As a large portion of the American population ages, management of the dementing illnesses is becoming increasingly important to the general psychiatrist and to the geriatric specialist. Progressive cognitive decline along with psychiatric and behavioral disturbances characterize the most common of these illnesses, dementia of the Alzheimer type (DAT). The behavioral symptoms that accompany Alzheimer's disease include agitation, sundowning, wandering, social inappropriateness. sexual impulsiveness, and repetitive purposeless activity. Common psychiatric disturbances include depression and psychosis. To assist the clinician in management of these behavioral disturbances, this article reviews the recent literature on the prevalence and treatment oí' both psychosis and troublesome behaviors associated with DAT.
Successful management of disruptive behaviors can make caring for the demented patient less burdensome to his or her caregivers. Rabins et al assessed this burden by interviewing the families of 55 demented patients.1 Of the caregivers interviewed, 879? complained of fatigue, anger, and depression directly linked to caring for the demented family member. Among the most distressing behaviors cited were physical agitation, accusatory behavior, and suspiciousness. Thus, controlling problem behaviors will not only improve the quality of ife for the patient but also will lessen the burden experienced by caregivers.
The prevalence of troublesome behaviors in DAT establishes the importance of successful treatment of these problems. Swearer et al studied the prevalence of behavioral problems in demented patients using a retrospective approach via primary caregiver questionnaires.2 Of the 126 patients involved in the study. 57 met the NINCDS-ADRDA criteria for primary Alzheimer's disease. Only four of the remaining patients had a history of psychiatric disorder. One or more of the problem behaviors were reported in 839Í of the patients. Angry outbursts were most frequent at SI'/i, followed by dietary and sleep disturbance reported at 46^t and 45^f, respectively. Paranoia was present in 32% of patients, whereas violence and hallucinations and delusions were present in slightly more than 20Vf.
In their study of 127 DAT patients meeting DSM-III criteria for Primary Degenerative Dementia. Teri et al demonstrated an increasing prevalence of behavioral problems with decreasing mental status.3 As cognitive impairment worsened, a higher number of problems occurred in each individual patient. Three or more behaviors were present in only 8% of the mildly impaired as opposed to 41% and 88'% in the moderate and severe groups, respectively. Both of these studies relied on the subjective accounts of caregivers and nonstandardized behavioral assessment; however, each study presents strong evidence that problematic behaviors are present in the majority of patients with DAT.
Because psychosis is a psychiatric phenomenon rather than a behavioral problem, it has been considered separately in a few studies. Burns et al performed one such assessment involving 178 patients meeting the NlNCDSADRDA criteria for possible or probable Alzheimer's disease.4 This study used a clinically diverse sample population consisting of 74 patients living at home, 78 iiipatients, and 26 in residential care. This group also relied on subjective caregiver reports and found that delusions (15.7%), visual hallucinations (139?, auditory hallucinations (9.6%), and misidentification syndromes (30.3(7f i were common in this population.
In addition to the high prevalence of troublesome behavior presented previously, this study demonstrates that a fraction of DAT patients will also experience psychosis during the course of their illness. It is therefore important to develop efficacious treatment options for these problems.
ASSESSMENT OF BEHAVIORAL PROBLEMS
Standardization of behavioral assessment allows for comparison and consistency when assessing the efficacy of different management options. Cohen-Mansfield recently reviewed the current methods of clinical assessment/'1 The methods discussed include caregiver rating, direct observation, and technological devices such as videotaping. Because the caregiver opinion commonly guides clinical management, the clinical importance of caregiver assessment was noted. In other words, treatment success is often directly proportional to the happiness of the caregiver. Thus, while the objective collection of data obtained through observational methods may provide more accurate data for research purposes, the caregiver ratings remain highly applicable to clinical management.
The same article goes on to discuss the Cohen-Mansfield Agitation Inventory (CMAI).5 This tool for standardized assessment cunXains 29 behaviors that the caregiver rates according to a 7-point scale in which 7 is the highest frequency of occurrence. This scale is currently the only well-studied one that focuses primarily on troublesome behaviors. Cohen-Mansfield divides agitated behavior into three main categories: physically aggressive, physically nonaggressive, and verbally agitated. In contrast to the prior generalized use of the term agitation, definition using the CMAI improves the investigative consistency and clinical applicability of this term. Because the identification of target symptoms is at the foundation of management of psychiatric disorders, well-defined agitated behaviors are important to clinical decisionmaking. To assess the literature critically and apply successful treatment options to individual patients, target symptoms responsive to intervention must be identified. A standardized approach to rating behavioral problems, such as the CMAI, will ensure clinical applicability only if individual target symptoms are mentioned in addition to the overall rating scale. Improvement in specific target symptoms is often more clinically relevant than improvement in a standardized rating. Once target symptoms are identified, management of these behaviors can be initiated accordingly.
NONPHARMACOLOGIC MANAGEMENT OF BEHAVIORAL PROBLEMS
The cause of behavioral problems is often difficult to discern because of the level of cognitive impairment and resulting lack of communication skills. In the demented patient, common causes can be directly related to the inability to impart their needs. It is thus important to identify any underlying triggers before proceeding to pharmacologie treatment of problem behaviors. A good history and physical examination are invaluable in the initial approach to these behaviors. Possible causes include hunger, pain, boredom, overstimulation, loneliness, fear, acute confusion or disorientation, and conflicts with other residents or staff. Manipulation or treatment of any of these triggers should be attempted as the first step in management of behavioral disturbances.
Psychotic symptoms should likewise receive a thorough evaluation before the initiation of medical management. Any prior history of a psychiatric disorder should be determined. Toxicity from either prescribed or illicit drugs should be assessed. Benzodiazepine withdrawal and alcohol withdrawal or intoxication are also possible causes of psychosis in these patients. Initial laboratory and radiographie screening may be a warranted step in evaluating the psychotic DAT patient. Underlying medical causes of delirium that should be ruled out include infection, dehydration, hypoxia, thyroid disease, B12 and thiamine deficiency, hyperadrenalism, hypo- or hyperglycemia, hypercalcemia, sodium and potassium imbalance, stroke, or intracranial pathology.
In a recent review of behavioral approaches to management, Gugel stressed the basic needs of the individual as causing all behavior.0 This article outlined human needs as originally developed by Maslow. The physiologic needs of "hunger, thirst, stimulation, relaxation, elimination, sex, sleep, and bodily integrity"6 are at the most primitive level. These basic needs are followed in order by safety needs, need for love, and finally the need for self-esteem. An initial behavioral approach to these problematic DAT patients will therefore assess whether these needs are being met sufficiently.
The behavioral treatment outlined by Gugel, opérant conditioning, attempts to exchange problematic behaviors for more acceptable behaviors. This method identifies and alters reinforcers of problematic behaviors to reinforce positive behaviors. Gugel points out that demented elderly patients are usually responsive to positive reinforcers such as praise rather than scolding or negative reinforcement.
Discerning modifiable behaviors from those that usually respond to pharmacologie treatment is important when treating DAT patients. This issue has been addressed in a previous review.7 Maletta lists the following "behaviors" as generally responsive to medication: delusions, hallucinations, verbal and physical agitation, and regressed behavior. He further described behaviors generally not responsive to medication management. These problems include wandering, socially inappropriate activities, purposeless repetitive activities, difficult personalities, hoarding, and stealing. Although categorization of behaviors may be helpful, the initial evaluation for modifiable causes should not be abandoned in the pharmacologically amenable group.
Wandering behavior is traditionally not responsive to medication and may simply be best remedied by providing an enclosed area where the patient can walk. The repetitive purposeless behaviors can be lessened by providing daily activities for the DAT patients. This strategy may work with any of the problem behaviors that are caused by underlying boredom or lack of stimulation. Sundowning behavior may be improved by increasing daytime activity, preventing daytime napping, and enforcing a regular sleep schedule. When problem behaviors are directly related to cognitive loss, a reassuring, empathie caregiver is the key to managing the medication-resistant behaviors. To allow efficacious use of environmental, social, and behavioral modification, nonpharmacologic treatments need to be tested further through randomized, controlled clinical trials. After the aforementioned modifications are exhausted, the clinician usually must proceed to pharmacologie treatment options,
PHARMACOLOGIC MANAGEMENT OF BEHAVIORAL PROBLEMS
As previously mentioned, the first step in initiating pharmacologie treatment is identification of target symptoms. With target symptoms in mind, medication should be used based on current knowledge of efficacy and adverse effects. In keeping with a medical model, the underlying pathophysiology of specific disturbances should guide not only research trials but also clinical prescription of medication. This approach is obviously not possible with some of the ill-defined problematic behaviors; however, one of the most troublesome behaviors, aggression, has some fairly well-defined neurocbemical correlates.4
Eichelman reviewed the role of certain neurotransmitters including acetylcholine, dopamine, GABA, norepinephrine, and serotonin. Although both acetylcholine and dopamine increase aggressive behavior in animal studies. GABA and serotonin have been shown to inhibit animal aggression, Serotonin has had the strongest corrélation with human aggression research, whereas the effect of norepinephrine has been equivocal in both animal and human studies. Therefore, serotonei'gic, GABAergic, and dopamine-blocking medications appear to be intuitive choices corresponding to these mechanisms. Because a deficiency in acetylcholine has been implicated in the pathogenesis of DAT, ariti - cholinergic treatment is not an acceptable option. Because tryptophan is a serotonin precursor, Eichelman points out. that the use of lithium takes advantage of the neurochcmistry by increasing tryptophan uptake into the brain. This article also mentioned a possible role of iiinbic kindling in episodic aggressive behavior. The anticonvulsant carbamazepine has a speculated limbi e anti-kindling effect possibly explaining its efficacy in treating aggressive behavior. The use of neuroleptics in the agitated DAT patient has generally been for the sedative properties of this class of drugs; however, with dopamine implicated in the neurochemistry of aggression, these drugs may act more directly on pathophysiology.
The selection of a specific medication must always be made by weighing its benefits against its inherent risks. The elderly patients that comprise the DAT population are more sensitive to adverse effects of medication than are their younger, healthy counterparts that are seen by the general psychiatrist. This fact should be kept in mind when determining which medication and dose to use. The clinician should individualize titration of psychotropic medications to an effective minimal dose.
The neuroleptic medications have traditionally been used to treat the troublesome behaviors of the DAT patient. Although neuroleptics are the most efficacious drugs for managing psychosis in these patients, their long-term adverse effects outweigh their efficacy in treating behavioral problems. The metaanalysis done by Schneider et al revealed that only 18 of every 100 patients in the analyzed studies benefited from treatment with a neuroleptic.'·1 This previously belabored, modest efficacy pales in comparison to the risk of anticholinergic, autonomie, and cardiac side effects associated with the lowpotency neuroleptics and risk of extrapyramidal symptoms and tardive dyskiiiesia associated with long-term use of the high-potency neuroleptics. ?"·? Because the high-potency neuroleptics have less of the anticholinergic and cardiac side effects, short-term use of these medications, when necessary, may be acceptable. The high-potency medications such as haioperidol can be effective when initiated at low doses, such as 0.5 mg, and then titrated upward until symptomatic improvement occurs.11 The novel antipsychotic risperidone may be particularly useful in this patient population, having the benefits of the high-potency neuroleptics but with lower extra pyramidal side effects.
The clinician should always consider trials of decreased dose or discontinuation of the medication when target symptoms are under control. The clinician should also keep in mind that environmental, social, and hehavioral modifications are often helpful in the DAT patients with psychotic symptoms. Although large, randomized, controlled trials are lacking in support of alternative medications, initial case studies and small trials appear promising.
The serotonergic agent trazodone seems to be an efficacious alternative in the treatment of the agitated DAT patient. The proposed mechanism is that of serotonergic inhibition of aggressive behavior. A retrospective study of 16 patients meeting NINCDS criteria for probable Alzheimer's disease revealed that trazodone treatment resulted in at least mild benefit in 13 of these patients12; however. 5 of the 16 patients had to discontinue use secondary to possible adverse effects. The listed adverse effects were hypotension, ileus, increased agitation, and worsened confusion. Pedal edema was noted in all patients receiving trazodone but was not sufficient to warrant discontinuation. Because all of the patients studied had comorbid medical illnesses and multiple medication regimens, the adverse effects could not be solely attributed to trazodone. Although four patients were identified as much improved, the target symptoms that improved were not noted in this report. In this study, the dose ranged from 100 to 300 mg petday, with a mean dose of 180 ± 70 mg per day.
In another report of six demented patients treated with trazodone and adjunctive tryptophan, four showed definite improvement.13 The four improved patients showed a reduction in aggression, noisy behavior, and temper tantrums. Trazodone was started at 50 mg twice daily and tryptophan 500 mg twice daily. Both medications were titrated upward until target symptom improvement or side effects occurred. Obviously, the literature on trazodone treatment of behavioral problems is in its early stage. Because this medication is generally well tolerated, large, long-term, controlled trials are needed to confirm the efficacy of this option in pharmacologie treatment of agitated behavior.
Buspirone is another medication with proposed serotonergic activity that has been used in DAT patients with problematic behaviors. Successful case study reports led Sakauyc et al to perform an open-label study involving 10 patients who met NINCDS-ADRDA criteria for Alzheimer's disease.1'1 They used the GohenMansfield Agitation Inventory (CMAI) to assess behavioral response to treatment with buspirone. A 22% reduction m the baseline CMAI was achieved with an initial dose of 5 mg three times per day increased at 5 mg per day to a maximum of 60 mg per day. Through clinical judgment, they observed maximal benefit from a dose of 30 to 40 mg per day at weeks 5 through 7. The only side effect mentioned in the study, nausea, resulted in one patient dropping out. Similar to trazodone, buspirone is usually well tolerated; however, good long-term trials are also lacking with this medication.
Like the neuroleptics, this class of drugs traditionally has been employed for its sedative effects. Previously reviewed literature revealed no studies that tested the use of benzodiazepines solely in DAT patients.1·"' The studies reviewed supported the use of short half-life agents with the specific target symptom of agitation due to an underlying anxiety showing the best results. The most efficacious of the short-acting benzodiazepines has been oxazepam, with effective doses ranging from 15 to 80 mg per day. Because it is well absorbed intramuscularly, lorazepam has generally been useful for short-term treatment of severely agitated patients; however, this use may not be applicable to the DAT patient because of the decrease in cognitive function and increase in confusion the benzodiazepines are known to cause. This effect can lead to a vicious cycle when the agitated DAT patient receives repeated, as-needed doses causing increased confusion and increased agitation. The long-term risks of dependence and withdrawal also limit the use of this medication. Clinical trials specifically dealing with DAT patients are needed to justify the use of these medications for indications other than anxiety.
Beta-blocking medications such as propranolol have reported efficacy in younger organically impaired patients."* Again, large, randomized, controlled trials are lacking involving the DAT patient population. The largest doubleblind, placebo-controlled study to date assessed the effects of pindolol on 11 organically impaired patients with behavioral abnormalities.17 In comparison with placebo, pindolol significantly reduced assaultiveness, hostility, un communicativeness, uncooperativeness, and repetitive behaviors. These results wei'e observed within 2 weeks of the start of treatment and at an optimal dose of 40 to 60 mg per day. Furthermore, the patients improved without the occurrence of significant bradycardia or hypotension. The inherent sympathomimetic effect of pindolol results in improved sympathetic tone when compared with propranolol. Initial reports have demonstrated significant efficacy and minimal side effects.
Carbamazepine has been employed recently in the treatment of agitated behavior. Gleason and Schneider used carbamazepine in treating nine patients who met the NINCDS-ADRDA criteria for probable Alzheimer's disease.18 According to change in Brief Psychiatric Rating Scale l BPRS), substantial improvement occurred in five of the patients. They further cited specific improvement in tension, hostility, uncooperativeness, and agitation. Doses ranged from 200 to 1000 mg per day with a corresponding range in serum drug levels from 2.3 to 9.6 pg/mL. Ataxia was the most common side effect and was present in three of the patients studied. One of the affected patients had to discontinue the medication as a result. In a more recent p i ace bo con trolled trial, carbiirnazepine reduced BPRS scores significantly more than placebo.19 Clinical global impression improved on 16 of 25 agitated, demented nursing homo patients without a decrease in Mini-Mental State Examination (MMSE). The preservation in cognitive function is an important asset to any medication treating the behavioral problems in DAT patients. The doses in this study ranged from 100 to 800 mg per day with an average dose near 300 mg per day. Despite the heterogeneous demented sample, this study is a good start toward establishing carbamazepine as a proven pharmacologie treatment option in treating DAT patients with agitation.
Another anticonvulsant, sodium valproate, has been efficacious in initial case reports.-" Mellow and colleagues report significant improvement in two of four agitated, demented patients. The two responsive patients met diagnostic criteria for DAT and benefited from a dose ranging from 1000 to 2500 mg per day. The medication was very well tolerated by all four participants. This initial repoli warrants further investigation of sodium valproate as a treatment option in agitated DAT patients.
When using either of these anticoiivulsant medications, the clinician should periodically check liver enzymes because both drugs can cause hepatotoxicity, although rarely. Carbamazepine also rarely causes leukopenia or can precipitate aplastic anemia. Thus, the clinician is justified in periodic assessment of complete blood counts.
Lithium has also been used in the treatment of behavioral problems in the DAT patient. Although some initial case studies showed significant benefit, more recent smail trials have been equivocal.21,22 As with the previous alternatives, large, randomized, controlled trials are needed to establish efficacy of lithium use in DAT patients.
Bright-light treatment has been specifically targeted at the behavior of sundowning. In a study of 10 patients meeting NINCDS-ADRDA criteria for Alzheimer's disease, daily exposure to 2 himrs of bright light reduced clinical rating of sundowning behavior in eight of the 10 patients. 2y A light box containing three U-shaped fluorescent bulbs provided 1500 to 2000 lux of light treatment in each 2-hour period. This appears to be a very simple and benign treatment for the often troublesome behavior of sundowning. It is hoped that further clinical trials will build on these promising initial results.
In a case report of two elderly demented men. estrogen significantly reduced physically aggressive behavior without occurrence of any adverse effects.2"1 Because testosterone has been shown to increase aggression, the use of estrogen seems intuitive in treating aggressively agitated DAT patients. Clinical trials involving estrogen will need to be done to support initially positive case reports.
In addition to the medication options listed previously, short-term psychiatric hospitalization has been beneficial for the agitated DAT patient.-1"' In a prospective study of 120 Alzheimer's disease patients, short-term iripatient stays were effective in returning patients to home and preventing institutionalization. These brief visits were used to reevaluate pharmacologie regimens and initiate comprehensive treatment plans. Both the patient and the burdened caregiver may benefit from brief separation and reorientation of management. The clinician should keep this option in mind when dealing with stressed caregivers and their agitated patients or family members afflicted with Alzheimer's disease.
A majority of patients with Alzheimer's disease will develop behavior problems during the course of this chronic, debilitating disease. In addition, a smaller portion of these patients will also present with psychotic symptoms. Identification of specific target symptoms will allow the critically thinking clinician to apply up-todate research to his or her individual patients. Once the modifiable environmental and social !actors have been identified, manipulation of these triggers should be attempted in the first step of treatment. Because aggressive agitation and psychotic symptoms are the most pharmacologically responsive behavioral disturbances, medication trials are warranted when these target symptoms are present. Whether employing behavioral modification or medication, preservation of cognitive functioning should be a primary goal in this group of sensitive patients. Although clinicians historically have taken advantage of the sedative properties of neuroleptics in the agitated patient, options with less severe side-effect profiles and more specific actions on proposed pathophysiologic mechanisms are available and under further investigation.
1. Rabins P. Mace N. Lucas MJ. The impart of dementia on the family. JAMA. 1982; 248:333-335.
2. Swearer J, Drach man D. O 'Dotinell B, et al. Troublesome anil disruptive behaviors in dementia. JAn? Gcriatr Soc 1988; 36:784-790.
3. Teri L. Larson E, Reiner B. Behavioral disturbance in demi.mtia of l Iu1 Alzheimer's type. -J Am Geriatr SIK: 1988; 36:1-6.
4. Burns A. -lacnby R, Levy K, Psychiatric phenomena in Alzheimer's disease. Br j Psyrhialry. 1990; 157:72-93.
5. Cohen-Mansfield J. Assessment of disruptive behavior/agitation in the elderly: function, methods, and difficulties. J Geriatr Psychiatry Nenrol. 1995: S:52-6Q.
6. Gugel RN. Behavioral approaches IOr managing patients with Alzheimer's disease and related di sordera. Mf ti Clin North Am. 1994; 78:861-867.
7. Maletta G. Pharmacologie· treatment and management of the aggressive demented patient. Psychiatrie Annal. 1994; 20:446-455.
8. Eichelman B. Neurochemical buses of aggressive behavior. Psychiatric Annals. 1987; 17:371-374.
9. Schneider L. Pollock V, Lyness n. A metaanalysis of controlled trials of neuroleptic treatment in dementia- J Am Geiiatr SM-. 1990; 38:553-563.
10. Grospberg G. The pitfalls of metaanalysts. J Am Cferiair Sw. 1990; 38:607.
11. Grossberg G. The older patient with psychotic symptoms. Psych in t r Sen·. 1995; 46:55-59.
12. Aisen P, Johannessen D, Marin D. Trazudone for behavioral disturbance in Alzheimer's disease. .4m ./ Geriatr Psychiatry. 1993; 4:349-350.
13. Wilc.ock G, Stevens J, Perkins A. Trazodono/tryptophan for aggressive behavior. Lainvt. 1987. April 18:929-931).
14. Sakauye K. Camp C. Ford P. Effects of buspirone on agitation associated with demeritan. Am -J Geriatr Psychiatry. 1993; 1:82-84.
15. Stern R. Duffeimeyer M, Zemishlani Z. et al. The USL- of henzoeliazepines in the management of behavioral symptoms in demented patients. Pnychiatr Clin North Am 1991; 4:375-384.
16. Elliot F. Propranolol for the control of belligerent behavior following acute bruin damage. Ann Neu rol. 1977; 1:489-491.
17. Greedciyke R, Kanter D. Therapeutic effects of pindolol on behavioral disturbances associated with organic brain disease: a double-blind study. J Ciin Psychiatry. 1986; 47:423-426.
18. Gleason R. Schneider L. Carbamazepine treatment of agitation in Alzhtimer's outpatients refractory to neuroleptics. J Clin Psychiatry. 1990; 51:115-118.
19. Tariot P. Erb R, Leibovici A, et al. Carbarnazepíne treatment of agitation in nursing home patients with dementia: a preliminary nt udv. J Am GcricitrSw. 1994; 42:1160-1166.
20. Mellow A, Siilann-Lopez C. Davis S. Sodium valproale in the treatment of behavioral disturbance in dementili. J Ceriatr Psychiatry Neural. 1993; 6:205-209.
21. Havens W. Cole J. Successful treatment of dementia with lithium. -7 Clin Psychupliarmacol. 1982; 2:71-72,
22. Holton A, George K. The use of lithium carbonate in severely demented patients with behavioral disturbance. Rr -J P.ivchtatry. 1985: 146:99-100.
23. Satlin A, Voliera- L. Ross V. et al. [!right light treatment of behavioral ;md sleep disturbances in patients with Alzhc-imers disease. Am -J Psychiatry. 1992; 149:10281032.
24. Kyomen H, Nobel K, Wei J. The use of cstrogen to decrease ag(*Tesi-ivc physical behavior in elderly men with dementia. JAm Geriatr Soc. 1991; 39:1110- ? 12.
25. Zuhenko G. Rosen J. Sweet R. et a!. Impact of psychiatric hoapitalization 'in behavioral complications of Alzheimer's disease.Am J Psychiatry. 1992; 149:1484-1491.