Alzheimer's disease, a neurodegenerative brain disease, is the most common cause of dementia. It currently afflicts about 4 million Americans and is the fourth leading cause of death in the United States. Alzheimer's disease is the leading cause of mental impairment in elderly people and accounts for a large proportion of admissions to assisted living homes, nursing homes, and other long-term care facilities.
Many Alzheimer's disease patients have concomitant psychiatric symptoms during the course of illness. Alois Alzheimer, in his original report,1 described a 51-year-old woman with memory impairment together with suspiciousness, paranoid delusions, and auditory hallucinations. This patient showed pathologic jealousy of her husband, believed her doctor was trying to stab her, thought others were trying to kill her, and appeared to experience auditory hallucinations.
Symptoms such as these have been known to occur in Alzheimer's disease for a number of years. However, they have received much less attention than the cognitive impairment of Alzheimer's disease. Green et al.2 reported that it is not the cognitive impairment or the functional incapacity of the demented patient but the behavioral changes, such as delusions, that cause the greatest stress to the patients and their caregivers. Burns et al.3 in 1990 reported that delusional symptoms in Alzheimer's disease are important because 1) they place strain on caregivers, 2) they may indicate subtypes of Alzheimer's disease, and 3) neuropathological correlations in Alzheimer's disease patients with and without delusions may shed light on the neuropathology of the disease and its symptoms. To this we would add another reason why delusions are important in Alzheimer's disease: they are often treatable.
For our literature review we performed a computerized medline search, extending from 1966 to 1997, for English-language citations on Alzheimer's disease and delusions. We obtained 66 articles. We used the bibliographies of these papers to identify other articles, which were also reviewed. We then focused on articles specifically reporting on Alzheimer's disease and delusions. However, articles discussing Alzheimer's disease and psychosis were also included.
This article synthesizes current knowledge of the association between Alzheimer's disease and delusions, guided by the literature review. We first define important terms, then discuss the clinical epidemiology and classification of delusions in Alzheimer's disease; the next section presents our approach to thinking about the etiology of delusions in Alzheimer's disease; and finally we discuss treatment and prognosis. Our main goal was to provide a succinct overview of delusions in Alzheimer's disease. The structure of this review might also serve as a method to approach the other behavioral disturbances of Alzheimer's disease.
According to DSM-IV,4 a delusion is a false belief, based on incorrect inference about an external reality, that is firmly sustained despite what almost everyone believes and despite evidence constituting incontrovertible and obvious proof to the contrary. The belief is not ordinarily accepted by other members of the person's culture or subculture. Simply stated, delusions are fixed false and idiosyncratic beliefs. Delusions must be differentiated from disorientation, confabulation, overvalued ideas, and misidentification.
Disorientation is a disturbance in which there is confusion about time, place, or person. Any structural or toxic brain disturbance may cause disorientation. Disorientation to time is often the first to occur, and disorientation to self is usually seen only in advanced dementia such as Alzheimer's disease. Disorientation is different from delusions in that disoriented people may present incorrect responses to questions, but the inaccuracies are not consistent and are not firmly held despite evidence to the contrary.
Confabulation is characterized by untrue responses to questions, but without a deliberate intention to mislead. They are the unconscious filling of gaps in memory by imagined or untrue experiences that patients believe, but that often have no basis of fact. Confabulated responses may also be drawn from a patient's actual experience but produced out of context. For example, when asked about present occupation, a patient may give a detailed but untrue description of previously held jobs. Confabulation is prominent in certain amnestic syndromes such as Wernicke-Korsakoff syndrome and disorders of the mamillary bodies, thalamus, and frontal lobes.
Overvalued ideas are false beliefs that are neither firmly sustained nor idiosyncratic. Misidentifications are also false beliefs, but they are transient and are neither fixed nor firmly held. Examples of misidentification are believing that people and situations on TV are real or believing that the mirror image of a person is a different person.
Estimates of the prevalence of delusions in Alzheimer's disease have ranged from 10% to 73%.5 These estimates vary because of the different ways by which delusions and dementia were diagnosed in different studies. They also vary by study population. Studies of patients from general hospital clinics give higher values compared with autopsy-proven cases of Alzheimer's disease. The methods of diagnosis of delusions have also varied; some studies depended on direct interview with the patients,6,7 others on telephone interviews with caregivers,3 and yet others on face-to-face interviews with caregivers.3 In our opinion, the best estimate from a community population comes from Burns et al.3 in a study of 178 patients with Alzheimer's disease who were drawn from one catchment area and satisfied the NINCDS/ADRDA criteria for clinical diagnosis of Alzheimer's disease. In this study, the prevalence of delusions over the course of the dementing illness was 16%. The prevalence in the past 12 months was 11%.
There is little information available about the incidence of delusions in Alzheimer's disease. Only the study by Flynn et al.,8 in 1991, used systematized criteria for the diagnosis of delusions and conducted sequential interviews. They estimated an annual incidence of 2.4%.
Cummings9 proposed four types of delusions in his study of 20 patients "with organic delusions." They are 1) simple persecutory delusions, 2) complex persecutory delusions, 3) grandiose delusions, and 4) delusions associated with specific neurological defects—nosoagnosia and reduplicative paramnesia.
Simple persecutory delusions consist of elementary, loosely structured beliefs that are usually transient, such as believing that possessions or money are being stolen or that one's spouse is unfaithful. Cummings9 found that simple delusions were often encountered in patients with Alzheimer's disease and vascular dementia.
Complex delusions have a more rigid and stable structure, and they are associated with substantial, though distorted, observation. Schneider's first-rank symptoms of schizophrenia and monosymptomatic delusions such as the Capgras syndrome are included in this category. Examples of complex delusions include patients' beliefs that the radio or TV is sending out messages to them, that the street lights are being focused on them, that others around them are talking about them, or that people are plotting to steal their homes. Patients with Capgras syndrome have the false belief that significant people have been replaced by identical-appearing impostors. Two other syndromes that share features with the Capgras syndrome are the Fregoli syndrome, where patients identify their persecutor in several different persons (the persecutor being accused of changing faces like an actor), and the intermetamorphosis syndrome, where patients believe that people have taken on the physical appearance of others. Cummings9 found that complex delusions occur in patients with hepatic encephalopathy, Parkinson's disease, posttraumatic encephalopathy, and idiopathic calcification of the basal ganglia. Grandiose delusions were seen only in one patient with Huntington's disease.
Delusions associated with specific neurological conditions include reduplicative paramnesia—a state in which patients believe that they are simultaneously in two geographical locations. It is usually seen during recovery from acute cerebral lesions, as in posttraumatic encephalopathy or cerebrovascular disease. Reduplicative paramnesia has been associated with right parietal or combined right parietal and bilateral frontal dysfunction.9 Other delusions with neurological defects include denial of blindness, as in Anton syndrome secondary to posterior bilateral cerebral artery occlusion, and denial of hemiparesis (neglect syndrome), seen with right hemispheric infarction.
Cummings9 observed that the complexity of delusions is related to the severity of intellectual impairment. He commented that the intellect is in the service of delusions. Patients who are moderately to severely cognitively impaired display simple delusions secondary to their poor judgment and reasoning capacity. However, patients who are much less cognitively impaired have greater ability to think and reason and thus have complex, elaborate delusions.
Cutting10 used Cummings's definitions as a basis for classifying delusions in 35 patients with "organic psychosis." He found 8 patients with simple persecutory delusions and 9 with mood-congruent delusions that were equivalent to the grandiose delusions defined by Cummings. The other 18 had delusions that did not fit into Cummings's classification, and Cutting thus called them complex, bizarre, or multiple delusions.
Burns et al.,3 in a study of 178 Alzheimer's patients, reported that delusions of theft were the most common, followed by delusions of suspicion—for example, believing that they were being watched or that their spouse was being unfaithful. These delusions were classified as simple persecutory delusions. Six subjects had other delusions, such as a conviction that a pet had been burnt; the idea that the patient had become a father at age 90; the belief that the patient was in Korea fighting a war; the belief that people came in at night and slit the patient's throat or added objects to the patient's house; and the belief that the patient was getting messages from a tape machine. Since these patients could not be grouped according to Cummings's classification, Burns et al. used Cutting's classification and thus divided delusions of Alzheimer's disease into two categories: 1) simple persecutory delusions and 2) complex, bizarre, or multiple delusions.
In a study by Rubin11 of 110 patients with Alzheimer's disease, 31% had paranoid delusions. The most common were delusions of theft in 28%. Delusion of suspicion was seen in 9%, more complex and systematized delusions were seen in 3.6%, and 1 patient developed an erotomanic-like delusional symptom. Rubin concluded from his study that most delusions in Alzheimer's disease are uncomplicated; however, systematized complex delusions can occur.
Deutsch et al.12 found that of 170 patients with probable Alzheimer's disease, 45% had delusions. Delusions of persecution were the most common. Seventy-three percent of the delusional patients had delusions of reference, jealousy, grandiose delusions, and somatic delusions, in descending order of frequency. The next most common was "other delusions," seen in 28%. Examples of these included the belief that people in the house were conducting drug sales and a patient's belief that he was in his seventies when he was in his eighties, which led him to assert that he needed to go to the bureau of vital statistics to change his age.
In summary, delusions in Alzheimer's disease are most commonly delusions of theft or suspicion that can be classified as simple persecutory delusions. The other delusions can be grouped as complex, bizarre, or multiple delusions.
Risk factors for the development of delusions in Alzheimer's disease may be studied in relation to 1) sociodemographic characteristics, 2) past history, 3) genetic makeup, 4) cognitive impairment, 5) functional impairment, 6) disease stage, 7) neurological signs and symptoms, and 8) other behavioral disturbances.
Cummings9 in his study of 20 patients with "organic psychosis" proposed a number of predisposing factors for the development of delusions in Alzheimer's disease. He reported that genetic constitution, early life experiences, exact location and extent of the brain lesion, personality characteristics, age at onset of dementia, rate of functional disruption, and cultural background might contribute to the development and elaboration of delusions. These factors might also influence the idiosyncratic delusional content unique to each patient and explain why delusions are not a uniform feature of an otherwise widespread central nervous system disorder.
Associations between delusions in Alzheimer's disease and gender, education, or race have been reported, although not replicated or extensively studied. Rockwell et al.13 found a significant association between female gender and the presence of delusions. In contrast, Burns et al.3 described a significant excess of delusions in males—35%, compared with 11% in females. Flynn et al.8 and Bullard et al.14 reported an association between lower levels of education and delusions. Deutsch et al.12 reported greater prevalence of delusions in African Americans—60%, compared with 40% in whites. No study has reported associations between delusions and socioeconomic status, marital status, duration of dementia, or age at onset of illness.
Kotrla et al.15 in their study of 57 patients with Alzheimer's disease found that the "psychotic" and "nonpsychotic" groups did not differ in their personal history of depression, alcohol dependence, family history of depression or schizophrenia, and personal or family history of mania. Bullard et al.14 reported that deafness and life events were associated with delusions.
The role of apolipoprotein E (APOE) gene in the development of psychiatric symptoms was studied by Lyketsos et al.16 They did not find any association between the APOE allele type and the presence of delusions.
Several studies have addressed the relationship between cognition and psychotic features in Alzheimer's disease patients. Only a few have studied the association between cognition and delusions, specifically. The results appear contradictory. Binnetti et al.,17 Jeste et al.,18 Drevets and Rubin,19 Rosen and Zubenko,20 and Stern et al.21 reported that psychosis is associated with more severe cognitive impairment. Conversely, studies by Kotrla et al.15 and Bylsma et al.22 showed no relationship between severity of cognitive impairment and psychosis. Additionally, Deutsch et al.,12 Becker et al.,23 and Burns et al.3 found no association between delusions and severity of cognitive impairment. More studies are needed to specifically address the relationship of delusions and cognitive impairment in Alzheimer's disease.
Almost all studies available have reported that "psychosis" is associated with greater functional impairment in patients with dementia or Alzheimer's disease.14,15,18—21,23,24
Alzheimer's Disease Stage
Alzheimer's disease is typically divided into three stages, based on severity of cognitive impairment: the early, middle, and late stage.25 Delusions are present at all stages of Alzheimer's disease. They have been associated with early disease,26 late disease,27 and middle-stage disease.19 In a longitudinal study11 of 110 patients with Alzheimer's disease, delusions were present during all stages of illness. These patients were divided by early, middle, and late-stage dementia according to the Clinical Dementia Rating (CDR) scale developed at Washington University.28,29 Psychotic symptoms were most active during the moderate stage of dementia. Thirty-five percent of patients who were followed longitudinally from the mild to the moderate stage of illness displayed delusions. However, when a group was recruited for a cross-sectional study during the moderate stage (with CDR 2), 52% were found to have delusions. The authors noted that this increase over time is probably secondary to an increased interest in participating in the study on the part of the families of moderately demented patients with psychosis.
Neurological Signs and Symptoms
Mayeux et al.30 reported that Alzheimer's disease patients with extrapyramidal signs (EPS) had psychotic symptoms twice as frequently as patients without EPS—50% versus 26%. When drug-induced EPS were considered separately, the frequency of psychotic symptoms was three times higher in those with EPS than in those without. A similar pattern was observed for myoclonus:30 45% of patients with myoclonus, versus 37% of patients without myoclonus, had psychotic symptoms.
Other Behavioral Disturbances
Deutsch et al.12 reported that delusions are associated with physical aggression in Alzheimer's disease. In fact, delusions frequently preceded episodes of aggression; of 169 patients, 30% were found to be aggressive, and 60% of the aggressive patients were delusional.
Cummings and Victoroff31 outlined four general explanations for the presence of delusions in Alzheimer's disease. Delusions may be the result of a logical attempt by patients to understand their environment; they may be secondary to mood changes; they may be a coincidental diagnosis with causes unrelated to the dementing disease; or they may be directly caused by underlying brain damage from the dementing disease.
Berrios32 proposed four reasons for the occurrence of "psychiatric phenomena" in conjunction with a dementia syndrome: 1) they may be due to an intercurrent confusional state; 2) they may be due to disinhibition of cortical functions resulting in "released" symptomatology; 3) they may be due to the pathoplastic effect of personality in the clinical expression of the disorder; and 4) they may represent a separate mental disorder that coexists with the dementia syndrome.
We propose that the etiology of delusions in Alzheimer's disease is best considered by using the "four perspectives of psychiatry" as described by McHugh and Slavney.33 These are four methods or types of logic that can be used to explain the cause of mental disturbances. The four perspectives are 1) the disease perspective; 2) the dimensional perspective; 3) the behavior perspective; and 4) the life-story perspective.
The disease perspective explains mental phenomena (in this case delusions) as being due to an abnormality in the brain (a "broken part") that causes the mental disturbance. The dimensional perspective sees delusions as arising out of the interaction between a person's vulnerability (such as the cognitive impairment or the patient's personality) and the particular current environment to which it is mismatched. The behavior perspective emphasizes the goal-directed aspects of behavior and attempts to show how behavior is gradually shaped by environmental consequences. The life-story perspective rests on the logic of narrative, in which a distressed state of mind is the result of a disturbing life experience. The etiology of delusions in Alzheimer's disease can be explored from any of these perspectives.
Delusions may reflect dysfunction of the limbic system. Cummings34 in his neurobiological model for delusions in neurological diseases reported that the two regions in the brain associated with delusions are the caudate nucleus and the temporal lobes, both portions of the limbic system. They are connected and innervated by the dopaminergic projections from brainstem nuclei. The cortical structures of the limbic system include the subcallosal gyri, cingulate gyri, parahippocampal gyri, hippocampus, posterior orbitofrontal cortex, and anterior insular region. The nuclear structures of the limbic system include the amygdaloid complexes, septal nuclei, hypothalamus, epithalamus, and anterior thalamic nuclei.34
The limbic system augments arousal and mediates environmental surveillance. Cummings34 reported that dysfunction of the limbic system interferes with assessment of environmental threat, produces incorrect assessment of danger, and causes inappropriate fear and threatened behavior. This is often manifested as paranoia. The individual may suspect the intentions of others and perceive menace and threat where none exist, resulting in persecutory ideation and inappropriate fear. Thus, there is an abnormal linkage between emotional content, current input, and past memories.
At a neurobiochemical level, the dopaminergic/cholinergic balance may be crucial in the genesis of delusions. One study35 has shown that psychosis is rare in untreated Parkinson's or progressive supranuclear palsy, both of which are characterized by marked dopamine deficiency. On the other hand, patients treated with dopaminergic agents and dopamine-treated Parkinson's patients have increased rates of psychosis.
In Alzheimer's disease, dopaminergic function is relatively preserved, and there is deficiency of choline acetyl transferase with resultant deficiency of acetylcholine.36 The severe deficiency of cholinergic function creates a marked imbalance between the two transmitters, causing a relative hyperdopaminergic state that may lead to the genesis of delusions. Cholinergic deficiency is most severe in the limbic regions, including the postcingulate gyrus, fusiform gyrus, and anterior and inferior temporal regions.36
In addition to Cummings's34 hypothesis and the hyperdopaminergic hypothesis, autopsy studies, computed tomographic (CT) studies, electroencephalographic (EEG) findings, and single-proton emission computed tomographic (SPECT) scan studies are available showing distinct abnormalities in Alzheimer's disease patients with delusions as compared with those without delusions.
Autopsy studies of 27 cases of Alzheimer's disease with "psychosis" have shown that there is significant increase in density of senile plaques and neurofibrillary tangles in the prosubiculum (allocortex), middle frontal cortex (neocortex), entorhinocortex, and superior temporal cortex compared with patients who did not have psychosis.37 The increased density of senile plaques and neurofibrillary tangles in the cerebral cortex of patients with psychosis may explain the association with more rapid cognitive decline.19—21
"Psychosis" is also associated with a greater level of norepinephrine in the substantia nigra, with similar trends in the caudate nucleus, thalamus, amygdala, and prosubiculum. These results suggest that a minimum threshold of norepinephrine functioning is required for the expression of psychosis.37
CT studies in Alzheimer's disease patients with delusions have shown that patients with systematized delusions have smaller lateral ventricles and more extensive calcification of the basal ganglia than patients without delusions.38 The intensity of delusions in dementia is inversely related to cortical atrophy. Thus, patients who are moderately to severely demented and have significant cortical atrophy often display simple delusions.
Visual and spectral EEG studies39 have shown more severe abnormalities in Alzheimer's disease patients with delusions and hallucinations than in those without. Visual EEG analysis39 showed that delusional patients had significantly greater amounts of moderately abnormal EEG. Spectrum analysis confirmed the increased amount of delta and theta activity.37
Starkstein et al.40 compared SPECT scans of 16 Alzheimer's disease patients with delusions and 29 without delusions. The two groups were comparable in age, years of education, duration of illness, and severity of dementia. Alzheimer's disease patients with delusions had significantly lower cerebral blood flow in both left and right temporal lobes as compared with the patients with no delusions. Deficits were similar in both the superior and the inferior temporal regions in patients with delusions. The study supported the suggestion by Cummings that damage to the temporal lobe may underlie the presence of delusions in Alzheimer's disease.
Thus, several studies of the brain have reported limbic system dysfunction in Alzheimer's disease patients with delusions. This finding indicates that delusions in Alzheimer's disease are best thought of as a result of a "broken part" in the brain, a consequence of disease progression to specific regions. However, this approach only partly illuminates the genesis of delusions, leaving open many questions about their timing, mechanism, and specific content.
The Dimensional Perspective
One aspect of the dimensional perspective focuses on cognition and emphasizes that those who have impairments in cognition are more vulnerable to environmental stressors and may react to them in extreme or unusual ways. As Alzheimer's patients experience cognitive deterioration they may become unable to draw sensible conclusions from their surroundings and may develop a disturbed sense of the world around them, much as mentally retarded adults may become suspicious and misinterpret the surrounding world because they cannot understand it well.
The dimensional perspective also looks at the effect of personality vulnerabilities on behavior. The relevant question is whether certain personality vulnerabilities make the Alzheimer's patient more vulnerable to the development of delusions. Two common affective dimensions of personality, extroversion—introversion and stability—instability, have been associated with emotional reactions to different types of stress.33 Individuals with different personalities may react differently to various stressors associated with the disease. Those with more introverted temperaments33 who function best in structured, organized, and predictable settings may be more likely to develop delusions when they are faced with the poor organizational skills that result from cognitive deterioration. They may begin to suspect that others are stealing from them or are trying to hurt them. In contrast, patients with more extroverted temperaments33 may be less likely to develop delusions, since they are better able to adapt to uncertainty and lack of structure. Personality may also be more likely to influence the content of delusions. So far no published data are available about how premorbid temperament influences the development of delusions.
Delusions might be the result of the shaping of a person's behavior over time based on the consequences of his or her actions. Goal-directed behaviors such as eating, drinking, sleeping, and sexuality all depend on learning and maturation. These behaviors may lead to a disorder when carried to excess. Many of these goal-directed behaviors are maintained by rewarding consequences. The explanation of delusions as "comfort" phenomena by Ballard and Oyebode41 or as "security" phenomena by Miesen42 may be explained by this perspective. These authors suggest that some of the psychotic symptoms involving family members may be comfort phenomena, giving the patient a sense of security, such as when patients mistake their sister for their deceased mother or believe that the doctor or nurse is actually related to them. The positive feedback that patients receive as a result of such false beliefs might reinforce the thought and eventually lead to action driven by the thought.
Delusions might also be shaped and sustained by other positive reinforcements derived from a particular behavior. A patient may notice that blaming a spouse for infidelity makes the spouse more attentive and caring or that accusing someone of stealing may lead to things being searched for and given to the patient as soon as requested. However, we can only explain some delusions in this way on the basis of their content. Other delusions, such as a patient's belief that a pet has been killed or that the patient is fighting in a world war, are less satisfactorily explained by this perspective.
The Life-Story Perspective
The life-story perspective takes into account the numerous events that have taken place in a person's life and tries to explain the person's emotional state on the basis of a narrative that incorporates past and present experiences. In this perspective, mental states are understood as arising out of a person's current predicament, and explanations of them depend on the ability to empathize with this predicament, to feel "what it would be like" to have Alzheimer's disease.
Rubin's11 hypothesis is an example of narrative reasoning. He suggested that delusions may be a defense mechanism exhibited by dementia patients when confronted with their limitations. Normal people tend to become suspicious when things are missing; they might suspect the items have been stolen, but because their insight and reasoning ability are intact, they are capable of realizing that this is not true and will search for the items until they find them. Dementia patients, however, have poor insight and are cognitively limited, so their reasoning ability is compromised and they tend to believe that missing items are stolen.
Lopez et al.39 have suggested that paranoid delusions in dementia may be an adaptive response resulting from progressive cognitive deterioration and the patient's inability to grasp reality and interpret it appropriately. Devanand et al.43 suggested that many types of delusions may be considered as types of confabulation, filling in the gaps as a result of memory impairment.
To summarize, a review based on these four perspectives suggests that limbic system dysfunction may initiate the formation of false beliefs, but a number of mediating factors—such as culture, life experiences, behavior, temperament, and cognitive function—influence the content, complexity, and timing of delusional beliefs.
Neuroleptics are effective for the treatment of delusions and hallucinations in a variety of psychotic disorders. On this basis, the presumption has been made that neuroleptics are effective for the treatment of delusions in dementia. Neuroleptics have been studied more widely than any other class of agents in the treatment of dementia. However, only a few studies, such as those by Steele et al.,44 Gotleib et al.,45 and Tune et al.,46 have been restricted to Alzheimer's disease. Most of these studies relate to the treatment of "agitated" behavior and not specifically delusions. To date there has been no published, randomized, double-blind placebo-controlled study of neuroleptics to treat only delusions in Alzheimer's disease. Most of the studies available have been done on inpatients and nursing home patients, and very little literature exists on the study of outpatients in this respect.
In a 4-week trial of psychiatric inpatients, thiothixene at a dose less than 15 mg per day showed marginal superiority over placebo in the treatment of 42 "agitated" patients with organic brain syndrome.47 In two other studies48,49 of patients with dementia, global clinical improvement was significantly greater with haloperidol, loxapine, or thioridazine compared with placebo, but patients on neuroleptics developed more side effects. The flaws in these studies were diagnostic heterogeneity, nonstratified assignment of psychotic patients to treatment, short trial duration, and high placebo rate of response. Devanand and Levy,50 in a study of 9 patients with Alzheimer's disease "with behavioral complications,", found that haloperidol in the dose of 1 to 5 mg was most efficacious, but some patients could not tolerate these doses because of the development of EPS. Also, it was found that haloperidol treatment was associated with small but significant cognitive decline on the Mini-Mental State Examination.
Two studies46,48 have shown that the most common side effects of neuroleptics in dementia patients include EPS and sedation. Other common side effects include orthostatic hypotension, confusion, constipation, and urinary retention. The rare side effects include neuroleptic malignant syndrome, hepatotoxicity, retinopathy, arrhythmias, and agranulocytosis.50 Poor compliance with medication and interactions with other drugs are additional complications noted in different studies. There are two studies51,52 that compared low-potency neuroleptics with high-potency neuroleptics. They did not find any significant difference in efficacy. In one open study45 of 10 Alzheimer's disease patients, depot neuroleptic treatment with low-dose fluphenazine decanoate (intramuscular, 1.25 to 3.75 mg semimonthly) was efficacious, with few side effects observed, during a 16-week trial.
We could find no published double-blind placebo-controlled or otherwise controlled studies of clozapine, risperidone, or olanzapine in the treatment of demented patients with behavioral complications. However, recent reports at international meetings have presented large, multicenter, placebo-controlled studies of olanzapine53 and risperidone54 in the treatment of agitation and psychosis in Alzheimer's patients, suggesting efficacy clearly superior to placebo with relatively fewer adverse events. Prado et al.55 conducted an open clinical trial with risperidone 0.5 to 4 mg daily in 18 demented inpatients with agitation and found it to be efficacious. Two patients discontinued the medication because of orthostatic hypotension. In an another study, Goldberg and Goldberg56 used 0.25 to 5 mg of risperidone in 64 nursing home patients with dementia; 41% showed marked improvement and 26% showed moderate improvement.
Benzodiazepines have also been used in the treatment of "agitated behavior" in demented patients. Burgio et al.57 showed no difference with treatment of haloperidol (0.5—3 mg) or oxazepam (10—30 mg) in 21 patients with behavior complications. However, the disadvantage of benzodiazepines was the development of tolerance, dependence, or deterioration in cognition.
Several other medications have been reported to reduce "agitation" in dementia. These include trazodone, buspirone, selective serotonin reuptake inhibitors, lithium, carbamazepine, divalproex sodium, and beta-blockers. Their efficacy specifically for delusions in Alzheimer's disease is unknown because double-blind placebo-controlled studies with these medications are not available.
Cummings et al.58 compared the antidelusional efficacy of physostigmine, an acetylcholine esterase inhibitor, with that of haloperidol in 2 patients. Physostigmine ameliorated the delusions and produced no side effects. However, it was difficult to maintain steady blood levels of physostigmine because it had to be given every 2 hours 6 times a day. It is possible that the newer cholinesterase inhibitors donepezil and tacrine have activity for delusion in Alzheimer's. Metrifonate, a novel cholinesterase inhibitor awaiting FDA approval, has shown efficacy for delusions relative to placebo in a recent randomized controlled trial.59
One of the greatest challenges in caring for delusional patients with Alzheimer's disease is continuing to provide respect, understanding, and objectivity while managing anger and other emotions evoked in the caregiver. The caregiver needs to be educated and trained about providing individualized yet flexible care. Videotapes, publications, and support groups can help caregivers learn more about delusions. Even though it is difficult to educate Alzheimer's disease patients about behavior modification strategies because of their poor learning capacity, "disruptive behaviors" can be reduced if patients, physicians, and caregivers are able to meet together in a reality-orienting or problem-solving group. Other nonpharmacological interventions that may be effective include changing the surrounding environment, reducing noise level, brief one-to-one contact by staff for reassurance, and family education. Whether these strategies specifically affect delusions is unknown.
A study by Devanand et al.60 has reported that "psychotic" symptoms in Alzheimer's disease show moderate persistence over time. Other studies11,19 have shown that even though Alzheimer's disease patients with delusions deteriorate more rapidly both functionally and in cognitive impairment, they do not die more rapidly. In fact, two studies have reported that there is a trend for lower mortality.11,18 The possible explanations for this range from psychosocial to neurochemical causes. It may be that delusional patients receive more care and support from their families, or they may be admitted to a hospital or nursing home much earlier and this may contribute to their decreased mortality. Looking at it from a different angle, patients with Alzheimer's disease and delusions have relative preservation of norepinephrine in the substantia nigra, thalamus, amygdala, caudate, and prosubiculum,37,61 despite greater degeneration in the limbic system. This selective preservation of brainstem nuclei that control vital centers such as deglutition, respiration, and cardiovascular homeostasis may lead to their increased longevity.
Literature review on delusions in Alzheimer's disease has revealed several advances in our understanding of the clinical epidemiology, etiology, and treatment of this syndrome. Nevertheless, several questions remain open. The incidence of delusions in Alzheimer's disease and its incidence in the different stages of the disease are not clear. More studies are needed on the predisposing risk factors for the development of delusions in Alzheimer's disease. The association of cognitive decline and delusions in Alzheimer's disease is still controversial and requires confirmation, specifically about whether different types of delusions occur at different severities of cognitive impairment and at different stages of disease progression. Further information is also needed to understand the etiology of delusions in Alzheimer's. The association of temperament, current stressors, and environmental reinforcements with delusions has barely been studied. The precise mechanism of how limbic dysfunction leads to delusions also merits further investigation. Finally, more studies need to be conducted on the treatment of delusions in Alzheimer's disease. Placebo-controlled double-blind studies using larger numbers of patients in various settings are called for. Both low- and high-potency neuroleptics and the newer agents need to be studied with regard to their dosage, side effects, compliance, and optimum duration of treatment. Nonpharmacologic treatments for delusions also merit study. Understanding the benefits of treatment on primary variables, such as reduction of delusions, as well as on secondary variables, such as caregiver status, aggression, and nursing home placement, is of critical importance.