Sleep and Breathing

, Volume 16, Issue 3, pp 717–722

Mal-effects of obstructive sleep apnea on the heart


  • Song-qing Yang
    • Department of ElectrodiagnosticsFirst Hospital of Jilin University
  • Li-li Han
    • Department of Respiratory MedicineFirst Hospital of Jilin University
  • Xiao-lu Dong
    • Department of Respiratory MedicineFirst Hospital of Jilin University
  • Chun-yong Wang
    • School of Public Health of Jilin University
  • Huan Xia
    • Department of Respiratory MedicineFirst Hospital of Jilin University
  • Pan Liu
    • Department of Respiratory MedicineFirst Hospital of Jilin University
  • Jing-hua Wang
    • Department of Respiratory MedicineFirst Hospital of Jilin University
  • Ping-ping He
    • Department of Respiratory MedicineFirst Hospital of Jilin University
  • Sheng-nan Liu
    • Department of Respiratory MedicineFirst Hospital of Jilin University
    • Department of Respiratory MedicineFirst Hospital of Jilin University
Original Article

DOI: 10.1007/s11325-011-0566-1

Cite this article as:
Yang, S., Han, L., Dong, X. et al. Sleep Breath (2012) 16: 717. doi:10.1007/s11325-011-0566-1



This study aims to examine the impact of chronic intermittent hypoxia on hearts in patients with obstructive sleep apnea (OSA).


Two hundred twenty patients were divided into groups based on (1) severity of the disease, (2) years of disease history, and (3) with or without secondary hypertension. All subjects underwent blood pressure measurements, polysomnogram monitoring, and cardiac Doppler ultrasound examinations.


The left ventricular ejection fraction (LVEF), fractional shortening (FS), and the ratio of early to late diastolic filling (E/A) in patients with severe OSA were lower than in those with moderate OSA and in healthy controls. The inner diameters of the main pulmonary artery (inD of MPA), the inner diameters of the right cardiac ventricle (inD of RV), and the thickness of anterior wall of the right ventricle (TAW of RV) were increased in patients with severe OSA compared to those with moderate disease and worsened as a function of time with disease. The tissue Doppler imaging-derived Tei index and pulmonary artery systolic pressure were also increased along with the severity of OSA. LVEF and FS in patients who had suffered from OSA for >10 years were decreased compared with those suffering from OSA for a shorter time. LVEF and FS in patients with secondary hypertension were decreased significantly relative to non-hypertensive OSA patients and healthy controls. E/A was decreased in OSA patients whether they had secondary hypertension or not.


OSA affected the left ventricular diastolic function in the early stage of the disease. Extended exposure to OSA resulted in left ventricular dysfunction with increased hypertension. Right ventricle dysfunction and abnormalities became more severe as the disease progressed.


OSAHeart diseaseCardiovascular diseaseHypoxemia


Obstructive sleep apnea (OSA) is characterized by repeated hypoxemia due to the collapse of the upper airway during sleep. This kind of chronic hypoxia is a potentially important and dangerous factor of poly-organ lesion including cardiovascular disease (CVD). OSA may increase the risk for CVD and induce symptoms of heart failure such as dyspnea, edema, and orthopnea. Overnight sleeping in OSA patients is associated with the development of heart subclinical systolic dysfunction and exaggerated diastolic dysfunction. Studies have shown that chronic hypoxia evokes changes in left ventricle (LV) and right ventricle (RV) functions [14]. Therefore, OSA is a strong predictor of fatal cardiovascular events in patients with CVD [5]. OSA is known to be dangerous to the heart; however, the concrete changes of the heart due to OSA are not fully understood. Here, we evaluated 220 patients with OSA and healthy controls to determine the effect of OSA severity, time with disease, and secondary hypertension on heart structure and function.

Materials and methods

Clinical records

A total of 220 patients with OSA, including 179 males and 41 females aged 27–83 years (mean 58.4 ± 0.7 years), were enrolled in this study. All were outpatients or inpatients in the First Hospital of Jilin University between 2006 and 2008. This study was approved by the Ethics Committee of First Hospital of Jilin University.

The diagnosis of OSA was based on the report of an American Academy of Sleep Medicine Task Force [6]. Subjects complained of symptoms such as snoring, awakening due to shortness of breath, and daytime sleepiness. They were subsequently found to have many physical symptoms such as pharyngeal hyperemia, pharyngeal cavity narrowing, tongue body hypertrophy, and tongue base postpositioning or micrognathia. Of the OSA patients, 61 patients were cigarette smokers and 37 reported a consistent use of alcoholic beverages (drinking alcoholic beverages containing more than 0.25 kg alcohol per day). The average BMI [weight (kg)/height square (m2)] of patients with OSA was 27.39 ± 5.74 kg/m2.

Based on the Apnea Hypopnea Index (AHI) and the pulse oxygen saturation nadir in sleep (SpO2 nadir), patients in the OSA group were categorized as mild (AHI 5–15, SpO2 nadir 85–89%, n = 77), moderate (AHI 15–30, SpO2 nadir 80–84%, n = 81), and severe (AHI > 30, SpO2 nadir < 80%, n = 62).

Based on the time the patients had suffered from snoring and apnea during sleep, as reported by spouses or people of the same house, they were divided into groups of <5 years (n = 65), 5–10 years (n = 79), and >10 years (n = 76).

There were 78 hypertensive and 142 non-hypertensive patients. Hypertension occurred about 4 to 5 years after the sleep apnea began. Hypertension was diagnosed using the WHO-recommended criteria, which is based on multiple measurements taken on different days. The patients with systolic pressure ≥ 18.6 kPa (140 mmHg) at rest state, and/or diastolic pressure ≥ 12 kPa (90 mmHg), were classified as hypertensive [7].

Patients taking sleeping pills, sedatives, and anti-hypertensive drugs were excluded. Patients were also excluded if they suffered from central sleep apnea syndrome, valvular heart disease, congenital heart disease, or acute left or right ventricular dysfunction or if they had a recent history (<1 month) of myocardial infarction or stroke or diabetes mellitus. None of the patients included had a history of asthma, chronic obstructive pulmonary disease, pulmonary interstitial fibrosis, or other debilitating symptoms such as cerebral apoplexy or dementia. None had lived at high altitudes for more than 3 months. Patients were excluded if they had used central nervous system excitatory drugs, such as methylphenidate (Ritalin), amphetamine, or coramine within a week of the study. Patients were also excluded if they had a history of chronic alcoholism.

The healthy control group consisted of a total of 75 health volunteers who were retirees of Jilin University. They were physically examined every year and no abnormalities were confirmed to be present by electrocardiogram or echocardiography. This group did not have diabetes, respiratory disorders, cardiovascular disease, or malignancies. They were also examined by polysomnography to exclude OSA. There were 61 males and 14 females (59.8 ± 1.1 years old). There were 20 smokers in the control group and 12 used alcohol. The average BMI of the controls was 26.32 ± 4.57 kg/m2. The differences between control and OSA groups in age, sexuality, smoking, alcohol use, and BMI were not significant (p>0.05; Table 1). Both the patients and control group signed the informed consent.
Table 1

Characterization of 220 OSA patients


Control (n=75)

OSA (n=220)




Male (%)






Smoking (%)



Alcohol user (%)



BMI = weight (kg)/height square (m2); p>0.5 for the comparison of age, male, BMI, smoking and alcohol user between control and OSA

Polysomnography and ultrasonography

Subjects were examined by a polysomnograph (PSG) (Embla s7000, Embla Systems Inc., USA). EEG (C3-A2, C4-A1), electrooculogram, electromyogram, and ECG were continuously recorded overnight. Breathing was monitored using a nasal cannula with a pressure transducer system, oronasal thermistors, and thoracoabdominal strain gauges. SpO2 nadir was monitored with a pulse oximeter.

All 295 subjects (OSA and control) were examined by the same technician using echocardiography (Vivid 7, GE Healthcare, US). Multiple parameters were measured including left ventricular ejection fraction (LVEF), fractional shortening (FS), the early and late diastolic filling peaks, the ratio of early to late diastolic filling (E/A) ratio, the inner diameters of the right cardiac ventricle (inD of RV), the inner diameters of the right cardiac ventricle (inD of MPA), and the thickness of anterior wall of right ventricle (TAW of RV). All images were video-recorded, analyzed offline, and measured for three consecutive cardiac cycles to derive the average values.

Additionally, the tissue Doppler imaging (TDI)-derived Tei index was measured. From the TDI recordings, the time interval from the end to the onset of mitral annular velocity pattern during diastole (a) was determined. The duration of the Sm (b) was measured from the onset to the end of the Sm. The Tei index was calculated as (a−b)/b as reported [8]. Additionally, the pulmonary artery systolic pressure (mmHg) was determined.

Statistical analysis

The measured values are given as mean ± standard deviation. Numerical data were analyzed in mono-agent of completely random design with SPSS13.0 (SPSS, Chicago, IL, USA). SNK-q was adopted in the comparison of interclass data. p < 0.05 indicates statistical significance and p <0.01 indicates being highly significant.


SpO2 and sleep structure

The SpO2 nadir of patients with OSA was significantly lower than that of control (p < 0.001). Total sleep time, sleep onset latency, and sleep efficiency did not differ significantly between the OSA group and control group. Compared with control, patients with OSA had an increased arousal index (P < 0.001), defined as the number of times awakened after sleep onset (P = 0.01), and reduced proportion of slow-wave sleep (P = 0.019) (Table 2).
Table 2

The PSG data of 220 patients


Control (n = 75)

OSA (n = 220)

SpO2 nadir (%)


81.1 ± 5.5**

Total time asleep (min)

289.2± 14.1

301.5 ±6.1

Sleep onset latency (min)

20.6 ±3.1

21.4± 2.3

Sleep efficiency (%)

71.3 ±2.5

72.1 ± 1.5

Sleep stage (%)

 Stage 1

6.5 ± 0.7

7.4 ±0.8

 Stage 2

60.8 ± 2.1

58.9± 0.9

 Slow wave

15.2 ± 1.8



11.0 ±1.1

14.5 ±0.7


2.9± 2.0

20.0 ± 5.6**


15.1 ±1.5

26.3 ±1.9**

Number of wake times after sleep onset

18.5 ± 2.3

27.2 ± 1.6*

Values are expressed as mean ± standard deviation

REM rapid eyes movement, AHI Apnea–Hypopnea Index, OSA obstructive sleep apnea

*P < 0.05 control vs. OSA; **P < 0.001 control vs. OSA.

Comparison of cardiac parameters based on OSA severity

Left heart parameters, such as the LVEFs, FS percentages, and E/A ratios from mild, moderate, and severe OSA groups and in the control group, are given in Table 3. LVEF, FS, and the E/A ratio in patients with the most severe OSA were lower than those in the other groups (p < 0.05). The E/A ratio decreased with a degree of OSA severity.
Table 3

Cardiac parameters of OSA patients with varying disease severities


Number of patients, N

LVEF (%)

FS (%)


inD of MPA (mmHg)

inD of RV (mmHg)

TAW of RV (mmHg)



63.11 ± 6.59

39.23 ± 4.56

1.29 ± 0.14

21.03 ± 2.45

18.84 ± 1.72

4.14 ± 1.06



63.23 ± 5.07

38.91 ± 5.45

1.15 ± 0.06*

21.46 ± 2.47

18.86 ± 1.65

4.71 ± 1.09



61.52 ± 9.15

37.82 ± 4.02

0.94 ± 0.18**

24.08 ± 1.98**

21.94 ± 1.99**

6.46 ± 2.06**



57.53 ± 8.15*

33.55 ± 5.42*

0.83 ± 0.09**

26.81 ± 1.97**

23.69 ± 2.85**

7.50 ± 1.68**

*P < 0.05; **P < 0.01 (for patients with a mild disease compared with controls, for patients with a moderate disease compared with those with a mild disease, and for patients with a severe disease compared with those with a moderate disease)

Right heart parameters such as inD of MPA, inD of RV, and the TAW of RV are also listed in Table 3. These parameters indicate that patients with moderate and severe OSA had compromised right ventricles.

The Tei indexes of different groups are listed in Table 4. Obviously, the Tei index markedly increased along with the severity of OSA. Notably, similar changes of pulmonary artery systolic pressure were observed (Table 5).
Table 4

Changes of Tei index in OSA patients


Number of patients, N

Tei index



0.49 ± 0.07



0.52 ± 0.10



0.57 ± 0.06*



0.66 ± 0.12**

*P < 0.05 (vs. mild); **P < 0.01 vs. (moderate)

Table 5

Changes of pulmonary artery systolic pressure in OSA patients


Number of patients, N

Pulmonary artery systolic pressure (mmHg)



22 ± 5



28 ± 6



38 ± 6*



55 ± 11**

*P < 0.05 (vs. mild); **P < 0.01 (vs. moderate)

Comparison of cardiac parameters in OSA patients based on time with disease

Patients who had suffered from OSA the longest had impaired left ventricle function. The LVEFs, FS percentages, and E/A ratios in patients were grouped based on time with disease in Table 6. Right ventricle changes were also correlated with time with disease. The inD of MPA, the inD of RV, and the TAW of RV worsened as time with disease increased (Table 6).
Table 6

Cardiac parameters of OSA patients based on time with disease


Number of patients, N

LVEF (%)

FS (%)


inD of MPA (mm)

inD of RV (mm)

TAW of RV (mm)



63.11 ± 6.59

39.23 ± 4.56

1.29 ± 0.14

21.03 ± 2.45

18.84 ± 1.72

4.14 ± 1.06

<5 years


62.86 ± 8.87

38.99 ± 5.85

1.19 ± 0.22*

22.74 ± 1.51

19.62 ± 1.31

4.69 ± 1.19

5–10 years


60.47 ± 7.59

37.19 ± 5.25

0.93 ± 0.22**

23.87 ± 2.34*

21.58 ± 1.79**

6.21 ± 1.21**

>10 years


55.44 ± 6.66*

32.76 ± 6.68*

0.85 ± 0.29**

25.55 ± 2.37*

23.56 ± 2.20**

7.59 ± 1.51**

*P < 0.05; **P < 0.01 (for OSA patients with disease for <5 years compared with control, for patients with disease for 5–10 years compared with those with disease for <5 years, and for patients with disease for >10 years compared with those with disease for 5–10 years)

Left ventricular function is more severely impaired in OSA patients with secondary hypertension

Compared with the control subjects and non-hypertensive OSA patients, both LEVF and FS in the OSA patients with secondary hypertension were reduced (p < 0.05 and p < 0.01, respectively; Table 7). The E/A ratios in both secondary hypertension and without hypertension groups differed from the control group (1.29 ± 0.14) (p < 0.01; Table 7).
Table 7

Cardiac parameters of OSA patients with or without hypertension


Number of patients, N

LVEF (%)

FS (%)




63.11 ± 6.59

39.23 ± 4.56

1.29 ± 0.14

Non-hypertensive OSA


62.45 ± 6.61

37.59 ± 5.99

1.18 ± 0.21**

Hypertensive OSA


54.56 ± 8.86*

33.56 ± 5.44**

0.85 ± 0.71**

*P < 0.5; **P < 0.01 (for comparison of left ventricular diastolic function between the control group and OSA patient)


OSA is associated with cardiovascular diseases such as hypertension and heart failure (HF). In a mouse model of OSA, time-dependent alterations that increase the susceptibility of the heart to oxidative stress have been observed. The prolonged exposure of mice to chronic hypoxia reversed the myocardial damage [9]. OSA excites the adrenergic nerves and may result in changes to the cardiovascular architecture. These alterations may occur before the clinical symptoms manifest. LV global dysfunction is associated with myocyte hypertrophy and apoptosis at the cellular level. OSA could conceivably worsen heart function through increased sympathetic activity, hemodynamic stress, hypoxemia, and oxidative stress [1012]. In this study, the impact of chronic hypoxia on cardiac structure and function was evaluated by comparing cardiac structure and function in healthy volunteers to patients with obstructive sleep apnea.

Doppler echocardiography at the level of the mitral or tricuspid valve reveals M-shaped double peaks that represent the ventricular diastolic phase. The first peak is the E peak, also called the crest time of the diastolic rapid filling phase. The second peak is the A peak, corresponding to atrial contraction, or the left ventricle late diastolic slow filling phase. Under normal conditions, the A peak is always lower than the E peak. The E/A ratio represented an important measure of left ventricular diastolic function. In addition, LVEF and FS are both strong indicators of the pumping function of the LV [13, 14]. In our study, decreases in these parameters usually reflect a reduction of myocardial contraction. Decreases in the E/A ratio, LVEF, and FS were correlated with severity and time with OSA, indicating that extended exposure to hypoxemia results in left ventricular dysfunction. It was notable that the E/A ratio was altered in OSA patients who had mild symptoms and suffered from OSA for less than 5 years, thus the left ventricular diastolic dysfunction very early in the disease course. The contractile function of LV in patients with mild and moderate OSA was decreased slightly relative to that of healthy controls. The contractile function of LV in patients with severe OSA was decreased significantly when compared with the control group. This result is consistent with data from a previous study [11].

A study by Zoccal et al. suggested that an enhanced respiratory–sympathetic coupling is the major factor contributing to hypertension in rats exposed to hypoxemia [15]. Hypoxemia induces increased chemoreflex and depressed baroreflex, resulting in sympathoadrenal hyperactivity, early hemodynamic alterations with proximal histological remodeling, and delayed changes in peripheral vasoreactivity [16]. Other data also support the idea that hypertension is induced by long-term hypoxia and preceded by alterations in the autonomic balance of heart rate variability and associated with an enhanced chemoreflex sensitivity [17]. In our study, more than 35% of OSA patients (78/220) exhibited secondary hypertension. Compared with the non-hypertensive OSA group, OSA patients with hypertension had decreases in E/A, LVEF, and FS, suggesting that hypertension may have complex effects on OSA-associated LV systolic and diastolic dysfunction. OSA patients without hypertension showed a reduction in the E/A ratio compared with controls. Secondary hypertension may exacerbate [18] the diastolic dysfunction caused by OSA and thereby the hypertension-related disorders such as cerebral stroke [19]. Our results agree with the findings by Sajkov who found that hypertension contributes to the adverse effects of hypoxemia via hypertension-induced LV remodeling and reduction in LV compliance.

Right heart catheterization is the most reliable method for the measurement of pulmonary artery blood pressure. However, it is not usually performed on patients due to inherent risk. Therefore, we used color Doppler ultrasound to measure the inD of MPA as an indirect measure of pulmonary blood pressure. Our results indicate that the inD of MPA, the inD of RV, and the TAW of RV were increased in OSA patients compared to healthy controls, consistent with previous observations by Gottlieb [12]. Three parameters correlated with disease duration and severity of OSA suggested that OSA patients have cardiac structural changes in the right ventricle and these abnormalities become more severe as the disease progresses and the time with disease increases.

The increases in the inner diameter of the pulmonary artery suggest pulmonary hypertension. It is possible that hypoxemia initially induces extensive contraction of pulmonary arterioles or may cause severe pulmonary vasospasms that subsequently lead to pulmonary hypertension. The long-term consequences of hypoxemia include pulmonary vascular structural changes and remodeling of the pulmonary artery and eventually right ventricle hypertrophy. OSA is associated with repetitive nocturnal hypoxemia and hypercapnia, large intrathoracic negative pressure swings, and acute advancing in pulmonary artery pressure. However, previous studies had not demonstrated that nocturnal hypoxemia associated with OSA was sufficient to cause significant changes in humans. A recent study showed that 20% to 40% of patients with OSA had pulmonary hypertension and observed reductions in pulmonary artery pressure in patients with OSA after nocturnal continuous positive airway pressure treatment [18]. Moreover, we found that the Tei index and pulmonary artery systolic pressure increased along with the severity of OSA, suggesting that OSA impaired cardiac function.

Patients are at risk for heart failure if there is coexisting left-sided heart disease or chronic hypoxic respiratory disease [19, 20]. In our study, the extent of heart dysfunction was correlated with disease duration, severity of OSA, and whether or not the patient suffered from complicating secondary hypertension. Early diagnosis and treatment of OSA may prevent the development of a heart disease, and patients who report daytime sleepiness, snoring, or repetitive sleep apnea should be referred to a sleep specialist for diagnosis and treatment.


This work is supported by National Natural Sciences Foundation (China) grant V30270502.

Conflicts of interest

The authors declare that they have no conflicts of interest.

Copyright information

© Springer-Verlag 2011