The spectrum of behavioral changes in Alzheimer's disease

Alzheimer's disease (AD) is accompanied by prominent behavioral disturbances that may be the presenting complaint or may emerge in the course of the disease. ^[1-3] Some behavioral alterations (notably delusions) predict rapid decline. ^[4-8] Behavioral dysfunction is a major source of caregiver stress ^[9] and an important contributor to the decision to institutionalize AD patients. ^[10-16] Treatment of behavioral abnormalities can improve the quality of life of the patient and caregiver. Thus, behavioral abnormalities are important in the diagnosis, prognosis, and management of AD. There is limited information about the behavioral disturbances in AD. Impediments to our understanding have included the absence of clearly operationalized definitions of many behaviors, limited availability of tools that conveniently and reliably evaluate a wide range of behaviors, an imperfect understanding of pathophysiologic changes responsible for the behavioral alterations, and a relative absence of data on behavioral responsiveness to treatment.

Past studies have consistently shown a high frequency of behavioral changes in AD, although the proportion of affected patients has varied among studies. Personality change, which is often among the earliest behavioral changes in AD, ^[17] most commonly includes disengagement and apathy. ^[1,18,19] Delusions affect 10 to 73% of AD patients, ^[20-26] hallucinations 3 to 49%, ^[6,21,24-29] and elation 5 to 17%. ^[30] Depression has been particularly difficult to assess in the setting of dementia and little consensus has emerged with regard to its frequency. ^{[21,24,25,31]} Contributing to this lack of agreement is the failure of behavioral rating scales to distinguish between apathy and sadness. ^{[25,29,32,33]} Apathy is more prominent than depression in AD patients ^[34] and may be more frequent in patients with a later age of dementia onset, or with extrapyramidal signs. ^[35] Beyond this, little is known about apathy or its relationship with other behavioral changes in AD. Agitation has been defined broadly to include aberrant motor behavior, physical aggression, or verbal abuse. Past studies have generally shown a wide variation in the frequency of these behaviors, i.e., motor restlessness occurring in 21 to 60% of AD patients, ^[22,36-38] wandering in 10 to 61%, ^{[22,25,29,38-40]} and aggression or assaultiveness in 18 to 65%. ^[9,36,40-43] Variability in the reported frequencies of behavioral change among studies reflect differing inclusion criteria, methodologies, and referral populations, and the changing profile of behaviors as cognitive deterioration progresses.

We used a recently developed instrument, the Neuropsychiatric Inventory (NPI), in the current study to assess behavioral abnormalities in AD patients. The NPI has proven validity and reliability. ^[44] The NPI is a convenient instrument that evaluates both the frequency and severity of the following 10 behaviors: delusions, hallucinations, agitation, dysphoria, anxiety, euphoria, apathy, disinhibition, irritability, and abnormal motor output. We determined the presence of changes in these behavioral domains and explored their relationship to each other and to cognitive decline.

Methods.

The study subjects consisted of 50 consecutive outpatients, presenting for dementia evaluation at the University of California, Los Angeles, or West Los Angeles Veterans Affairs Medical Center dementia clinics, who met all study criteria and were willing to be interviewed. Diagnostic evaluation included complete medical history, physical and neurologic examination, neuropsychological testing, magnetic resonance imaging of the brain, electro-encephalogram, and routine blood tests (including thyroid-stimulating hormone, vitamin B₁₂ level, and serological test for syphilis). All subjects met National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer's Disease and Related Disorders Association (NINCDS/ADRDA) criteria for probable AD. ^[45] In addition, inclusion criteria for dementia required that subjects have an acquired persistent decline involving at least three of the following domains: language, memory, visuospatial skills, cognition (calculation, abstraction, judgment, and so on), and emotion or personality. ^[46] Exclusion criteria were delirium, other active systemic or neurologic diseases, history of alcohol or substance abuse, history of head trauma with loss of consciousness, and history of psychiatric disorder preceding the onset of memory loss. An added selection criterion was that the primary caregiver must live with the patient and be willing to provide information necessary for scoring the NPI.

Caregivers were interviewed with the NPI by following procedures previously described ^[44] in which screening questions for each behavior were posed first. The caregiver was asked if the behavior represented a change from that exhibited by the patient before the onset of the dementia and was present during the past month. If a positive response was obtained from the screening questions, then the behavioral domain was explored with scripted questions that focused on specific features of the behavioral disturbance. The caregiver then rated the behaviors; scores from 1 through 4 were obtained for the frequency and 1 through 3 for the severity of each behavior (a composite score for each domain was the product of the frequency and severity subscores, maximum = 12). The Mini-Mental State Examination ^[47] (MMSE) was also administered within 1 month of the NPI, in most cases the two studies were administered on the same day.

Controls were 40 residents from a retirement community (Leisure World) in Southern California. Spouses of nondemented individuals were interviewed with the NPI and their partners were examined with the MMSE on the same day. Inclusion criteria for the controls were a score of 25 or above on the MMSE and the absence of any intellectual-decline (as reported by the spouse). Comparable with the patient caregivers, control interviewees were asked about behaviors that differed from the spouse's usual behavior and had occurred within the previous 4 weeks.

To better compare the change in behaviors as disease progresses and cognition declines, patients were divided into three groups based on their MMSE scores (scores 0 to 10 = Severe; 11 to 20 = Moderate; 21 to 30 = Mild). For each behavior, mean subscores were compared among the three MMSE groups via analysis of variance and significance was established using the Tukey least significant difference criterion. Spearman correlations among the 10 NPI behaviors in the entire group were computed for the frequency, severity, and composite subscores for each behavioral domain.

Results.

The mean age of the 50 subjects was 75 years (SD, 6.6 years); their mean MMSE score was 16.2 (SD, 7.5). Twenty-four were female with a mean age of 74 years (SD, 7.4 years) and a mean MMSE score of 16.3 (SD, 7.4); 26 were male with a mean age of 75.3 years (SD, 5.9 years) and a mean MMSE score of 16.3 (SD, 7.7). The mildly impaired group (n = 17) was composed of eight males and nine females, mean age 75 years (range, 65 to 88 years) with a mean MMSE score of 23.9 (range, 21 to 29); the moderately impaired group (n = 20) was composed of 12 males and eight females, mean age 75 years (range, 56 to 88 years) with a mean MMSE score of 16.1 (range, 12 to 20); the severely impaired group (n = 13) was composed of six males and seven females with a mean age of 73.2 (range, 63 to 88 years) and a mean MMSE score of 5.7 (range, 0 to 10). The 40 control subjects were composed of 20 females with a mean age of 72.9 years (range, 63 to 88 years) and a mean MMSE score of 28.5 (range, 25 to 30); the 20 males had a mean age of 73.8 years (range, 62 to 86 years) with a mean MMSE score of 28.4 (range, 26 to 30). There was no significant difference between the ages of the patients and the control subjects.

Five AD patients in the mild group and one in the moderate MMSE group had no scorable behavioral change revealed by the NPI. Eighty-eight percent of AD patients had scorable psychopathology. The most common behavior was apathy, which was exhibited by 72% of all patients, followed by agitation (60%), anxiety (48%), irritability (42%), dysphoria (38%), aberrant motor behavior (38%), disinhibition (36%), delusions (22%), and hallucinations (10%). The least common behavior was euphoria expressed by only 8% of all AD patients. The percentage of AD subjects exhibiting the 10 NPI behaviors for the three MMSE groups are shown in Table 1.

The 40 control subjects had detectable changes in only three behavioral domains. Three control subjects had an increase in dysphoric behavior, two exhibited disinhibition, and one (a control subject with a positive dysphoria score) was also irritable. All three of these NPI scores were low and significantly less than the AD subjects (p < 0.01).

Age and sex were analyzed separately to determine if any behavior had a significant relationship with these variables. No significant correlation was found between the 10 behaviors and age. Analysis of the relationship between behavior and sex revealed a difference in one domain; i.e., agitation was significantly greater among male AD patients than female patients (mean composite score for males, 3.0; females, 1.2; p = 0.012).

The mean values for the total NPI scores frequency (maximum = 40), severity (maximum = 30), and composite (maximum = 120) for each MMSE group are shown in Table 2. Mean values for the total NPI scores increased across each MMSE group with a significant difference (p < 0.01) between the mild and severely impaired groups for all three NPI subscores; the severity score in the moderate group was significantly increased (p < 0.05) over the mildly impaired group.

To better understand the relationship of behavioral change to cognitive decline, we analyzed the NPI scores within the three MMSE groups. Mean frequency scores of each behavior (maximum = 4) and mean severity scores of each behavior (maximum = 3) were analyzed separately from the composite (product of frequency x severity; maximum = 12) scores for the three MMSE groups (data on file with the National Auxiliary Publications Service. See note at end of article). Frequency, severity, and composite subscale scores behaved similarly across the MMSE terciles and we describe only the relationship of the composite scores to the MMSE groups. Five composite scores (agitation, dysphoria, anxiety, apathy, and aberrant motor behavior) increased with dementia severity (Figure 1). Delusions and hallucinations decreased from the moderate to the severe group; euphoria and irritability were lowest in the moderate group, and disinhibition decreased across each group. Composite scores for agitation (p < 0.05), dysphoria (p < 0.05), and apathy (p < 0.01) showed a significant increase from the mild to the severe MMSE group, whereas aberrant motor behavior showed a significant increase in the severe group (p < 0.01) compared with the other cognitive groups.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Mean composite subscores of the 10 Neuropsychiatric Inventory behaviors across the three Mini-Mental State Examination (MMSE) groups. Del = delusions; Hal = hallucinations; Agit = agitation; Dys = dysphoria; Anx = anxiety; Euph = euphoria; Apa = apathy; Dis = disinhibition; Irrit = irritability; Motor = aberrant motor behavior; Mild = MMSE scores 21-30; Moderate = MMSE scores 11-20; Severe = MMSE scores 0-10. dagger significantly different from all groups (p < 0.01); # significantly different from mild group (p < 0.05); double dagger significantly different from mild group (p < 0.01).

To explore the relationships among behaviors, Spearman correlation coefficients were calculated for the composite scores of each behavior in a 10 x 10 matrix for all the AD patients. This exploratory analysis was not meant to establish strict significance, and thus a false positive (type 2 error) was deemed preferable to missing a possible correlation (a false negative or type 1 error). For this reason no Bonferroni adjustment was made; but an adjusted alpha level of 0.01 was required for significance. Agitation correlated with irritability (r = 0.51; p = 0.0002). Dysphoria correlated with anxiety (r = 0.40; p = 0.004) and irritability (r = 0.41; p = 0.003). Apathy correlated with abnormal motor output (r = 0.36; p = 0.01). Disinhibition correlated with irritability (r = 0.38; p = 0.007).

Discussion.

The present study demonstrated that noncognitive behavioral changes are ubiquitous in AD, occurring in 88% of a typical clinical population. The range of psychopathology observed in AD is wide; i.e., all 10 behaviors assessed by the NPI were present in at least a few of the AD patients. This contrasts with the findings in normal control subjects in whom changes in behavior were rare and of low intensity. Apathy was the most frequent behavioral change in AD, occurring in 72% of patients. This was followed by agitation, which occurred in 60%, and anxiety in 48%; dysphoria and aberrant motor behavior were also common. Agitation, dysphoria, apathy, and aberrant motor behavior increased significantly with worsening cognitive function. The consistency across many studies identifying agitation, aggression, or wandering, ^[25,38,40] delusions, paranoia, and hallucination, ^[25,48] as significantly increasing with cognitive decline supports the validity of the current findings.

Delusions and hallucinations did not always occur together across the three MMSE groups. Although delusions increased in frequency and severity, hallucinations declined in the more severely impaired group. The independence of these symptoms may point to differing underlying pathophysiologic processes. The lower frequency of these symptoms in the current study compared with other studies may result from our strategy of sampling only behaviors that had occurred in the 4 weeks before the interview rather than throughout the entire course of a patient's illness.

Agitation, broadly defined to include aberrant motor behavior and physical or verbal aggression, poses a major clinical challenge. The NPI emphasizes non-compliance, refusal to cooperate with the caregiver, obstinence, resistance, crying, kicking, and being "hard to handle" within the agitation section. Aberrant motor behavior, ie, pacing, picking at clothing, and so on, are scored separately. As the disease progresses, agitation worsens. Determinants of this worsening agitation may be varied. There was a steady increase in agitation with worsening MMSE, from nearly 50 to 85% of patients. Agitation correlated with irritability when analysis included all the AD patients but showed correlations with other symptoms when analyzed across the three MMSE groups. For the mildly impaired patients, a correlation with anxiety, disinhibition, and irritability was found; the moderate group showed a correlation with delusions and hallucinations; no correlation with other behaviors was found in the severely impaired patients. These varied associations may reflect changing determinants of agitation across the spectrum of AD severity. By following both the frequency and severity of agitation with the NPI, the clinician may be able to accurately predict causative influences and treatment response.

Dysphoria increased from more than 10% in the mild group to more than 60% in the severe group. Dysphoria is not equivalent to depression and most of the patients in our study would not meet criteria for a major depressive episode. The low occurrence of major depression is consistent with past studies. The dysphoria domain of the NPI excludes features common to both dementia and depression (e.g., vegetative symptoms) and reflects the patient's mood state more specifically than commonly used depression rating scales. The criteria for major depression, as defined by the Diagnostic and Statistical Manual of Mental Disorders, ^[49] does not require sadness; the diagnosis can be made when the patient is anhedonic and manifests five other common features of depression. Anhedonia has many features of apathy, and thus patients with AD may meet the primary criterion for major depression because of the high prevalence of apathy in AD. Dementia also produces disturbances common in depression, i.e., weight loss, agitation, and insomnia, allowing many patients with AD to meet diagnostic criteria for major depression even in the absence of mood changes. The NPI facilitates the identification of mood changes by emphasizing crying and statements of worthlessness, hopelessness, and guilt and by excluding vegetative features from the dysphoria subscale. The current study found no relationship between dysphoria and apathy indicating that the two are dissociable and should not be used interchangeably when attempting to identify mood changes in AD patients. Dysphoria was correlated with anxiety and irritability in the current study; an association also noted in idiopathic depression ^[50] and Parkinson's disease. ^[51]

Anxiety has often been subsumed within the broader category of motor restlessness reported in 21 to 60% of AD patients. ^[22,36-38] The NPI defines anxiety as worried or frightened behavior displayed for no apparent reason, or tense and fidgety behavior. In the current study, anxiety correlated with dysphoria, and in mildly impaired patients, with agitation. Although pharmacologic treatment for anxiety in dementia has not been systematically addressed, anxiolytic agents may also be useful for ameliorating some of the troublesome behaviors encountered.

Irritability, a common behavior in AD, was present in nearly one-half of the patients in this study and increased with progressive cognitive decline. Irritability is distinct from agitation, being defined as rapid emotional fluctuations between frustration and impatience with the patient becoming easily disturbed. Irritability correlated with agitation, dysphoria, and disinhibition but showed no correlation with agitation in the moderate and severe patients when analysis was conducted across MMSE groups. Because agitation, irritability, and aberrant motor behavior are major challenges in the care of the demented patient, monitoring their individual response to treatment is important. Aberrant motor behavior, although nearly certain to occur in the severely impaired patient, was rare in the mildly impaired group. If aberrant motor behaviors, such as pacing or other stereotypes, occur in the early stages of AD, a superimposed delirium should be suspected.

There was euphoria in only two cases in this study; its presence in a demented patient may suggest an alternate diagnosis (e.g., a frontotemporal dementia).

Most behavioral changes in AD are not determined by the degree of a patient's intellectual decline. Some patients with modest cognitive change demonstrated marked behavioral disturbances; whereas a few with significant cognitive decline showed no behavioral dysfunction (few behaviors showed continuous worsening across all three MMSE groups). In addition, behavioral abnormalities are varied and often transient throughout the course of AD, showing a more irregular temporal pattern than the continuous cognitive decline; some behaviors may be more common early in the disease whereas others occur later. The decrease in frequency or severity of behaviors such as delusions, hallucinations, and disinhibition with increasing cognitive dysfunction points to differing bases for these signs and symptoms. AD pathology may differentially disrupt behavioral and cognitive networks through the course of the disease.

A drawback of the present study is the lack of a longitudinal assessment of the behavioral changes in AD. Studying patients longitudinally will allow the characterization of the temporal progression of neuropsychiatric changes in individual AD patients. Although the NPI evaluates a wide spectrum of behavioral change, it does not sample all possible abnormal behaviors in neuropsychiatric disorders. Eating disorders, sexual changes, vegetative symptoms, and changes in diurnal patterns-all common in dementia-are not assessed by the NPI. Applying the NPI to other demented groups, such as frontotemporal dementias, progressive supranuclear palsy, or vascular dementia, will clarify the patterns of behavioral changes related to various neuropathologies. Understanding the relationship between neuropsychiatric changes and the patterns of cognitive and functional decline observed as a disease evolves enable a prediction of the character and rate of decline some behaviors may herald. Although our center evaluates patients referred for a second opinion or with difficult symptoms, most patients are brought in by family members for an initial evaluation; thus, extrapolating our results to AD patients treated in the community should be possible.

Patients suffering from AD manifest abnormalities in multiple domains; neurologic, cognitive, neuropsychiatric, and basic daily functions are all affected. Refining our understanding of the multidimensional aspects of AD is necessary to respond effectively to the demands of an aging society and the increasing number of dementia patients.

Note.

Readers can obtain 2 pages of supplementary material from the National Auxiliary Publications Service, c/o Microfiche Publications, PO Box 3513, Grand Central Station, New York, NY 10163-3513. Request document no. 05276. Remit with your order (not under separate cover), in US funds only, $7.75 for photocopies or $4.00 for microfiche. Outside the United States and Canada, add postage of $4.50 for the first 20 pages and $1.00 for each 10 pages of material thereafter, or $1.75 for the first microfiche and $.50 for each fiche thereafter. There is a $15.00 invoicing charge on all orders filled before payment.