Quantitative Models of Periodic Breathing and Cheyne–Stokes Respiration
I. Introduction Cheyne-Stokes respiration (CSR) is an exaggerated form of periodic breathing (PB) in which the ventilatory pattern displays a cyclic variation between periods of hyperpnea and periods of apnea or hypopnea. The classic pattern of CSR is one in which the apneic phase is followed by breaths that gradually wax and subsequently wane in amplitude until the next period of apnea occurs. CSR has long been associated with both congestive heart failure (CHF) and neurologic disease (1-3). As such, there has been a long-running controversy on whether the pathogenetic mechanisms underlying CSR are primarily neurological or cardiovascular in origin (3). Guyton et al. (4) induced CSR in anesthetized dogs by artificially lengthening the carotid arteries, thereby prolonging lung-tobrain circulation time (CT). Since it was known that patients with CHF commonly have prolonged CTs (5), these workers concluded that their results supported a cardiovascular mechanism for CSR. On the other hand, careful analysis of their results shows that only one-third of the experimental preparations developed CSR in spite of extremely long circulation times that ranged from two to five minutes. Furthermore, with such drastic interventions, neurological damage in some of the preparations could not be ruled out. In support of the neurological mechanism, Brown and Plum (6) found that 5 of the 28 patients with CSR that they studied were free of heart disease, although all had some kind of neurological lesion. On the other hand, numerous reports have indicated that PB can also appear in healthy subjects under a variety of circumstances. For instance, it is now well known that PB is a common occurrence in normal sojourners to high altitude, particularly when these subjects are asleep (7-11). This condition disappears when the subjects return to sea level. Douglas and Haldane (12) found that PB could be induced transiently following forced hyperventilation for about two minutes or during administration of a hypoxic gas mixture. PB is also commonly observed during sleep onset and in the light stages of sleep (13,14). Furthermore, PB frequently occurs in apparently healthy infants in the first few months of life but becomes uncommon thereafter (15,16).