This review describes the common effects of psychotropic drugs on the cardiovascular system and offers guidance for practical management. Selected reports from the literature describing common side effects associated with psychotropic drugs are reviewed, and suggestions for further reading are given throughout the text. Orthostatic hypotension is the most common adverse autonomic side effect of antipsychotic drugs. Among the atypical antipsychotics the risk of orthostatic hypotension is highest with clozapine and among the conventional drugs the risk is highest with low potency agents. Rarely, orthostatic hypotension may result in neurocardiogenic syncope. QTc prolongation can occur with all antipsychotics but an increased risk is seen with pimozide, thioridazine, sertindole and zotepine. QTc prolongation is a marker of arrhythmic risk. Torsade de pointe, a specific arrhythmia, may lead to syncope, dizziness or ventricular fibrillation and sudden death. Heart muscle disease presents most commonly in the elderly as chronic heart failure, but myocarditis and cardiomyopathy, although relatively rare, are devastating, but potentially reversible complications of psychotropic drug therapy have been particularly linked to clozapine treatment. Patients with severe mental illness (SMI) are a 'high risk' population with regard to cardiovascular morbidity and mortality. It is probable that many patients accumulate an excess of 'traditional' risk factors for the development of cardiovascular disease, but other mechanisms including psychotropic drugs may also be influential in increasing risk in this vulnerable group. These risks need to be seen in the context of the undoubted therapeutic efficacy of the psychotropic armamentarium and the relief that these drugs bring to those suffering from mental disorder.

Individuals with serious mental illness experience excess morbidity and mortality, including an increased prevalence of diabetes mellitus and cardiovascular disease. Cardiovascular disease is the leading cause of death in persons with serious mental illness, and the elevated prevalence of obesity in this population is of particular concern. Obesity is an independent cardiometabolic risk factor that impacts morbidity and mortality and contributes to the development of other cardiometabolic risk factors, such as dyslipidemia and hypertension. In addition, obesity is a major risk factor for type 2 diabetes, with the relative risk of diabetes increasing with body mass index. Increased abdominal fat is strongly associated with insulin resistance, which can lead to impaired glucose regulation. Abdominal obesity, hyperglycemia, hypertension, and dyslipidemia are key components of the metabolic syndrome, a constellation of cardiometabolic risk factors linked by their common association with insulin resistance. Evidence from large clinical samples indicates a high prevalence of metabolic syndrome and all of its components in persons with serious mental illness, particularly in patients with schizophrenia. In addition, psychotropic agents, including some antipsychotic medications, are associated with substantial weight gain, as well as with adiposity-dependent and possibly adiposity-independent changes in insulin sensitivity and lipid metabolism, which increase the risk of diabetes and cardiovascular disease. Among the second-generation antipsychotics, clozapine and olanzapine are associated with the highest risk of substantial weight gain, similar to the weight gain potential associated with low-potency first-generation antipsychotics such as thioridazine or chlorpromazine, as well as with an increased risk of diabetes and dyslipidemia. Various strategies for monitoring cardiometabolic risk factors in patients with mental illness are discussed in this review.

Many clinicians have become concerned about the safety of new antipsychotics particularly in view of the association of agranulocytosis with clozapine and of aplastic anaemia with remoxipride. The Committee on Safety of Medicines and Medicines Control Agency 'yellow card' post-marketing surveillance data were analysed for reports of haemopoietic disorders with the 16 antipsychotics in common use. Corrections for relative risk were made in three separate ways: (i) control for degree of use, using Northern Ireland prescribing data for 1995; (ii) percentage of total reports from 1963 to 1996; and (iii) examination of the first 5 years' post-marketing data only. After clozapine and remoxipride the highest risks of haemopoietic reactions appeared to be associated with the aliphatic phenothiazine derivatives thioridazine and chlorpromazine. There is therefore no evidence of any increased risk with high-potency drugs such as haloperidol or pimozide or with the newer drugs such as sulpiride or risperidone. Continued vigilance, however, is necessary as more new atypicals become available and begin to be widely prescribed.

Induction of bradykinesia by SM-9018, a novel 5-HT2 and D2 antagonist, was compared with that of other neuroleptics using the pole test in mice. Neuroleptics including SM-9018, haloperidol, chlorpromazine, and thioridazine dose dependently induced bradykinesia in the pole-descending behavior of mice with relative potencies consistent with those for catalepsy induction. SM-9018 was about 70 times weaker than haloperidol and twice as weak as thioridazine in inducing bradykinesia. Other CNS drugs such as barbiturates and antidepressants had no effects. Haloperidol-induced bradykinesia was significantly attenuated by a cholinergic muscarinic antagonist (i.e., trihexyphenidyl) and 5-HT2 antagonists (i.e., ritanserin and cyproheptadine) whereas that caused by SM-9018 was relatively resistant to the 5-HT2 antagonists. These findings suggest that SM-9018 is weaker than other neuroleptics in inducing extrapyramidal side effects and that the 5-HT2 blocking activity of SM-9018 may contribute to its atypical neuroleptic property.

Direct injections of muscimol into the ventral tegmental area (VTA) or substantia nigra zona reticulata (SNR) have been used to selectively stimulate the mesolimbic and nigro-striatal dopamine pathways respectively. Such injections induced locomotor activity, rearing, sniffing and in some animals an intermittent grooming response. These responses were rapid in onset, dose-related and relatively short lasting (less than 40 min). Selective increases in dopamine turnover were seen in the nucleus accumbens and in the striatum following VTA and SNR injections of muscimol (100 ng) respectively. Haloperidol inhibited the behavioural consequences of VTA and SNR injections of muscimol with similar potency (ED50S 0.01-0.03 mg/kg IP), and fluphenazine did likewise (ED50S 0.05-0.16 mg/kg IP). However, thioridazine (ED50S VTA: 1.45-2.04 mg/kg IP, SNR 8.50-9.20 mg/kg IP) and in particular clozapine (ED50S VTA: 0.24-0.58 mg/kg IP, SNR: 6.10-9.70 mg/kg IP) were more potent at inhibiting the locomotor activity and sniffing responses due to VTA rather than SNR administered muscimol. Since dopamine D2 antagonists are believed to exert their anti-psychotic effects via an action on mesolimbic dopaminergic systems, and their ability to induce extrapyramidal side effects (EPS) is thought to be due to an action on nigro-striatal dopamine systems, these results suggest that the behavioural models described can be used to predict efficacy and side-effect liability of potential neuroleptic drugs.

From 1976 to 1980, there were 3,193 admissions due to acute drug overdosage at the resuscitation centre of the Rudolf-Virchow-Hospital, Free University of Berlin. The frequency and the characteristics of selfpoisoning with antidepressants and low-potency neuroleptic drugs (mainly perazine and thioridazine) were determined. These drugs were involved in 92 cases (3%) during this five-year period. Amitriptyline--in combination with a benzodiazepine--was the most common antidepressant taken by the patients. Ten of the patients required assisted ventilation. Complete ECG recordings were carried out in 24 patients. The most common abnormality was a prolonged QTc interval (21 patients) followed by a QRS duration of 0.11 seconds or longer (17 patients). Sinus tachycardia was present in half of the cases. In no cases did the medical records describe convulsions or cardiac arrhythmias requiring special treatment. The percentage of patients showing ECG changes and respiratory depression was higher when other drugs such as ethanol were ingested along with antidepressants than when only antidepressants were taken. One patient died after six weeks in hospital due to a complicated abscess. From 1976 to 1980, the incidence of antidepressant selfpoisoning was relatively low compared with findings from other studies. Data from other studies suggest that, in the period 1980-1987, there was an increase in the incidence of antidepressant selfpoisoning, at least in the West Berlin and Munich areas. However, these figures are much lower than those reported by British, American, and Australian authors.

A simple and rapid screening method, where the ability of neuroleptic drugs to antagonise the abnormal pattern of exploration induced by a low dose of d-amphetamine in a 10 min test, was evaluated. The d-amphetamine 2 mg/kg pretreatment induced both an increased locomotion, thought to reflect an increased dopamine transmission in the nucleus accumbens, and weak stereotypies, thought to reflect an increased dopamine transmission in the neostriatum. Haloperidol, chlorpromazine and thioridazine blocked all ongoing behaviours while clozapine and sulpiride, regarded as causing less extrapyramidal side effects in the clinic, only antagonised the d-amphetamine induced locomotion. The findings support the notion that the common site of action for anti-psychotic drugs is blockade of dopamine receptors outside the neostriatum while the blockade of dopamine receptors within the striatum probably are related to the propensity of these drugs to induce the extrapyramidal side effects. It seems possible with this method to screen neuroleptic drugs for their relative potency in blocking limbic and striatal dopamine receptors simultaneously in one short experiment. The method might be used when new anti-psychotic drugs with low incidences of extrapyramidal side effects are sought for.

The Medicare Utilization Review Committee conducted a survey to determine the use of drugs with anticholinergic effects in confused elderly nursing home patients. Twenty-nine patients (34.5 percent) were receiving anticholinergic drugs, predominantly from the antidepressant and antipsychotic class. Thioridazine (Mellaril) was the most frequently used antipsychotic. No patients received higher than the equivalent recommended daily dose of atropine when calculated in terms of relative anticholinergic potency. No statistically significant correlation could be found between the presence of confusion and the amount of anticholinergics administered. Confusion and cognitive deterioration frequently are encountered when caring for aged patients. Clinical experience and recent studies suggest that anticholinergics may increase the risk of or exacerbate existing confusion, and this possible effect of anticholinergic activity should be considered when prescribing for elderly patients.

Using the radioligand binding assay, the in vitro potency of three neuroleptic drugs at dopaminergic ([3H]spiroperidol), alpha adrenergic ([3H]WB-4101) and muscarinic cholinergic ([3H]quinuclidinyl benzilate) receptor binding sites was compared to the potency of 12 metabolites of these drugs at the same receptors. Metabolites resulting from the oxidation of the ring sulfur of either thioridazine or fluphenazine were very weak at all three receptors. On the other hand, two metabolites, thioridazine-S-sulfoxide (mesoridazine) and thioridazine-S-sulfone (sulforidazine) were more potent than thioridazine at both the dopaminergic and alpha adrenergic receptors but less potent at muscarinic receptors. The binding data correlated very well with published clinical data on the relative potency and incidence of autonomic and extrapyramidal side effects of these three drugs. Since several of the metabolites which are potent receptor blockers accumulate in the plasma during chronic neuroleptic treatment, it appears that both the therapeutic and side effects of neuroleptic drug treatment are due in large part to metabolites of the administered drug.

In rats, a single injection of antipsychotic drugs produced a transitory change in the kinetic state of tyrosine hydroxylase (TH) in striatum and nucleus accumbens. The affinity of TH for 2-amino-4-hydoxy-6,7-dimethyl-5,6,7,8-tetrahydropterine (DMPH4) and the Vmax with respect ot tyrosine were increased. The relative potencies of anti-psychotics to change the kinetics of TH in striatum and nucleus accumbens when injected into rats were measured in the presence of 0.4 mM DMPH4. The doses of methiothepin, pimozide and halopridol which increased the affinity for DMPH4 of striatal TH were lower than those required to produce a similar change in nucleus accumbens. In contrast, thioridazine and clozapine were more effective in nucleus accumbens than in striatum. Chlorpromazine was equally active in these two tissues. Haloperidol increased the turnover rate of dopamine in striatum with doses that are relatively smaller than those required in the nucleus accumbens. Clozapine was more active in increasing turnover rate of dopamine in nucleus accumbens; the activity of chlorpromazine in these two tissues was equal. These results suggest that antipsychotics with high incidence of extrapyramidal side effects affect the nigrostriatal dopaminergic pathway selectively.