Diverse autonomic regulation of pupillary function and the cardiovascular system during alcohol withdrawal

Drug and Alcohol Dependence 159 (2016) 142–151

Contents lists available at ScienceDirect

Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep

Full length article

Diverse autonomic regulation of pupillary function and the cardiovascular system during alcohol withdrawal Thomas Jochum a , Johannes Hoyme a , Steffen Schulz b , Markus Weißenfels a , Andreas Voss b , Karl-Jürgen Bär a,∗ a b

Psychiatric Brain & Body Research Group Jena, Department of Psychiatry and Psychotherapy, Jena University Hospital, Jena, Germany Department of Medical Engineering and Biotechnology, Ernst-Abbe-Hochschule Jena, University of Applied Sciences, Jena, Germany

a r t i c l e

i n f o

Article history: Received 8 September 2015 Received in revised form 7 December 2015 Accepted 8 December 2015 Available online 4 January 2016 Keywords: Pupil Vagal function Locus coeruleus Heart rate variability Cardiac mortality Alcohol dependence Withdrawal

a b s t r a c t Background: Previous research indicated the complexity of autonomic dysfunction during acute alcohol withdrawal. This study aimed to investigate the pupillary light reflex as an indicator of midbrain and brainstem regulatory systems in relation to cardiovascular autonomic function. Methods: Thirty male patients were included in the study. They were investigated during acute alcohol withdrawal syndrome and 24 h later during clomethiazole treatment and compared to healthy controls. Parameters of pupillary light reflex of both eyes as well as heart rate variability, blood pressure variability and baroreflex sensitivity (BRS) were studied. Results: We observed significantly reduced sympathetic (small diameter, e.g., left eye: 5.00 in patients vs. 5.91 mm in controls) and vagal modulation (e.g., prolonged latencies, left eye: 0.28 vs. 0.26 ms) regarding both pupils during acute alcohol withdrawal syndrome. Cardiovascular parameters showed reduced vagal modulation (e.g., b-slope of BRS: 7. 57 vs. 13.59 ms/mm Hg) and mixed results for sympathetic influence. After 24 h, autonomic dysfunction improved significantly, both for the pupils (e.g., left diameter: 5.38 mm) and the heart (e.g., b-slope of BRS: 9.34 ms/mm Hg). While parameters obtained from the pupil correlated with cardiac autonomic function (e.g, BRS and left diameter: r = 0.564) in healthy controls, no such pattern was observed in patients. Conclusion: Results obtained from the pupil during acute alcohol withdrawal do not simply mirror autonomic dysfunction regarding the heart. Pupillary and cardiovascular changes after 24 h indicate state dependencies of the results. The findings are discussed with respect to autonomic mechanisms and potentially involved brain regions. © 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Alcohol withdrawal syndrome is a common health hazard that occurs when alcohol dependent people either suddenly stop or significantly reduce their alcohol intake. They experience a combination of physical and emotional symptoms such as restlessness, anxiety, nausea or tremor. In addition, alcohol withdrawal syndrome is associated with obvious autonomic changes such as tachycardia, hypertension, an increased body sweat rate and an elevated cardiac output (Kahkonen, 2004).

∗ Corresponding author at: Psychiatric Brain & Body Research Group Jena, Department of Psychiatry and Psychotherapy, University Hospital, Philosophenweg 3, 07743 Jena, Germany. Fax: +49 3641 9390452. E-mail address: [email protected] (K.-J. Bär). http://dx.doi.org/10.1016/j.drugalcdep.2015.12.030 0376-8716/© 2016 Elsevier Ireland Ltd. All rights reserved.

The underlying mechanisms of autonomic dysfunction during alcohol withdrawal syndrome are complex, as various central and peripheral regulatory systems are thought to be involved. In addition to the obvious sympathetic over-activity and vagal withdrawal, neuroendocrine abnormalities are of major importance for the cardiovascular system (Potter et al., 1984; Whitworth et al., 2000). In particular, tachycardia or elevated blood pressure might be the result of increased central and peripheral adrenergic activity. It has been shown that the highest cerebrospinal fluid concentrations of noradrenaline are present at the onset of alcohol withdrawal syndrome (Eisenhofer et al., 1990; Kahkonen, 2004). Furthermore, it has been demonstrated that the concentration and metabolism of dopamine, noradrenaline, adrenaline, and homovanillic acid are higher in alcohol-dependent patients compared with healthy controls (Hawley et al., 1994; Potter et al. 1983). Various studies have investigated changes of cardiovascular regulation during alcohol withdrawal using sophisticated methods

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

143

Table 1 Pupillary light reflex and cardiovascular parameters.

Pupillometry

Parameter

Definition

Autonomic division

Dimension

Pupillary diameter

Greatest horizontal diameter of the darkness-adapted pupil Elapsed time between light stimulus and beginning of constriction Amplitude (difference between baseline pupil diameter & maximal constriction divided by baseline diameter Velocity of 40–80% interval of the amplitude Mean heart rate at rest Spectral power in high frequency range (0.15–0.4 Hz) normalized by total power below 0.4 Hz Ratio of low (0.04–0.15 Hz) and high frequency power (0.15–0.4 Hz)

SNS (1)

mm

PNS (1)

ms

PNS (2)

%

PNS (1) PNS and SNS (4) PNS (4)

mm/s beats/minute

Latency Relative amplitude

Heart rate variability (HRV)

Constriction velocity Heart rate High frequency (HFn) LF/HFHRV

Heart rate complexity Pre-ejection period

Compression entropy (Hc) PEP

Blood pressure variability (BPV)

Blood pressure LF/HFBPV

Baroreflex sensitivity (BRS)

b-slope t-slope

Time between ventricle contraction onset and blood ejection through the aortic valve Mean systolic blood pressure Ratio of low (0.04–0.15 Hz) and high frequency power (0.15–0.4 Hz) Bradycardic baroreflex sensitivity Tachycardic baroreflex sensitivity

SNS/PNS (4) rather PNS (5) SNS (6)

ms

PNS and SNS (7) SNS/PNS (7)

mmHg

PNS (8) PNS (8)

ms/mmHg ms/mmHg

PNS: parasympathetic nervous system; SNS: sympathetic nervous system; LF: low frequency, HF: high frequency; References in brackets: (1) Heller et al., 1990; (2) Levy et al., 1992; (3) Smith and Smith, 1983; (4) Berntson et al., 1997; (5) Baumert et al., 2004; (6) Smith et al., 1989; (7) Floras 2013; (8) Malberg et al., 1999.

such as heart rate variability (HRV) or blood pressure variability. HRV has become a widely used and very sensitive tool to describe mainly parasympathetic (vagal) influences on the heart. In particular, it indicates the extent of heart rate fluctuations around its mean. In simple terms, high parasympathetic modulation is associated with a high degree of HRV and is a marker of well-being and longevity (Zulfiqar et al., 2010). In contrast, reduced HRV is observed during stress, in various cardiac and psychiatric conditions and is associated with increased cardiac morbidity in many conditions. The extent of heart rate fluctuations (HRV) might be assessed by a variety of measures. In particular, the calculation of time domain or frequency domain parameters is commonly used. For instance, the parameter high frequency (HF, see method section for details) is a measure of vagal modulation, whereas the ratio of low and high frequency (LF/HF) indicates the sympatho-vagal balance (Table 1). Another set of parameters has been developed to describe the regularity or complexity of heart rate fluctuations. The application of these novel analyses has led to a higher sensitivity for detecting autonomic dysfunctions (Baumert et al., 2004). In this study, compression entropy (Hc) was used. Hc indicates to which degree heart rate data can be compressed. The more frequent certain sequences occur in the signal – and therefore the more regular these series are – the higher the compression rate and the lower the Hc value. Previous studies have shown that HRV is decreased in alcoholdependent subjects compared to nondependent users; however, this variation improves to some extent after prolonged abstinence (de Zambotti et al., 2015; Irwin et al., 2006; Karpyak et al., 2014). In addition, it has been suggested that HRV may provide an important marker for individual differences regarding the risk of relapse. Interestingly, low HRV seems to predict levels of overall craving and is associated with self-regulatory effort (Quintana et al., 2013). We have previously shown that the acute phase of withdrawal is associated with a pronounced reduction of vagal modulation most likely associated with sympathetic over-activity (Bär et al., 2006a,b, 2007). This was observed by means of time or frequency domain parameters. Staying in line with the clinical pattern, we found some improvement of vagal function after 24 h of treatment. In addition to variability measures, complexity parameters such as compression entropy (Hc) or heart rate turbulence indices revealed that heart rate regulation is significantly less complex (more regular) in patients during alcohol withdrawal (Bär et al., 2008b; Jochum et al.,

2012). In addition, reduced physical fitness was partly attributed to reduced HRV leading to less efficient regulation of physiological processes (Herbsleb et al., 2013). The concept of HRV using heart rate as signal has been transferred to short-term blood pressure changes. Hence, the term blood pressure variability (BPV) was coined. In contrast to HRV, it is generally assumed that increased BPV indicates amplified sympathetic modulation and measures of BPV are considered as novel risk marker for cardiovascular diseases (Floras, 2013; Parati et al., 2012). In addition, the baroreceptor-heart rate reflex (baroreflex sensitivity, BRS) is a key mechanism contributing to the neural regulation of the cardiovascular system (Table 1). For instance, a rise in blood pressure elicits reflex parasympathetic activation and sympathetic inhibition with a subsequent decrease of heart rate. The sensitivity of this negative feedback loop signifies the quality of the system. BRS measurements are established tools for assessing the autonomic control of the cardiovascular system. In the last three decades, clinical evidence has been accumulated showing that cardiovascular diseases are often accompanied by an impairment of baroreflex mechanisms with an imbalance in the sympatheticvagal outflow to the heart (La Rovere et al., 2008). It has been shown that reduced BRS sensitivity is present in patients during alcohol withdrawal (Bär et al., 2006a) adding evidence to the notion of reduced vagal modulation during alcohol withdrawal. As indicated above, most information gained from indices of HRV and BRS analysis describes vagal modulation. To add a sympathetic marker to our analysis, we have calculated the pre-ejection period (PEP, Table 1). It is the measure of choice to monitor changes in cardiac sympathetic activity non-invasively. Under conditions of stable preload and afterload, changes in PEP reflect changes in contractility, which are influenced by sympathetic but not parasympathetic activity (Newlin and Levenson, 1979). So far, the loss of vagal functioning has been demonstrated mainly in relation to the cardiovascular system in patients suffering from alcohol dependence. However, it is important to finally assess the neural basis. Pupillary function can be assessed during functional magnetic imaging and can be related to the activity of important structures such as the Edinger–Westphal nucleus or the locus coeruleus. Therefore, we analyzed pupil function during withdrawal and tested the hypothesis that the loss of vagal modulation is apparent in the pupil’s reaction to light, which is strongly

144

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

Table 2 Demographic and clinical data of participants. Controls

Patients

Number of participants

30

30

Demographic data Age (years) Ethnicity/race

44.3 ± 12.3 All caucasians

46.3 ± 9.3 All caucasians

Education Primary (≤10 years at school, n) Secondary (>10 years at school, n) Alcohol consumption (g/day) Body mass index (kg/m2 ) Number of smokers Smoking (p/y)

12 18 46.3 ± 63.2 25.4 ± 3.3 25 22.9 ± 14.9

26 4 153.0 ± 40.4 24.3 ± 3.6 26 23.5 ± 13.9

LARS MALT AW scale (acute, unmedicated) AW scale (medicated after 24 h)

n.a. n.a. n.a. n.a.

13.8 ± 11.1 27.9 ± 8.5 5.4 ± 1.2 2.5 ± 1.2

Clomethiazole (mg) after 24 h

n.a.

3360 ± 1254

p/y: packyears (cigarette packs per day multiplied with the number of smoking years); LARS: Lübecker alcohol withdrawal risk scale; MALT: Munich Alcoholism Test; AW scale: Alcohol Withdrawal scale; BDI: Beck Depressions Inventory. Data presented as mean ± SD. n.a.: not applicable.

influenced by the autonomic nervous system. Previous studies have demonstrated the feasibility that both adrenergic activation and parasympathetic withdrawal can be estimated by the pupillary light reflex assessment (Keivanidou et al., 2010). While the acute intake of nicotine or morphine produces miosis, the immediate effect of alcohol consumption rather leads to enlarged pupils (Bokoch et al., 2015; Castro et al., 2014). Changes of pupil sizes and function during alcohol withdrawal have not been reported to our knowledge previously. This is of specific importance since the Edinger–Westphal nucleus, which influences pupil constriction, has been shown to be involved in stress responses as well as reward response and shows close associations with the dopaminergic ventral tegmental area (VTA) and the noradrenergic locus coeruleus (LC) (Ryabinin et al., 2013). In this study, we evaluated the pupillary light reflex, a putative marker of both central sympathetic and parasympathetic balance, in addition to the assessment of HRV, BPV and BRS indices. We hypothesized a sympathetic over-activity and vagal withdrawal for the pupil and the heart that are most pronounced during acute withdrawal when compared to controls and to the re-evaluation after 24 h (Fig. 1). In particular, the relation between heart and eye might reveal important insights into reduced vagal function and sympathetic over-activity during acute alcohol withdrawal. In previous studies, we found close associations between measures of HRV and pupillary light reflex measures both in healthy subjects (Bär et al., 2009) and in patients with depression or schizophrenia (Bär et al., 2008a, 2004). However, associations were not always as expected. For instance, sympathetic functioning as indicated by the diameter of the pupil correlated positively with complexity measures of heart rate (compression entropy) which predominately indicates vagal function in healthy subjects (Bär et al., 2009).

2. Materials and methods 2.1. Participants Forty-two male patients admitted for alcohol detoxification were screened (Fig. 1). According to inclusion criteria, thirty patients were included in our study (for participants’ characteristics, see Table 2) and compared to thirty healthy controls. All patients had a history of alcohol dependence according to DSM-

IV criteria and suffered from a severe acute alcohol withdrawal syndrome. In addition to a clinical interview on the detoxification unit performed by a staff psychiatrist, several scales such as the Munich Alcoholism Test (Feuerlein et al., 1979) and the Lübecker alcohol withdrawal risk scale (Wetterling et al., 1997) were used for screening and confirmation of the diagnosis (Fig. 1). The Munich Alcoholism Test consists of two parts (self and external assessment) in order to validate the diagnosis. Patients scoring lower than 11 points in the Munich Alcoholism Test were not included in the study. The Lübecker alcohol withdrawal risk scale indicates the expected severity of withdrawal symptoms during screening (Table 2). In addition, to quantify the severity and course of the withdrawal syndrome, the “Alcohol Withdrawal Scale” (AWscale; Wetterling et al., 1997) and the Banger score (Banger et al., 1992) were used. The AW scale considers autonomic and mental symptoms and is a suitable instrument to quantify alcohol withdrawal (Table 2; Wetterling et al., 1997). The Banger score is a clinical scale to quantify the severity of the alcohol withdrawal syndrome. The rating of absent to severe autonomic, neurological and mental symptoms results in a sum score which allows the application of an adequate clomethiazole dose. The Banger score was used according to in-house guidelines for 6 h (one rating per hour) to determine the potential dose of clomethiazole based upon the severity of withdrawal symptoms (4–6 points: 6 ml clomethiazole per hour; 6–8 points: 12 ml clomethiazole per hour). Patients were included as soon as they reached at least 4 points at the Banger score. Clomethiazole treatment was started after the first assessment according to Banger values (Banger et al., 1992; Jochum et al., 2010, 2011). To distinguish between the influences of medication and the disease, patients were investigated twice; once during the acute state without any medication and after 24 h when they had received pharmacological treatment (Fig. 1). Measurements during the acute phase were compared to healthy controls and to re-assessment indices. In addition to a physical examination and routine ECG, blood chemistry and standardized toxicological investigations (blood and urine) for licit and illicit drugs were performed. To avoid any interference with autonomic assessment, patients with a history of drug or substance abuse or any evidence of drugs or other illegal substances in the toxicological screening were excluded from the study. Similarly, patients with a history of severe alcohol-related diseases such as liver cirrhosis, peripheral neuropathy, diabetes mellitus or any signs of cardiomyopathy or ophthalmological diseases were not included in the screening procedure. According to our laboratory guidelines, smokers were asked to refrain from smoking one hour before the assessment and not to change smoking habits during the study to avoid additional nicotine withdrawal symptoms. Controls were investigated only once. Control subjects were recruited from the general community via advertisement and matched to patients. Controls were also interviewed to assure the absence of a psychiatric disorder or any present or previous substance abuse related disease. A routine clinical examination was performed to exclude any further interfering disease. Exclusion criteria of patients were similarly applied to controls. This study was carried out in accordance with the Declaration of Helsinki. All participants had to be sober to participate in the study and gave written informed consent to a protocol approved by the Ethics Committee of the University Hospital, Jena. Patients were not compensated for participating in the study. 2.2. Data acquisition and pre-processing All subjects were studied between 1 p.m. and 6 p.m. in a quiet room, which was kept comfortably warm (22–24 ◦ C). Patients and

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

145

Fig. 1. The flow chart gives an overview on the study protocol.

controls were asked to refrain from smoking, heavy eating or exercising 2 h before the investigation. Participants were asked to relax, breathe regularly and move as little as possible. Before the ECG and blood pressure were recorded for 30 min, a 10 min adaptation period to the surroundings was allowed. After disconnecting the recording device, participants were left for another 10 min before pupillometry was performed according to the protocol below.

2.5. Non-linear compression entropy (Hc) of heart rate variability The application of this method on RR time series has been described in detail elsewhere (Baumert et al., 2004). In brief, the entropy (complexity) of a given text is defined as the smallest algorithm that is capable of generating the text. In this study, such compression was generated by applying the LZ77 algorithm for loss-less data compression (Ziv and Lempel, 1977). The ratio of the compressed and the original time series length represents an index of entropy and is referred to as compression entropy Hc.

2.3. Heart rate and blood pressure recordings The high resolution electrocardiogram (sampling rate: 1000 Hz) was recorded for 30 min from two separate adhesive monitoring electrodes (CNSystems® , Medizintechnik GmbH, Austria), which were placed on the chest wall to assume maximal R-wave amplitude. The device automatically extracted the RR-intervals (beat to beat interval, BBI). Values of average diastolic and systolic blood pressure were calculated by means of continuous blood pressure, which was simultaneously recorded non-invasively from the third and fourth finger using the vascular unloading technique (Penaz et al., 1976). RR-interval time series were filtered and interpolated for ectopic beats and artifacts. Respiration was measured using impedance data recorded by the Task Force Monitor® .

2.6. Baroreflex sensitivity (BRS) The BRS was assessed using the sequence method (Bertinieri et al., 1985; Malberg et al., 1999). A detailed description has been published previously (Bär et al., 2006a). In brief, spontaneous sequences of at least three consecutive beats were analyzed, when increased systolic blood pressure (SBP) of at least 1 mmHg caused an increased beat-to-beat interval (BBI) of at least 5 ms (bradycardic sequence). For each sequence, the regression between the three SBP values and three BBI values was calculated and the slope (bradycardic slope: bslope) of the regression line was used as an index of BRS (Table 1).

2.7. Pre-ejection period (PEP) 2.4. Time and frequency domain parameters of heart rate variability (HRV) In accordance with the suggestions of the HRV study task force [Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology 1996], we computed measures of HRV in the frequency domain (Table 1). The normalized high frequencies (HFn) as well as the ratio of low frequency and high frequency of HRV (LF/HF) were calculated (Berntson et al., 1997).

PEP is commonly used as a measure of sympathetic control (Berntson et al., 1994; Cacioppo et al., 1994). This parameter was derived from impedance cardiography and ECG data. It is quantified as the time interval in milliseconds from the onset of the ECG Q wave to the B point of the dZ/dt wave derived from impedance data. PEP is equal to the period between electrical invasion of the ventricular myocardium and the opening of the aortic valve. PEP depends on the time development of the intraventricular pressure and myocardial contractility, which is largely under sympathetic control. Therefore PEP is commonly used as a noninvasive measure

146

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

of sympathetic tone, whereas lower PEP values represent higher levels of sympathetic control (Smith et al., 1989). 2.8. Pupillary light reflex Pupillary light reflex was performed as previously reported (Bär et al., 2005). In brief, subjects were placed comfortably in an adjustable chair in front of a Compact Integrated Pupillograph (CIP Version 8.0, AMTech, Weinsheim, Germany) in a dimly lit room (background illumination: 15 lx) for at least 10 min. Subjects then placed their head on chin and forehead rests of the apparatus located on a table in front of them. To examine their left eye, they were requested to fix their gaze on a black spot (diameter: 0.5 cm) located 10◦ to the right at a distance of 2 m. To examine their right eye, participants were instructed to fix their gaze on a similar spot 10◦ to the left. Subjects were instructed to sit quietly while keeping their eyes open and focused on an object without blinking. The measuring system consisted of: (1) a charged coupled device line-scan camera connected to a frame grabber of a personal computer and (2) an image-processing software with a 250 Hz realtime analysis. The system detected the pupil margins electronically under infrared illumination and tracked the pupil diameter using standardized lighting (200 ms; 4 × 103 cd/m−2 ) continuously. The following pupillary light reflex parameters (Table 1; Heller et al., 1990; Levy et al., 1992; Smith and Smith, 1983) were calculated (mean from 5 consecutive examinations): the pupil diameter (mm), constriction latency (ms), constriction velocity (mm/s) and the relative amplitude of constriction (%). The order of which eye was tested first was randomized between subjects. 2.9. Statistical analyses For statistical analysis, SPSS for Windows (version 13.0, SPSS Inc., Chicago, IL, USA) was used. First, the Kolmogorow–Smirnow test was applied to all parameters in order to control for normal distribution. In the first analysis, all parameters obtained from acute patients were compared with data from healthy controls. Here, an overall MANOVA was performed applying the between-subject factor GROUP (patients (acute) versus healthy controls) including the parameters heart rate (HR), LF/HFHRV , normalized HF (HFn), b-slope, t-slope, respiratory rate, systolic and diastolic blood pressure, compression entropy (Hc), pre-ejection period (PEP), LF/HFBPV as well as pupillary diameter, latency, constriction velocity and relative amplitude of both eyes. Then, univariate analyses were performed for single parameters. The second analysis compared data of patients during the unmedicated state with data obtained during clomethizole treatment after one day. Here, a paired t-test was performed to compare autonomic parameters of acute patients versus 24 h data. In previous studies, we observed differences regarding the association between cardiac and pupillary function (Bär et al., 2008a). Therefore, selective parameters as assessed from pupillometry (diameter and latency of the left eye) were correlated with HRV and BRS parameters (HFn, t-slope and compression entropy) using Pearson correlation coefficient for patients and controls separately. Significance level was set at p < 0.05. To compare the correlation coefficients of controls and patients, the Fisher r-to-z transformation was used (http://vassarstats.net/rdiff.html). 3. Results 3.1. Autonomic and pupil parameters of patients and controls during the first investigation The Kolmogorov–Smirnov test indicated a normal distribution of parameters. The applied MANOVA showed an overall difference

for obtained parameters at admission between patients and controls [ = 0.272; F(40,19) = 7.208; p < 0.0001]. Follow-up univariate ANOVAs showed significant differences between patients and controls for most autonomic parameters with the exception of HFn and LF/HFHRV (Fig. 2, Table 3) as well as for all included pupil parameters on both sides except for the relative amplitude (Fig. 3, Table 3). Overall, parameters obtained indicated significantly reduced vagal modulation for the heart and the pupil. However, mixed results were obtained for sympathetic modulation. While, increased sympathetic modulation was observed at the level of the heart, a reduction was found at the pupil. 3.2. Changes in autonomic and pupil function after treatment The results for all parameters of the pairwise t-test comparing values of patients at baseline and after treatment are displayed in Figs 2 and 3 and Table 3. Most autonomic cardiovascular parameters and values of pupillary light reflex of both sides showed an improvement of autonomic dysfunction after 24 h. Sympathetic modulation regarding the pupil increased significantly as well. 3.3. Correlation analysis Using the Pearson’s correlation analysis, we investigated the association between pupil and autonomic regulation. Results are depicted in Fig. 4 and Table 4. With regard to the control group, we found a positive correlation between all cardiovascular parameters (HFn, tslope, Hc) and left pupil diameter. The same autonomic parameters correlated negatively with the parameter latency. In patients, however, there was no correlation between cardiovascular and pupillary light reflex parameters. Results of the statistical comparison of correlation coefficients of patients and controls are displayed in Table 4. 4. Discussion The clinical autonomic pattern during alcohol withdrawal seems to be dominated by sympathetic predominance. In this study, we found both reduced vagal as well as decreased sympathetic modulation regarding the pupils of patients and reduced vagal activity at the level of the heart. Results obtained for cardiac sympathetic activity indicated not unequivocally increased cardiac sympathetic modulation in our patients. In addition, we observed interesting correlations between autonomic function of the pupil and the heart in healthy subjects while no such associations were observed in patients during acute alcohol withdrawal. In general, we observed reduced parasympathetic modulation at the level of the eye and heart during alcohol withdrawal. This was indicated for the pupil by prolonged latencies or decreased constriction velocities. In addition, reduced baroreflex sensitivity or reduced complexity of the heart rate indicated reduced vagal cardiac modulation as previously shown. However, and more intriguing, we observed decreased diameters of pupils (reduced sympathetic tone) and some signs of increased sympathetic modulation regarding the heart (PEP value). Our results of reduced vagal activity observed at the pupil might point to a central origin. It has been described that cholinergic neurons of the Edinger–Westphal nucleus control the tension of the sphincter muscle of the iris (Loewenfeld, 1999). Edinger–Westphal nucleus neurons are spontaneously active pacemaker cells that have a high intrinsic firing rate, leading to pupillary constriction (Ichinohe and Shoumura, 2001). Inhibition of the Edinger–Westphal nucleus results in pupillary dilation. Thus, our results of significantly decreased pupil sizes might suggest an increased activity of these pacemaker cells, while reduced light reflex parameters might point toward a reduced activity of these

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

147

Fig. 2. Significantly increased values of heart rate (A), systolic blood pressure (B) and the respiratory rate were obtained during 30 min in patients during AWS. Heart rate and respiratory rate of patients decreased significantly after 24 h. Reduced vagal modulation during AWS in patients is indicated by reduced baroreflex sensitivity (D) and reduced compression entropy (E). The paired t-test shows some increase after 24 h. Significantly increased sympathetic modulation at the heart is indicated by reduced values of the pre-ejection period in patients (F). ***p < 0.001; **p < 0.01; *p < 0.05, n.s. not significant.

148

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

Table 3 Results of pupillary light reflex (left eye) and cardiovascular parameters during acute alcohol withdrawal syndrome and after 24 h. Controls (C) Cardiovascular parameters (mean ± standard deviation) 0.27 ± 0.15 HFn LF/HFHRV 4.47 ± 5.00 bslope (ms/mm Hg) 13.59 ± 6.39 80.62 ± 10.1 dBP (mm Hg) 7.99 ± 4.5 LF/HFBPV

unmedicated patients (UP) 0.23 4.94 7.57 95.07 4.21

± ± ± ± ±

0.13 3.94 6.07 12.02 3.0

Pupillary light reflex parameters (mean ± standard deviation) of the left eye 5.91 ± 5.99 5.00 ± 0.85 Pupillary diameter (mm) Latency (ms) 0.26 ± 0.26 0.28 ± 0.03 Constriction velocity (mm/s) 4.58 ± 4.64 3.85 ± 1.16 Relative amplitude (%) 24.27 ± 24.13 24.30 ± 6.65

C vs. UPF; p value

medicated patients (MP)

UP vs. MP p value

0.01; p = 0.9 0.24; p = 0.623 17.5; p < 0.001 23.9; p < 0.001 8.21; p < 0.006

0.31 3.26 9.34 90.33 4.16

± ± ± ± ±

0.14 2.53 6.88 10.35 4.7

p = 0.9 p < 0.03 p < 0.03 p < 0.04 p = 0.67

8.06; p < 0.006 6.97; p < 0.01 6.12; p < 0.016 0.12; p = 0.91

5.38 0.28 3.59 22.45

± ± ± ±

0.94 0.03 1.18 7.03

p < 0.003 p < 0.004 p < 0.024 p < 0.04

HFn: normalized high frequency of heart rate variability; LF/HFHRV : low frequency–high frequency ratio of heart rate variability; bslope: bradycardic slope of baroreflex sensitivity; dBP: diastolic blood pressure; LF/HFBPV : low frequency–high frequency ratio of blood pressure variability; Pupillary diameter: greatest horizontal diameter of the darkness-adapted pupil; latency: elapsed time between light stimulus and beginning of constriction; relative amplitude: amplitude (difference between baseline pupil diameter maximal constriction) divided by baseline diameter; constriction velocity: gradient of pupillometric graph in the constriction phase at the 40–80% interval of the amplitude.

Fig. 3. Values of the pupillary light reflex of the right eye are displayed. Significantly reduced pupil diameters in patients are depicted (A) during acute AWS with some increase after 24 h. The latency (elapsed time between light presentation and beginning of constriction) is prolonged in patients (B) indicating reduced vagal modulation. Similarly, the reduced constriction velocity indicates reduced vagal tone at the pupil (C). The relative amplitude (difference between baseline pupil diameter and pupil diameter after constriction in relation to the baseline value) indicated no difference between vagal modulation in acute patients and controls (D). The decrease after 24 h might indicate a reduction in vagal tone. ***p < 0.001; **p < 0.01, *p < 0.05; n.s. not significant.

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

149

Fig. 4. The close correlation between pupil diameter and baroreflex sensitivity (A) as well as pupil diameter and compression entropy (B) in controls is depicted in red circles. No such correlation was observed in patients during alcohol withdrawal (black triangles, A and B).

Table 4 Correlations between cardiovascular and puillometric parameters. Pupil diameter (left eye) HFn

Controls Unmedicated patients tslope Controls Unmedicated patients Hc Controls Unmedicated patients

Comparison of correlation coefficients r = 0.491** r = 0.101 r = 0.564** r = 0.058 r = 0.422* r = 0.129

Latency (left eye) p = 0.055 p = 0.017 p = 0.12

Comparison of correlation coefficients r = −0.445* r = 0.319 r = −0.525* r = 0.071 r = −0.453* r = 0.324

p = 0.003 p = 0.016 p = 0.0024

Hc: non-linear compression entropy of heart rate variability; baroreflex sensitivity: tachycardic slope: tslope; HFn: normalized high frequencies; significant correlations according to Spearman rank correlation (*p < 0.05; **p < 0.01) are depicted in bold; correlation coefficients were compared using the Fisher r-to-z transformation (*p < 0.05).

pacemaker cells. To discuss this contradiction a possible involvement of some brain regions such as the locus coeruleus (LC), the ventral tegemental area (VTA) and nuclei in the periaqueductal grey (PAG) matter should be taken into account. The LC is an important pre-autonomic nucleus exerting an excitatory influence on sympathetic and an inhibitory influence on parasympathetic outflow (Szabadi and Bradshaw, 1996). Therefore, the general autonomic consequences of LC activation are pupil

dilatation, inhibition of the kinetics of the light reflex responses, increases in blood pressure and heart rate (Hou et al., 2005; Szabadi and Bradshaw, 1996). Thus, apart from small pupils, all peripheral signs of LC activation are typical signs of alcohol withdrawal and previous studies have described increased LC activity during alcohol withdrawal (Engberg and Hajos, 1992; Knapp et al., 1998). To understand enlarged pupils, it might be necessary to comprehend that separate populations of LC neurons may function as sym-

150

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151

pathetic and parasympathetic pre-motor neurons (Samuels and Szabadi, 2008a,b). Thus, we would like to speculate that sympathetic pre-motor neurons of LC for the pupil might be inhibited resulting in reduced pupillary diameter while parasympathetic pre-motor neurons dwarf the activity of Edinger–Westphal pacemaker cells leading to prolonged latencies. Nevertheless, this assumption needs to be verified by future studies. A further area of interest is the ventral tegmental area (VTA). An excitatory influence of a meso-pupillomotor pathway, originating from the dopaminergic neurons of the VTA and projecting to the Edinger–Westphal nucleus has been proposed (Samuels et al., 2006). Increased firing of DA neurons is associated with increased pupil sizes. Reduced firing rates of spontaneously active DA neurons in the VTA during acute alcohol withdrawal might suggest a reduced influence of this meso-pupillomotor pathway on Edinger–Westphal neurons leading to the observed smaller pupils (Diana et al., 1992; Shen and Chiodo, 1993). However, the explanatory power of a simple interaction between VTA and Edinger–Westphal neurons is limited since we have not observed increased parasympathetic modulation of the pupillary light reflex. In addition, activity of nuclei in the PAG might influence the observed pattern of smaller pupils. It has been demonstrated that neurons present in the PAG have an inhibitory effect on the Edinger–Westphal nucleus (Smith et al., 1968). These neurons produce pupillary dilation when stimulated electrically (Smith et al., 1968). In addition, opioid induced miosis is caused by the activation of GABAergic inhibition on these neurons in the PAG (Bokoch et al., 2015; Vaughan et al., 1997). Possibly, alcohol withdrawal is involved in regulatory dysbalances between GABAergic and glutamatergic neurons in the PAG. Taking this evidence together, our results obtained for the pupil are likely to show summarized effects of various sources and do not mirror a simple effect of increased neuronal activity of either the LC, the VTA or the PAG on Edinger–Westphal neurons. The re-examination of patients after 24 h during clomethiazole treatment indicated that most measures of pupillary function seem to normalize during the medicated state. Although the specific effects of clomethiazole on pupillary function are unknown, it is tempting to attribute observed changes after 24 h to the reduction of withdrawal symptoms (Table 2). Observed changes at the re-examination, however, indicate that small pupils during acute alcohol withdrawal are not the result of chronic changes of neurotransmitters or altered interactions of brain regions due to the long-lasting use of alcohol. Results of heart rate variability (HRV) and blood pressure variability (BPV) are comparable with results obtained in previous studies by our group (Bär et al., 2006a,b). Both HRV and BPV results do not indicate the expected increase in sympathetic activity unequivocally. Only, the pre-ejection period, one of the best sympathetic parameters obtained at the heart, shows increased sympathetic modulation. In contrast, BPV results are not in favor of this assumption. However, the question still remains, which other mechanisms apart from glucocorticoids and catecholamines might have the ability to drive heart rate and blood pressure. In healthy subjects, we observed clear correlations between pupillary function and cardiovascular regulation (Table 4). As previously reported (Bär et al., 2009), some of these associations are counter intuitive. However, it is interesting that no such relation was observed during alcohol withdrawal in our patients. This might point to regulatory diversities regarding the pupil and heart during withdrawal and argues against a simple sympathetic overstimulation. Some limitations need to be addressed. Only male patients were investigated during alcohol withdrawal. Results can therefore not be generalized to both genders and polysubstance dependent individuals. Furthermore, controls were only matched with respect to body mass index and age and the achieved level of education.

In addition, patients were only investigated in the resting state and not during stimulation paradigms. Moreover, pupil parameters were obtained not simultaneously with cardiovascular indices. In conclusion, reduced sympathetic and vagal modulation regarding the pupil accompanies autonomic dysfunction for the heart characterized by low vagal modulation and some increase in sympathetic function. A further elucidation of mechanisms might help to understand and prevent the development of severe autonomic dysfunction during delirium tremens. Role of funding Source This work was funded by internal funds of the University hospital Jena only. Conflict of interest There is no potential or actual conflict of interest for any of the authors. Contributors Thomas Jochum contributed on the design of the study, patient recruitment and writing of the manuscript. Johannes Hoyme and Markus Weißenfels contributed on measurement of physiological parameters. Steffen Schulz contributed on calculation of HRV, BPV and BRS parameters. Andreas Voss contributed on supervision of HRV, BPV and BRS parameters. Karl-Jürgen Bär contributed on design of the study, writing of the manuscript and supervision of the study. All authors contributed to and have approved the final manuscript. References Bär, K.J., Greiner, W., Jochum, T., Friedrich, M., Wagner, G., Sauer, H., 2004. The influence of major depression and its treatment on heart rate variability and pupillary light reflex parameters. J. Affect. Disord. 82, 245–252. Bär, K.J., Boettger, M.K., Till, S., Dolicek, J., Sauer, H., 2005. Lateralization of pupillary light reflex parameters. Clin. Neurophysiol. 116, 790–798. Bär, K.J., Boettger, M.K., Boettger, S., Groteluschen, M., Neubauer, R., Jochum, T., Baier, V., Sauer, H., Voss, A., 2006a. Reduced baroreflex sensitivity in acute alcohol withdrawal syndrome and in abstained alcoholics. Drug Alcohol Depend. 85, 66–74. Bär, K.J., Boettger, M.K., Neubauer, R., Groteluschen, M., Jochum, T., Baier, V., Sauer, H., Voss, A., 2006b. Heart rate variability and sympathetic skin response in male patients suffering from acute alcohol withdrawal syndrome. Alcohol. Clin. Exp. Res. 30, 1592–1598. Bär, K.J., Boettger, M.K., Koschke, M., Boettger, S., Groteluschen, M., Voss, A., Yeragani, V.K., 2007. Increased QT interval variability index in acute alcohol withdrawal. Drug Alcohol Depend. 89, 259–266. Bär, K.J., Boettger, M.K., Schulz, S., Harzendorf, C., Agelink, M.W., Yeragani, V.K., Chokka, P., Voss, A., 2008a. The interaction between pupil function and cardiovascular regulation in patients with acute schizophrenia. Clin. Neurophysiol. 119, 2209–2213. Bär, K.J., Boettger, M.K., Schulz, S., Neubauer, R., Jochum, T., Voss, A., Yeragani, V.K., 2008b. Reduced cardio-respiratory coupling in acute alcohol withdrawal. Drug Alcohol Depend. 98, 210–217. Bär, K.J., Schulz, S., Koschke, M., Harzendorf, C., Gayde, S., Berg, W., Voss, A., Yeragani, V.K., Boettger, M.K., 2009. Correlations between the autonomic modulation of heart rate, blood pressure and the pupillary light reflex in healthy subjects. J. Neurol. Sci. 279, 9–13. Banger, M., Philipp, M., Herth, T., Hebenstreit, M., Aldenhoff, J., 1992. Development of a rating scale for quantitative measurement of the alcohol withdrawal syndrome. Eur. Arch. Pychiatr. Clin. Neurosci. 241, 241–246. Baumert, M., Baier, V., Haueisen, J., Wessel, N., Meyerfeldt, U., Schirdewan, A., Voss, A., 2004. Forecasting of life threatening arrhythmias using the compression entropy of heart rate. Method Inform. Med. 43, 202–206. Berntson, G.G., Cacioppo, J.T., Quigley, K.S., Fabro, V.T., 1994. Autonomic space and psychophysiological response. Psychophysiology 31, 44–61. Berntson, G.G., Bigger Jr., J.T., Eckberg, D.L., Grossman, P., Kaufmann, P.G., Malik, M., Nagaraja, H.N., Porges, S.W., Saul, J.P., Stone, P.H., van der Molen, M.W., 1997. Heart rate variability: origins, methods, and interpretive caveats. Psychophysiology 34, 623–648.

T. Jochum et al. / Drug and Alcohol Dependence 159 (2016) 142–151 Bertinieri, G., di Rienzo, M., Cavallazzi, A., Ferrari, A.U., Pedotti, A., Mancia, G., 1985. A new approach to analysis of the arterial baroreflex. J. Hypertens. 3, S79–S81. Bokoch, M.P., Behrends, M., Neice, A., Larson, M.D., 2015. Fentanyl, an agonist at the muopioid receptor, depresses pupillary unrest. Auton. Neurosci. Basic 189, 68–74. Cacioppo, J.T., Uchino, B.N., Berntson, G.G., 1994. Individual differences in the autonomic origins of heart rate reactivity: the psychometrics of respiratory sinus arrhythmia and preejection period. Psychophysiology 31, 412–419. Castro, J.J., Pozo, A.M., Rubino, M., Anera, R.G., Jimenez Del Barco, L., 2014. Retinal-image quality and night-vision performance after alcohol consumption. J. Ophthalmol. 2014 (704823). Diana, M., Pistis, M., Muntoni, A., Rossetti, Z.L., Gessa, G., 1992. Marked decrease of A10 dopamine neuronal firing during ethanol withdrawal syndrome in rats. Eur. J. Pharmacol. 221, 403–404. Eisenhofer, G., Szabo, G., Hoffman, P.L., 1990. Opposite changes in turnover of noradrenaline and dopamine in the cns of ethanol-dependent mice. Neuropharmacology 29, 37–45. Engberg, G., Hajos, M., 1992. Alcohol-withdrawal reaction as a result of adaptive-changes of excitatory amino-acid receptors. Naunyn Schmiedebergs Arch. Pharmacol. 346, 437–441. Feuerlein, W., Ringer, C., Kufner, H., Antons, K., 1979. The diagnosis of alcoholism—the Munich Alcoholism Test (MALT). Int. J. Rehabil. Res. 2, 533–534. Floras, J.S., 2013. Blood pressure variability: a novel and important risk factor. Can. J. Cardiol. 29, 557–563. Hawley, R.J., Nemeroff, C.B., Bissette, G., Guidotti, A., Rawlings, R., Linnoila, M., 1994. Neurochemical correlates of sympathetic activation during severe alcohol-withdrawal. Alcohol. Clin. Exp. Res 18, 1312–1316. Heller, P.H., Perry, F., Jewett, D.L., Levine, J.D., 1990. Autonomic components of the human pupillary light reflex. Invest. Ophthalmol. Vis. Sci. 31, 156–162. Herbsleb, M., Schulz, S., Ostermann, S., Donath, L., Eisentrager, D., Puta, C., Voss, A., Gabriel, H.W., Bär, K.J., 2013. The relation of autonomic function to physical fitness in patients suffering from alcohol dependence. Drug Alcohol Depend. 132, 505–512. Hou, R.H., Freeman, C., Langley, R.W., Szabadi, E., Bradshaw, C.M., 2005. Does modafinil activate the locus coeruleus in man? Comparison of modafinil and clonidine on arousal and autonomic functions in human volunteers. Psychopharmacology 181, 537–549. Ichinohe, N., Shoumura, K., 2001. Marked miosis caused by deafferenting the oculomotor nuclear complex in the cat. Auton. Neurosci. 94, 42–45. Irwin, M.R., Valladares, E.M., Motivala, S., Thayer, J.F., Ehlers, C.L., 2006. Association between nocturnal vagal tone and sleep depth, sleep quality, and fatigue in alcohol dependence. Psychosom. Med. 68, 159–166. Jochum, T., Reinhard, M., Boettger, M.K., Piater, M., Bär, K.J., 2010. Impaired cerebral autoregulation during acute alcohol withdrawal. Drug Alcohol Depend. 110, 240–246. Jochum, T., Weissenfels, M., Seeck, A., Schulz, S., Boettger, M.K., Voss, A., Bär, K.J., 2011. Endothelial dysfunction during acute alcohol withdrawal syndrome. Drug Alcohol Depend. 119, 113–122. Jochum, T., Schulz, S., Schein, M., Schroder, R., Voss, A., Bär, K.J., 2012. Heart rate turbulence during acute alcohol withdrawal syndrome. Drug Alcohol Depend. 122, 253–257. Kahkonen, S., 2004. Mechanisms of cardiovascular dysregulation during alcohol withdrawal. Prog. Neuropsychopharmacol. Biol. Psychiatry 28, 937–941. Karpyak, V.M., Romanowicz, M., Schmidt, J.E., Lewis, K.A., Bostwick, J.M., 2014. Characteristics of heart rate variability in alcohol-dependent subjects and nondependent chronic alcohol users. Alcohol. Clin. Exp. Res. 38, 9–26. Keivanidou, A., Fotiou, D., Arnaoutoglou, C., Arnaoutoglou, M., Fotiou, F., Karlovasitou, A., 2010. Evaluation of autonomic imbalance in patients with heart failure: a preliminary study of pupillomotor function. Cardiol. J 17, 65–72. Knapp, D.J., Duncan, G.E., Crews, F.T., Breese, G.R., 1998. Induction of Fos-like proteins and ultrasonic vocalizations during ethanol withdrawal: further evidence for withdrawal-induced anxiety. Alcohol. Clin. Exp. Res. 22, 481–493. Levy, D.M., Rowley, D.A., Abraham, R.R., 1992. Portable infrared pupillometry using Pupilscan: relation to somatic and autonomic nerve function in diabetes mellitus. Clin. Auton. Res. 2, 335–341.

151

Loewenfeld, I., 1999. The Pupil: Anatomy, Physiology and Clinical Applications. Butterworth-Heinemann, Woburn, MA, 01801. Malberg, H., Wessel, N., Schirdewan, A., Osterziel, K.J., Voss, A., 1999. Dual sequence method for analysis of spontaneous baroreceptor reflex sensitivity in patients with dilated cardiomyopathy. Z. Kardiol. 88, 331–337. Newlin, D.B., Levenson, R.W., 1979. Pre-ejection period: measuring beta-adrenergic influences upon the heart. Psychophysiology 16, 546–552. Parati, G., Ochoa, J.E., Bilo, G., 2012. Blood pressure variability, cardiovascular risk, and risk for renal disease progression. Curr. Hypertens. Rep. 14, 421–431. Penaz, J., Voigt, A., Teichmann, W., 1976. Contribution to the continuous indirect blood pressure measurement. Z. Gesamte Inn. Med. 31, 1030–1033. Potter, J.F., Bannan, L.T., Saunders, J.B., Ingram, M.C., Beevers, D.G., 1983. Blood pressure and pressor mechanisms during alcohol withdrawal. J. Hypertens. 1 (Suppl), 97–99. Potter, J.F., Bannan, L.T., Beevers, D.G., 1984. The effect of a non-selective lipophilic beta-blocker on the blood-pressure and noradrenaline, vasopressin, cortisol and renin release during alcohol withdrawal. Clin. Exp. Hypertens. A 6, 1147–1160. Quintana, D.S., Guastella, A.J., McGregor, I.S., Hickie, I.B., Kemp, A.H., 2013. Heart rate variability predicts alcohol craving in alcohol dependent outpatients: further evidence for HRV as a psychophysiological marker of self-regulation. Drug Alcohol Depend. 132, 395–398. Ryabinin, A.E., Cocking, D.L., Kaur, S., 2013. Inhibition of VTA neurons activates the centrally projecting Edinger–Westphal nucleus: evidence of a stress-reward link? J. Chem. Neuroanat. 54, 57–61. Samuels, E.R., Szabadi, E., 2008a. Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function part I: principles of functional organisation. Curr. Neuropharmacol. 6, 235–253. Samuels, E.R., Szabadi, E., 2008b. Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function part II: physiological and pharmacological manipulations and pathological alterations of locus coeruleus activity in humans. Curr. Neuropharmacol. 6, 254–285. Samuels, E.R., Hou, R.H., Langley, R.W., Szabadi, E., Bradshaw, C.M., 2006. Comparison of pramipexole and modafinil on arousal, autonomic, and endocrine functions in healthy volunteers. J. Psychopharmacol. 20, 756–770. Shen, R.Y., Chiodo, L.A., 1993. Acute withdrawal after repeated ethanol treatment reduces the number of spontaneously active dopaminergic-neurons in the ventral tegmental area. Brain Res. 622, 289–293. Smith, S.A., Smith, S.E., 1983. Reduced pupillary light reflexes in diabetic autonomic neuropathy. Diabetologia 24, 330–332. Smith, J.D., Ichinose, L.Y., Masek, G.A., Watanabe, T., Stark, L., 1968. Midbrain single units correlating with pupil response to light. Science 162, 1302. Smith, J.J., Muzi, M., Barney, J.A., Ceschi, J., Hayes, J., Ebert, T.J., 1989. Impedance-derived cardiac indices in supine and upright exercise. Ann. Biomed. Eng. 17, 507–515. Szabadi, E., Bradshaw, C.M., 1996. Autonomic pharmacology of alpha(2)-adrenoceptors. J. Psychopharmacol. 10, 6–18. Vaughan, C.W., Ingram, S.L., Connor, M.A., Christie, M.J., 1997. How opioids inhibit GABA-mediated neurotransmission. Nature 390, 611–614. Wetterling, T., Kanitz, R.D., Besters, B., Fischer, D., Zerfass, B., John, U., Spranger, H., Driessen, M., 1997. A new rating scale for the assessment of the alcohol-withdrawal syndrome (AWS scale). Alcohol Alcohol. 32, 753–760. Whitworth, J.A., Mangos, G.J., Kelly, J.J., 2000. Cushing, cortisol, and cardiovascular disease. Hypertension 36, 912–916. Ziv, J., Lempel, A., 1977. Universal algorithm for sequential data compression. IEEE Trans. Inf. Theory 23, 337–343. Zulfiqar, U., Jurivich, D.A., Gao, W., Singer, D.H., 2010. Relation of high heart rate variability to healthy longevity. Am. J. Cardiol. 105, 1181–1185. de Zambotti, M., Willoughby, A.R., Baker, F.C., Sugarbaker, D.S., Colrain, I.M., 2015. Cardiac autonomic function during sleep: effects of alcohol dependence and evidence of partial recovery with abstinence. Alcohol 49, 409–415.