Dr. Beaser: I'm Dr. Richard Beaser, and I'd like to welcome you to Joslin Diabetes FOCUS. It is now recognized that the person with schizophrenia or bipolar disorders is at increased risk for developing diabetes or one of the other manifestations of the metabolic syndrome, such as obesity, hypertension, or dyslipidemia. And the mental health provider is often at the center of the resulting treatment issues and decisions.

What is the etiology of this increased risk? Is it lifestyle? Is it genetics? Is it a side effect of the antipsychotic medications? The potential for some antipsychotic medications to play a role in the etiology of the metabolic syndrome has been suggested by recent reports in the literature. While interpreting the true implications of these reports and studies can be confusing, the mental health provider is often faced with making crucial decisions regarding the potential for medication-induced risk based on study data. These decisions often impact medication choice or the need for medication change.

In addition, people with schizophrenia or bipolar disorders often see their mental health providers more frequently than their primary care provider, if they have a primary care provider at all. The mental health provider is often the one who identifies the at-risk patient and evaluates the impact of antipsychotic treatment design. This clinician is often the first to recognize signs or symptoms of the metabolic syndrome or acute manifestations, such as diabetic ketoacidosis (DKA), and thus must ensure that timely preventive or treatment strategies are initiated. Primary care providers and diabetes specialists must also understand the relationship between these psychiatric disorders and the metabolic syndrome, and appreciate the challenges of initiating preventive strategies in this patient population.

Therefore, the challenge we face is to reduce the risk of developing the metabolic syndrome while offering effective antipsychotic therapy. Integral to treatment is an understanding of what the current state of knowledge gleaned from recent literature reports does and does not tell us. Once any components of the metabolic syndrome are present, aggressive strategies to prevent macrovascular damage must begin. The presence of any components of this syndrome should signal the initiation of aggressive strategies to prevent diabetic complications, including lifestyle changes, use of medications to treat these components, and the minimization of other factors that could increase insulin resistance.

Today's program seeks to explore current perspectives on the etiology of this increased macrovascular risk among those with schizophrenia or bipolar disorders, highlighting the interpretation and implications of recent literature reports. An overview of preventive and treatment strategies will be highlighted. I would like to start by introducing our faculty. Dr. Philip Levy is Clinical Professor of Medicine at the University of Arizona College of Medicine in Tucson. And Dr. John Newcomer is Associate Professor of Psychiatry at the Washington University School of Medicine in St. Louis, Missouri.

Dr. Beaser: John, there's certainly been a lot of recent attention to the observation that people with schizophrenia or bipolar disorders seem to have an increased risk of developing diabetes, as well as other components of the metabolic syndrome. Perhaps you could open our discussion with an overview of the epidemiology of this association.

Dr. Newcomer: Severe and persistent mental disorders, such as schizophrenia and bipolar disorder, can have a devastating impact on psychological, social, and work functioning, and can also increase the risk of morbidity and mortality from major medical disorders. Antipsychotic treatment can improve psychological, social, and work functioning. Newer antipsychotic treatments have demonstrated superior or additional effects in several areas of outcome. However, antipsychotic treatment may also contribute to medical risks. The major goal of treatment for individuals with schizophrenia and bipolar disorder is to improve psychiatric outcomes and to minimize medical risk factors. In addition to recognizing and treating psychiatric symptoms, the role of the psychiatric professional includes assuring screening for major modifiable risk factors, and understanding the contributions of patient factors, such as nutrition and activity level, and treatment factors, such as weight gain, incontributing to medical risk. Finally, the psychiatric professional must understand how and when to respond to medical risk factors with interventions and appropriate referrals.

Schizophrenia affects approximately 1 out of every 100 individuals worldwide. It is a chronic and disabling illness that has an impact on patients, caregivers, and society. Approximately 70% of patients improve or respond to treatment with antipsychotic drugs. However, the treatment response is often partial or suboptimal, and outcomes can be complicated by medical morbidity and mortality. Treatment compliance with antipsychotic medications tends to be poor and worsened by adverse events. With older or typical antipsychotic medications, approximately half of patients may be poorly compliant after 1 year. Individuals with schizophrenia have an approximately 20% shorter life expectancy in comparison with the general population.

Dr. Newcomer: They experience higher standardized mortality rates for cardiovascular disease, diabetes mellitus, respiratory disease, and infectious diseases. Cardiovascular disease, referring to the combination of cerebrovascular and coronary heart disease, accounts for most of the excess mortality in schizophrenia. This is due to schizophrenia patients having 2 times the general population rates, the general population in developed nations having cardiovascular disease as the leading cause of mortality. While clinicians routinely work to address the 10% of patients with schizophrenia who will die by suicide, modifiable risks for cardiovascular disease remain prevalent. There's growing appreciation that primary prevention for cardiovascular disease may be especially important in patients with schizophrenia and bipolar disorder.

Dr. Newcomer: Major risk factors for cardiovascular disease include overweight and obesity, hypertension, smoking, dyslipidemia, and hyperglycemia. Having any 1 risk factor for cardiovascular disease is associated with a 2-fold increase in the odds ratio for disease, but multiple risk factors produce additive or greater risk. I have highlighted 3 of the 5 risk factors to underscore that patients taking medications that increase body weight and adiposity can experience 3 of 5 major risk factors. A major contributor to mortality from cardiovascular disease and diabetes mellitus is increased body fat, or adiposity.

Dr. Newcomer: Patients with schizophrenia suffer an increased prevalence of overweight and obesity, as well as other major modifiable risk factors for cardiovascular disease. There is particularly well-established evidence that the prevalence of obesity and overweight in schizophrenia is increased in comparison with the general population and that medication effects can contribute to this problem. Increased adiposity can disturb glucose and lipid metabolism, further increasing macrovascular and cardiovascular risk.

Dr. Beaser: The metabolic syndrome seems to be at the heart of the problem for these people. Phil, perhaps you could provide more details about this constellation of metabolic abnormalities and its role in increasing macrovascular and cardiovascular risk.

Dr. Levy: There are several components of the insulin resistance syndrome or metabolic syndrome. Diabetes is a common component. This syndrome leads to premature atherosclerosis and premature death, and sedentary lifestyle and obesity play important contributory roles. Components also include hypertension and dyslipidemia. The characteristic dyslipidemia is that of low high-density lipoprotein cholesterol and high triglycerides.

Dr. Levy: A critical point about diabetes was observed in a study by Haffner and colleagues. In this study, patients with diabetes were compared with an equivalent population without diabetes. The 7-year incidence of myocardial infarction (MI) in patients with diabetes and no previous history of heart disease was the same as a nondiabetic population with a previous history of MI, indicating that the person with type 2 diabetes has heart disease until proven otherwise. While diabetes mellitus is now considered a risk equivalent for cardiovascular disease, the insulin resistance syndrome is a major risk factor as well.

Dr. Levy: The leading cause of death in diabetes is ischemic heart disease. Diabetes is also associated in many cases with asymptomatic heart disease. Insulin resistance and hyperinsulinemia occur for many years preceding the onset of diabetes mellitus. As long as the pancreas can make enough insulin, the blood glucose levels will remain normal and the patient will not have diabetes until a relative or absolute insulin deficiency occurs. The pathophysiologic defects of type 2 diabetes include both hyperinsulinemia and insulin resistance and insulin deficiency. The insulin resistance syndrome is an integral part of type 2 diabetes in the vast majority of patients. Type 2 diabetes, therefore, has a dual pathophysiologic defect. The defect that occurs first is the insulin resistance followed by a decrease in insulin secretory capacity. Several studies have been done to try to prevent diabetes in people who are at high risk. Lifestyle changes seem to be the most effectivestrategy, but may also be very difficult to achieve in the schizophrenic population.

Dr. Levy: The insulin resistance syndrome has both inherited and acquired influences. The inherited influences cannot be modified, but many of the acquired influences can be modified, including overeating, obesity, inactivity, and use of diabetogenic medications.

Dr. Levy: In addition to clinical manifestations, the insulin resistance syndrome is associated with biochemical abnormalities, which can lead to premature atherosclerosis, including carbohydrate, lipid, and clotting abnormalities.

Carbohydrate abnormalities include glucose intolerance, hyperinsulinemia, and insulin resistance.

Dr. Levy: And finally, the National Cholesterol Education Program has given us defining characteristics for the metabolic syndrome. Waist circumference is important. In fact, some say that if the patient is preceded in the examining room by the abdomen, then that patient has the metabolic syndrome.

Any 3 of the 5 risk factors make the diagnosis by these criteria: 20% to 25% of the population at large has abnormalities consistent with the metabolic syndrome and 60% of the obese population has these abnormalities.

Dr. Beaser: So increased weight and adiposity may be one factor that contributes to why people with schizophrenia and bipolar disorders are more likely to develop diabetes and the other components of the metabolic syndrome. But what are some of the other possible contributing factors? This question has been the center of much debate in the last couple of years, and there is still much that is not resolved. It has been suggested that the increased incidence of diabetes and the metabolic syndrome is a result of a genetic link. Others suggest that it results from a schizophrenic lifestyle, which might include variations in nutrition and activity. Still others argue that the increased risk is brought on by some of the medications that are used to treat these conditions. While all of these factors may well play a role, it is difficult to determine to what degree each has an impact and what should be the clinical response to this concern. Further compounding the confusion is thatmuch of the literature consists of case reports, poorly controlled studies, or retrospective analyses of populations where the definition of diabetes and/or the means to detect it may be inconsistent or based on older standards. Yet this is an important enough issue that we should clarify what evidence exists to explain the etiology of this problem. John, perhaps you can start by discussing some of the factors, including lifestyle changes, which may increase metabolic risk in these patient populations.

Dr. Newcomer: There is intriguing evidence that major psychiatric conditions such as depression, bipolar disorder, schizophrenia, and Alzheimer's disease are associated with increased prevalence of insulin resistance or type 2 diabetes mellitus. Several large epidemiological studies support this conclusion for depressive disorders, with limited evidence -- some prior to the introduction of antipsychotics -- that schizophrenia patients have more insulin resistance or diabetes compared with controls.

Lifestyle factors can certainly contribute to the risk of developing diabetes mellitus or metabolic complications in individuals with schizophrenia and bipolar disorder, with some speculation that disease-specific factors, genetic or otherwise, could also contribute. The so-called negative symptoms of schizophrenia, involving loss of social skills, will, drive, and motivation, may be important. Many patients with schizophrenia suffer a downward drift in socioeconomic status with adverse effects on diet and access to medical care and medical education. While some patients have increased activity levels related, for example, to a homeless lifestyle, many more patients suffer social isolation and reduced activity levels along with greatly increased rates of smoking. Food intake in these individuals may be one of the few pleasures in their lives. We do know that adverse events associated with antipsychotic treatment can contribute to metabolic complications. The most clearlyestablished relevant adverse treatment effect is weight gain.

Dr. Newcomer: Prevalence estimates for type 2 diabetes mellitus in schizophrenia and bipolar disorder are generally 2 times greater than age-corrected general population rates, with no evidence for more type 1 diabetes. A single report describes increased prevalence of glucoregulatory disturbance, not only in schizophrenia patients but also in their first-degree nonpsychotic relatives, suggesting either genetic or familial effects that may contribute to this problem. We know that adiposity and family history are major risk factors for developing diabetes. Unfortunately, these factors were not quantified in early reports concerning drug-naive patients. There is clearer evidence that the introduction of antipsychotic medications increased prevalence rates of diabetes. Those early reports concerning medication effects on prevalence rates typically involved the first antipsychotic, chlorpromazine, and other low-potency phenothiazines, which were all associated with prominentweight gain. By the 1980s, most patients with schizophrenia were treated with higher-potency typical antipsychotics, such as haloperidol and fluphenazine, with modest effects on weight and limited evidence for effects on glucose metabolism.

Dr. Newcomer: Even prior to the introduction of the newer atypical antipsychotics, early public health service data indicated a higher prevalence of overweight and obesity in individuals with schizophrenia vs the general population. It is presumed that lifestyle disturbances and the potential medication effects of the older conventional drugs on weight would contribute to this problem.

Dr. Newcomer: Other elements of the metabolic syndrome, such as dyslipidemia and hypertension, have received far less study in patients with schizophrenia and bipolar disorder. We know, however, that increases in weight are associated with dyslipidemia and that increases in weight, insulin resistance, and smoking are associated with increased risk of hypertension.

Dr. Newcomer: Cigarette smoking is another major risk factor that is prevalent in patients with schizophrenia. Best estimates suggest that 60% to 75% of patients may smoke in comparison to an overall United States prevalence of about 25%. This is not only a risk factor for cardiovascular disease and respiratory disease, but smoking is also associated with increased insulin resistance and may contribute to metabolic complications in this population.

Dr. Newcomer: In general, we know less about medical risk factors in patients with bipolar disorder than we do in patients with schizophrenia. Increased rates of obesity, hyperglycemia, type 2 diabetes, and dyslipidemia have been reported in bipolar patients. Disturbances in diet, activity, and other lifestyle factors are also reported to be similar to patients with schizophrenia. In addition, medication-induced weight gain is similar across studies involving patients with schizophrenia and bipolar disorder.

Dr. Beaser: Clearly, obesity goes hand in hand with the metabolic syndrome, especially in this population. Phil, could you elaborate on the role that obesity plays in the etiology of the metabolic syndrome?

Dr. Levy: The incidence of diabetes and obesity are increasing at a rapid rate in the United States, and they tend to parallel each other. Here we see the change in the incidence of obesity in the United States from a 10-year period, 1991 to 2001, using telephone survey data, which probably underestimate the numbers. There does not appear to be one state in our union that has escaped this unfortunate trend.

Dr. Levy: And here we see the change in the incidence of diabetes over that same 10-year period. Although the diagnostic criteria for both diabetes and obesity have changed, there is no question that there is an epidemic increase in both of these entities.

Dr. Levy: There is also a relationship of body mass index (BMI) and the relative risk of diabetes. Obesity today is classified as a BMI of 30 kg/m² or greater. Much of the schizophrenic population falls into this category. Therefore, the population should be considered at high risk for developing diabetes.

Dr. Levy: The American Diabetes Association (ADA) has criteria for those who should be screened for diabetes. Note that all adults aged 45 years and older should be screened, and that people who are 20% or more above ideal body weight should be screened as well. High-risk individuals may have to be screened at much younger ages. The easiest way to do a screen is to do a fasting blood glucose, but a random blood glucose can be used as an alternative. A fingerstick blood glucose is not diagnostic, but if elevated would suggest a need for further testing. Although obesity is not necessarily a causative factor for the metabolic syndrome or diabetes, it certainly contributes. Most patients with type 2 diabetes are overweight, and modest weight loss can improve the metabolic factors of diabetes and, in some cases, can reverse diabetes. Diabetes and obesity are on a rampage in this country, and they go together like supersizing and gluttony. It is important for us as healthcareproviders to be aware of this epidemic and to make efforts to reverse it. Diabetes and obesity lead to premature atherosclerosis and death. Lifestyle changes and decreased calorie intake can prevent this lethal progression.

Dr. Beaser: Antipsychotic medications have also been postulated to play a role in the etiology of the metabolic syndrome in this population. John, could you outline the evidence for and against this postulate?

Dr. Newcomer: The role of antipsychotic medications in causing or contributing to metabolic adverse events can be understood to a large degree in terms of medication effects on weight and adiposity, previously described effects of adiposity on insulin resistance, as well as evidence that treatment with certain antipsychotic medications is associated with insulin resistance, hyperglycemia, dyslipidemia, type 2 diabetes, or diabetic complications.

Dr. Newcomer: This illustrates one of several analyses concerning the weight gain potential of antipsychotic medications, in this case calculating mean weight change after 10 weeks of standardized drug treatment with 95% confidence intervals for various treatment conditions. On the right-hand side of the slide, the analysis demonstrates that treatment with olanzapine, clozapine, and low-potency phenothiazines for 10 weeks produces approximately 4 kg of mean weight gain. Atypical antipsychotics, such as risperidone and quetiapine, can be found in the middle of the slide, producing approximately 2 kg of mean weight gain over this same time period, with agents such as ziprasidone, haloperidol and, not shown on this slide, a newer agent, aripiprazole, all producing 1 kg or less weight gain over 10 weeks.

Dr. Newcomer: The US Food and Drug Administration (FDA) approved package label defines weight gain as a greater than or equal to 7% increase in body weight during treatment. This slide illustrates pooled multicenter clinical trial data for the newer antipsychotic medications during 4- to 8-week clinical trials. For each medication, the appropriate comparison is with its own placebo control, offering a relative risk incidence estimate. Moving from left to right, aripiprazole and ziprasidone treatment are associated with approximately twice the placebo rate of weight gain. Despite higher absolute incidence numbers, risperidone treatment is similarly associated with approximately twice the placebo rate of weight gain. Quetiapine treatment produces over 3 times the placebo rate of weight gain. Notably, olanzapine treatment produces approximately 10 times the placebo rate of weight gain using the FDA definition.

Dr. Newcomer: There is growing evidence that antipsychotic-induced weight gain occurs largely due to an increase in appetite, and that histamine type 1 receptor affinity is a strong predictor of weight gain potential. Weight gain can be understood as an imbalance between caloric intake and caloric expenditure. The excess caloric intake required to gain 4 kg over 10 weeks, as seen with olanzapine, is predicted to produce substantial additional weight gain over subsequent weeks and months. Long-term clinical trial data confirm these predictions, with the largest datasets coming from pivotal trials detailed in the package inserts. Olanzapine treatment at standard doses produces more than 10 kg of mean weight increase over the first year of treatment, in contrast to approximately 2-3 kg of mean weight increase over 1 year of treatment with risperidone or quetiapine, and only 1 kg of mean weight increase over 1 year of treatment with aripiprazole or ziprasidone. Even small dailyimbalances between caloric intake and expenditure can lead to weight change. For example, excess intake of 125 kcal per day, a small candy bar, can lead to a 6-kg increase in body fat over 1 year.

Dr. Newcomer: Convergent evidence from case reports and case series, prospective observational studies, retrospective database analyses, and controlled analytic studies including randomized clinical trials, support the conclusion that medications that induce greater weight gain, such as olanzapine and clozapine, are associated with greater risk for insulin resistance and diabetes mellitus. This is consistent with several decades of research concerning the adverse effects of increasing adiposity.

Dr. Newcomer: This slide summarizes available case reports based on published literature, including recent analyses of FDA MedWatch postmarketing surveillance data, as well as unpublished postmarketing data from manufacturers. In general, there are more cases of new-onset diabetes mellitus, DKA and related complications, and deaths associated with clozapine and olanzapine treatment. Counting cases is a limited method, for example, due to potential bias in case reporting and particularly the problem of under-reporting. Only 1% to 10% of adverse events are reported to the FDA MedWatch system, and only in the first couple of years after a medication is launched. It has not been possible to explain away case counts based on prescription frequency or prescription years. Risperidone, for example, was launched before olanzapine and maintains a greater history of prescription years, but also has relatively smaller numbers of reported metabolic adverse events.

Dr. Newcomer: There are more than 15 retrospective database analyses on this topic, many of them industry sponsored, unpublished, and not peer reviewed. The basic methodology is to look within a large administrative or insurance database and measure the association between antipsychotic drug treatment and an insensitive, unreliable surrogate indicator of type 2 diabetes, such as the prescription of a hypoglycemic agent or a relevant International Classification of Diseases, 9th edition (ICD-9) code. None involve prospective testing of validated plasma glucose indicators of diagnosis.

Approximately two thirds of the studies, such as this report from Koro et al, indicate that drugs that induce more weight gain, for example olanzapine, are associated with increased risk of diabetes mellitus in comparison with either no treatment, conventional treatment, or a drug producing less weight gain, for example risperidone. No study indicates that a drug with high weight gain potential, such as olanzapine, has lower risk of diabetes in comparison with any alternative treatment. Approximately one third of 18 studies find no differences between groups or generally increased association for treated groups vs untreated controls. Unfortunately, some of the unpublished results, even when they are inconsistent with the majority of previous analyses, can end up receiving considerable coverage in the media, leading to confusion.

Dr. Newcomer: There are a number of limitations to these retrospective database analyses. For example, a limited history of prior drug exposure makes it difficult to control for prior treatment effects (eg, a 20-kg weight gain on prior treatment certainly has an impact on current risk of diabetes, but may not be reflected in this kind of analysis). The most important problem with retrospective database analyses involves the use of insensitive, surrogate diagnostic indicators for diabetes (eg, the prescription of a hypoglycemic drug or ICD-9 codes). Because the ADA estimates that more than 33% of individuals with diabetes are undiagnosed, these individuals will have no hypoglycemic prescription or ICD-9 code to use as identifiers. The error rate in making the diagnosis may be greater than the effect that you're looking for.

Dr. Newcomer: A growing number of studies that utilize sensitive measures of insulin resistance and glucose regulation report adiposity-related decreases in insulin sensitivity. In this National Institute of Mental Health-funded study, we examined the relationship between BMI and a gold-standard measure of insulin sensitivity, using a euglycemic clamp procedure in chronically treated patients with schizophrenia and untreated healthy controls. Humans with schizophrenia are, not surprisingly, physiologically similar to untreated healthy human controls, with all subjects showing adiposity-related decreases in insulin sensitivity.

Dr. Newcomer: Based on reports indicating that approximately 25% of the new-onset type 2 diabetes occurs in the absence of substantial weight gain or obesity, several groups have looked for adiposity-independent differences in insulin sensitivity during treatment with antipsychotics. Here we studied chronically treated patients and untreated healthy controls with fasting oral glucose tolerance tests, with all groups matched on the key risk variables of adiposity, age, and ethnicity. Despite equivalent adiposity and other risks, we detected treatment-related differences in fasting and postload plasma glucose and insulin resistance. This result is similar to that of Henderson and colleagues using intravenous (IV) glucose tolerance tests and another validated measure of insulin sensitivity. These combined results suggest that some patients can experience drug-related differences in glucose regulation independent of changes in adiposity. Most patients, however, will remainvulnerable to well-characterized, adiposity-related decreases in insulin sensitivity.

Dr. Newcomer: There are several ongoing, prospective, randomized antipsychotic treatment studies involving sensitive metabolic measures. This Pfizer-funded study compared 6 weeks of double-blind treatment with ziprasidone vs olanzapine. Despite no differences in efficacy for psychiatric outcomes, significant differences were detected on several measures of metabolic outcome. Significant increases in weight, BMI, fasting insulin, and homeostasis model assessment measures of insulin resistance were detected in the olanzapine treatment arm. No significant increase in fasting plasma glucose was observed during olanzapine treatment in these relatively young, healthy patients over the brief 6-week treatment. No significant increase in any metabolic parameter was observed during ziprasidone treatment.

Dr. Newcomer: Several decades of research indicate that plasma lipids tend to increase with body weight and adiposity. Not surprisingly, therefore, increases in plasma cholesterol or triglycerides have been reported in association with treatment using some of the newer antipsychotic agents. Neutral effects on plasma lipids have been reported with those newer antipsychotic agents associated with minimal weight gain.

Dr. Newcomer: Looking again at the Pfizer-funded, 6-week, double-blind comparison of olanzapine and ziprasidone treatment, fasting lipid profiles were examined at baseline and endpoint. Along with previously mentioned increases in weight, adiposity, and insulin resistance, olanzapine induced significant increases in both total and low-density lipoprotein (LDL) cholesterol, as well as triglycerides.

Dr. Newcomer: Koro and colleagues reported on a study from the United Kingdom involving more than 18,000 individuals in a population-based sample. The odds of developing hyperlipidemia during olanzapine treatment were approximately 5 times higher than for individuals with no exposure to the antipsychotic, and approximately 3 times higher than for individuals taking conventional antipsychotic agents. Risperidone treatment was not associated with significant increases in the odds of hyperlipidemia in comparison with either untreated persons or those receiving conventional antipsychotics.

Dr. Beaser: From what has been said so far, one thing does seem clear. We need to pay close attention to diabetes and its related metabolic abnormalities, and when present, control them aggressively. Phil, can you review for our audience how they can do this?

Dr. Levy: The control of diabetes has been shown to be important in reducing the microvascular complications of retinopathy, nephropathy, and neuropathy. Diabetes control is also associated with a decrease in macrovascular complications, such as heart attack and stroke, although this decrease has not been statistically significant. The United Kingdom Prospective Diabetes Study (UKPDS) is the largest prospective study of type 2 diabetes ever done. The study started in the United Kingdom in the 1970s with more than 5000 patients with diabetes, and this group is still being observed today.

Hemoglobin A1C reflects the average blood glucose over a 3-month period preceding the blood draw. UKPDS data show the correlation of A1C values over a number of years with both microvascular and macrovascular complications, in this case, MI. As the A1C drops, the incidence of both micro- and macrovascular complications shows a progressive drop. Although the drop is higher at higher A1C levels, the drop continues and actually never becomes level.

Dr. Levy: This slide shows the results of 3 studies that looked at complications of diabetes. The studies showed better control and reduced microvascular complications, but no significant effect on macrovascular disease. Glucose is relevant, but there are other factors that contribute to macrovascular disease.

Dr. Levy: And here we see the risk of death related to A1C levels. Even in the normal range of 4% to 6%, there is an increased risk of death as the A1C levels rise within that range. People with postprandial hyperglycemia, who are not yet overtly diabetic, show increased cardiovascular risk.

Dr. Levy: The mental health provider must recognize the patient with diabetes who is at risk for vascular problems. Shown here are some of the signs and symptoms of cardiovascular disease in patients with type 2 diabetes. Be aware that many of these patients may be asymptomatic and that these patients might require an early cardiology referral. Their symptoms can be atypical, outcomes are worse in women, and aggressive risk factor reduction is needed.

Dr. Levy: For the mental health provider, these are the patients who should be screened in the psychiatric population. All patients with schizophrenia or bipolar disorder need to be screened whether they are symptomatic or not. A fingerstick glucose can be done easily in less than a minute if no other screening techniques are available. Patients should also be screened if they experience weight gain or unexplained weight loss, increased thirst or urination, erectile dysfunction, or if they are taking diabetogenic medications.

Dr. Levy: Here are the goals of therapy: a hemoglobin A1C of less than 7% by the ADA or less than or equal to 6.5% by the American College of Endocrinology. It is important to measure lipids and blood pressure because treatment can prevent cardiovascular complications and death. Aims of lipids and blood pressure control are aggressive. We would like the blood pressure to be at 130/80 mm Hg or less and the LDL cholesterol to be 100 mg/dL or less. All patients with type 2 diabetes should receive 75-325 mg of aspirin daily to reduce clotting, unless contraindicated. Also, microalbumin is measured annually and should be treated with an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker if elevated.

Dr. Beaser: So the mental health provider really does seem to be on the front lines of this issue, and needs to be aware of the potential for the metabolic syndrome, as well as how to identify it. With that in mind, John, could you review the various antipsychotic medications and, in particular, address their relative efficacy, selection, and titration, and the impact on the metabolic issues that we've been discussing?

Dr. Newcomer: The history of psychopharmacological approaches to the treatment of schizophrenia is briefly summarized on this slide, spanning the time from the introduction of chlorpromazine in 1954, through the introduction of clozapine in 1989 as the first atypical antipsychotic agent, on to the currently available first-line newer antipsychotics.

Dr. Newcomer: The psychopharmacology of treatments for bipolar disorder has undergone a similar evolution, beginning with the use of lithium salts and now for many patients involving, at least at some phase of treatment, the use of newer antipsychotic agents.

Dr. Newcomer: Newer antipsychotics have become first-line treatment for schizophrenia as well as other psychotic disorders. These agents are characterized by broad efficacy for so-called positive symptoms, including hallucinations, delusions, and paranoid thinking; evidence for superior efficacy for negative symptoms; as well as superior efficacy for improvement of mood and some of the cognitive impairments associated with schizophrenia, including improvements in memory, attention, and executive functions. The newer agents have been labeled atypical based on the reduced risk of extrapyramidal symptoms, including tardive dyskinesia, in comparison with conventional agents. There is evidence that newer agents can be superior to conventional drugs for the prevention of relapse, and the general improvement in tolerability for newer agents can lead to improved compliance. Given broad equivalents in psychiatric efficacy in large randomized pivotal trials, the major distinguishingdifferences between the newer antipsychotics involve differences in adverse event profiles.

Dr. Newcomer: Elevation of plasma prolactin is a characteristic of drugs that block dopamine type 2 receptors.This effect was commonly observed with conventional agents and is, in general, less common with most of the second-generation antipsychotics. The elevation of plasma prolactin may or may not be clinically relevant in a given patient, leading to the recommendation, "Treat the patient not the level."

Dr. Newcomer: While some of the newer antipsychotic medications, like older agents, have alpha-1 adrenergic activity and can, therefore, lower blood pressure, in general there are no sustained adverse effects on blood pressure observed across this class of newer medications. All newer antipsychotic agents can modestly prolong the corrected QT interval, with ziprasidone producing modestly more QTc increase in comparison with other agents. However, this has not led to demonstrated clinical significance to date based on almost 500,000 United States patient exposures to ziprasidone, with no dose-related QTc prolongation, no cardiac events in the existing overdose experience, no reports of torsades de pointes, and no increase in the rate of sudden or unexpected cardiovascular deaths.

Dr. Newcomer: At the time of ziprasidone's launch, there was considerable concern about the theoretical risks associated with QTc prolongation. It tended to overshadow an appreciation for active and clinically significant risks associated with obesity, metabolic disturbances, and cardiovascular risk. That perception is slowly giving way to medical reality.

Dr. Newcomer: Optimizing the psychiatric clinical response should involve the use of serial trials of monotherapy, with adequate doses and duration of trial prior to switching to the next agent. Switching between antipsychotics most prudently involves establishing adequate doses or plasma levels of the incoming agent in advance of the gradual taper of the previous drug. If at the conclusion of a series of trials of individual agents there is still an inadequate clinical response, some clinicians have pursued combination medication strategies, although at this point in time there is little evidence to support these approaches, and the risk of adverse events generally increases with the number of concomitant medications.

Dr. Newcomer: Treatment strategies concerning the medical risks under discussion should be focused on prevention. This will necessarily involve continual patient education; nutritional and exercise interventions; appropriate screening for risks; and the monitoring of weight, glucose, lipids, and blood pressure; and, finally, knowing when to refer.

Dr. Newcomer: In summary, the treatment of schizophrenia and bipolar disorders should involve careful consideration of both medical and psychiatric issues. Medications should be selected with reference to the side effect profile of individual agents. Treatment should be carefully optimized, using monotherapy wherever possible. Patients should undergo ongoing screening, monitoring, and education to prevent or minimize the medical risks that often occur during treatment of these severe conditions.

Dr. Beaser: So John, from a practical perspective, a patient comes to you with a family history of diabetes, and the patient is overweight and has dyslipidemia. How do these factors impact your selection of antipsychotic therapy?

Dr. Newcomer: I think the key concept is risk reduction. When we have preexisting, established risk factors, the goal is to certainly not make them worse, and ideally to make them better. So we're going to be thinking about selecting a medication that has a lower risk of contributing to weight gain and these other metabolic disturbances. It's really based on the idea that multiple risk factors are worse than single risk factors.

Dr. Beaser: So you look at the whole picture is what you're saying. Let's take another perspective. What if the patient has no preexisting risk factors, no dyslipidemia, no family history, and is not obese? Does that change the way you approach the situation?

Dr. Newcomer: Certainly, with no preexisting risk factors, the clinician has more latitude. On the other hand, if we have a teenager who isn't a smoker, we don't want them to start. The idea of even a single risk factor is still important, and it's really the idea of primary prevention that we've talked about.

Dr. Beaser: So even in the patient who doesn't have risk factors that are obvious to you, considerations need to be made in selecting medications.

Dr. Newcomer: I think you want to avoid starting up a problem that doesn't exist.

Dr. Beaser: Let's look at this from another perspective. If a patient is already on an antipsychotic medication and develops any of the signs of the metabolic syndrome, how would you deal with this well into the treatment course?

Dr. Newcomer: I can imagine 3 scenarios. One is where this metabolic problem starts in the setting of the use of a higher-risk medication. And here, in addition to whatever lifestyle and other interventions one might make, one of the possible interventions is to shift away from the higher-risk drug and switch to a drug with lower risk. Another scenario is that this metabolic problem emerges during treatment with a lower-risk medication. And in that setting, obviously you don't have a lot of choice in terms of switching to another medication. And the third scenario is that this is a patient who has established treatment resistance to front-line agents and they're already taking clozapine, the gold-standard drug for treatment-resistant patients. This is a drug with higher risk and, in that setting, you really don't have a lot of flexibility in terms of what drug you can use.

Dr. Beaser: What you're basically saying is in that scenario, you might then need to deal with the issues of the metabolic syndrome.

Dr. Beaser: In a patient for whom switching antipsychotic medications is not an immediate option and in whom we're faced with a possibility of diabetes, what would be your approach to this person in both evaluation and management?

Dr. Levy: We'll start with the diagnostic criteria for diabetes. This graph shows the glucose tolerance categories after an oral 75-g glucose load. If the fasting glucose is 126 mg/dL or greater, the diagnosis of diabetes is confirmed. A 2-hour value of 200 mg/dL or greater after a glucose load also confirms the diagnosis. The 2-hour value is more sensitive, but it's often easier to obtain a fasting glucose rather than doing a glucose tolerance test. A 2-hour value of 140-199 mg/dL is called impaired glucose tolerance (IGT), and these people are at risk for developing diabetes at an increased rate. They also have an increased risk of cardiovascular disease. Although a glucose level of 110 mg/dL is considered the upper limit of normal for a fasting blood glucose, any value of 100 mg/dL or more warrants careful patient follow-up. In fact, the lowering of the actual diagnostic level for impaired fasting glucose to 100 mg/dL is currently being considered.

Dr. Levy: The ADA classifies diabetes into types, with the 2 major types being type 1 and type 2. Remember that type 2 diabetes makes up about 90% to 95% of the diabetic population, so most likely these are the types of patients with diabetes that the mental health provider will be seeing.

Dr. Levy: There are several oral agents for treating type 2 diabetes, and here we see their various mechanisms and sites of action. The oral agents target the pathophysiologic defects.

Dr. Levy: Let's review the different classes of oral medications for type 2 diabetes. The insulin secretagogues consist of sulfonylureas, repaglinide, and nateglinide. These drugs work on the pancreas to stimulate insulin secretion.

Dr. Levy: The biguanide, metformin, has its primary action on the liver, decreasing hepatic glucose production. The thiazolidinediones, or glitazones, increase insulin sensitivity by promoting the uptake of glucose by skeletal muscle and adipose tissue.

Dr. Levy: And the alpha-glucosidase inhibitors slow glucose absorption from the gut.

We can also use insulin by injection. Insulin is deficient in type 2 diabetes, and exogenous insulin by injection can correct this insulin deficiency.

Thiazolidinediones, insulin, and some of the sulfonylureas may produce weight gain.

Dr. Levy: These are some important considerations in the treatment of type 2 diabetes in those with schizophrenia or bipolar disorders.

The underlying etiology of diabetes in most patients is probably no different in this population from any other population. Also, remember the comorbid factors, such as obesity, hypertension, dyslipidemia, and insulin resistance syndrome, need to be treated.

Dr. Levy: Lifestyle changes are critical, but also may be more difficult here than in the general population. As a general rule, clinicians must balance the ideal goals of therapy with the practical limitations in the patient with schizophrenia.

Dr. Levy: A critical medical issue is DKA, which can occur rapidly in the schizophrenic population and can be fatal. Some clinical points about this condition are listed on this slide.

Key points to remember include the following:

• There's a sudden and unpredictable onset.
• It can happen in people with type 2 diabetes or people without the diagnosis of diabetes at all.
• It can happen soon after starting on antipsychotic medication.
• Mental status changes can be a sign, which may be difficult to discern in a schizophrenic patient.
• Patients should go to the emergency room immediately if there is any question of DKA. Delays in therapy that may be common in psychiatric populations may result in increased mortality.

Dr. Levy: And there are other situations that warrant immediate care or consultation. Both very high and very low glucose levels should raise concern. Keep in mind that some patients with very low glucose levels may be asymptomatic, and that in this population symptoms may be subtle or the patient may not complain. Persistent nausea or vomiting, somnolence, or a change in mental status also are signs of a serious problem.

Dr. Levy: Prevention is the key to success. Here are some preventive strategies for the patient with mental illness who is at risk for diabetes:

• Screen for diabetes, especially before treatment, using either fasting plasma glucose, casual glucose in the setting of symptoms, or an oral glucose tolerance test. Although not technically diagnostic, an elevated fingerstick glucose can be highly suggestive.
• Minimize weight gain through appropriate medication selection and lifestyle changes, if possible.
• Check weight parameters regularly, including BMI and waist circumference.
• Measure lipid levels and blood pressure and institute early treatment.
• Beware of DKA.
• Clinicians should also screen frequently for diabetes in these patients. In this case, an ounce of prevention is indeed worth a pound of cure.
• It is extremely important for the psychiatrist to work closely with the medical care provider to develop a joint therapeutic plan.

Dr. Beaser: John and Phil, let's take another clinical scenario. A patient has recently been started on one of the antipsychotic medications with higher risk for developing weight gain and metabolic disturbances, and a family member calls you to report that the patient is more somnolent than usual with no immediate change in medication dose. What would you do?

Dr. Newcomer: In the setting of reliable information about an acute mental status change that cannot be explained by obvious factors, such as increased dose of a sedating psychotropic, an acute metabolic disturbance has to be ruled out.

Dr. Beaser: It sounds like you're putting the ball in Phil's court on this one. Phil, how would you approach it?

Dr. Levy: If possible, one could try to elicit some of the classic symptoms of DKA, including polyuria, polydipsia, nausea, vomiting, dehydration, sudden weight loss, etc. In addition to the mental status changes, if any family member has diabetes, a quick fingerstick glucose would be helpful. If there's any doubt, the patient should be evaluated immediately in a hospital emergency facility; early diagnosis and IV therapy can be lifesaving in this situation.

Dr. Beaser: I think the key take-home message is: if in doubt get them to the emergency room sooner rather than later.

Dr. Beaser: Let's let each of you give a sense for the key take-home points that we've talked about.

Dr. Newcomer: I think well-established treatment-induced increases in weight and adiposity are going to lead to predictable increases in insulin resistance and this, in turn, ups your chance of hyperglycemia, dyslipidemia, and cardiovascular disease. In addition, we've seen that antipsychotic medications can be associated with disturbances in glucose metabolism, sometimes independent of increases in weight or adiposity. All of this argues for primary prevention.

Dr. Beaser: And in terms of whether this is something related to drugs -- all drugs -- some drugs, what is the literature saying right now?

Dr. Newcomer: I think this is an at-risk population for many reasons. And on top of whatever those multiple factors are, you have groups of medications, some of them capable of causing substantial weight gain, and weight gain in an at-risk population is always a prescription for some difficulties.

Dr. Beaser: Phil, from your perspective, if someone develops these problems, how would you approach it? Is this something that is easy in terms of dealing with these metabolic disturbances?

Dr. Levy: It's not easy, as we've discussed already. Prevention is the best thing that one can do. If the disturbances occur, then you have to treat them. I think that the mental healthcare provider should be aware of diabetes, should be screening for diabetes, should screen all the patients with schizophrenia and bipolar disorders for diabetes, should also be testing them periodically for diabetes, and should be aware of the fact that they do have a higher incidence of the metabolic syndrome and a higher incidence of type 2 diabetes, and that they are a population at risk and should be looked at and treated as such.

Dr. Beaser: In terms of medication adjustment and changing the medications, are there any caveats that you would give?

Dr. Newcomer: In general, in most situations we're still not advising someone to suddenly pull the carpet out from underneath a patient. We're looking for opportunities to reduce risk. So when patients are hospitalized, when they become stable in an outpatient setting, these are opportunities to look at ways that we can improve the long-term medical outcome.

Dr. Beaser: What I think I hear you both saying is that if you take the psychiatric considerations and the metabolic concerns and synthesize them together, ultimately the right decisions can be made.

The metabolic syndrome and its increased risk of macrovascular damage can have a significant impact on the lives of people with schizophrenia or bipolar disorders. The mental health provider at the front lines of therapy is uniquely positioned to address issues such as lifestyle and antipsychotic medication selection in hope of reducing these risks. Appreciating the impact of antipsychotic treatments in the etiology of the metabolic syndrome allows appropriate consideration of the therapeutic risks vs benefits when designing antipsychotic treatments. Recent literature reports have provided some perspective, but much discussion on the role of antipsychotic medications in the etiology of diabetes in the metabolic syndrome is ongoing.

Hopefully we've provided you with some insight into how to interpret these reports and draw appropriate academic and clinical conclusions. The mental health practitioner is also in a position to identify the medically at-risk patient and trigger preventive or therapeutic interventions in a timely manner. Medical concerns must now become an integral part of a comprehensive care program for people with schizophrenia or bipolar disorders. Thus, we have also highlighted key clinical issues to consider when psychotic disorders and their treatments have an impact on medical and metabolic conditions.