Heart rate variability and its impact on pulmonary dysfunction in rats subjected to hemorrhagic shock

Introduction: The activity of autonomic nervous system and its association with organ damage have not been entirely elucidated in hemorrhagic shock. The aim of this study was to investigate heart rate variability (HRV) and pulmonary gas exchange in hemorrhagic shock during unilateral subdiaphragmatic vagotomy. Methods: Male Sprague Dawley rats were randomly assigned into groups of sham, vagotomized (Vag), HS and Vag+HS. HS was induced in conscious animals by blood withdrawal until reaching to mean arterial blood pressure (MAP) of 40±5 mm Hg. Then, it was allowed to MAP returning toward the basal values. MAP and heart rate (HR) were recorded throughout the experiments, HRV components of low (LF, sympathetic index), high (LH, parasympathetic index), and very low (VLF, injury index) frequencies and the LF/HF ratio (sympathetic index) calculated, and the lung histological and blood gas parameters assessed. Results: In the initial phases of HS, the increase in HR with no change in MAP were observed in both HS and Vag+HS groups, while LF increased only in the HS group. In the second phase, HR and MAP decreased sharply in the HS group, whereas, MAP decreased only in the Vag+HS group. Meanwhile, LF and HF increased relative to their baselines in the HS and Vag+HS groups, even though the values were much pronounced in the HS group. In the third phase, HR, MAP, LF, HF, and the LF/HF ratio were returned back to their baselines in both HS and Vag+HS groups. However, in the Vag+HS group, the VLF was lower and HR was higher than those in the other groups. Furthermore, blood gas parameters and lung histology indicated the impairment of gas exchange in the Vag+HS group. Conclusions : The sympathetic activity is predominant in the first phase, whereas the parasympathetic activity is dominant in the second and third phases of hemorrhagic Furthermore, shock with may be linked to lung injury and


Introduction
Hemorrhagic shock (HS) is one of the leading causes of death in the world [1]. The inability to supply the oxygen demand of the body organs in HS may lead to a severe cell damage, metabolic disorders, or even cell death in untreated patients [2]. HS is categorized into four classes. In classes I and II, despite the blood loss, systolic blood pressure, pulse pressure, and plasma levels of bicarbonate and lactic acid remain in the normal range or may change slightly. The constant blood pressure can be linked to the increase in heart rate (HR) or vascular resistance because of the neural activity and humoral factors [3]. A substantial decrease in cardiac output and blood pressure occurs by 30-40% of blood loss in reversible class III. However, if hemorrhage persists without any therapeutic intervention, the compensatory mechanisms may be failed which results in deadly irreversible shock in class IV [4].
The hemodynamic responses to blood loss in classes I and II of hemorrhagic shock in conscious animals consist of three phases [5]. In the initial compensation phase, the arterial blood pressure is maintained despite a decrease in cardiac output [6]. In the second phase namely decompensatory phase, both arterial blood pressure and HR drop sharply in parallel with losing 15 to 20% of total blood volume [7]. In the third phase or recompensatory phase, the arterial blood pressure, and HR recover gradually when the hemorrhage is terminated [8]. However, mechanisms of different phases of compensation have not been fully elucidated thus far.
The compensatory response to hemorrhagic shock depends on the integrity of afferent fibers of the vagus nerve originate from the arterial baroreceptors [5,9,10]. Furthermore. the subdiaphragmatic vagus nerves may be involved in regulating the autonomic nervous 4 system's negative feedback [11]. It may also play a role in immune responses through anti-inflammatory cholinergic pathways [12]. However, no study has clarified the role of subdiaphragmatic vagus nerves in class II hemorrhagic shock.
Inappropriate treatment of hemorrhagic shock may lead to body tissue and multi-organ failures. The lung is very vulnerable to the damage because it receives the total cardiac output per minute. Therefore, it may be affected by many substances, which are released from the injured tissues passing through the circulation to the lung. In a few studies, the increase in pulmonary capillary permeability and the infiltration of immune cells into the pulmonary interstitial space have been shown in hemorrhagic shock [13]. However, the role of subdiaphragmatic vagus nerve in lung injury induced by hemorrhagic shock remains unclear.
The time intervals between two consecutive heartbeats are frequently influenced by the activity of autonomic nervous system, physical activity, metabolism and homeostatic challenges [14]. These fluctuations are defined by analyzing heart rate variability (HRV) [15]. The analyses of HRV may provide useful information about the autonomic nervous system activity in three phases of compensation in HS. To date, a few studies have indicated HRV can reflect the autonomic activity variations in the early and compensatory phases of shock [16]. A study by Porter et al. identified that HRV depends on the rate of bleeding [17]. HF band is an index for parasympathetic activity and respiratory cyclerelated heart rate [18]. LF band is an index of the activities of sympathetic system and baroreceptors [15,19]. The LF/HF ratio is used to assess the sympathetic balance and VLF is related to the metabolic regulatory mechanisms and injury [20,21].
Based on the results of the noted researches, the present study aimed at investigating the alterations of hemodynamic and HRV at three phases of class II hemorrhagic shocks in conscious rats. The relation between the HRV, lung tissue damage and blood gas 5 parameters in the three phases of hemorrhagic shock were evaluated. Furthermore, the role of the subdiaphragmatic vagus nerve was assessed at the above conditions. This study was designed in conscious rats to exclude the effect of anesthetic drugs on the respiratory system-related heart rate variability.

Study design
This study was conducted based on the Medical Ethics Committee Regulations in accordance with the declaration of Helsinki on 24 male Sprague-Dawley rats weighing 250-300g. Animals were housed in standard cages under controlled laboratory temperature, humidity, and 12:12 hours of light/dark cycles. They had free access to water and standard food, and were fasted a few hours before starting the experiments, and randomly assigned into four groups of Sham (n = 5), vagotomized (Vag, = 5), hemorrhagic shock without (HS, n = 7) and with vagotomy (HS+Vag, n = 7). Animals were anesthetized by intraperitoneal injection of 50 mg/kg sodium pentobarbital (Sigma).
Additional doses used if needed. The femoral vein was then cannulated by a 120-PE catheter. It was passed through a hole in the rear skin of the tail, and fixed by two sutures. Then, a 50 PE catheter was inserted into the tail artery, and fixed firmly.
Thereafter, the area of surgery was rinsed by 1% lidocaine (Sigma) to minimize the postoperative pain. Then, the conscious animals were transferred to a dark metabolic cage with a constant temperature. In order to manage the stress and anxiety of animals, the experiments were performed under a calm with minimum noise conditions. Animals' tails were fixed outside the cage. Therefore, they could relatively move in a cage without damaging the catheter, and disturbing the recording setup. The arterial catheter was connected through a pressure transducer (MLT844) to a data acquisition system 6 (Powerlab, PL26T04, ADinstruments, Australia). The mean arterial blood pressure (MAP) was recorded throughout the experiments and HR calculated accordingly. The femoral vein was used for blood sampling, and blood withdrawal during induction of hemorrhagic shock.

Subdiaphragmatic vagotomy
After anesthesia, an incision was made in the upper abdominal skin. The fascia and muscles were dissected, and the left subdiaphragmatic vagus nerve, separated from the surrounding tissues, and cut apart. Then, muscles and skin were sutured, and the surgical areas were rinsed with 1% lidocaine. In the Sham group, animals underwent the arterial and venous cannulation and abdominal laparotomy, whereas, the vagus nerve remained intact.

Study protocol: Induction of hemorrhagic shock
After 70 minutes of the steady-state period, the arterial and venous blood samples were taken for analyzing the blood gas parameters. Then, groups of HS and Vag+HS were subjected to hemorrhagic shock, according to the previous studies and our pilot experiments [17]. Briefly, the blood withdrawal with a constant flow rate (0.5 ml/min/rat) was taken from the femoral vein until the mean blood pressure reached to 40± 5 mm Hg.
After that, the blood withdrawal was ceased so that it was allowed MAP returned to the pre-hemorrhage level. The amount of blood volume (BV) was estimated using the Lee and Blaufox equation: EBV ml = 0.06×body weight g+0.77 [22].
Two hours later, the arterial and venous blood samples were taken for blood gas analysis.
MAP and HR were recorded continuously throughout the experiments. Finally, animals were killed by high doses of pentobarbital and the lungs were prepared for histological analyses. All mentioned procedures were performed in the groups of Sham and Vag in the 7 same time course as in the HS groups, except for the induction of hemorrhagic shock ( Figure 1).

HRV Analysis
Kubios HRV Premium/Animal software (ver. 3.2) was used for HRV analysis. The interpolated techogram of 10 Hz [23], the Welch's period diagram window width for 512 pieces and the hanning window with 50% overlap were selected. Based on previous studies, frequency-domain components of very low (VLF: 0 to 0.2 Hz), low (LF: 0.20 to 0.75 Hz), high-frequency (HF: 0.75 to 3.0 Hz) [24,25], and LF/HF ratio were considered in this study. We used HRV analysis in three phases of class II hemorrhagic shock. The duration of calculations in the first and third phases was within 5 minutes, and in the second phase was for one minute. At each time interval, the pulse interval was calculated offline from the intervals of the peak systolic blood pressure. Interval pulse data were exported from the Kubios software to the Excel. Finally, the HRV components of LF, HF, and VLF data were expressed as logarithm values (log).

Arterial and venous blood gas parameters
The 200 µl of blood samples were taken during the baseline period, and at the end of the experiments for the blood gas analyses using an easy blood gas device (Medica, USA). The arterial and venous oxygen contents were calculated based on the following equation.

Histological evaluations
At the end of the experiments, the chest was opened and the lung removed and fixed in 4% formalin. Then, the samples were dehydrated by different concentrations of ethanol 8 and xylol and embedded in paraffin. Tissue sections were prepared by a microtome (pfm medical, UK) and stained with hematoxylin and eosin [13]. All slides of histology were evaluated in a blinded manner by a pathologist. The pathological components of the focal thickening of the alveolar membranes, vascular congestion, perivascular neutrophil infiltration, and alveolar hemorrhage were evaluated. Each index was graded with a score of 0 to 3 based on the absence (0), mild (1), moderate (2) or severe (3) [13].

Statistical Analysis
Data are given as mean±SE. Analysis of variance (ANOVA) with Tukey's post hoc test was used for multiple comparisons. Also, repeated-measure ANOVA was used for comparison of data during the different time course of experiments. We used the paired T-Test for comparing the data of blood gas variables at the onset and offset of the experiments, and other variables if needed. Furthermore, we compared data of histology using the nonparametric Mann-Whitney test. All analysis was performed using the software of SPSS 18.
Significance was assumed when P<0.05 and the confidence limits used were the 95% intervals.

Results
Animals in the HS and Vag+HS groups underwent hemorrhage with the volume of 26.07±3.07% and 23.37±1.68% of EBV, respectively. Also, the duration of blood withdrawal were 8.92 ± 0.94 and 8.42 ± 0.52 minutes in the HS and Vag+HS groups, respectively. No significant difference was detected in the amount or duration of blood withdrawal at above-mentioned groups. Also, the pattern of those variables in the In-Com phase was similar to those in the HS group. That is, MAP was maintained and HR was higher than those in their baselines (p˂0.01). In the De-Com phase, alterations of MAP in the Vag+HS group were similar to that in the HS group. However, unlike the HS group, the HR in the Vag+HS group did not show a significant difference with respect to its baseline, or those ones in the Sham and The LF/HF ratio showed a significant reduction in the De-Com phase of the Vag+HS group as compared with the other groups (p˂0.05), and was higher than the baseline in minutes 100-110 (p˂0.05) (Figure 3). Table 1, and Table 2  At the end of the experiments, the Hb concentrations in both HS groups were lower than those in the Vag and Sham groups. However, it was only significant in the Vag+HS group compared with the Vag group. The PaO 2 /FIO 2 ratio in the Vag+HS group was lower than that in the Sham group (normal value of PaO 2 /FIO 2 ratio is about 300±10 mm Hg) (p˂0.05). Also, the decrease in PaO 2 were along with the decrease in arterial oxygen content (CaO 2 ) (p˂0.05, compared with the Sham and Vag groups). In addition, the venous oxygen content (CvO 2 ) in the HS group was lower than those in the Sham and Vag group (p˂0.05) and CvO 2 in the Vag+HS group was lower than those in the Sham and Vag group (P˂0.01) ( Table 3).
Vascular congestion in the Vag+HS group was higher than those in the Sham, Vag and HS groups (p˂0.05). There was no alteration between histological parameters between the Sham, Vag, and Vag+HS groups ( Figure 4).

Discussion
A few studies have reported the spectral analysis of HRV, and the role of subdiaphragmatic vagus nerve in class II hemorrhagic shock in conscious animals.
Besides, the impact of HRV and acute lung injury has not been illustrated at the mentioned conditions [26]. Therefore, in the present study, hemodynamic parameters and HRV were evaluated during three phases of class II hemorrhagic shock. Furthermore, the association between the level of the autonomic nervous system activity and lung injury was evaluated. The contribution of autonomic nervous system differs in three phases of class II hemorrhagic shock. The sympathetic activity in the In-Com phase shifts to the parasympathetic activity in the De-Com phase. Also, the parasympathetic activity is much pronounced in the Re-Com phase of hemorrhagic shock. Although subdiaphragmatic vagotomy does not change the baseline hemodynamic parameters, it increases the heart rate during the hemorrhagic shock, particularly in the Re-Com phase, and leads to lung injury probably linked to interfering with anti-inflammatory effect of the cholinergic pathway.
No alteration was identified in the values of MAP and HR at baseline. Therefore, all groups enter the study with identical conditions. In addition, no variation in the MAP and HR between the vagotomized and non-vagotomized animals, suggest the subdiaphragmatic vagus nerve may not regulate directly the systemic hemodynamic. In fact, in order to minimize the direct effect of the vagus nerve on hemodynamic and heart rate variability components, we did subdiaphragmatic vagotomy instead of cervical vagotomy. Likewise, no alteration was detected in MAP and HR during the time course of experiments in the Sham and Vag groups. Therefore, the unilateral subdiaphragmatic vagotomy per se could not affect the hemodynamic and cardiac function.
The constant arterial blood pressure in the In-Com phase of blood withdrawal in the HS group can be due to the increase in HR, which is also supported by the results of previous studies [5]. Other studies have indicated that the systemic vascular resistance, but not HR, increases markedly in conscious rats subjected to hemorrhagic shock [8]. However, in our study, we did not measure vascular resistance because of limitations in the experimental preparations. Therefore, it can be suggested that at least a part of a constant blood pressure is linked to the increase in HR. There was not a significant variation in the parasympathetic component of HF, though it tended to increase. However, a significant increase in LF may be due to the increase in sympathetic activity [19]. On the other hand, the LF/HF ratio showed no change, which can be related to an insignificant increase in HF. Together, it is expected that the parasympathetic activity decreases and the sympathetic activity increases in this phase. In line with our results, Porter et al.
showed that the increase of HR in the compensatory phase of hemorrhage, is mainly mediated by the increase in sympathetic activity, but not the parasympathetic drive [17].
By contrast, Troy et al have reported that the constant heart rate in the initial phase of hemorrhagic shock is related to the decrease in parasympathetic drive, with no change in 13 the sympathetic activity [5]. The difference in the experimental conditions may be influential in this dissimilarity results.
In our study, the unilateral subdiaphragmatic vagotomy did not alter the MAP in the In-Com phase in the Vag+HS group, whereas it increased HR. There was no difference between LF and HF in the Vag+HS group compared with those in the Vag and Sham groups. Therefore, the increase in HR may be linked to the decrease in parasympathetic activity, increase in sympathetic activity, or both.
The blood withdrawal resulted in a sudden drop in MAP and HR in the De-Com phase of the HS group similar to some studies [5], whereas, others have reported the opposite results [27]. HF increased in the De-Com phase of the HS group in agreement with the report of Porter et al [17]. Besides, an increase in LF was observed in the De-Com phase of HS in our study, causing the LF/HF ratio to remain unchanged. The increase in LF may be related to the increase in sympathetic activity. This conclusion is in contrast with the reports of Potas & Dampney that the sympathetic system is inhibited in hemorrhagic shock linked to the reduction of venous return and cardiac filling, and leads to the reduction of heart rate [28]. However, the question remains why the increase in LF was not associated with the increase of HR in our study. There are some evidences, indicating that the LF band cannot be considered entirely as a sympathetic index. In one study, it has been reported that when parasympathetic activity is high, the LF band cannot be used as a reliable index for the sympathetic activity [29]. Furthermore, angiotensin II increases a few minutes after induction of hemorrhagic shock. Angiotensin II increases the LF band, whereas, administration of angiotensin II receptor antagonist has an opposite effect [30,31].
Therefore, it can be argued that at least parts of increased LF band in the De-Com phase of both hemorrhagic shock groups are linked to the increases in plasma angiotensin II. To sum up, since the parasympathetic index of HF increases and HR decreases in the De-Com phase of HS, the dominant effect of the parasympathetic activity is suggested in this phase.
Although, the reduction in MAP in the De-Com phase of both HS and Vag+HS groups were similar, HR did not decrease in the Vag+HS group. However, the amount of blood withdrawn in this group was identical to that in the HS group. Therefore, it can be suggested that the main cause of pressure drop in this phase may be the decrease in vascular resistance, but not reduced HR. We did not find a similar study indicating the effect of subdiaphragmatic vagotomy on HRV components during hemorrhagic shock. However, it has been shown that any drug, surgery, or physiological interventions which disrupts the parasympathetic activity could reduce HRV [32]. Also, the subdiaphragmatic vagal nerve contains both afferent and efferent fibers that may play a role in regulating the negative feedback in sympathetic activity and adrenal medullary secretion. Thus, the interference in vagal subdiaphragmatic pathways may increase the plasma epinephrine following the disinhibition of the adrenal gland [11]. Consequently, it may prevent of bradycardia during the De-Com phase of hemorrhage in the Vag+HS group. Furthermore, angiotensin II released in this phase may decrease the R-R interval and increase HR [31].
As a result, increased HR in the De-Com phase of the Vag+HS group may be linked to the effects of angiotensin II and epinephrine. On the other hand, HF and LF were high in the De-Com phase of the Vag+HS group, even though these parameters were lower than those in the HS group. Also, the LF/HF ratio was low in the Vag+HS group. Taken together, the decrease in MAP, with no change in HR, concomitant with the results of HRV analyses suggest the less sympathetic activity compared with the parasympathetic activity in the De-Com phase of class II hemorrhagic shock.
In the third phase, which lasted for 110 minutes, the HR and MAP gradually returned back to the baseline in the HS group, and did not indicate a significant difference with the ones in the Sham group. All components of HRV were gradually recovered. Although at the end of the record, LF appeared to be higher than those in the Sham group and their baselines, these differences were not significant. A few studies have addressed the hemodynamic alterations and HRV in the Re-Com phase of hemorrhagic shock. In the study of Porter et al. the recovery was assessed within 30 minutes after hemorrhage [17], which was not enough time to return MAP and HR to their baselines. Our study was performed in a long time, so that the hemodynamic parameters returned to their baselines and being stable up to the end of the experiments. However, HR increased after minute 10 of Re-Com phase in the Vag+HS group, being significant from 30 minutes to the end of the experiments. There was no significant variation between HF during the time course of Re-Com phase in the Vag+HS group. As a result, the cause of increased HR cannot be directly linked to the decrease in the parasympathetic drive. It may be related to the effect of subdiaphragmatic vagotomy on the medullary adrenal gland and the release of norepinephrine [11]. In addition, the LF/HF ratio at the end of experiments in the Vag+HS group was higher than that in the HS group, which may suggest a gradual increase in sympathetic drive in this time.
Our results showed a significant increase in VLF in the De-Com phase and at the onset of the Re-Com phase, and it returned to the baseline level thereafter in the HS group.
Although, the VLF increased in the De-Com phase of Vag+HS group, it was significantly lower than that in the HS group. However, VLF returned to the baseline level in the Vag+HS group faster than the HS group. Our data suggest that the relationship between the HR and the level of VLF could be inversely proportional. The decrease in VLF leads to the increase in HR. The increase in VLF band may be also due to the increase of the renin angiotensin system activity, temperature regulatory mechanisms [33] or endocrine and hypothalamic systems involved in metabolic rate [21]. Numerous studies have suggested that among the HRV components, the VLF band is a powerful indicator in predicting different diseases and injuries [34]. In our study, VLF in the Re-Com phase of the Vag+HS group was lower than that in the Sham group. Therefore, it can be suggested that there is an inverse relationship between the level of VLF and organ dysfunctions including the lung in hemorrhagic shock.
There was no significant alteration in the blood gas parameters at the end of the experiments in the Sham, Vag and HS groups. However, the subdiaphragmatic vagotomy In the hemorrhagic shock, the arterial-venous oxygen content difference increases due to the increased oxygen demand of the affected tissues. Therefore, the decrease in PvO 2 in the Vag+HS group along with increased the arterial-venous oxygen content difference may be linked to the increase in tissue oxygen uptake by body organs including the heart in order to repaying the oxygen debt [37].

Conclusions
In this study, we indicated the contribution of autonomic nervous system in three phases of class II hemorrhagic shock. The sympathetic system is predominant in the first phase, whereas the parasympathetic system is dominant in the second and third phases of hemorrhagic shock. Also, the subdiaphragmatic vagotomy leads to lung injury in hemorrhagic shock. Taken together, the analyzing the HRV during different phases of hemorrhagic shock may pave the way for rapid diagnosis and treatment of the patients during critical condition in the hospital. Also, the administration of anti-cholinergic drugs may not be recommended because of worsening the lung injury.