Definition of Drug-Induced Cognitive Impairment in the Elderly

Donna M. Lisi, PharmD, BCPS, BCPP, CGP, FASCP


June 14, 2000

In This Article

Drugs Associated With Cognitive Impairment

Taking a thorough drug history is one of the first steps that should be performed when assessing an older patient with changes in cognitive function. This history should include prescription drugs, over-the-counter medications, illicit substances, alcohol use, herbs, vitamins, nutraceuticals, homeopathic products, and naturopathic remedies, including the use of home remedies as well as other forms of complementary or alternative medicine. In the elderly, an increased number of medications may have a greater negative impact on orientation and memory as opposed to concentration and judgment.[28] The more complex a drug regimen, the more difficult it may be to identify the specific drug(s) that may be causing cognitive impairment. It is important to note that in the elderly, drug-induced cognitive impairment may occur even in the presence of nontoxic or therapeutic levels of a drug. Further, there may be intraclass differences in the propensity to induce cognitive impairment.

Numerous drugs have been identified in The Medical Letter on Drugs and Therapeutics as causing a multitude of psychiatric symptoms, including hallucinations, fearfulness, insomnia, paranoia, depression, delusions, bizarre behavior, agitation, anxiety, panic attacks, manic symptoms, hypomania, depersonalization, psychosis, schizophrenic relapse, aggressiveness, nightmares, vivid dreams, excitement, disinhibition, rage, hostility, mutism, hypersexuality, suicidality, crying, hyperactivity, euphoria, dysphoria, lethargy, seizures, Tourette-like syndrome, obsessiveness, fear of imminent death, illusions, emotional lability, sensory distortions, impulsivity, and irritability, which can impact on mental capacity. Further, there are a number of medications that may be linked to causing cognitive impairment by inducing delirium, confusion, disorientation, memory loss, amnesia, stupor, coma, or encephalopathy. Among these drugs are: acyclovir, anticholinergics and atropine, anticonvulsants, tricyclic antidepressants, asparaginase, baclofen, barbiturates, benzodiazepines, beta-blockers, buspirone, caffeine, chlorambucil, chloroquine, clonidine, clozapine, cytarabine, digitalis glycosides, disulfiram, dronabinol, ganciclovir, histamine-2 antagonists, ifosfamide, interleukin-2, ketamine, levodopa, maprotiline, mefloquine, methyldopa, methylphenidate, metrizamide, metronidazole, pergolide, phenylpropanolamine, pilocarpine, propafenone, quinidine, salicylates, seligiline, sulfonamides, trazodone, and trimethoprim-sulfamethoxazole. Often these medications produce more than 1 type of psychiatric symptom.[29]

A simple mnemonic to help remember the drugs or drug classes that are associated with acute changes in mental status in the elderly is ACUTE CHANGE IN MS (Table 2).[30]

Many of these drugs have already been recognized as being potentially inappropriate for use in the elderly. In 1991, Beers and colleagues[31] published explicit criteria for determining inappropriate medication use in nursing home residents. These criteria were derived by expert consensus using the Delphi method. The risk-benefit profile of specific agents within various drug classes, including sedative-hypnotics, antidepressants, antipsychotics, antihypertensives, nonsteroidal anti-inflammatory agents, oral hypoglycemics, analgesics, dementia treatments, platelet inhibitors, H2-blockers, antibiotics, decongestants, iron, muscle relaxants, gastrointestinal antispasmodics, and antiemetics, were examined. Many of the drugs were cited because of potential CNS adverse effects.[31] This list was later updated in 1997 to include drug-disease combinations that may also be inappropriate for use by the elderly.[32] In 1999, the Health Care Financing Administration drafted new nursing facility survey procedures and interpretative guidelines based on these 2 articles. Under these new guidelines, which went into effect on July 1, 1999, drugs that were considered to have a high potential for severe CNS adverse outcomes were pentazocine, long-acting benzodiazepines, amitriptyline, doxepin, meprobamate, disopyramide, digoxin, methyldopa, chlorpropamide (if hypoglycemia results), gastrointestinal antispasmodic drugs, and barbiturates (Table 3).

Other drugs that were considered to be potentially inappropriate, but were thought to produce less severe adverse outcomes, were identified. Among the medications that may produce adverse CNS effects are indomethacin, reserpine, diphenhydramine, muscle relaxants, sedating antihistamines, and trimethobenzamide (which can cause extrapyramidal effects). Lastly, drugs were identified that may exacerbate insomnia. This list of medications included decongestants, theophylline, desipramine, selective serotonin reuptake inhibitors (SSRIs), methylphenidate, monoamine oxidase inhibitors, and beta-agonists.[33]


Both anesthetics and preoperative medications such as anticholinergic agents used to dry up secretions or sedative premedication (barbiturate or benzodiazepine) have been associated with the development of delirium postoperatively. Since it is thought that the residual effects of anesthetics on cognitive function may remain 48-72 hours after surgery, the choice of the anesthetic drug is important. Newer medications with shorter elimination half-lives may be preferred in the elderly.[16] The type of anesthesia (ie, general vs spinal) does not seem to affect the occurrence rate of postoperative delirium.[14]


Although sepsis is one of the main risk factors for delirium, antibiotics and anti-infective agents may also produce changes in mental status. Among the agents that have been associated with delirium are aminoglycosides (eg, gentamicin, tobramycin, streptomycin), penicillins, cephalosporins, sulfonamides, and fluoroquinolones (eg, ciprofloxacin, ofloxacin).[10,34] Inhibition of GABA may be involved in fluoroquinolone- and penicillin-induced delirium. Penicillin can induce psychosis and encephalopathy. Risk factors for drug-induced delirium include renal impairment, increased permeability of the blood-brain barrier, high antibiotic dosage, intrathecal or intravenous administration of antibiotics, prior psychiatric illness, severe medical illness, slow acetylator status, and advanced age. Overall, however, this class of drugs has a low risk of inducing cognitive changes.[18] Other anti-infectives that have been associated with drug-induced cognitive impairment are erythromycin, clarithromycin, ketoconazole, amphotericin B, isoniazid, rifampin, quinacrine, chloroquine, quinine, trimethoprim/sulfamethoxazole, amantadine, acyclovir, and zidovudine.[2,30] Trimethoprim-sulfamethoxazole can cause acute psychosis and a catatonic depressive-like reaction.[30]

Anticholinergic Agents

This class includes drugs with known anticholinergic properties such as the first-generation, sedating antihistamines (eg, diphenhydramine, hydroxyzine, chlorpheniramine, meclizine), antispasmodics (eg, belladonna, diphenoxylate, clinidium, dicyclomine, hyoscyamine), oxybutynin, trazodone, ipratropium bromide, tricyclic antidepressants (which are discussed separately under antidepressants), phenothiazines (eg, thioridazine, prochlorperazine, promethazine, chlorpromazine, fluphenazine), muscle relaxants (cyclobenzaprine, orphenadrine), mydriatics (atropine, homatropine, tropicamide), diphenoxylate/atropine, antiparkinsonian agents (eg, benztropine, trihexyphenidyl), and antiarrhythmics (eg, disopyramide, quinidine, procainamide). Further, other drugs which may have possible anticholinergic effects include codeine, colchicine, warfarin, digoxin, furosemide, haloperidol, isosorbide dinitrate, meperidine, nifedipine, cimetidine, ranitidine, prednisolone, quinidine, and theophylline.[10,35,36,37] Many drug classes starting with the prefix "anti" have anticholinergic properties (eg, antihistamines, antidepressants, antipsychotics, antispasmodics, antiparkinsonian drugs, and some antihypertensives) and may help alert the practitioner to drugs that may be a source of confusion in their patients.[38]

Anticholinergic agents have been causally linked to the development of memory impairment in healthy subjects. Memory impairment may be associated with basal forebrain cholinergic pathways, whereas changes in consciousness seen in delirium may be attributable to alterations in pontine cholinergic pathways projecting into the frontal cortex and brain stem. Acetylcholine is also involved with attention, the sleep-wake cycle, and other aspects of cognitive functioning.[8,13]

In a study that was published in 1983, approximately 60% of nursing home residents and 23% of ambulatory patients were receiving drugs with anticholinergic properties. In some cases, patients may have received 3 or more anticholinergic medications concurrently.[39]

Tune and others[36] examined the anticholinergic effects of drugs commonly prescribed for the elderly as a potential means for assessing risk of delirium (Table 4). Using a standard concentration of 10-8 M of 25 compounds and an anticholinergic radioreceptor assay, they assessed these substances against an internal standard of atropine. Atropine equivalents represented in nanograms per milliliter of equivalent amounts of atropine were compared to the test drug. Of the 25 drugs tested, 14 produced detectable anticholinergic effects with 10 of these 14 medications, resulting in anticholinergic levels that have been associated with significant deficits in memory and attention in normal elderly.

Medications that were not associated with anticholinergic effects in this study included hydrochlorothiazide, propranolol, salicylic acid, nitroglycerin, insulin, methyldopa, ibuprofen, diltiazem, atenolol, metoprolol, and timolol.[36]

In an earlier paper, Tune and colleagues[40] had found that postoperative cardiac surgery patients who had experienced delirium had high serum levels of anticholinergic drugs and that impairment in cortical function was related to this elevated level. This group later examined the cumulative anticholinergic effects of drug regimens among surgical intensive care unit patients.[41] They have since expanded their work to examine the anticholinergic effects of 48 commonly prescribed medications.[42]

Flacker and colleagues[35] analyzed the association of serum anticholinergic activity with delirium in medical patients aged 75 or older. Delirium was associated with a higher serum anticholinergic activity quintile. The number of symptoms of delirium were also associated with higher serum anticholinergic activity. Mach and colleagues[43] demonstrated the resolution of delirium in an elderly population upon discontinuation of medications, which resulted in a reduction of serum anticholinergic levels. Only 5 of 17 medications discontinued were known to have in vitro anticholinergic activity. Even topically administered anticholinergic ophthalmic preparations have been associated with the development of delirium.[44,45] Other investigators have reported the presence of high serum anticholinergic levels among patients who have not received a drug that blocks acetylcholine, which raises the possibility of an endogenous source of anticholinergic activity that may possibly increase during times of stress.[35,46] Among elderly nursing home residents, serum anticholinergic activity seems to increase during illness and declines upon recovery, regardless of medication changes.[47]

In the presence of central anticholinergic toxicity, the use of physostigmine (a 1- to 2-mg test dose) may rapidly improve mental status. However, this drug has many severe side effects, including increased secretions, bronchospasm, vomiting, aspiration, and bradycardia, so its routine use cannot be advocated in the elderly.[10] The value of acetylcholinesterase inhibitors such as donepezil in this setting is unclear. Often, removing the causative agent and offering supportive care may be sufficient.

In summary, the likelihood of developing delirium following ingestion of an anticholinergic is unpredictable and may depend on other concomitant medications that exert anticholinergic effects, baseline cognitive status, pharmacokinetic or pharmacodynamic effects, specific agent used, and the total anticholinergic burden.[18]

It should also be stated that despite all of this evidence, the association between anticholinergic drugs and the development of delirium is not universally accepted. Francis and coworkers [4] and Schor and colleagues[48] failed to demonstrate causality between the use of these agents and the development of delirium in elderly medical inpatient populations. Yet others have felt that the lack of association between delirium and anticholinergic drugs in epidemiologic studies is one of misclassification of drug effects rather than the inability of the anticholinergic effects of drugs not to produce delirium.[14]


All anticonvulsants can affect cognition, even in the presence of therapeutic drug levels. They may cause drug-induced delirium or dementia. These effects appear to be dose related. Further, repeated episodes of uncontrolled seizures can adversely affect cognition. Phenobarbital, primidone, and clonazepam have a greater negative impact on cognition than do valproic acid, carbamazepine, or phenytoin. The mental status changes of phenytoin, phenobarbital, and primidone may be attributable to interference with normal folate mechanism.[30] In the elderly, it is important to check both serum albumin and serum creatinine when dosing phenytoin, because both hypoalbuminemia and an elevated serum creatinine necessitate lowering the dose administered. Newer anticonvulsants may also have a lower risk of cognitive impairment.[1,18] The Neurotoxicity Scale has been developed to help assess the adverse effects of anticonvulsants on cognitive function. The revised version of the Neurotoxicity Scale consists of 24 questions. Among the domains tested are fatigue, slowing, memory, concentration, language, and motor coordination. Although this scale may be useful for identifying the presence or absence of an antiepileptic drug-induced side effect, it is unsuitable for determining the type or severity of this event because it produces a global or "all or none" evaluation of whether a person on an antiseizure medication is experiencing cognitive impairment. This scale is self administered by the patient. Further, it has been tested primarily in younger patients (average 34.1 years). Therefore, it may have limited utility in assessing the drug-induced cognitive impairment of an elderly person who is already confused or delirious or who may be on a complex medication regimen.[49] Use of monotherapy and maintenance of serum concentrations within the therapeutic range (for older agents with therapeutic drug monitoring available) may help to minimize cognitive changes.


It is important to note that in the elderly, depression may present as pseudodementia. Therefore, cognitive impairment can be induced by the disease process itself. However, tricyclic antidepressants are notorious for producing adverse CNS side effects such as delirium, disorientation, and memory impairment in the elderly owing to their highly anticholinergic properties. The most common and specific feature of tricyclic-induced cognitive impairment in the elderly is impaired short-term recall memory.[50] Other types of impairment include reduced reaction time, impaired retrieval from secondary memory, and impaired information processing.[1]

Confusion or agitation developed in approximately 5% of elderly depressed patients who received either amitriptyline or imipramine.[51] The former agent has been associated with impaired cognitive performance. Preskorn and Jerkovich[52] found that 6% of patients administered tricyclics experienced CNS toxicity. Tricyclic antidepressants can also induce a Creutzfeldt- Jakob-like dementia.[19]

The use of tricyclic antidepressants has fallen out of favor for use among patients in this age group because of their side-effect profile and the availability of newer, safer classes of antidepressants. However, if tricyclic antidepressants are to be used in the elderly, 2 agents have been preferred because of their more favorable risk-to-benefit ratio. These drugs are nortriptyline and desipramine. Kutcher and Shulman[53] describe the first case report of desipramine-induced delirium in an elderly woman with a subtherapeutic serum desipramine concentration. This 68-year-old woman had initially been started on 25 mg of desipramine. After 1 week her dose was increased to 50 mg. Within 3 days of the dosage increase, this woman started experiencing bouts of confusion, demonstrated inattentiveness and hypoalertness, and had disorganized speech. Her serum desipramine level, which was drawn 13 hours after her last dose, was 112 nmol/L (therapeutic range: 500-1000 nmol/L). The drug was discontinued and 3 days later, she was back at her baseline mental state.

In general, risk factors for drug-induced delirium are high tricyclic antidepressant plasma concentration, advanced age, and female gender.[18]

Trazodone, a nontricyclic antidepressant, is also associated with impaired cognition.[1] Confusion is one of the most common side effects of nefazodone, a compound structurally related to trazodone.[30]

Fortunately, newer medications that are devoid of anticholinergic properties, such as SSRIs and reversible inhibitors of monoamine oxidase ( not yet available in the United States) may actually improve cognitive function as witnessed by improved vigilance, attention, memory, and psychomotor performance in some studies. This effect may be unrelated to their antidepressant properties.[50] Yet, when these drugs are combined with other medications, caution may be advised.[54] Whereas the reversible monoamine oxidase inhibitors may have less effects on cognition, older monoamine oxidase inhibitors such as tranylcypromine have been associated with adverse CNS effects.[2] Fluoxetine has been associated with the development of acute organic brain syndrome.[55] Caution is also advocated in the face of antidepressant-induced electrolyte imbalances (eg, SSRI-induced hyponatremia). In the case of SSRIs, one also needs to be concerned about the development of serotonin syndrome, which is characterized by delirium, autonomic instability, hyperreflexia, ankle clonus, tremor, diarrhea, and rigidity.[9,18] Serotonin syndrome may occur when SSRIs are combined with tramadol.[30]

Antiparkinsonian Agents

Approximately 20% to 30% of patients with Parkinson's disease have a concomitant dementia.[1] As with patients with other neuropsychiatric conditions, Parkinson's patients may be especially prone to the development of drug-induced cognitive impairment. One of the drugs that is most often associated with changes in mental status is levodopa. About 5% of patients develop delirium from the use of this drug,[56,57] although cognitive symptoms may occur in up to 60% of patients.[30] Yet, not all mental status changes are delirium; patients may experience isolated hallucinations while maintaining a clear state of consciousness, and this would not be considered delirium. Early clues to possible worsening cognitive function may include abnormal dreaming and sleep disturbances.[30] If these signs occur, lowering the dose of medication may be helpful. A relative excess of dopamine has been proposed as a possible cause of delirium.[13] Risk factors for drug-induced confusion include increasing age, dementia, and high doses of antiparkinsonian drugs.[1] As mentioned earlier, anticholinergic drugs used in Parkinson's disease can cause cognitive impairment. If dementia is present, Parkinson's patients on anticholinergic agents may be more than twice as likely to develop delirium compared with nondemented Parkinson's patients.[58] Amantadine's adverse cognitive effects may be dose dependent. The dose needs to be reduced in the elderly because of decreased renal function. High-potency dopamine agonists, such as pergolide, may be associated with higher rates of delirium than levodopa, with altered mental function occurring in 11% to 33% of patients. Bromocriptine can induce mental status changes even when used in low doses. Drug-induced delirium is also common with selegiline. Psychiatric side effects to these medications may become more common as the disease progresses. If these medications were to be ranked by their potential to cause cognitive changes, anticholinergic Parkinson's drugs would have the highest propensity, whereas bromocriptine, levodopa, and selegiline would be associated with medium degree of risk.[18] If a patient develops drug-induced cognitive impairment while on multiple antiparkinsonian agents, it may be beneficial to slowly withdraw the anticholinergics, selegiline, and amantadine before removing other agents from the regimen.[1]


As with other psychoactive medications, the risk of developing drug-induced cognitive impairment may be dose related. However, age may also be a significant risk factor for the development of this condition. Many traditional antipsychotics possess anticholinergic properties (eg, thioridazine, chlorpromazine, trifluoperazine), which may partly explain the predisposition of this class of drugs to the development of delirium and accelerated cognitive decline. One of the newer atypicals, clozapine, is also highly anticholinergic. Other atypicals that are devoid of significant anticholinergic effects, such as risperidone, appear less likely to cause drug-induced delirium. Such drugs as thioridazine and chlorpromazine may have a medium potential to induce cognitive changes, whereas risperidone has a low risk of such an event. The possibility of neuroleptic malignant syndrome should also be ruled out in patients in whom delirium develops shortly after the administration of an antipsychotic. Neuroleptic malignant syndrome is characterized by delirium, fever, autonomic dysfunction, extrapyramidal syndrome, and recent history of antipsychotic use.[9,18] One flaw in some of the studies on delirium and major tranquilizer use is that they fail to distinguish whether antipsychotics were the cause of delirium or were used to treat the delirium.

Cardiac Medications/Antihypertensives

This category includes the antiarrhythmics (eg, digoxin, amiodarone, lidocaine, disopyramide, procainamide, quinidine, flecainide, mexiletine, propafenone, tocainide), dipyridamole, and antihypertensives such as beta-blockers (eg, propranolol), methyldopa, clonidine, reserpine, calcium channel blockers, and angiotensin-converting enzyme inhibitors (ACEIs).[5,10,18] It is important to keep in mind that hypertension itself is a risk factor for vascular dementia and that aggressive lowering of blood pressure may also have a deleterious effect on cognition. Uncontrolled blood pressure and plasma lipids may lead to vascular dementia.

Among the antihypertensives that historically have been associated with significant adverse CNS effects (both delirium and dementia) is methyldopa. This drug produces cognitive impairment and decreased visual motor performance.[4] Methyldopa acts like a false neurotransmitter being converted to alpha-methyl-noradrenaline. In general, centrally acting antihypertensives such as clonidine and guanabenz are associated with more adverse cognitive effects. Reserpine irreversibly damages noradrenergic storage granules, thereby inducing altered mental function.[19] Dipyridamole has been associated with decreased Mini-Mental Status Examination scores.[59] CNS effect may be the first and only manifestation of digoxin toxicity and may be even more common than cardiac effects.[60] Both delirium and dementia can be signs of digoxin toxicity.

Cognitive changes can occur even in the presence of therapeutic digoxin levels.[61] Amiodarone's long half-life may promote prolonged confusion. Diuretics can cause fluid and/or acid-base imbalances, which can result in confusion, especially in the postoperative patient. CNS toxicity is common with lidocaine. Beta-blockers can be associated with pseudodementia. The incidence of neuropsychiatric toxicity ranges from 1% to over 20%.[30] Although controversial, less lipophilic beta-blockers may be preferred over highly hydrophilic agents as a way to reduce possible CNS adverse effects. Topical beta-blockers used for glaucoma have also been associated with the development of delirium.[2]

For drugs such as ACEIs, calcium channel blockers, and amiodarone, drug-induced delirium may represent an idiosyncratic event. The risk of cognitive impairment remains low for such drugs as diuretics and ACEIs. Other drugs, including quinidine, digoxin, methyldopa, alpha-blockers, postganglionic blockers, and beta-blockers, may have a medium risk of inducing such changes.[2,18]

Chemotherapeutic Agents

Drugs, either alone or when combined with other treatment modalities in cancer in the presence of a compromised host, can cause adverse CNS effects. For example, cognitive impairment induced by methotrexate is enhanced when this drug is administered to a patient undergoing cranial radiation. Among the chemotherapeutic agents that have been identified as causing delirium are carmustine, vincristine, vinblastine, L-asparaginase, ifosfamide, intrathecal procarbazine, high-dose cytosine arabinoside, methotrexate, 5-fluorouracil, hexamethylmelamine, etoposide, nitrogen mustard, lomustine, dacarbazine, and cytarabine.[2,5] Adjunctive agents such as antiemetics, cyclosporin, biologic response modifiers (interferon, interleukins) and corticosteroids are causally related to the production of mental status changes. Interleukins (eg, IL-2) may produce drug-induced dementia by increasing the blood-brain barrier's permeability to neurotoxins; by activating inappropriate central neuropeptidergic systems that impair attention, registration and memory; or by a direct neurotoxic effect. Cyclosporin's adverse CNS effects may be attributable to similar mechanisms, as it inhibits IL-1 and IL-2.[19] The actual propensity for each drug to cause cognitive impairment is unclear because these medications are often used in combination as part of treatment protocols.[2,5]


One of the proposed theories of what causes delirium is increased CNS cortisol levels. Exogenously administered corticosteroids may produce a similar effect. Corticosteroids can induce both delirium and chronic cognitive impairment as well as psychosis. Use of high-dose steroids (> 80 mg/day of prednisone), long duration of use, or the abrupt discontinuation of these hormonal agents can induce mental status changes. Even brief exposure to high doses of steroids can reversibly affect neuronal activity in the hippocampus, the area of the brain associated with memory; with continued use, permanent injury occurs. Overall, there is a medium risk of cognitive-induced impairment secondary to this class of drugs.[18] In addition to high dose, female gender and concomitant neuropsychiatric disease are predisposing risk factors for drug-induced mental status changes.[30]

Herbal Products

There is a misconception among consumers that because a product is natural or herbal it is without toxicity. A recent report has linked the use of St. John's Wort to the development of mania.[62] In another report, 2 patients developed encephalopathy and neuropathy following the ingestion of a Chinese herbal broth that contained podophyllin.[63] Melatonin use may be associated with the development of confusion.[64,65,66] Most recently, the FDA has warned of the potential neurotoxic effects of GHB or gamma-butyrolactone, a substance whose uses include sleep induction, release of growth hormone, enhancement of sexual activity and athletic performance, relief of depression, and prolongation of life.[67]

H2 Antagonists

All histamine-2 (H2) receptor antagonists have been associated with acute CNS toxicity, including delirium.[18,68] The drug that has received the most attention as being associated with medication-induced delirium is cimetidine. Cimetidine is thought to possess anticholinergic properties. Whether or not this explains it, its association with the development of delirium is unclear. However, cimetidine-induced delirium has been reversed with the use of physostigmine.[1,69] Cantu and Korek[70] found that there was no difference among the H2-blockers in their propensity to cause CNS changes. Among hospitalized patients, about 1% to 2% develop drug-induced cognitive changes compared with 15% to 80% of intensive care unit patients.[18] Advanced age and impaired renal function may be risk factors for the drug-induced CNS changes. Nonetheless, the overall risk of H2-antagonist-induced cognitive impairment is low.

Hypoglycemic Agents

Insulin and oral hypoglycemic agents may cause both reversible and irreversible brain damage secondary to hypoglycemia, which may result in cognitive loss.[71]


Lithium may impair memory and psychomotor performance. It is also associated with the development of delirium. Lithium has a high risk of inducing cognitive impairment. It may induce a Creutzfeldt- Jakob-like dementia. Its ability to produce dementia may be related to its inhibition of protein kinase C, which results in interference of regulatory processes of neuronal growth and differentiation. Lithium's toxicity is potentiated by drugs such as thiazide diuretic and nonsteroidal anti-inflammatory agents, which interact with this drug to produce higher lithium levels.[1,2,19,72,73]

Narcotic Analgesics

It is important to recognize that untreated pain itself can cause delirium. However, narcotics can also induce this condition, especially among postoperative patients. Narcotics are among the primary causes of delirium in the postoperative patient. The risk of drug-induced delirium may be highest with meperidine. In one study, among individual narcotic agents studied, only meperidine was significantly associated with the development of delirium (odds ratio 2.7) among postoperative patients aged 50 or older.[46] Meperidine has long been recognized as a drug that should not be given to older persons because this age group undergoes an age-related decline in renal function, which allows for accumulation of normeperidine, a neurotoxic substance. The delirium induced by meperidine has been characterized by fluctuations in levels of awareness, confusion, disorientation, illusions, visual and auditory hallucinations, persecutory delusions, and seizures. Further, both meperidine and normeperidine have anticholinergic properties. This drug was originally developed as an antispasmodic alternative to atropine during the 1930s. Meperidine's toxicity may be more pronounced when this drug is combined with the enzyme inhibitor cimetidine or with other drugs possessing anticholinergic activity.[74] Francis and colleagues[4] and Schor and others[48] also found a correlation between the use of narcotics and the development of delirium. The route of administration (eg, intramuscular vs patient-controlled analgesia) may also influence the risk of developing drug-induced delirium. Epidural and intramuscular administration may be more problematic than patient-controlled analgesia.[1] Even tramadol has been associated with drug-induced confusion.[30]

Nonsteroidal Anti-inflammatory Agents (Including Salicylates)

Aspirin use may pose a problem in the elderly because older patients may not even consider this substance a medication. This age group is more prone to having pains and aches and is therefore more likely to use this drug. Delirium is the major manifestation of salicylate toxicity. Confusion can also occur at therapeutic doses. Acetaminophen, while safe in usual doses, may also cause cognitive impairment in an overdose situation. Drug-induced cognitive effects from nonsteroidal anti-inflammatory agents range from delirium with indomethacin (medium risk for cognitive changes) and sulindac to disturbances in memory and concentration with naproxen and ibuprofen (low risk for cognitive changes).[18] However, in light of recent data that nonsteroidal anti-inflammatory agents may be protective against the development of Alzheimer's disease, the role of these agents in inducing cognitive impairment needs to be clarified. It may be that high doses (not therapeutic doses) of nonsteroidal anti-inflammatory agents have an adverse effect on cognition.[1]

Over-the-Counter Products

The elderly consume a large amount of over-the-counter medications. These medications, which are often less expensive than prescription drugs, may be used by older adults in an attempt to save money and to help maintain their independence. However, these medications, especially cough/cold products, sleep aids, and antinausea agents, contain potent anticholinergic substances that can induce delirium in older persons. Oral decongestants such as phenylpropanolamine and pseudoephedrine can also cause delirium in the elderly. Mental status changes associated with the use of decongestants may occur with low doses and topical administration.[30]

Promotility Agents

Metoclopramide has been associated with the development of drug-induced delirium.[75] This drug crosses the blood-brain barrier and affects both dopaminergic and cholinergic systems. Cisapride, a newer promotility agent, may have fewer CNS effects; however, it is associated with very serious drug interactions, so caution is advised when using this agent.

Proton Pump Inhibitors

Omeprazole may be associated with neuropsychiatric adverse effects, especially in older patients and in patients with liver disease.[30,76]


This class of drugs includes benzodiazepines such as flurazepam and diazepam, barbiturates, meprobamate, chloral hydrate, and sedating antihistamines, which are found in over-the-counter sleep aids. Long-acting benzodiazepines, such as flurazepam, especially if used in high doses, are the most likely drugs to cause or exacerbate dementia. Shorter-acting drugs, such as diazepam or temazepam, have a medium risk of causing drug-induced cognitive impairment.[18] CNS toxicity is often dose dependent.

In one study, exposure to long-acting benzodiazepines was significantly associated with the development of delirium (odds ratio 3.0) among postoperative patients aged 50 or older.[46] Another study found that 11% of older patients admitted to a general hospital developed cognitive impairment following benzodiazepine use.[77] Benzodiazepines have been associated with impaired learning of verbal and visual information,[1] immediate and delayed memory, and psychomotor performance.[78] The psychomotor and cognitive impairment may be persistent with long-term use of benzodiazepines. Anterograde amnesia occurs more commonly with higher potency and shorter-acting benzodiazepines, thereby limiting the usefulness of these medications.[1]

Barbiturates can cause chronic cognitive impairment, which may mimic Alzheimer's disease. The sedation produced by sedative-hypnotics may lower the elderly person's threshold for developing drug-induced delirium or dementia.[18] Even newer agents such as zolpidem are associated with adverse cognitive effects similar to those seen with triazolam. Zolpidem produces memory impairment that corresponds to its peak blood concentration.[79]


Although theophylline may be associated with drug-induced cognitive impairment, it is unlikely to occur when this drug is used in usual doses.[1] Most adverse cognitive effects ("theophylline madness") occur in an overdose situation. If overdose occurs, one must be very watchful for seizures, which may soon develop if they are not present already.[30]

Urinary Antispasmodics

These drugs (eg, oxybutynin, flavoxate) induce delirium either via their anticholinergic effects or by causing urinary retention ("cystocerebral syndrome"). This latter condition is thought to be related to an increase in adrenergic tone, which leads to increased peripheral and CNS catecholamine levels. Risk factors for this condition include benign prostatic hypertrophy, dementia, and diabetes associated with autonomic dysfunction.[30]

Withdrawal Effects

Delirium associated with the withdrawal of centrally active psychotropics such as benzodiazepines, barbiturates, or alcohol may be attributable to understimulation of the inhibitory neurotransmitter GABA, which leads to symptoms of hyperactivity.[13] In the surgical patient, withdrawal from alcohol resulting in delirium may not manifest until 12-48 hours after surgery.[16] In the elderly, mortality associated with alcohol withdrawal-induced delirium tremens may be as high as 27%.[80] It is important to keep in mind that although the discontinuation of anticholinergic drugs is encouraged, rapid withdrawal of these agents may result in cholinergic rebound. This has been noted with clozapine, among other drugs.[81]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.