Reviews/Commentaries/Position Statements REVIEW ARTICLE

Patients on Atypical Drugs Another high-risk group for type 2

1 MICHAEL E.J. , MA, MB, BCHIR, MD, FRCP is an expensive medical 2 FRANK-GERALD PAJONK, MD condition to health care systems and to society in general (5,6). The introduction of after 1956 transformed the lives of many sufferers. Several major chemical classes ABSTRACT — Patients with schizophrenia are more likely than the general population to of antipsychotic drugs were developed, develop diabetes, which contributes to a high risk of cardiovascular complications; individuals with schizophrenia are two to three times more likely to die from cardiovascular disease than the principally the (including general population. The risk of diabetes, and hence cardiovascular disease, is particularly in- chlorpromazine itself), the butyrophe- creased by some of the new atypical antipsychotic drugs. Individuals taking an atypical antipsy- nones (e.g., ), and the thiox- chotic drug, particularly younger patients under 40 years of age (odds ratio 1.63, 95% CI anthines (e.g., flupenthixol). All these 1.23–2.16), represent an underrecognized group at high risk of . The mechanisms “conventional neuroleptics” are effective responsible for antipsychotic-induced diabetes remain unclear. Hypotheses include these drugs’ because they are D antago- potential to cause weight gain, possibly through antagonism at the H , 5-HT , or 5-HT 2 1 2A nists (7), but they all have major draw- receptors. Other mechanisms independent of weight gain lead to elevation of serum leptin and . Patients with psychoses have difficulties with diet and lifestyle interventions backs and contribute their own stigmata for diabetes and weight management. If develops, withdrawal from antipsychotic to schizophrenia. These drugs share a set medication will often be inappropriate, and a change to an atypical antipsychotic drug with lower of Parkinsonian-like movement-disorder diabetogenic potential should be considered, especially in younger patients. Management of side effects, the “extrapyramidal side ef- psychoses should routinely include body weight and blood glucose monitoring and steps to fects” (EPSs), resulting from antagonism promote exercise and minimize weight gain. Careful collaboration between the psychiatric and at dopamine D2 receptors in the basal diabetology teams is essential to minimize the risk of diabetes in patients taking atypical anti- ganglia. psychotic medication and for effective management when it develops. This collaboration will also help minimize the already high risk of cardiovascular disease in individuals with Many conventional neuroleptics schizophrenia. cause excessive daytime sedation, and many are muscarinic antagonists, causing Diabetes Care 26:1597–1605, 2003 dry mouth, blurred vision, and constipa- tion, and may precipitate glaucoma in older patients. he number of individuals in the half of these individuals, the illness will be Newer drugs with lower EPS liability, population receiving antipsychotic lifelong, probably requiring long-term the “atypical ,” are increas- T drugs is surprisingly high. The most medication. Schizophrenia causes social ingly replacing the conventional neuro- common reason is schizophrenia, al- and also carries a high mortali- leptics. The reason why these drugs have though antipsychotic drugs are widely ty—approximately twice as high as in the better efficacy and side-effect profiles is used in other psychiatric conditions (e.g., general population (1). It has a strikingly not fully understood. One suggestion for and Alzheimer’s disease). high suicide rate of 10% (2,3), but the their lower EPS liability is lower occu- Schizophrenia is far more common than most common cause of death is acceler- pancy of dopamine D2 receptors in the is generally appreciated. For example, as ated heart disease (two to three times that , but the evidence for this is con- many as 1 in 100 individuals in the pop- in the general population) (4). Despite its flicting. However, it is widely accepted ulation will suffer one or more episodes of relatively low prevalence, the early age at that antipsychotic activity depends on an- schizophrenia in a lifetime, and for at least onset and its chronic nature means that tagonism at central D2 receptors and that the threshold for antipsychotic activity is ϳ ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● 65% occupancy of D2 receptors for both conventional neuroleptics and atyp- From the 1Department of Human Nutrition, University of Glasgow, Glasgow, U.K.; and the 2Department of Psychiatry and Psychotherapy, the Saarland University Hospital, Homburg, Germany. ical antipsychotics (8). Address correspondence and reprint requests to Professor M.E.J. Lean, Professor of Human Nutrition, As early as the mid-1960s, associa- University of Glasgow, Glasgow Royal Infirmary, Glasgow G31 2ER, U.K. E-mail: mej.lean@clinmed. tions between diabetes and conventional gla.ac.uk. neuroleptic treatment were reported (9– Received for publication 9 September 2002 and accepted in revised form 12 February 2003. M.E.J.L. has received honoraria for speaking engagements from Roche, Janssen-Cilag, Abbott, and Merck 18), but evidence has accumulated that and research funds from Sanofi-Synthelabo, Roche, Janssen-Cilag, GlaxoSmithKline, and Alizyme Thera- the risk is even higher for some of the peutics. F.-G.P. is employed by Janssen Cilag. atypical antipsychotics. This review sum- Abbreviations: EPS, extrapyramidal side effect; FDA, Food and Drug Administration. marizes the evidence for a causal link, dis- A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion factors for many substances. cusses some possible mechanisms, and © 2003 by the American Diabetes Association. concludes by reviewing some of the spe-

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1597 Diabetes and antipsychotic drugs

Table 1—Diabetogenic potential of atypical antipsychotic drugs during the first year of treatment with clo- zapine are 7.44 (95% CI 1.603–34.751) Reports Diabetes Diabetes compared with patients with of diabetes Reported resolved when improved when not receiving antipsychotics. In a 5-year Date (by January cases of drug stopped drug stopped naturalistic study of 82 patients being introduced 2002) ketoacidosis or switched or switched‡ treated with , 52% experienced one or more and 23% two or more epi- Clozapine 1989† 26 10 8‡ 10 (5) sodes of hyperglycemia; 30% were diag- 1994 3 1§ 3 — nosed as having type 2 diabetes (46). 1996 21 8 10 7 (3) Koller et al. (43) analyzed the U.S. FDA’s 1998 2 1 1 1 MedWatch surveillance program for the 2001 0 0 ——period of January 1990–February 2001 Data were compiled from surveys (29–32) plus other case reports found in Medline using the search terms and pooled these data with published “diabetes” and “antipsychotic or clozapine or olanzapine or quetiapine or risperidone or ziprasidone” pub- cases. They found 384 cases of clozapine- lished between January 1999 and January 2002 (33–42). Data in parentheses indicate cases where the dose associated hyperglycemia, with 242 cases of antipsychotic was reduced and/or diabetes was controlled with drugs and/or diet. †Reintroduced with safeguards to detect potential . ‡In one of these patients, diabetes resolved completely but of confirmed diabetes. Of these, 80 had recurred when clozapine was restarted and did not resolve when clozapine was subsequently.stopped ketosis or acidosis, and 25 patients died §Patient was HIV positive and taking a protease inhibitor concomitantly with risperidone; protease inhibitors during hyperglycemic episodes. On the are known to cause diabetes. other hand, Wang et al. (47) carried out a case-control study in patients taking “psy- cial problems associated with manage- cally relevant hyperglycemia or diabetes chiatric drugs” in the government- ment in this patient population. according to internationally recognized sponsored drug benefit program in New criteria. It is worth noting that in psychi- Jersey (7,227 case subjects with newly IS THERE A LINK BETWEEN atric circles, it is now generally recognized treated diabetes and 6,780 control sub- SCHIZOPHRENIA AND that diabetes is a risk in patients treated jects). They reported that 1.3% of individ- DIABETES? — Even before antipsy- with antipsychotics, particularly cloza- uals who developed diabetes took chotic drugs appeared, there was evi- pine. This means that the number of clozapine versus 1.7% of control subjects. dence that diabetes was more common in published case reports probably under- Considering that the minimum age for in- patients with schizophrenia (17,19–26). represents the true prevalence of antipsy- clusion was Ͼ20 years, the mean age of A more recent study of patients with chotic-associated diabetes. After our individuals in their study was 63.6 years schizophrenia in the U.S. Veterans Ad- survey was completed, other publications for diabetes case subjects and 61.9 years ministration health care system found on the subject have appeared, most nota- for control subjects, marking the popula- that the rate of diabetes was 6.2–8.7% bly those by Koller and her colleagues, tion as highly unusual. Nevertheless, they (27), compared with 1.1% in U.S. men which include results from the U.S. Food found no significant association between aged 20–39 years without schizophrenia and Drug Administration’s (FDA) Med- clozapine use and developing diabetes, (28). These figures probably overestimate Watch Drug Surveillance System (43,44). whereas use of “nonclozapine antipsy- the incidence because most patients had We have not included their data into our chotic medication” was significantly asso- already been exposed to antipsychotic analysis because the new analyses cover ciated with diabetes. In addition, less than medication. Nevertheless, the rate is strik- different periods from our analysis, use half of patients had “psychotic disorders” ingly high. different criteria for inclusion, and give (rather than “affective disorders,”“ insufficient information for us to deter- disorders,” or “other psychiatric disor- DIABETES AND ATYPICAL mine which cases we have already incor- ders”), and the analysis did not look for ANTIPSYCHOTIC porated and which are new. We therefore any association between diagnosis, anti- MEDICATION — We carried out a comment on these new analyses separately. psychotic use, and diabetes. This study is search in January 2002 in Medline using interesting, but more work is needed to the terms “diabetes” and “antipsychotic” Clozapine exclude the many potential confounding and identified four recent systematic re- Clozapine (available on a limited basis factors. views (29–32). We then compiled a table since the 1980s) was associated with 26 (Table 1) summarizing the cases con- published case reports of diabetes. Ap- Risperidone tained in these reports, to which we proximately one-third of cases were re- We found three published cases of new- added the results of a further search using ported as diabetic ketoacidosis, although onset diabetes associated with risperi- the terms “diabetes” and “antipsychotic or not all reports give detailed values for pH, done since 1993 (35,41). Mahmoud et al. clozapine or olanzapine or risperidone or blood glucose, blood gases, bicarbonate, (45) estimated that the odds of type 2 di- quetiapine or ziprasidone” for publica- and ketones and may therefore not strictly abetes in a patient treated with risperi- tions between January 1999 and January fulfil the criteria for full-blown ketoacido- done are 0.88 (95% CI 0.372–2.070) 2002 that had not been cited by the four sis. It is also worth noting that patients compared with patients not receiving an- reviews listed above (33– 42). We in- taking clozapine are very prone to gain tipsychotic drugs during the first year of cluded all reported cases in which suffi- weight (see WEIGHT GAIN below). treatment (i.e., not significantly different cient detail was given for us to determine Mahmoud et al. (45) calculated that from the risk in an untreated comparable that patients had developed either clini- the odds of developing type 2 diabetes population). Switching to risperidone

1598 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Lean and Pajonk may also restore near-normal glycemic Table 2—Logistic regression analysis of the association between atypical and conventional control in patients who develop hypergly- neuroleptics and diabetes in patients of all ages with schizophrenia with reference to diabetes cemia on other atypical antipsychotic in a nonschizophrenic population drugs (48). Risperidone tends to cause weight gain, but less so than either cloza- n Odds ratio Rank order 95% CI P pine or olanzapine (see WEIGHT GAIN be- low). Any atypical 22,648 1.09 — 1.03–1.15 0.002 Clozapine 1,207 1.25 2 1.07–1.46 Ͻ0.005 Olanzapine Risperidone 9,903 1.05 4 0.98–1.12 0.15 We found 21 case reports between 1996 Olanzapine 10,970 1.11 3 1.04–1.18 Ͻ0.002 and 2002 of new-onset diabetes in pa- Quetiapine 955 1.31 1 1.11–1.55 Ͻ0.002 tients taking olanzapine, with 8 described Data are from Sernyak et al. (27). Low rank order equates high risk. as presenting with diabetic ketoacidosis. Mahmoud et al. (45) estimated that the prescriptions for risperidone (31.4%), done were 2.13 (1.36–3.35; P Ͻ 0.002), odds of type 2 diabetes in the first year of ϭ treatment with olanzapine are 3.10 (95% 8,788,000 for olanzapine (26.2%), 1.64 (1.23–2.21; P 0.0009), 1.51 4,184,000 for quetiapine (12.5%), (1.12–2.04; P Ͻ 0.008), and 1.82 (1.05– CI 1.620–5.934) compared with the risk Ͻ in patients with psychosis not receiving 2,222,000 for clozapine (6.6%), and 3.15; P 0.04), respectively. antipsychotics. Olanzapine is similar to 491,000 for ziprasidone (1.5%) (49). A Retrospective analyses are compli- clozapine in its high weight gain potential study of the U.S. Veterans Administration cated by the possibility that duration of (see WEIGHT GAIN below). In 2002, the health care system analyzed the records of treatment will affect the risk of developing Committee on the Safety of Medicines re- individuals diagnosed as having schizo- diabetes and that diabetes may persist af- ported a number of diabetic case subjects phrenia over a 4-month period (27). Of ter a switch to an antipsychotic drug with presenting with ketoacidosis or coma the 38,632 patients, 1,207 were pre- low diabetes-inducing liability, thus as- from “yellow card” reporting in the U.K. scribed clozapine, 10,970 olanzapine, signing an inappropriately high risk to the Most, but not all, also experienced weight 9,903 risperidone, and 955 quetiapine. “benign” antipsychotic. To deal with gain. In their search of the literature up to Overall, there was a significant associa- these problems, Chue and Welch (50) May 2001 and from the U.S. FDA Med- tion between the development of diabetes controlled for duration of , the Watch surveillance program for the pe- and prescription of quetiapine, clozapine, reason for any switches to current therapy riod of January 1994–May 2001, Koller and olanzapine, but not risperidone. (e.g., weight gain), age, and concomitant and Doraiswamy (44) found 237 cases of Odds ratios (and 95% CIs) for the devel- medication. They looked at the effect of olanzapine-associated hyperglycemia. Of opment of diabetes for quetiapine, cloza- clozapine, olanzapine, risperidone, and these cases, 188 were new-onset diabetes, pine, olanzapine, and risperidone were conventional neuroleptic on a variety of 1.31 (1.11–1.55; P Ͻ 0.002), 1.25 (1.07– metabolic risk factors. Although they did 80 involved metabolic ketosis or acidosis, Ͻ Ͻ and 15 patients died. 1.46; P 0.005), 1.11 (1.04–1.18; P not give actual prevalence values, they 0.002), and 1.05 (0.98–1.12; P ϭ 0.15), concluded that the prevalence of diabetes Quetiapine respectively (Table 2). The strongest ef- was highest with clozapine, lower for Ͻ We found three cases of quetiapine- fect was in patients aged 40 years. In olanzapine (which had a higher preva- associated diabetes reported since 1998, this group, the odds ratios (and 95% CIs) lence than conventional antipsychotics), one of these presenting as diabetic keto- for the development of diabetes for cloza- and lowest with risperidone. Clozapine acidosis (Table 1). Quetiapine is interme- pine, quetiapine, olanzapine, and risperi- and olanzapine were associated with the diate between risperidone and olanzapine in weight gain potential (see WEIGHT GAIN Table 3—Rank order of risk of antipsychotic drugs for diabetes-related factors (in order of below). decreasing value), adjusted for diagnosis, duration of antipsychotic treatment, other medica- tions, family history of diabetes, ethnicity, and smoking habits Ziprasidone Ziprasidone was introduced in 2001, and there have not been any reports of diabe- Conventional tes, but its introduction is probably too Clozapine Olanzapine Risperidone neuroleptics recent for any drug-associated cases to Prevalence of diabetes 1 2 4 3 have appeared in the literature. Ziprasi- Hyperglycemia (fasting) 1 2 3 4 done has a relatively low weight gain po- Hyperinsulinemia (fasting) 2 1 3 4 tential (see WEIGHT GAIN below). Elevated total cholesterol 1 3 4 2 Elevated triglycerides 1 2 4 3 Comparison studies Elevated BMI 2 1 3 4 These published reports of diabetes must Elevated plasma uric acid 1 3 2 4 be considered in relation to the relative Sum of ranks* 9 14 23 24 numbers of patients treated with each of Data are from Chue and Welch (50). *The parameters assessed are not equivalent in their contribution to the the agents. In the U.S., it has been esti- pathology of diabetes or its cardiovascular complications. However, no attempt has been made to weight the mated that in 2001, there were 10,547,000 sums of rank orders. Low rank order or rank sum equates high prevalence or risk.

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1599 Diabetes and antipsychotic drugs

Table 4—Weight gain liabilities for atypical antipsychotic drugs

Clozapine Olanzapine Risperidone Quetiapine Ziprasidone Haloperidol Wirshing et al. (59)* 6.9 Ϯ 0.8† 6.8 Ϯ 1.0** 5.0 Ϯ 0.6 ——3.7 Ϯ 0.6 Meyer (53)‡ 5.3–6.3 6.8–11.8 2.0–2.3 2.77–5.6 0.23 — Czobor et al. (52)§ 4.2 Ϯ 4.7¶ 5.4 Ϯ 4.6*† 2.3 Ϯ 2.8# ——0.2 Ϯ 0.2 Data are in kilograms.*Maximal weight gain Ϯ SE. Maximal weight gains were adjusted by controlling for age, treatment duration, and initial body weight. These weight gains occurred over treatment periods of 24–73 months, and the corrected values are adjusted for duration of treatment. The expected weight gain for the normal population not receiving antipsychotic drugs over this period would be ϳ2–3 kg. For clozapine, weight gain ceased at 25 months, compared with 21 months for olanzapine, 15 months for risperidone, and 18 months for haloperidol. Patients in this study were counselled about diet and exercise and were referred to a clinical nutritionist if their weights increased by Ͼ4.5 kg. †P Յ 0.01 vs. haloperidol (pairwise comparison, controlled for age, treatment duration, and initial weight). ‡Range of weight gain over 1 year (6 months for ziprasidone). §Mean weight gain Ϯ SD after 14 weeks’ treatment; P Ͻ 0.05 vs. baseline; ¶P Ͻ 0.05 vs. haloperidol; #no significant difference from haloperidol. greatest increase in insulin resistance, (52,53,57,59). However, patients taking sible that the apparent correlation be- BMI, and lipids. By assigning a rank order antipsychotic drugs can develop diabetes tween weight gain potential and diabeto- to each of the agents and summing these without significant weight gain (60) or genicity results from a common pharma- ranks for all the risk factors examined (the can lose weight (61). Furthermore, their cological action rather than diabetes being lower the rank sum, the higher the risk), it diabetes usually improves rapidly when an indirect effect caused by weight gain. can be concluded that the overall risk was the antipsychotic drug is withdrawn, highest for clozapine and slightly less for without significant reduction in body olanzapine. The overall risk was almost weight, and often recurs rapidly if the Receptor antagonism the same for risperidone and conven- drug is started again. The antipsychotic activity of both atypical tional neuroleptics (Table 3). Cagliero et al. (62) used a frequently and conventional antipsychotics is medi- Koro et al. (51) used a population- sampled intravenous tolerance test to in- ated by antagonism at central dopamine based nested case-control design to exam- vestigate the acute effect of clozapine, D2 receptors (7,8). Consequently, if dia- Ͼ ine the data held on 3.5 million patients olanzapine, and risperidone on insulin re- betes were related only to antagonist po- in England and Wales over a 13-year pe- sistance in a small group of nonobese pa- tency at D2 receptors, all antipsychotics riod. Olanzapine significantly increased tients (BMI Ͻ30 kg/m2 at baseline) with would be expected to have similar diabe- the risk of diabetes compared with no an- schizophrenia. Clozapine produced sig- tes-inducing potential. This is clearly not tipsychotic (adjusted odds ratio 5.8 [95% nificantly higher 20-min glucose levels the case. The dosages of antipsychotic drugs CI 2.0–16.7]). The risk associated with than risperidone, and individuals taking differ greatly, but the effective antipsy- risperidone was less (2.2 [0.9–5.2]). They risperidone had a significantly higher in- chotic concentration in plasma or cere- also found a small increase in risk associ- sulin sensitivity than those taking olanza- brospinal fluid is closely correlated with ated with conventional neuroleptics (1.4 pine or clozapine. Newcomer et al. (63) [1.1–1.7]). The risk associated with olan- their antagonist potency at D receptors performed modified oral glucose toler- 2 zapine was also significantly increased (8). To compare the receptor profiles of ance tests in patients with schizophrenia compared with conventional neuroleptics the different antipsychotic drugs, it is and healthy control subjects matched for (4.2 [1.5–12.2]), whereas the risk with therefore necessary to examine their po- BMI and age. Compared with patients risperidone was not significantly greater tencies at different receptors relative to taking conventional neuroleptics or un- than that with conventional neuroleptics their potencies at the receptors through (1.6 [0.7–3.8]). treated normal control subjects, olanzap- which their antipsychotic effects are pri- ine and clozapine produced significant marily mediated—namely D2 receptors elevations in plasma glucose. Risperi- (8). No clear patterns emerge (Table 5). MECHANISMS FOR done, however, only produced significant Antagonism at 5-HT receptors. High ANTIPSYCHOTIC- elevations when compared with un- antagonist potency at 5-hydroxytrypta- treated control subjects. ASSOCIATED DIABETES mine 5-HT2A receptors combined with Intriguingly, a number of the re- slightly lower potency at D2 receptors Weight gain ported cases of new-onset diabetes asso- may be one prerequisite for the low EPS Weight gain is common with conven- ciated with atypical antipsychotic use are liability and extra efficacy of the atypical tional neuroleptics and atypical antipsy- described as presenting with diabetic ke- antipsychotics (Table 5) (64). It is un- chotics (52,53), and excessive body toacidosis, although most did not ulti- likely to be the reason for antipsychotic- weight is a clearly established risk factor mately need insulin. Few of the reports induced diabetes, however, because for type 2 diabetes (54–56). It is tempting give detailed clinical and laboratory infor- risperidone has a 5-HT2A/D2 potency ra- to think that antipsychotic-induced dia- mation, but some of the patients were tio similar to that of clozapine and olan- betes is a consequence of weight gain. For clearly very ill. zapine, but has lower propensity to cause example, clozapine and olanzapine have These observations suggest a direct diabetes. the highest propensity to cause both metabolic effect rather than an indirect 5-HT2C receptors are probably in- weight gain and diabetes (Table 4) effect secondary to weight gain. It is pos- volved in the regulation of food intake

1600 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Lean and Pajonk

Table 5—Potency of antipsychotic drugs at different receptors

Risperidone (1–5 mg/day) Haloperidol (5–10 mg/day) Olanzapine (5–15 mg/day) Relative potency vs. Relative potency vs. Relative potency vs.

Ki (mol/l) D2 receptors Ki (mol/l) D2 receptors Ki (mol/l) D2 receptors ϫ Ϫ9 ϫ Ϫ9 ϫ Ϫ9 D2 4.07 10 1 2.04 10 1 63.10 10 1 ϫ Ϫ9 ϫ Ϫ9 ϫ Ϫ9 H1 33.11 10 0.123 1,202.26 10 0.0017 1.20 10 52.5 ϫ Ϫ9 ϫ Ϫ9 ϫ Ϫ9 5-HT1A 426.58 10 0.0096 1,621.81 10 0.0013 3,235.94 10 0.020 ϫ Ϫ9 ϫ Ϫ9 ϫ Ϫ9 5-HT2A 0.41 10 10.0 302.00 10 0.0068 2.34 10 26.9 ϫ Ϫ9 ϫ Ϫ9 5-HT2C 75.86 10 0.054 * * 18.20 10 3.47 ϫ Ϫ9 ϫ Ϫ9 AChm * * 3,467.37 10 0.00059 54.95 10 1.15 ␣ ϫ Ϫ9 ϫ Ϫ9 ϫ Ϫ9 1-Adrenoceptors 2.45 10 1.66 26.30 10 0.078 57.54 10 1.10 ␣ ϫ Ϫ9 ϫ Ϫ9 ϫ Ϫ9 2A-Adrenoceptors 21.88 10 0.19 1,047.13 10 0.0020 426.58 10 0.15

Absolute values are given as Ki values (mol/l): the higher the Ki, the lower the potency. However, to take into account the different dosages, it is more useful to compare the potencies of these drugs in terms of their potency at the receptors through which their antipsychotic activity is mediated—namely D2 receptors. Values Ͻ Ͼ of relative potency 1.0 therefore mean that the drug has lower potency at that receptor type than at D2 receptors, and, conversely, values 1.0 mean that the drug is more potent at that receptor than at D2 receptors. Typical daily doses are also included. Data are from Leysen et al. (64). *Potency at this receptor was too low to measure.

(65). Relative antagonist potency at weight gain nor new-onset diabetes have Leptin 5-HT2C receptors roughly matches been reported in psychiatric patients pre- Leptin levels are elevated in patients with weight gain potential, except perhaps for scribed muscarinic antagonists for exces- antipsychotic-induced new-onset diabe- ziprasidone and quetiapine (Table 5). An- sive EPS. tes (75,76) and in many patients taking tagonism at 5-HT2C receptors is therefore clozapine or olanzapine who have not an appealing explanation for clozapine- been diagnosed with diabetes (72,77). Acute or olanzapine-induced weight gain and Leptin is released from adipocytes and is Several cases of new-onset diabetes attrib- diabetes. However, two 5-HT –selective believed to reduce appetite and stimulate 2C uted to clozapine and olanzapine were as- agonists, (ϩ/Ϫ)-2,5-dimethoxy-4-iodo- catabolism of fat and/or inhibit fat syn- sociated with acute pancreatitis (71). It is and ␣-methyl-5-HT, both thesis in adipocytes, although serum lev- possible, therefore, that antipsychotic- cause hyperglycemia rather than hypo- els are elevated in obese humans, induced diabetes results from chemical glycemia in rats (66). If antagonism at indicating leptin resistance (78). How- damage to the pancreas. However, diabe- 5-HT receptors is involved in antipsy- ever, the rapidity and the disproportion- 2C tes associated with atypical antipsychotics chotic-induced diabetes, it is probably ate magnitude of the increase in leptin is associated with hyperinsulinemia not the only mechanism. levels when clozapine is started (75,76) rather than failure of insulin release The role of 5-HT receptors on glu- suggest that it may be a direct effect and 1A (46,72,73). Hyperlipidemia has been re- cose homeostasis is complex (67–70), but not a response to antipsychotic-induced ported in several studies (46,73) and ele- 5-HT blockade is unlikely to be respon- weight gain. Raised leptin and subsequent 1A vated serum triglycerides is an almost sible for new-onset diabetes because the downregulation of hypothalamic leptin universal finding in type 2 diabetes (46), relative potencies of atypical antipsychot- receptors or altered transport dynamics consistent with the view that diabetes is a ics at 5-HT receptors do not match their could explain the weight gain and diabe- 1A complex metabolic disturbance involving diabetogenic potential (Table 5). tes in patients taking certain antipsychot- carbohydrates, fats, and amino acids. Antagonism at H receptors. ics (72). Against this are the results of a 1 Pancreatitis associated with antipsychotic Antagonism at central histamine H re- study comparing leptin levels in 59 pa- 1 use could therefore be an indirect effect ceptors has been suggested as the reason tients with chronic schizophrenia with caused by hyperlipidemia. for antipsychotic-induced weight gain those from a group of healthy subjects (59). However, quetiapine has relatively matched for sex, age, and BMI (79). There low weight gain potential (Table 4), yet is Insulin resistance was no difference between leptin levels in 87 times more potent at H1 receptors than Dwyer et al. (74) have shown that some patients taking chronic antipsychotic at D2 receptors (Table 5). antipsychotics inhibit glucose transport medication (37 conventional and 17 Antagonism at muscarinic acetylcho- into PC12 cells in culture and increase atypical) and matched control subjects. line receptors. Although both clozapine cellular levels of the glucose transporter The relevance of this is uncertain because and olanzapine produce significant isoforms GLUT1 and GLUT3. This sce- of the small numbers of patients taking blockade of muscarinic ACh receptors at nario would lead to hyperglycemia, atypical antipsychotics. A definitive study antipsychotic doses, muscarinic ACh re- which in turn would cause a homeostatic of the putative correlation between anti- ceptor blockade can be discounted as the increase in insulin release. Prolonged hy- psychotic intake and leptin would require cause of either weight gain or diabetes. perinsulinemia could eventually lead to an antipsychotic-naive population, a con- Risperidone has no measurable affinity insulin resistance and further hyperglyce- trol group given placebo, and several test for muscarinic ACh receptors (Table 5) mia as a result of downregulation of insu- groups each given different antipsychot- but causes some weight gain, and neither lin receptors. ics with different diabetes-inducing

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1601 Diabetes and antipsychotic drugs

Table 5—Continued

Clozapine (200–450 mg/day) Quetiapine (300–450 mg/day) Ziprasidone (20–160 mg/day) Relative potency vs. Relative potency vs. Relative potency vs.

Ki (mol/l) D2 receptors Ki (mol/l) D2 receptors Ki (mol/l) D2 receptors 177.82 ϫ 10Ϫ9 1 724.44 ϫ 10Ϫ9 1 6.76 ϫ 10Ϫ9 1 1.07 ϫ 10Ϫ9 166.0 8.32 ϫ 10Ϫ9 87.10 64.56 ϫ 10Ϫ9 0.11 190.55 ϫ 10Ϫ9 0.93 416.87 ϫ 10Ϫ9 1.74 5.50 ϫ 10Ϫ9 1.23 6.31 ϫ 10Ϫ9 28.2 275.42 ϫ 10Ϫ9 2.63 2.09 ϫ 10Ϫ9 3.24 12.59 ϫ 10Ϫ9 14.1 2,454.71 ϫ 10Ϫ9 1.10 6.46 ϫ 10Ϫ9 1.05 33.11 ϫ 10Ϫ9 5.37 1,071.52 ϫ 10Ϫ9 7.94 2,454.71 ϫ 10Ϫ9 0.0028 22.39 ϫ 10Ϫ9 7.94 52.48 ϫ 10Ϫ9 13.8 13.18 ϫ 10Ϫ9 0.51 53.70 ϫ 10Ϫ9 3.31 1,778.28 ϫ 10Ϫ9 0.41 186.21 ϫ 10Ϫ9 0.036 potentials. Nevertheless, although imper- returned to normal is not always clear in of the literature up to 2001, Cohen (61) fect, the existing evidence does suggest these reports. In some cases, it was re- found 22 cases of new-onset diabetes that that elevated leptin may play a part in the markably quickly—within 2–3 days of resolved and 6 that did not when the an- etiology of antipsychotic-induced diabe- stopping or switching—although some- tipsychotic was stopped. In a survey of tes (Fig. 1). times oral hypoglycemic agents or insulin diabetes associated with clozapine (43), were used. In nearly all the reports, blood glycemic control improved after cloza- REVERSIBILITY OF glucose levels were normal when mea- pine was stopped in 78% of individuals ANTIPSYCHOTIC-INDUCED sured 2–3 weeks after stopping the anti- who developed diabetes; 62% of these pa- DIABETES — In most reported cases psychotic drug. In a few cases, hyper- tients no longer required hypoglycemic of hyperglycemia or diabetes associated glycemia persisted after stopping or drugs. Of 12 patients who were restarted with antipsychotics, the antipsychotic switching but was usually less marked on clozapine, 9 developed hyperglycemia (usually clozapine or olanzapine) was ei- than before, or the blood glucose concen- again. For diabetes associated with olan- ther stopped completely or substituted tration became manageable with an oral zapine, Koller and Doraiswamy (44) re- with another antipsychotic. The speed hypoglycemic agent, when insulin was ported that 78% of patients had improved with which blood glucose concentrations previously needed (Table 1). In a survey glycemic control once olanzapine was

Figure 1—Possible mechanisms for antipsychotic-induced new-onset dia- betes. Possible sites of action for anti- psychotics in causing diabetes include, among other possibilities, direct ␤-cell damage, appetite stimulation, or stim- ulation of leptin release.

1602 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Lean and Pajonk stopped or the dosage decreased; hyper- before the drug was started. Withdrawal itor their own blood glucose concentra- glycemia recurred in 8 of 10 patients of clozapine must therefore be carried out tions, calculate insulin doses, manage when olanzapine was restarted. slowly over a period of several weeks or their own food intake, or self-inject. Com- months while replacement antipsychotic pliance with prescribed oral hypoglyce- MANAGEMENT OF is slowly introduced. As patients with clo- mic drugs is also likely to be poor. DIABETES IN PATIENTS zapine are usually severely ill and have Unfortunately patients with schizo- WITH PSYCHOSIS usually failed to respond to other agents, phrenia often find it difficult to attend out- it may be necessary to persist with cloza- patient clinics regularly and frequently fail Relapse prevention and switching pine and manage the diabetes. to adhere to treatments. The medical out- antipsychotic drugs look for a schizophrenic patient with dia- Diabetes is a serious medical development Drug interactions: antipsychotic and betes is therefore particularly bad and is that requires immediate intervention and oral hypoglycemic medications reflected in their greatly increased rates of possibly lifelong management, often with Clozapine is metabolized mainly by coronary heart disease (63). Management increasing antidiabetic medication. How- CYP1A2 and CYP3A4, olanzapine mainly of diabetes therefore presents special ever, schizophrenia is also a serious ill- by CYP1A2 and CYP2D6, quetiapine and problems requiring close supervision to ness, the management of which usually ziprasidone almost exclusively by CYP3A4, avoid acute problems, such as hyper- or requires continuation of antipsychotic drugs. and risperidone by CYP2D6 (82,83). All hypoglycemia and ketoacidosis. The effective management of both condi- are moderately protein bound, but this Although their primary use is in tions demands a careful and committed does not pose a significant interaction schizophrenia, the atypical antipsychotics collaboration between the two medical risk. Although all the sulfonylureas bind are used in a variety of other illnesses: teams—psychiatry and diabetology. strongly to plasma protein and can dis- behavioral and psychological symptoms The course of schizophrenia is usu- place weak acids, such as aspirin, they do of (e.g., Alzheimer’s disease ally multiple acute episodes of frank psy- not displace the atypical antipsychotics and Lewy body disease), bipolar disorder, chosis and disability interspersed with from their binding sites (82,83). The sul- and a variety of psychiatric disorders with periods of milder symptoms. Some pa- fonylureas tolbutamide, glipizide, and psychotic features. Patients with demen- tients may be essentially normal between glibenclamide are metabolized by CYP2C9, tias are older and are therefore at much acute episodes. Acute episodes tend to be- so that there is no reason to expect hepatic higher risk of developing diabetes than come more severe over time and the in- interference (84). None of the oral hypo- young patients with schizophrenia. Atyp- terval between episodes progressively glycemic agents have been reported to in- ical antipsychotics with low diabetes- shorter (80). The prognosis is definitely teract with any of the atypical antipsychotics inducing liability should therefore be better in patients who continue to take (82,83). Metformin is excreted largely un- particularly preferred in this context. antipsychotics between acute episodes, changed and is therefore unlikely to cause even when they are symptom free. Taken pharmacokinetic interaction with any of CONCLUSIONS — Patients on atyp- in this way, antipsychotics reduce both the atypical antipsychotics. ical antipsychotic medication for schizo- the frequency and intensity of relapses phrenia or other illnesses should be and therefore protect against the deterio- Managing diabetes in patients with considered a high-risk group for diabetes ration associated with repeated acute ep- schizophrenia or taking atypical and vascular disease. Use of atypical anti- isodes (80,81). antipsychotics psychotics is associated with a generally Therefore, although stopping an anti- A high-fat diet combined with physical high risk of type 2 diabetes, but the risk is psychotic drug might resolve the diabetes inactivity contributes to weight gain and lower with some of these drugs than with it has triggered, effective antipsychotic predisposes susceptible individuals to others. The mechanisms include the therapy, preferably with a less diabeto- type 2 diabetes. Lifestyle management is drug-induced weight gain that is common genic drug, must be continued to prevent therefore also central to long-term care. with antipsychotics, but there is also evi- psychotic relapse and long-term deterio- For patients with type 2 diabetes, the ma- dence for a direct metabolic effect. This ration. Some conventional neuroleptics jor pathological hazard is accelerated cor- may be related to antagonism at the have low potential to cause diabetes, but onary heart disease and . The 5-HT2C or histamine H1 receptors or to replacing an atypical with a conventional frequent smoking habit of patients with elevation of serum leptin beyond that in- neuroleptic might reduce compliance and schizophrenia greatly aggravates this duced by increased body weight alone. usually results in motor side effects and problem (85). It is therefore important to Stopping the antipsychotic com- increasing severity of negative symptoms, monitor coronary risk factors, such as hy- monly allows the diabetes to resolve. such as social withdrawal, poverty of pertension and dyslipidemia regularly. Given the compounding effects of weight thought, and lack of initiative. Among the Managing diabetes in patients with gain and diabetes on coronary heart dis- atypical antipsychotics, risperidone ap- schizophrenia is complicated by their ease (the major cause of premature death pears to have the least propensity to cause lack of insight, loss of initiative, and cog- in schizophrenia), aggravated by smoking diabetes, and time will tell whether this nitive deficits that are central features of and inactivity (frequent features of schizo- also applies to quetiapine and ziprasidone. the illness. Even in the supervised envi- phrenia), antipsychotics with low poten- Withdrawal from clozapine is partic- ronment of psychiatric units, it can be dif- tial for weight gain and diabetes should be ularly difficult because a so-called re- ficult to ensure that patients follow preferred, provided their efficacy in bound effect may develop, in which the dietary advice. Patients with active psy- schizophrenia is adequate. Among the patient’s condition becomes worse than chosis are also unlikely to be able to mon- atypical antipsychotics, risperidone has

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1603 Diabetes and antipsychotic drugs been shown to reduce the long-term risk diabetes in nonpsychotic medical pa- 31. Muensch J, Carey M: Diabetes mellitus as- of relapse compared with the conven- tients. Dis Nerv Syst 30:341–344, 1969 sociated with atypical antipsychotic med- tional neuroleptic haloperidol (81). 14. Hiles B: Hyperglycaemia and glycosuria ications: new case report and review of the Diabetologists and psychiatrists need following chlorpromazine therapy (Let- literature. J Am Board Fam Pract 14:278– to work together to monitor patients pre- ter). J Am Med Assoc 162:1651, 1956 282, 2001 15. McKee HA, D’Arcy PF, Wilson PJ: Diabe- 32. Wirshing DA, Spellberg BJ, Erhart SM, scribed atypical antipsychotics to detect tes and schizophrenia: a preliminary study. Marder SR, Wirshing WC: Novel antipsy- impaired glucose tolerance and manage J Clin Hosp Pharm 11:297–299, 1986 chotics and new onset diabetes. Biol Psy- diabetes. This will help reduce the high 16. Newcomer JW, Craft S, Fucetola R, chiatry 44:778–783, 1998 risk of cardiovascular disease in patients Moldin SO, Selke G, Para L, Miller R: Glu- 33. Bonanno DG, Davydov L, Botts SR: Olan- with schizophrenia. Particular attention cose-induced increase in memory perfor- zapine-induced diabetes mellitus. Ann should be paid to patients taking cloza- mance in patients with schizophrenia. Pharmacother 35:563–565, 2001 pine or olanzapine. Management of schizo- Schizophr Bull 25:321–335, 1999 34. Brugman NJ, Cohen D, De Vries RH: Di- phrenia in general should include a greater 17. Thonnard-Neumann E: Phenothiazines abetes mellitus ontstaan na behandeling attention to medical risks, and effective and diabetes in hospitalized women. Am J met clozapine. Ned Tijdschr Geneeskd 144: diet and exercise programs are needed. Psychiatry 124:978–982, 1968 437–439, 2000 18. Waitzkin LA: A survey for unknown dia- 35. Croarkin PE, Jacobs KM, Bain BK: Dia- betes in a mental hospital. Diabetes 15: betic ketoacidosis associated with risperi- 164–172, 1966 done treatment? (Letter) Psychosomatics References 19. Benson V, Marano MA: Current estimates 41:369–370, 2000 1. Brown S: Excess mortality of schizophre- from the National Health Interview Sur- 36. Procyshyn RM, Pande S, Tse G: New-on- nia: a meta-analysis. Br J Psychiatry 171: vey [article online], 1995. Available from set diabetes mellitus associated with 502–508, 2001 http://www.cdc.gov/nchs/data/series/sr_ quetiapine. Can J Psychiatry 45:668–669, 2. Axelsson R, Lagerkvist-Briggs M: Factors 10/sr10_199acc.pdf 2000 predicting suicide in psychotic patients. 20. Braceland FJ, Meduna LJ, Vaichulis JA: 37. Rigalleau V, Gatta B, Bonnaud S, Bour- Eur Arch Psychiatr Clin Neurosci 241:259– Delayed action of insulin in schizophre- geois ML, Vergnot V, Gin H: Diabetes as a 266, 1992 nia. Am J Psychiatry 102:108–110, 1945 result of atypical antipsychotic drugs: a 3. Cohen LJ, Test MA, Brown RL: Suicide 21. Freeman H: Resistance to insulin in men- report of three cases. Diabet Med 17:484– and schizophrenia: data from a prospec- tally disturbed soldiers. Arch Neurol Psy- 486, 2000 tive community treatment study. Am J chiatry 56:74–78, 1946 38. Roefaro J, Mukherjee SM: Olanzapine-in- Psychiatry 147:602–607, 1990 22. Harris MI: Impaired glucose tolerance in duced hyperglycaemic nonketotic coma. 4. Mortensen PB, Juel K: Mortality and the U.S. population. Diabetes Care 12: Ann Pharmacother 35:300–302, 2001 causes of death in first admitted schizo- 464–474, 1989 39. Selva KA, Scott SM: Diabetic ketoacidosis phrenic patients. Br J Psychiatry 163:183– 23. Kasanin J: The blood sugar curve in men- associated with olanzapine in an adoles- 189, 1993 tal disease. Arch Neurol Psychiatry 16: cent patient. J Pediatr 138:936–938, 2001 5. Davies LM, Drummond MF: The eco- 414–419, 1926 40. Smith H, Kenney-Herbert J, Knowles L: nomic burden of schizophrenia. Psychiatr 24. Langfeldt G: The insulin tolerance test in Clozapine-induced diabetic ketoacidosis. Bull 14:522–525, 1990 mental disorders. Acta Psychiatr Scand 80: AustNZJPsychiatry 120–121, 1999 6. Rupp A, Keith SJ: The cost of schizophre- 189–200, 1952 41. Wirshing DA, Pierre JM, Eyeler J, Wein- nia. Psychiatr Clin North Am 16:413–423, 25. Lorenz WF: Sugar tolerance in bach J, Wirshing WC: Risperidone-asso- 1993 praecox and other mental isorders. Arch ciated new-onset diabetes. Biol Psychiatry 7. Seeman P, Chau-Wong M, Wong K: An- Neurol Psychiatry 8:184–196, 1922 50:148–149, 2001 tipsychotic drug doses and neuroleptic/ 26. Mukherjee S, Schnur DB, Reddy R: Fam- 42. Wu G, Dias P, Wu C, Li G-J, Kumar S, dopamine receptors. Nature 261:717–719, ily history of type 2 diabetes in schizo- Singh S: Hyperglycemia, hyperlipidaemia 1976 phrenic patients (Letter). Lancet 1:495, and periodic paralysis: a case report of 8. Seeman P: Atypical antipsychotics: mech- 1989 new side effects of clozapine. Prog Neuro- anism of action. Can J Psychiatry 47:27– 27. Sernyak MJ, Douglas DL, Alarcon RD, Psychopharmacol Biol Psychiatry 24:1395– 38, 2002 Losonczy MF, Rosenheck R: Association 1400, 2000 9. Adams PF, Marano MA: Current estimates of diabetes mellitus with use of atypical 43. Koller EA, Schneider B, Bennett K, Du- from the National Health Interview Sur- neuroleptics in the treatment of schizo- bitsky G: Clozapine-associated diabetes. vey [article online], 1994. Available from phrenia. Am J Psychiatry 159:561–566, Am J Med 111:716–723, 2001 http://www.cdc.gov/nchs/data/series/sr_ 2002 44. Koller EA, Doraiswamy PM: Olanzapine- 10/sr10_193acc.pdf 28. Harris MI, Flegal KM, Cowie CC, Eber- associated diabetes mellitus. Pharmaco- 10. Bhide M, Tiwari N, Balwani J: Effect of hardt MS, Goldstein DE, Little R, Wieder- therapy 22:841–852, 2002 chlorpromazine on peripheral utilization meyer H, Byrd-Holt D: Prevalence of 45. Mahmoud R, Gianfrancesco F, Grogg A, of glucose. Arch Int Pharmacodyn 156: diabetes, impaired fasting glucose toler- Nasrallah HA: Differential effects of anti- 166–171, 1965 ance in US adults: the Third National psychotics on type 2 diabetes: findings 11. Charatan FBE, Bartlett NG: The effect of Health and Nutrition Examination Sur- from a large health plan database. In Pro- chlorpromazine (“Largactil”) on glucose vey. Diabetes Care 21:518–524, 1998 ceedings of the 39th Annual Meeting of the tolerance. J Mental Sci 191:351–353, 1955 29. Liebzeit KA, Markowitz JS, Caley CF: American College of Neuropsychopharma- 12. Dixon L, Weiden P, Delahanty J, Gold- New onset diabetes and atypical antipsy- cology, San Juan, Puerto Rico, Dec. 10–14, berg R, Postrado L, Lucksted A, Lehman chotics. European Neuropsychopharmacol- 2001. p. 199 A: Prevalence and correlates of diabetes in ogy 11:25–32, 2001 46. Henderson DC, Cagliero E, Gray C, Nas- national schizophrenia samples. Schizo- 30. Mir S, Taylor D: Atypical antipsychotics rallah RA, Hayden DL, Schoenfeld DA, phr Bull 26:903–912, 2000 and hyperglycaemia. Int Clin Psychophar- Goff DC: Clozapine, diabetes mellitus, 13. Dynes JB: Diabetes in schizophrenia and macol 16:63–74, 2001 weight gain, and lipid abnormalities: a

1604 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Lean and Pajonk

five-year naturalistic study. Am J Psychia- of the literature. Pharmacopsychiatry.In 74. Dwyer DS, Pinkofsky HB, Liu Y, Bradley try 157:975–981, 2000 press RJ: Antipsychotic drugs affect glucose up- 47. Wang PS, Glynn RJ, Ganz DA, Schnee- 62. Cagliero E, Borba CP, Hayden DL, take and the expression of glucose trans- weiss S, Levin R, Avorn J: Clozapine use Schoenfeld DA, Goff DG, Henderson DC: porters in PC12 cells. Prog Neuro- and risk of diabetes mellitus. J Clin Psycho- Clozapine and olanzapine induce insulin Psychopharmacol Biol Psychiatry 23:69– pharmacol 22:236–243, 2002 resistance in patients with schizophrenic 80, 1999 48. Berry SA, Mahmoud RA: Normalization of disorders (Abstract). Diabetes 50 (Suppl. 75. Bromel T, Blum WF, Ziegler A, Schulz E, olanzapine-associated abnormalities of 2):A91, 2001 Bender M, Fleischhaker C, Remschmidt insulin resistance and insulin release after 63. Newcomer JW, Haupt DW, Fucetola R, H, Krieg JC, Hebebrand J: Serum leptin switch to risperidone: the Risperidone Melson AK, Schweiger JA, Cooper BP, levels increase rapidly after initiation of Rescue Study (Poster). Presented at the Selke G: Abnormalities in glucose regula- clozapine therapy. Mol Psychiatry 3:76– American College of Neuropsychophar- tion during antipsychotic treatment of 80, 1998 macology meeting, Hawaii, December, schizophrenia. Arch Gen Psychiatry 59: 76. Kraus T, Haack M, Schuld A, Hinze-Selch 2001 337–345, 2002 D, Kuhn M, Uhr M, Pollmacher T: Body 49. IMS Health National Prescription Audit 64. Leysen JE, Janssen PMF, Hellen L, Gom- weight gain and leptin plasma levels dur- Plus, 2001 meren W, Van Gomple P, Lesage AS, Me- ing treatment with antipsychotic drugs. 50. Chue P, Welch R: Investigation of the gens AAHP, Schotte A: Receptor inter- Am J Psychiatry 156:312–314, 1999 metabolic effects of antipsychotics in pa- actions of new antipsychotics: relation to tients with schizophrenia (Poster). Pre- pharmacodynamic and clinical effects. Int 77. Melkersson KI, Hulting AL, Brismar KE: sented at the 51st Annual Meeting of the J Psychiatry Clin Pract 2:S3–S17, 1998 Different influences of classical antipsy- Canadian Psychiatric Association, Que- 65. Tecott LH, Sun LM, Akana SF, Strack AM, chotics and clozapine on glucose-insulin bec, October–November 2001 Lowenstein DH, Dallman MF, Julius D: homeostasis in patients with schizophre- 51. Koro CE, Fedder DO, L’Italien GJ, Weiss Eating disorders and epilepsy in mice nia or related psychoses. J Clin Psychiatry SS, Magder LS, Kreyerbuhl J, Revicki DA, lacking 5-HT2c receptors. Na- 60:783–791, 1999 Buchanan RW: Assessment of indepen- ture 374:542–546, 1995 78. Woods SC, Kaiyala K, Porte D, Schwartz dent effect of olanzapine and risperidone 66. Chaouloff F, Laude D, Baudrie V: Effects MW: Food intake and energy balance. In

on risk of diabetes among patients with of the 5-HT1C/5 5-HT2 receptor agonist Diabetes Mellitus. Porte D, Sherwin RS, schizophrenia: population based nested DOI and ␣-methyl-5-HT on plasma glu- Eds. Stamford, CT, Appleton & Lange, case-control study. BMJ 325:243–245, cose and insulin levels in the rat. Eur 1997, p. 175–192 2002 J Pharmacol 187:435–443, 1990 79. Herra´n A, Garcia-Unzueta MT, Amado 52. Czobor P, Volavka J, Sheitman B, Linden- 67. Baudrie V, Chaouloff F: Repeated treat- JA, de la Maza MT, A´ lvarez C, Va´zquez- meyer J-P, Citrome L, McEvoy J, Cooper ment with the 5-HT1A receptor agonist, Barquero JL: Effects of long-term treat- TB, Chakos M, Lieberman JA: Antipsy- isapirone, does not affect 8-OH-DPAT- ment with antipsychotics on serum leptin chotic-induced weight gain and therapeu- and stress-induced increases in plasma levels. Br J Psychiatry 179:59–62, 2001 response: a differential association. levels in the rat. Eur J Pharma- 80. Davis JM, Andriukutis S: The natural J Clin Psychopharmacol 22:244–251, 2002 col 198:129–135, 1991 course of schizophrenia and effective 53. Meyer JM: Effects of atypical antipsychot- 68. Sugimoto Y, Yamada J, Kimura I, Horiska maintenance therapy. J Clin Psychophar- ics on weight gain and serum lipid levels. K: The effects of the serotonin 1A receptor macol 6:2–10, 1986 J Clin Psychiatry 62:27–34, 2001 agonist on tolbutamide-in- 81. Csernansky JG, Mahmoud R, Brenner R, 54. Harris MI: Prevalence of cardiovascular duced hypoglycemia in rats. Biol Pharm for the Risperidone-USA-79 Study Group: risk factors at the diagnosis of type 2 dia- Bull 18:1296–1298, 1995 A comparison of risperidone and haloper- betes. Clin Invest Med 18:231–239, 1995 69. Uvnas-Moberg K, Ahlenius S, Alster P, idol for the prevention of relapse in pa- 55. Pi-Sunyer FX: Medical hazards of . Hillegaart V: Effects of selective serotonin tients with schizophrenia. N Engl J Med Ann Intern Med 119:655–660, 1993 and dopamine agonists on plasma levels 346:16–22, 2002 56. UKPDS Study Group: Characteristics of of glucose, insulin, and glucagon in the 82. Burns MJ: The pharmacology and toxicol- newly presenting type 2 diabetic patients: rat. Neuroendocrinology 63:269–274, 1996 ogy of atypical antipsychotic agents. Clin male preponderance and obesity at differ- 70. Wozniak KM, Linnoila M: Hyperglycemic Toxicol 39:1–14, 2001 ent ages. Diabet Med 5:154–159, 1998 properties of serotonin receptor antago- 83. Ereshefsky L: Pharmacokinetics and drug 57. Allison DB, Mentore JL, Heo M, Chandler nists. Life Sci 49:101–109, 1991 LP, Cappelleri JC, Infante MC, Weiden PJ: 71. Goldstein LE, Sporn J, Brown S, Kim H, interactions: update for new antipsychot- Antipsychotic-induced weight gain: a Finkelstein J, Gaffey GK, Sachs G, Stern ics. J Clin Psychiatry 57:12–15, 2002 comprehensive research synthesis. Am J TA: New-onset diabetes mellitus and dia- 84. Indiana University Department of Medi- Psychiatry 156:1686–1696, 1999 betic ketoacidosis associated with olanza- cine Cytochrome P450 Drug Interaction 59. Wirshing DA, Wirshing WC, Kysar L, pine treatment. Psychosomatics 40:438– Table [article online], 2003. Available Berisford MA, Goldstein D, Pashdag J, 443, 1999 from http://medicine.iupui.edu/flockhart/. Mintz J, Marder SR: Novel antipsychotics: 72. Melkersson KI, Hulting AL, Brismar KE: Accessed 2 July 2002 comparison of weight gain liabilities. Elevated levels of insulin, leptin, and 85. Neaton JD, Wentworth D: Serum choles- J Clin Psychiatry 60:358–363, 1999 blood lipids in olanzapine-treated pa- terol, blood pressure, cigarette smoking, 60. Wirshing DA: Adverse effects of atypical tients with schizophrenia or related psy- and death from coronary heart disease: antipsychotics. J Clin Psychiatry 62:7–10, choses. J Clin Psychiatry 61:742–749, 2000 overall findings and differences by age for 2001 73. Meyer JM: Antipsychotics and severe hy- 316,099 white men: Multiple Risk Factor 61. Cohen D: Atypical antipsychotics and perlipidemia. J Clin Psychopharmacol 21: Intervention Trial Research Group. Arch new-onset diabetes mellitus: an overview 369–374, 2001 Intern Med 152: 56–64, 1992

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1605