American Journal of Emergency Medicine (2012) 30, 651-656

Original Contribution

Depressed sympathovagal balance predicts mortality in patients with subarachnoid hemorrhage^?,??

Te-Fa Chiu MD ^a,b, Chien-Cheng Huang MD ^c,d, Jiann-Hwa Chen MD, MPH ^c,d,

Wei-Lung Chen MD, PhD ^c,d,?

^aDepartment of Emergency Medicine, Chang Gung Memorial Hospital Linkou Branch, Tao-Yuan 333, Taiwan

^bChang Gung University College of Medicine, Tao-Yuan 333, Taiwan

^cDepartment of Emergency Medicine, Cathay General Hospital, Taipei 106, Taiwan

^dFu-Jen Catholic University School of Medicine, Taipei 242, Taiwan

Received 11 January 2011; revised 26 February 2011; accepted 27 February 2011

Abstract

Objectives: The objective of this study is to investigate the role of sympathovagal balance in predicting Inhospital mortality by assessing power spectral analysis of Heart rate variability among patients with nontraumatic subarachnoid hemorrhage in an emergency department (ED).

Methods: A cohort of 132 adult patients with spontaneous SAH in an ED was prospectively enrolled. A continuous 10-minute electrocardiography for off-line power spectral analysis of the HRV was recorded. Using the inhospital mortality, the patients were classified into 2 groups: nonsurvivors (n = 38) and survivors (n = 94). The HRV measures were compared between these 2 groups of patients. Results: Having compared the various measurements, the very low-frequency component, low- frequency component, normalized low-frequency component (LF%), and low-/high-frequency component ratio (LF/HF) were significantly lower, whereas the normalized high-frequency component was significantly higher among the nonsurvivors than among the survivors. A multiple logistic regression model identified LF/HF (odds ratio, 2.16; 95% confidence interval [CI], 1.18-3.97; P = .013) and LF% (odds ratio, 0.78; 95% CI, 0.69-0.88; P b .001) as independent variables that were able to predict inhospital mortality for patients with SAH in an ED. The receiver operating characteristic area for LF/HF in predicting inhospital mortality was 0.957 (95% CI, 0.914-1.000; P b .001), and the best cutoff points was 0.8 (sensitivity, 92.1%; specificity, 90.4%).

Conclusions: Power spectral analysis of the HRV is able to predict inhospital mortality for patients after SAH in an ED. A tilt in the sympathovagal balance toward depressed sympathovagal balance, as indicated by HRV analysis, might contribute to the poor outcome among these patients.

(C) 2012

^? Competing interests: None of the authors have any conflicts to disclose.

^?? Authors’ contributions: The study was designed by T.F.C. and W.L.C. Acquisition of the data was performed by C.H.H., J.H.C., and W.L.C. Analysis and interpretation of data were done by T.F.C. and W.L.C. The statistical analysis was done by J.H.C. and W.L.C. The manuscript was drafted by T.F.C. Critical revision of the manuscript for important intellectual content was done by W.L.C. Final approval of the manuscript was done by T.F.C., C.H.H., J.H.C., and W.L.C.

* Corresponding author. Department of Emergency Medicine, Cathay General Hospital, Taipei, Taiwan. Tel.: +886 2 27082121×3761; fax: +886 2

27021428.

E-mail address: weilung.chen@msa.hinet.net (W.-L. Chen).

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2011.02.037

Introduction

Subarachnoid hemorrhage (SAH) remains a devastating disease with a mortality rate approaching 50%, although early surgical or endovascular protection of the ruptured aneurysms and aggressive postoperative care have improved the outcome [1,2]. Many studies have reported that patients with intracranial events exhibit several changes in heart rate, blood pressure, and overall cardiovascular control since the “Cushing responses” began to be used as the clinical signs of an intracranial event [3-6]. Furthermore, these changes are reported to appear markedly in patients with SAH who have a poor outcome or are subsequently declared brain dead [6-12]. Despite the clinical variables and the definition of prognostic criteria [13-15], individual outcome prediction remains difficult to establish for patients with SAH.

Heart rate variability analysis is a Noninvasive tool that is able to evaluate autonomic nervous modulation of the heart. Power spectral analysis of the HRV provides an assessment of the degree of sympathetic and parasympathetic modulation of the heart over a relatively short period [16,17]. The power spectrum of the HRV is often categorized into a high-frequency component (HF) and a low-frequency component (LF). The HF is related to respiratory sinus arrhythmia and cardiac vagal activity, whereas the LF and normalized LF (LF%) are jointly modulated by the neural activities of both vagal and sympathetic nerves. Power spectral analysis of the HRV has gained popularity and has been broadly applied as a functional indicator of the autonomic nervous system [16,17]. Moreover, a reduction of HRV and impaired sympathovagal balance, as represented by a depressed low-/high-frequency component ratio (LF/HF), have been suggested to be indicative of poor prognosis in critically ill patients [18] and in patients with intracranial events, such as acute stroke, head injury, and neurosurgical illness [19-24]. Because Cardiac abnormalities have been suggested to be associated with the prognosis for patients with SAH [10-12], we, therefore, hypothesize that impaired cardiac variability may correlate with outcome in these patients.

This study aimed to explore the role of power spectral analysis of HRV in predicting inhospital mortality for patients with spontaneous SAH in an emergency department (ED).

Materials and methods

Study design

This was a prospective cohort study that investigated the role of autonomic nervous modulation, as indicated by power spectral analysis of HRV, in predicting inhospital mortality for adult patients with spontaneous SAH in an ED. The study protocol was approved by the institutional review board of the hospital. Written informed consent was obtained from the

patients themselves or from their next of kin before they were enrolled in the study.

Study setting and selection of participants

This study was conducted in a 700-bed university- affiliated medical center with a 40-bed ED staffed with board-certified emergency physicians that provides care for approximately 55 000 patients per year. From July 2005 to June 2010, adult patients with nontraumatic aneurysmal and spontaneous nonaneurysmal SAH diagnosed by Computed tomographic scans of brain or xanthochromia (visual decision) of the cerebrospinal fluid if the computed tomography was nondiagnostic were consecutively enrolled. Patients with persistent arrhythmia, with cardiac pacing, or younger than 18 years were excluded.

Study protocol and outcome measures

All subjects were studied in a supine position and in an air-conditioned resuscitation room with a constant temper- ature of around 25?C. After obtaining written informed consent, a 10-minute lead II electrocardiographic recording was performed immediately after diagnosis was made (ECG 100C, ECG Amplifier; BIOPAC Systems, Inc, Goleta, CA). The time interval from diagnosis to capture of the 10-minute ECG recording was less than 30 minutes. The output signals were digitized at 1000 Hz by an A/D converter (MP150WSW, Starter System for Desktop, and Notebook PCs; BIOPAC Systems, Inc). The digitized ECG signals were subsequently stored in a notebook computer for later power spectral analysis of HRV.

The following demographic data and clinical variables were recorded at the same time for all patients: age, sex, laboratory data, underlying diseases, and comedication that could affect autonomic nervous activity. Subjects who received sedatives, nicardipine, or mechanical ventilation at the same time of 10-minute ECG recording were recorded. To evaluate the severity of the illness of these patients, the Hunt-Hess scale [13] and World Federation of Neurolog- ical Surgeons (WFNS) class [14] were also recorded for these patients. After hospital discharge, the inpatient medical record was reviewed to complete the data collection, including the etiology and mortality. Patients discharged from the hospital in less than 28 days or who remained alive for more than 28 days were classified as survivors in this study; otherwise, the patients were referred to as nonsurvivors.

Heart rate variability analysis

The method used for power spectral analysis of HRV has been described elsewhere [20] and adhered to the standards developed by the Task Force of the European Society of Cardiology and the North American Society of Pacing and

Electrophysiology [17]. In brief, the digitized ECG signals were retrieved to measure the consecutive RR intervals, which are the time intervals between successive pairs of QRS complexes, using the software that was developed for the detection of the R wave (Matlab 6.5; MathWorks, Inc, Natick, MA). All artifacts or ectopic beats were removed, and the resultant missing data (b5% per record) were replaced by interpolated beats derived from the nearest valid data. If the percentage of deletion was more than 5%, then the patient was excluded from the study. The last 512 stationary RR intervals were then used for the HRV analysis. The power spectrum of these RR intervals was obtained by means of fast Fourier transformation (Mathcad 11; Mathsoft, Inc, Cambridge, MA). The area under the spectral peaks within the range 0.01 to 0.04, 0.04 to 0.15, 0.15 to 0.4, and 0.01 to 0.4 Hz were defined as the very low-frequency component (VLF), LF, HF, and total power (TP), respec- tively [16,17]. It is generally accepted that the efferent vagal activity is a major contributor to the HF fluctuations at the respiratory frequency. Thus, the HF in the power spectrum of RR intervals is often used to denote the vagal modulation of the subject. On the other hand, the LF in the power spectrum is modulated by both vagal and sympathetic activities of the patient. Although the interpretation of the ratio of LF to HF remains controversial, it is generally used to reflect the balance between sympathetic and vagal modulations. Therefore, the normalized HF [HF% = 100 x HF/(TP – VLF)] was used as an index of vagal modulation, the LF and normalized LF [LF% = 100 x LF/(TP – VLF)] were used as indices of sympathetic and vagal modulation, and the LF/HF

was used as an index of sympathovagal balance [16,17].

Statistical analysis

As nearly all parameters derived from power spectral analysis of HRV were not normally distributed, therefore, these values are presented as medians (25th-75th interquartile range). ?2 Tests or Fisher exact tests, where appropriate, were used for the statistical analysis of categorical variables. Continuous variables are presented as mean (SD) or median (25th-75th interquartile range) and compared using the independent-sample t test (assuming a normal distribution) or the Mann-Whitney U rank sum test (assuming a nonnormal distribution). For statistical purposes, the Clinical scores used in this study were dichotomized in good and poor groups (WFNS 1 and 2 vs WFNS 3-5, Hunt-Hess 1 and 2 vs Hunt- Hess 3-5). The forward selection multiple logistic regression model with Hosmer-Lemeshow goodness of fit was per- formed to identify the factors that might be used in an ED as the independent predictors of inhospital mortality in these patients. The clinical variables and spectral powers of HRV with a univariate comparison of P b .2 between 2 groups were eligible for inclusion in the model. A receiver operating characteristics (ROC) curve of the statistically significant variables associated with inhospital mortality was drawn up. P b .05 was considered statistically significant. Statistical

analyses were performed using a common statistical package (SPSS 16.0 for Windows; SPSS, Inc, Chicago, IL).

Results

During a 5-year study period, a total of 139 adult patients with nonTraumatic SAH were treated in the ED. Five of 139 patients were excluded from the study because they had arrhythmia (atrial fibrillation) or a permanent pacing rhythm, such that the power spectral analysis of HRV could not be performed. Off-line power spectral analysis of the HRV yielded acceptable recording (ectopic ventricular beats b5%) in 132 of 134 enrolled subjects. Based on their inhospital mortality, these 132 eligible patients, aged 19 to 77 years, were classified as nonsurvivors (n = 38) or survivors (n = 94). The basic characteristics of both groups of patients are shown in Table 1. There were no significant differences in age, sex, vital signs, etiology, comedications (?-blocker,

Table 1 Demographic and clinical characteristics of the patients with SAH

Nonsurvivors Survivors P (n = 38) (n = 94)

Age (y), mean +- SD 50 +- 11 52 +- 12 .471 Sex, n .556 Male/female 13/25 39/55 Vital signs, mean +- SD MAP (mm Hg) 124 +- 11 126 +- 13 .515 Heart rate 88 +- 16 94 +- 13 .423 (beat per minute) Etiology, n (%) .716 Aneurysm 29 (76) 75 (80) AVM 4 (11) 11 (12) Others 5 (13) 8 (9) WFNS class, n (%) b.001 ? 1 and 2 10 (26) 77 (82) 3 to 5 28 (74) 17 (18) Hunt-Hess scale, n (%) b.001 ? 1 and 2 10 (26) 80 (85) 3 to 5 28 (74) 14 (15) Laboratory data, mean +- SD WBC (per mm3) 11850 +- 2668 9953 +- 2481 b.001 ? Glucose (mg/dL) 117 +- 45 103 +- 25 .069 Comedications, n (%) ?-Blocker 4 (11) 14 (15) .588 Sedatives 6 (16) 5 (5) .077 Nicardipine 9 (24) 14 (15) .310 Mechanical ventilation, 5 (13) 4 (4) .119 n (%) Underlying disease, n (%) Hypertension 7 (18) 15 (16) .798 CAD 3 (8) 8 (9) 1.000 DM 5 (13) 9 (10) .544

MAP indicates mean arterial pressure; AVM, arteriovenous malformation; CAD, coronary artery disease; DM, diabetes mellitus. * P b .05 between 2 groups.

sedative drugs, or nicardipine), use of mechanical ventila- tion, or underlying diseases between these 2 groups of patients. However, the WFNS class, Hunt-Hess scale, and white blood cell count were significantly higher in the nonsurvivors compared with the survivors.

Table 2 lists the frequency domains of the HRV measures of both groups of patients. The TP and HF were not significantly different between the 2 groups, whereas the VLF, LF, LF%, and LF/HF were significantly lower and the HF% was significantly higher among the nonsurvivors when compared with the survivors.

Multiple logistic regression model analysis was per- formed to analyze the risk of inhospital mortality (the dependent variable). The independent variables included in the analysis were WFNS class, Hunt-Hess scale, WBC, glucose, use of sedative drugs or mechanical ventilation, TP, VLF, LF, LF%, HF%, and LF/HF. The results showed that LF%, LF/HF, and Hunt-Hess scale were the significant independent predictors of inhospital mortality for adult patients with spontaneous SAH in an ED. The odds ratios (ORs) and P values for LF%, LF/HF, and Hunt-Hess scale are list in Table 3.

With more than 1 of OR (2.16; 95% confidence interval [CI], 1.18-3.97; P = .013) in prediction of inhospital mortality, that is, the OR of LF/HF was greater than 1 with 95% CI that does not include 1, the ROC curve of LF/HF was constructed. As depicted in Fig. 1, the area under the ROC curve of LF/HF for prediction of inhospital mortality was 0.957 (95% CI, 0.914-1.000; P b .001); and the best cutoff

point was 0.8 (sensitivity, 92.1%; specificity, 90.4%).

Discussion

By assessing power spectral analysis of the HRV, the present study demonstrated that impaired sympathovagal balance was an independent predictor of inhospital mortality in adult patients with nontraumatic SAH. We found that the LF% and LF/HF were significantly lower in patients with SAH with a poor prognosis. These results suggested that a tilt

Table 2 Spectral power of HRV of the patients

Nonsurvivors (n = 38)		Survivors (n = 94)	P
TP (ms2)	179.1 (12.2-230.1)	162.6 (26.6-618.8)	.184
VLF (ms2)	33.8 (7.5-58.2)	76.0 (13.0-199.5)	.004 ?
LF (ms2)	30.7 (2.2-38.5)	53.9 (8.1-229.9)	.004 ?
HF (ms2)	59.4 (2.4-145.7)	41.7 (1.4-156.7)	.260
LF% (nu)	36.9 (16.0-40.6)	62.3 (55.6-82.8)	b.001 ?
HF% (nu)	63.2 (59.4-84.0)	37.7 (17.2-44.4)	b.001 ?
LF/HF	0.6 (0.2-0.7)	1.7 (1.3-4.8)	b.001 ?
Values are median and interquartile range (25th-75th percentile). nu indicates normalized unit. * P b .05 between 2 groups (Mann-Whitney U rank sum test).

OR (95% CI)		P
H-H scale	9.39 (2.13-41.40) a	.003
LF/HF	2.16 (1.18-3.97)	.013
LF%	0.78 (0.69-0.88)	b.001

in the sympathovagal balance toward sympathetic depres- sion, as represented by power spectral analysis of HRV, might contribute to the poor outcome in adult patients with spontaneous SAH.

Table 3 Statistically independent variables associated with mortality in the multiple logistic regression models

H-H scale indicates Hunt-Hess scale.

^a The OR for Hunt-Hess scale represents the OR for each 1-point increases from low grade of Hunt-Hess scale (1 and 2) to high grade of Hunt-Hess scale (3 to 5).

The relationship between acute central nervous system events and various cardioVascular abnormalities has been known since the description of Cushing phenomenon at the turn of the last century [3]. The central nervous system modulates cardiac function via 1 of 2 pathways: an indirect effect via humoral mediators such as epinephrine and norepinephrine and a direct effect via afferent and efferent connections with the sympathetic and vagal nervous system. Previous studies have demonstrated that SAH, because it causes a local cerebral arteriolar spasm, can give rise to ischemic lesions in the hypothalamus and surrounding area [25], which may cause sympathetic stimulation of the heart via elevated plasma catecholamine [26]. This, in turn, would seem to be responsible for the cardiac abnormalities [27,28].

Fig. 1 The ROC curve of LF/HF in predicting inhospital mortality in patients with SAH. The area under the ROC curve for LF/HF was 0.957 (95% CI, 0.914-1.000; P b .001), and the best cutoff points was 0.8 (sensitivity, 92.1%; specificity, 90.4%).

Studies have further reported that an elevated catecholamine level may be strongly associated with a poor prognosis in these patients [27,29]. However, power spectral analysis of HRV in this study showed that the LF% and the LF/HF were significantly lower in patients with a poor prognosis, suggesting that there was a relative reduction in Sympathetic activity or a depression of the sympathovagal balance in patients with SAH with a poor outcome. Thus, there would seem to be a paradox between the concentration of plasma catecholamine, as reported by previous studies, and the results of the power spectral analysis of HRV, as described in the present study. Su et al [20] have reported that a nearly abolished HRV and dramatically decreased LF% and LF/HF was present at brain death in patients with head injury, although increasing severity of the brain stem damage was accompanied by a slight increase in sympathetic drive and a slight decrease parasympathetic drive, as indicated by increasing LF% and LF/HF and decreasing HF, respectively. In addition, Haji-Michael et al [24] have found that a lowered LF/HF is associated with a poor recovery or death after neurosurgical illness. We, therefore, speculate that a critical illness may lead to a stress response with increased sympathetic activity; however, an impaired sympathovagal balance or loss of sympathetic tone might represent an inadequate response of the autonomic nervous system to the stress associated with a more severe condition and, furthermore, indicate a poor prognosis in critically ill patients [18,30].

Previously, loss of HRV has been reported to be prognostically unfavorable in patients with coronary artery diseases and critical illness [17,18]. Sajadieh et al [31] have further found that the combination of HRV measures and C-reactive protein can predict death in community subjects. In addition, loss of cardiac variability and impaired LF were found in patients with Brain infarction [32]. Moreover, impaired LF/HF was reported to be as indicator of severity in patients with head injury [19,20], acute stroke [33], and neurosurgical illness [24]. In this study, power spectral analysis of HRV was shown to be useful in predicting the outcome for patients with SAH; and both LF% and LF/HF have been identified to be independently predictors of inhospital mortality. To our best knowledge, the present study is the first study to investigate the role of power spectral analysis of HRV in predicting outcome for patients with SAH; and the results should provide emergency physicians with valuable information that will help them identify the patients with potential worst outcomes, although the exact physiologic mechanisms responsible for the various HRV components are still incompletely understood.

There are several limitations in this study. Firstly, HRV analysis cannot be used to assess patients with a nonsinus rhythm, a significant degree of arrhythmia, or cardiac pacing [17]. This restriction may limit the applicability of the results to the broader spectrum of patients with SAH in an ED. Secondly, the study was exploratory in nature, that is,

patients who were currently taking heart rate-altering medication (eg, ?-blocker) were not excluded from this study. In addition, patients who received necessary treat- ments such as sedative drugs, nicardipine, or mechanical ventilation were not excluded from this study. This may further limit the present results in implications. For statistical purpose, the Clinical scores (Hunt-Hess scale and WFNS class) used in this study were dichotomized in good and poor groups with grades 1 and 2 and 3 to 5, respectively, which may affect the analysis and the results. This is the third limitation in the present study. Finally, the sample size of this study is relatively small, which could mean that the predicting power provided here is not sufficiently meaning- ful; and further investigations with a large sample size are needed before the results are applied clinically.

Conclusions

This study demonstrated that spectral power of HRV, especially LF/HF, would seem to be useful as independent predictor of inhospital mortality for adult patients with spontaneous SAH in an ED. A tilt in the sympathovagal balance toward depressed sympathetic modulation, as indicated by power spectral analysis of HRV, may contribute to the poor outcome in these patients.

References

1. Johnston SC, Selvin S, Gress DR. The burden, trends, and demographics of mortality from subarachnoid hemorrhage. Neurology 1998;50:1413-8.
2. Wijdicks EF, Kallmes DF, Manno EM, Fulgham JR, Piepgras DG. Subarachnoid hemorrhage: neurointensive care and aneurysm repair. Mayo Clin Proc 2005;80:550-9.
3. Cushing H. Experimental and clinical observations concerning states of increased intracranial pressure. Am J Med Sci 1902;124:375-400.
4. Norris JW. Cardiovascular manifestations of acute neurological lesions. In: Aminoff MJ, editor. Neurology and general medicine. The neurological aspects of medical disorders. New York: Living- stone/Churchill; 1989. p. 159-67.
5. Chen HI, Wang DJ. Systemic and pulmonary Hemodynamic responses to Intracranial hypertension. Am J Physiol 1984;247:H715-21.
6. Samuels MA. The brain-heart connection. Circulation 2007;116:77-84.
7. Claassen J, Kreiter KT, Kowalski RG, Du EY, Ostapkovich N, Fitzsimmons BM, et al. Effect of acute physiologic dearrangements on outcome after subarachnoid hemorrhage. Crit Care Med 2004;32: 832-8.
8. Bernardini GL, Mayer SA. Subarachnoid hemorrhage: clinical presentation and neuropsychological outcome. Med Update Psychia- trists 1998;3:71-6.
9. Kuo TBJ, Yien HW, Hseu SS, Yang CCH, Lin YY, Lee LC, et al. Diminished vasomotor component of systemic arterial pressure signals and baroreflex in brain death. Am J Physiol 1997;273:H1291-8.
10. Coghlan LA, Hindman BJ, Bayman EO, Banki NM, Gelb AW, Todd MM, et al. Independent associations between electrocardiographic abnormalities and outcomes in patients with Aneurysmal subarachnoid hemorrhage: findings from the intraoperative hypothermia aneurysm surgery trial. Stroke 2009;40:412-8.
11. Zaroff JG, Rordorf GA, Newell JB, Ogilvy CS, Levinson JR. Cardiac outcome in patients with subarachnoid hemorrhage and electrocardio- graphic abnormalities. Neurosurgery 1999;44:34-9 discussion 39-40.
12. Frontera JA, Parra A, Shimbo D, Fernandez A, Schmidt JM, Peter P, et al. Cardiac arrhythmias after subarachnoid hemorrhage: risk factors and impact on outcome. Cerebrovasc Dis 2008;26:71-8.
13. Hunt WE, Hess RM. Surgical risk as related to time of intervention in the repair of Intracranial aneurysms. J Neurosurg 1968;28:14-20.
14. Drake CG, Hunt WE, Sano K, Kassell N, Teasdale G, Pertuiset B, et al. Report of world federation of neurological surgeons committee on a universal subarachnoid hemorrhage grading scale. J Neurosurg 1988;68:985-6.
15. Wartenberg KE, Schmidt JM, Claassen J, Temes RE, Frontera JA, Ostapkovich N, et al. Impact of Medical complications on outcome after subarachnoid hemorrhage. Crit Care Med 2006;34:617-23.
16. Pomeranz B, Macaulay RJ, Caudill MA, Kutz I, Adam D, Gordon D, et al. Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol 1985;248:H151-3.
17. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 1996;93:1043-65.
18. Kleiger RE, Stein PK, Bigger Jr JT. Heart rate variability: measurement and clinical utility. Ann Noninvasive Electrocardiol 2005;10:88-101.
19. Rapenne T, Moreau D, Lenfant F, Vernet M, Boggio V, Cottin Y, et al. Could heart rate variability predict outcome in patients with severe head injury? A pilot study. J Neurosurg Anesthesiol 2001;13:260-8.
20. Su CF, Kuo TBJ, Kuo JS, Lai HY, Chen HI. Sympathetic and parasympathetic activities evaluated by heart-rate variability in head injury of various severities. Clin Neurophysiol 2005;116:1273-9.
21. Gujjar AR, Sathyaprabha TN, Nagaraja D, Thennarasu K, Pradhan N. Heart rate variability and outcome in acute severe stroke: role of power spectral analysis. Neurocrit Care 2004;1:347-53.
22. Lacquaniti LG, Irone M, Barbacini S, Merlo F, Demo P, Pellegrin C, et al. Heart rate variability and severe brain damage: preliminary data. Int J Clin Monit Comput 1993;10:181-5.
23. Baillard C, Vivien B, Mansier P, Mangin L, Jasson S, Riou B, et al. Brain death assessment using instant spectral analysis of heart rate variability. Crit Care Med 2002;30:306-10.
24. Haji-Michael PG, Vincent JL, Degaute JP, van de Borne P. Power spectral analysis of cardiovascular variability in critically ill neurosurgical patients. Crit Care Med 2000;28:2578-83.
25. Crompton MR. Hypothalamic lesions following the rupture of cerebral berry aneurysms. Brain 1963;86:301-14.
26. Myers MG, Norris JW, Hachniski VC, Sole MJ. Plasma norepineph- rine in stroke. Stroke 1981;12:200-4.
27. Benedict CR, Loach AB. Clinical significance of plasma adrenaline and noradrenaline concentrations in patients with subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry 1978;41:113-7.
28. Zaroff JG, Pawlikowska L, Miss JC, Yarlagadda S, Ha C, Achrol A, et al. Adrenoceptor polymorphisms and the risk of cardiac injury and dysfunction after subarachnoid hemorrhage. Stroke 2006;37:1680-5.
29. Grad A, Kiauta T, Osredkar J. Effect of elevated plasma norepineph- rine on electrocardiographic changes in subarachnoid hemorrhage. Stroke 1991;22:746-9.
30. Muhlnickel Von B. Variability of heart rate as a diagnostic criterion for diagnosing brain death. Anaesthesiol Reanim 1989;14:71-81.
31. Sajadieh A, Nielsen OW, Rasmussen V, Hein HO, Hansen JF. C- reactive protein, heart rate variability and prognosis in community subjects with no apparent heart disease. J Intern Med 2006;260: 377-87.
32. Korpelainen JT, Sotaniemi KA, Huikuri HV, Myllyla VV. Abnormal heart rate variability as a manifestation of Autonomic dysfunction in hemispheric brain infarction. Stroke 1996;27:2059-63.
33. Gujjar AR, Sathyaprabha TN, Nagaraja D, Thennarasu K, Pradhan N. Heart rate variability and outcome in acute severe stroke: role of power spectral analysis. Neurocritical care 2004;3:347-54.