318
T. Penzel et al.
the greatest differences arising between deep sleep on the one hand—with virtually
uncorrelated beat-to-beat regulation—and REM sleep on the other, with extensively
correlated beat-to-beat regulation of the heart rate.
20.4 Cyclical Variation of Heart Rate with Sleep Apnea
Very shortly after Guilleminault described sleep-related breathing disorders, it
became apparent that characteristic alterations in heart rate take place during obstructive apnea events [14]. This phenomenon was described as cyclical variation in heart
rate, and the proposal arose to utilize this characteristic pattern for diagnosis of sleep
apnea [14]. Many diagnostic devices were developed to detect and diagnose sleep
apnea outside the context of a sleep lab utilizing this heart-rate pattern [25].
During each individual apnea phase, relative bradycardia becomes apparent, with
relative tachycardia occurring during each subsequent increased respiratory activity.
These patterns in heart rate—arising from alterations in sympathetic nervous activity
during each apnea event—are, in their temporal course, directly linked to breathing.
As a result, it is possible to count the number of apnea events based on heart rate
data alone. For example, the MESAM system and its successors, as well as new
polygraph devices, employ heart rate for diagnosis of sleep apnea in an outpatient
setting [25].
Since the cyclical variation in heart rate presents an impressive periodic pattern,
it became obvious to apply methods of frequency and spectral power analysis once
again to quantitatively assess cyclical variations in heart rate—and thus, potentially
determine the degree of severity of sleep-related breathing disorders. To this end,
various teams have employed the method of Fourier analysis, however, limitations
arose in the attempt to evaluate the periodic patterns automatically [17]. This is
because the cyclical variations are not strictly periodic since apneas and hypopneas
show large variation in duration especially when occurring in various sleep stages.
Period analyses in the time domain or by classical spectral methods quickly reach
their limits. The cyclical variation in heart rate remains dependent to greater or
lesser degree on physical training condition, age, weight, and concomitant diseases
(e.g., diabetes). Often, there are individual characteristic patterns of bradycardia and
tachycardia that are influenced by several factors such as concomitant cardiologic
diseases, arrhythmia, pacemaker ECGs, and heart failure—which make them difficult
to be interpreted. For these reasons, a fully automated assessment of sleep-related
breathing disorders cannot take place reliably if it is based on cyclical heart-rate
variation alone. Determination of oxygen saturation, changes in ECG morphology
and more direct recordings of breathing disorders are essential for greater reliability.
T. Penzel et al.
the greatest differences arising between deep sleep on the one hand—with virtually
uncorrelated beat-to-beat regulation—and REM sleep on the other, with extensively
correlated beat-to-beat regulation of the heart rate.
20.4 Cyclical Variation of Heart Rate with Sleep Apnea
Very shortly after Guilleminault described sleep-related breathing disorders, it
became apparent that characteristic alterations in heart rate take place during obstructive apnea events [14]. This phenomenon was described as cyclical variation in heart
rate, and the proposal arose to utilize this characteristic pattern for diagnosis of sleep
apnea [14]. Many diagnostic devices were developed to detect and diagnose sleep
apnea outside the context of a sleep lab utilizing this heart-rate pattern [25].
During each individual apnea phase, relative bradycardia becomes apparent, with
relative tachycardia occurring during each subsequent increased respiratory activity.
These patterns in heart rate—arising from alterations in sympathetic nervous activity
during each apnea event—are, in their temporal course, directly linked to breathing.
As a result, it is possible to count the number of apnea events based on heart rate
data alone. For example, the MESAM system and its successors, as well as new
polygraph devices, employ heart rate for diagnosis of sleep apnea in an outpatient
setting [25].
Since the cyclical variation in heart rate presents an impressive periodic pattern,
it became obvious to apply methods of frequency and spectral power analysis once
again to quantitatively assess cyclical variations in heart rate—and thus, potentially
determine the degree of severity of sleep-related breathing disorders. To this end,
various teams have employed the method of Fourier analysis, however, limitations
arose in the attempt to evaluate the periodic patterns automatically [17]. This is
because the cyclical variations are not strictly periodic since apneas and hypopneas
show large variation in duration especially when occurring in various sleep stages.
Period analyses in the time domain or by classical spectral methods quickly reach
their limits. The cyclical variation in heart rate remains dependent to greater or
lesser degree on physical training condition, age, weight, and concomitant diseases
(e.g., diabetes). Often, there are individual characteristic patterns of bradycardia and
tachycardia that are influenced by several factors such as concomitant cardiologic
diseases, arrhythmia, pacemaker ECGs, and heart failure—which make them difficult
to be interpreted. For these reasons, a fully automated assessment of sleep-related
breathing disorders cannot take place reliably if it is based on cyclical heart-rate
variation alone. Determination of oxygen saturation, changes in ECG morphology
and more direct recordings of breathing disorders are essential for greater reliability.