Abstract
QT interval dispersion (QTd) reflects inhomogeneity of repolarisation. Delayed cardiac repolarisation leading to the prolongation of the QT interval is a well-characterised precursor of arrhythmias. Obstructive sleep apnoea syndrome (OSAS) can cause cardiovascular complications, such as arrhythmias, myocardial infarction, and systemic and pulmonary hypertension. The aim of this study was to assess QTd in OSAS patients without hypertension.
A total of 49 subjects without hypertension, diabetes mellitus, any cardiac or pulmonary diseases, or any hormonal, hepatic, renal or electrolyte disorders were referred for evaluation of OSAS. An overnight polysomnography and a standard 12-lead ECG were performed in each subject. According to the apnoea–hypopnoea index (AHI), subjects were divided into control subjects (AHI <5, n = 20) and moderate–severe OSAS patients (AHI ≥15, n = 29). QTd (defined as the difference between the maximum and minimum QT interval) and QT-corrected interval dispersion (QTcd) were calculated using Bazzet's formula.
In conclusion, the QTcd was significantly higher in OSAS patients (56.1±9.3 ms) than in controls (36.3±4.5 ms). A strong positive correlation was shown between QTcd and AHI. In addition, a significantly positive correlation was shown between QTcd and the desaturation index (DI). The AHI and DI were significantly related to QTcd as an independent variable using stepwise regression analysis.
The QT-corrected interval dispersion is increased in obstructive sleep apnoea syndrome patients without hypertension, and it may reflect obstructive sleep apnoea syndrome severity.
The variability in the QT interval duration between the different leads of a surface 12-lead ECG reflects local differences in recovery time of the myocardium 1, 2. In addition, QT interval dispersion (QTd) is increased in patients with a priori myocardial infarction, who have a susceptibility to ventricular tachyarrhythmias, most obviously via a re-entry mechanism 3. It has been suggested that increased QTd may be a marker of arrhythmic risk in patients with impaired left ventricular function 4.
Obstructive sleep apnoea syndrome (OSAS) is characterised by repetitive collapse of the upper airway during sleep 5. The obstructive apnoeic event is associated with considerable breathing efforts against the totally or partially occluded upper airway. The apnoea is terminated by an arousal and heavy snoring as airflow is restored. Severity of OSAS is described according to the total number of apnoeas and hypopnoeas per hour of sleep, which is termed the apnoea–hypopnoea index (AHI).
Cardiovascular disturbances are the most serious complications of OSAS. These complications include nocturnal arrhythmias 6, acute myocardial infarction 7, heart failure 8, stroke 9, and systemic 10, 11 and pulmonary hypertension 12. All of these cardiovascular complications increase the morbidity and mortality of OSAS. Sleep apnoea is currently accepted as one of the identifiable causes of hypertension, as described in the JNC 7 report 13.
The aim of the present study was to assess QTd in OSAS patients without hypertension.
METHODS
Study population
Ninety-six subjects were admitted to a sleep clinic with nocturnal snoring and/or excessive daytime symptoms, and a detailed sleep and cardiovascular anamnesis of the patients was recorded. Sleep cycle, nutritional status, medications, alcohol usage and family anamnesis were also established. All patients were asked questions from the Epworth sleepiness scale (ESS) 14, and patients with high scores (ESS ≥10) were accepted into the sleep study.
A physical examination was performed on all subjects. Systolic (BPs) and diastolic (BPd) blood pressures were measured in the sitting position from the right arm using a sphygmomanometer (Erka, Bad Tölz, Germany), after ≥5 min of rest. Heart rate per minute (HR) was measured in the sitting position, and the body mass index (BMI) of each patient was calculated as weight divided by height squared (kg·m−2). After an overnight fast (≥12 h), routine biochemical parameters (e.g. blood glucose, lipids, urea, creatinine, electrolytes, hepatic enzymes) were analysed using commercial kits (Abbott Laboratories, Abbott Park, IL, USA) by an autoanalyser (Aeroset; Abbott Laboratories).
Pulmonary function tests (SensorMedics 2400; SensorMedics, Bilthoven, The Netherlands) and arterial blood gas analysis (ABL 30; Radiometer, Copenhagen, Denmark) were performed in all patients at rest. In the morning, a 12-lead surface ECG was taken from every subject: all were in sinus rhythm, and no subjects had intraventricular conduction delay or a prolonged QRS complex, which would thereby increase the QT interval. In addition, all of subjects underwent treadmill exercise testing, which was normal for each subject. None of the subjects was taking medications that could potentially prolong the QT interval.
To be eligible, none of the 96 subjects was permitted to suffer from: 1) any known cardiac or pulmonary disease; 2) hypertension (blood pressure >140/90 mmHg or taking antihypertensive therapy); 3) diabetes mellitus or impaired glucose tolerance (fasting blood glucose >110 mg·dL−1 or taking antidiabetic therapy); 4) angina pectoris; 5) bundle branch block, atrial fibrillation, any arrhythmias or fewer than six measurable leads in the resting ECG; 6) chronic renal or hepatic diseases (by both self-report and serum analysis); or 7) serum electrolytes imbalances.
Polysomnography
Polysomnography 15 was applied to all subjects during the diagnostic night. The portable, limited sleep study, performed with an Embletta device 16, consisted of the following: 1) nasal pressure detection using a nasal cannulae/pressure-transducer system, which recorded the square root of pressure as an index of flow; 2) thoraco-abdominal movement detection, via two piezoelectric belts; 3) finger pulse oximetry; and 4) body position detection.
An apnoea was defined as total obstruction of oronasal airflow for ≥10 s, a hypopnoea was defined as a decrease of airflow of ≥50% and desaturations were accepted as a ≥4% decrease in oxygen saturation 17. The desaturation index (DI) was identified as the number of oxygen desaturation events per hour of sleep. Subjects with an AHI ≥5 were diagnosed as OSAS 18. According to the AHI, subjects were divided into two groups: 1) control subjects (AHI <5, n = 20); and 2) patients with moderate–severe OSAS (AHI ≥15, n = 29). There were no mild OSAS (AHI 5–15) patients in the study population.
Measurement of QT interval and dispersion
In the morning, after ≥10-min rest in the supine position, a 12-lead ECG was recorded at a paper speed of 50 mm·s−1 on a six-channel recorder. Two consecutive cycles were measured from each of the standard 12 leads and, subsequently, a mean QT was calculated from the two values. The QT intervals were manually measured from the onset of the QRS to the end of the T-wave, defined as the return to TP isoelectric baseline, by a tangential method 3. Only monophasic well-defined T-waves were accepted for measurement. When U-waves were present, the QT was measured to the nadir of the curve between the T and U waves, with the aid of a tangent. If the end of the T-wave could not be reliably determined, or if T-waves were isoelectric or of very low amplitude, the lead was not included in the analysis. A minimum of six leads, at leads three precordial, was required for inclusion in the study. The measurements were performed manually by an experienced observer blinded to the clinical data of the patients. The QTd was defined as the difference between the maximum and minimum QT values, and the mean value of two consecutive cycles was calculated. Bazett's formula was used to obtain HR-corrected (c) values of the QT intervals and dispersions 19, as follows:
To estimate the intra-observer variability, two photocopies of a random sample of 10 ECGs were taken and the QT interval was measured again. The relative difference (mean absolute difference in percentage of mean measured value) was 1.5% for the QT interval and 4.2% for the QTd.
Statistical analysis
Results are presented as mean±sd and the Mann-Whitney U-test was used to compare the two groups. Stepwise regression analysis was used to identify possible determinants of QT-corrected interval dispersion (QTcd). The correlations among QTcd and the variables of OSAS severity were investigated by Pearson correlation analysis. A p-value <0.05 was considered statistically significant.
RESULTS
A total of 27 males and 22 females were included the study. None of them was using alcohol and 45% of them smoked cigarettes. The mean ESS score of the study population was 16.8±6.1 (range 10–24). Basic characteristics and QTcd of the moderate–severe OSAS patients and controls are shown in table 1⇓. There were no significant differences between moderate–severe OSAS patients and controls according to sex, age, BMI, BPs, BPd and HR (p>0.05). AHI and DI were significantly higher in moderate–severe OSAS patients than in controls (p<0.0001), although, in contrast, these patients had the lowest average and minimum nocturnal saturation of arterial oxygen (Sa,O2; p<0.0001). The percentage of sleep duration where Sa,O2 was <90% was highest in the OSAS patients, whereas it was lowest in controls.
Basic characteristics and QT-corrected interval dispersion (QTcd) in moderate–severe obstructive sleep apnoea syndrome (OSAS) patients and controls
The QTcd was significantly higher in moderate–severe OSAS patients (56.1±9.3 ms) than in controls (36.3±4.5 ms; p<0.001). A strong positive correlation was shown between QTcd and AHI, reflecting the severity of OSAS (p<0.001, r = 0.954). Correlation between QTcd and AHI in OSAS patients without hypertension is shown in figure 1⇓. In addition, a significant positive correlation was shown between QTcd and DI (r = 0.485, p<0.01). However, statistically negative correlations were found between QTcd and minimum nocturnal Sa,O2 (r = −0.494, p<0.05) and average nocturnal Sa,O2 (r = −0.452, p<0.05).
Correlation between QT-corrected interval dispersion (QTcd) and apnoea–hypopnoea index (AHI) in obstructive sleep apnoea patients without hypertension. R = 0.954; p<0.01.
In OSAS patients, multiple regression analysis was performed to evaluate the relationships described previously when the effects of confounding factors such as age, BMI, HR, BPs and BPd were taken into account, although all of these factors were not significantly different between moderate–severe OSAS patients and controls. The AHI (R2 = 0.25, p = 0.0001), DI (R2 = 0.22, p = 0.0001), minimum and average nocturnal Sa,O2 (%; R2 = −0.20, p = 0.002 and R2 = −0.18, p = 0.002, respectively) were significantly related to the QTcd as independent variables using stepwise regression analysis.
DISCUSSION
QTd has been shown to be a useful noninvasive method for the detection of inhomogeneity of ventricular recovery times 1, 2. In addition, experimental studies provide powerful evidence for the significance of the dispersion of myocardial recovery times in respect to the occurrence of ventricular arrhythmias 20, 21. However, there are some problems when trying to achieve an accurate measurement of the QT interval 22: low T-wave amplitude and the presence of the U-wave create difficulties in identifying the end of the T-wave. Careful selection of the leads from which the T-wave can reliably be identified is required.
It has been proposed that the inter-lead differences in ventricular recovery time could be due to technical artefacts, such as differential tissue attenuation or differences in unipolar versus bipolar leads. Day et al. 23 have suggested that the QTd could be affected by the number of measurable leads. In the current study, it was found that the number of evaluable leads was, on average, 10.
Recovery time dispersion has been studied using both invasive and time-consuming methods, such as monophasic action potentials 24 or body surface mapping 25, respectively. QTcd is preferable to QTd when simultaneous 12-lead recording is not available, and inter-lead variations in cycle length could introduce errors in calculation.
The current authors have previously used Bazett's formula 19 for calculation of QTc. To evaluate and compare QTc formulae in 21 healthy subjects, Molnar et al. 26 used 24-h Holter monitoring, as it allows the assessment of QT intervals over a large range of rates. QT–RR relationships for individuals and the group were fitted by regression analysis to five QT prediction formulas: simple Bazett's, modified Bazett's, linear (Framingham), modified Fridericia's and exponential (Sarma's). It was shown that there were no significant differences in mean squared residuals between formulae. When using individually calculated regression parameters, each formula gave good or acceptable QTc over the entire range of RR intervals. The simple Bazett's formula, which uses no regression parameters, was unreliable at high rates. When group-based regression parameters were applied to individuals, no formula had a clear advantage over simple Bazett's. It was concluded that any formula that invokes regression parameters unique to each individual provides satisfactory QTc. When individual parameters cannot be determined, as in a 12-lead ECG, no formula provides an advantage over the familiar simple Bazett's.
The present study showed that QTcd is increased in patients with moderate–severe OSAS when compared with controls. In previous studies, QTd has been shown to be increased in patients after acute myocardial infarction 3, and in patients with chronic heart failure 4, long QT syndrome 27 and hypertrophic cardiomyopathy 28.
OSAS is closely associated with obesity 29 and ageing 30. A strong relationship between systemic hypertension and OSAS has been indicated in some epidemiological studies 10, 11, 31–33 and, presently, sleep apnoea is accepted as one of the identifiable causes of hypertension, as shown in the JNC 7 report 13. It is well known that the risk of developing systemic hypertension increases depending on the severity of OSAS 34. In addition, strong relationships have been established between the severity of systemic hypertension and AHI, DI and minimum nocturnal Sa,O2 in several studies 33, 35. In addition, it has been suggested that blood pressure decreases when the treatment of OSAS is optimal 36. In the current study, the proportion of sleep time where Sa,O2 was <90% was highest in the OSAS patients, whereas it was the lowest in the controls. Therefore, the moderate–severe OSAS patients had longer periods of hypoxia during sleep than control subjects. In the present study, the presence of systemic hypertension, diabetes mellitus and coronary artery disease in patients was excluded, and there were no significant differences in sex, age, BMI and arterial blood pressure between subjects. Thus, the QTcd was assessed in uncomplicated (isolated) patients with OSAS. In addition, positive correlations were found between QTcd, and AHI and DI, reflecting the severity of OSAS (r = 0.954, p<0.001 and r = 0.485, p<0.01, respectively). AHI (R2 = 0.25, p = 0.0001) and DI (R2 = 0.22, p = 0.0001) were also significantly related to the QTcd, as independent variables, using stepwise regression analysis. However, there were statistically negative correlations between QTcd, and minimum and average nocturnal Sa,O2 (r = −0.494, p<0.05 and r = −0.452, p<0.05, respectively) as expected. Therefore, it might be suggested that increased QTcd in OSAS patients is related to the severity of OSAS and, thus, to hypoxaemia.
Although arrhythmic susceptibility was not assessed in the patients with OSAS, the present observations suggest that degree of AHI is significantly associated with increased inhomogeneity in repolarisation (r = 0.954, p<0.001), probably predisposing the patients to life-threatening arrhythmias. Several studies have investigated the prevalence of nocturnal arrhythmias in patients with OSAS 6, 37, 38. Two were prospective studies, which followed-up consecutive referrals and included a control group: the prevalence of arrhythmias in these prospective studies was similar to that observed in healthy adults 37, 38. The study with the most valid measurement and classification of arrhythmias found no difference between the groups 37.
In this study, the reasons for increased QTcd in patients with OSAS compared with controls have not been clarified. Altered autonomic cardiac control is known to predispose individuals to ventricular arrhythmias under several experimental and clinical conditions 39, 40; increased sympathetic and/or reduced vagal tone may facilitate arrhythmogenesis by a re-entrant mechanism, triggered activity and increased automaticity. In the future, this may be confirmed if HR variability in patients with OSAS is investigated in longitudinal studies.
Conclusion
The present study demonstrates that increased AHI and DI in patients with OSAS may result in inhomogeneity of repolarisation, favouring a propensity towards ventricular tachyarrhythmias. A significant positive correlation was found between repolarisation inhomogeneity (QTcd) and severity of OSAS (AHI). In the future it will be important to establish whether repolarisation inhomogeneity may be corrected by continuous positive airway pressure treatment aimed at reducing upper airway obstruction in patients with OSAS.
Study limitation
This study is affected by fundamental limitations in the use of QT interval and QT dispersion measurements. For comparison with other studies, the QT interval was corrected for heart rate. It is well known that Bazett's formula is inaccurate at higher heart rates 26. It overcorrects at faster heart rates and undercorrects at lower heart rates. In addition, there were only a small number of patients in the present study population, which is an important study limitation. Furthermore, the sleep clinic population used here may not reflect the findings in the general community, and the results should be further confirmed with several longitudinal studies.
- Received June 7, 2004.
- Accepted December 23, 2004.
- © ERS Journals Ltd
References
