Elsevier

Schizophrenia Research

Volume 80, Issue 1, 1 December 2005, Pages 19-32
Schizophrenia Research

Prevalence of the metabolic syndrome in patients with schizophrenia: Baseline results from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial and comparison with national estimates from NHANES III

https://doi.org/10.1016/j.schres.2005.07.014Get rights and content

Abstract

One important risk factor for cardiovascular disease is the metabolic syndrome (MS), yet limited data exist on its prevalence in US patients with schizophrenia.

Methods

Using baseline data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Schizophrenia Trial, assessment of MS prevalence was performed based on National Cholesterol Education Program (NCEP) criteria, and also using a fasting glucose threshold of 100 mg/dl (AHA). Subjects with sufficient anthropometric data, data on use of antihypertensives, hypoglycemic medications or insulin, and fasting glucose and lipid values > 8 h from last meal were included in the analysis. Comparative analyses were performed using a randomly selected sample from NHANES III matched 1 : 1 on the basis of age, gender and race/ethnicity.

Results

Of 1460 CATIE baseline subjects, 689 met analysis criteria. MS prevalence was 40.9% and 42.7%, respectively using the NCEP and AHA derived criteria. In females it was 51.6% and 54.2% using the NCEP and AHA criteria, compared to 36.0% (p = .0002) and 36.6% (p = .0003), respectively for males. 73.4% of all females (including nonfasting subjects) met the waist circumference criterion compared to 36.6% of males. In a logistic regression model with age, race and ethnicity as covariates, CATIE males were 138% more likely to have MS than the NHANES matched sample, and CATIE females 251% more likely than their NHANES counterparts. Even when controlling for differences in body mass index, CATIE males were still 85% more likely to have MS than the NHANES male sample, and CATIE females 137% more likely to have MS than females in NHANES.

Conclusions

The metabolic syndrome is highly prevalent in US schizophrenia patients and represents an enormous source of cardiovascular risk, especially for women. Clinical attention must be given to monitoring for this syndrome, and minimizing metabolic risks associated with antipsychotic treatment.

Introduction

There is a burgeoning literature on metabolic issues in patients with schizophrenia coinciding with a renewed focus on physical health care for this population (Meyer and Nasrallah, 2003, Marder et al., 2004). One basis for this concerns rests in data which demonstrate undertreatment of common medical conditions in patients with schizophrenia, and increased standardized mortality from natural causes (Allebeck, 1989, Mortensen and Juel, 1990, Newman and Bland, 1991, Simpson and Tsuang, 1996, Brown, 1997, Harris and Barraclough, 1998, Ruschena et al., 1998, Salazar-Fraile et al., 1998, Brown et al., 1999, Brown et al., 2000, Druss et al., 2000, Druss et al., 2001, Osby et al., 2000, Curkendall et al., 2001, Hiroeh et al., 2001). Among all causes of death, cardiovascular (CV) disease is responsible for as much as 50% of the excess mortality associated with the diagnosis of schizophrenia (Osby et al., 2000). There are many reasons why patients with schizophrenia have elevated CV risk, but one source of CV risk which may be overrepresented in this population is the cluster of findings that define the metabolic syndrome (MS), a core feature of which is abdominal obesity.

Patients with schizophrenia are at risk for developing obesity due to many factors including inactive lifestyle, poor dietary choices, and side effects of psychotropic medications (Wirshing and Meyer, 2003). The latter has received particular attention in the literature as accrued experience with atypical antipsychotics demonstrates marked weight gain liabilities with certain agents, particularly olanzapine and clozapine (Meyer, 2001). Yet, the correlation between obesity and schizophrenia, especially among female patients, is one that antedates the availability of atypical antipsychotics. Allison and colleagues', analysis of 1989 data on 150 individuals with schizophrenia from the 1989 National Health Interview Survey showed that females with schizophrenia had a significantly greater body mean mass index (BMI) than age-matched US norms from the same study (27.36 vs. 24.50 kg/m2, respectively, p < .001), and a trend, which did not reach statistical significance, for greater obesity among males with schizophrenia (Allison et al., 1999). Subsequent data for the decade 1987–96 confirms the observation that obesity is, at minimum, equally prevalent among male schizophrenia patients compared to the general population, while females with schizophrenia are at greater risk (Homel et al., 2002).

The MS, also called the dysmetabolic syndrome or Syndrome X, has been discussed in the cardiology and endocrinology for over two decades, but the last five years has seen an explosion of publications in this area. The metabolic syndrome is defined by a cluster of clinical features that include increased abdominal or visceral adiposity (measured by waist circumference), atherogenic dyslipidemia (low high density lipoprotein (HDL) and elevated fasting triglycerides), hypertension, and impaired fasting glucose or overt diabetes mellitus (DM) (Expert Panel on Detection , 2001). The definition in Table 1 from the National Cholesterol Education Program (NCEP) is commonly used, although recent consensus panels [National Heart, Lung and Blood Institute and American Heart Association (hereafter referred to as AHA)] suggest incorporating the new lower threshold for impaired fasting glucose of 100 mg/dl (Grundy et al., 2004). In the United States, increased attention to MS followed the NCEP publication of its third Adult Treatment Protocol (ATPIII) in 2001 (Expert Panel, 2001). By highlighting MS as a condition worthy of clinical attention due to its association with increased CV risk, and providing clinicians with easily verifiable clinical criteria, ATPIII facilitated identification of persons with the syndrome. It is worth noting that there is a modified version, promulgated by the World Health Organization (WHO), which uses elevated fasting insulin levels as one criterion, and is less applicable in routine clinical settings. There is broad overlap between the WHO and NCEP definitions, and a reanalysis of the third National Health and Nutrition Examination Survey (NHANES III) data using both definitions found that, for the 8608 subjects ages 20 and up, MS status was classified identically by both criteria for 86.2% of subjects. Prevalence estimates between the two definitions differed for some subgroups, notably for African-American men (Ford and Giles, 2003).

The MS is highly prevalent, with 47 million persons in the US meeting criteria based upon estimates derived from NHANES III (Ford et al., 2002). The age-adjusted prevalence of MS from NHANES III was 23.7%, with the lowest prevalence (6.7%) in the cohort ages 20–29, and the highest (43.5%) in those ages 60 and over. Ethnicity was also an important predictor of MS prevalence, with Hispanic subjects having the highest age-adjusted prevalence of MS (31.9%). Importantly, the effect of ethnicity persisted even after adjustment for age, BMI and socioeconomic status, indicating that certain individuals may have a biological predisposition towards storing excess fat intraabdominally (Park et al., 2003). Overall, gender was not as predictive of increased risk, with nearly identical age-adjusted prevalences for women (23.4%) and men (24.0%); however, African American and Hispanic women were notable exceptions, with prevalences 57% and 26% greater, respectively than their male counterparts (Park et al., 2003). More recent data from NHANES 1999–2000 indicate that female gender may be associated with increased risk among all ethnicities. The age-adjusted MS prevalence in NHANES 1999–2000 showed a 23.5% increase among females compared to NHANESIII, but only 2.2% increase for males (Ford et al., 2004).

While diabetes mellitus represents a disease entity with high CV risk, the presence of MS defines a cohort who may be at even greater risk. Cross-sectional data from NHANES III found the prevalence of coronary heart disease (CHD) to be significantly higher among nondiabetic patients with MS (13.9%) than in diabetic patients who did not meet criteria for the syndrome (7.5%) (Alexander et al., 2003). Moreover, data from a large Scandinavian trial revealed that a diagnosis of MS was associated with a 3-fold increased risk for both CHD and stroke over the median follow-up of 6.9 years (Isomaa et al., 2001).

The MS significantly increases risk for CV disease, and, in those without diabetes, represents a prediabetic state which, over time, progresses to overt diabetes in a significant proportion of individuals. Evidence for this progression comes from NHANES III, which found that only 13% of diabetics did not meet MS criteria among the cohort over age 50 (Alexander et al., 2003). Embedded in the choice of waist circumference as one criterion of MS is the concept that abdominal or visceral fat deposition represents a greater risk for future development of diabetes than other phenotypes of obesity (i.e., fat deposition on the hips and buttocks). This concept is illustrated by the Hoorn Study, a prospective trial of 1342 nondiabetics ages 50–75 residing in Hoorn, Netherlands. After 6.4 mean years of follow-up, central obesity, as measured by waist–hip ratio (WHR), was a significant predictor of conversion to type 2 DM, whereas BMI was not (De Vegt et al., 2001).

What is the connection between visceral adiposity (as measured directly by imaging, or indirectly by waist circumference) and glucose intolerance that places individuals with this phenotype at greater risk for future DM than those who deposit fat on the thighs and buttocks? Aside from the epidemiological evidence, results from studies examining measures of insulin sensitivity and disposition point to a clear biological correlation between intraabdominal or visceral fat mass and glucose–insulin homeostasis (Wagenknecht et al., 2003). One important result of the Insulin Atherosclerosis Family Study was the finding that body mass index (BMI) was not a significant contributor to the linear model of insulin sensitivity, while visceral adiposity (i.e., intraabdominal fat), and its interaction with subcutaneous fat, contributed to 27% of the variance in insulin sensitivity (Wagenknecht et al., 2003). This result confirms the findings of numerous other smaller studies which demonstrate that visceral adiposity is a significant predictor of insulin sensitivity, even more than BMI (Goodpaster et al., 1999).

Diabetes mellitus has garnered much of the recent attention in the schizophrenia metabolic literature due to data associating atypical antipsychotics, especially clozapine and olanzapine, with new onset diabetes or diabetic ketoacidosis (American Diabetes Association et al., 2004). Studies indicate that DM may be twice as prevalent among patients with schizophrenia (approximately 14%) as in the general population (7%) (Dixon et al., 2000), but there are emerging data which show that MS is even more prevalent than DM among schizophrenia patients. As of this writing, there are but three published prevalence studies of MS in schizophrenia patients. Heiskanen and colleagues, using the NCEP criteria for MS, were the first to publish MS prevalence data among the mentally ill using a sample of 35 Finnish outpatients with schizophrenia (Heiskanen et al., 2003). The sample prevalence of 37% was 2 to 4 times higher than the prevalence reported for the surrounding geographical area in Eastern Finland. Another study by Basu et al. of 33 outpatients with schizoaffective disorder (mean age 44.5 years) enrolled in a clinical trial noted a prevalence of 42.4% (Basu et al., 2004), nearly double the prevalence for the cohort ages 40–49 in NHANES III (Ford et al., 2002). The best published estimate is from Cohn and colleagues Canadian study of 240 subjects with schizophrenia or schizoaffective disorder (65% male, mean age 43.3 years), which found a prevalence of 42.6% for males and 48.5% for females using the NCEP criteria (Cohn et al., 2004). Cohn also reported equal prevalences for those under age 45 (43.8%) and those ages 45 and over (45.8%), a finding in stark contrast to the age-related increase in MS prevalence noted in the general population.

There are other data sets available, mostly unpublished, summarized in Table 2. Except for the Cohn study, the sample sizes are small, limiting the generalizability, and in one instance, subjects were selected for the study on the basis of having olanzapine-related obesity (Pandina et al., 2004). Nonetheless, when one considers that the mean age of the schizophrenia patients studied in these data sets is roughly 43 years old, the cited prevalences are all greater than the MS prevalence in NHANES III for the cohort ages 40–49 of 24% for males and 20% for females (Ford et al., 2002).

There is currently a need for data on MS prevalence from large samples of patients with schizophrenia to accurately assess the prevalence of this phenomenon cross-sectionally, and examine the features of MS in this patient cohort, especially when compared with the general population. With the broad recruitment strategy of the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Schizophrenia Trial, its multi-site design, and large sample size, the baseline data obtained offers an excellent opportunity to explore the demographics of MS among US patients with schizophrenia, and perform comparisons with matched general population data. Specifically, we sought to test the hypothesis that MS is more prevalent among patients with schizophrenia than demographically matched counterparts in the general population, and that this increased risk would still be evident even after controlling for differences in body mass between the two cohorts.

Section snippets

Methods

The methods for the CATIE Schizophrenia Trial have been published in detail previously (Stroup et al., 2003). Briefly, the CATIE Schizophrenia Trial is a national, multisite, NIMH-sponsored prospective trial of antipsychotic effectiveness in patients with schizophrenia that broadly assesses metabolic, symptom, neurocognitive and functional outcomes. Institutional Review Board approval was obtained at each site, and subjects voluntarily enrolled after having been provided informed consent in

Results

There were 1460 subjects enrolled in the CATIE Schizophrenia trial at baseline, and 686 had sufficient data for assessment of MS status using NCEP criteria, (687 using the AHA modified criteria), with fasting laboratory measures obtained 8 or more hours since last meal. Table 3 lists the distribution by race and ethnicity, which were coded as separate variables in the database. A demographic comparison of the fasting cohort with those who had random laboratory values (< 8 h from last meal),

Discussion

Presented here is the largest data set on MS in patients with schizophrenia, with results which confirm the findings of smaller studies that this metabolic disorder is more prevalent in these patients than among matched individuals from the general population. Female patients with schizophrenia appear to be particularly vulnerable both to central obesity and MS, and represent a very high risk group not previously identified in the general literature on MS. While females of all races and

References (50)

  • S. Brown et al.

    The unhealthy lifestyle of people with schizophrenia

    Psychological Medicine

    (1999)
  • S. Brown et al.

    Causes of the excess mortality of schizophrenia

    British Journal of Psychiatry

    (2000)
  • T. Cohn et al.

    Characterizing coronary heart disease risk in chronic schizophrenia: high prevalence of the metabolic syndrome

    Canadian Journal of Psychiatry-Revue Canadienne de Psychiatrie

    (2004)
  • S.M. Curkendall et al.

    Increased cardiovascular disease in patients with schizophrenia

  • F. de Vegt et al.

    Relation of impaired fasting and postload glucose with incident type 2 diabetes in a Dutch population: the Hoorn study

    JAMA

    (2001)
  • L. Dixon et al.

    Prevalence and correlates of diabetes in national schizophrenia samples

    Schizophrenia Bulletinn

    (2000)
  • B.G. Druss et al.

    Mental disorders and use of cardiovascular procedures after myocardial infarction [see comments]

    JAMA

    (2000)
  • B.G. Druss et al.

    Quality of medical care and excess mortality in older patients with mental disorders

    Archives of General Psychiatry

    (2001)
  • Expert Panel on Detection and Evaluation and Treatment of High Blood Cholesterol in Adults

    Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III)

    JAMA

    (2001)
  • E.S. Ford et al.

    A comparison of the prevalence of the metabolic syndrome using two proposed definitions

    Diabetes Care

    (2003)
  • E.S. Ford et al.

    Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey

    JAMA

    (2002)
  • E.S. Ford et al.

    Increasing prevalence of the metabolic syndrome among US adults

    Diabetes Care

    (2004)
  • B.H. Goodpaster et al.

    Effects of weight loss on regional fat distribution and insulin sensitivity in obesity

    Diabetes

    (1999)
  • S.M. Grundy et al.

    Definition of metabolic syndrome: report of the National Heart, Lung, And Blood Institute/American Heart Association Conference on scientific issues related to definition

    Circulation

    (2004)
  • E.C. Harris et al.

    Excess mortality of mental disorder

    British Journal of Psychiatry

    (1998)
  • Cited by (1030)

    • Diabetes and Mental Health

      2023, Canadian Journal of Diabetes
    View all citing articles on Scopus

    Supported by NIMH grant #N01MH90001.

    View full text