| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Psychiatry, Adolescent Clinic (L.J.M.v.N., L.d.H., H.B., T.v.A.), Social Psychiatric Service Centre (J.G.S.), and Departments of Endocrinology and Metabolism (R.M.E.B., G.A., E.F., M.J.M.S., H.P.S.), Radiology and Medical Physics (H.W.V.), and Clinical Chemistry (M.T.A.), Laboratory of Endocrinology and Radiochemistry, Academic Medical Center, University of Amsterdam, 1105 AZ, Amsterdam, The Netherlands
Address all correspondence and requests for reprints to: Jitschak Storosum, Department of Psychiatry, SPDC, Academic Medical Center Amsterdam, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. E-mail: j.g.storosum{at}amc.uva.nl.
| Abstract |
|---|
|
|
|---|
Design: Seven antipsychotic medication-naive patients fulfilling the DSM IV A criteria for schizophrenia/schizoaffective disorder were matched for body mass index, age, and sex with seven control subjects. We measured endogenous glucose production and peripheral glucose disposal using a hyperinsulinemic euglycemic clamp (plasma insulin concentration
200 pmol/liter) in combination with stable isotopes. Fat content and fat distribution were determined with a standardized single-slice computed tomography scan and whole body dual-energy x-ray absorptiometry.
Results: Endogenous glucose production during the clamp was 6.7 µmol/kg·min (SD 2.7) in patients vs. 4.1 µmol/kg·min (SD 1.6) in controls (P = 0.02) (95% confidence interval –5.2 to 0.006). Insulin-mediated peripheral glucose uptake was not different between patients and controls. The amount of sc abdominal fat in patients was 104.6 ± 28.6 cm3 and 63.7 ± 28.0 cm3 in controls (P = 0.04) (95% confidence interval 4.4–77.2). Intraabdominal fat and total fat mass were not significantly different.
Conclusions: Antipsychotic medication-naive patients with schizophrenia or schizoaffective disorder display hepatic insulin resistance compared with matched controls. This finding cannot be attributed to differences in intraabdominal fat mass or other known factors associated with hepatic insulin resistance and suggests a direct link between schizophrenia and hepatic insulin resistance.
| Introduction |
|---|
|
|
|---|
Recently, it was shown that first-episode, relatively old, antipsychotic-naive patients with schizophrenia had higher levels of plasma glucose, insulin, and cortisol than age- and sex-matched controls (5). This finding can be attributed to differences in body composition between patients with schizophrenia and healthy controls since a higher waist-to-hip ratio and higher visceral fat content, as assessed by computed tomography (CT) scan, were reported in two controlled studies (6, 7). By contrast, no differences in glucose and insulin concentrations and visceral fat mass were reported in younger antipsychotic-naive schizophrenic patients compared with healthy controls (8, 9). A recent study, using minimal model analysis of an iv glucose tolerance test, showed insulin resistance in a rather small number of drug free patients compared with healthy controls (10). These conflicting results may be due to differences in study design or differences in the definition of insulin sensitivity.
Here, we report a matched case-control study comparing glucose metabolism in antipsychotic medication-naive schizophrenic or schizoaffective patients with matched controls, using the hyperinsulinemic euglycemic clamp technique in combination with stable isotopes, which is considered to be the gold standard for measuring insulin sensitivity in vivo (11). We hypothesized that patients have hepatic and peripheral insulin resistance caused by higher plasma free fatty acids (FFAs) (12, 13) and an altered secretion of adipokines (14, 15, 16, 17).
The main reason for the recent study, conducted in drug-naive patients with schizophrenia, is to investigate whether an impaired glucose metabolism is a part of the pathophysiology of the primary disorder and is not secondary to the use of medication.
| Subjects and Methods |
|---|
|
|
|---|
Seven antipsychotic medication-naive patients, who were experiencing their first psychotic episode (hallucinations and/or delusions), were included in a cross-sectional study, comparing cases with matched controls. These patients were recruited from the Psychiatric in- and outpatient Clinic of the Academic Medical Center (AMC) of the University of Amsterdam, from February 2004 until March 2006. The patients were eligible for the study if they fulfilled the Diagnostic and Statistical Manual of Mental Disorders (DSM) IV A criteria for schizophrenia or schizo-affective disorder. Characteristic symptoms are two (or more) of the following, each present for a significant portion of time during a 1-month period: delusions; hallucinations; disorganized speech (e.g. frequent derailment or incoherence); grossly disorganized or catatonic behavior; and negative symptoms, i.e. affective flattening, alogia, or avolition. Patients were eligible if they had no other axis I or II DSM IV diagnosis that needed treatment, were older than 18 yr, understood the objective of the study, and were competent to give informed consent. This competency was evaluated by the nursing team and independent resident involved in the treatment of the patient. Exclusion criteria were: 1) DM or a medical history or family history for type 2 DM; 2) a recent history (6 months or less) of substantial alcohol abuse, or DSM IV criteria for alcohol dependence disorder, psychoactive substance abuse, or dependence disorder; 3) alcohol use in the last 3 d before the start of the study; 4) cannabis use in the last month before the start of the study; 5) the use of antipsychotic medication or any other medication except paracetamol (acetaminophen); 6) somatic illness, including neoplasm, metabolic or endocrine disorders, active infection, or gross structural abnormalities on magnetic resonance imaging of the brain; and 7) no informed consent. The diagnosis of schizophrenia according to DSM IV was made by an experienced psychiatrist at baseline and reevaluated after 6 months during a clinical consensus meeting with three psychiatrists. Substance (ab)use and physical health were self-reported and confirmed by caregivers; no urine tests were performed. The control group consisted of seven healthy subjects (medical students) matched for body mass index (BMI), age, and sex.
Because we expected to find differences in fat distribution, and no differences in total fat percentages, patients and controls were matched for BMI, not according to the dual-energy x-ray absorptiometry (DEXA) or CT fat measures. Patients and controls were not matched for lifestyle parameters or dietary intake. None of the subjects was taking any form of prescribed or over-the-counter medication.
The study was approved by the Medical Ethical Committee of the AMC. After a complete description of the study was given, written informed consent was obtained.
Hyperinsulinemic euglycemic clamp
Subjects were admitted to the Metabolic Clinical Research Unit of the AMC and studied in the supine position. After a 13-h fast since 1900 h the day before, a catheter was inserted into the dorsal vein of the hand or distal vein of each arm. One catheter was used for sampling of arterialized blood using a heated hand box (60 C). The other catheter was used for infusion of [6,6-2H2]glucose, glucose 20%, and insulin. At 0800 h (t = –2.5 h), after drawing a blood sample for background enrichment of plasma glucose, a continuous infusion of [6,6-2H2]glucose (>99% enriched; Cambridge Isotopes, Andover, MA) was started at a rate of 0.11 µmol/kg·min after a priming dose equivalent to 80-min infusion. After 120, 130, 140, and 150 min, blood samples were drawn for determination of glucose enrichments. Subsequently, at t = 0 h, a primed continuous infusion of insulin (Actrapid 100U/ml; Novo Nordisk Farma, Alphen a/d Rijn, The Netherlands) was started for 2.5 h at a rate of 20 mU/m2 body surface area·min, aiming for a plasma insulin concentration of approximately 200 pmol/liter. Plasma glucose was measured every 5 min (Beckman glucose analyzer 2; Beckman, Palo Alto, CA), and glucose 20% was infused at a variable rate to maintain plasma glucose at 5.0 mmol/liter. [6,6-2H2]glucose was added to the 20% glucose solution to achieve glucose enrichments of 1% to minimize changes in isotopic enrichment due to changes in the infusion rate of exogenous glucose, and, thus, to allow for accurate quantification of glucose kinetics. During the last hour of the clamp, blood samples were drawn at 5-min intervals for determination of glucose enrichments. At t = 0 h and t = 2.5 h, samples for the measurement of insulin, glucagon, cortisol, catecholamines, FFA, retinol-binding protein 4 (RBP4), and adiponectin were drawn. During the study the participants were only allowed to drink water.
Laboratory assays
Stable isotope analysis ([6,6-2H2]glucose enrichment) was measured as described earlier (18). Cortisol in plasma was determined with a chemiluminescent immunoassay (Immulite 2000; Diagnostic Products Corp., Los Angeles, CA). Insulin was determined with a chemiluminescent immunometric assay (Immulite 2000). Glucagon was determined with the LINCO 125I RIA (LINCO Research, Inc., St. Charles, MO). Norepinephrine and epinephrine were determined with an in-house HPLC method. Adiponectin was determined by a RIA (LINCO). RB4 was determined with an ELISA kit (Immundiagnostik, Bensheim, Germany). The FFA concentration was determined with an enzymatic colorimetric method (NEFA-C test kit; Wako Chemicals GmbH, Neuss, Germany).
Two consecutive 24-h urine samples were collected separately at home. During this collection, the urine was kept refrigerated. Of the two 24-h urine samples, total volume, as well as concentration of free cortisol (in-house HPLC method) and creatinine (Jaffé method, Hitachi 917; Roche, Indianapolis, IN), were measured. To minimize the effects of unreliable collections, we calculated cortisol to creatinine ratios in all 24-h samples (19).
Indirect calorimetry
Oxygen consumption (VO2) and carbon dioxide production (VCO2) were measured with the ventilated hood technique (model 2900; Sensormedics, Anaheim, CA). VO2 and VCO2 were measured continuously during the final 30 min in the basal state and during the hyperinsulinemic euglycemic clamp. The mean rates of VO2 and VCO2 during the final 20 min were used for calculation of glucose and fat oxidation as described by Frayn (20). Nonoxidative glucose disposal was calculated as the difference between total glucose disposal and glucose oxidation.
Abdominal fat measurements
Total, as well as regional, fat mass was quantified in all patients by DEXA (Hologic QDR-4500W; Hologic, Inc., Bedford, MA; software version whole-body v8.26A: 5), providing a quantitative assessment of peripheral and truncal fat mass in kilograms. A standardized single-slice abdominal CT scan (Mx8000 Quad; Philips Medical Systems, Best, The Netherlands) using 120 kV, 100 mAs, and a slice thickness of 1 cm was performed. On the survey image, the level of the fourth lumbar vertebra was chosen, which is the level of the umbilicus in most patients (21). It was shown that the fat volume in a slice at this level is a valid predictor of total abdominal fat in men (22, 23). The volume of total adipose tissue, intraabdominal adipose tissue, and sc adipose tissue was determined by summing the volumes of the voxels with CT values within the range of –170 to –30 Hounsfield units, and expressed in cubic centimeters. Care was taken to exclude intracolonic contents with Hounsfield units within the same range (24). Radiologists were blind to the subject groups.
Calculations
Endogenous glucose production (EGP) and peripheral glucose uptake (Rd) were calculated using the modified form of the Steele equations as described previously (25).
Data analysis
Mann-Whitney U tests (two-tailed) were used to compare results between patients and controls. All results are expressed as means and SDs. The data were analyzed using SPSS, version 11 (SPSS, Inc., Chicago, IL). The overall significance level was set at P = 0.05 (two-tailed).
| Results |
|---|
|
|
|---|
Basal plasma glucose concentration did not differ between patients. Fasting insulin concentrations in patients tended to be higher as compared with controls: 49 ± 18 pmol/liter (8.2 ± 3 µU/ml) vs. 29 ± 12 pmol/liter (4.8 ± 2 µU/ml); P = 0.07 (95% CI 2.4–35.9). Basal EGP was not different between patients and controls. During the hyperinsulinemic euglycemic clamp, plasma glucose and insulin concentrations were not different between patients and controls. EGP during the clamp was significantly higher in patients [6.7 µmol/kg·min (SD 2.7) vs. 4.1 µmol/kg·min (SD 1.6) in controls; P = 0.02; 95% CI –5.2 to 0.006]. Rd, glucose oxidation, and nonoxidative disposal were not different between patients and controls.
|
Basal plasma FFAs tended to be lower in patients [0.31 mmol/liter (SD 0.12) vs. 0.54 mmol/liter (SD 0.27); P = 0.07; 95% CI –0.48 to 0.02]. During the clamp they were not significantly different. Basal lipid oxidation and lipid oxidation during the clamp were not different.
|
Glucoregulatory hormones were not significantly different between groups. The mean cortisol to creatinine ratio in 24-h urine collections in patients was 8.8 ± 7.0 nmol/mmol and in controls 5.0 ± 1.6 nmol/mmol (P = 0.57; 95% CI –2.8 to 10.2).
Basal plasma RBP4 concentrations were not different between patients and controls. Basal adiponectin concentrations and adiponectin concentrations during the clamp were not different between patients and controls.
Abdominal fat measurements
Results from CT scan and DEXA measurements were obtained in six patients and six controls; one patient did not show up for the imaging study, and one control subject withdrew consent because of fear of radiation exposure in the DEXA measurement, despite the text in the informed consent form. The amount of sc abdominal fat was significantly different in patients (104.6 ± 28.6 cm3) compared with controls (63.7 ± 28.0 cm3) (P = 0.04; 95% CI 4.4–77.2), as well as total abdominal fat tissue (141.1 ± 36.3 cm3 in patients vs. 93.3 ± 39.6 cm3 in controls; P = 0.04; 95% CI –1.0 to 96.7). Intraabdominal fat was not significantly different between groups: 36.6 ± 17.1 cm3 in patients vs. 29.6 ± 12.4 cm3 in controls (P = 0.42; 95% CI –12.2 to 26.3). Total fat mass as measured with DEXA was not significantly different: 11.1 (15%) ± 3.3 kg in patients vs. 8.3 (12%) ± 2.4 kg in controls; 95% CI –0.8 to 6.4). Total truncal fat was 4.6 ± 1.5 kg in patients vs. 3.5 ± 1.0 kg in controls (P = 0.12; 95% CI –0.5 to 2.7). Total arm fat was 1.3 ± 4.5 in patients vs. 0.8 ± 0.2 kg in controls (P = 0.05; 95% CI 0.05–1.0). Total leg fat was 4.1 ± 1.6 kg in patients vs. 3.0 ± 1.2 kg in controls (P = 0.25; 95% CI –0.6 to 3.0).
| Discussion |
|---|
|
|
|---|
Previous studies investigating insulin resistance in antipsychotic-naive, first-episode psychotic patients showed conflicting results (5, 8, 10). Those studies were conducted with the homeostasis model assessment, which is less precise and less reproducible than the hyperinsulinemic euglycemic clamp technique, especially in small sample-sized studies (11). Our finding of a trend for higher fasting insulin levels was reported earlier (5). Other studies could not confirm this finding, but these studies did not match the controls for BMI (8, 9), although after correction with BMI as a covariate, antipsychotic free patients showed significantly higher fasting insulin levels compared with antipsychotic-naive patients and healthy controls (8). The reduced hepatic insulin sensitivity in the patients cannot be attributed to cortisol because neither plasma cortisol levels nor cortisol to creatinine ratios were significantly different between the two groups.
Adiponectin levels have been described to correlate positively with (whole body) insulin sensitivity in human subjects (10, 26). However, there was no difference in adiponectin levels between our patients and healthy controls. We did not measure high and low molecular weight adiponectin and, therefore, cannot exclude a significant difference in especially high molecular weight adiponectin between patients and controls, explaining the difference in EGP. Plasma RBP4 (27) and FFA concentrations (28, 29) are additional important modulators of insulin sensitivity but are unlikely to play a role in the induction of hepatic insulin resistance in schizophrenia because we found no differences between patients and controls.
Epidemiological studies have reported an association between abdominal obesity and insulin resistance (30). Ever since, much research has focused on visceral fat and hepatic insulin sensitivity. A remarkable finding in the present study was the difference in abdominal fat distribution between our patients and controls. Patients showed a higher sc abdominal fat mass without increase in visceral fat mass, which contradicts reported data in the literature (6, 7). An increase in visceral fat in antipsychotic free and antipsychotic-naive patients was reported earlier (6, 7). In one of these studies (7), significant differences in life styles between patients and controls (higher saturated fat intake, lower fiber intake, and less exercise in the patients) could explain this finding. Patients and controls in our study were not matched for dietary intake, or for smoking or exercise patterns. This may be a confounding variable because smoking per se may affect insulin sensitivity (31). However, insulin sensitivity as noticed in smokers could be related to an increase in counter-regulatory hormones, but to the best of our knowledge, nicotine use has not been associated with selective hepatic insulin resistance. Subcutaneous fat is described to be associated with insulin resistance, especially in patients with upper body obesity (13), though a lack of significant effect on hepatic glucose production by sc fat was also reported (32). Increased FFA release from visceral adipocytes draining directly into the portal vein is thought to explain the correlation between visceral fat and hepatic insulin resistance (12). The source of increased systemic plasma FFA in upper body obesity seems to be the upper body sc fat compartment (13). This suggests that hepatic insulin resistance is induced by increased FFA delivery from enhanced lipolysis from visceral adipocytes and peripheral insulin resistance by increased systemic FFA availability derived from enhanced lipolysis from upper body sc adipocytes. However, these mechanisms are not relevant for our study because our subjects showed no differences in truncal fat, and the patients in our study did not have increased plasma FFA levels. Despite similar levels of visceral fat, patients displayed hepatic insulin resistance, suggesting that other mechanisms are responsible for our finding. Obviously, we did not measure portal FFA delivery, and, therefore, cannot exclude any differences in lipolytic activity in visceral adipocytes between patients and controls. Another explanation for the increased sc abdominal fat mass could be leptin (33) because elevated plasma leptin levels correlate with sc fat deposition.
Finally, neuronal input to the liver may be responsible for our findings (34). Recent experiments in rats have shown a very important role for multisynaptic autonomic pathways originating in the hypothalamus, and reaching the liver via the brainstem, in the regulation of hepatic glucose production. The infusion of either insulin or small-molecule insulin mimetics in the third ventricle suppresses glucose production independent of circulating levels of insulin and of other glucoregulatory hormones (35), though this finding might be species specific because no significant role of brain insulin on hepatic glucose production was found in dogs (36). A distal lesion of the parasympathetic input to liver prevents this, resulting in striking insulin resistance with increased hepatic glucose production (37). Other experiments in rats have shown a pivotal role for vagal input to white adipose tissue, with an anabolic role in the local regulation of insulin sensitivity (38). Together, these studies have demonstrated a cross talk between the central nervous system on one hand and adipose tissue and liver on the other, coupling central nutrient sensing to peripheral nutrient production. If indeed this cross talk might be disturbed in schizophrenia, this might represent a whole new perspective in metabolic research in schizophrenia (39). In addition, alterations in the central nervous system lipid regulation of patients with schizophrenia, such as myelin and fatty acid biosynthesis dysfunction (40), could be accompanied by alterations in liver and fat tissue lipid regulation, and thus explain both the hepatic insulin resistance and altered fat distribution.
In conclusion, antipsychotic-naive, first-episode psychotic patients with a diagnosis of schizophrenia or schizoaffective disorder show hepatic insulin resistance compared with matched controls that cannot be attributed to an increase in visceral fat mass, or differences in plasma FFAs, plasma adiponectin, or plasma RBP4 concentrations. Our findings suggest a direct link between schizophrenia and hepatic insulin resistance.
| Acknowledgments |
|---|
| Footnotes |
|---|
First Published Online November 20, 2007
Abbreviations: BMI, Body mass index; CI, confidence interval; CT, computed tomography; DEXA, dual-energy x-ray absorptiometry; DM, diabetes mellitus; EGP, endogenous glucose production; FFA, free fatty acids; RB4, retinol-binding protein 4; Rd, peripheral glucose uptake; VCO2, carbon dioxide production; VO2, oxygen consumption.
Received May 29, 2007.
Accepted November 8, 2007.
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |