REVIEW / SYNTHESE

Respiratory modulation of the autonomic nervoussystem during Cheyne–Stokes respiration1

Richard S.T. Leung, John S. Floras, and T. Douglas Bradley

Abstract: Cheyne–Stokes respiration (CSR) is associated with increased mortality among patients with heart failure.However, the specific link between CSR and mortality remains unclear. One possibility is that CSR results in excitation ofthe sympathetic nervous system. This review relates evidence that CSR exerts acute effects on the autonomic nervous sys-tem during sleep, and thereby influences a number of cardiovascular phenomena, including heart rate, blood pressure, at-rioventricular conduction, and ventricular ectopy. In patients in sinus rhythm, heart rate and blood pressure oscillate duringCSR in association with respiratory oscillations, such that both peak heart rate and blood pressure occur during the hyper-pneic phase. Inhalation of CO2 abolishes both CSR and the associated oscillations in heart rate and blood pressure. In con-trast, O2 inhalation sufficient to eliminate hypoxic dips has no significant effect on CSR, heart rate, or blood pressure. Inpatients with atrial fibrillation, ventricular rate oscillates in association with CSR despite the absence of within-breath res-piratory arrhythmia. The comparison of RR intervals between the apneic and hyperpneic phases of CSR indicates that thisbreathing disorder exerts its effect on ventricular rate by inducing cyclical changes in atrioventricular node conductionproperties. In patients with frequent ventricular premature beats (VPBs), VPBs occur more frequently during the hyper-pneic phase than the apneic phase of CSR. VPB frequency is also higher during periods of CSR than during periods ofregular breathing, with or without correction of hypoxia. In summary, CSR exerts multiple effects on the cardiovascularsystem that are likely manifestations of respiratory modulation of autonomic activity. It is speculated that the rhythmic os-cillations in autonomic tone brought about by CSR may ultimately contribute to the sympatho-excitation and increasedmortality long observed in patients with heart failure and CSR.

Key words: central sleep apnea, heart rate variability, atrial fibrillation.

Resume : La respiration de Cheyne–Stokes (RCS) est associee a une augmentation du taux de mortalite chez les patientspresentant une insuffisante cardiaque. Toutefois, le lien specifique entre la RCS et la mortalite demeure obscur. Une possi-bilite serait que la RCS provoque une excitation du systeme nerveux sympathique. La presente synthese rend compte deresultats qui demontrent que la RCS a des effets aigus sur le systeme autonome nerveux durant le sommeil et qu’elle in-fluence ainsi plusieurs phenomenes cardiovasculaires, tels que la frequence cardiaque, la pression arterielle, la conductionatrio-ventriculaire et l’ectopie ventriculaire. Chez les patients en rythme sinusal, la frequence cardiaque et la pression arte-rielle oscillent durant la RCS en association avec les oscillations respiratoires, de sorte la frequence cardiaque et la tensionarterielle de crete se produisent durant la phase d’hyperpnee. L’inhalation de CO2 supprime la RCS et les oscillations asso-ciees de la frequence cardiaque et de la tension arterielle. A l’oppose, l’inhalation d’O2 en quantite suffisante pour eliminerles ralentissements hypoxiques n’a pas d’effet significatif sur la RCS, la frequence cardiaque et la tension arterielle. Chezles patients en fibrillation auriculaire, la frequence ventriculaire oscille en association avec la RCS malgre l’absenced’arythmie respiratoire. La comparaison des intervalles RR entre les phases apneique et hyperpneique de la RCS indiqueque ce trouble respiratoire influe sur la frequence ventriculaire en reduisant les variations cycliques des proprietes deconduction du nœud atrio-ventriculaire. Chez les patients presentant de frequents battements ventriculaires prematures(BVP), les BVP se produisent plus frequemment durant la phase hyperpneique que durant la phase apneique de la RCS. Lafrequence des BVP est aussi plus elevee durant les periodes de RCS que durant les periodes de respiration reguliere, avecou sans correction de l’hypoxie. En resume, les multiples effets de la RCS sur le systeme cardiovasculaire sont probable-ment des manifestations de la modulation respiratoire de l’activite autonome. On suggere que les oscillations rythmiques dutonus autonome induites par la RCS pourraient ultimement contribuer a l’excitation du sympathique et a l’augmentation dutaux de mortalite observees depuis longtemps chez les patients presentant une insuffisance et une RCS.

Received 15 August 2005. Published on the NRC Research Press Web site at http://cjpp.nrc.ca on 24 February 2006.

R.S.T Leung,2 J.S. Floras, and T.D. Bradley. University of Toronto Centre for Sleep and Chronobiology, Toronto, ON M5B 1W8,Canada.

1This paper is one of a selection of papers published in this Special Issue, entitled Young Investigator’s Forum.2Corresponding author (e-mail: richard.leung@utoronto.ca).

61

Can. J. Physiol. Pharmacol. 84: 61–66 (2006) doi: 10.1139/Y05-145 # 2006 NRC Canada

Mots cles : apnee centrale du sommeil, variabilite de la frequence cardiaque, fibrillation auriculaire.

[Traduit par la Redaction]

______________________________________________________________________________________

IntroductionCheyne–Stokes respiration with central sleep apnea (CSR)

is a form of periodic breathing in which apneas and hypo-pneas alternate with ventilatory periods having a crescendo–decrescendo pattern of tidal volume. CSR is common amongpatients with congestive heart failure; it is present in 30%–40% in the 2 largest reported series (Javaheri et al. 1998;Sin et al. 1999), and its presence is associated with increasedmortality (Hanly and Zuberi-Khokhar 1996; Lanfranchi et al.1999; Sin et al. 2000). One possible mechanism linking CSRwith poor prognosis is through intermittent surges in sym-pathetic nervous activity targeting the heart and peripheralvasculature (Naughton et al. 1995; Trinder et al. 2000).

We sought to explore the manner in which CSR modu-lates the autonomic nervous system by examining its effecton a number of cardiovascular phenomena. At the level ofthe sinus node and peripheral vasculature, we examined theeffects of CSR on heart rate (HR) and blood pressure (BP)(Leung et al. 2003). At the level of the atrio-ventricular(AV) node, we examined whether CSR would influence theventricular response to atrial fibrillation (AF) by altering AVconduction properties (Leung et al. 2005). Finally, at thelevel of the ventricles, we explored the hypothesis that CSRprovokes ventricular ectopic beats (Leung et al. 2004).

Heart rate and blood pressureIt has been previously demonstrated, both during volun-

tary periodic breathing under normoxic conditions in healthysubjects and during spontaneous CSR during wakefulness inpatients with heart failure, that periodic breathing generatesaccompanying oscillations in HR and BP (Lorenzi-Filho etal. 1999a; Trinder et al. 2000). Therefore, neither hypoxianor arousals from sleep are necessary for inducing periodicoscillations in these hemodynamic variables during wakeful-ness, suggesting that periodic breathing alone is a sufficientstimulus for these effects. However, it was not knownwhether this was true for spontaneous CSR in patients withheart failure during sleep. To answer this question, we madeuse of the fact that inhalation of CO2 acutely eliminatesCSR (by raising PaCO2 above the apneic threshold), whereasO2 inhalation sufficient to abolish hypoxic dips does not(Lorenzi-Filho et al. 1999b). Therefore, to determine whetherBP and HR oscillations are mainly related to ventilatory os-cillations or recurrent dips in oxyhemoglobin saturation(SaO2), we administered CO2 and O2 by inhalation to heartfailure patients with CSR during overnight polysomnography.

In 10 subjects with systolic heart failure (left ventricu-lar ejection fraction <45%) and spontaneous CSR (apnea–hypopnea index >15 events per h, of which >80% werecentral), the periodic breathing pattern during sleep was ac-companied by oscillations in SaO2 and transcutaneous CO2and by oscillations in HR and BP, in which peaks in HRand BP occurred during hyperpnea and troughs during ap-nea (Fig. 1). Using spectral analysis techniques, it was de-

termined that the HR and BP oscillations were highlycoherent with the oscillations in respiration and that thepeak fluctuations in HR and BP were practically synchro-nous with each other and followed corresponding cycles inbreathing by approximately 10 s. Since HR and BP roseconcomitantly, it is clear that the HR changes were notbaroreflex-mediated.

Inhalation of CO2 in these 10 subjects (sufficient to raisetranscutaneous CO2 by an average of 2 mmHg) completelyabolished CSR and restored regular breathing (Lorenzi-Filhoet al. 1999b). HR and BP oscillations were eliminated alongwith the periodic breathing pattern. However, in 6 subjectswho received supplemental O2 at doses sufficient to com-pletely eliminate O2 desaturation, CSR and its associated os-cillations in HR and BP persisted in all subjects. Sinceelimination of CSR by CO2 inhalation abolished oscillationsin HR and BP, whereas alleviation of apnea-related dips inSaO2 by O2 inhalation did not, these phenomena appear tobe more tightly linked to oscillations in ventilation than tofluctuations in SaO2.

The precise mechanism linking CSR with BP and HR os-cillations remains to be determined. In particular, it is un-clear whether CSR acts directly on the autonomic nervoussystem or whether intermediate signals such as lung stretch,PaO2, or PaCO2 play a role. However, it appears that CSR-induced oscillations in HR and BP are dependent primarilyon periodic oscillations in ventilation, rather than episodichypoxia. Since fluctuations of PaCO2 above and below theapneic threshold (Naughton et al. 1993) are necessary forthe perpetuation of CSR, cardiovascular oscillations mightalso be linked to oscillations in PaCO2.

Ventricular response to atrial fibrillation

HR has long been known to vary with the phase of respi-ration in subjects in sinus rhythm (a phenomenon known asrespiratory sinus arrhythmia). However, the influence of res-piration on heart rate had been thought to be lost during AF.Since sinoatrial rate and AV nodal conduction are modu-lated by the autonomic nervous system in a similar fashion(Hayano et al. 1997; O’Toole et al. 1984; Wallick et al.1982), we reasoned that CSR might also be capable of or-ganizing the seemingly random ventricular response to AF.We therefore hypothesized that this breathing pattern wouldinfluence ventricular rate in AF by causing cyclical changesin the electrophysiological properties of the AV node.

It is important to realize that even under normal circum-stances, the ventricular response to AF is not completelyrandom. For example, the ventricular rate displays a circa-dian variation, being higher during the day and lower atnight (Hayano et al. 1998). These non-random alterations inrate have been attributed to modulation of the electrophysio-logical properties of the AV node, mediated by the auto-nomic nervous system (Hayano et al. 1997, 1998).

During AF, the ventricular rate is determined chiefly by

62 Can. J. Physiol. Pharmacol. Vol. 84, 2006

# 2006 NRC Canada

the AV refractory period and the degree of concealed AVconduction. The AV refractory period determines the short-est possible time between successive atrial impulses that canbe successfully transmitted to the ventricles (Toivonen et al.1990). Concealed AV conduction is the phenomenon where-by atrial impulses incompletely penetrate the AV junctionand do not reach the ventricles, but affect the transmissionof subsequent impulses (Langendorf et al. 1965; van denBerg et al. 1994). In turn, changes in AV node refractory pe-riod and concealed conduction are considered reflections ofalterations in autonomic input to the AV node (Hayano et al.1997, 1998; O’Toole et al. 1984; Wallick et al. 1982).

In 13 subjects with AF, systolic heart failure, and CSR,we observed that the mean ventricular rate was approxi-mately 5 bpm higher during the hyperpneic phase than theapneic phase of CSR (Fig. 2) (Leung et al. 2005). Again, byemploying spectral analysis techniques, we found that theventricular rate was highly coherent with periodic breathing,implying a strong relationship between the two. In contrast,there was no evidence of the normal respiratory sinus ar-rhythmia during periods of regular breathing. Comparisonof the distribution of RR intervals between hyperpnea andapnea was consistent with the concept that CSR caused al-terations in the electrophysiologic properties of the AVnode (Hayano et al. 1998), such that the AV refractory pe-riod was shorter and the degree of concealed conductionwas less during the hyperpneic phase of CSR. HR changeswere found to lag behind changes in respiration by 8 s, al-most identical to the 10 s lag observed between respirationand HR and BP changes in patients in sinus rhythm.

Ventricular ectopyVentricular premature beats (VPBs) are a risk factor for

arrhythmic death in patients with ischemic heart disease(Hartikainen et al. 1996). A number of previous reportshave established an association between CSR and ventricu-lar ectopy (Javaheri and Corbett 1998; Lanfranchi et al.2003), but a cause-effect relationship has yet to be defini-tively established. Notably, Javaheri (2000) observed thatthose patients whose CSR was alleviated by application ofcontinuous positive airway pressure (CPAP) also experi-enced a reduction in the frequency of ventricular ectopy.However, given that CPAP exerts hemodynamic effects in-dependent of its effects on breathing pattern, it remained un-clear whether CSR actually provokes ventricular ectopy.

If CSR indeed precipitates ventricular ectopy, we hy-pothesized that VPBs would occur more frequently duringepisodes of CSR than during periods of regular breathing inthe same patients. We further hypothesized that the fre-quency of VPBs during CSR would be greater during thehyperpneic phase when respiratory activation of the auto-nomic nervous system is at its peak and when its effects onHR, BP, and AV nodal conduction also manifest.

In 23 subjects with systolic heart failure, CSR, and fre-quent ventricular ectopy ( >30 VPBs), it was observed thatVPBs did not occur randomly through the CSR cycle(Leung et al. 2004). Instead, as hypothesized, VPBs weretwice as frequent during the hyperpneic phase than the ap-neic phase of CSR (Fig. 3). Furthermore, VPBs were overallmore frequent during periods of CSR than during periods of

regular breathing during the night. We also administeredsupplemental O2 sufficient to correct hypoxia in 4 of thesesubjects, with no accompanying reduction in the frequencyof VPBs. On the other hand, administration of CO2 suffi-cient to abolish CSR and normalize breathing resulted in asignificant reduction in VPB frequency.

Evidence that CSR is a cause of VPBs was previously

Fig. 1. Recording from a representative patient in sinus rhythm.The patient displays spontaneous Cheyne–Stokes respiration (CSR)consisting of the classic crescendo–decrescendo pattern of hyperp-nea alternating with periods of apnea. Oscillations in heart rate(HR) and blood pressure (BP) are clearly associated with the venti-latory oscillations, such that peaks in HR (i.e., troughs in RR inter-val) and BP occur during the hyperpnea and troughs in HR and BPduring the apnea. ILV, instantaneous lung volume.

Fig. 2. Recording from a subject with atrial fibrillation. Tremen-dous variability in RR intervals reflects the underlying chaoticrhythm. However, ventricular rate is higher (i.e., RR interval isshorter) during the hyperpnea of CSR and lower during the apnea.Moreover, there is less variability during the hyperpnea than apnea.Together, these findings suggest that CSR is influencing the elec-trophysiology of the atrio-ventricular node. ILV, instantaneous lungvolume.

Leung et al. 63

# 2006 NRC Canada

provided by Javaheri (2000), who demonstrated that the fre-quency of VPBs could be reduced by the overnight applica-tion of CPAP. However, by increasing intrathoracic pressure,CPAP exerts a number of direct effects on the heart that areunrelated to normalization of breathing. Direct effects ofCPAP that might reduce ventricular irritability include re-ductions in heart size, left ventricular afterload, and walltension (Kaye et al. 2001; Mehta et al. 2000; Naughtonand Bradley 1998; Tkacova and Bradley 2000).

On the other hand, our study showed that VPBs do notoccur randomly during the CSR breathing cycle, but areclustered during the hyperpneic phase. Such a temporal rela-tionship suggests that VPBs are not only associated withCSR, but are directly provoked by CSR. We also showedthat the frequency of VPBs was reduced by inhalation of alow concentration of CO2, an intervention that eliminatesCSR, but unlike CPAP, does not have direct effects on intra-thoracic pressure or cardiac size. Taken together, these find-ings provide further compelling evidence that CSR provokesventricular ectopy and sheds light on the likely mechanism.

CSR and the autonomic nervous systemIn this series of experiments, we have demonstrated that

CSR exerts cyclical effects on BP and on the electrophysiol-ogy of 3 different structures of the heart: the sinus node, theAV node, and the ventricles. These cyclical variations inHR, BP, AV nodal conduction, and ventricular ectopy mostlikely result from intermittent surges in central sympatheticoutflow phase-linked to oscillations in the central respiratorydrive (Franklin et al. 1997; Guyenet et al. 1993; van deBorne et al. 1998). Sympathetic and parasympathetic activ-ity have obvious effects on the rate of sinoatrial node dis-charge and hence HR, while the observed BP changes canbe explained by alterations in sympathetic vasomotor tone.It is also well known that the AV refractory period is short-

ened by sympathetic activity and lengthened by parasympa-thetic activity (Mazgalev et al. 1999; O’Toole et al. 1984;Toivonen et al. 1990; Wallick et al. 1982).

The advantage, under normal circumstances, of such co-activation of respiratory and cardiovascular autonomic sys-tems would be to match lung perfusion, through alterationsin HR and cardiac output, with ventilation to optimize gasexchange during changing metabolic demands. However,this interaction of the 2 systems might also lead to deleteri-ous effects during CSR, such as excessive sympathetic acti-vation and increased ventricular irritability in the setting of adiseased or ischemic myocardium. Indeed, CSR in patientswith heart failure is associated with increased sympatheticactivity (Naughton et al. 1995; van de Borne et al. 1998),which is in turn related to ventricular arrhythmias (Meredithet al. 1991).

There is evidence for direct connections between respira-tory and cardiovascular sympathetic neurons in the brain-stem. Activation of the respiratory neurons can co-activatethese sympathetic neurons (Guyenet et al. 1993). Severalstimuli, including chemostimulation, arousals, or voluntarycortical influences could contribute to linked cardiorespira-tory outflow. However, arousals have been to shown havelittle effect on BP during CSR (Trinder et al. 2000). Chemo-stimulation by hypoxia might also be expected to elicit sym-pathetic discharge, but our results indicate little role forhypoxia in generating these oscillations.

While changes in both sympathetic and parasympathetictone might be responsible for our findings, several lines ofevidence suggest a greater importance for sympathetic influ-ences. Our subjects had heart failure, which is associatedwith greatly diminished parasympathetic modulation of heartrate (Floras 1993). Indeed, the loss of the normal parasym-pathetic modulation of heart rate was the earliest autonomicabnormality described in both experimental and human heartfailure (Amorim et al. 1981; Binkley et al. 1991; Eaton et al.1995; Saul et al. 1988). Considering that respiratory modula-tion of both sympathetic and parasympathetic outflow hasbeen shown to be dependent on the level of pre-existingtone (Eckberg et al. 1988), the observed changes in AVproperties that correspond with oscillations in BP are morelikely to be the result of sympathetic than of parasympa-thetic modulation. However, the larger tidal volumes accom-panying CSR might conceivably engage parasympatheticreflexes (Taha et al. 1995) not active during regular breath-ing in these patients. Further evidence for a sympathetic,rather than parasympathetic, mechanism for our findingscomes from our observation of synchronous blood pressureoscillations, because apnea-induced increases in BP are dueto increases in sympathetic outflow (Katragadda et al. 1997).

Another way in which we might distinguish between thesympathetic and parasympathetic nervous systems is to ex-amine the latency of the responses. The oscillations in meanHR and BP lagged the changes in respiration by approxi-mately 8–10 s. Thus, there appears to be a relatively longtime constant between the generation of central respiratorydrive, ventilation, and the subsequent ventricular rate andvasomotor responses. This 8–10 s delay is again more inline with the behavior of the sympathetic nervous systemrather than the parasympathetic nervous system, which hasa shorter time constant (Pomeranz et al. 1985).

Fig. 3. Recording from a subject with CSR and frequent ventricu-lar ectopy. Ventricular premature beats (VPBs) occur more fre-quently during the hyperpneic phase of CSR and are relativelyabsent during the apneic phase. Oxygen saturation did not fall be-low 90% during the period of recording. ILV, instantaneous lungvolume; TcCO2, transcutaneous PCO2.

64 Can. J. Physiol. Pharmacol. Vol. 84, 2006

# 2006 NRC Canada

Alterations in blood gases might be a necessary inter-mediate step by which CSR exerts its cyclical effects on theautonomic nervous system. The sympatho-excitatory effectsof CO2 are likely to be particularly relevant in patients withcongestive heart failure (CHF) and CSR, in whom there isevidence of an enhanced muscle sympathetic nerve responseto CO2 (Narkiewicz et al. 1999). Dips in oxygen saturationalso accompany CSR, and because of the circulatory delay,such dips are maximal in the middle of the hyperpnea andshould cause stimulation of sympathetic activity at thattime. However, oscillations in HR and BP and provocationof VPBs occur during periodic breathing even in the absenceof hypoxia.

In conclusion, we have shown that CSR exerts a numberof acute cardiovascular effects, including fluctuations in HRand BP, alterations in AV conduction, and facilitation ofventricular ectopy. These effects are not explainable by dipsin SaO2 alone and appear to be manifestations of respiratorymodulation of the autonomic nervous system. We speculatethat the intense respiratory drive underlying the hyperpneicphase of CSR results in enhanced cardiac and sympatheticoutflow, which is responsible for the effects observed. Oscil-lations in autonomic tone brought about by CSR may ulti-mately contribute to the sympatho-excitation and increasedmortality long observed in patients with heart failure andCSR.

Acknowledgements

This work was supported by the Canadian Institutes ofHealth Research (grants MOT 11607 and UI 14909). R.S.T.Leung is the recipient of a Canadian Institutes of Health Re-search Clinician Scientist Phase II Award, and T.D. Bradleyis supported by a Canadian Institutes of Health Research Se-nior Scientist Award.

ReferencesAmorim, D.S., Dargie, H.J., Heer, K., Brown, M., Jenner, D., Ol-

sen, E.G., Richardson, P., and Goodwin, J.F. 1981. Is there auto-nomic impairment in congestive (dilated) cardiomyopathy?Lancet, 1: 525–527. PMID: 6111634.

Binkley, P.F., Nunziata, E., Haas, G.J., Nelson, S.D., and Cody,R.J. 1991. Parasympathetic withdrawal is an integral componentof autonomic imbalance in congestive heart failure: demonstra-tion in human subjects and verification in a paced canine modelof ventricular failure. J. Am. Coll. Cardiol. 18: 464–472. PMID:1856414.

Eaton, G.M., Cody, R.J., Nunziata, E., and Binkley, P.F. 1995.Early left ventricular dysfunction elicits activation of sympa-thetic drive and attenuation of parasympathetic tone in the pacedcanine model of congestive heart failure. Circulation, 92: 555–561. PMID: 7634470.

Eckberg, D.L., Rea, R.F., Andersson, O.K., Hedner, T., Pernow, J.,Lundberg, J.M., and Wallin, B.G. 1988. Baroreflex modulationof sympathetic activity and sympathetic neurotransmitters in hu-mans. Acta Physiol. Scand. 133: 221–231. PMID: 3227916.

Floras, J.S. 1993. Clinical aspects of sympathetic activation andparasympathetic withdrawal in heart failure. J. Am. Coll. Car-diol. 22: 72A–84A. PMID: 8376699.

Franklin, K.A., Sandstrom, E., Johansson, G., and Balfors, E.M.1997. Hemodynamics, cerebral circulation, and oxygen satura-

tion in Cheyne-Stokes respiration. J. Appl. Physiol. 83: 1184–1191. PMID: 9338427.

Guyenet, P.G., Koshiya, N., Huangfu, D., Verberne, A.J., and Ri-ley, T.A. 1993. Central respiratory control of A5 and A6 pontinenoradrenergic neurons. Am. J. Physiol. 264: R1035–R1044.PMID: 8322954.

Hanly, P.J., and Zuberi-Khokhar, N.S. 1996. Increased mortalityassociated with Cheyne-Stokes respiration in patients with con-gestive heart failure. Am. J. Respir. Crit. Care Med. 153: 272–276. PMID: 8542128.

Hartikainen, J.E., Malik, M., Staunton, A., Poloniecki, J., andCamm, A.J. 1996. Distinction between arrhythmic and nonar-rhythmic death after acute myocardial infarction based on heartrate variability, signal-averaged electrocardiogram, ventriculararrhythmias and left ventricular ejection fraction. J. Am. Coll.Cardiol. 28: 296–304. doi: 10.1016/0735-1097(96)00169-6.PMID: 8800101.

Hayano, J., Yamasaki, F., Sakata, S., Okada, A., Mukai, S., and Fu-jinami, T. 1997. Spectral characteristics of ventricular responseto atrial fibrillation. Am. J. Physiol. 273: H2811–H2816. PMID:9435618.

Hayano, J., Sakata, S., Okada, A., Mukai, S., and Fujinami, T.1998. Circadian rhythms of atrioventricular conduction proper-ties in chronic atrial fibrillation with and without heart failure.J. Am. Coll. Cardiol. 31: 158–166. doi: 10.1016/S0735-1097(97)00429-4. PMID: 9426035.

Javaheri, S. 2000. Effects of continuous positive airway pressure onsleep apnea and ventricular irritability in patients with heart fail-ure. Circulation, 101: 392–397. PMID: 10653830.

Javaheri, S., and Corbett, W.S. 1998. Association of low PaCO2with central sleep apnea and ventricular arrhythmias in ambula-tory patients with stable heart failure. Ann. Intern. Med. 128:204–207. PMID: 9454528.

Javaheri, S., Parker, T.J., Liming, J.D., Corbett, W.S., Nishiyama,H., Wexler, L., and Roselle, G.A. 1998. Sleep apnea in 81 am-bulatory male patients with stable heart failure. Types and theirprevalences, consequences, and presentations. Circulation, 97:2154–2159. PMID: 9626176.

Katragadda, S., Xie, A., Puleo, D., Skatrud, J.B., and Morgan, B.J.1997. Neural mechanism of the pressor response to obstructiveand nonobstructive apnea. J. Appl. Physiol. 83: 2048–2054.PMID: 9390980.

Kaye, D.M., Mansfield, D., Aggarwal, A., Naughton, M.T., and Es-ler, M.D. 2001. Acute effects of continuous positive airwaypressure on cardiac sympathetic tone in congestive heart failure.Circulation, 103: 2336–2338. PMID: 11352880.

Lanfranchi, P.A., Braghiroli, A., Bosimini, E., Mazzuero, G., Co-lombo, R., Donner, C.F., and Giannuzzi, P. 1999. Prognostic va-lue of nocturnal Cheyne-Stokes respiration in chronic heartfailure. Circulation, 99: 1435–1440. PMID: 10086966.

Lanfranchi, P.A., Somers, V.K., Braghiroli, A., Corra, U., Eleuteri,E., and Giannuzzi, P. 2003. Central sleep apnea in left ventricu-lar dysfunction: prevalence and implications for arrhythmic risk.Circulation, 107: 727–732. doi: 10.1161/01.CIR.0000049641.11675.EE. PMID: 12578876.

Langendorf, R., Pick, A., Edelist, A., and Katz, L.N. 1965. Experi-mental demonstration of concealed AV conduction in the humanheart. Circulation, 32: 386–393. PMID: 5830603.

Leung, R.S., Floras, J.S., Lorenzi-Filho, G., Rankin, F., Picton, P.,and Bradley, T.D. 2003. Influence of Cheyne-Stokes respirationon cardiovascular oscillations in heart failure. Am. J. Respir.Crit. Care Med. 167: 1534–1539. doi: 10.1164/rccm.200208-793OC. PMID: 12626347.

Leung, R.S., Diep, T.M., Bowman, M.E., Lorenzi-Filho, G., and

Leung et al. 65

# 2006 NRC Canada

Bradley, T.D. 2004. Provocation of ventricular ectopy byCheyne-Stokes respiration in patients with heart failure. Sleep,27: 1337–1343. PMID: 15586786.

Leung, R.S., Bowman, M.E., Diep, T.M., Lorenzi-Filho, G., Floras,J.S., and Bradley, T.D. 2005. Influence of Cheyne-Stokes re-spiration on ventricular response to atrial fibrillation. J. Appl.Physiol. 99: 1689–1696. doi: 10.1152/japplphysiol.00027.2005.PMID: 15994246.

Lorenzi-Filho, G., Dajani, H.R., Leung, R.S., Floras, J.S., andBradley, T.D. 1999a. Entrainment of blood pressure and heartrate oscillations by periodic breathing. Am. J. Respir. Crit. CareMed. 159: 1147–1154. PMID: 10194159.

Lorenzi-Filho, G., Rankin, F., Bies, I., and Douglas Bradley, T.1999b. Effects of inhaled carbon dioxide and oxygen on cheyne-stokes respiration in patients with heart failure. Am. J. Respir.Crit. Care Med. 159: 1490–1498. PMID: 10228116.

Mazgalev, T.N., Garrigue, S., Mowrey, K.A., Yamanouchi, Y., andTchou, P.J. 1999. Autonomic modification of the atrioventricularnode during atrial fibrillation: role in the slowing of ventricularrate. Circulation, 99: 2806–2814. PMID: 10351976.

Mehta, S., Liu, P.P., Fitzgerald, F.S., Allidina, Y.K., and DouglasBradley, T. 2000. Effects of continuous positive airway pressureon cardiac volumes in patients with ischemic and dilated cardio-myopathy. Am. J. Respir. Crit. Care Med. 161: 128–134. PMID:10619809.

Meredith, I.T., Broughton, A., Jennings, G.L., and Esler, M.D.1991. Evidence of a selective increase in cardiac sympatheticactivity in patients with sustained ventricular arrhythmias. N.Engl. J. Med. 325: 618–624. PMID: 1861695.

Narkiewicz, K., Pesek, C.A., van de Borne, P.J., Kato, M., andSomers, V.K. 1999. Enhanced sympathetic and ventilatory re-sponses to central chemoreflex activation in heart failure. Circu-lation, 100: 262–267. PMID: 10411850.

Naughton, M.T., and Bradley, T.D. 1998. Sleep apnea in conges-tive heart failure. Clin. Chest Med. 19: 99–113. doi: 10.1016/S0272-5231(05)70435-4. PMID: 9554221.

Naughton, M., Benard, D., Tam, A., Rutherford, R., and Bradley,T.D. 1993. Role of hyperventilation in the pathogenesis of cen-tral sleep apneas in patients with congestive heart failure. Am.Rev. Respir. Dis. 148: 330–338. PMID: 8342895.

Naughton, M.T., Benard, D.C., Liu, P.P., Rutherford, R., Rankin,F., and Bradley, T.D. 1995. Effects of nasal CPAP on sympa-thetic activity in patients with heart failure and central sleep ap-nea. Am. J. Respir. Crit. Care Med. 152: 473–479. PMID:7633695.

O’Toole, M.F., Wurster, R.D., Phillips, J.G., and Randall, W.C.1984. Parallel baroreceptor control of sinoatrial rate and atrio-ventricular conduction. Am. J. Physiol. 246: H149–H153.PMID: 6696125.

Pomeranz, B., Macaulay, R.J., Caudill, M.A., Kutz, I., Adam, D.,

Gordon, D., et al. 1985. Assessment of autonomic function inhumans by heart rate spectral analysis. Am. J. Physiol. 248:H151–H153. PMID: 3970172.

Saul, J.P., Arai, Y., Berger, R.D., Lilly, L.S., Colucci, W.S., andCohen, R.J. 1988. Assessment of autonomic regulation inchronic congestive heart failure by heart rate spectral analysis.Am. J. Cardiol. 61: 1292–1299. doi: 10.1016/0002-9149(88)91172-1. PMID: 3376889.

Sin, D.D., Fitzgerald, F., Parker, J.D., Newton, G., Floras, J.S., andBradley, T.D. 1999. Risk factors for central and obstructivesleep apnea in 450 men and women with congestive heart fail-ure. Am. J. Respir. Crit. Care Med. 160: 1101–1106. PMID:10508793.

Sin, D.D., Logan, A.G., Fitzgerald, F.S., Liu, P.P., and Bradley,T.D. 2000. Effects of continuous positive airway pressure oncardiovascular outcomes in heart failure patients with and with-out Cheyne-Stokes respiration. Circulation, 102: 61–66. PMID:10880416.

Taha, B.H., Simon, P.M., Dempsey, J.A., Skatrud, J.B., and Iber,C. 1995. Respiratory sinus arrhythmia in humans: an obligatoryrole for vagal feedback from the lungs. J. Appl. Physiol. 78:638–645. PMID: 7759434.

Tkacova, R., and Bradley, T.D. 2000. Therapy of obstructive andcentral sleep apnea in patients with congestive heart failure. InSleep apnea: implications in cardiovascular and cerebrovasculardisease. Edited by T.D. Bradley and J.S. Floras. Marcel Dekker,New York. pp. 461–494.

Toivonen, L., Kadish, A., Kou, W., and Morady, F. 1990. Determi-nants of the ventricular rate during atrial fibrillation. J. Am.Coll. Cardiol. 16: 1194–1200. PMID: 2229766.

Trinder, J., Merson, R., Rosenberg, J.I., Fitzgerald, F., Kleiman, J.,and Douglas Bradley, T. 2000. Pathophysiological interactionsof ventilation, arousals, and blood pressure oscillations duringcheyne-stokes respiration in patients with heart failure. Am. J.Respir. Crit. Care Med. 162: 808–813. PMID: 10988087.

van de Borne, P., Oren, R., Abouassaly, C., Anderson, E., andSomers, V.K. 1998. Effect of Cheyne-Stokes respiration on mus-cle sympathetic nerve activity in severe congestive heart failuresecondary to ischemic or idiopathic dilated cardiomyopathy.Am. J. Cardiol. 81: 432–436. doi: 10.1016/S0002-9149(97)00936-3. PMID: 9485132.

van den Berg, M.P., Crijns, H.J., Haaksma, J., Brouwer, J., and Lie,K.I. 1994. Analysis of vagal effects on ventricular rhythm in pa-tients with atrial fibrillation. Clin. Sci. (Lond.), 86: 531–535.PMID: 8033507.

Wallick, D.W., Martin, P.J., Masuda, Y., and Levy, M.N. 1982. Ef-fects of autonomic activity and changes in heart rate on atrio-ventricular conduction. Am. J. Physiol. 243: H523–H527.PMID: 7124960.

66 Can. J. Physiol. Pharmacol. Vol. 84, 2006

# 2006 NRC Canada