Antipsychotics should be used when BPSD are not effectively controlled by all the non-pharmacological measures. Therefore, before antipsychotic treatment, the possible trigger factors (fever, constipation, urinary retention, pain, hunger, neurosensitive deficits) should be explored and controlled. One also needs to exclude environmental (i.e., strong lights, noise, etc.) and iatrogenic factors (i.e., use of anticholinergics, benzodiazepines).

Prior to initiating treatment with an antipsychotic medication, the patient should be assessed for movement disorders, orthostatic blood pressure and pulse, and the following baseline laboratory studies should be checked: fasting blood sugar, lipid battery, white blood cells (WBC) count with differential, liver function battery, and electrocardiogram with corrected QT interval (QTc) read or calculated. Table 3 reports the factors to be assessed before prescribing an antipsychotic drug.

There are generally two thresholds for stopping dose titration; either a satisfactory response is acquired, or distressing adverse effects occur. Depending on the pharmacodynamics and pharmacokinetic interactions between the antipsychotic and the individual patient's central nervous system and clinical status, doses higher or lower than those recommended in clinical trials may be required to achieve a therapeutic response. This may be more likely in patients not generally accepted into clinical trials, because of extreme elderly, medical co-morbidities, or taking multiple medications.18

For all elderly patients with dementia, antipsychotic dose is started at the lowest feasible dose, and some patients are sufficiently controlled by this minimum dose. Others will require upward dose titration, depending on factors such as age, renal function, hepatic function, and diagnosis. In general, dementia-related psychotic symptoms require lower doses of antipsychotics than those with schizophrenia.192021

In younger patients with schizophrenia, antipsychotic are generally titrated as rapidly as tolerated to target doses presumed to be therapeutic. In practice, initial titration usually proceeds at a rate of one to two dose increases per week. For frail elders or those who are medically compromised, titration should be even slower. For medications with a wide dosing range (e.g., low potency drugs like quetiapine), increases are made after the steady state target dose is attained in order to avoid overshooting the minimal therapeutic dose for the individual patient. During the initial stages of antipsychotic dose titration, it may be advisable to split doses (two to three times a day) or retarded formulation to minimize problems from side effects such as orthostatic hypotension and sedation. For clozapine, quetiapine, ziprasidone, and low-potency typical agents, continued divided dosing (usually two times a day) may be recommended. For olanzapine, risperidone, aripiprazole, and high-potency typical antipsychotics, single bedtime doses may be used.22

In schizophrenia, the therapeutic drug-level of clozapine, haloperidol, fluphenazine, thiothixene, and perphenazine were established, but whether these therapeutic drug levels are equally applicable to the elderly patients with dementia are uncertain. However, extreme values of serum level may be helpful to monitor the elderly patients with dementia: a zero level could confirm suspected noncompliance; conversely, a very high level could confirm suspected toxicity. Therefore, if clinically the non-compliance or toxicity are suspected, checking drug level may be helpful.

Initial management of apparent treatment resistance follows a logical sequence of confirming the diagnosis, ruling out comorbid substance abuse or other confounding conditions, ascertaining compliance, and adjusting the medication dose. Dose adjustments should be usually made on the basis of clinically observed therapeutic effects and adverse effects, but drug levels can be checked to determine extreme values (zero or far above therapeutic range). The patient experienced EPS or akathisia due to antipsychotics may benefit more from a dose decrease than a dose increase, since the adverse effect can be overwhelming the therapeutic benefit.

Until now, clinical guideline for treatment resistance in elderly patients with dementia is not yet established. In schizophrenia, if true treatment resistance is occurred, antipsychotics should first be switched to another atypical agent other than clozapine and maintained on an adequate dose of that agent for 6-12 weeks. If there is a partial response, the trial should be maintained up to 16 weeks. If there is still an inadequate response, a trial of clozapine should be considered. This trial should be continued, if possible, for a period of 6-12 months. Adjunctive therapies such as mood stabilizers could be used to augment a partial response.15

If patients do not respond to dose adjustment of antipsychotics and modulation of other aggravating factor, then the clinician should consider switching antipsychotics. However, switching antipsychotics can be a potentially destabilizing event. The risk for rebound and withdrawal phenomena is greatest when the pre- and post-switch antipsychotics are considerably different regarding binding affinity for specific receptors (i.e., pharmacodynamic rebound) or regarding their respective half life (i.e., pharmacokinetic rebound).23 When one antipsychotics are switched to another, it is generally safer to taper and discontinue pre-switch antipsychotics before starting the second. When this is not clinically feasible because of severe psychotic symptoms, the post-switch drug can be added to the pre-switch drug, which can then be tapered and discontinued (cross-tapering). During the course of antipsychotic cross-tapering, careful attention should be paid to potential adverse effects such as cardiac conduction dysfunction, orthostatic hypotension and EPS. One-step switching, in which the first antipsychotic is stopped and the second started (often at a therapeutic or near-therapeutic dose), has not been demonstrated to be safe in elderly patients. It has been reported to be done safely in nonelderly patients with aripiprazole.24 It cannot be done with clozapine.

In the course of cross-tapering to switch from an antipsychotic (e.g., conventional) to other antipsychotics (e.g., atypical), the patients can be clinically aggravated as the cross-taper proceeds. In this case, it is not easy to know whether the problem is the decreasing level of previous drug, the increasing level of switching drug, or rebound phenomena. It is in these patients more than in those switched for lack of efficacy where rebound phenomena are readily noticeable. Of note, however, the clinician should not prematurely conclude that the apparent reduced efficacy or symptomatic worsening is definitely due to the lack of efficacy of the newly initiated antipsychotic. Rather, intra-switch pharmacokinetic or pharmacodynamic rebound phenomena should be also considered. These can be managed by slowing down the switch schedule, which can maintain a sufficient dopamine blockade during the switch. Maintenance of sufficient blockade of dopamine or other relevant receptors can be achieved by either increasing the pre-switch antipsychotic back to a higher level and/or increasing the newly initiated antipsychotic to a more therapeutic dose. Alternatively, calming medications, such as benzodiazepines, medications with antihistaminic effects, sleep aides, etc. can be given transiently to overcome potentially concerning or destabilising, but frequently transient rebound phenomena. The easiest way to minimize the chance for adverse events during a switch between antipsychotics that differ strongly pharmacokinetically or pharmacodynamically is to implement an overlapping or "plateau" cross titration or switch.

If clinically possible, it may be best to complete the cross-taper to give the second medication an adequate trial as monotherapy. This is consistent with current guidelines advocating antipsychotic monotherapy as the gold standard of treatment. Although the combination of a conventional and an atypical antipsychotic is not recommended in the treatment of elderly patients with dementia, it is sometimes used if the clinical response is not adequate in monotherapy.

The usual duration of antipsychotic medication treatment is various. The duration of treatment is practically indefinite in primary psychosis like schizophrenia and delusional disorders. Elders with schizophrenia are best maintained long-term with a stable, low dose of antipsychotic-preferably an atypical agent. For elderly patients with dementia, the usual duration of treatment in is 3 months. After this time, periodic evaluation of the necessity of antipsychotics should be done and attempts to reduce the dose of the antipsychotic medication are recommended and most patients are controlled even in minimal dosage of antipsychotics. Unless adverse effects necessitate abrupt discontinuation, dose reduction should proceed slowly to minimize withdrawal symptoms.

Conventional antipsychotics given at the recommended geriatric (low) doses should be tapered over 2 weeks, at a minimum. Clozapine withdrawal should be carefully proceeded because of the anticholinergic and antiadrenergic activity as well as by the potential for rebound psychosis, such that the taper is best performed over a minimum of 4 weeks in elderly patients. An only slightly less protracted withdrawal period may apply to thioridazine because of its anticholinergic properties.

'Rapid tranquillisation' is a commonly practiced in patients with BPSD. The Royal College of Psychiatrists takes for its definition and its clinical guidance on managing violence: 'All medication given in the short-term management of disturbed/violent behavior should be considered as part of rapid tranquillisation'.25 The management of violence in patients with dementia should be a diagnosis-driven, and medication should be targeted not at the behavior itself, but at the underlying pathophysiology. According to this view, though, the pathophysiology of violent behavior is not completely understood in BPSD, antipsychotics, especially conventional antipsychotcis (e.g., haloperidol injection) is justifiable in aggressive patients associated with BPSD. These usages can not only target the underlying pathophysiology, but also augment the previously prescribed antipsychotics. These treatments may be useful for patients with minimal effective dose, and presenting intermittent burst of psychotic episodes.

Before symptomatic care of delirium, definitive treatment of the cause of delirium (e.g., correction of hyponatremia or treatment of urinary tract infection) should take precedence over symptomatic treatment. Antipsychotic medications can be used to treat psychosis (hallucinations, delusions, or thought disorganization) or severe and persistent agitation in the delirious patient when these symptoms cause suffering and pose a danger. Though, there is currently no consensus clinical guideline as to which antipsychotic agent is preferred; the most commonly used drugs are haloperidol, quetiapine, and risperidone.

Delirium is the only medical condition in geriatrics for which a conventional antipsychotic (haloperidol) continues to be used routinely. In the medically frail patients, haloperidol is useful because this has no significant hemodynamic effects at usual therapeutic doses, little or no respiratory depression, little effect on glucose regulation, and availability in intravenous form (so that low doses can be used with good effect and little adverse effect). A recent study reported an association of haloperidol use with lower hospital mortality in mechanically ventilated patients.26 Haloperidol is also less expensive than comparator drugs. The disadvantage of haloperidol is that high doses can prolong the QTc interval. In addition, this drug can cause well-documented EPS, neuroleptic malignant syndrome and tardive dyskinesia when used long-term. These risks have been thought to be lower for the atypical agents.

The advantage of quetiapine is that it is highly sedating, and thus can calm down agitated patients with delirium, but this medication can result in overly sedating in the frail elderly patient. In addition, quetiapine can cause hypotension and glucose dysregulation, even with short-term use. Risperidone can be effective in reducing psychosis and agitation in delirium when dosed appropriately. Unfortunately, this medication is too often dosed and titrated as it is for the younger population, resulting in akathisia in the delirious elder. Risperidone also can cause hypotension.

Regardless of which agent is used for the treatment of delirium, the duration of treatment should be limited; the medication should be used until symptoms abate and then should be tapered off over 3-5 days.

BPSD may be the consequence of secondary to cognitive impairment or underlying primary pathological abnormalities, which change the clinical expression of dementia.27 Agitation frequently occurs in dementia and antipsychotics can be effective, even if their use is off-label. A review performed in 2012 comparing the efficacy of off-label use of atypical antipsychotics in dementia suggested that olanzapine, aripiprazole, and risperidone have a moderate-to-high efficacy in agitation.15 Although the use of antipsychotics for dementia is off-label, antipsychotics is probably the best option for short-term treatment (6-12 weeks) of severe, persistent, and resistant aggression.28 Serious possible adverse events are a major limitation to long-term therapy, thus recommending that dosage be decreased and treatment discontinued whenever behavioral symptoms have been sufficiently controlled. At present, olanzapine, quetiapine, and risperidone present the similar efficacy in acute stages of disease, without inducing the less neurological adverse effects, which frequently seen in haloperidol.

Conventional antipsychotics have been widely used for BPSD, even if some studies showed an efficacy comparable to placebo.34 Atypical antipsychotics (in particular risperidone and olanzapine) showed an efficacy superior to placebo in randomized studies in BPSD treatment, with a better tolerability profile versus conventional drugs.1352930 Such tolerability profile is due to their characteristic mechanism of action,467 as previously described. Eleven randomized, placebo-controlled trials were carried out using conventional antipsychotics for the treatment of BPSD; they mostly involved small sample sizes and were performed over periods of 4 and 12 weeks (only 1 up to 16 weeks' duration).28313233 A significant but modest advantage over placebo for behavioral symptoms has been reported (59% vs. 41%).28 Some trials showed a significant improvement in aggression with haloperidol compared with placebo, but not in other symptoms of agitation. However, the use of typical antipsychotics is limited by the several severe adverse effects, such as parkinsonism, tardive dyskinesia, dystonia, and QTc interval prolongation. Furthermore, the use of several drugs in elderly patients, who are often affected by a number of comorbidities, raises the risk of drug interactions and consequent fatal adverse events.5 The QTc interval prolongation was most often reported following the use of thioridazine and droperidol, both of which have been withdrawn.

Eighteen placebo-controlled studies have examined the efficacy of atypical antipsychotics in patients with Alzheimer's dementia (AD) over a 6- to 12-week treatment period. Among them, only 3 trials were carried out for 6 to 12 months.34353637 Meta-analyses have been conducted of risperidone and aripiprazole and showed the efficacy of these agents in the short-term management (6-12 weeks) of aggression.3536 Both atypical drugs showed similar benefit. In other atypical antipsychotics (such as clozapine, sertindole, zotepine, amisulpride), no trials have investigated in AD yet.28

Clinical Antipsychotic Trials of Intervention Effectiveness-Alzheimer's Disease (CATIE-AD) was a cornerstone clinical study performed in 421 elderly AD outpatients with psychosis or agitated/aggressive behavior from 45 centers.38 The purpose of this study was to compare the effectiveness and adverse events of atypical antipsychotics and placebo in patients with AD and psychosis or agitated/aggressive behavior. The patients were randomly allocated into groups receiving olanzapine, quetiapine, risperidone, or placebo for up to 36 weeks. The primary end point was the medication discontinuation time for any reason (phase 1). The secondary end points were improvement on the Clinical Global Impression of Change scale and the medication discontinuation time due to ineffectiveness or intolerance. This study revealed that olanzapine and risperidone were significantly more efficacious than quetiapine and placebo when the discontinuation time of treatment due to a lack of efficacy was analyzed (22.1 and 26.7 weeks with olanzapine and risperidone, respectively, vs. 9.1 and 9.0 weeks with quetiapine and placebo, respectively). The placebo group was superior to the 3 drugs on the analysis of treatment interruption due to intolerability

This trial demonstrated that these drugs showed a high rate of adverse events and this can offset their efficacy, but some portion of patients can be benefited without adverse symptoms.38 Furthermore, secondary analysis of CATIE-AD results showed that risperidone was effective in a variety of behavioral areas compared with placebo at the time of discontinuation, whereas patients medicating olanzapine improved in areas of hostility and suspiciousness, but they showed a worsening of depressive symptoms compared with placebo.39

The results of the CATIE-AD have provoked the hot debate about the role of atypical antipsychotics in influencing cognitive performance in patients with AD. Some trials show that risperidone reduces the Mini-Mental State Examination scores in a statistically significant way compared with placebo, which is not the case with olanzapine and quetiapine. On the contrary, other trials indicate that quetiapine and olanzapine are responsible for the greater cognitive decline. The samples of individual clinical trials are not enough to determine whether the cognitive decline varies with antipsychotic use; however, this decline has been evident for all the molecules compared to placebo.40

A 2006 Cochrane Collaboration review of placebo controlled trials concluded that risperidone and olanzapine may improve aggression compared with placebo and risperidone that may improve psychosis relative to placebo.41

A meta-analysis performed in 2005 including 17 placebo-controlled studies of 4 drugs (risperidone, olanzapine, quetiapine, and aripiprazole) enrolled a combined 5110 elderly patients with dementia.9 Studies revealed a 4.5% mortality rate among patients who had been treated with atypical antipsychotics and have demonstrated an approximately 1.6-1.7 fold increase compared to placebo treated patients in mortality in these studies. Cardiac (i.e., heart failure and sudden death) or infectious (pneumonia) events were the main causes of death.944

However, other meta-analysis of randomized controlled trials36 also found that haloperidol was associated with about 2-fold increased mortality versus placebo. A number of observational studies have been completed in the last few years23414447 and these studies also showed that the risk was even higher than that associated with the use of atypical drugs. A retrospective cohort study based on health insurance database on 22890 patients 65 years or older showed that conventional antipsychotics were associated with a significantly higher risk of mortality than were atypical antipsychotics at all intervals studied. The risk of death was higher within the first 40 days and with higher dosages of conventional antipsychotics.3 Similarly, Liperoti et al.44 showed that, among 6524 new users of atypical antipsychotics and 3205 new users of conventional antipsychotics residents of nursing homes in 5 US states during the years 1998-2000, the rate of death was 22% higher for users of conventional versus atypical antipsychotics [hazard ratio, 1.26; 95% confidence interval (CI), 1.13-1.42]. A retrospective cohort study of 37241 elderly people residents in British Columbia showed a greater incidence in mortality associated with high dosages of typical antipsychotics and during the first 40 days of treatment.48 A population-based cohort study of data on Medicaid, Medicare, the Minimum Data Set, the National Death Index, and a national assessment of nursing home quality was carried out to evaluate the risks of mortality associated with the use of antipsychotic drugs in elderly nursing home residents. The study was performed on 75524 elderly patients with behavioral disorders who were treated with haloperidol, aripiprazole, quetiapine, risperidone, olanzapine, and ziprasidone. The risk of mortality increased following higher doses of these drugs and was highest with haloperidol and least for quetiapine.45

On the other side, a nested case-control study on 2385 elderly demented patients from the Dutch general practice database reported no statistically significant differences in the risk for all-cause mortality among users of conventional and atypical drugs.47 Similarly, a retrospective cohort study of Medicaid patients in Tennessee of 44218 and 46089 baseline users of single typical and atypical drugs, respectively, and 186600 matched controls of antipsychotic drugs revealed higher rates of sudden cardiac death for users. Former users of antipsychotic drugs had no significantly increased risk. For both classes of drugs, the risk for current users increased significantly with an increasing dose.2 The risk is increased in elderly patients for the high number of drugs taken for comorbidities.

Cardiac adverse effects including sinus tachycardia, atrial and ventricular extrasystoles, QTc interval prolongation, T-wave inversion, ST-segment depression, and atrioventricular blocks have been reported for atypical and conventional antipsychotics.4849 Among them, QTc prolongation associated with antipsychotics is most clinically worrisome for frail elderly patients.50 As shown in Fig. 2, the QT interval spans the time from the start of ventricular depolarization (the Q wave of the QRS complex) to the end of ventricular repolarization (the T wave). The QT shortens as the heart rate increases and is corrected for heart rate (QTc) by one of several algorithms. The Bazett correction is most widely used formula and this is defined as $\text{QTcB}=\frac{\text{QT}}{\sqrt{\text{RR}}}$ Usually, QTc prolongation results from delayed repolarization, which may enable early after-depolarizations to occur and set the stage for extrasystoles and eventual torsades de pointes. QTc prolongation is defined as a QTc value >450 ms for men or >470 ms for women.51 The QTc intervals >500 ms are particularly worrisome, since they are associated with a significant risk of ventricular dysrhythmia, torsade de pointes and sudden death.525354 Antipsychotic medications are known to affect the QTc in a dose and level-dependent manner.51 In one large epidemiologic study, patients taking antipsychotics in dose-equivalents of more than 100 mg thioridazine had a 2.4-fold risk of sudden death.55 Pfizer/FDA reported that thioridazine had the greatest risk of prolonging QTc, following in descending order by ziprasidone, quetiapine, chlorpromazine, risperidone, olanzapine, aripiprazole, and haloperidol. Indeed, some epidemiologic studies have widely shown the direct relationship between conventional antipsychotics and the risk of sudden death;256 the risk is also increased with higher doses, both for conventional and atypical antipsychotic drugs2 and with concurrent treatments in elderly people. Table 4 reports the degree of QT prolongation by antipsychotic drugs.535758 Fig. 3 reports a practical algorithm expressing arrhythmogenic hazard risk according to QTc values.59 Moreover, the studies on antipsychotic drugs have shown that atypical medications appear to be more "cardiosafe" than conventionals; only clozapine has been linked to serious cardiac adverse effects, such as myocarditis.5

Liperoti et al.60 conducted a case-control study among nursing home residents in 6 US states using the Systematic Assessment of Geriatric drug use via Epidemiology database. Cases were residents hospitalized for ventricular arrhythmias or cardiac arrest between July 4, 1998, and December 30, 1999. The sample consisted of 649 cases and 2962 controls. The use of conventional antipsychotics was associated with a nearly 2-fold increase in risk of hospitalization for ventricular arrhythmias or cardiac arrest [adjusted odds ratio (OR), 1.86; 95% CI, 1.27-2.74]. There was no increased risk associated with the use of atypical antipsychotics (OR, 0.87; 95% CI, 0.58-1.32). The risk of hospitalization for ventricular arrhythmias or cardiac arrest was 3.27 times higher among conventional users with cardiac disease (95% CI, 1.95-5.47).60

Urgent treatment of prolonged QTc includes discontinuing all offending medications; suppressing early after-depolarizations by administration of intravenous magnesium, potassium, and lidocaine; and increasing the heart rate to shorten the QT interval by pacing or administration of isoproterenol. The patient may require sedation during the treatment period.

However, a recent article by Pariente et al.61 has shown that antipsychotic use is associated with a modest and time-limited increase in the risk of myocardial infarction among community-dwelling older patients treated with cholinesterase inhibitors. The risk is higher in the first 30 days of treatment (hazard ratio, 2.19; 95% CI, 1.11-4.32) and decreases over time.

Some of the atypical drugs were reported to cause congestive heart failure; however, this effect occurred infrequently following olanzapine (1/100 and 1/1000) and rarely following clozapine and quetiapine (<1/1000). Hypotension is a common adverse effect encountered with atypical drugs, especially following clozapine (9%), quetiapine (7%), and risperidone and olanzapine (5%); ziprasidone and haloperidol rarely cause hypotension (1% of the cases reported).62 Therefore, this adverse effect has to be taken into account in elderly patients treated with antihypertensive drugs, for the possible pharmacodynamic interaction.

Though, the FDA has issued a warning for stroke risk among elderly patients with BPSD who are medicated with risperidone, olanzapine, or aripiprazole, it is not yet clear whether other atypical or conventional antipsychotics should be included in this warning. A retrospective cohort study on 11400 people older than 65 years pooled results from 11 randomized controlled trials; it revealed increased incidences of cerebrovascular adverse events compared with placebo. Other studies reported a similar risk of stroke among users of both conventional and atypical antipsychotic drugs.6364 Interestingly, differential risk was reported for subgroups of conventional antipsychotics, such as phenothiazines and butyrophenones compared with benzamides and atypical antipsychotics.45 On the contrary, a UK-based electronic primary care records in the general practice research database reported that all antipsychotics are associated with an increased risk of stroke. The risk might be higher in patients receiving atypical antipsychotics than those receiving typical antipsychotics and in people with dementia compared with people without dementia.65

Potential mechanisms of stroke by atypical antipsychotics suggest cardiovascular effects (e.g., orthostatic hypotension, arrhythmias), thromboembolic effects, excessive sedation resulting in dehydration, and hemoconcentration and hyperprolactinemia.66 In particular, α1-receptor blockade may cause orthostatic hypotension and tachycardia. Elderly people with a preexisting cerebrovascular disease might experience a transient ischemic attack or even a cerebrovascular adverse event as a consequence of hypotension, which on its turn worsens the deficit in cerebral perfusion.48 Also, the rebound excess of catecholamines following orthostatic hypotension may cascade the consequent vasoconstriction and further aggravation of cerebral vascular deficit.

Other putative mechanisms are thromboembolism and hyperprolactinemia. Thromboembolism is probably related to immobility due to sedation for H1 receptor blockade; this enhances venous stasis, dehydration, and the consequent hemoconcentration. Hyperprolactinemia is frequently associated with the increase in endothelial dysfunction, platelet aggregation, and the decrease in insulin sensitivity.

EPSs are by far more common in elderly demented patients because brain levels of dopamine are usually decreased in these patients. A substantial amount of literature supports the belief that EPS is occurred significantly lower in atypical antipsychotic agents than typical agents.67 Because of the dose-dependent D2-receptor blockade, among atypical antipsychotics, risperidone is associated with the highest risk, whereas clozapine and quetiapine are associated with the lowest risk of EPSs.68 Risperidone is poised between the typical and atypical agents in its extrapyramidal profile; at doses above 4 mg in elders, it may have effects as pronounced as those of the conventional agents, whereas at low doses (1 mg/day or less), it has little extrapyramidal effect, except perhaps akathisia. The EPS in elderly patients tend to differ clinically from that found in younger populations. The risk of akathisia and dystonia decreases with age, with a rate of 3% and 15% respectively found in those over 65 years of age.69 On the other hand, drug-induced parkinsonism is more common among antipsychotic treated elders. This syndrome resembles idiopathic parkinsonism, with symptoms of bradykinesia, rigidity, impaired righting reflex, dysphagia, sialorrhea, autonomic instability, and even seborrhea. Symptoms do not usually occur immediately, but after 1-2 weeks of treatment.69

When EPS is developed in antipsychotic-treated elderly, discontinuing or lowering the dose of medication or switching to an atypical agent should be tried. If the offending antipsychotic is discontinued, EPS are usually improved within several weeks, but in some cases, signs and symptoms persist for a year or more. If parkinsonism is persisted, amantadine, dopamine reuptake blocker and D2 receptor agonist, can be used.

Tardive dyskinesia is an abnormal involuntary movements associated with antipsychotic medication treatment. This occurs more frequently among elderly patients, women, and patients with diabetes. All conventional and all atypical antipsychotics have been reported in association with tardive dyskinesia, but the risk introduced by atypical antipsychotics appears to be lower than the risk with conventional agents. As regards tardive dyskinesia, the use of atypical antipsychotics is associated with an incidence of 1% per year compared with 5% for conventional drugs.70

Peripheral cholinergic receptor blockade result in tachycardia, dry mouth, blurred vision, exacerbation of glaucoma, constipation, and urinary retention, any of which could be troublesome in the elderly patient. Central muscarinic receptor blockade cause drowsiness, irritability, disorientation, impaired memory and other cognitive dysfunction, aggressiveness, and delirium. The patients with Alzheimer's disease, dementia with Lewy bodies characterized by cholinergic hypofunction are likely to be particularly affected. If these symptoms are misdiagnosed as a new-onset or exacerbating psychosis, the dose of antipsychotic may be increased, resulting in a vicious worsening cycle. Among conventional antipsychotics, chlorpromazine, thioridazine, and mesoridazine are the most anticholinergic, and among atypical antipsychotics, clozapine and olanzapine are the most anticholinergic. Although in younger patients the anticholinergic effects of olanzapine are considered negligible at usual therapeutic doses, this may not be the case in the elderly patients, particularly among those affected with dementia.71

Metabolic risk (weight gain, diabetes, obesity, dyslipidemia, metabolic syndrome) is another important adverse events of atypical antipsychotics.45 Weight gain is especially observed during the first 4 to 12 weeks of treatment following most of the atypical antipsychotic drug use.4564727374 After this phase, weight gain continues, but at a slower rate or even stabilizes, whereas it continues over a long period following clozapine and olanzapine.72 Weight gain might be the result of increased food intake, due to the possible drug interactions with neuronal dopamine, serotonin, and histamine1 receptors.68 Obesity might also be the result of 5-HT2 receptor blockade by atypical antipsychotic drugs.

A number of studies have reported that both conventional and atypical antipsychotic drugs may increase the plasma glucose level in patients with or without a prior history of diabetes mellitus.45757677 This especially occurs with regard to olanzapine and clozapine and in elderly patients for the previously mentioned reasons. Clozapine and olanzapine are associated with, respectively, a 1.4- and 1.3-fold higher risk of diabetes than conventional antipsychotics, whereas the relative risk appears to be slightly lower following quetiapine and risperidone.75 The suggested mechanisms other than weight gain have been postulated in elderly patients:45 histamine H1-receptor antagonism, serotonin 5-HT2C-receptor antagonism, reduced responsiveness of pancreatic cells, pancreatic cell damage and, as a consequence, reduced insulin secretion, and increased insulin resistance.

Some antipsychotics can also cause hypertriglyceridaemia.45787980 This is more evident following olanzapine and clozapine use in the elderly.81 In a retrospective chart reviewed comparative study, treatment with various antipsychotics resulted in high triglyceride serum levels in 56% of clozapine, 39% of olanzapine, and 21% of risperidone compared with none of haloperidol and 8% of fluphenazine-treated patients.80 As shown by Koro et al.,82 there was a 3-fold higher risk of hyperlipidemia related to conventional antipsychotics compared with a control population without antipsychotic exposure.

Conversely, Rondanelli et al.83 suggested that the treatment with low dose of atypical antipsychotics is not associated with weight gain or increased risk of diabetes mellitus or lipid metabolism abnormalities. The study involved only 36 patients affected with Alzheimer disease. A study on 95 patients affected with dementia receiving mostly olanzapine or risperidone showed no effect on any of the parameters of the metabolic syndrome based on the National Cholesterol Education ProgramY Adult Treatment Panel III criteria.84

The CATIE-AD showed a clear effect of all atypical antipsychotic drugs to increase weight gain and body mass index, with olanzapine having the greatest risk followed by risperidone. On the contrary, no effect was shown on glucose, total cholesterol, and triglyceride levels.38

A decline in WBC counts can be seen with typical antipsychotics, particularly low-potency phenothiazines. Agranulocytosis (absolute neutrophil count <500/mm³), a potentially fatal condition, is associated with clozapine45 and much rarer with other phenothiazines (chlorpromazine, fluphenazine, perphenazine, thioridazine, and trifluoperazine). Some cases of olanzapine- and risperidone-induced neutropenia and agranulocytosis have also been reported.85 In a study by Rettenbacher et al.85 neutropenia (neutrophil count G2000/HL) was found in the mixed group in 17.6% and in 11.8% of patients treated with clozapine during the first 6 months. There was no statistically significant difference between those groups with respect to the risk of developing neutropenia during the investigation period. However, there was no case of agranulocytosis, and neutropenia was transient in all patients. Eosinophilia occurred in some patients who developed neutropenia later on but had no significant predictive value. Another case of quetiapine-induced neutropenia was recently described.8687 The risk of agranulocytosis with clozapine and other medications is highest in the first few months of treatment, but it can occur at any time and is not a dose dependent development. The overall risk is less than 1% but may be higher in the elderly patients.

Pneumonia, deep venous thrombosis are recently found adverse effects. Age-related functional dependence and hematologic changes make these effects more apparent in elderly demented people. A population-based, nested case-control study by Trifirò et al.88 was conducted on a cohort of people older than 65 years who used an antipsychotic drug, who were registered in the Dutch Integrated Primary Care Information database from 1996 to 2006. The study showed that the elderly patients with taking either atypical or typical antipsychotic drugs are associated with risk for community-acquired pneumonia in a dose dependent manner. The risk is high when treatment is started, and rapidly disappears when treatment is stopped. In particular, 25% of the case patients died in 30 days, and their disease was considered fatal. Current use of either atypical (OR, 2.61; 95% CI, 1.48-4.61) or conventional (OR, 1.76; CI, 1.22-2.53) antipsychotic drugs was associated with a dose-dependent increase in the risk for pneumonia compared with past use of antipsychotic drugs. The incidence rates of pneumonia were similar during past use of atypical and typical antipsychotic drugs, which supports the notion that the risk difference between the 2 classes was minimal. Only atypical antipsychotic drugs were associated with an increase in the risk for fatal pneumonia (OR, 5.97; CI, 1.49-23.98).

Pneumonia risk might be due to the anticholinergic action and antihistaminic H1-receptor blocking effect of antipsychotic drugs (more evident for some of these antipsychotic drugs). In fact, the anticholinergic effects may be responsible for dryness of mouth, impaired oropharyngeal bolus transport, dysphagia, and, as a consequence, aspiration pneumonia.88 Furthermore, sedation as a result of H1-receptor blocking in the central nervous system is a well-known cause of swallowing problems, which could cause aspiration pneumonia. In fact, antipsychotic drugs with higher H1-receptor affinity, such as phenothiazines, show increased risk for pneumonia compared with butyrophenones.

Also the use of conventional antipsychotic drugs may be a risk factor for aspiration pneumonia, because of such extrapyramidal effects as akinesia, and the adverse effect is by far more evident in elderly demented people who can experience dysphagia, especially in the late stages of dementia.

There is an association between the use of antipsychotic drugs and the risk of venous thromboembolism (VTE).8990 In a large primary care population composed of 25532 eligible cases (15975 with deep vein thrombosis and 9557 with pulmonary embolism) and 89491 matched control subjects from a study population of 7267673, individuals who were prescribed antipsychotic drugs in the previous 24 months had a 32% greater risk of VTE than nonusers, despite adjustment for potential risk factors (OR, 1.32; 95% CI, 1.23-1.42).83 The increased risk was more noted among new users and those prescribed atypical antipsychotic drugs.

Recent studies also showed an increased risk of pulmonary embolism with antipsychotic treatment, varying with type of antipsychotic drug and dependent on dose. Clozapine was associated with the highest risk.91 Atypical antipsychotic agents appear to increase the risk of VTE. However, these events are rare, and in clinical practice, the absolute risk should be weighed against the effectiveness of these medications in the elderly population. On the other hand, another study showed no evidence of an increased risk of VTE in elderly patients using antipsychotic drugs.91 The high risk for VTE is probably linked to the increase in platelet aggregation and to changes in fibrinolysis and coagulation.7092 In particular, it is known that some conventional antipsychotics such as haloperidol, chlorpromazine, and fluphenazine increase platelet aggregation. Some atypicals such as risperidone, clozapine, and olanzapine can inhibit platelet 5-HT2A receptors, thus modulating receptor density and affinity, even if they do not uniformly potentiate platelet adhesion and aggregation.92

Atypical antipsychotics have a weak potential to cause hyperprolactinemia, except for risperidone and amisulpride. Plasma prolactin increase is usually dose related and is more common in women, but is independent of age.9394 As previously mentioned, hyperprolactinemia is one of the suggested mechanisms for cerebrovascular effects induced by antipsychotics.