Prevalence of diabetes mellitus in chronic schizophrenic inpatients in relation to long-term antipsychotic treatment

Abstract

Background

Many reports indicate that the incidence and prevalence of diabetes mellitus is increased in schizophrenic patients and related to antipsychotic treatment. In an exploratory cross-sectional study we assessed the prevalence of type 2 diabetes mellitus in 266 chronic schizophrenic and schizoaffective inpatients and investigated whether the duration of antipsychotic treatment was related to the development of diabetes mellitus.

Method

We measured the non-fasting plasma glucose level in 266 inpatients with DSM IV diagnosis of schizophrenia or schizoaffective disorder in 5 different long-stay wards in the Netherlands. Measured variables were: age, sex, ethnicity, BMI, current antipsychotic treatment, duration of illness and duration of antipsychotic treatment.

Results

The overall prevalence of type 2 diabetes mellitus was 9%, which is significantly higher than the prevalence of 4.9% in the general population (OR 1.89, CI 1.14–3.13; p < 0.014). The prevalence was increased in two age cohorts: 30–39 years (3.8% vs. 0.3%, OR = 13.29, CI = 2.17–81.36, p = < 0.005) and 40–49 years (9.3% vs. 1.5%, OR = 6.74, CI = 2.77–16.38, p = 0.000). No new cases of diabetes mellitus were detected during the course of the study. The increased prevalence was found to be related to overweight and obesity. The time of exposure to antipsychotic treatment was not significantly correlated with the prevalence of diabetes mellitus when adjusted for age (F = 0.804, df = 1, p = 0.371, respectively, F = 0.194, df = 1, p = 0.660). Both typical and atypical antipsychotics contributed equally to the prevalence of diabetes mellitus.

Conclusion

No significant relation between long-term antipsychotic treatment and prevalence of diabetes mellitus was found. The high prevalence of diabetes mellitus in schizophrenic patients warrants screening of these patients already at young age for glucose disturbance.

Introduction

The relation between schizophrenia and diabetes mellitus has puzzled clinicians and researchers for more than a century. In 1899, long before the introduction of antipsychotic drugs, Sir Henry Maudsley wrote in “The pathology of mind” that “diabetes is a disease which often shows itself in families in which insanity prevails (Mukherjee et al., 1989)”. Since then, researchers have reported an increased prevalence of hyperglycemia and/or type 2 diabetes mellitus in schizophrenic patients (Kooy, 1919, Braceland et al., 1945, Freeman, 1946). The introduction of chlorpromazine (Delay et al., 1952) marked the beginning of a new era in this research: 3 years after its introduction, the first report of a diabetogenic effect of chlorpromazine appeared (Charatan and Bartlett, 1955) with case reports on new onset diabetes mellitus type 2 and worsening of existing diabetes mellitus following suit (Hiles, 1956, Amdisen, 1964). The case reports led, mainly but not exclusively in the USA, to the screening of entire hospital populations which, as a rule, showed a strong increase in the prevalence of diabetes mellitus in schizophrenic patients (Waitzkin, 1966a, Waitzkin, 1966b, Clayer and Dumbrill, 1967, Keskiner et al., 1973). The results of the investigation by Thonnard-Neumann (1968) can be taken as an example of this type of research: the prevalence of diabetes mellitus among 405 female psychiatric patients increased in 10 years from 4.2% to 17.2%. The only known exception is a study by Schwarz and Munoz (1968), who found normal prevalences of diabetes mellitus (2.5%) and hyperglycemia (3.2%) in 805 hospitalized psychiatric patients. In the nineties, case-reports on diabetes mellitus as a side effect of the atypical antipsychotics clozapine (Kamran et al., 1994, Dawn et al., 1997, Colli et al., 1999, Mohan et al., 1999, Brugman et al., 2000, Wehring et al., 2000, Wehring et al., 2003, Koller et al., 2001), olanzapine (Wirshing et al., 1998, Fertig et al., 1998, Lindenmayer and Patel, 1999, Goldstein et al., 1999, Bettinger et al., 2000, Melkersson et al., 2000, Koller and Doraiswamy, 2002), quetiapine (Sobel et al., 1999, Procyshyn et al., 2000, Domon and Cargile, 2002, Meyer et al., 2004, Misawa et al., 2004, Koller et al., 2004, Cohen, 2005), and risperidone (Mohan et al., 1999, Croarkin et al., 2000, Koller et al., 2003) were published, thereby renewing the interest in the subject. Analysis of the case-reports indicate that in the vast majority of the cases, clozapine and olanzapine are involved (Liebzeit et al., 2001, Lean and Pajonk, 2003, Cohen, 2004). However, more systematic studies on the diabetogenic effect of typical (Mukherjee et al., 1996, Dixon et al., 2001), atypical (Hägg et al., 1998, Melkersson et al., 1999, Henderson et al., 2000, Lund et al., 2001, Koro et al., 2002, Sernyak et al., 2002, Wang et al., 2002) or both typical and atypical antipsychotics (Gianfrancesco et al., 2002, Gianfrancesco et al., 2003, Wilson et al., 2002, Buse et al., 2003, Kornegay et al., 2002), gave conflicting results, suggesting that it is not only the atypical antipsychotics but also the classical antipsychotics, i.e., antipsychotic medication in general that is able to disturb glucose metabolism.

Weight gain, a well documented problem in the treatment of schizophrenia, both with classical (Allison et al., 1999) as with atypical drugs (Allison et al., 1999, Wetterling, 2001), would provide a logical explanation of this finding, but the involved pathophysiological mechanisms may be more complicated. Several explanations have been offered for drug-induced weight gain: increment of appetite through H1 receptor blockade alone (Wirshing et al., 1999, Kroeze et al., 2003) or a combination of receptor blockades (Baptista et al., 2002), prolactine (Baptista, 2002), a genetic polymorphism of the 5-HT_2c receptor (Reynolds et al., 2002, Reynolds et al., 2003) or the receptor genes for 5HT2c, β₃, α_1a or TNF-α (Basile et al., 2001). Few studies address energy expenditure as an explanation with inconclusive results: a study of 10 patients on olanzapine showed no effect on resting energy expenditure (Gothelf et al., 2002) while a case study found a decrease in both basal and 3-h energy expenditure (Virkkunen et al., 2002).

Animal research indicated that several mechanisms are involved in antipsychotic induced glycaemic disturbance: disruptions in insulin release (Ammon et al., 1973, Hermansen, 1978, Joost et al., 1983), increased insulin release (Melkersson et al., 2001, Melkersson, 2005), disturbed peripheral glucose utilization (Bonaccorsi et al., 1964, Bhide et al., 1965), altered expression of glucose transporters (Dwyer, 1999, Dwyer et al., 2001, Dwyer and Donohoe, 2003) and the catecholaminergic receptor system (Gupta et al., 1960, Ghafghazi et al., 1968, Jori and Carrara, 1966). More recently, the adequacy of the current animal model in rats has become the subject of debate (Baptista et al., 2004a, Baptista et al., 2004b).

Whatever the exact pathophysiological mechanism, treatment with antipsychotic medications in clinical practice seems to constitute an important factor in the onset of type 2 diabetes mellitus in schizophrenia. We decided to investigate whether the duration of antipsychotic exposure was related to the development of diabetes mellitus. We carried out a cross-sectional clinical investigation of the long-stay schizophrenic inpatients, who have sufficiently long duration of antipsychotic exposure to investigate such relationship. We have earlier reported on 93 long-stay inpatients (all included in the present study), in which we assessed the prevalence of disturbed glucose metabolism (Cohen et al., 2003). The present study, an extension of the previous study, tests the hypothesis that duration of treatment with antipsychotic medication would be related to the occurrence of type 2 diabetes mellitus.