Treating Aggression in Patients With Dementia - Treatme...

Treating Aggression in Patients With Dementia

After reading this article, you will be familiar with:

* Behavioral distrubances associated with dementia and a conceptual model to categorize them;
* Nonpharmacological modalities used to reduce negative behaviors;
* Pharmacological interventions and their place in treatment.

Who will benefit from reading this article?

Psychiatrists, primary care physicians, neurologists, nurse practitioners, psychiatric nurses and other mental health care professionals. Continuing medical education credit is available for most specialties. To determine if this article meets the CE requirements for your specialty, please contact your state licensing board.


Funmilayo Rachal, MD, and Mark E. Kunik, MD, MPH. Treating Aggression in Patients With Dementia. Psychiatric Times 2006; 25(6)

Dementia is characterized as a progressive and chronic decline in cognitive function, not limited to memory impairment, which significantly interferes with baseline daily functioning and frequently involves behavioral disturbances. Although there is limited longitudinal data available, it is known that behavioral problems in dementia negatively affect patients and caregivers. These disturbances lead to institutionalization (Evans and Strumpf, 1988), increased costs and caregiver burden, and a poorer prognosis (Lyketsos et al., 1999; Rabins et al., 1999).

Several studies have attempted to conceptualize disruptive behaviors within theoretical frameworks and categorize them into subtypes to further define and standardize assessment tools. Although clinicians and researchers have not yet reached a consensus, the most commonly used definitions are from the Cohen-Mansfield Agitation Inventory. The inventory defines disruptive behavior on a spectrum of aggressive to nonaggressive, manifested either verbally or physically (Figure 1). Examples include verbally nonaggressive behavior, such as complaining or negativism, to more verbally aggressive behavior, such as cursing, screaming or verbalizing sexual advances. Physically nonaggressive behaviors include pacing, inappropriate disrobing or hoarding versus more physically aggressive behaviors, such as spitting, kicking, biting or hurting self or others (Cohen-Mansfield, 2000). Aggressive behaviors are of concern to caregivers because these behaviors often complicate daily care and make patient management difficult. However, behavioral disturbances are not limited to aggression but are often confounded by apathy, psychosis and depressive symptoms.

Although few studies have compared the frequency of particular behavioral disturbances in patients with dementia versus that in the general population, two large studies confirmed that up to 75% of study participants with dementia had at least one Neuropsychiatric Inventory (NPI) symptom and that 62% of these were clinically significant (Lyketsos et al., 2002). The most common disturbance reported in 27% to 36% of study participants with dementia was apathy, followed by up to 32% with depression and 30% with agitation and aggression (Figure 2) (Lyketsos et al., 2000). Both studies give caregivers and clinicians a framework of common behaviors to anticipate. More importantly, these studies suggest that behavioral disturbances are a central component of dementia and contribute significantly to morbidity. Various treatment modalities, including nonpharmacologic and pharmacologic approaches, have been used in an attempt to decrease the morbidity and mortality associated with behavioral disturbances in patients with dementia.

Assessment and Treatment

A multidimensional behavioral model was developed to assist clinicians and caregivers in understanding the modifiable and non-modifiable patient, caregiver and environmental factors that affect behavior (Figure 3). The patients' factors influence their predisposition and expression of unmet needs, and the caregivers' factors influence their perception and management of a behavioral disturbance. Use of this model allows differentiation of behaviors that are part of the dementia process from those that are exacerbated by other factors, ultimately aiding clinicians in developing a more effective individualized treatment plan (Braun and Kunik, 2004).

Nonpharmacologic Interventions

Nonpharmacologic interventions in patients with dementia are often focused on individualizing care and altering personal, behavioral or environmental factors that may contribute to inappropriate behaviors. It has been well established that nonpharmacologic interventions play an important role in the management of behavioral disturbances in patients with dementia (Table). However, many challenges have been identified in implementing nonpharmacologic interventions in the nursing home setting. Some of these challenges are inherent in the organization of nursing homes, including staff treatment beliefs and training levels, lack of technology and the physical setting itself. Other factors are more external, such as low professional staffing, a punitive legal climate and concerns regarding regulatory agencies and surveys. Together, these pose significant restraints on meeting the more individualized needs of patients with dementia and educating family and staff about identifying and coping with behavioral disturbances (Cody et al., 2002).


___
Table -- Comparison of nonpharmacological interventions for treating disruptive behavior in patients with dementia

Behavioral therapy

Piagetian Therapy
- Significantly decreased disruptive behavior at 12-18 months
- Facility-level parallel study

Behavior Therapy
- Significantly decreased depressive symptoms
- RCT


Engaging activities

Exercise study 1
- 22% decrease in agitation; 150% increase in agitation with control
- RCT

Exercise study 2
- Improved depression and physical role functioning at 3 months
- RCT

Activities
- 71% decrease in agitation in activities vs 49% in control
- RCT

Audio stimulation
- No significant change in agitated behavior
- RCT


Psychomotor

Study 1
- No significant change in apathy or depression
- RCT

Study 2
- No significant change in apathy or depression
- Significanty decreased aggression
- RCT


Sensory

Music/light/massage
- 25%-54% improvement in agitated behavior
- Observational

---
RCT = randomized controlled trial
___


Behavioral therapy. Behavior-based therapy aims to detect situations in which behaviors occur and provide instruction in anticipating and coping with these situations. Both behavioral modification and behavioral therapy have been studied as treatment options for patients with dementia.

Teri and colleagues (1997) conducted a study of behavior therapy and found statistically significant reductions in depressive symptoms in participants taught to identify behavior-pleasant events, increase the frequency of these events and confront behaviors that interfered with these events. A review of the literature by Snowden and colleagues (2003) found six studies that reported improvements in disruptive behavior, but only two of these studies had large sample sizes. One large study used a Piaget-based theory that developed stage-based treatment plans for patients, grouped by developmental stages (Matteson et al., 1997). This three-year study reported lower disruptive behavior scores for the intervention group at 12 months to 18 months but not at three months. These studies suggest that behavioral modification and therapy play a role in decreasing the frequency of disruptive behavior. A recent review reported that behavioral management techniques geared toward caregiver psychoeducation and interaction may positively affect patient behavior for several months (Livingston et al., 2005). Nevertheless, additional, larger trials are needed to provide more solid evidence.

Engaging activities. Camp and colleagues (2002) proposed that providing social support and contact, engaging in activities, and offering relief from discomfort with individualized treatment plans would decrease disruptive behavior because they address the patient's unmet needs. However, studies reviewing efficacy of these approaches have been mixed. In their review, Snowden and colleagues (2003) reported a 12% to 60% improvement in agitated behaviors while engaging subjects in activities, ranging from a rather passive audio/video-simulation to more active exercise-based activities. The only randomized control trial on audio simulation reported no significant difference in behavior during direct observation and weekly staff survey, but improved agitation of 67% with intervention compared to 59% and 46% with usual care and placebo, respectively (Camberg et al., 1999). However, a six-month activity-based randomized control trial conducted by Rovner and colleagues (1996) of 81 patients provided mentally and physically stimulating activities showed a 71% decrease in agitation in the activities group compared with a 49% decrease in the control group. Two exercise-based activity studies reported decreased agitation and improved depression and physical role functioning. One study found a 22% decrease in the mean observation of agitation in the intervention group compared with a 150% increase in mean observed agitation in the control group (Alessi et al., 1999). The second community-based exercise trial combined a three-month, home-based exercise program with behavioral management training and reported improved depression based on the Cornell Depression Scale and improved scores for physical role functioning rated on the Sickness Impact Profile's Mobility subscale (Teri et al., 2003).

Together, these studies validate that engaging understimulated patients with age-appropriate exercise and socially engaging activities may decrease agitated behaviors.

Sensory stimulation. It is challenging to provide proper stimulation for patients with dementia, and many support the theory that too little or too much stimulation is often an underlying source of agitation and disruptive behavior. Consequently, many studies have investigated the role of music, light and touch, with mixed results. A review of these interventions has found that, of the reported 25% to 54% improvement in agitated behavior, the results were statistically significant in only six of 20 trials. The study designs have been observational, so further studies are needed to confirm the benefits of these alternative interventions as a standard of care (Snowden et al., 2003).

Psychomotor therapy. Patient stimulation is not restricted to the sensorium but can also be psychosocial or cognitive in nature. Psychomotor therapy is designed to help patients with dementia cope with changes associated with the disease process via psychosocial and cognitive stimulation, such as group activities and games. Little evidence is available on psychomotor therapy. Of two studies reviewed by Verkaik et al. (2005), one found significantly lower aggression scores rated on a Dutch subscale of aggression but no difference in apathy and depression (Droes, 1991). The other study reported no statistical difference in apathy or depression, although there was some improvement in cognitive functioning (Hopman-Rock et al., 1999). Therapies aimed at increasing cognitive functioning have not shown promising results, but those aimed more at improving psychosocial functioning have been more promising. This is probably because the focus of these psychosocial interventions is to meet the individualized needs of the patient.

Multiple nonpharmacologic treatments, including emotion-, cognitive- or simulation-oriented methods, are not discussed in this article. These studies have small sample sizes and observational designs, and thus these modalities need further investigation. However, nonpharmacologic interventions may be further investigated and used given the limited efficacy of pharmacologic interventions and the recent reports of serious side effects associated with long-term use of antipsychotics.

Pharmacologic Interventions

Currently, there is no consensus among clinicians and researchers or good evidence for the use of pharmacologic treatments for disruptive behaviors. Despite the lack of evidence, sometimes nonpharmacologic interventions are ineffective, and pharmacologic interventions become necessary. In choosing these interventions, the goals should be clear as to whether the agitation is acute or chronic and what factors may be precipitating the behavior. Antipsychotics have commonly been used to control aggressive behavior associated with both acute and chronic agitation. Long-term use in patients is often a result of staff perception of medication effectiveness, although it is not based on evidence or clinical practice (Cohen-Mansfield et al., 1999). However, the institution of the Consolidated Omnibus Budget Reconciliation Act affected the prescribing practices of psychotropics, and several studies reported a decrease in antipsychotic administration (Borson and Doane, 1997; Lantz et al., 1996). Evidence in the literature supports that drug withdrawal is safe and does not lead to rapid deterioration or exacerbation of disruptive patient behavior (Cohen-Mansfield et al., 1999).

A U.S. Food and Drug Administration (2005) public health advisory recently warned that patients with dementia treated with atypical antipsychotics were at increased risk of death compared with those treated with placebo. This warning was issued after evaluation of 17 studies with aripiprazole (Abilify), olanzapine (Zyprexa), quetiapine (Seroquel) or risperidone (Risperdal) that demonstrated a 1.6- to 1.7-fold increase in mortality. The cause of death was heart-related, either by heart failure, sudden death or infections (mostly pneumonia).

A recently published meta-analysis of 15 studies conducted by Schneider and colleagues (2005) calculated a relative risk of 1.65 for death by pooling data of antipsychotic use for brief trial periods of eight to 12 weeks in patients with dementia. (This risk corresponds to the aforementioned FDA estimates.) The risk of death was not statistically significant for each drug, but the overall odds ratio for death in patients treated with antipsychotics was 1.54% (95% confidence interval [CI] 1.06-2.23, p=0.02) when compared with that for patients treated with placebo. The absolute risk difference found in this meta-analysis was 1% excess deaths with atypical treatment versus placebo. When examining the number needed to treat, ranging from four to 12 in these studies, the authors reported that for every nine to 25 subjects helped in these trials there would possibly be one death. Thus, the FDA warning, coupled with the meta-analysis data, suggest that antipsychotic use in elderly patients with dementia is not without significant risk.

Antipsychotics-acute agitation. Historically, antipsychotics have been prescribed for treating psychosis as well as acutely violent and agitated behavior. The common complication of disruptive behavior in patients with dementia-not limited to aggression and agitation but also psychosis-has led to widespread use of these agents as a long-term treatment in patients with dementia, especially those in the institutional setting. For treating acute behavioral disturbances that pose a risk to patient and staff, it is evident that rapid-acting psychotropics are superior to placebo. A double-blind study of intramuscular olanzapine reported significant results with the use of 2.5 mg and 5 mg at two hours and 24 hours, when compared with placebo (Meehan et al., 2002).

A meta-analysis conducted by Lonergan and colleagues (2002) illustrated that doses of the conventional antipsychotic haloperidol (Haldol) at 1.2 mg to 3.5 mg were superior to placebo for the treatment of aggression alone. They concluded:

1. Evidence suggests that haloperidol was useful in the control of aggression, but was associated with increased side effects; there was no evidence to support the routine use of this drug for other manifestations of agitated dementia. 2. Similar dropout rates among haloperidol and placebo treated patients suggested that poorly controlled symptoms, or other factors, may be important in causing treatment discontinuation. 3. Variations in degree of dementia, dosage and length of haloperidol treatment, and in ways of assessing response to treatment suggested caution in the interpretation of reported effects of haloperidol in the management of agitated dementia. 4. The present study confirmed that haloperidol should not be used routinely to treat patients with agitated dementia. Treatment of agitated dementia with haloperidol should be individualized and patients should be monitored for side effects of therapy.

These findings have been confirmed in other studies, and a significantly higher incidence of extrapyramidal symptoms has also been reported (Allain et al., 2000). A recent randomized controlled trial comparing oral olanzapine and haloperidol supported these findings and reported comparable efficacy of these agents in treating aggression (Verhey et al., 2005). Consequently, these agents may be more effective for short-term use in treating aggression; with long-term use, patients are more likely to develop extrapyramidal symptoms.

Antipsychotics-long-term use. Although antipsychotics have been shown to be effective in the acutely aggressive patient, long-term use of atypical antipsychotics is still controversial. Many clinicians remain divided over the continued use of atypical antipsychotics in patients with dementia as a result of the varying efficacy reported in the literature and the documented increased risk of cerebrovascular and cardiovascular events, especially in patients with known risk factors. Of six randomized control trials of atypical antipsychotics reviewed in a study conducted by Sink and colleagues (2005), four found a significant decrease in neuropsychiatric symptoms. These four studies were conducted in the institutional setting for up to 12 weeks on participants with moderate-to-severe dementia. Of the trials conducted with olanzapine, one study had oral administration. Olanzapine was administered intramuscularly in the others. Orally administered olanzapine at doses of 5 mg/day and 10 mg/day, rather than 15 mg/day, was reported superior to placebo in reducing agitation/aggression, hallucinations and delusion, as measured on the NPI scale (Street et al., 2000). Another study found no significant difference in behavior measured on the subscales of the NPI, the Brief Psychiatric Rating Scale (BPRS) and the Clinical Global Impression of Change (CGIC) except on secondary analysis (De Deyn et al., 2004). Increased adverse effects, including extrapyramidal symptoms, somnolence and cerebrovascular events, were reported for both olanzapine and risperidone as compared with placebo (Sink et al., 2005).

As with olanzapine, studies of risperidone show varying efficacy. A comparison study conducted by De Deyn and colleagues (1999) of haloperidol, risperidone and placebo found no significant difference between these interventions in primary outcome measures by Behavioral Pathology in Alzheimer's Disease Rating Scale (BEHAVE-AD). Significant results were reported on secondary outcome. These findings conflicted with those reported in other risperidone studies.

In one study conducted by Katz and colleagues (1999) of doses ranging from 0.5 mg to 2 mg/day risperidone, it was concluded that 1 mg/day to 2 mg/day was superior to placebo in treating aggression on multiple behavioral scales with a greater than 50% reduction in BEHAVE-AD scores compared with a 33% reduction with placebo. All studies comparing risperidone with placebo reported increased adverse effects of somnolence; in addition, increased treatment-emergent extrapyramidal symptoms with risperidone compared with placebo and olanzapine were reported by Deberdt and colleagues (2005). Reports in a study by Brodaty and colleagues (2003) of serious adverse effects, including of cerebrovascular events, in 16.8% of the treatment group versus 8.8% in the placebo group were the catalyst for mandated warnings about using risperidone in elderly patients with dementia. Although the patients had known risk factors for cerebrovascular disease, these results made clinicians more cautious about using antipsychotics, and ultimately led to the FDA's public warnings on the serious side effects of these medications. Other studies have confirmed this increased risk of adverse events, especially with atypical antipsychotics (Percudani et al., 2005). Currently there remains no consensus within the literature on whether there is a difference in safety among first- or second-generation antipsychotics.

Collectively, these studies of atypical antipsychotics illustrate a dilemma: treating behavioral disturbances in patients with dementia may increase the risk of adverse events. Before prescribing antipsychotics to elderly patients with dementia, clinicians must assess whether the patient's distress is significant or may result in harm to the patient or others (Rabins and Lyketsos, 2005). If distress is significant and it is necessary to start antipsychotics, initial doses should be low and titration should be slow. Furthermore, clinicians must discuss with caregivers the increased risk of adverse effects, including akathisia, tremors, bradycardia, orthostatic hypotension, sedation, tardive dyskinesia and cerebrovascular events.

Antidepressants. Many studies have investigated the use of antidepressants for treating neuropsychiatric symptoms because of their lower side-effect profiles in comparison with antipsychotics. There is limited evidence in the literature of their efficaciousness. A Cochrane review of trazodone (Desyrel) found no significant benefits of trazodone to placebo (Martinon-Torres et al., 2004). One randomized controlled study of selective serotonin reuptake inhibitors (SSRIs) found that 20 mg/day of citalopram (Celexa) significantly improved agitation and lability, compared with placebo in hospitalized patients with dementia (Pollock et al., 2002). Trial limitations included a brief study period of 17 days and high dropout rate, with about half the participants in both the citalopram and placebo groups discontinuing the trial, either due to adverse effects or lack of efficacy (there was no statistical difference among the groups in the amount or reason for discontinuation). However, the authors felt that their results are generalizable for the effects of citalopram in the acute hospitalization setting. Further studies have found no significant effects of antidepressants on neuropsychiatric symptoms but have found these agents efficacious in treating concomitant depression and anxiety in patients with dementia. Further studies of SSRIs, especially sertraline (Zoloft), have found no significant effects on neuropsychiatric symptoms but have found these agents efficacious in treating concomitant depression and anxiety in patients with dementia (Finkel et al., 2004, Lyketsos et al., 2003).

Cholinesterase inhibitors and other agents. Although cholinesterase inhibitors are often used to augment cognitive function in Alzheimer's patients by blocking acetylcholinesterase, an enzyme responsible for the breakdown of acetylcholine, it is believed that these agents may also decrease neuropsychiatric symptoms (Ballard and O'Brien, 1999). In particular, cholinesterase inhibitors target hallucinations, believed to be associated with cholinergic depletion. The appeal of cholinesterase inhibitors lies in their tolerability in elderly patients and decreased risk of mortality compared with atypical antipsychotics (Ballard and O'brien, 1999; Sink et al., 2005).

Trials of cholinesterase inhibitors have found conflicting results, with either no clinically significant improvement or minor improvement in neuropsychiatric symptoms. A community-based randomized control study of subjects with Alzheimer's comparing donepezil (Aricept) with placebo reported significantly improved functional and behavioral symptoms in the intervention group at 24 weeks (a 5.6-point NPI score difference between donepezil at 5 mg/day to 10 mg/day versus placebo) (Feldman et al., 2001). Another commonly used cholinesterase inhibitor, galantamine (Reminyl), has been reported to be slightly significantly superior to placebo in studies with daily doses beginning at 16 mg (Herrmann et al., 2005). A study by Olin and Schneider (2002) confirmed that, for the most part, doses starting at 8 mg/day of galantamine were significantly better than placebo; but doses above 8 mg/day increased adverse gastrointestinal symptoms. The researchers concluded, "It appears that doses of 16 mg/d were best tolerated in the single trial where medication was titrated over 4 week periods, and because this dose showed statistically indistinguishable efficacy with higher doses, it is probably preferable initially."

Current studies of cholinesterase inhibitors have been limited by their statistically small effects and recruitment of patients without severe neuropsychiatric symptoms. Further studies of these agents with larger numbers of study participants and inclusion of patients with significant neuropsychiatric symptoms are needed to truly evaluate their effectiveness and generalizability in treating behavioral disturbances in the inpatient hospital setting.

This has led some researchers to study other potentially more tolerable agents, such as anticonvulsants and trazodone (Ballard and O'Brien, 1999; Martinon-Torres et al., 2004). However, modest effects have consistently been reported and confirm the need for further research into nonpharmacologic interventions.

Conclusion

The evidence for nonpharmacologic and pharmacologic interventions is weak, but the prevalence and consequences of behavioral disturbances require intervention. Historically, antipsychotics have been first-line treatment for disruptive behavior in patients with dementia. However, further studies are warranted before the use of antipsychotics is deemed the gold standard. Furthermore, given all the negative data and the warnings of serious side effects, there is a need to examine alternatives. This dilemma demands reliance on available evidence and use of interventions least likely to cause harm as a first-line strategy.

It is proposed that clinicians use nonpharmacologic interventions-such as behavioral, psychomotor and engaging activities-as first-line treatments for disruptive behavior. Using these techniques and developing individualized care plans to address the unmet needs (such as comfort, adequate pain control and a relaxed environment) of demented patients are critical first interventions. If results are not satisfactory, it may be beneficial to then supplement the nonpharmacologic interventions with a trial of pharmacological agents, such as cholinesterase inhibitors or antidepressants. Further use of antipsychotics for treatment-resistant patients may be an alternative after discussing the risks and benefits with the caregiver.


___
Dr. Rachal is research coordinator for the Houston Center for Quality of Care and Utilization Studies, Health Services Research and Development Service.

Dr. Kunik is a geropsychiatrist and serves as the associate director for research training for the South Central Mental Illness Research Education and Clinical Center (MIRECC) of the Veterans Administration. Dr. Kunik is also the associate director for the Houston Center for the Quality of Care and Utilization Studies and associate professor at the Meninnger department of psychiatry at Baylor College of Medicine.

Neither author has anything to disclose regarding the subject matter of this article.