Dynamic analysis of autonomic nervous system

Research Paper Clinical Autonomic Research 4, 141-148 (1994)

THE dynamics and relationship o f s low cortical activity (am- plitude modulat ion o f e lectroencephalograph at 0.02--0.05 Hz) and cardiovascular fluctuations (0.01-0.05 Hz range) was studied by t ime-frequency mapping (modif ied Wigner distri- bution) in a patient wi th Alzheimer's disease. The amplitude modulating at rest basal alpha and them activity (lead Oz) was exaggerated compared wi th control subjects. Fluctuations at respiratory and nonrespiratory (0.01-0.05 Hz) frequencies in R-R intervals and blood pressure were present and within normal range. Spontaneous hypotens ion (by 20 to 80 mmHg lasting 15-20 s) accompanied by cardioacceleration occurred repeatedly in the supine pos i t ion and during hyperventila- tion. Slow cortical activity and nonrespiratory fluctuations (0.01-0.05 Hz ) in blood pressure increased concurrently wi th hypotens ions .

No signs of peripheral autonomic dysfunction or malfunction of baroreceptors were observed and the vasomotor instability appeared to be o f central origin. The results suggested that s low cortical activity was functional ly related to the central autonomic nervous system and reflected cortico-thalamo- brainstem interaction.

Key words: Slow brain activity, Alzheimer's disease, Autonomic nervous system, Electroencephalography, Blood pressure, Wigner distribution

Time-frequency analysis of slow cortical activity and cardio- vascular fluctuations in a case of Alzheimer's disease

Peter Novak 1, Vera Novak 1,cA, Zhixing Li 2 and Guy Remillard a

1Lady Davis Institute for Medical Research and Jewish General SBMD Hospital, 3755 Chemin de la C6te-Ste-Catherine, McGill University, Montreal, H3T 1E2; 2Research Center, Hospital Notre Dame, and aDepartment of Neurology, Sacre Coeur Hospital, Montreal, Quebec, Canada.

CACorresponding Author

Introduction

The central integration of autonomic nervous sys- tem (ANS) functions is accomplished by an interac- tion of many brainstem, mesencephalic and cortical regions (prefrontal cortex, forebrain, amygdala, insula and hypothalamus)? Coordination of sympa- thetic and vagal outflow to cardiovascular and respi- ratory motoneurones is controlled by brainstem net- work through rather complex feedback loops. Time and frequency domain analysis have recently shown that ANS activity can be determined from 0.01-0.5 Hz rhythms in heart rate and blood pressureY The baseline 2-6 Hz rhythm, which underlines the tonic activity of sympathetic preganglionic neurones is transformed in a quasiperiodic pattern to all post- ganglionic sympathetic nerves. It is believed that this activity is representative of synaptic transmission and forms a basis for different sympathetic spinal rhythms. <5 Slower rhythms in efferent sympathetic activity (in rats (0.15-0.6 Hz) were found to correlate well with blood pressure oscillations. 6 By analogy, similar rhythms in humans and dogs would be the rhythms nonrespiratory frequencies (NONRF) (0.01-0.05 Hz) in blood pressure. 7,8

Oscillations in R-R intervals at respiratory frequen- cies (RF) reflect over 0.01-0.5 Hz the parasympathetic activity 9,m and are proportional to the frequency and respiratory cycle length. Amplitude of R-R interval fluctuations at RF can be significantly reduced by a centrally acting muscarinic blocker" and therefore it was proposed as an index central vagal activity.

Parasympathetic and sympathetic efferent activity participate on spectral powers at lower frequencies, 12 as it was suggested by pharmacological studies.

Experiments in cats showed that respiratory rhythms propagates both caudally to cardiomotor neurones and sympathetic neurones, but also to upper regions and neocortex prefrontal cortex and limbic system23,~4 Occasional synchrony between respiration and 0.1 Hz modulation of spectral powers in alpha band in humans was described by Pfurtscheller. 15 This pilot observation was recently confirmed using a modern approach of dynamic spectral analysis. Modulation of electroencephalo- graph (EEG) at low frequencies was observed in healthy subjects during waking state in the 0.015-0.028 Hz in 98%, 0.04-0.07 Hz in 95% and in the respiratory 0.09-0.15 Hz in 65% cases in alpha and theta bands) 6 The magnitude of slow EEG modulation increased during deactivation after hy- perventilation and correlated well with an aug- mented amplitude basal EEG in both bands. ~7,~8 Low frequency R-R interval rhythms at frequencies similar to slow EEG rhythms were found in 60% of these subjects (unpublished observations). The aetiology of rhythmic modulation of EEG activity is not precisely known. It was suggested that these rhythms reflect activity of central ANS functions and originate in brainstem neural network.

In this study we present a patient with Alzheimer's disease in whom spontaneous hypotensions in su- pine position were observed during routine auto-

© Rapid Communications of Oxford Ltd. Clinical Autonomic Research • vol 4 • 1994 141

P. Novak et al.

nomic nervous system testing. This phenomenon has, to our knowledge, not been previously re- ported. Time-frequency mapping a method for si- multaneous evaluation of EEG and peripheral cardio- vascular rhythms is used as a tool to noninvasively study the relationship between slow modulation of cortical activity and peripheral ANS rhythms. The dynamic approach used in this study enables more informative cross-analysis between the EEG and blood pressure.

Results

Case report: The 54-year-old man with Alzheimer's disease underwent evaluation in our laboratory for EEG and autonomic functions. The diagnosis was based on a two-year progressive worsening of memory and intellect associated with aphaso-apraxo- gnosia. Computer tomography (CT), magnetic reso- nance imaging (MRI) scans and SPECT confirmed diffuse cortical and subcortical atrophy suspected from mildly abnormal EEG findings. The patient had no abnormality of the cranial nerves and had no evidence of peripheral neuropathy. The vibratory sensitivity and tendon reflexes at all sites, including the Achilles tendon reflex, were normal. The respira- tory and cardiovascular examinations were within normal limits. Repeated electrocardiograph (ECG) including Holter monitoring, routine blood and urine investigations including blood profile, glucose, thy- roid hormones, B12 and folic acid, were within normal limits.

Methods

Protocol: Resting recordings for 10 min in supine position with eyes closed was followed by 4 min of voluntary hyperventilation. This protocol was re- peated after 6 months.

Data acquisition and analysis: EEG (Neuroscience, Brain Imager IIs), ECG (lead II), beat-to-beat blood pressure from the third finger (Finapres Ohmeda) and respiration (nasal thermistor) were recorded simultaneously. The EEG was recorded using an EEG mapping system (Brain Imager IIs, Neuroscience Ltd). Monopolar recordings referenced to linked ears are made using the standard ten-twenty system of electrode placement. The impedance of Ag-AgC1 electrodes is kept below 4.7 kOhm. The EEG data are filtered (bandpass 0.3-40 Hz), digitized at a sampling frequency of 200 Hz and are written to an optical disk. R-R intervals (consecutive QRS complexes) and corresponding systolic and diastolic pressure values were off-line extracted from the original signals. R-R intervals, systolic, diastolic pressures and respiration were equidistantly resampled at 1-s interVals using

142 Clinical Autonomic Research • vol 4. 1994

linear interpolation algorithm. The trend at frequen- cies below 0.005 Hz was removed using moving polynomial. This nonlinear filtering has preserve near DC level frequencies without the distortion common in linear filtering.

The EEG signal from midline occipital area (Oz lead) was analysed as follows. Dominant EEG activ- ity was obtained as the square root of peak power within the band 2.8-12 Hz (theta and alpha band) on the Wigner distribution for each 100 ms time instant. The subsequent ten values of the dominant EEG activity were averaged resulting in time resolution one second. This process can be viewed as time- varying filtering of EEG signal with adaptive bandpass for the dominant frequency of EEG. 16J7,19

Time-frequency mapping 19 is based on the modi- fied Wigner distribution which decomposes the sig- nal as a function of time into the function of time and frequency. Thus both time and frequency values are simultaneously available. This high-resolution method was applied for both cardiovascular and EEG signals and described in detail in previous studies? In these studies, time-frequency mapping was shown to give a reliable estimate in nonstationary time-series. The Wigner distributions were computed with the same parameters applied uniformly for dominant EEG activity, cardiovascular and respira- tory signals (n = 128, M = 8, Gaussian window with alpha = 2.8 in the equation in Ref. 19). Cross Wigner distributions between blood pressure and EEG sig- nals were computed to evaluate common spectral content and time profile of slow rhythms. Data are expressed as mean _+ SE.

Results

Fluctuations in R-R intervals and blood pressure at RF (R-R intervals 2392 _+ 1391.7 ms2/Hz, systolic pres- sure 9.9 + 3.1 mmHg2/Hz, diastolic pressure 1.0_+0.07 mmI-Ig2/I-Iz) and at NONRF (0.01-0.05 Hz) (R-R intervals 362.88 _+ 5.8 ms2/Hz, systolic pressure 91.1 _+ 12.2 mmHg2/Hz and diastolic pressure 30.9_+7.8 mmHg~/Hz) were continuously present and within normal limits compared with our refer- ence data (Table 1), 9 only fluctuations in NONRF in R-R intervals seemed attenuated. Mean R-R interval 1124 _+7 ms, systolic and diastolic pressures 134 -+ 3 and 65 _+ 0.75 mmHg) were within normal limits.

Table 1. Reference values for spectral powers in cardiovascular signals at rest from WD 9

NONRF RF

R-R intervals (ms 2 Hz) 2736 + 47.6 2293 _+ 23.5 SBP (mmHg2/Hz) 34.3 + 0.9 7.7 ± 0.1 DBP (mmHg2/Hz) 10.3 ± 0.6 2.9 ± 0.2

n=15. NONRF = nonrespiratory frequencies; RF = respiratory frequen- cies.

Dynamic analysis of autonomic nervous system

Three sequential episodes of hypotensions (SBP decreased by 24, 22 and 82 mmHg, DBP decreased by 18, 20 and 68 mmHg) occurred spontaneously during the resting recording (Fig. 1). Hypotension was characterized by an initial increase of blood pressure over nearby values, rapid decline within 4, 9 and 22 s and recovery. Blood pressure drops were compensated for a moderate cardioacceleration. The third episode lasted 40 s and blood pressure de- creased to 48 mmHg in systolic and to 16 mmHg in diastolic pressure, while R-R intervals shortened to 980 ms. Fluctuations at RF in R-R intervals (= 0.24 Hz) diminished rapidly with the beginning of the hypo- tensive reactions and remained at minimal ampli- tudes until blood pressure recovery (Fig. 2A and B). Respiratory frequency and amplitude remained un- changed. The NONRF (0.01-0.05 Hz) fluctuations in both systolic and diastolic pressure blood pressure sharply increased to the maximum concomitant with blood pressure minimum (Fig. 3B). Similarly, oscilla- tions at the NONRF in R-R intervals increased in parallel with blood pressure falls.

Basic EEG activity showed diffuse occurrence of theta waves 6-7 Hz and short-lasting bilaterally syn- chronous bursts across temporoparietal regions. Ba- sal rhythms was alpha 8-10 Hz, reactive to eyes opening. EEG findings suggested mild cortical dys- function with abnormality in pons or thalamus. Al- pha and theta EEG (2.8-12 Hz) activity from Oz lead was further evaluated for presence of slow modula- tion (0.01-0.07 Hz) rhythms. The Wigner distribution from the resting period confirmed the presence of slow EEG modulation at frequencies of 0.02 Hz and 0.05 Hz (Fig. 3A). The basal values of this slow cortical activity were comparable with those in con- trol subjects (20-40 years old) from our reference value# 7 (see Table 2).

Cross Wigner distribution between EEG-Oz and systolic pressure were computed to evaluate the temporal profile of slow EEG activity and NONRF oscillations in blood pressure (Fig. 3C). Time-frequency maps showed that the slow cortical activity peaked with the falls in blood pressure. With blood pressure recovery following the first hypoten- sion, its magnitude declined to control values. How- ever, the amplitude of slow cortical activity persisted, was elevated after the second hypotension and reached its maximum at the beginning of the third and most pronounced hypotension.

Table 2. Reference values for global spectral powers in EEG 17

Peak 1 Peak 2

Frequency Power Frequency Power (Hz) (uV2/Hz) (Hz) (uV2/Hz)

Theta (Cz) 0.02 ± 0.0004 3.2 ± 2.8 0.069 ± 0,0004 2.87 +_ 2.11 Alpha (Oz) 0.02 ± 0.0002 3.7 +_ 1.4 0.066 _+ 0.0005 3.25 +_ 2.15

5.2 EEG - 0z

4 . 4

:~ 3.6

2.8

2.0

1500 R - R interval

1 2 0 0

~ 1100

1 0 0 0

9 0 O

2" 14o ~ ~ E 100 E

60

20

7O

E 4O I Diost. pressure

25

I0

2500 Respiration

2 5 0 0

E = 21 0 0

1 9 0 0

I ~ i i i i I 1700 0 100 200 500 400 500 600

T i m e ( s )

FIG.1. Signals of EEG theta activity (lead Qz), R-R intervals, systolic and diastolic pressure and respiratory signals at rest during the first visit. Arrows indicate spontaneous hypotensions.

During the second recording (rest and hyperventi- lation) hypotensive events with cardioacceleration and parallel increases of slow cortical activity were also present and become more frequent and more severe (Fig. 4A and B). Hypotension occurred in intervals of between 50 and 170 s. They were ob- served during spontaneous respiration and persisted during and after hyperventilation. Corresponding Wigner distributions confirmed again simultaneous increases of slow cortical activity and NONRF blood pressure rhythms (Fig. 4C and D). The patient did not experience nausea or dizziness or agitation; however a significant progressive worsening of cognitive functions was noted.

Clinical Autonomic Research. vol 4. 1994 143

P. Novak et al.

%- 212 o

A

o

o • 0

~0 0

2 " *

0 cu

interval

FIG. 2. Detail of Wigner distributions of the R-R intervals (A) and respiratory movement (B) (signals from Fig. 1) cen- tred at respiratory frequencies. Note a decrease of vagally mediated respiratory fluctuations in R-R intervals = 0.24 Hz in coincidence with brood pressure drops.

Discussion

In this study, spontaneous hypotensions in supine position concomitant with augmentation of the slow cortical activity (modulation of basal theta rhythm at 0,01-0.02 Hz) were described in a patient with Alzheimer's disease. Follow-up of this patient (after 6 months) showed progression in frequency and severity of hypotensions together with progressive cognitive impairment. Since no abnormalities of pe- ripheral baroreceptor functions were observed, data indicate that repetitive spontaneous hypotensions suggest progressive deficit of central ANS functions.

The loss of cholinergic neurones in Alzheimer's disease has been receiving considerable attention over past years and was considered as a causal factor for cortical atrophy and loss of cognitive functions. 2° In contrast, little is known about the involvement of central or peripheral ANS functions in Alzheimer's disease. Recently a scatter report evaluating the R-R interval power spectra in elderly controls and Alzheimer's disease patients appeared and data showed significantly higher spectral powers at slow frequencies (around 0.03 Hz) than at respiratory frequencies in Alzheimer's disease patients. We inter- preted these data as an increase of sympathetic

efferent tone secondary to a loss of cholinergic func- tion with involvement of peripheral cholinergic path- ways. 2~ However, respiratory fluctuations in R-R in- tervals are proportional amplitude and cycle lengths 94° and therefore interpretation of these data is limited. On the contrary, augmentation of the slow 0.03 Hz rhythm can be secondary to an underlying vasomotor instability, hence suggesting a greater variability of sympathetic efferent activity. Such cases can be seen in patients with neurally mediated syn- cope, in whom an increased power at low frequen- cies in blood pressure and R-R intervals were asso- ciated by a diminished vasomotor tone. 22 Moreover, the 0.01-0.05 Hz rhythm is under combined sympa- thetic and parasympathetic influence and its augmen- tation could be secondary to a loss of inhibition from upper cortical regions.

The temporal cortex and frontal inferior parietal regions demonstrated a severe synapse (about 30%) and cell loss in Alzheimer's disease. 23 These findings are not isolated to specific layers, but have a rather diffuse character? 4 A decreased cholinergic input to the cortical neurones is associated with a slowing of background EEG activity and augmentation of pow- ers in the theta and delta. 23 Spectral analysis of temporal regions during REM sleep showed a promi-

144 Clinical Autonomic Research • vol 4 . 1994

Dynamic analysis of autonomic nervous system

EEG - Oz

J

o el-

/

400

320

240

l 6O

0 , 0 0 0 . 0 4 0.08 0,1 2 0.1 6

Frequency (Hz)

o i i i i 3 2 0 . . . . . . ~ . . . . . . " . . . . . . . . :-

/

t . . . . . . . . : . . . . . . . . i . . . . . . . . . , . . . . . . . . :.

0.00 0,04 0.08 0, l 2 0.1 6

Frequency (Hz)

O~

0

O O,-

/

4 0 0 . . . . . . . . . . . '- . . . . . . . . i-

I r e / q ,

U : . . . . . . . . . . . . . . . . . : . . . . . . . . . . 320 . . . .

2 4 0

° 0 i ~,0 - - - ~ - g . . . . . ~ . . . . . . . . . . . . . . . . .

0,00 0.04 0.08 0.12 0.1 ~. , ~ Frequency (Hz)

FIG. 3. A detail of Wigner and cross Wigner distribution between EEG and systolic pressure signals and their contour plots of signals from Fig. 1, centred at slower frequencies. Hypotensions are indicated by arrows, Slow EEG activity manifests as increased spectral powers at 0.02 Hz and 0,05 Hz (A). in systolic pressure, the dominant spectral power cover band 0.02-0.04 Hz (B). Cross Wigner distributions show a common spectral content between EEG and systolic pressure during hypotensions.

nent slowing of EEG and prevalence of theta rhythm especially in temporal region and during wakeful- ness. During wakefulness a significant slowing of EEG was observed, with a prominent asymmetry between hemispheres. 2s In our patient diffuse EEG abnormalities (bilateral synchronous bursts of theta

waves) over temporal lobe region were also noted. and cerebral atrophy was confirmed by CT and SPEC examination. Cardiac and respiratory rhythms can be significantly affected by a function loss or atrophy of mesiotemporal lobe structures (amygdala and hippocampus) 26 as it is known from data from pa-

Clinical Autonomic Research • vol 4 . 1994 145

P. Novak e t al.

tients with temporal lobe epilepsy. Moderate-to- severe cell loss favours the ipsilateral correlations of unit discharges with the cardiac cycle (55 vs. 22%), respiratory period (46 vs. 3%) and inspiratory time (27 vs. 7%) ~7 and potentiates ANS activation. Tachyarrhythmias or in contrast bradycardia and syn- cope can occur during partial seizures even during

A 7.0 Rest 1 Hypervent i lot ion

EEO - Oz 5.5

~. 4.0

2.5

T E E

B

1.0

1 80

150

120

90

0 60 0 21 0 420 630 840 1 050 1 260 rO

I~" Time (s) 3; ~q / c4 04 & ~'C° ~ /~ EEG-Oz

t o ~ +- ~ t Syst. pressure

subclinical seizure discharges} 7-3~ Abnormalities in blood pressure regulation can then be explained by loss and atrophy of mesiotemporal structures.

Animal experiments have shown that blood pres- sure waves in the 0.01-0.05 Hz range result from rhythmically modulated firing of the vasomotor neu- rones 32 and that ascending portion of blood pressure

FIG. 4. Signals of EEG theta activity (lead Oz) (A) and systolic pressure (B) at rest and during hyperventilatlon at rest 6 month~ later. Arrows indicate frequent spontaneous hypotensions. Wigner distribu- tions (C, D) show a parallel augmentation of spectral powers (<0.04 Hz) in both EEG and blood pressure.

146 Clinical Autonomic Research-vol 4. 1994

Dynamic analysis of autonomic nervous system

wave is preceded by bursts of neuronal activity. Firing of reticular neurones is increased by a variety of stressful stimuli, most consistently by a cerebral ischaemia or hypoperfusion. 33 Animal studies related the prominent EEG oscillations at frequencies below 1 Hz with and recurrence of sleep spindles with a frequency of 3-10 s (0.2-0.3 Uz) 34 tO changes of arousal during sleep. This activity could be abolished by activation of a variety of ascending activating systems and release of ANS neurotransmitters.

In our patient, both respiratory and nonrespiratory fluctuations in R-R intervals were continuously de- tected and within normal limits, suggesting that in- tegrity of baroreflex feedback loop was preserved. Furthermore, the cardioacceleration during blood pressure drops with concomitant diminution of the respiratory frequencies in R-R intervals suggest a compensatory withdrawal of cardiovagal activity. At the same time, no changes in frequency or amplitude of respiration were observed, which could explain diminishing of powers at RF in R-R intervals. No abnormalities of cardiac rhythms or diminished ve- nous rhythms, which could explain blood pressure drops, were observed. Hypotensive reactions associ- ated with bradycardia and loss of consciousness triggered by emotional or painful stimuli have been frequently reported in patients with vasodepressor syncope. These reactions are believed to originate in pathways descending from prefrontal or medial insu- lar cortex? 5 Moreover, during spontaneous supine hypotension, unlike orthostatic hypotensions or mechanical occlusion of vena cava by pregnant uterus 36 the venous return was not compromised. Also the fact that vasovagal syncope can be induced in cardiac transplant patients suggest involvement of other mechanisms than activation of peripheral mechanoreceptors.37

This study showed that the slow EEG activity can be functionally related to the central ANS. Hence the observed vasomotor instability is presumably of cen- tral nervous system origin. The slow cortical activity would then correspond to negative feedback be- tween brainstem-thalamus-cortex.

References 1. Shields RW Jr. Functional anatomy of the autonomic nervous

system. J Clin Neurophysiol 1993; 10: 2-13. 2. Kitney Rl. An analysis of the nonlinear behaviour of the human

thermal vasomotor control system. J Theor Biol 1975; 52: 231-248.

3. DeBoer RW, Karemaker JM, Strackee J. Hemodynamic fluctua- tions and baroreflex sensitivity in humans: a beat-to-beat analy- sis. Am J Physiol 1987; 253: 680-689.

4. Gebber GL, Barman SM Basis for 2-6 Hz rhythm in sympa- thetic nerve discharge. Am J Physiol 1980; 239: R48-R56.

5. Zhong S, Kenney M J, Gebber GL. High power, low frequency components of cardiac, renal, splenic and vertebral sympathetic nerve activities are uniformly reduced by spinal cord transection. Brain Res 1991; 556: 130-134.

6. Persson PB, Stauss H, Chung O, Wittmann U, Unger T. Spec- trum analysis of sympathetic nerve activity and blood pressure in conscious rats. Am J Physiol 1992; 263: H1348-H1355.

7. Kienzle MG, Fergusson DW, Birkett CL, Myers GA, Berg W J, Mariano J. Clinical, hemodynamic and sympathetic neural corre- lates of heart variability in congestive failure. Am J Cardiol 1992; 69: 761-767.

8. Pagani M, Lombardi F, Guzzetti S, Rimoldi O et al. Power spec- tral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 1986; 59: 178-193.

9. Novak V, Novak P, deChamplain J, LeBlanc R, Martin R, Nadeau R. The effect of respiration on heart rate and blood pressure fluctuations. J Appl Physiol 1993; 74: 617-626.

10. Novak V, Novak P, deChamplain J, Nadeau R. Altered cardiorespiratory transfer in hypertension. Hypertension 1994; 23: 104-113.

11. Lacroix D, Logier R, Kacet S, Hazard JR, Dagano J, Lekieffre J. Effects of consecutive administration of central and peripheral anticholinergic agents on respiratory sinus arrhythmia in normal subjects. J Auton Nerv Sys 1992; 39: 211-218.

12. Japundzic N, Grichios ML, Zitoun P, Laude D, EIghozi JL. Spec- tral analysis of blood pressure and heart rate in conscious rats: effects of autonomic blockers. J Auton Nerv Sys 1990; 30: 90-100.

13. Hugelin A. Forebrain and midbrain influence on respiration. In: Cherniack NS, Widdicombe, eds. The Respiratory System. Handbook of Physiology Sect 3, vol II part 1. Bethesda, MD: Am. Physiol. Society, 1986: 69-91.

14. Hukuhara T Jr, Miyakawa M, Kimura N, Takano K, Kate F. Periodic variation of electrocorticogram in relation to respiratory rhythm and long-term periodic fluctuations of the renal sympa- thetic activity. In: Sieck GC, Gandevia SC, Cameron WE, eds. Respiratory Muscles and Their Neuromotor Control. New York: Liss, 1987: 121-125.

15. Pfurtscheller G. UItralangsame Swankungen innerhalb der rhythmischen Aktivitaet im Alpha-Band und deren moegliche Ursachen. Pflugers Arch 1976; 367: 55-66.

16. Novak P, Lepicovska V, Dostalek C. Periodic amplitude modu- lation of EEG. Neurosci Lett 1992; 138: 213-215.

17. Novak P, Lepicovska V. Slow modulation of EEG. NeuroReport 1992; 3: 189-192.

18. Novak P, Lepicovska V. Increase of slow modulation of EEG in a patient with AIzheimer's disease. Physiol Res 1992; 41: 293-297.

19. Novak P, Novak V. Time-frequency mapping of the heart rate, blood pressure and respiratory signals. Med Biol Eng Comput 1993; 31: 103-110.

20. Whitehouse P J, Price DL, Struble RG et ai. Alzheimer's disease and senile dementia: loss of neurons in basal forebrain. Science 1982; 215: 1237-1239.

21. Aharon-Peretz J, Harel T, Revach M, Ben-Haim SA. Increased sympathetic and decreased parasympathetic cardiac innerva- tion in patients with Alzheimer's disease. Arch Neuro11992; 49: 919-922.

22. Lepicovska V, Novak P, Nadeau R. Time-frequency mapping of neurally mediated syncope. J Cfin Autonom Res 1992; 2: 317-326.

23. Mira SS, Heyman A, McKeel D, et al. The consortium to estab- lish a registry for Alzheimer's disease (CERAD)II. Standardiza- tion of the neuropathologic assessment of Alzheimer's disease. Neurology 1991; 41: 479-486.

24. Scheff SW, Price DA. Synapse loss in the temporal lobe in AIzheimer's disease. Ann Neurol 1993; 33: 190-199.

25. Petit D, Montplaisir J, Lorrain D, Gauthier S. Spectral analysis of the rapid eye movement sleep electroencephalogram in right and left temporal regions: a biological marker of Alzheimer's disease. Ann Neurol 1992; 32: 172-176.

26. Frysinger RC, Harper RM, Havkel RJ. State-dependent cardiac and respiratory changes associated with complex partial epi- lepsy. In: Engel J Jr, Ojemann GA, Luders He, Williamson PD, eds. Fundamental Mechanisms of Human Brain Function. New York: Raven Press, 1987: 219-226.

27. Frysinger RC, Harper RM. Cardiac and respiratory correlations with unit discharge in epileptic human temporal lobe. Epilepsia 1990; 31: 162-171.

28. Liedholm L J, Gudjonsson O. Cardiac arrest due to partial epi- leptic seizures. Neurology 1992; 42: 824-829.

29. Lathers CM, Schraeder PL, Weiner FL. Synchronization of car- diac autonomic neural discharges with epileptogenic activity: the Iockstep phenomenon. EEG Clin Neurophysiol 1987; 67: 247-259.

30. Lathers CM, Schraeder PL. Review of autonomic dysfunction cardiac arrhythmias and epileptogenic activity. J Clin Pharmacol 1987; 27: 346-356.

31. Hughes JR, Schlagenhauff RE. Electro-clinical correlation in temporal lobe epilepsy with emphasis on inter-areal analysis of the temporal lobe. EEG Clin Neurophys 1961; 13: 333-339.

Clinical Autonomic Research • vol 4 • 1994 147

P. Novak e t al.

32. Preiss G, Pol0sa C. Patterns of sympathetic neuron activity associated with Mayer waves. Am J Physiol 1974; 228: 724-730.

33. Miyakawa K, Takeuchi T, Shimizu T, Haashi M, Kanai K. Mecha- nism of blood pressure waves of the third order. In: Miyakawa K, Koepchen HP, Polosa C, eds. Mechanisms of Blood Pressure Waves. Berlin: Springer-Veriag, 1984: 85-117.

34. Steriade M, McCormick DA, Sejnowski T. Thalamocortical oscil- lations in the sleeping and aroused brain. Science 193; 262: 679-685.

35. Van Lieshout VT, Wieiing W, Karemaker JM, Eckberg D. The vasovagal response. Clin Sci 1981; 81: 575-586.

36. Calvin S, Jones OW, Knieriem K, Weinstein L. Oxygen satura- tion in the supine hypotensive syndrome. Obstet Gynaecol 1992; 79: 27-31.

37. Lightfoot JT, Rowe SA, Fortney SM. Occurrence of presyncope in subjects without ventricular innervation. Clin Sci 1993; 85: 695-700.

Received 16 October 1992; accepted with revision 7 March 1994.

148 Clinical Autonomic Research • vol 4 . 1994