Brief Communication

Insulin Resistance and Adiponectin Levels in Drug-Free Patients With Schizophrenia: A Preliminary Report
Tony A Cohn, MBChB, FRCPC1, Gary Remington, MD, PhD, FRCPC2,3, Robert B Zipursky, MD, FRCPC2,4, Azar Azad, PhD5, Philip Connolly, PhD6, Thomas MS Wolever, MD, PhD7
Objective: To compare the insulin sensitivity and adiponectin levels of medication-free patients suffering from schizophrenia or schizoaffective disorder with that of matched healthy volunteers. Method: We evaluated 9 nondiabetic patients aged 26.6 years (median 26 years, range 17 to 41 years) and matched volunteers, using the frequently sampled intravenous glucose tolerance test, minimal model analysis, and fasting adiponectin levels. Results: The mean insulin sensitivity index of the patients was 42% lower than that of the healthy volunteers (P = 0.026), with inadequate compensation in insulin secretion. Patients with schizophrenia tended to have reduced adiponectin levels (P = 0.055). Conclusions: By direct measurement, this study provides evidence of insulin resistance and susceptibility to type 2 diabetes in patients with schizophrenia who are free of antipsychotic drugs. (Can J Psychiatry 2006;51:382386) Information on funding and support and author affiliations appears at the end of the article.

Clinical Implications Patients with schizophrenia may be at increased risk for developing type 2 diabetes before the prescription of antipsychotic medications. Susceptibility to diabetes may be influenced by constitutional factors, antipsychotic medications, diet, physical inactivity, and stress. Patients with schizophrenia should be regularly screened for emerging diabetes according to current diabetes clinical practice guidelines. Limitations The small number of patients we studied limits the generalizability of this study. We did not measure visceral adiposity. We used a nonvalidated self-report scale to assess physical activity.

Key Words: schizophrenia, drug-free, insulin resistance, adiponectin, diabetes susceptibility, frequently sampled intravenous glucose tolerance test, Bergmans minimal model atients with schizophrenia have increased rates of type 2 diabetes (1). Recent attention has focused on the contribution of second-generation antipsychotic medications to the risk of diabetes. It is equally important, however, to examine whether patients with schizophrenia are susceptible to type 2

P
382

diabetes independent of antipsychotic treatment. The Clinical Practice Guidelines for the Management of Diabetes in Canada lists schizophrenia as a risk factor for type 2 diabetes (2). However, there is only limited empirical evidence for this assertion.
W Can J Psychiatry, Vol 51, No 6, May 2006

Insulin Resistance and Adiponectin Levels in Drug-Free Patients With Schizophrenia: A Preliminary Report

Historical evidence prior to the 1950s introduction of antipsychotic medication suggests an association between mental illness and diabetes (3). An increased rate of impaired fasting glucose was recently reported in drug-naive individuals with schizophrenia (4). In contrast, by indirect measures of fasting glucose and insulin (that is, the HOMA-IR), others found relative insulin resistance in previously treated, but not in antipsychotic-naive, individuals with schizophrenia (5). No research has been published in drug-free patients with schizophrenia, using dynamic, rigorous, and validated measures of insulin sensitivity and insulin secretion. The FSIGTT is widely used and considered valid for the simultaneous assessment of insulin secretion and insulin sensitivity (6,7). In addition, adiponectin, an adipocytederived peptide, is independently associated with apparently healthy individuals risk of developing type 2 diabetes (8). The goal of our study was to use the FSIGTT and fasting adiponectin levels to compare drug-free patients with schizophrenia and matched healthy volunteers.

We performed a 3-hour FSIGTT with Bergmans minimal model analysis, using an insulin-modified protocol (11). Following baseline blood samples, a bolus of 50% glucose (25.1ml/m2 body surface area) was rapidly injected intravenously at time 0 minutes, and further blood samples were taken. At 20 minutes, we injected participants with intravenous insulin (1.6 U/m2). We obtained additional blood samples from 20 minutes to 180 minutes. We determined Si by kinetic analysis of resulting blood glucose and insulin concentrations, using the minimal model (6). Acute insulin response to glucose, a measure of insulin secretion, was calculated from the area under the insulin curve from 0 minutes to 10 minutes. Disposition index was calculated from the product of Si and acute insulin response to glucose. Additionally, we measured fasting adiponectin prior to the FSIGTT, morning plasma cortisol, dietary intake (with a 3-day food diary), and physical activity (with a nonstandard ized self-report questionnaire). We compared dependent variables with unpaired t tests, using the SPSS statistical software package, version 12.1.

Methods
We studied 9 consecutively presenting patients to the First Episode Psychosis Program at the Centre for Addiction and Mental Health in Toronto who were willing to participate. They were drug-free, met American Psychiatric Association DSM-IV SCID criteria for either schizophrenia or schizoaffective disorder (9), and were capable of providing informed consent. We also studied 9 healthy matched volunteers. Healthy volunteers were matched with the patients for age, sex, ethnicity, smoking status, BMI, and waist circumference; they had no psychiatric disorder on structured interview. Patients and healthy volunteers were medically screened, were physically healthy, and did not have diabetes or impaired fasting glucose on a fasting glucose determination according to American Diabetes Association criteria (10). None of the patients and one of the healthy volunteers had a first-degree relative with type 2 diabetes. All participants gave informed consent to the Research Ethics Boardapproved study protocol.

Results
Of the 9 patients, 7 had never been treated with psychotropic medications and 2 had no antipsychotic treatment for more than 1 year. Seven patients were diagnosed with schizophrenia and 2 with schizoaffective disorder. One patient was a hospitalized inpatient and 8 were outpatients. Patients scored on average 53 (SD 10) on the Brief Psychiatric Rating Scale, which is consistent with moderate levels of symptom intensity. The patient group and the control group were well matched, and each consisted of 8 male subjects and 1 female subject: 5 individuals of European descent, 2 of South Asian descent, and 2 of Afro-Caribbean descent. The age (26.6 years, SD 8.7, and 27.1 years, SD 6.2, respectively), BMI (23.7 kg/m, SD 2.9, and 23.7 kg/m, SD 2.7, respectively), and smoking status (that is, 4 patients with schizophrenia and 3 volunteers who smoked) of the patient and control groups were similar. According to self-report of sleeping, sitting, walking, and exercise time, patients and healthy volunteers had similar levels of physical activity. Patients reported consuming fewer calories than healthy control subjects (1590 Kcal, SD 969, compared with 2669 Kcal, SD 869; P = 0.029), and patients tended to consume less dietary fibre (6.1g/100 Kcal, SD 6.2, compared with 10.7 g/100 Kcal, SD 5.3; P = 0.122); percentage of energy from total and saturated fat was similar. Morning cortisol was similar between patients and healthy volunteers (369 nmol/L, SD 153, compared with 347 nmol/L, SD 100).
383

Abbreviations used in this article

BMI body mass index FSIGTT frequently sampled intravenous glucose tolerance test HOMA-IR homeostatic model of insulin resistance SCID Si Structured Clinical Interview for DSM-IV insulin sensitivity index

Can J Psychiatry, Vol 51, No 6, May 2006 W

The Canadian Journal of PsychiatryBrief Communication

Figure 1 Average time course of plasma glucose and insulin during the FSIGTT(SEM) in 9 drug-free patients with schizophrenia and 9 matched healthy volunteers

Inset A mean (SEM) insulin sensitivity(Si)

Inset B mean (SEM) acute insulin response to glucose

*P < 0.05 **P < 0.01 ***P < 0.005

SEM = standard error of mean

Drug-free patients with schizophrenia Matched healthy volunteers

Glucose disposal differed between patients and healthy volunteers, with patients having higher glucose levels between 12 minutes and 60 minutes after injection of intravenous glucose (Figure 1), whereas insulin levels were similar at all time points. Mean Si was 42% lower in patients than in control subjects (P = 0.026) (Inset A), while acute insulin response to glucose was similar (P = 0.752) (Inset B). There was a trend toward reduced disposition index in patients (1370, SD 872, compared with 2019, SD 744; P = 0.108). Fasting adiponectin tended to be lower in patients (20.3 ng/ml, SD 8.40, compared with 29.3 ng/ml, SD 10.0; P = 0.055).
384

We conducted a subanalysis on the 7 medication-naive patients and matched healthy volunteers. Here, insulin sensitivity in the patients was reduced by 49.8% (P = 0.025), and adiponectin was reduced by 33.01% ( P = 0.034).

Discussion
The results indicate that drug-free patients with schizophrenia are insulin-resistant with failed compensation in insulin secretion needed to maintain glucose disposal in the face of reduced insulin sensitivity. This suggests that schizophrenia itself or associated lifestyle factors are associated with insulin resistance and may also influence b-cell compensation. Additionally, increased susceptibility to diabetes in these patients is
W Can J Psychiatry, Vol 51, No 6, May 2006

Insulin Resistance and Adiponectin Levels in Drug-Free Patients With Schizophrenia: A Preliminary Report

supported by the trend toward lower adiponectin levels. These findings are unrelated to previous antipsychotic treatment and remain consistent when comparing only drug-naive patients with matched control subjects. Increased insulin resistance may be a consequence of emotional stress and disturbed pituitary-adrenal function and therefore not specific to schizophrenia. However, in the patients we studied, morning cortisol, a crude measure of stress, was no different from that of healthy volunteers. Lifestyle factors, such as a diet high in saturated fat and low in fibre and physical inactivity, may contribute to insulin resistance. However, we found no significant difference between patients and control subjects in physical activity or fat intake. Patients tended to consume less dietary fibre, which is consistent with other populations that have an increased prevalence of diabetes (12). This may account for the increased insulin resistance, but our small sample size limits a definitive statement to that end. Increased visceral adiposity, as reported by other authors (13), may account for the increased insulin resistance, and an association between visceral adiposity and low adiponectin levels has been reported. Although patients and healthy volunteers had similar waist circumference, we did not include a direct measure of visceral fat. We conclude that, as a group, patients with schizophrenia are insulin-resistant in the absence of treatment with antipsychotic medication, which suggests that the disease itself, or associated lifestyle and environmental factors, is associated with an increased susceptibility to diabetes. Thus it appears that antipsychotic medication is not a necessary but rather a contributing factor in accounting for the increased diabetes risk. Antipsychotic-emergent diabetes should be studied as an interaction between illness-related vulnerability and antipsychotic treatment. Adiponectin may be a useful marker of diabetes risk in patients with schizophrenia, and future studies should include measures of visceral adiposity. While the small number of study participants may limit generalizability, our study is strong in that it provides clear evidence, by direct measurement, of relative susceptibility to diabetes in a group of patients with schizophrenia who are free of antipsychotic treatment. Further research with larger study samples is needed to replicate these findings and to further explain the association between schizophrenia and insulin resistance.

Funding and Support This study was funded by a Research in Psychiatry Grant from the Associates in Psychiatry, Centre for Addiction and Mental Health, Toronto, Ontario. This study was conducted within the Schizophrenia Program at the Centre for Addiction and Mental Health.

References
1. Dixon L, Weiden P, Delahanty J, Goldberg R, Postrado L, Lucksted A, and others. Prevalence and correlates of diabetes in national schizophrenia samples. Schizophr Bull 2000;26:9032. 2. Canadian Diabetes Association Guidelines Expert Committee. Canadian diabetes association 2003 clinical practice guidelines for the prevention and management of diabetes in Canada. Can J Diabetes 2003;27(Suppl 2):S1S152. 3. Kohen D. Diabetes mellitus and schizophrenia: historical perspective. Br J Psychiatry Suppl 2004;47:S64 6. 4. Ryan MC, Collins P, Thakore JH. Impaired fasting glucose tolerance in first-episode, drug-naive patients with schizophrenia. Am J Psychiatry 2003;160:284 9. 5. Arranz B, Rosel P, Ramirez N, Duenas R, Fernandez P, Sanchez JM, and others. Insulin resistance and increased leptin concentrations in noncompliant schizophrenia patients but not in antipsychotic-naive first-episode schizophrenia patients. J Clin Psychiatry 2004;65:1335 42. 6. Bergman RN, Prager R, Volund A, Olefsky JM. Equivalence of the insulin sensitivity index in man derived by the minimal model method and the euglycemic glucose clamp. J Clin Invest 1987;79:790 800. 7. Ferrari P, Alleman Y, Shaw S, Riesen W, Weidmann P. Reproducibility of insulin sensitivity measured by the minimal model method. Diabetologia 1991;34:52730. 8. Spranger J, Kroke A, Mohlig M, Bergmann MM, Ristow M, Boeing H, and others. Adiponectin and protection against type 2 diabetes mellitus. Lancet 2003;361:226 8. 9. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, fourth ed, text revision. Washington (DC): American Psychiatric Association; 2000. 10. Diagnosis and classification of diabetes mellitus. Diabetes Care 2005;28(Suppl1):S37S42. 11. Finegood DT, Hramiak IM, Dupre J. A modified protocol for estimation of insulin sensitivity with the minimal model of glucose kinetics in patients with insulin-dependent diabetes. J Clin Endocrinol Metab 1990;70:153849. 12. Salmeron J, Ascherio A, Rimm EB, Colditz GA, Spiegelman D, Jenkins DJ, and others. Dietary fiber, glycemic load, and risk of NIDDM in men. Diabetes Care 1997;20:54550. 13. Thakore JH, Mann JN, Vlahos I, Martin A, Reznek R. Increased visceral fat distribution in drug-naive and drug-free patients with schizophrenia. Int J Obes Relat Metab Disord 2002;26:13741.

Acknowledgements We thank Daniel Bois for his assistance with the experimental procedures and with data collection. This paper was presented at the International Congress on Schizophrenia Research, Savannah, Georgia, April 2005.

Manuscript received September 2005, revised, and accepted February 2006. 1 Lecturer, Department of Psychiatry, Faculty of Medicine, University of Toronto, Toronto, Ontario; Staff Psychiatrist, Schizophrenia Program, Centre for Addiction and Mental Health, Toronto, Ontario. 2 Professor, Department of Psychiatry, Faculty of Medicine, University of Toronto, Toronto, Ontario. 3 Clincial Director, Schizophrenia Program, Centre for Addiction and Mental Health, Toronto, Ontario. 4 Director, Medication Assessment Program for Schiophrenia, Schizophrenia Program, Centre for Addiction and Mental Health, Toronto, Ontario. 5 Codirector, Core Laboratory, Banting and Best Diabetes Centre, Faculty of Medicine, University of Toronto, Toronto, Ontario. 6 Director, Lipid Research Laboratory, JA Little Lipid Research Laboratory, St Michaels Hospital, Toronto, Ontario. 7 Professor, Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario. Address for correspondence: Dr Tony Cohn, Centre for Addiction and Mental Health, 1001 Queen Street West, Toronto, ON, M6J1H4; Tony_Cohn@camh.net

Can J Psychiatry, Vol 51, No 6, May 2006 W

385

The Canadian Journal of PsychiatryBrief Communication

Rsum : La rsistance linsuline et les concentrations dadiponectine chez des patients non mdicaments souffrant de schizophrnie : un rapport prliminaire
Objectif : Comparer la sensibilit insulinique (SI) de patients non mdicaments souffrant de schizophrnie ou de trouble schizo-affectif avec celle de volontaires en sant jumels. Mthode : Nous avons valu 9 patients non diabtiques gs de 26,6 ans (moyenne dge 26 ans, de 17 41 ans) et jumel des volontaires, laide du test de tolrance au glucose intraveineux chantillonn frquemment, de la nalyse de modle minimal, et des concentrations dadinopectine jeun. Rsultats : Lindice moyen de SI des patients tait 42 % plus bas que celui des volontaires en sant (P = 0,026), avec une compensation inadquate de scrtion dinsuline. Les patients souffrant de schizophrnie tendaient avoir des concentrations rduites dadiponectine (P = 0,055). Conclusions : Par une mesure directe, cette tude fournit la preuve de la rsistance linsuline et la susceptibilit au diabte de type 2 chez les patients souffrant de schizophrnie exempts dantipsychotique.

386

W Can J Psychiatry, Vol 51, No 6, May 2006