| Objective: Our aims were to determine in patients with HF: 1) the prevalences of obstructive sleep apnea (OSA) and central sleep apnea (CSA) and predictors of them, and whether these factors were independent to the morbidity of OSA and CSA; 2) the factors related to the severity of sleep apnea and whether daytime sleepiness was still available for assessment of sleep apnea; 3) the relationship between the transition in types of sleep apnea (from OSA to CSA or from CSA to OSA) and cardiac function.Methods : Inclusion criteria for heart failure patients were: 1) men and women aged≥18 years; 2) HF due to ischemic or non-ischemic dilated cardiomyopathy for≥6 months; 3) left ventricular (LV) systolic dysfunction (LVEF≤45% by radionuclide angiography or echocardiography performed within 1 month prior to polysomnography) and New York Heart Association (NYHA) class II-IV after optimization of medical therapy; and 4) stable clinical status on stable optimal medical therapy for≥1 month before entry.Polysomnography: Overnight attended polysomnography was performed using standard techniques and criteria for recording and scoring of sleep stages and arousals from sleep. Variables included EEG, ECG, EMG, EOG, chest and abdominal movements, mouth and nasal air flow, SaO2 and transcutaneous PCO2(PtcCO2). Patients were divided into categories according to the frequency of apnea and hypopnea per hour of sleep (i.e., apnea-hypopnea index [AHI]): No Sleep Apnea (NSA: AHI <5 per hour of sleep), CSA (AHI≥5 per hour of sleep with >50% central events) and OSA (AHI≥5 per hour of sleep with >50% obstructive events).Questionnaire before Polysomnography : Sleep Diary, Epworth Sleepiness Scale and Questionnaire before Sleep.Results: 1 Prevalence and Physiological Predictors of Sleep-Related-Disordered Breathing in HF1.1 Prevalence of sleep apneaUsing AHI cutoffs of≥5,≥10, and≥15 per hour of sleep, the overall prevalences of sleep apnea were 82%, 58%, and 42%, respectively; of OSA were 54%, 36%, and 24%, respectively; and of CSA were 28%, 22%, and 18%, respectively.1.2 Factors associated with OSA and CSAMultivariable analysis revealed that OSA was independently associated with older age, male sex, and greater BMI, whereas CSA was independently associated with older age, male sex, atrial fibrillation, lower awake PCO2, and diuretic use. However, beta-blocker use, LVEF, NYHA class, and HF etiology were not independently associated with OSA or CSA.1.3 Independent predictors of OSA and CSAThe best cutoff value of BMI, the most important continuous variable associated with OSA was >29.1 kg/m2 (area under the curve 0.68±0.03, P<0.001, sensitivity 64.5%, specificity 68.7%), whereas the best cutoff value of PCO2, the most important continuous variable associated with CSA, was <38.7 mm Hg (area under the curve 0.75±0.04, P<0.001, sensitivity 60.1%, specificity 81.2%).1.4 Sex-specific clinical characteristics of patients with heart failureThe prevalences of sleep apnea, OSA, and CSA were approximately 2.5 fold higher in men than in women (P<0.001, P=0.012 and P=0.034, respectively), despite no differences in age, BMI, LVEF, and NYHA class between them (Table 4). The only differences between men and women were the greater proportion of men with ischemic etiology of HF, and a higher proportion of slow-wave sleep in women. 2 Subjective Sleepiness in HF Patients with SRDB2.1 Subjective sleepinessPatients with OSA or CSA presented with no significantly higher ESS scores compared with NSA (P=0.105 and P=0.235 respectively, Table 1), revealing OSA and CSA patients have no significantly subjective sleepiness. The ESS score had no significant correlation with AHI (r=?0.018; P=0.795), LVEF (r=?0.003; P=0.965), BMI (r=?0.096; P=0.170) and sleep onset latency (r=?0.104; P=0.138), also had no association with total sleep time, (r=?0.028; P=0.687) sleep efficiency (r=0.001; P=0.986) and sleep stages (Stage 1: r=?0.029, P=0.680; Stage 2: r=?0.003, P=0.971; SWS: r=0.114, P=0.104 and Rapid Eyes Movement stage: r=0.008, P=0.913, respectively).2.2 Sleep structureCompared with NSA, patients with OSA or CSA had increased arousal index (ArI) ( P<0.001 respectively), more awake number after sleep onset ( P=0.01 and P<0.001 respectively), furthermore, reduced proportion of slow-wave sleep (P=0.024 respectively, Figure 1), and patients with CSA had a greater percentage of stage 1 sleep than those in the other two groups (P=0.002 and P=0.024 respectively) and there was also a significant difference in the awake number between the CSA and OSA group (P=0.016).2.3 PCO2 and BMIBivariate correlation analysis revealed that in patients with CSA, there was significant correlation between the severity of CSA, as assessed by the AHI, and the degree of PCO2 during sleeping or wakefulness (r=-0.35, P=0.006 ; r=-0.337, P=0.008 respectively), on the contrary, in OSA patients, the AHI was significant associated with the degree of the BMI (r=0.208, P=0.028),but not with the degree of PCO2 during sleeping or wakefulness (r=-0.037, P=0.704; r=-0.088, P=0.374 respectively). 3 Alteration in Sleep-Related-Disordered Breathing type and its related factors in HF3.1 Polysomnography DataAs the result of the analysis, no significant differences were found in the total time asleep, sleep structure, mean and mean lowest SaO2 between the two nights in any of the three groups. In the group with unstable apnea type, there was a significant change of the type of sleep apnea events during two polysomnography tests from OSA to CSA or CSA to OSA (p<0.001). However, during the two tests, there was no significant difference between the proportion of the supine position( OSA to CSA:50.8±9.6% Vs 48.2±12.7%, P=0.769;CSA to OSA:61.2±11.6% Vs 53.8±10.5%, P=0.542).3.2 PCO2 and periodic breathing cycle lengthIn the 12 subjects with unstable apnea type changed from OSA to CSA, mean wake and sleep PtcCO2 was significantly lower (41.0±1.3 Vs 34.9±1.0,P=0.001;42.1±1.2 Vs 36.3±1.1, P=0.006), PBCL and LFCT were significant longer (51.9±2.1 Vs 62.3±1.9,P<0.001;54.4±1.8 Vs 65.3±1.6,P<0.001), LVEF decreased (32.1±2.5 Vs 19.9±3.5 P=0.018), LVEF correlated inversely with the change in PBCL and LFCT (r=0.687, P=0.014;r=-0.591, P=0.043); In the 6 subjects with unstable apnea type from CSA to OSA, mean wake and sleep PtcCO2 significantly increased (37.7±0.6 Vs 41.8±0.9,P=0.007;39.2±0.5 Vs 42.7±1.0, P=0.017), PBCL and LFCT were significant shorter (61.5±3.4 Vs 49.7±2.8,P=0.025;66.1±2.1 Vs 52.1±1.6,P=0.003), PtcCO2 correlated inversely with the change in PBCL(r=-0.586, p=0.045), not with LFCT(r=-0.381, p=0.222).In the stable OSA and stable CSA groups, there were no significant differences in mean sleep and week PtcCO2, between the baseline and follow-up sleep studies. In the stable CSA group, AHI correlated inversely with the change in LVEF (r=-0.474,P=0.047), not with the other two groups(r=0.206, P=0.413; r=0.548, P=0.260 respectively). Conclusions:1 In HF patients on optimal contemporary therapy, we found a high prevalence of both OSA and CSA. The prevalences of OSA and CSA did not change despite increasing use of beta-blockers and spironolactone. Male, greater age and higher BMI were independent predictors for OSA. Low awake PCO2, atrial fibrillation, male, greater age, and diuretic use were independent predictors for CSA. BMI >29.1kg/m2 and PCO2<38.7 mm Hg may be the predictors for OSA and CSA. The sensitivity and specificity were high. Therefore, BMI>29.6 kg/m2 and PCO2<38.7 mm Hg might be the shortcut way for the clinic to dignosis OSA and CSA. It has been proved that OSA and CSA can be contribute to the prognosis of HF, and both are the independent risk factors to the mortality of HF. Our study suggests that a large proportion of HF patients, even those receiving optimal contemporary therapy, might benefit from effective therapy of these sleep-related breathing disorders. .2 In HF patients receiving optimal contemporary HF treatment, prevalence of SRBD was high, and there was not a significant relationship between the severity of SRBD and the ESS, and the SRBD patients had poorer sleep quality compared with those without SRBD. These observations suggest that a large proportion of HF patients, even those receiving optimal contemporary therapy, might not benefit from effective therapy of these SRBD. Because the absence of subjective sleepiness is not a reliable indicator of who does not have SRBD, therefore, deletion sleepiness may not be the primary target of treatment for SRBD in patients with HF. According to our study, we suggest that BMI and the wake PCO2 may be the clinic sign to diagnose OSA and CSA. .3 The type of sleep apnea was not stable in HF patients with periodic OSA or CSA. It would be changed from OSA to CSA when the wake and sleep PCO2 was decreased, while PBCL and LFCT was increased; on the contrary, when the wake and sleep PCO2 increased, while PBCL and LFCT decreased, it would be shifted from CSA to OSA. This alteration might be related to the chang of heart function... |
PDF Full Text Request