There are many articles we have shared with you about diabetes and depression, eating disorders, stress, and anxiety. In this article we want to share with you how to read articles about mental illness and diabetes. This is a complex issue but important for those of us who deal or have dealt with those with mental illness. It is very important for primary care givers who are asked at times to prescribe these medications and for psychiatrists who are called in for consults. We now know that diabetes and hyperglycemia are fairly common in the general population (read our article on the diabetes epidemic). The estimated prevalence of diabetes from 1988-1994 in adults in the US was 5.1% with another 2.7% have undiagnosed diabetes. An additional 6.9% had borderline fasting glucose levels, bringing the total prevalence of adult Americans with elevated glucose to 14.7%, a staggering percentage. Numerous variables influence the etiology of diabetes, including obesity and having a sedentary life style (see this month's medical abstracts), both of which are common in the mentally ill. Indeed, mental illness itself may be a risk factor for the development of diabetes. We share excerpts from this article from Critical Breakthroughs in Psychiatry for you to get a feeling about where physicians are in this field. That is, do you treat the mental illness? If you do, will you make a risk factor cause type 2 diabetes? Is there a real causative association that has been proven?
Diabetes and the Mentally Ill.
The rates of diabetes and impaired glucose tolerance(IGT) are relatively high in patients with schizophrenia and bipolar disorder. The rate of Type 2 diabetes in schizophrenia has been reported at two to four times that of the general population. Similarly, rates in hospitalized bipolar patients have also been reported as two to three times the rate of the general population.
Reports of elevated rates in the mentally ill have spurred an ongoing examination of possible causative actors. Currently, a prolific amount of literature debates the involvement of antipsychotic medication in the development of diabetes. Indeed, there are numerous accounts in the literature of diabetes, hyperglycemia and hyperinsulinemia in patients taking antipsychotic medication. However, it is not yet clear whether antipsychotics are to blame, either directly or indirectly. In fact, Mukherjee et al reported a small sample of patients in Italy with schizophrenia who were not receiving psychotropics but who still had an elevated rate of diabetes. These patients actually had a higher rate of diabetes than patients with schizophrenia who had received medication.
Although no decisive connection has been shown between any psychotropic and diabetes, it is possible that some antipsychotic medications play a role in elevated rates of diabetes in the mentally ill. Chlorpromazine was the first psychotropic implicated in diabetes, and evidence for and against the involvement of various agents continues to mount as spates of studies move through the literature.
Much of the latest literature focuses on atypical agents in particular, and there are conflicting reports comparing the rates of nonnormal glucose levels in patients taking older antipsychotics vs. newer antipsychotics. However, the evaluation of historical data suggests a long existing association between diabetes and mental illness. Indeed, elevated rates of glycemic abnormalities were reported in psychiatric patients even before the introduction of antipsychotics or mood stabilizers. A study by Dixon et al estimated the prevalence of diabetes in patients with schizophrenia before the widespread use of atypical agents (1991-1996) and reported much higher rates among patients with schizophrenia than seen in the general population. These data support the argument that the higher prevalence of patients with schizophrenia predates the widespread use of atypical agents.
The use of antipsychotics and mood stabilizers is a necessity for the comprehensive treatment of mental illness. Conventional agents-typical antipsychotics-are associated with higher rates of side effects such as extrapryamidral symptoms and tardive dyskinsea. Atypical antipsychotics have proven to provide greater efficacy than the typical agents while offering a lower-risk side effect profile. Therefore, it is important to carefully evaluate the recent studies pertaining to the use of antipsychotics and the incidence of diabetes. Each study differs in design, definitions, purpose, and reliability. As such, several criteria must be kept in mind while considering these data. One criteria is a definition of diabetes. According to the American Diabetes Association (ADA), the diagnostic criteria for diabetes includes fasting plasma glucose (FGPG) >or=126 mg/dl; a random plasma glucose level of > or = 200 mg/dl; or a 2 hr plasma glucose of > or = 200 mg/dl glucose level during an oral glucose tolerance test. These designations are contingent upon confirmation on different days. IGT and impaired fasting glucose (IFG) refer to a metabolic stage between normal glucose homeostasis and diabetes. IGT is characterized by FPG > 100 mg/dl and < 126 mg/dl, and IFG by FPG > 110 mg/dl but < 140 mg/dl.
Other methodological concerns are more complicated but are also of use when evaluating clinical trials. A recent paper by Kraemer et al (2001) examines these concerns in detail and defines five different clinically important ways in which risk factors may work together to influence study outcomes. Although a thorough review of their thesis is too long for this paper, a brief synopsis may help to guide a reader of current literature. Several definitions are integral to these classifications, such as risk (the probability of an outcome), correlate (a measure associated in some way with the outcome), and causal risk factor (a correlate shown to precede an outcome).
The key message of this thesis is that analysis of risk must meet carefully considered and highly specific conditions before causative links can be established. In this sense, chains of causality must be built, much as biochemical interactions are modeled, through understanding one element at a time. The terminology used in risk analysis is often improvised and may lead to confusing interpretations; not all factors are risk factors or are causative. For example, diabetes could be linked to the severity of schizophrenia, and particular antipsychotics may be used more frequently in severe antipsychotics. It might appear, therefore, that these particular antipsychotics are linked to the development of diabetes, while in fact that they are merely associated with the causal risk factor of the severity of schizophrenia. Only prospective, randomized trials can address these potential complications.
In summary, there are a broad range of studies examining the influence of antipsychotics on glucose and insulin including epidemiological, and incidence and prevalence. Many of these studies are primary and limited to a small sample size or methological mistakes. To date, no pattern has emerged in the literature regarding diabetes associated with antipsychotic use. The numerous potential causative factors have not been parsed in a consistent or empirical manner to allow for clear interpretation. By holding to strict criteria for risk factors according to Dr. Kraemer et al over each study, it becomes clear that no causative relationships can be developed.
While diabetes' probable association with schizophrenia and its possible association with certain antipsyphotic agents should be considered, medication decisions should not be based on this factor according to Critical Breakthrough in Psychiatry. It is thought that clinicians should make decisions that are efficacious in all symptom domains, and that the treatment be safe and tolerable for the patient giving an optimal functioning outcome.
BSP
