Author + information
- Received July 15, 2004
- Revision received November 29, 2004
- Accepted December 6, 2004
- Published online June 21, 2005.
- Kengo Usui, MD, PhD,
- T. Douglas Bradley, MD, FRCPC,
- Jonas Spaak, MD, PhD,
- Clodagh M. Ryan, MB, BCh,
- Toshihiko Kubo, MD, PhD,
- Yasuyuki Kaneko, MD and
- John S. Floras, MD, DPhil, FRCPC, FACC⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. John S. Floras, Suite 1614, 600 University Avenue, Toronto, Ontario M5G 1X5, Canada.
Objectives This study was designed to determine whether reductions in morning systolic blood pressure (BP) elicited by treatment of moderate to severe obstructive sleep apnea (OSA) in heart failure (HF) patients are associated with a reduction in sympathetic vasoconstrictor tone.
Background Daytime muscle sympathetic nerve activity (MSNA) is elevated in HF patients with coexisting OSA. In our recent randomized trial in HF, abolition of OSA by continuous positive airway pressure (CPAP) increased left ventricular ejection fraction (LVEF) and lowered morning systolic BP.
Methods Muscle sympathetic nerve activity, BP, and heart rate (HR) of medically treated HF patients (EF <45%) and OSA (apnea-hypopnea index ≥20/h of sleep) were recorded on the morning after overnight polysomnography, and again one month after patients were randomly allocated nocturnal CPAP treatment or no CPAP (control).
Results In nine control patients, there were no significant changes in the severity of OSA, MSNA, systolic BP, or HR. In contrast, in the 8 CPAP-treated patients, OSA was attenuated, and there were significant reductions in daytime MSNA (from 58 ± 4 bursts/min to 48 ± 5 bursts/min; 84 ± 4 bursts/100 heart beats to 72 ± 5 bursts/100 heart beats; p < 0.001 and p = 0.003, respectively), systolic BP (from 135 ± 5 mm Hg to 120 ± 6 mm Hg, p = 0.03), and HR (from 69 ± 2 min−1to 66 ± 2 min−1; p = 0.013).
Conclusions Treatment of coexisting OSA by CPAP in HF patients lowers daytime MSNA, systolic BP, and HR. Inhibition of increased central sympathetic vasoconstrictor outflow is one mechanism by which nocturnal CPAP reduces awake BP in HF patients with moderate to severe OSA.
Obstructive sleep apnea (OSA), with ≥15 apneic or hypopneic events per hour, is present in 4% to 9% of the adult North American population aged 30 to 60, but its reported prevalence in heart failure (HF) patients with impaired left ventricular (LV) systolic function is substantially higher—between 11% and 37% (1). Heart failure is a condition in which sympathetic activation in the awake state has been linked to premature mortality and sudden death (2). Obstructive sleep apnea-induced apnea, hypoxia, hypercapnia, and arousal trigger surges in central sympathetic vasoconstrictor outflow, peripheral resistance, and blood pressure (BP) during sleep (1). Individuals with normal LV systolic function and OSA exhibit increased muscle sympathetic nerve activity (MSNA) even while awake (3,4). This carryover effect is also evident in HF patients, in whom coexisting OSA is associated with higher daytime MSNA (5).
In our recent randomized trial involving 24 HF patients (mean LV ejection fraction [LVEF] 27%) with coexisting OSA, 1 month of therapy with nocturnal continuous positive airway pressure (CPAP) caused a increase in daytime LVEF and a significant reduction in morning systolic BP (6). We considered attenuation of daytime sympathetic vasoconstrictor discharge a likely mechanism for this fall in BP. We therefore recruited additional patients to test the hypothesis that abolition of coexisting OSA in HF by CPAP would lower daytime MSNA.
Following ethics board approval and informed consent, we studied men and women with 1) HF of >6 months duration; 2) LVEF ≤45% (radionuclide angiography); 3) >3 months of stable optimal drug therapy at highest tolerated dose; 4) moderate to severe OSA (≥20 apneas and hypopneas/h of sleep) with >50% of events obstructive (1,7); and 5) sinus rhythm. Exclusion criteria included: 1) primary valvular heart disease; 2) cardiac pacing; and 3) unstable angina, myocardial infarction, or cardiac surgery within three months (6).
All subjects underwent a second baseline overnight polysomnographic study, with sleep stages, apneas, hypopneas, and arousals defined and scored as previously described (6,7). Oxyhemoglobin saturation (SaO2) was monitored by pulse oximetry. The apnea-hypopnea index (AHI) was calculated as the frequency of apneas and hypopneas/h of sleep.
Measurement of HR, BP, and MSNA
The next morning, the electrocardiogram, BP (digital photoplethysmography; Finapres 2300, Ohmeda, Englewood, Colorado) and peroneal MSNA were recorded over 15 min, with subjects awake, resting quietly supine, and breathing without apnea, as confirmed by respiratory inductance plethysmography. Muscle sympathetic nerve activity was expressed as burst frequency (bursts/min) and burst incidence (bursts/100 cardiac cycles) (8,9).
Subjects were allocated randomly to either a control group (optimal HF drug therapy) or to a group that, in addition, received CPAP. The night after the baseline sleep study, the latter subjects underwent overnight CPAP titration, during which pressure was adjusted to abolish apneas and hypopneas or to the highest tolerated level. They were then provided a metered CPAP machine to document hours of use and were instructed to apply this for >6 h nightly. After one month the sleep study and the awake study were repeated.
All data were acquired and analyzed by investigators blinded to the sequence of studies and treatment. Data are expressed as mean ± SEM. Analyses were performed using SigmaStat 2.03 (SPSS Inc., Chicago, Illinois). Baseline characteristics were compared by two-tailed unpaired ttests for continuous variables and the Fisher exact test for nominal variables. Two-way repeated measures analyses of variance, followed by Tukey’s test, were used to compare within- and between-group differences in variables obtained one month apart. A p value <0.05 was considered significant.
High-quality MSNA data from both sessions were acquired from 17 subjects, 9 randomized to control and 8 to CPAP (Table 1).Groups were similar for age; body mass index; LVEF; AHI (>80% of respiratory events were obstructive); minimum SaO2; sleep structure; arousal frequency; and use of digoxin (41%), diuretics (76%), angiotensin-converting enzyme inhibitors (100%), and beta-receptor antagonists (59%).
Neither drug use nor body mass index changed significantly between the baseline and one-month studies. In control patients, there were no significant changes in total or obstructive AHI, or any other polysomnographic variable. In contrast, CPAP, at a pressure of 7.5 ± 0.5 cm H2O (used 6.0 ± 0.6 h/night), reduced total AHI (from 40.4 ± 7.9 events/h to 8.8 ± 1.5 events/h; p < 0.001 within- and p < 0.01 between-group interaction), obstructive AHI (from 33.3 ± 4.5 events/h to 4.2 ± 1.0 events/h; p < 0.001 within- and p < 0.005 between-group interaction), and arousal frequency (from 32.1 ± 8.7/h to 14.3 ± 3.5/h; p < 0.025), and increased minimum SaO2during sleep (90.8 ± 0.5% vs. 78.4 ± 4.6%; p < 0.05).
In the control group, hemodynamic and microneurographic variables were similar on the two study days (Table 2).In CPAP-treated subjects, awake systolic BP fell by 15.4 ± 4.9 mm Hg (p = 0.015; p = 0.04 for the between-group comparison), as did HR (p = 0.013; p = 0.004 for between-group comparison). In contrast with control patients, there were significant reductions in MSNA burst frequency in all subjects (p < 0.001; p < 0.009 for between-group comparison). The MSNA burst incidence also fell (p = 0.003) after one month of CPAP treatment (Table 2, Figs. 1 and 2).⇓⇓
This is the first randomized trial to examine the impact of treating OSA on MSNA. In those patients with coexisting systolic HF, treating OSA with nocturnal CPAP for one month caused significant reductions in morning MSNA, systolic BP, and HR. Remarkably, these cardiovascular effects of CPAP were observed on a background of medical HF therapy (including angiotensin-converting enzyme inhibition and beta-blockade) that should attenuate the impact of any additional intervention on BP or HR. Because ventricular systolic function is affected adversely by increases in cardiac afterload (10,11), these decreases have important therapeutic implications.
These findings also provide new insight into the contribution of sympathetic vasoconstrictor tone to BP regulation in HF patients with coexisting OSA. Muscle sympathetic nerve activity is increased in most patients with HF, in whom it relates directly to resistance in the vascular bed distal to the recording electrode (12). However, there is considerable interindividual variation in sympathetic nerve firing rates in patients with systolic HF that cannot be explained simply on the basis of reflex responses to their altered hemodynamics (2,8). One potential source of such variation is the additional sympatho-excitatory influence of coexisting OSA. Upper airway obstruction not only activates the sympathetic nervous system of HF patients reflexively by eliciting apnea, hypoxia, hypercapnia, a fall in cardiac output, and arousal from sleep (1,9), but exposure to brief episodes of intermittent hypoxia evokes sympathetic activation and BP elevation that persist long after recovery from the hypoxic stimulus (13,14). Thus, recurrent obstructive apneas at night in HF patients could elicit sustained aftereffects on MSNA and BP that carry-over into the awake state. Consistent with this concept, in our series of 301 patients with HF, those with OSA had higher daytime systolic BP than those without, despite their receiving more antihypertensive therapy; systolic BP correlated directly with the AHI (15). We proposed that their higher awake BP was due in part to increased daytime sympathetic vasoconstrictor tone (5) that might be reduced if OSA were treated with CPAP.
In HF patients, acute abolition of obstructive apnea during sleep with CPAP causes immediate reductions in systolic BP and HR (16). In subjects with OSA but without HF, long-term treatment with CPAP lowers awake systolic BP (17). In our recent randomized trial involving medically treated HF patients with OSA, the addition of nocturnal CPAP increased mean daytime LVEF from 25% to 34%, yet at the same time lowered morning systolic BP by 10 ± 4 mm Hg (6). The present study is an extension of that trial, with the objective of testing the new hypothesis that abolition of obstructive apnea during sleep would abrogate its daytime sympatho-excitatory pressor aftereffects (5). Our demonstration of reductions in MSNA, systolic BP, and HR after one month of treatment of OSA by CPAP are consistent with this hypothesis and indicate that inhibition of sympathetic vasoconstrictor discharge is one mechanism by which nocturnal CPAP lowers daytime systolic BP in HF patients with moderate to severe OSA.
A similar randomized trial detected an absolute increase in daytime LVEF of 5% after 3 months of CPAP therapy, but no change in mean arterial BP (systolic BP was not reported) (18). Importantly, these authors recruited subjects with milder OSA (AHI >5 events/h) and HF (LVEF ≤55%) and presumably less baseline sympathetic activation. This cannot be confirmed, however, because neither MSNA nor daytime plasma norepinephrine were measured.
Four major risk factors for ventricular remodeling, myocyte loss, disease progression, and premature mortality in patients with impaired LV systolic function and OSA are 1) abrupt increases in negative intrathoracic and LV transmural pressures (i.e., afterload) and distending forces during obstructive apneas, resulting in increased myocardial energy requirements; 2) nocturnal hypoxia and oxidative stress; 3) nocturnal and daytime sympathetic activation; and 4) nocturnal and daytime hypertension. Treatment of HF patients with coexisting OSA by nocturnal CPAP addresses each of these risk factors and therefore has the potential to improve LV systolic function via several mechanisms. Compared with HF patients with normal breathing patterns during sleep, those with OSA are exposed to the adverse effects on the failing heart and circulation of increased central sympathetic outflow during sleep and when awake. By removing this additional apnea-induced stimulus to sympathetically mediated vasoconstriction, and thereby lowering systolic BP through specific therapy of coexisting OSA, nocturnal CPAP might improve the prognosis of HF patients with moderate to severe OSA. This hypothesis merits specific testing in a longer term outcome trial.
Supported by Canadian Institutes of Health Research (CIHR) Grant MOP 11607. Drs. Usui, Spaak, Ryan, and Kubo received unrestricted research fellowships, respectively, from Respironics, Inc.; the Heart and Stroke Foundation (HSF) of Canada; the Toronto Rehabilitation Institute and Respironics, Inc; and the Japan Information Center for Respiratory Failure Patients. Dr. Kubo was supported in part by CIHR grant MT9721. Dr. Bradley is a CIHR senior scientist. Dr. Floras holds the Canada Research Chair in Integrative Cardiovascular Biology and is an Ontario HSF career investigator. Drs. Bradley and Floras have been awarded a CIHR-University-Industry clinical trial grant involving patients with heart failure and central apnea, partially supported by manufacturers of CPAP devices.
- Abbreviations and Acronyms
- apnea-hypopnea index
- blood pressure
- continuous positive airway pressure
- heart failure
- heart rate
- left ventricular
- left ventricular ejection fraction
- muscle sympathetic nerve activity
- obstructive sleep apnea
- oxyhemoglobin saturation
- Received July 15, 2004.
- Revision received November 29, 2004.
- Accepted December 6, 2004.
- American College of Cardiology Foundation
- Bradley T.D.,
- Floras J.S.
- Egri Z.J.,
- Spaak J.,
- Yu E.,
- et al.
- Rechtschaffen A.,
- Kales A.
- Notarius C.F.,
- Atchison D.J.,
- Floras J.S.
- Cohn J.N.
- Hara K.,
- Floras J.S.
- Arabi Y.,
- Morgan B.J.,
- Goodman B.,
- Puleo D.S.,
- Xie A.,
- Skatrud J.B.
- Xie A.,
- Skatrud J.B.,
- Puleo D.S.,
- Morgan B.J.
- Tkacova R.,
- Rankin F.,
- Fitzgerald F.S.,
- Floras J.S.,
- Bradley T.D.
- Becker H.F.,
- Jerrentrup A.,
- Ploch T.,
- et al.