|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the University of Oklahoma Health Sciences Center (Drs. Dernaika, Tawk, Younis, and Kinasewitz), Oklahoma City, OK; and University of Arkansas for Medical Sciences (Dr. Nazir), Little Rock, AR.
Correspondence to: Tarek Dernaika, MD, 920 Stanton L. Young Blvd, WP 1310, Oklahoma City, OK 73104; e-mail: Tarek-Dernaika{at}ouhsc.edu
Abstract
Objective: To determine whether central sleep apnea (CSA) occurring during continuous positive airway pressure (CPAP) titration in patients with obstructive sleep apnea (OSA) reflects subclinical congestive heart failure (CHF), and whether these events will improve with CPAP therapy.
Design: Cross-sectional analysis of patients with suspected sleep-related breathing disorders referred for split-night polysomnography
Patients and methods: Forty-two OSA patients with and without CPAP-related CSA were analyzed. All CSA patients (n = 21) and control subjects (n = 21) underwent echocardiography, pulmonary function testing, and arterial blood gas (ABG) analysis. Repeat polysomnography with CPAP was performed 2 to 3 months after adequate CPAP therapy in CSA group patients.
Results: Demographic, Epworth sleepiness scale, pulmonary function test, ABG, and baseline diagnostic polysomnography findings were similar in both groups. There was no difference in the prevalence of subclinical left ventricular systolic dysfunction in the CSA group vs the control group. CSA patients had decreased sleep efficiency (SE), increased sleep stage 1 percentage, sleep stages shift, wake time after sleep onset (WASO), and total arousals compared to control subjects. Twelve of 14 patients (92%) in the CSA group demonstrated complete or near-complete resolution of CSA events on follow-up polysomnography and showed improvement in SE, WASO, and total arousals compared to their baseline study.
Conclusions: CSA events occurring during CPAP titration are transient and self-limited. They may be precipitated by the sleep fragmentation associated with initial CPAP titration and are not associated with an increased prevalence of occult CHF compared to OSA patients without CPAP-related CSA.
Key Words: central sleep apnea Cheyne-Stokes respiration continuous positive airway pressure heart failure obstructive sleep apnea regulation of respiration sleep-disorders therapy
Central sleep apnea (CSA) affects patients with congestive heart failure (CHF),1 sleep-related hypoventilation syndromes,2 and neurologic disorders.34 Premature, full-term infants5 and healthy subjects at sleep onset or living at high altitude6 are also at risk for CSA. Instability of the respiratory control system during sleep with lowering of the PaCO2 below the apneic threshold, thereby resulting in central apnea, appears to play the most important role in the pathophysiology of CSA.7
In addition, CSA is occasionally observed as an isolated phenomenon developing during continuous positive airway pressure (CPAP) titration studies in patients with obstructive sleep apnea (OSA). Central apneas may occur following arousals from respiratory effort-related arousals (inadequate pressure) or arousals secondary to excessive pressure or a mouth leak. Nevertheless, the outcome and implication of these events remain undetermined.
Central apneas during sleep cause arousals, disrupted sleep, increased sympathetic activity, and are associated with increased morbidity and mortality in patients with CHF.8 We hypothesized that patients who demonstrate CSA during CPAP titration have subclinical CHF, and sought to determine whether these events are persistent or represent a transient and self-limited instability of the ventilatory control. To our knowledge, this has not been systematically studied before.
Materials and Methods
Patients
Patients were considered eligible for the study if they were
18 years old and seen from September 2004 through December 2005 with OSA identified using split-night polysomnography, and who also demonstrated CSA events exclusively during the CPAP titration portion of the sleep study (CSA group). CSA was defined as a CSA index (CSAI)
5/h. During CPAP titration, the pressure was raised only to eliminate obstructive and not central events. Subjects were excluded if they demonstrated CSA during the diagnostic portion of the sleep study, or had known CHF by history or previous echocardiography, COPD or other significant lung disease, daytime hypercapnia or hypoxemia, cerebrovascular disease, seizure disorder, or a history of benzodiazepines, narcotics, or illicit drug use. For each subject, a control patient with OSA but no CSA during the CPAP titration portion of a successful split-night study who underwent a transthoracic echocardiography (TTE), pulmonary function testing, and arterial blood gas (ABG) analysis within 4 weeks of the sleep study was identified by review of sleep laboratory logs. The study protocol was reviewed and approved by the institutional review board, and signed informed consent was obtained from all study participants.
Procedures
Polysomnography included a standard EEG, electrooculography, and submental electromyography for sleep staging. Respirations were monitored using chest and abdominal impedance plethysmography. Airflow was assessed with an oronasal thermistor as well as nasal pressure cannula. Arterial oxygen saturation was monitored using pulse oximetry (Nellcor N-595; Nellcor; Pleasanton, CA).
Heart rate and rhythm were recorded and monitored with continuous ECG. Periodic limb movements were monitored using bilateral tibial electromyogram leads. All polysomnography was administered and scored by experienced sleep laboratory technicians. Apnea was defined as a complete cessation of airflow lasting
10 s. Hypopnea was defined as reduction in respiratory airflow of > 30%, lasting
10 s, and accompanied by a decrease of
3% in oxygen saturation and/or an arousal. The apnea-hypopnea index (AHI) was defined as the number of apneas and hypopneas per hour of sleep and was used to diagnose OSA. OSA was defined as an AHI
5/h. Arousals were defined according to the recommendations of the American Sleep Disorders Association Atlas task force.9 Sleep data were staged according to the system of Rechtschaffen and Kales.10
An AHI
20/h and a minimum of 2 h of sleep during the diagnostic sleeping portion of the study were essential for patients to meet criteria for split-night polysomnography. A minimum of 3 h of sleep during the CPAP titration portion of the test was required. The starting CPAP was 4 to 5 cm H2O. The pressure was increased gradually every few minutes with a maximum waiting time of 20 min if only obstructive apneas, hypopneas, and flow limitation were obvious. Successful CPAP titration was defined as an obstructive AHI < 10/h on optimal CPAP pressure.
Each patient underwent TTE performed by a cardiologist blinded to the results of the sleep study, to assess left and right systolic and diastolic ventricular function. Diagnostic criteria of diastolic dysfunction were the following: (1) presence of normal left ventricular ejection fraction (LVEF)
50%; and (2) evidence of abnormal left ventricular relaxation, diastolic distensibility, or diastolic stiffness, whereas systolic dysfunction was defined as LVEF
50%.11 ABG analysis and pulmonary function testing were performed for all patients before the baseline sleep study. Patients were in the sitting position, awake, and not stressed or in pain when performing arterial puncture.
Repeat ABG analysis and polysomnography were performed in the CSA group after 8 to 12 weeks of adequate CPAP therapy. During the follow-up study, patients slept with their CPAP machines set at pressures determined by the baseline CPAP titration study. Patients with ventricular dysfunction at baseline underwent a repeat TTE after CPAP therapy.
Statistical Analysis
The distribution of data was analyzed for normality. Standard parametric tests were used to detect differences between two different groups if homogeneity of variance was present. Otherwise, nonparametric tests were used. The Fisher exact test was used to evaluate the distribution of categorical variables in a 2 x 2 contingency table when applicable. A p value < 0.05 was considered statistically significant.
Results
A total of 116 patients with OSA who met criteria for a split-night study were evaluated. Twenty-three patients (19.8%) showed CSA events during CPAP titration. Two subjects with CPAP-related CSA were further excluded from the study because of an unsuccessful titration to eliminate obstructive events, leaving 21 subjects with CPAP-related CSA (CSA group; CSAI, 22.1 ± 13.8). These patients were compared to an equal number of control subjects (CSAI, 0.5 ± 0.9) identified by review of sleep logs as indicated in methods. The first 21 patients who met these criteria formed the control group. All included subjects had evidence of OSA on their diagnostic portion of the split-night study as defined earlier. Findings of baseline characteristics, Epworth sleepiness scale, ABG results, and the baseline diagnostic portion of polysomnography were similar in both groups (Table 1 ).
|
|
|
|
|
|
CHF is associated with Cheyne-Stokes breathing during wakefulness and sleep. This type of respiration is quite distinct with a crescendo and decrescendo ventilatory pattern with an apnea or hypopnea at the nadir. Javaheri and Parker1 found that up to 45% of patients with CHF (LVEF
40%) had > 20 central apneas or hypopneas per hour of sleep.
CPAP therapy has been proven highly effective in treating OSA, leading to improvement in both subjective and objective daytime alertness as well as quality of life.12 CPAP has also been shown to be useful in the treatment of CSA with or without presence of Cheyne-Stokes respirations.1314 Nasal CPAP may cause acute reduction in CSA events by increasing PCO2 above the apneic threshold15 and/or a reduction in the severity of oxygen desaturation in patients with CHF. Additionally, several studies1617 have shown that treatment with CPAP produces long-term improvements in cardiac function in CHF patients that eventually lead to a decrease in the amount of periodic breathing and CSA.
Sustained improvements in central apneas were also described after tracheostomy in patients with OSA.1819 In a study by Gilleminault et al,19 reversible alterations in ventilatory control and improvement in ventilatory response to CO2 were found in five male patients 31 to 57 years of age with OSA after tracheostomy monitored before and several times during the 6 months after tracheostomy. A dramatic decrease in the overall apnea index was seen immediately after surgery, but the number of central apneas and the central apnea index did not reach low values until several months after tracheostomy. This improvement may represent adaptation of the central chemoreceptors to the reduction of hypercapnea overnight or reflects changes in the quality of sleep, since sleep deprivation and fragmentation results in a significant deterioration of the hypercapnic ventilatory responses.20
Whether CSA occurring exclusively during CPAP titration reflects any subclinical cardiac dysfunction has never been studied before. In this study, the mean ejection fraction and the prevalence of both systolic and diastolic dysfunction were similar in the study and control groups. While occult sleep-disordered breathing, predominantly CSA, has been documented in patients with CHF, the reverse has been shown to be true as well.21 Although, we did not specifically address that question, our study shows that patients with OSA and CSA noted during CPAP titration do not have a higher prevalence of occult CHF when compared to patients who do not manifest CPAP-related CSA. Nevertheless, abnormalities in cardiac function and structure revealed by echocardiography were not uncommon in patients with sleep-disordered breathing (50% overall prevalence among all study groups).
Awake hypocapnia is an important determinant of CSA in CHF regardless of the degree of cardiac dysfunction.22 In our study, patients in the CSA group had a lower daytime PCO2 compared to control subjects (p = 0.07), whereas a trend toward a higher PCO2 following CPAP therapy was observed in the 14 subjects with CPAP-related CSA who underwent repeat polysomnography. Chaouat et al23 also reported a slight but statistically significant increase in daytime PCO2 in patients with OSA treated with CPAP. Since we did not examine the ventilatory response in this particular group of patients, the role of awake hypocapnea in patients with CPAP-related CSA cannot be ascertained based on our results and remains an open question. The mechanism underlying the increase in daytime PCO2 in these patients after CPAP therapy is also unclear but may be related to a change in ventilatory pattern or drive.1524
Outcome of CPAP-Related CSA
Sleep onset and early stages of sleep represent a particularly susceptible period for the development of respiratory instability and CSA. Sleep-state variation between wakefulness and light sleep may lead to fluctuations in PCO2 level below the apneic threshold required to maintain rhythmic breathing, resulting in the development of CSA.2526 This pattern may occur even in normal healthy adults and is more prevalent during the early stages of non-rapid eye movement sleep.27 The recovery from apnea is usually associated with transient wakefulness or arousal and hyperventilation. Sleep state and breathing continue to oscillate until sleep is consolidated, a higher PCO2 set point is established, and PCO2 is maintained above the apnea threshold.
CPAP is generally initiated during an attended split-night polysomnography with the pressure level being titrated to eliminate obstructive respiratory events. However, this may be bothersome to patients who are receiving CPAP for the first time, resulting in frequent awakenings, arousals, and sleep stages shifts, thus facilitating ventilatory instability and potentially leading to the generation of CSA. However, in our study, the difference in optimal CPAP pressure was small between the two groups; and the CPAP titration protocol, the total sleep time on CPAP, and the time spent at optimal pressure were similar.
There was a decrease in central apnea events at optimal CPAP, which we attribute to the predominance of rapid eye movement (REM) sleep (40% of total sleep time at optimal pressure). During REM sleep, the occurrence of periodic breathing and central apneas is markedly reduced or absent.2829 An immediate improvement in ventilatory control in theses subjects while receiving optimal CPAP is less likely but cannot be excluded.
Nearly all patients with CPAP-induced CSA at baseline had complete resolution of the central events on follow-up when studied on the prescribed level of CPAP used to eliminate OSA. The patients were studied on the same effective pressure for the obstructive apnea that was determined on the first night of sleep study. Our findings suggest that CPAP-related CSA may be explained by ventilatory control instability resulting from sleep fragmentation as demonstrated by decreased SE, increased sleep stage 1, WASO, and arousals in the study group during the baseline study. These central respiratory events were consistently eliminated or significantly reduced on follow-up polysomnography, and this was associated with a significant improvement in total sleep time, as well as reduced light sleep and total arousals. This could be attributed to a progressive improvement in the hypercapnic ventilatory response, as OSA patients treated with CPAP, have improved sleep quality, fewer sleep state transitions, and eventually more restorative sleep.
CPAP-related CSA appears to represent a benign and transient phenomenon and is likely related to sleep fragmentation and sleep stage shifts that occur with initial CPAP titration. Bilevel positive airway pressure (BPAP) is often used in the treatment of Cheyne-Stokes respirations in patients with CHF,30 and is frequently considered as an alternative mean to treat CSA events occurring during CPAP titration. However, reports3132 have suggested that BPAP may actually worsen CSA with or without periodic breathing mediated by an overshoot in ventilation causing decreased PCO2 below the apneic threshold. Treating CPAP-induced central apnea with BPAP could also carry the potential of worsening these self-limited events. Finally, the persistence of CSA despite adequate CPAP therapy may warrant further investigation for occult cardiovascular disease. Larger prospective studies are needed for assessing the incidence and prevalence of CHF in patients with sleep-disordered breathing, in particular CSA.
Footnotes
Abbreviations: ABG = arterial blood gas; AHI = apnea-hypopnea index; BPAP = bilevel positive airway pressure; CHF = congestive heart failure; CPAP = continuous positive airway pressure; CSA = central sleep apnea; CSAI = central sleep apnea index; LVEF = left ventricular ejection fraction; OSA = obstructive sleep apnea; REM = rapid eye movement; SE = sleep efficiency; TTE = transthoracic echocardiography; WASO = wake time after sleep onset
The authors have no conflicts of interest to disclose.
Received for publication October 18, 2006. Accepted for publication March 19, 2007.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |