Predictive Factors, Prognostic Analysis And Impact On Renal Function Of Sleep Apnea In Heart Failure

Part 1:Sleep apnea in heart failure:A reviewSleep apnea(SA)is highly prevalent in heart failure(HF),typically classified as obstructive SA(OSA)and central SA(CSA).In OSA,there is pharyngeal collapse with consequent upper airway obstruction.In CSA,the underlying abnormality is in the brainstem respiratory centers.SA is associated with a series of pathophysiological consequences,including negative intrathoracic pressure,intermittent hypoxemia,activated sympathetic nervous system and enhanced inflammation,resulting in cardiac remodeling and dysfunction.Evidence shows that both OSA and CSA are independent risk factors of adverse outcomes in HF.For OSA,continuous positive airway pressure(CPAP)ventilation can improve quality of sleep and cardiac function,however,CPAP showed no survival benefit in HF with reduced ejection fraction(HFrEF)in some randomized controlled trials.For CSA,adaptive servoventilation(ASV),a minute ventilation targeted positive airway therapy,was associated with increased risk of adverse cardiovascular events in HFrEF.Thus,further evidence is needed to determine the effect of positive airway pressure ventilation in patients with HF and SA.In addition,some novel therapies might provide new insight into the treatment of HF with SA.Part 2 Prevalence and predictors of sleep apnea in patients hospitalized with heart failureObjective:Data regarding clinical predictors of sleep apnea(SA)in patients hospitalized with heart failure(HF)is limited.We aimed to determine the independent clinical predictors of SA in patients hospitalized with HF based on a large cohort and establish a predicting model.Methods:Sleep studies were performed on patients hospitalized with HF from January 2015 to Feburary 2019 using a multichannel cardiopulmonary polygraphy.Patients with a minimum valid recording time of 4 hours were included in analysis.SA was defined as the apnea-hypopnea index(AHI)? 15 events/h and severe SA was defined as AHI? 30 evnets/h.Stepwise multivariate logistic regression analysis was used to determine the independent predictors of SA and severe SA,respectively.Results:A total of 959 patients were enrolled in the study.The prevalences of SA and severe SA were 49.3%and 23.9%,respectively.Multivariate logistic analysis showed that the independent predictors of SA were male sex(OR = 2.479,95%CI:1.741?3.532,P<0.001),older age(per 10 years increase:OR=1.298,95%CI:1.162?1.449,P<0.001),higher body mass index(BMI)(per 5 kg/m2 increase:OR=1.674,95%CI:1.397?2.006,P<0.001),hypertension(OR = 1.557,95%CI:1.149?2.109,P=0.004),higher diastolic blood pressure(DBP)(per 5 mmHg increase:OR=1.063,95%CI:1.003?1.126,P=0.038),increased N-terminal-pro brain natriuretic peptide(NT-proBNP)(lnNT-proBNP:OR=1.159,95%CI:1.035?1.298,P=0.011)and decreased left ventricular ejection fraction(LVEF)(per 5%decrease:OR=1.132,95%CI:1.070?1.198,P<0.001).The predicting model of SA had a C-statistic of 0.723(95%CI:0.691?0.755)and a chi-square P of 0.753 in Hosmer-Lemeshow good-of-fit test,respectively.The independent predictorts of severe SA were male sex(OR=2.547,95%CI:1.613?4.023,P<0.001),higher BMI(per 5 kg/m2 increase:OR=1.547,95%CI:1.287?1.859,P<0.001),history of hypertension(OR=1.726,95%CI:1.228?2.427,P=0.002),higher DBP(per 5 mmHg increase:OR = 1.084,95%CI:1.017?1.155,P=0.013)and increased NT-proBNP(lnNT-proBNP:OR=1.339,95%CI:1.182?1.517,P<0.001).The predicting model of severe SA had a C-statistic of 0.717(95%CI:0.680?0.753)and a chi-square P of 0.734 in Hosmer-Lemeshow good-of-fit test,respectively.Conclusions:SA was highly prevalent in patients hospitalized with HF.Clinicl factors independently correlated with S A were age,gender,BMI,history of hypertension,DBP,NT-proBNP and LVEF,with severe SA were gender,BMI,hypertension,DBP and NT-proBNP.Part 3 Phenotypes and correlated clinical factors of sleep apnea in patients hospitalized with heart failureObjective:According to pathogenesis,sleep apnea(SA)is typically classified as predominantly obstructive(OSA)and predominantly central(CSA).The prevalence and associated clinical factors of OS A and CSA have been reported in stable heart failure(HF),however,such data is lack in paitients hospitalized with HF.We aimed to determine the prevalence and associated clinical factors of SA phenotypes in paitients hospitalized with HF.Methods:Patients hospitalized with HF from March 2019 to January 2020 were enrolled in this study.All enrolled patients underwent one-night sleep study using a multichannel cardiopulmonary polygraphy,among whom,those with a minimum valid recording time of 4 hours were included in analysis.SA was defined if the apnea-hypopnea index? 15 events/h.SA was classified as CSA if? 50%apneas and hypopneas were central origin,otherwise,was classified as OSA.Results:Of 314 patients included in analysis,there were 56(17.8%)with OSA and 121(38.5%)with CSA,respectively.Multivariate linear analysis showed obstructive AHI was correlated with age,gender and body mass index(BMI),whereas central AHI was correlated with age,BMI,systolic blood pressure,N-terminal-pro brain natriuretic peptide(NT-proBNP)and partial pressure of arterial carbon dioxide(PaCO2).Multinomial logistic regression analysis showed that OSA was associated with older age(per 10 years increase:OR=1.405,95%CI:1.113?1.774,P=0.004),higher BMI(per 5 kg/m2 increase:OR=2.445,95%CI:1.577?3.792,P<0.001)and increased NT-proBNP(lnNT-proBNP:OR=1.372,95%CI:1.040?1.811,P=0.025),CSA was associated with male sex(OR = 2.120,95%CI:1.139?3.948,P=0.018),older age(per 10 years increase:OR=1.479,95%CI:1.227?1.783,P<0.001),higher BMI(per 5 kg/m2 increase:OR=1.857,95%CI:1.283?2.687,P=0.001),increased NT-proBNP(lnNT-proBNP:OR=1.375,95%CI:1.095?1.727,P=0.006)and decreased PaCO2(per 1 mmHg decrease:OR=1.059,95%CI:1.011?1.108,P=0.015).Conclusion:CSA was the predominant phenotype of SA in patient hospitalized with HF.Clinical factors correlated with OSA were age,BMI and NT-proBNP,with CSA were age,gender,BMI,NT-proBNP and PaCO2.Part 4 Prognostic value of sleep apnea frequency and nocturnal hypoxemia in patients hospitalized with heart failureObjective:Nocturnal hypoxemia is one predominant pathophysiological consequence of sleep apnea(SA).It was reported that compared with SA frequency,nocturnal hypoxemia was more robust to influence the prognosis of stable heart failure(HF),however,their prognostic value in patients hospitalized with HF remains unknown.We aimed to compare the prognostic effect of SA frequency and nocturnal hypoxemia in a cohort of patients hospitalized with HF.Methods:Patients hospitalized with HF from January 2015 to December 2017 were enrolled in the study and underwent sleep tests.The apnea-hypopnea index(AHI),oxygen desaturation index(ODI),mean oxygen saturation(MeanSO2),minimal oxygen saturation(MinSO2)and the percentage of time with saturation below 90%(T90%)during night were recorded by a multichannel cardiopulmonary polygraphy.Patients with a minimum valid recording time of 4 hours were included in analysis.SA frequency was assessed by AHI and the value of 15 events/h was used as the threshold.The median values were used as the thresholds for ODI,MeanSO2,MinSO2 and T90%.The endpoint was defined as the first event of all-cause death,heart transplantation,left ventricular assist device implantation,unplanned hospitalization for worsening heart failure,acute coronary syndrome(ACS),significant arrhythmias(sustained ventricular tachycardia,ventricular fibrillation and asystole)and stroke.Results:Of 382 patients included in analysis,189(49.5%)had AHI?15 events/h,the median value of T90%was 3.6%.A total of 185 endpoint events occurred.Fifty-one patients were with all-cause death,15 with heart transplantation,104 with worsening heart failure,3 with nonfatal ACS,4 with significant arrhythmias and 8 with strokes.Kaplan-Meier curves showed that the risk of endpoint did not differ between patients with AHI?15 events/h and those with AHI<15 events/h(52.4%vs 44.6%,log-rank P=0.353),however,patients with T90%?3.6%had a significantly higher risk of endpoint compared to those with T90%<3.6%(54.5%vs 42.4%,log-rank P=0.023).Stepwise multivariate Cox regression analysis showed that T90%was independently associated with endpoint[hazard ratio(HR)=1.008,95%confidence interval(CI):1.001?1.016,P=0.033).Compared to those with T90%<3.6%,the risk of endpoint increased by approximate 40%in patients with T90%?3.6%(HR=1.408,95%CI:1.030?1.925,P=0.032).However,AHI was not one independent risk factor of adverse outcomes in multivariate Cox analysis.Conclusion:Nocturnal hypoxemia had a greater impact than SA frequency in the prognosis of pateients hospitalized with HF.Whether or not treatment of nocturnal hypoxemia can improve the prognosis of HF and S A needs to be determined.Part 5 Association of sleep apnea and worsening renal function in patients hospitalized with heart failureObjective:Worsening renal function(WRF)is one risk factor associated with adverse outcomes in heart failure(HF).Sleep apnea(SA)is responsible for renal dysfunction by multiple pathophysiological mechanisms,however,the association between SA and WRF in HF remains unknown.The present study aimed to determine the association between S A and WRF in patients hospitalized with HF.Methods:This retrospective study enrolled hospitalized patients with HF from 2015 to 2017.Sleep studies were performed during hospitalization using a multichannel cardiopulmonary polygraphy.The apnea-hypopnea index(AHI)and the oxygen desaturation index(ODI)were recorded.SA defined as AHI?15/h.Patients with a minimum valid recording time of 2 hours were included in analysis.Serum creatinine(SCr)was routinely measured on admission,at discharge and at a 2-or 3-day interval during hospitalization.WRF was defined as an absolute increase in SCr of?26.5 ?mol/L at any time during hospitalization from admission.Results:There were a total of 462 patients included in the study,whose median AHI and ODI were 14.9 events/h and 19.4 events/h,respectively.There were 230(49.8%)patients with SA.WRF occurred in 125(27.1%)patients and was observed more in those with AHI?15 events/h(32.2%vs 22.0%,P=0.014)or with ODI?20 events/h(32.6%vs 21.8%,P=0.009).In multivariate logistic regression analysis,AHI and ODI were independently associated with WRF.The possibility of WRF increased by 12.4%for every 5 events/h increase in AHI[odds ratio(OR)=1.124,95%confidence interval(CI):1.054-1.199,P<0.001];patients with AHI?15 events/h had a significant higher possibility of WRF than those with AHI<15 events/h(OR=1.549,95%CI:1.004?2.389,P=0.048).Similarily,the possibility of WRF increased by 12.4%for every 5 events/h increase in ODI(OR=1,124,95%CI:1.050?1.203,P=0.001);patients with ODI?20 events/h had a significantly higher possibility of WRF than those with ODI<20 events/h(OR=1.708,95%CI:1.111?2.625,P=0.015).Conclusion:SA is associated with in-hospital WRF and might be a predictor for develpment of WRF in HF patients.