Neurohormonal and metabolic effects of medetomidine compared with xylazine in beagle dogs

Abstract

This study was aimed to investigate and compare the effects of medetomidine and xylazine on the blood level of some stress-related neurohormonal and metabolic variables in clinically normal dogs, especially focusing on time and dose relations of the effects. A total of 9 beagle dogs were used for 9 groups, which were treated with physiological saline solution (control), 10, 20, 40, and 80 μg/kg medetomidine, and 1, 2, 4, and 8 mg/kg xylazine, intramuscularly. Blood samples were taken at 10 times during 24 h from a central venous catheter. Plasma norepinephrine, epinephrine, cortisol, glucose, insulin, glucagon, and non-esterified fatty acid concentrations were determined. Both medetomidine and xylazine similarly and dose-dependently inhibited norepinephrine release and lipolysis. Medetomidine suppressed epinephrine release dose-dependently with greater potency than xylazine. Xylazine also tended to decrease epinephrine levels dose-dependently. The cortisol and glucagon levels did not change significantly in any treatment group. Both drugs suppressed insulin secretion with similar potency. Both medetomidine and xylazine increased glucose levels. The hyperglycemic effect of medetomidine, in contrast with xylazine, was not dose-dependent at the tested dosages. The results suggested that the effect of medetomidine on glucose metabolism may not be due only to α₂-adrenoceptor-mediated actions.

Introduction

The α₂-adrenoceptor agonists, medetomidine and xylazine, are widely used in veterinary practice for different purposes. They are mainly used as sedative, muscle relaxant and analgesic agents in anesthesia of different species (1,2). By reducing gastric and intestinal motility, α₂-agonists are useful for gastrointestinal surgery or endoscopy (1). These drugs are also reliable emetics for small animals (3,4). In addition, xylazine is used as a diagnostic agent for congenital or acquired hyposomatotropism in dogs and cats (5).

Although α₂-agonists are multipotent drugs, they should be used carefully, because unexplained and sometimes fatal accidents may be associated with their use in the healthy small animal patient, even without painful intervention (6,7). These are usually associated with the cardiovascular side effects of these drugs (1,2). However, whether the neuroendocrine and metabolic effects of α₂-agonists are involved in the causes of these accidents is not fully understood. There are limited data on the usage of medetomidine and xylazine for anesthesia of patients with cardiovascular or respiratory illness (1,8). In spite of the fact that α₂-agonists strongly interfere with the neuroendocrine system, there is still no report to prove the effects of these agents on different endocrine or metabolic diseases, such as diabetes mellitus, Cushing's, and Addison's diseases. Numerous studies have shown that α₂-agonists such as xylazine and clonidine decrease plasma catecholamine and cortisol levels (9,10,11), inhibit insulin release (12,13) and lipolysis (14,15), and increase plasma glucose (16,17) and glucagon (18) levels in various species. However, specific data on the effects of medetomidine and xylazine especially time and dose relations are still insufficient in dogs. The effects of different dosages of medetomidine and xylazine on basal plasma cortisol and nonesterified fatty acid (NEFA) levels have not yet been examined in dogs. Such a study may provide useful information for the usage of α₂-agonists under different pathologic conditions or stress associated with surgical intervention.

The purpose of this study was to investigate and compare the effects of medetomidine and xylazine on some stress-related neurohormonal and metabolic biochemicals (catecholamines, cortisol, glucose, insulin, glucagon, and NEFA) at blood levels in dogs. This study also aimed to examine the dose relation of the effects induced by the 2 drugs. Because both medetomidine and xylazine exert their actions mainly on α₂-adrenoceptors, we hypothesized that there are no differences between their effects on the blood levels of the examined biochemicals.

Materials and methods

Animals

Nine healthy beagle dogs of either sex, aged 32.2 ± 13.8 mo (mean ± SD) and weighing 11.9 ± 3.1 kg (mean ± SD) were used. All dogs were housed in our laboratory for at least 1 mo before the experiment, and fed standard dry dog food. Routine hematological and plasma biochemical tests had been performed before the experiment. All values were within the normal physiological range. One day before the experiment, the animals were placed into separate cages in the experimental room controlled at 25°C by air conditioning. Food and water were withheld for 12 h before the drug injection, and water was offered again after complete recovery from sedation. The feeding time was always 9 o'clock pm, after the last blood sampling of that day. The experimental protocols were approved by the Animal Research Committee of Tottori University.

Experimental protocol

Two experiments were included in this study. The first experiment consisted of 7 groups with an intramuscular (IM) treatment in each. The treatments were physiological saline solution (0.5 mL); 10, 20, 40, and 80 μg/kg medetomidine HCl (1% solution, Domitor; Meiji Seika Kaisha, Tokyo, Japan); and 1 and 2 mg/kg xylazine HCl (2% solution, Celactal; Bayer, Tokyo, Japan). Five beagle dogs were used repeatedly in each of the 7 groups at weekly intervals, according to a randomized design. The second experiment was designed to provide additional data about the effects of 4 and 8 mg/kg xylazine, especially on blood glucose. Four other beagle dogs were used for the 2 additional groups at a week interval in randomized order. The treatments will be referred to as MED-10, -20, -40, and -80 for the medetomidine-treated groups and XYL-1, -2, -4, and -8 for xylazine-treated groups.

The basis of the comparison of selected dosages in this study was that in our preliminary investigations, the MED-10 and -20 showed similar level and duration of sedation induced by XYL-1 and -2, respectively. Consequently, the sedative effect, when assessed by lateral and thoracic recumbency position, lasted for 65 ± 23 and 61 ± 13 min (mean ± SD) in MED-10, and XYL-1, 112 ± 32 and 116 ± 40 min in MED-20, and XYL-2, 210 ± 35 and 181 ± 65 min in MED-40 and XYL-4, and 270 ± 77 and 209 ± 54 min in MED-80 and XYL-8 groups, respectively. Therefore, these pairs of MED and XYL produced similar duration of sedation in this study. However, higher dosages of xylazine induced more alertness, sometimes anxiety and muscle rigidity. That is why 4 and 8 mg/kg xylazine dosages are not recommended for clinical practice.

This study was designed to model clinical conditions (except for XYL-4 and -8). Because α₂-agonists are more often used IM (7), this route was preferred in our study. The femoral biceps muscle was used for injections. The concentrations of the medicines were kept at commercially recommended levels (1% for medetomidine and 2% for xylazine) as they are normally used in practice. Subsequently, the injected volumes were different between the treatments. This might have affected the speed and completeness of absorption, but this bias was also present under clinical conditions.

Instrumentation

Under local anesthesia with 2% lidocaine (Xylocaine; Fujisawa Pharmaceutical, Osaka, Japan), a 16-gauge central venous (CV) catheter was introduced into the jugular vein up to the central vein area. The CV catheter was flushed with 0.5 mL of heparinized physiological saline solution, capped, and fixed. The catheter was placed in the evening before the experiment and removed after the last blood sampling. There was no remarkable inflammatory sign at the catheterized site during the course of the experiment.

Sample collection

In the first experiment, blood samples were collected from the CV catheter for the following 10 times: 0 (initial value before drug administration), 0.5, 1, 2, 3, 4, 6, 8, 12, and 24 h after drug injection. The initial 1 mL of blood collected from the CV catheter was discarded to avoid contamination by heparin. The following 5 mL of blood was used as a sample. One mL of each blood sample was mixed with Trasylol (Bayer, Leverkusen, Germany) separately for glucagon measurement, and the remaining 4 mL blood was mixed with ethylenediamine tetraacetic acid (EDTA) for other measurements. Both samples were centrifuged immediately at 4°C, then the plasma was separated and frozen at −80°C and analyzed within 3 mo. Plasma catecholamines (norepinephrine and epinephrine), cortisol, glucose, insulin, glucagon, and NEFA were measured in all samples.

The blood samples for the second experiment (XYL-4 and -8) did not include Trasylol samples for glucagon measurements at any time, and the EDTA blood samplings at 12 h were omitted.

Analytical methods

Catecholamines were extracted on activated alumina according to the method described by Bouloux et al (19), and measured by a high performance liquid chromatography (LaChrom; Hitachi, Tokyo, Japan) combined with an electrochemical detector (Coulochem II; ESA, Chelmsford, Massachusetts, USA). Cortisol was measured by single antibody radioimmunoassay (RIA) technique using a commercially available kit (I-AE16; Eiken Chemical, Tokyo, Japan). Insulin and glucagon were measured by double antibody RIA technique (I-AJ16; Eiken Chemical and Glucagon kit Daiichi; TFB Stock Company, Tokyo, Japan). Glucose and NEFA values were determined by use of a spectrophotometer (Auto Sipper Photometer U-1080; Hitachi).

Data evaluation

All data obtained were analyzed together using the statistical software (Statview v.4.1; Abacus Concepts, Berkeley, California, USA). One-way analysis of variance (ANOVA) for repeated measures was used to examine the time effect within each group, and one-way ANOVA for group effect at each time point. When a significant difference was found, the Tukey test was used to compare the means.

Quadratic curves were fitted to the data of each group at the recovery period, and the slope was used to compare the speed of recovery among the treatments. When the time effect was significant by ANOVA and Tukey test (norepinephrine, epinephrine, glucose, insulin, and NEFA), the normalized area under curve (AUC) was calculated. The AUC was measured by calculating the sum of the trapezoids formed by the data points and the x-axis from 0 h to 6 or 8 h. The difference between the mean AUC of the control group and the AUC of a certain individual was defined as the normalized AUC. The normalized AUC data were plotted versus dose, in either XYL or MED treatment, and simple linear regression analysis was applied in each. When significance was found, the effect of XYL or MED on the plasma level of the examined biochemical was claimed to be dose-related.

The normalized AUC data were also used for general comparison of medetomidine and xylazine. The AUC data of every variable were divided into 2 groups, medetomidine and xylazine, regardless of the applicated dose. These 2 groups were compared with unpaired t-test. This method provides an overall comparison between the potency of the effects of medetomidine and xylazine treatments. The level of significance of all tests was set at P 0.05.

Results

The norepinephrine levels significantly decreased at 0.5 h after the injection of medetomidine (Figure 1A). This effect was the shortest in MED-10, in which norepinephrine levels decreased until 1 h and then returned to its initial value gradually. The norepinephrine values in MED-20 at 0.5 to 3 h were significantly lower than the initial value. The norepinephrine release in both MED-40 and -80 treatments was similarly suppressed until 4 h, and then slowly returned to baseline. These findings, the results of the slope of the recovery phase and the regression analysis of the normalized AUC data, all indicated that medetomidine suppressed the norepinephrine release dose-dependently. The norepinephrine values decreased after the xylazine treatments, and remained significantly low until 2 h in XYL-1, 3 h in XYL-2, 4 h in XYL-4, and 6 h in XYL-8 groups (Figure 1B). The value at 24 h increased significantly over baseline in XYL-8. The slopes of the recovery phases and the regression analysis of the AUC determinations proved that xylazine also decreased plasma norepinephrine levels dose-dependently. The overall comparison of the AUC data by t-test indicated that MED and XYL suppressed norepinephrine level to a similar degree.

Figure 1. Plasma norepinephrine levels following the administration of medetomidine (MED μg/kg, A) and xylazine (XYL mg/kg, B) in dogs. a: significantly different from the initial value (P 0.05). Each point and vertical bar represent the mean and the standard error of mean (SEM) (n = 4 to 5).

The mean values of epinephrine in all treatment groups (except for XYL-1) decreased to a significantly low level at 0.5 h and 1 h post-injection and gradually returned to baseline (Figure 2). There was a tendency to decrease at 0.5 and 1 h in XYL-1, but it was not significant. The suppression of epinephrine release was similar in both MED-80 and XYL-8, according to the AUC results, and the significantly decreased levels continued until 4 h in both groups. The linear regression of AUC data was significant in the MED groups, indicating that MED decreased the plasma epinephrine levels dose-dependently. Although the effect of XYL on plasma epinephrine was not significant by regression analysis, the AUC data of epinephrine were significantly different between XYL-1 and XYL-8 by Tukey test. The slopes of the recovery phases also decreased dose-dependently in both MED and XYL. Comparing the AUC data by t-test we found that MED more potentially suppressed epinephrine release than XYL.

Figure 2. Plasma epinephrine levels following the administration of medetomidine (MED μg/kg, A) and xylazine (XYL mg/kg, B) in dogs. a: significantly different from the initial value (P 0.05). Each point and vertical bar represent the mean and SEM (n = 4 to 5).

Glucose values increased significantly after injection in all groups except for both MED-10 and control groups (Figure 3). Values at both 1 and 2 h (124 ± 8 and 123 ± 19 mg/dL, mean ± SD) in MED-10 group tended to increase from the baseline value (103 ± 11 mg/dL), but it was not significant. In the medetomidine groups, the highest glucose level shifted to the right as the dosage increased. Namely, the glucose level peaked at 2 h (161 ± 19 mg/dL) in MED-20, at 3 h (147 ± 22 mg/dL) in MED-40 and at 4 h (151 ± 43 mg/dL) in MED-80. The regression analysis of the AUC data was non-significant (Figure 4A). Therefore, medetomidine did not show dose-related increase in blood glucose level, and higher doses of medetomidine resulted in even lower glucose levels than MED-20.

Figure 3. Plasma glucose levels following the administration of medetomidine (MED μg/kg, A) and xylazine (XYL mg/kg, B) in dogs. a: significantly different from the initial value (P 0.05). Each point and vertical bar represent the mean and SEM (n = 4 to 5).

Figure 4. The normalised AUC (0–8 h) data of the plasma glucose graphs are plotted versus the dose of medetomidine (A) and xylazine (B). Simple linear regression analysis was applied. P 0.001 proves that the regression in the graph B is highly significant.

On the other hand, this was not the case in the xylazine groups. Xylazine increased glucose values dose-dependently. Especially, the glucose levels in XYL-4 and XYL-8 groups were greatly increased when compared with the medetomidine groups. The regression analysis of the AUC data was also highly significant (Figure 4B). However, just like in medetomidine groups, the highest glucose levels tended to shift right as the dosage of xylazine increased. The peaks were observed at 2 h in both XYL-1 and XYL-2 (126 ± 16 and 164 ± 26 mg/dL), and at 4 h in both XYL-4 and XYL-8 groups (256 ± 111 and 281 ± 78 mg/dL) after injection. The overall comparison of the AUC data showed that the hyperglycemic effect of XYL was significantly greater than that of MED.

The insulin values significantly lowered at 0.5 h after injection of MED or XYL, and then gradually returned to baseline in all treatment groups (Figure 5). Significant decreases in the insulin level lasted for 2 and 3 h in MED-40 and MED-80 groups, and for 2 h in XYL-4 and XYL-8 groups. The slopes of the recovery phases indicated that the plasma insulin levels returned to baseline in a dose-related manner after MED injection, and the slope was smaller in XYL-8 than XYL-1, -2, and -4 groups. The regression analysis of the AUC data was not significant in either MED or XYL treatment. The overall comparison also failed to prove significant difference between MED and XYL on plasma insulin levels. The high variances among the data may explain why the regression analysis was non-significant. This may result from the biases in absorption after IM injections in this experiment.

Figure 5. Plasma insulin levels following the administration of medetomidine (MED μg/kg, A) and xylazine (XYL mg/kg, B) in dogs. a: significantly different from the initial value (P 0.05). Each point and vertical bar represent the mean and SEM (n = 4 to 5).

The NEFA levels significantly decreased at 0.5 h in all groups treated with medetomidine and xylazine, and gradually returned to baseline (Figure 6). Levels significantly decreased until 1, 2, 3, and 4 h in the MED-10, -20, -40, and -80 groups, respectively. On the other hand, NEFA levels were significantly low until 2 h in all XYL-treated groups. The NEFA levels in XYL-4 and -8 tended to decrease until 6 and 8 h respectively, but it was not significant. There were significant increases over baseline in some of the MED groups at 6 to 12 h and at 24 h in XYL-8. The regression analysis of the AUC data and the slope data of the recovery phases indicated that the effects of both MED and XYL on plasma NEFA levels were dose-related. The overall comparison failed to prove significant difference between MED and XYL treatments in the degree of NEFA suppression.

Figure 6. Plasma NEFA levels following the administration of medetomidine (MED μg/kg, A) and xylazine (XYL mg/kg, B) in dogs. a: significantly different from the initial value (P 0.05). Each point and vertical bar represent the mean and SEM (n = 4 to 5).

Both cortisol and glucagon values did not change significantly during the course of experiment in any of the treatment groups (data are not shown).

Discussion

The α₂-adrenoceptor agonists are well known to inhibit the sympathetic outflow in the central nervous system through their actions on α₂-adrenoceptors, hence decreasing the level of circulating catecholamines (2,11). Our results support this theory because the plasma catecholamine levels decreased after an injection of either medetomidine or xylazine. There are several data demonstrating the inhibition of catecholamine release at blood levels associated with the usage of α₂-agonists. For example, Benson et al (20) found that medetomidine administered preoperatively reduced the catecholamine levels in both ovariohysterectomized and non-operated dogs under isoflurane anesthesia. Talke et al (21) reported that dexmedetomidine administered to human patients postoperatively using a computer-controlled infusion technique decreased the plasma levels of catecholamines during infusion. Benson et al (9) also reported that xylazine injected after onychectomy reduced plasma concentrations of catecholamines in cats. In addition, Flacke et al (22) reported that treatment with clonidine reduced the catecholamine level in dogs anesthetized with fentanyl-enflurane-nitrous oxide. However, there were no reports to compare the effects of medetomidine and xylazine at different dosages on the plasma catecholamine levels in dogs under basal conditions. In the present study, both medetomidine and xylazine suppressed norepinephrine release dose-dependently with similar potency. On the other hand, medetomidine dose-dependently suppressed the epinephrine secretion. Xylazine also tended to inhibit epinephrine release dose-dependently. However, the potency of medetomidine in reducing the plasma epinephrine levels was greater than that of xylazine.

The plasma cortisol levels are influenced by both the peripheral site at the adrenal cortex, and the central site through the release of corticotropin-releasing factor (CRF) and adrenocorticotropic hormone (ACTH) in the brain. The effects of a₂-agonists on the plasma cortisol level have been assessed under different experimental conditions in a variety of species. Maze et al (23) found that an IM injection of 80 μg/kg dexmedetomidine reduced the basal cortisol level and the cortisol release to ACTH stimulation at 3 h post-injection in dogs, and concluded that only high dosages of dexmedetomidine inhibit adrenal steroidogenesis. In humans, it was reported that 0.45 mg clonidine administered orally for 3 d reduced plasma cortisol level (24), but 0.1 and 0.2 mg clonidine for 4 d did not affect it under basal conditions (25). Haas et al (26) reported that an intraperitoneal administration of 1 mg/kg clonidine significantly decreased the hypothalamic CRF-like immunoreactivity in rats. Furthermore, Taylor et al (27) reported in ponies that an intravenous (IV) anesthesia using detomidine, ketamine, and guaifenesin decreased the cortisol level, but did not significantly change the ACTH concentration. On the other hand, there are several reports with respect to the influence of α₂-agonists on the cortisol release associated with surgical stimulation. Premedication with medetomidine was reported to reduce or delay the increase of plasma cortisol levels induced by ovariohysterectomy in dogs (20,28). Sedation with xylazine diminished the increase in the plasma cortisol level after intradermal testing in dogs (10). In addition, clonidine prevented the elevation of cortisol level during laparoscopy in humans (29). These findings indicate that treatments with α₂-agonists such as medetomidine, xylazine or clonidine, show inhibitory effect on the release of cortisol. However, whether it is due to the α₂-adrenoceptor-mediated specific action, other receptor-mediated actions, or the result of non-specific effects by providing sedation and analgesia which reduce stress response, is unknown. In this respect, recent studies have evidenced that imidazoline receptors, but not α₂-adrenoceptors, may play a role in the direct inhibition of cortisol release in the adrenal cortex. An in vitro study revealed that the imidazoline α₂-adrenergic agents, medetomidine, detomidine, and atipamezole, all suppressed the release of cortisol from porcine adrenocortical cells (30). As medetomidine and detomidine are selective α₂-agonists, and atipamezole is a selective α₂-antagonist, this effect is unrelated to their actions on α₂-adrenoceptors. Maze et al found (23) that the selective α₂-agonist, dexmedetomidine (D-medetomidine) and its clinically ineffective enantiomer, L-medetomidine, were equally effective in blocking the ACTH-stimulated corticosterone secretion in adrenocortical cells of rats. This finding also supports that imidazoline receptors may also be responsible for inhibition of adrenal steroidogenesis.

In the present study, we examined the plasma cortisol level under basal conditions without surgical stress or ACTH stimulation, and found that both medetomidine and xylazine failed to significantly alter the plasma cortisol levels at the examined dosages. The possible explanation why an imidazoline derivative, medetomidine did not significantly decrease the cortisol level may be that its applied dose was too little to decrease cortisol release (23).

The most important finding of this study was that medetomidine did not increase the plasma glucose level in a dose-related manner in the examined dosages, whereas xylazine increased it dose-dependently. Our finding that the hyperglycemic effect of medetomidine was not dose-dependent is the first report in dogs. On the other hand, the present study revealed that either medetomidine or xylazine decreased plasma insulin levels in all dosages used. The comparison of the AUC insulin data also suggested that the potencies of medetomidine and xylazine to suppress insulin release were similar. There are numerous reports as to the effects of α₂-agonists on the blood glucose and insulin levels in different species. Either clonidine or xylazine was found to dose-dependently increase blood glucose in cattle (16,17) and rats (31,32). It was reported in dogs that an IM injection of 2.2 mg/kg xylazine increased blood glucose and decreased insulin level (12), which were in agreement with the present results. Similar results have also been reported in cats (13) and sheep (18). Two reports are available on the effect of medetomidine in dogs. Burton et al (33) found that 10 and 20 μg/kg medetomidine administered IV to beagle dogs tended to elevate the plasma glucose level, but it did not reach the level of significance and remained within the normal physiological range. They observed a peak in glucose level of about 90 mg/dL at 3 h after 20 μg/kg medetomidine treatment. This is in contrast with our results because the glucose level in the MED-20 group of our study was significantly elevated at 2 and 3 h post-injection up to the level of 161 ± 19 and 144 ± 32 mg/dL (mean ± SD), respectively. In that study, they also found that the insulin levels decreased similarly to our results, and that those effects were not different between 10 and 20 μg/kg medetomidine treatment. In addition, Benson et al (20) have reported that an intramuscular injection of 15 μg/kg medetomidine induced a decrease in insulin level, but did not significantly alter the glucose level in the non-operated dogs under isoflurane anesthesia. The insulin levels in their study (20) were similar to those observed in MED-10 and -20 of our study, but in contrast with our results, the plasma glucose level almost did not change 75 min after 15 μg/kg medetomidine injection in their study. It is not clear whether isoflurane anesthesia in their study might have influenced the glucose response. The fact that insulin plasma levels decreased after treatment and the glucose levels did not increase (20) suggests that insulin might not be the only factor controlling the plasma glucose levels after medetomidine treatment. Our results also suggest that the difference in the hyperglycemic response between medetomidine and xylazine found in our study can not be explained only by the α₂-adrenoceptor-mediated inhibition of insulin release. There may be other factors to consider.

It is well known that α₂-agonists inhibit the insulin release through their actions on α₂-adrenoceptors in the pancreas β-cells (34,35). Clonidine at high concentrations has been reported to increase the glucose release from bovine and canine liver slices in vitro (16,36). The subtype of α-adrenoceptors involved in this action seems to be α₁ rather than α₂, because prazosin, a specific α₁-adrenoceptor antagonist blocked this glucose release more effectively than the α₂-antagonist, yohimbine (36). As both xylazine and clonidine have similar α₂/α₁ selectivity ratios (37), high dosages of xylazine might exert some α₁-adrenoceptor-mediated effect similar to clonidine. This could be one reason for the extreme increasing of plasma glucose levels in XYL-4 and -8 groups of our study.

The central effects of α₂-agonists are also obscure. Xylazine administered intracerebroventricularly (ICV) at a dosage of 0.5 mg did not change the blood glucose level in a cat (13). Clonidine administered ICV to rats dose-dependently increased the blood glucose level (32), but whether it was central effect or the result of systemic absorption remained unclear. On the other hand, central imidazoline receptors may have a role in the regulation of glucose metabolism, because ICV-injected agmatine, a putative endogenous ligand for imidazoline receptors, exerted anti-hyperglycemic effect (38). Medetomidine shows higher α₂/α₁ receptor selectivity than xylazine (37), but because of its imidazoline structure it may also have affinity to bind to the imidazoline receptors (23,30). A possible action on the putative imidazoline receptors might explain why the glucose levels did not increase further after administration of high dosages of medetomidine in our study.

Other hormones, such as growth hormone (GH) and glucagon, may also influence the blood glucose level. Xylazine is known to increase the blood level of GH in dogs and cats (5). However, GH values were not determined in the present study. On the other hand, the plasma glucagon levels did not change significantly in the treatment groups of the present study, indicating that it may not be related to the hyperglycemic effects of both medetomidine and xylazine. This was in agreement with previous reports that xylazine did not significantly change the plasma glucagon level in dogs (39) and cats (13). Therefore, further work may be necessary to clarify the differences in the hyperglycemic effects of medetomidine and xylazine in dogs.

Both medetomidine and xylazine similarly dose-dependently decreased the plasma NEFA levels in this study (Figure 6). According to the authors' knowledge this is the first report available on the effects of either medetomidine or xylazine on NEFA levels in dogs. The suppression of lipolysis in dogs may be mediated by both central and peripheral α₂-adrenoceptors (14,15).

In conclusion, the present study revealed that both medetomidine and xylazine similarly inhibited norepinephrine release and lipolysis dose-dependently. Medetomidine suppressed epinephrine release dose-dependently with greater potency than xylazine. The cortisol and glucagon levels did not change significantly in any treatment group. Both drugs suppressed insulin secretion with similar potency. Furthermore, this study demonstrated that the hyperglycemic effect of medetomidine, in contrast with xylazine, was not dose-dependent at the tested dosages, and suggested that the effect of medetomidine on glucose metabolism may not be only due to α₂-adrenoceptor mediated actions.