Cheyne-Stokes respiration (CSR) in severe stable congestive heart fail
ure (CHF) may be associated with significant nocturnal arterial oxygen
desaturation and sleep disruption. Previous investigations of inhaled
CO2 in CSR have been uncontrolled and of short duration, sleep has no
t been monitored electroencephalographically, and most patients studie
d have had neurological disease with or without cardiac disease. The p
urpose of our study was to document the effects of inhaled CO2 on CSR
in patients with severe stable CHF (left ventricular ejection fraction
<35% and NYHA class 3 or 4 dyspnea) in controlled all-night polysomno
graphic studies. Six patients were studied for 3 nights and days: adap
tation, control and inhalation of CO2. These patients received a const
ant F1CO2 = 0.03 in air (with a 4-5 mm Hg increase in PaCO2) on night
3. This caused virtual abolition of CSR as reflected by CSR duration/t
otal sleep time (62-2.2%; p = 0.0012) and CSR duration/nonrapid eye mo
vement (NREM) sleep time (73-2.4%; p = 0.00064), and NREM apnea index
was reduced from 33.5 to zero (p = 0.026). The apparatus used to accur
ately control F1CO2, however, was intrusive and some features of sleep
structure such as sleep latency were adversely affected. We conclude
that inhalation of CO2 with a constant F1CO2 = 0.03 virtually eradicat
es CSR in all-night polysomnographically monitored studies in patients
with severe stable CHF. The clinical significance of these findings r
emains to be determined.