| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Central Institute of Mental Health, Mannheim, Germany (M.G., A.W., D.K., A.N., M.D.); Department of Endocrinology and Metabolism, Otto-von-Guericke-University, Magdeburg, Germany (H.L.).
Address correspondence and reprint requests to Maria Gilles, MD, Central Institute of Mental Health, J5, 68159 Mannheim, Germany. E-mail: gilles{at}as200.zi-mannheim.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: Ten healthy male volunteers were included in a double-blind, placebo-controlled crossover study of a single dose of 40 mg of the 5-HT2 antagonist ketanserin versus placebo. Insulin sensitivity was measured by means of the euglycemic-hyperinsulinemic clamp technique. Subjects were treated with the 
-1 adrenergic antagonist phenoxybenzamine in both parts of the study to control for ketanserin’s effects at the level of this receptor.
Results: Compared with the placebo condition, subjects showed a significantly decreased insulin sensitivity after ketanserin (placebo: 9.4 ± 3.6 mg/kg/min; ketanserin: 7.7 ± 2.1 mg/kg/min; p = .047).
Conclusion: The selective 5-HT2 antagonist ketanserin impaired insulin sensitivity. This effect was possibly mediated by suppression of 5-HT2A receptor mediated glucose uptake in skeletal muscle.
Key Words: 5-HT2 antagonism • impaired insulin sensitivity • euglycemic-hyperinsulinemic clamp technique
Abbreviations: 5-HT = 5-hydroxytryptamine (serotonin); BMI = body mass index; FPG = fasting plasma glucose; IS = insulin sensitivity; BP = blood pressure.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This issue of major public health concern may limit the use of these agents, although they convey a reduced burden of extrapyramidal side effects and are believed to improve cognitive and negative symptoms of schizophrenia more effectively than conventional antipsychotics, thus favoring reintegration of patients (5,6).
One characteristic feature of most second generation antipsychotic drugs is their strong serotonin (5-HT2A) receptor antagonistic property. Thus, they differ from "older" drugs like haloperidol, flupenthixol, etc., which have a weak 5-HT2A) receptor antagonistic property.
The occurrence of diabetes mellitus in schizophrenic patients after treatment with second-generation antipsychotics has mostly been attributed to weight gain. However, direct receptor-mediated effects may also induce impairment of insulin sensitivity (7). The role of central nervous 5-HT2A regulation for the metabolic syndrome and physical activity is increasingly recognized (8). However, 5-HT2A does not only interact with energy homeostasis at the level of the brain but is also involved in the regulation of insulin action in the muscle. Specifically, 5-HT2A receptors have been found to be expressed both on adult rat skeletal muscle (9), and on human muscle cells (10), 5-HT2A leads to a rapid stimulation of glucose uptake in isolated rat skeletal muscle and especially in L6 myotubes (10). This effect is specifically mediated by 5-HT2A receptors, leading to an increase in the plasma membrane content of glucose transporters 1, 3, and 4 and is antagonized by the potent 5-HT2A receptor antagonist ketanserin (10).
On the other hand, 5-HT2A has been shown to interfere with insulin induced capillary recruitment in skeletal muscle, thus impairing glucose uptake (11,12). Therefore, the overall effect of 5-HT2A and 5-HT2A antagonism is unclear.
The objective of the present study was to determine the clinical significance of 5-HT2A antagonism on insulin sensitivity.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Subjects were admitted to our clinical research unit at 10:00 PM the evening before study medication was given. After admission, volunteers were kept fasting throughout the entire examination period. Tap water was allowed ad libitum. Bed rest was mandatory until the clamp examination was started. At 8:00 AM, a blood sample was drawn to determine serum potassium level. Because, in addition to its 5-HT2 antagonistic activity, high doses of ketanserin also exert -1-adrenergic antagonistic effects, it may lead to dilation of resistance vessels, increased delivery of glucose to muscle tissue, and increased insulin sensitivity (14). In order to exclude this confounding vascular effect of ketanserin, subjects were pretreated with a strong covalent binding -antagonistic agent, both under ketanserin and placebo conditions in order to maximally dilate resistance vessels in both treatment conditions and, thereby, override the -antagonistic effect of ketanserin. Phenoxybenzamine (Esparma, Magdeburg, Germany) was administered orally at 8:00 AM (10 mg) and again at 10:00 AM (5 mg). To prevent hypotension, isotonic saline infusion was started at 8:00 AM at a rate of 400 ml/h. Euglycemic, hyperinsulinemic clamp was performed using a commercially available clamp system (IMTM, Magdeburg, Germany, and EKF-Diagnostic, Barleben, Germany). Using a randomized, double-blind, balanced crossover approach, 40 mg ketanserin (Janssen-Cilag, Berchem, Netherlands) or placebo was given orally at 11.00 AM. After administration of ketanserin or placebo, a 3-hour insulin infusion was initiated at a constant rate of 1.0 mU/min/kg via an antecubital vein at 11 AM (H-Tronin, Aventis, Frankfurt, Germany). From a venous catheter at the contralateral arm, blood was withdrawn in 5-minute intervals for determination of glucose. In order to obtain arterialized blood, this arm had been warmed by a heating blanket. Blood glucose was measured at bedside by the standard analyzer of the clamp system, using the glucose oxidase method. Readings were then transferred online to the computer module, which calculated the glucose infusion rate necessary to maintain proband’s blood glucose at a constant level of 90 mg/dl (5 mmol/l) according to an algorithm based on the method by DeFronzo et al. (15). Glucose solution (20%) was administered together with insulin by the pump module of the clamp system, which was adjusted online by the computer module. If serum potassium was below 4.0 mmol/l, potassium chloride was added to insulin to obtain an infusion rate of 5 mmol/h. Use of the computer-based clamp system rather than manual clamp resulted in blood glucose profiles with little variation (mean blood glucose SD 8.8 mg/dl) and facilitated performance of the procedure and a more standardized insulin sensitivity measurement. Insulin sensitivity was calculated as glucose infusion rate during the last 30 minutes of the clamp, divided by body weight, as proposed by DeFronzo et al. (15).
Wilcoxon signed rank tests were used to compare fasting plasma glucose concentrations and insulin sensitivity, with p < .05 considered significant. Insulin sensitivity was calculated using average glucose infusion rate during the final 30 minutes of the clamp, when a steady state was achieved. Glucose infusion rate was divided by body weight and expressed as mg x kg–1 x min–1. All results are expressed as mean ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
All results are presented in Table 1.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This finding sheds new light on the association between second-generation antipsychotics and diabetes mellitus. All second-generation antipsychotics, except amisulpride, are antagonists at the 5-HT2 receptor (16). Therefore, several of these drugs may impair insulin sensitivity at the level of 5-HT2 receptor-mediated regulation of the muscle glucose transporters. So far, mainly olanzapine and clozapine had been linked with antipsychotic-induced diabetes (17). This may be due to the fact that these drugs are associated with significant weight gain (18), thus combining 2 disadvantageous effects on metabolism. However, our data indicate that not only these specific drugs but all substances antagonizing the 5-HT2 receptor may carry a risk with regard to insulin sensitivity.
Furthermore, our data are in agreement with favorable effects of selective serotonin reuptake inhibitors on the metabolic condition in diabetic patients. While fluoxetine is known to induce stimulation of 5-HT2c receptors and its short-term administration has been associated with weight loss, it also seems to improve insulin sensitivity beyond the effect mediated through weight loss by a possible stimulating effect on glycogen synthase activity in skeletal muscle tissue (19). Sertraline and paroxetine, though lacking direct 5-HT2c properties, may also improve glycemic control (20,21), but larger controlled studies are needed to confirm these findings.
Although our study sample was small and the finding requires an independent confirmation, the effect size and the compatibility of our result with in vitro findings render a false-positive finding unlikely.
One previous study by Fogari et al. (22) has examined the effects of ketanserin on insulin sensitivity measured by euglycemic, hyperinsulinemic clamp. In contrast to our findings, insulin sensitivity did not differ significantly between ketanserin and placebo in their study. However, blood pressure was significantly lower under ketanserin than under placebo conditions. This is not a surprising effect, since ketanserin is a dilator of resistance vessels, and in clinical practice it has been used as an antihypertensive agent. Vasodilatation of resistance vessels leads to an increased delivery of substrate to glucose utilizing tissue such as muscle and fat. Thus, a potential direct inhibiting effect of ketanserin on muscle glucose uptake may be masked by increased delivery of insulin and glucose to muscle.
In our study, this confounding factor was effectively addressed by pretreatment with the vasodilating receptor antagonist phenoxybenzamine and infusion of saline both under placebo and verum conditions. As evidenced by identical blood pressure in both study conditions, ketanserin did not exert an additional vasodilatory effect on top of antagonism. Additionally, blood pressure and pulse rate were not differentially modulated by both treatment conditions, and there was no other evidence for a different regulation of the sympathoadrenergic tone. Of course, the nature of our data does not allow us to exclude that ketanserin impaired insulin sensitivity due to effects of the central nervous system (CNS). However, in the CNS, 5-HT2A receptors stimulate sympathetic activity, and, therefore, blockade of this receptor with ketanserin would have rather lowered the sympathoadrenergic activity leading to improved insulin resistance.
Ketanserin is a competitive antagonist at 5-HT2A and, to a much lesser extent, 5-HT2C receptors (http://kidb.bioc.cwru.edu/pdsp.php). It also possesses -1-adrenergic properties at concentrations higher than those needed to act as an antagonist at 5-HT2 receptors and weak antihistaminergic and dopamine-blocking effects. The dosage used in our study is well below the dose range required for these confounding effects (23).
Our study cannot differentiate between direct effects of 5-HT2A on insulin-induced glucose uptake in muscle and adipose tissue on one hand and indirect vascular effects on the other hand. However, in rat hindlimb 5-HT2A receptor agonism with -methylserotonin has been shown to decrease insulin induced glucose uptake by a vascular mechanism, ie, capillary recruitment (12). Therefore, antagonism of 5-HT2 receptors, which was studied in the present work, would be expected to improve insulin sensitivity, if it were due to vascular effects. We hence conclude that the impairment of insulin sensitivity observed in the present study is most likely the consequence of direct action of ketanserin on muscular or adipose tissue.
Most second-generation antipsychotic drugs have a very heterogeneous receptor binding profile. In this study, we examined the effect of occupation of a single receptor type, which is characteristic to most second-generation antipsychotics. Other receptors involved in the action of antipsychotic drugs are known to have effects on insulin sensitivity as well, eg, muscarinic or -1 adrenergic receptors (24,25). Differences in diabetes propensity of various second generation antipsychotics may well be explained by their specific receptor binding profile, with 5-HT2A antagonism being only one modulator of insulin sensitivity. This may also explain the fact that there is no apparent correlation between 5-HT2A receptor affinity and diabetes risk under treatment with different antipsychotic drugs.
Insulin sensitivity is a major long-term determinator of lipoprotein composition (26). However, our study of an acute single administration of ketanserin cannot assess those long-term sequelae of insulin resistance.
Taken together, our study provides first evidence that the 5-HT2 receptor is directly involved in glucose metabolism in humans. This finding is relevant for pharmacological agents with 5-HT2 receptor antagonistic properties (second-generation antipsychotics, mirtazapine) which may deteriorate insulin sensitivity and, in susceptible patients or combined with weight gain, may increase the risk for impaired glucose tolerance and ultimately for diabetes mellitus. Additionally, our findings support the assumption that serotonergic antidepressants may directly improve insulin sensitivity.
We thank Ms. Waltraud VanSyckel for editorial and secretarial assistance.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Received for publication July 8, 2004; revision received May 9, 2005.
DOI:10.1097/01.psy.0000174994.91245.34
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
