hypertrophy were defined as LV mass at least 49.2 g/m2.7

for men and at least 46.7 for women. Relative wall thickness

of at least 0.43 identified patients with concentric LV

geometry.

Results Patients with moderate/severe OSA (n U 86) had a

higher body mass index and a higher prevalence of

paroxysmal atrial fibrillation than those (n U 51) with mild

OSA and controls (n U 20). Prevalence of hypertension,

diabetes, obesity, LV mass and blood pressure did not differ

between the groups. Relative wall thickness was positively

related to apnea–hypopnea index (r U 0.30; P U 0.003) and

the prevalence of concentric LV geometry was 20% in

controls, 12% in mild OSA and 58% in moderate/severe

OSA (P < 0.001). In logistic regression analysis concentric

LV geometry was associated with moderate/severe OSA

[odds ratio (OR) 7.6, P < 0.001], low stress-corrected midwal

shortening (OR 3.38, P U 0.004), and higher body mass

index (OR 1.09, P U 0.03).

Conclusions Moderate/severe OSA is associated with high

prevalence of concentric LV geometry. This increased

prevalence may in part explain the increased rate of

cardiovascular events in these patients. J Hypertens

28:000–000 Q 2010 Wolters Kluwer Health | Lippincott

Williams & Wilkins.

Journal of Hypertension 2010, 28:000–000

Keywords: apnea–hypopnea index, concentric hypertrophy, desaturations,left ventricular geometry, obstructive sleep apnea

Abbreviations: AHI, apnea–hypopnea index; ECG, electrocardiogram; LV,left ventricular; LVH, left ventricular hypertrophy; OSA, obstructive sleepapnea

aDepartment of Cardiology, Villa Bianca Hospital, Trento, Italy, bDepartment ofHeart Disease, Haukeland University, Bergen, Norway and cDepartment ofClinical and Experimental Medicine, Federico II, University Hospital, School ofMedicine, Naples, Italy

Correspondence to Giovanni Cioffi, MD, Echocardiography Laboratory,Villa Bianca Hospital, via Piave 78, 38100 Trento, ItalyTel: +39 0 461 916000; fax: +39 0 461 916874;e-mail: [email protected]

Received 12 August 2009 Revised 8 November 2009Accepted 18 December 2009

IntroductionObstructive sleep apnea (OSA) is a common under-

diagnosed pathological condition implicated in the

initiation and progression of cardiovascular diseases

[1,2]. OSA has several negative effects on the heart

including increase in left ventricular (LV) afterload,

venous return (influencing preload) and sympathetic

activity, all potential stimuli of LV hypertrophy (LVH)

[1,3,4]. The magnitude of these effects parallels the

severity of OSA. Patients with OSA have a high preva-

lence of arterial hypertension and, more importantly,

obesity, which represents the strongest predisposing

factor for the sleep abnormalities [5,6]. The relation

between hypertension and obesity and its impact on

LV geometry are known [7,8]. Less is known about

the impact of OSA on LV geometry. OSA should

hypothetically produce changes in LV geometry lead-

ing to a concentric remodeling through sympathetic

activation, vasoconstriction and elevation in blood

pressure.

Accordingly, we studied a group of patients with OSA to

test the hypothesis that higher degree of sleep disturb-

ance is related to concentric LV geometry.

MethodsStudy populationThe study group included participants above 18 years of

age with ascertained OSA by a formal sleep study, who

were addressed by their General Practitioners to our

center for a more precise stratification of their global

cardiovascular risk. All participants were free of symp-

toms and clinical signs of cardiac disease (New York

Heart Association functional class I). Recruitment period

Original article 1

0263-6352 � 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins DOI:10.1097/HJH.0b013e328336c90a

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

lasted from January 2007 to March 2008. Exclusion

criteria were presence of coronary heart disease (diag-

nosed by clinical, electrocardiographic and echocardio-

graphic evaluation at rest and by the results of exercise/

echo-stress test or coronary angiography), LV ejection

fraction less than 50%, LV wall motion abnormalities, or

more-than-mild valve regurgitation or any grade of

valve stenosis.

Study protocolAll patients underwent a complete clinical and echocar-

diographic evaluation which included Berlin question-

naire test [9], the measurement of body weight and

waist, arterial blood pressure, serum creatinine, hemo-

globin, haematocrit, microalbuminuria and reactive C

protein, polysomnography, transthoracic echocardiogra-

phy, maximal exercise test by cycloergometer, 24-h ECG

Holter monitoring and 24-h blood pressure Holter

monitoring. Physical and echocardiographic examin-

ations were performed the same morning blood samples

were collected. Polysomnography, exercise test, spirome-

try, 24-h ECG and blood pressure Holter monitoring

were performed within the next 3 days. Variables con-

sidered in the statistical analysis derived by 24-h blood

pressure Holter monitoring were mean day-time systolic

and diastolic blood pressure, mean nocturnal systolic

and diastolic blood pressure, mean diurnal systolic and

diastolic blood pressure, mean systolic and diastolic day/

night-time excursion. History of hypertension and num-

ber of antihypertension drugs were also considered in the

study analyses. As parameters of sympathetic activity

we considered baseline heart rate, mean day-time and

nocturnal heart rate and day-time night-time heart rate

excursion. The protocol of the present study was

approved by the local institutional review board and

written informed consent was obtained from all patients.

Polysomnography and definitionsAll participants underwent a standard overnight poly-

somnography while breathing room air (Hypno PTT

machine - Tayco HealthCare). The following parameters

were considered: apneas; hypoapneas; duration of apneas/

hypopneas (s), oxygen saturation (%) and arousals.

Apneas were defined as an absence (<20% baseline) of

airflow for at least 10 s; hypoapneas as a reduction in

airflow (20–50% baseline) producing a reduction in O2

saturation at least 3%; desaturations were recognized in

presence of a reduction of oxygen saturation by at least

3% from baseline and expressed as total number/hours of

sleep [10]; arousals were identified from the signals

received from the electrodes positioned on the chest

and were expressed as total number/hours of sleep.

Obstructive and central events were discriminated by

the analysis of the respiratory effort and the typical

respiratory pattern. Mixed events were considered as

obstructive. OSA was identified by the apnea–hypopnea

index (AHI¼ rate of apneasþhypopneas per hour of

sleep), which was also used for classifying patients

according to the severity of the syndrome. Patients were

thus categorized in two groups according to suggested

cut-points [6]: ‘mild OSA’ (AHI between 5 and 15) and

‘moderate/severe OSA’ (AHI >15). Patients with AHI

less than 5 were retained to have not OSA and were

considered as control group.

Hypertension was defined as pharmacologically treated

high blood pressure or a resting blood pressure at least

140/90 mmHg. Overweight was diagnosed according to

the clinical guidelines of National Institute of Health [11]

in presence of a body mass index at least 27.8 kg/m2

in men and at least 27.3 kg/m2 in women; obesity

was defined as body mass index equal or greater than

30 kg/m2. Paroxysmal atrial fibrillation was diagnosed by

24-h Holter monitoring as the presence of a burst of atrial

arrhythmia lasting more than 30 s [12]. Functional

capacity was determined by maximal exercise test

(cycloergometer) and measured in watts.

EchocardiographyThe echocardiographic evaluations were performed by

transthoracic approach using a commercially available

instrument (MEGAS Esaote Biomedical System, Flor-

ence, Italy) equipped with a 2.5–3.5 Mhz annular-array

transducer. Echocardiographic measurements were

blinded to the presence and magnitude of OSA.

Left ventricular chamber dimensions (normalized for

height), interventricular septum, and posterior wall thick-

ness were measured from M-mode tracings according to

the American Society of Echocardiography [13] and LV

mass calculated using a necropsy validated formula [14].

LV mass was normalized for height to the 2.7 power and

LVH was defined as LV mass at least 49.2 g/m2.7 for men

and at least 46.7 for women [15]. Relative wall thickness

was calculated as the end-diastolic ratio posterior wall

thickness/LV diameter and indicated concentric LV geo-

metry if at least 0.43 (the 97.5 percentile in normal

population) [16].

Concentric geometry included concentric LVH (if LV

mass index was increased) or concentric remodeling (if

LV mass index was normal). Similarly, in patients without

concentric LV geometry (relative wall thickness <0.43),

two patterns of LV geometry were identified: eccentric

LVH (if LV mass index was increased) and normal (if LV

mass index was normal).

Left ventricular volumes were estimated by the

Z-derived formulas [17] and used to generate ejection

fraction and stroke volume. Stroke volume was normal-

ized for height at the allometric power of 2.04 to account

for its nonlinear variation with body size [18]. Myocardial

contractility was assessed by the midwall shortening and

by end-systolic stress-corrected midwall shortening

2 Journal of Hypertension 2010, Vol 00 No 00

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

[19,20]. Stress-corrected midwall shortening less than

87% in men and less than 90% in women was indicative

of impaired LV myocardial contractility [21].

Statistical analysisData are reported as mean values� 1 standard deviation

for parametric variables and median and quartiles for

nonparametric variables (i.e. AHI and desaturations).

Between-group comparisons of categorical and continu-

ous variables were performed by x2 test and analysis of

variance (ANOVA), respectively. To assess whether

moderate/severe OSA was an independent determinant

of concentric LV geometry, the relations of concentric

LV geometry to severity of OSA, demographics, indexes

of LV systolic and diastolic function and measures of

hypertension were investigated by multiple logistic

regression analyses using a stepwise forward procedure

(SPSS/PC 11 Release, SPSS Inc. Chicago, IL). Measures

of hypertension were included in the model one by one to

avoid collinearity. A two-tailed value of P< 0.05 was used

to reject the null hypothesis.

ResultsDescription of the study populationThe study population consisted of 157 patients (mean age

of 61� 13 years, 17% women) and had a high prevalence

of arterial hypertension (72%), dyslipidemia (58%),

obesity (44%), and diabetes mellitus (18%). LVH was

detected in 45% of cases, and concentric geometry in

38%. LV myocardial contractility measured at midwall

level was impaired in 53% of patients. Spirometry results

were normal in 90% of patients, and the maximal func-

tional capacity to exercise test was 133� 40 watts. The

median AHI was 17 [9–31] and the median duration of

apnea–hypopnea episodes was 24 s [21–26]. OSA was

mild in 51 patients (32%) and moderate to severe in 86

patients (55%). Twenty patients (13%) had AHI less than

5 and formed the control group. The clinical character-

istics of these patients were similar to those with mild

OSA, but they were treated with a smaller number of

antihypertension drugs and less frequently received ACE

inhibitors/AT1-receptor blockers.

Clinical features of patients with mild andmoderate/severe obstructive sleep apneaPatients with moderate/severe OSA had a higher body

mass index and a higher prevalence of obesity and

paroxysmal atrial fibrillation than patients with mild

OSA, but had similar prevalence of dyslipidemia, dia-

betes mellitus and hypertension. Mean values of blood

pressure were not different between the groups both

when measured in the day-time and night-time but

patients with moderate/severe OSA patients used a

higher number of antihypertensive drugs and received

more frequently diuretics and vasodilators than those

with mild OSA (Table 1). Parameters of renal function,

inflammatory state, hematocrit, maximal functional

capacity and spirometric pattern were not different

between the groups (data not shown).

Left ventricular systolic performance and functionEchocardiographic features of the study groups are shown

in Table 2. Parameters of LV performance and function

were similar between controls and patients with mild

OSA. Patients with moderate/severe OSA exhibited a

decreased LV myocardial contractility, as measured at the

midwall level, than patients with mild OSA and controls,

with a significantly greater prevalence of LV systolic

dysfunction. Moderate/severe OSA was mostly found

among patients with impaired LV myocardial contracti-

lity and concentric geometry (Fig. 1).

Left ventricular geometryThe study groups had similar LV mass and prevalence of

hypertrophy. The AHI was directly related with relative

wall thickness. This significant linear correlation was

found both in the total population (r¼ 0.30, SEE 16.4,

P¼ 0.003) (Fig. 2) and in the subgroup of patients with

normal LV geometry (r¼ 0.38, SEE 12.5, P¼ 0.001), who

were significantly younger (56� 16 vs. 65� 9 years,

P< 0.0001), had smaller LV volumes and left atrial size,

and less frequently impaired LV myocardial contractility

(30 vs. 68%, P< 0.0001) than patients with abnormal LV

geometry. Figure 3 shows the prevalence of LV geo-

metric patterns in controls, in patients with mild and

moderate/severe OSA. Prevalence of concentric LV geo-

metry (concentric remodelingþ concentric LVH) was

similar in controls (four patients¼ 20%) and in patients

with mild OSA (six patients¼ 12%, P¼ns), whereas it

was significantly higher in patients with moderate/severe

OSA (50 patients¼ 58%, P< 0.0001). Approximately

two-thirds of these patients (80% with concentric remo-

deling and 74% with concentric LVH) had moderate/

severe OSA, which affected only 40% of patients with

nonconcentric LV geometry (38% with normal geometry

and 45% with eccentric LVH) (P¼ 0.007).

Determinants of concentric geometryConcentric LV geometry was positively associated in

univariate analysis with older age, female sex, obesity,

severity of OSA, number of antihypertension drugs and

reduced functional capacity. As expected, this condition

was closely related to smaller left ventricles, greater LV

mass and impaired diastolic and systolic function (data

not shown). Serum creatinine, protein C reactive,

parameters deriving from spirometry and 24-h ECG

and blood pressure monitoring were not related to con-

centric LV geometry. In particular, diurnal/nocturnal

systolic blood pressure excursion, diastolic blood pressure

excursion and heart rate excursion were 4.8� 5 mmHg,

6.5� 6 mmHg and 5.0� 5 b.p.m. vs. 4.9� 5 mmHg,

4.2� 4 mmHg and 4.3� 4 b.p.m. (all P¼ns), respect-

ively, in patients with and without concentric LV

geometry.

Left ventricular geometry in sleep apnea Cioffi et al. 3

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

4 Journal of Hypertension 2010, Vol 00 No 00

Table 1 General characteristics of 157 study patients divided according to the severity of obstructive sleep apnea in three groups [controls(AHI < 5), mild (AHI<— 15) and moderate/severe OSA (AHI >15)]

Variables Controls (20 patients) Mild OSA (51 patients) Mod/severe OSA (86 patients)

ClinicalAge (years) 56�15 60�13 63�12Male sex (%) 80 80 85Body weight (kg) 87�22 86�17MM 94�18Body mass index (kg/m2) 29�6 29�5MM 31�5Waist (cm) 95�12 103�13MM 110�13Obesity (%) 40 33MM 51SBP (mmHg) 134�14 133�16 136�17DBP (mmHg) 80�8 80�8 82�8Heart Rate (beats/min) 62�9 61�9 63�9Diabetes mellitus (%) 10 21 17COPD (%) 5 17 14Dyslipidemia (%) 60 55 59Chronic atrial fibrillation (%) 5 6 15History of paroxysmal atrial fibrillation (%) 25 25 20Hypertension (%) 50 71 78Duration of hypertension (months) 27�5 28�5 29�6Antihypertensive treatment (%) 65 73 80Number of antihypertension drugs 0.8�0.5M 1.31�1.1MM 1.59�1.1Diuretics (%) 15 25MM 36ACE inhibitors/ AT1-receptor blockers (%) 35M 65MM 70Calcium-antagonists (%) 10 20 22Beta-blockers (%) 20 39 23

PolysomnographyApneas (total number) 14 (10–21)M 39 (25–55)MM 170 (103–241)Hypopneas (total number) 10 (3–23)M 35 (23–53)MM 67 (46–91)AHI (apneasþ hypopneas/hours) 3.4 (1–10)M 10 (6–15)MM 33 (23–46)Obstructive/central apnea ratio per patient (%) 96% 86% 89%Duration of apneas/hypopneas (s) 24�4 23�4 25�5Oxigen saturation (%) 95�1.7 95�1.5 94�1.9Desaturations (number/hour) 0.35 (0–2)M 2 (0–4)MM 78 (36–109)Arousals (number/h) 19 (10–26)M 24 (18–34)MM 41 (35–56)

24-h blood pressure Holter monitoringMean SBP (mmHg) 107�16 121�14 124�20Mean DBP (mmHg) 72�6 76�6 77�5Mean nocturnal SBP (mmHg) 100�17 115�19 118�19Mean nocturnal DBP (mmHg) 68�8 73�6 75�7Mean diurnal–nocturnal SBP excursion (mmHg) 5.9�5 4.7�5 5.8�5Mean diurnal–nocturnal DBP excursion (mmHg) 4.5�5 4.1�5 5.5�5

24-h ECG Holter monitoringMean diurnal–nocturnal HR excursion (mmHg) 4.6�5 5.1�5 4.3�4Ectopic ventricular beats (in 24 h) 636 (130–1039) 1081 (320–4038) 1573 (488–5012)NSTV (% of patients) 0 12 16Ectopic single atrial beats (in 24 h) 19 (2–56) 183 (50–610) 118 (20–400)Paroxysmal atrial fibrillation (% of patients) 25 28MM 45

ACE, angiotensin-converting enzyme; AHI, apnea–hypopnea index; AT, angiotensin; COPD, chronic obstructive pulmonary disease; DBP, diastolic blood pressure; HR,heart rate; NSTV, nonsustained ventricular tachycardia; SBP, systolic blood pressure. MP<0.05 between controls and mild OSA. MMP<0.05 between mild and moderate/severe OSA.

Table 2 Echocardiographic characteristics of controls, patients with mild (apnea/hypoapnea index <15) and moderate/severe obstructivesleep apnea (apnea/hypoapnea index >15)

Variables Controls (20) Mild OSA (51) Mod/severe OSA (86)

LV systolic function and afterloadLV stroke volume (ml) 71�18 70�24 69�20LV stroke work (ml/mmHg) 138�41 133�49 135�44LV ejection fraction (%) 63�8 59�8 59�9LV midwall shortening (%) 16.3�2.6 17.1�3.6 15.5�3.0M

End-systolic stress (dynes/cm2) 139�45 130�34 127�36Stress-corrected midwall shortening (%) 88�13 92�18 83�15M

LV systolic dysfunction (at midwall level) (%) 38 39 62M

LV geometryLV end-diastolic diameter (cm/height) 2.60�0.2 2.64�0.3 2.47�0.3M

End-diastolic thickness of intraventricular septum (cm) 1.02�0.1 1.05�0.1 1.12�0.1M

End-diastolic thickness of posterior wall (cm) 0.99�0.1 1.00�0.1 1.08�0.1M

Relative wall thickness 0.39�0.07 0.39�0.06 0.43�0.06M

LV mass index (g/m2.7) 45�9 47�12 49�11LV hypertrophy (%) 7 (35%) 21 (41%) 43 (50%)Concentric geometry (%) 4 (20%) 6 (12%) 50 (58%)M

LV, left ventricular. MP<0.05 between mild and moderate/severe obstructive sleep apnea.

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

Left ventricular geometry in sleep apnea Cioffi et al. 5

Fig. 1

0.2 0.3 0.4 0.5 0.6 0.740

60

80

100

120

140

Str

ess-

corr

ecte

d m

idw

all s

hort

enin

g (

%)

Relative wall thickness

r = − 0.37P < 0.0001

Absent - Mild OSA

Moderate – severe OSA

Relation between stress-corrected midwall shortening (vertical axis) and relative wall thickness (horizontal axis). The reference lines indicate thesex-specific partition values for stress-corrected midwall shortening (87% in men, 90% in women, horizontal lines), and the upper limit of relative wallthickness (0.43, vertical line). Moderate/severe OSA was more prevalent in the lower right quadrant of patients with impaired LV myocardialcontractility and concentric geometry. The correlation line is shown.

Fig. 2

−10 0 10 20 30 40 50 60 70 80 900.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

0.60

0.65

0.70

Rel

ativ

e w

all t

hick

ness

Apnea/hypopnea index

Absent - Mild OSA

Moderate – severe OSA

r = 0.30P = 0.0001

Scatter plots show the linear correlations between apnea–hypopnea index (horizontal axis) and relative wall thickness (vertical axis) in the total studypopulation.

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

In logistic regression analysis, including age, sex, diabetes,

chronic obstructive pulmonary disease, body mass index,

systolic blood pressure, mean nocturnal systolic blood

pressure, hypertension, number of antihypertension drugs,

moderate/severe OSA, nocturnal heart rate, maximal work

capacity, LV end-diastolic diameter, impaired relaxation

pattern and stress-corrected midwall shortening (categor-

ized as impaired vs. normal), the variables independently

associated with concentric LV geometry were impaired

stress-corrected midwall shortening, moderate/severe

OSA and body mass index (Table 3).

DiscussionPrincipal findingsIn this investigation, we studied a group of middle-aged

patients with OSA and high prevalence of obesity, hy-

pertension and dyslipidemia but normal LV ejection

fraction to assess whether severity of OSA influences

LV structure leading to a concentric geometry. Our

results confirm this hypothesis, in line with the findings

of several other studies reporting high prevalence of

concentric hypertrophy or remodeling in population of

obese hypertensive patients [5,22]. We documented a

close relation between concentric LV geometry

and moderate/severe OSA, which was independent of

the presence of obesity and impaired LV myocardial

contractility, the other factors associated with concentric

LV geometry in our population. Among patients with

abnormal LV geometry, approximately three-fourths of

patients with moderate/severe OSA had a concentric

LV geometry. In contrast, among patients with mild

OSA and abnormal LV geometry, a concentric pattern

was detected in a minority of patients, being prevalent

the eccentric LVH, paralleled by a lower presence

of impaired LV myocardial contractility. Thus, the

documented effects of OSA on LV structure and function

[3–5,10] largely depend on the severity of sleep-

disordered breathing syndrome.

Furthermore, our analyses showed that the positive

association between LV relative wall thickness (index

of LV geometry) and severity of OSA was also present in

the subgroup of patients with normal LV geometry

(Fig. 2). Thus, in spite of their younger age and normal

LV mass, these patients had a propensity to remodel their

ventricles in a concentric fashion. This behavior, which

coexists in approximately half of cases with an impaired

6 Journal of Hypertension 2010, Vol 00 No 00

Fig. 3

Prevalence of the geometric left ventricular patterns in controls, patients with mild and moderate/severe OSA.

Table 3 Variables independently related to concentric leftventricular geometry: multivariate model (multiple logisticregression analysis)

Variables Odds ratio 95% CI P

Stress-corrected midwall shortening(impaired vs. normal)

3.38 1.46–7.8 0.004

Moderate-severe OSA(AHI > 15 vs. �15)

7.60 3.1–18.6 <0.0001

Body mass index 1.09 1.01–1.18 0.03

AHI, apnea–hypopnea index.

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

LV myocardial contractility, evidently heralds overt

cardiac structural abnormalities and represents a

phenomenon ‘ab initio’, which might identify, at an early

stage, the concentric LV geometry phenotype.

Potential mechanismsAn important limitation to most of previous studies which

evaluated LV geometry in the various clinical settings is

their focus on the patients as they present while awake.

This approach assumes that any determinants contribut-

ing in the patho-physiology or progression of LV geome-

try are inactive during the sleep [23]. Unfortunately, OSA

produces a series of mechanical, hemodynamic, chemical,

neural, and inflammatory responses (i.e. intermittent

apnea-induced hypoxia, hypercapnia, rises in sympath-

etic drive and LV myocardial afterload, daytime hyper-

tension, loss of vagal heart rate regulation and systemic

vasculature remodeling) [1,5,10,24,25] which strongly

impact the cardiovascular system, being all potent stimuli

to myocyte necrosis and apoptosis, myocardial ischemia,

and finally, development of concentric LV geometry.

These conditions may explain the extremely high preva-

lence of concentric LV geometry found in our patients

with moderate/severe OSA as well as in those recently

studied by Drager et al. [26] who showed a close associ-

ation between OSA, increased arterial stiffness and

increased LV mass and concentric geometry in a small

group of 15 patients with severe OSA and hypertension.

In contrast, Mylinski et al. [27] found a predominance of

eccentric LVH in 108 newly-diagnosed OSA patients.

Such apparent discrepancy can be explained by the

younger age and duration of OSA of these patients

and, more importantly, by the severity of OSA, which

was mild in a considerable portion of patients included in

the analysis.

The associations between sleep-disordered breathing

and LV geometry have been previously investigated

in a lot of studies conducted on patients with obesity

and hypertension and the results are irreconciliable

[2,5,7,22,28–30]. Both overweight and increased blood

pressure determine a state of chronic myocardial LV

overload and represent two of the main predisposing

factors for LVH and systolic dysfunction. In theory,

the two conditions are hemodynamically different and

should lead to different LV geometric adaptations

[5,25,28–31]. In clinical practice, such dichotomy is

less evident, because obesity and hypertension lead to

heterogenic and unpredictable changes in LV geometry

[5,22–24].

In our patients neither the presence of hypertension

(highly prevalent in all our study subgroups), systolic

or diastolic blood pressure, average 24-h blood pressure,

or nocturnal hypertension were different between

patients with mild and those with severe OSA despite

the ascertainment of hypertension was appropriate,

measurements of blood pressure were accurate and cru-

cial selection bias were reasonably absent. These findings

are not totally unexpected being in line with those

observed by several studies conducted both in patients

with [10,26] or without OSA [32] which clearly evidenced

that changes in LV geometry and LVH may precede the

onset of hypertension. This phenomenon may be related

to some consequences of severe OSA including signifi-

cant hypoxiemia, increased sympathetic drive and/or

transmural pressure due to the respiratory effort during

an apneic/hypoapneic event. Furthermore, it is important

to underline that our patients with moderate/severe OSA

patients were using a higher number of antihypertensive

drugs and receiving more frequently diuretics and vaso-

dilators than those with mild OSA or controls. These

differences in pharmacological treatment could explain

why the measures of hypertension were not related to the

magnitude of OSA in our population. However, type and

number of antihypertension drugs were not related to the

presence of concentric geometry in the multivaried

analysis, suggesting that other factors had a more relevant

effect on LV structure. The present study was limited by

lack of a sub-analysis on patients with OSA without

concomitant hypertension (not doable for the small

sample size) which could remove the possible confound-

ing effect of pharmacological treatment in evaluating the

role of blood pressure in explaining the study results.

Obesity was actually one of these factors. It was strongly

related to LV concentric geometry, independently of

severity of OSA in our population. In obese patients,

moderate/severe OSA may activate, more than in other

phenotypes, a variety of biological molecules including

lipids, proteins and DNA which can alter biological

functions such as myocardial energy metabolism [33],

endothelial dysfunction [34] and oxidative stress [35], all

stimuli leading to concentric LV geometry. Through

these mechanisms obesity may come undone from

OSA, representing both the main predisposing condition

for the sleep-disordered breathing and a primary deter-

minant of concentric LV geometry by itself.

Study limitation and clinical implicationsThe protocol of the present cross-sectional study did not

permit to generalize our results being in a selected

patient-based population with high prevalence of hyper-

tension, diabetes and dyslipidemia, and no conclusion

can be made on the causality of our findings. As

parameters of sympathetic activity we considered only

heart rate and its diurnal–nocturnal excursion. The

evaluation of plasma catecholamine levels, baroreflex

sensitivity or heart rate spectral analysis, not available

in our study, could be useful to better comprehend the

relations between LV geometry and OSA. However,

our hypothesis that OSA elicits concentric LVH could

be indirectly supported by the positive effects of

continuous positive airway pressure (CPAP) therapy on

Left ventricular geometry in sleep apnea Cioffi et al. 7

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

LV geometry. Several studies, indeed, demonstrated that

the application of nasal CPAP produced at medium-term

follow-up a significant regression of LV mass [10,36]

which was probably due to the decreased LV afterload,

sympathetic drive and normalization of oxygenation, all

potential underlying causative factors of LVH activated

by OSA. On the basis of Framingham data (8), the

shift from concentric to normal LV geometry and the

regression of LV mass represent two of the main thera-

peutic targets for these patients.

Also the relatively small sample size leads to consider our

data as hypothesis-generating results, not conclusive. In

this view, our study suggests that the assessment of OSA

could be useful to investigate the determinants of

changes in LV structure, performance and function in

patients with risk factors for cardiovascular disease. The

heterogenic expression of LV geometric abnormalities

characterizing patients with obesity and hypertension is

lost in those who have moderate/severe OSA. In most of

these patients, indeed, concentric LV geometry is the

pathological pathway elicited by the sleep-disordered

breathing. This increased prevalence may in part explain

the increased rate of cardiovascular events in these

patients.

AcknowledgementsConception and design: Giovanni Cioffi, Carlo Stefenelli,

Tiziano Edoardo Russo, Francesco Furlanello, Alessan-

dro Selmi, Eva Gerdts, Dana Cramariuc, Giovanni de

Simone. Generation of clinical data: Giovanni Cioffi,

Carlo Stefenelli, Tiziano Edoardo Russo, Francesco Fur-

lanello, Alessandro Selmi. Analysis and interpretation of

data, or both: Giovanni Cioffi, Eva Gerdts, Dana Cramar-

iuc, Giovanni de Simone. Drafting of the manuscript or

revising it critically for important intellectual content:

Giovanni Cioffi, Dana Cramariuc, Eva Gerdts, Carlo

Stefenelli, Giovanni de Simone. Final approval of the

manuscript submitted: Giovanni Cioffi, Eva Gerdts,

Dana Cramariuc, Carlo Stefenelli, Giovanni de Simone,

Alessandro Selmi, Tiziano Edoardo Russo.

References1 Shamsuzzaman ASM, Gersh BJ, Somers VK. Obstructive sleep apnea:

implications for cardiac and vascular disease. J Am Med Assoc 2003;290:1906–1914.

2 Phillipson EA. Sleep apnea: a major public health problem. N Engl J Med1993; 328:1271–1273.

3 Hedner J, Ejnell H, Caidahl K. Left ventricular hypertrophy independent ofhypertension in patients with obstructive sleep apnea. J Hypertension1990; 8:941–946.

4 Noda A, Okada T, Yasuma F, Nakashima N, Yokota M. Cardiac hypertrophyin obstructive sleep apnea syndrome. Chest 1995; 107:1538–1544.

5 Avelar E, Cloward TV, Walker JM, Farney RJ, Strong M, Pendleton RC, et al.Left ventricular hypertrophy in severe obesity: interaction among bloodpressure, nocturnal hypoxemia and body mass. Hypertension 2007;49:34–39.

6 Peppard PE, Young T, Palta M, Skadrut J. Prospective study of theassociation between sleep-disordered breathing and hypertension. N EnglJ Med 2000; 19:1378–1384.

7 Savage DD, Garrison RJ, Kannel WB, Levy D, Anderson SJ, Stokes J 3rd,et al. The spectrum of left ventricular hypertrophy in a general populationsample: the Framingham Study. Circulation 1987; 75:I26–33.

8 Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP.Prognostic implications of echocardiographically determined leftventricular mass in the Framingham Heart Study. N Engl J Med 1990;323:1706–1707.

9 Hiestand DM, Britz P, Goldman M, Phillips B. Prevalence of symptomsand risk of sleep apnea in the US population: results from the nationalsleep foundation sleep in America 2005 poll. Chest 2006; 130:780–786.

10 Cloward TV, Walker JM, Farney RJ, Anderson JL. Left ventricularhypertrophy is a common echocardiographic abnormality in severeobstructive sleep apnea and reverses with nasal continuous positive airwaypressure. Chest 2003; 124:594–601.

11 Clinical guidelines on the identification, evaluation, and treatment ofoverweight and obesity in adults: the evidence report. National Institutes ofHealth. Obes Res 1998; Suppl 2:51–209.

12 Vincenti A, Brambilla R, Fumagalli MG, Merola R, Pedretti S. Onsetmechanism of paroxysmal atrial fibrillation detected by ambulatory Holtermonitoring. Europace 2006; 8:204–210.

13 Sahn DJ, Demaria A, Kisslo J, Weyman A. The committee on M-Modestandardization on the American Society of Echocardiography:recommendations regarding quantitation in M-Mode echocardiography.Results of a survey study of echocardiographic measurements. Circulation1978; 58:1072–1083.

14 Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E,Sachs I, Reichek N. Echocardiographic assessment of left ventricularhypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450–458.

15 de Simone G, Devereux RB, Daniels SR, Koren MJ, Meyer RA, Laragh JH.Effect of growth on variability of left ventricular mass: assessment ofallometric signals in adults and children and their capacity to predictcardiovascular risk. J Am Coll Cardiol 1995; 25:1056–1062.

16 Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux RB. Association ofcarotid atherosclerosis and left ventricular hypertrophy. J Am Coll Cardiol1995; 25:83–90.

17 de Simone G, Devereux RB, Ganau A, Ahn RT, Saba PS, Mureddu G, et al.Estimation of left ventricular chamber and stroke volume by limited M-modeechocardiography and validation by two-dimensional and Dopplerechocardiography. Am J Cardiol 1996; 78:801–807.

18 de Simone G, Devereux RB, Daniels SR, Mureddu G, Roman MJ, KimballTR, et al. Stroke volume and cardiac output in normotensive children andadults. Assessment of relations with body size and impact of overweight.Circulation 1997; 95:1837–1843.

19 de Simone G, Devereux RB, Roman MJ, Ganau A, Saba PS, Alderman MH,Laragh JH. Assessment of left ventricular function by the midwall fractionalshortening/end-systolic stress relation in human hypertension. J Am CollCardiol 1994; 23:1444–1451.

20 de Simone G, Devereux RB, Koren MJ, Mensah GA, Casale PN,Laragh JH. Midwall left ventricular mechanics. An independentpredictor of cardiovascular risk in arterial hypertension. Circulation 1996;93:259–265.

21 Bella JN, Palmieri V, Roman MJ, Paranicas MF, Welty TK, Lee ET,et al. Gender differences in left ventricular systolic function inAmerican Indians (from the Strong Heart Study). Am J Cardiol 2006;98:834–837.

22 de Simone G, Devereux RB, Roman MJ, Alderman MH, Laragh JH. Relationof obesity and gender to left ventricular hypertrophy in normotensive andhypertensive adults. Hypertension 1994; 23:600–606.

23 Bradley TD, Floras JS. Sleep apnea and heart failure. Part I: obstructivesleep apnea. Circulation 2003; 107:1671–1678.

24 Somers VK, Dyken ME, Mark AL, Abboud FM. Sympathetic-nerve activityduring sleep in normal subjects. N Engl J Med 1993; 328:303–307.

25 Khatri IM, Freis ED. Hemodynamic changes during sleep. J Appl Physiol1967; 22:867–873.

26 Drager LF, Bortolotto LA, Figuereido AC, Silva BC, Krieger EM,Lorenzi-Filho G. Obstructive sleep apnea, hypertensione and theirinteraction on arterial stiffness and heart remodeling. Chest 2007;131:1379–1386.

27 Mylinski W, Duchna HW, Rasche K, Dichmann M, Mosiewicz J, Schultze-Werninqhaus G. Left ventricular geometry in patients with obstructive sleepapnea coexisting with treated systemic hypertension. Respiration 2007;74:176–183.

28 Hedner JA, Wilcox I, Sullivan CE. Speculations on the interaction betweenvascular disease and obstructive sleep apnea. In: Saunders NA, SullivanCE, editors. sleep and breathing, 2nd ed. Rev. New York: Marcel Dekker;1994. pp. 823–846.

29 Lauer MS, Anderson KM, Kannel WB, Levy D. The impact of obesity on leftventricular mass and geometry. The Framingham Heart Study. J Am MedAssoc 1991; 266:231–236.

8 Journal of Hypertension 2010, Vol 00 No 00

CE: Namrta; HJH/201793; Total nos of Pages: 9;

HJH 201793

30 Sukhija R, Aronow WS, Sandhu R, Kakar P, Maquire GP, Ahn C, LehrmanSG. Prevalence of left ventricular hypertrophy in persons with and withoutobstructive sleep apnea. Cardiol Rev 2006; 14:170–172.

31 Messerli FH, Sundgaard-Riise K, Reisin ED, Dreslinski GR, Ventura HO,Oigman W, et al. Dimorphic cardiac adaptation to obesity and arterialhypertension. Ann Intern Med 1983; 99:757–761.

32 Post WS, Larson MG, Levy D. Impact of left ventricular structure on theincidence of hypertension: the Framingham Heart Study. Circulation 1994;90:179–185.

33 Ambrosio G, Zweier JL, Flaherty JT. The relationship between oxygenradical generation and impairment of myocardial energy metabolismfollowing postischemic reperfusion. J Mol Cell Cardiol 1991; 23:1359–1374.

34 Schulz R, Seeger W, Grimminger F. Serum nitrite/nitrate levels inobstructive apnea. Am J Resp Crit Care Med 2001; 164:1997–1998.

35 Suzuki YJ, Jain V, Park AM, Day RM. Oxidative stress and oxidant signalingin ostructive sleep apnea and associated diseases. Free Radic Biol Med2006; 40:1683–1692.

36 Duchna HW, Myslinski W, Dichmann M, Rasche K, Schultze-WerninghausG, Orth M. Cardiac structure and function in patients with obstructive sleepapnea syndrome and co-prevalent arterial hypertension. Influence of CPAPtherapy. Med Klin 2006; 101:1–8.

Left ventricular geometry in sleep apnea Cioffi et al. 9

HJH Journal of Hypertension Typeset by Thomson Digital

for Lippincott Williams & Wilkins Manuscript No. 201793 Dear Author, During the preparation of your manuscript for typesetting, some queries have arisen. These are listed below. Please check your typeset proof carefully and mark any corrections in the margin as neatly as possible or compile them as a separate list. This form should then be returned with your marked proof/list of corrections to the Production Editor.

QUERIES: to be answered by AUTHOR/EDITOR QUERY NO. QUERY DETAILS RESPONSE

NO QUERY