|Home | About | Journals | Submit | Contact Us | Français|
To estimate the proportion of community dwelling older adults with dementia being prescribed a psychotropic and identify patient and caregiver factors associated with antipsychotics use.
Retrospective cohort study of The Aging, Demographics, and Memory Study (ADAMS) from 2002 to 2004 designed to assess dementia severity and service use among community-dwelling older adults. The frequency of psychotropic medication (antipsychotics, antidepressants, anticonvulsants and benzodiazepines) use was tabulated and weighted to the US population by dementia diagnosis. Logistic regression analysis identified factors associated with antipsychotic use.
The 307 participants of ADAMS had the following dementia diagnosis: Alzheimer’s disease (69.29%), vascular dementia (17.74%) and other dementia (12.39%). The proportion of participants prescribed a psychotropic medication broken down by therapeutic class was as 19.07% antipsychotics, 29.08% antidepressants, 9.84% benzodiazepines, and 8.85% anticonvulsants. Older adults with dementia were significantly more likely to receive an antipsychotic if they had moderate dementia (OR =7.4, p<0.05), or severe dementia (OR=5.80, p<0.05), compared to mild dementia or were diagnosed with Alzheimer (OR =6.7, p<0.05) dementia compared to vascular dementia.
Older adults with dementia who lived with their caregivers in were significantly less likely to be medicated with antipsychotics (OR= 0.19, p<0.05) compared to those who lived alone. Also, persons with dementia were significantly less likely to be prescribed an antipsychotic if the caregivers were clinically depressed (OR=0.03, p<0.05) compared to those who were not depressed.
We found psychotropic medication use is common among community-dwelling older adults with dementia. Caregivers appear to have a substantial impact on whether or not an antipsychotic is prescribed, which adds additional complexity to conversations discussing the risk-benefit ratio of this medication class.
The number of older adults afflicted with dementia is expected to increase three-fold by 2050. No cure for dementia exists and currently available therapies only slow progression for a limited period of time. Dementia is characterized by cognitive loss and occupational, social, and/or physical decline. Neuropsychiatric symptoms such as anxiety and depression, thought and perceptual problems, and activity disturbances –also accompaniment to the disease process with a prevalence reported between 60 to 80 percent.[4, 5]
Neuropsychiatric symptoms pose a substantial problem for persons with dementia and their families. The presence of neuropsychiatric symptoms has been associated with greater functional impairment, more rapid cognitive decline, and an overall poorer quality of life compared to persons without these symptoms. [6–8] At the same time, caregivers of persons with dementia exhibiting neuropsychiatric symptoms, report higher levels of depression and burden, in contrast to caregivers of persons with dementia without these symptoms. Medications such as antipsychotics, antidepressants, anticonvulsants, and benzodiazepines are frequently used to ameliorate neuropsychiatric symptoms, but the actual frequency of their use among community-dwelling persons with dementia residing in the community has not been well established.
Thought and perceptual problems and activity disturbances remain particularly problematic as no FDA approved medication currently exists for treatments and commonly used off-label therapies have noteworthy shortcomings. Controversy exists over the use of antipsychotics for neuropsychiatric symptoms in this population, due to the marginal increase in cerebrovascular events and death, even with low dose, short-term use. [11–13] However, debates surrounding the risk-benefit ratio of antipsychotic use have emphasized patient-related outcomes, while overlooking the caregiver’s health. Caregiver outcomes are extremely important to consider as poorly controlled neuropsychiatric symptoms are frequently cited as a determinant in placing loved ones into a nursing home.
The current investigation takes advantage of the Health and Retirement Study, a probability-based nationally-representative longitudinal study to address some of these current gaps in the literature. The first objective of the study is to quantify psychotropic medication use among community-dwelling persons with dementia. The second identifies patient and caregiver factors associated with antipsychotic use in this population. Taken together, the study findings detail the overall burden of psychotropic medication use among community-dwelling persons with dementia, and serve as a benchmark to compare practice patterns. Additionally, patient and caregiver factors associated with antipsychotic use are identified and help to better understand characteristics associated with its use in community-dwelling populations.
This retrospective cohort analysis that takes advantage of the Aging, Demographics, and Memory Study (ADAMS), a supplement to the Health and Retirement Study (HRS) conducted in waves 2002 and 2004, designed to assess dementia severity and service use. The HRS is a nationally representative, longitudinal study of health, retirement, and aging, which surveys the demographics, health problems, health care services use, family structure and health insurance for older adults, along with some caregiver information. Among HRS respondents surveyed in the 2002 Wave, 1,770 respondents aged 70 or older residing in the community were identified as having cognitive impairment [15, 16]. Among initial 1,770 respondents, 856 persons were recruited into ADAMS successfully and completed initial assessments [16, 17]. Of these, 307 received a diagnosis of dementia and represent our study sample .
The ADAMS in-person evaluation was a 3- to 4-hour structured assessment conducted in the participant’s residence by a nurse and neuropsychology technician, both specially trained in data collection for dementia evaluation. This protocol required participation of both the HRS respondent and an informant who was familiar with the participant’s daily activities and medical history, usually a close family member (e.g., spouse or child). The neuropsychology technician administered the neuropsychological battery to the participant, while the nurse met separately with the informant to obtain detailed information about the respondent, including cognitive and functional changes, medical and psychiatric history, current medication use, and current behavioral and psychiatric symptoms.
A geropsychiatrist, neurologist, neuropsychologist, and cognitive neuroscientist reviewed all information collected during the in-home assessment for each participant, and assigned a preliminary dementia diagnosis based upon the DSM-III-R and DSM-IV. Subsequently, the study geropsychiatrist reviewed available medical records and corroborated or revised the preliminary diagnosis .
Respondents were mailed a medication form prior to the visit and asked to have this information available when the assessment team arrived. Information collected on the form included names and dosages of prescription medication, vitamins/supplements, and other over-the-counter items taken in the previous two weeks. Other information included the purpose for which the medication was being taken, frequency of medication use, and how long the respondent had taken it. At the time of the clinical assessment, the nurse examined all medication containers to confirm the medication information. Participants were designated as taking an antipsychotic if the therapy was administered at any time, whether scheduled or as needed.
The medication file (drug files) contains drug name codes (AECODE). We matched these drug files with the participants other measures using their unique identifiers. Generic and brand name antipsychotics in the US were identified and matched to those listed in the participants drug file, see Appendix 1) [19, 20].
Demographic variables included age (younger than 85 years old, 85 years or older), race (White/non-Hispanic, black, Hispanic), gender, living arrangement (living with caregivers), and level of education (no formal, education to 1–11 years of education, high school or college graduate). Dementia diagnosis was categorized as Alzheimer’s disease, vascular dementia, and other dementia. Behavior issues were captured using the Neuropsychiatry Inventory (NPI) which includes an assessment of delusions, hallucinations, agitation/aggression, depression, apathy, elation, anxiety, disinhibition, irritability/lability, and aberrant motor behavior. Dementia stage was assessed using the Clinical Dementia Rating Scale (CDR) and participants were categorized as mild, moderate, or severe impairment). Physical function was ascertained by the number of Activity of Daily Living (ADL) and Instrumental Activity of Daily Living limitations (IADL). Other psychotropic medications categorized included antidepressants, anticonvulsants and benzodiazepines. Since antipsychotics are used by clinicians for a range of behavioral and psychological conditions in dementia, the total NPI score was used in the analysis of NPI. We used the total score of NPI in summing all items ranging from 0 to maximum 10 if they report any behavior problems of delusions, hallucinations, agitation/aggression, depression, apathy, elation, anxiety, disinhibition, irritability/lability, and aberrant motor behavior.
Dementia stage was assessed using the Clinical Dementia Rating Scale (CDR) which has been extensively validated and found reliable in community-dwelling populations. We categorized participants as having mild impairment (total CDR=0.5–1), moderate impairment (total CDR=2) and severe impairment (total CDR >=3) based upon the CDR total score.
The psychological well-being of the caregiver was measured using the short version of Center for Epidemiologic Studies Depression scale (CES-D) which includes . 8 items: “I felt depressed,” “I felt lonely,” “I felt sad”, “I was happy,” “I enjoy life,” “I felt everything I did was an effort,” “My sleep was restless,” and “I could not get going”. Response categories are “Yes” or “No” for each of the 8 items as caregivers consider their emotions over the past week. Scores are summed with a point for each “Yes” response, except the two well-being items which are reverse coded. The validity and reliability of the short version of the CES-D has been established, with a cutoff point of three or greater indicative of “active depression” [23, 24].
We again used the unique drug codes to other psychotropic medications frequently prescribed for behavioral issues in persons with dementia including antidepressants, anticonvulsants and benzodiazepines Available generic and brand names from each therapeutic class were identified and matched to each participant, see Appendix 1[19, 26].
Descriptive statistics were used to characterize the study population. Dementia diagnosis was separated into Alzheimer’s disease, vascular dementia, and other dementia. The proportion of participants by dementia diagnosis for each therapeutic class of psychotropic (antipsychotics, antidepressants, anticonvulsants, and benzodiazepines) was calculated and subsequently weighted to the overall US population of people age 70 and over residing in the community in the year 2000. We excluded participants categorized as “other dementia” from the multivariate analysis as only 279 participants were classified into this group which represented a wide range of diagnosis (e.g. alcohol related dementia, frontotemporal dementia, mixed dementia, and Lewy body dementia, etc.). Multivariate analysis employed logistic regression to examine the relationship between participant and caregiver factors and whether or not an antipsychotic was prescribed. Factors under consideration were selected a priori based upon literature review and expert consensus with the goal of generating the most clinical relevant and parsimonious model. We examined Akaike Information Information (AIC) and Schwarts Criterion (SC) values as criteria to assess model fit and collinearity was tested when we developed model using stepwise method. Statistical analysis was conducted using Stata Version 10.
The study participants, 307 community-dwelling older adults with a diagnosis of dementia, are summarized in Table 1. The majority of the population was female (68.30%), widowed (70.40%), and white (83.32%). The average age was 84.42years and approximately half had a high school diploma or college or graduate degrees. The most common dementia diagnosis was Alzheimer’s disease (69.29%), followed by vascular dementia (17.74%) and the remainder were classified as “other dementia” (12.39%). A total of 11.02% of caregivers were clinically depressed and a third of participants lived with a caregiver.
Community-dwelling older adults with a diagnosis of dementia and took antipsychotics had significantly greater physical disability in Activities of Daily Living (ADL) (3.64 vs. 3.02. p<0.05), and Instrumental Activities Daily Living (IADL) (4.31 vs.3.43, p<0.01) along with more severe dementia as measured by Cognitive Function Clinical Dementia Ration Score (CDR)(2.16 vs. 1.35, p<0.05) compared to persons with dementia who did not take antipsychotics. Those who took antipsychotics also had caregivers report more symptoms of apathy (32.26% vs. 20.12, p<0.01), and agitation (37.40% vs. 18.54%, p<0.001) compared to those who did not take antipsychotics. Also, caregivers depressive symptom were significantly less when persons with dementia were taking an antipsychotic (3.63 % vs. 12.75, p<0.001).
Table 2 displays the frequency of psychotropic medication use by dementia diagnosis. The proportion of participants prescribed a psychotropic medication broken down by therapeutic class was as follows: 19.07% antipsychotics, 29.08% antidepressants, 9.84% benzodiazepines, and 8.85% anticonvulsants.
The most frequently prescribed antipsychotics were olanzapine(7.88%), risperidone (6.65%), quetiapine (3.25%), and haloperidol (0.56%). The most common prescribed antidepressants were sertraline (8.79%), paroxetine (6.16%), and citalopram (3.39%). Phenytoin (5.03%) was the most frequent anticonvulsant and lorazepam (4.40%) and temazepam (2.84%) the most frequently prescribed benzodiazepines.
Multivariate logistic regression was used to examine patient and caregiver factors associated with antipsychotic use
Community-dwelling older persons with dementia were significantly more likely to receive an antipsychotics if they had moderate dementia severity (OR =7.4, p<0.05), or severe dementia (OR=5.80, p<0.05), or were diagnosed with Alzheimer (OR =6.7, p<0.05).
However, those who were older than 85 years old were significantly less likely to receive antipsychotic medications (OR=0.33 p<0.05). Participants who lived with their caregivers were significantly less likely to be medicated with antipsychotics (OR= 0.19, p<0.05). Dementia patients in the community were significantly less likely to be medicated with antipsychotics if the caregivers were clinically depressed (OR=0.03, p<0.05). Caregiver depression did not very significantly between those that did and did not live with the person with dementia for either those taking or not taking an antipsychotic, p=0.79 and 0.38, respectively.
Psychotropic medications were commonly prescribed for community-dwelling older adults with dementia. Antidepressants and antipsychotics represented the most commonly reported drug classes, in which approximately one in four and one and five participants took these therapies, respectively. Selective serotonin reuptake inhibitors represented the majority of antidepressants reported, whereas neuroleptic use was predominately atypical antipsychotics. Antipsychotic use was significantly more likely to be used among persons with more advanced dementia (moderate or severe compared to mild) and in participants diagnosed with Alzheimer dementia compared to vascular dementia. Caregivers of community-dwelling patients with dementia prescribed antipsychotics reported significantly lower rates of depression than those caring for persons not on this therapy. If the dyad lived together, antipsychotic use was significantly less likely compared to when they lived apart.
Our results suggest that persons with dementia experience substantial neuropsychiatric symptoms as evidence by the widespread use of psychotropic medications. Our findings mirror the experience of the United Kingdom which recently found a similar prevalence of psychotropic medication use in community-dwelling dementia populations. The percentage of persons prescribed a psychotropic from the reported drug classes was 28.7% for antidepressants, 17.7% for antipsychotics, and 16.7% for benzodiazepines. One key difference between studies was the proportion of persons on typical versus atypical antipsychotics. That is, less than 1% of persons with dementia residing in the community were prescribed a typical antipsychotic in the United States versus 37.5% in the United Kingdom. Potential reasons for this divergence in prescribing patterns may reflect differences in perceptions of efficacy and/or adverse effects, variations in pharmaceutical marketing, and/or insurance coverage for specific therapies.
The proportion of persons with dementia on a psychotropic in the community appears to be less than that reported in nursing homes. For example, a recently published study by Stevenson and his colleagues that used a nationally representative sample of nursing home residents reported 26% of residents on an antipsychotic and 13% a benzodiazepine. However, the actual difference in the proportion of community-dwelling and nursing home residents with dementia prescribed an antipsychotic or benzodiazepine was less than anticipated, questioning whether there is an overuse of these therapies among persons with dementia residing in the community. This finding questions whether the greater regulatory scrutiny for psychotropic medication use that occurs in nursing homes should be considered for adoption in the community as well.
Much debate has transpired over the last several years regarding the use of antipsychotics in the management of neuropsychiatric behaviors in persons with dementia. The preponderance of evidence from randomized controlled trials suggests that this class of medications is moderately effective for the treatment of psychotic symptoms and agitation and aggression in persons with dementia [29, 30]. The CATIE-AD study, in particular, was designed to assess the clinical efficacy of atypical antipsychotics (olanzapine, quetaipine, and risperdone) versus placebo in persons with dementia residing in the community that displayed clinically significant neuropsychiatric behaviors. The primary outcome was time to discontinuation of the assigned therapy for any reason with secondary outcomes including discontinuation due to lack of efficacy and adverse events or intolerability. While time to discontinuation for any reason did not vary among the groups, discontinuation because of lack of efficacy significantly favored placebo and quetiapine. At the same time, discontinuation of therapy due to adverse effects and intolerability favored the antipsychotic groups. This study supports the efficacy of antipsychotics at extremely low doses for neuropsychiatric behaviors in the setting of adverse effects that may offset this advantage. Caregiver outcomes were not considered as part of the main outcomes for this study or much of the research published in this area to date.
We found that caregiver depression to be significantly less likely if the person with dementia for which care was provided was prescribed an antipsychotic. This finding highlights the importance of measuring and considering caregiver-related outcomes when assessing the effectiveness of dementia-related interventions. We hypothesize that antipsychotics decrease neuropsychiatric symptoms, resulting in improvements in caregiver’s psychological well-being. The results fit within a growing literature examining the relationship between neuropsychiatric symptoms and caregiver well-being [31–39]. For example, a recent meta-analysis reported that pharmacologic treatment of neuropsychiatric symptoms decreases caregiver burden, and the use of antipsychotics specifically was associated with a decrease in behavioral disturbances while simultaneously reducing caregiver burden. The decline in caregiver burden is thought to contribute to the beneficial effects found in caregiver mood. The findings also suggest that studies designed to ameliorate neuropsychiatric symptoms should consider the incorporation of both patient and caregiver outcomes, as interventions may differentially impact each member of the dyad.
Strengths of the study include its representative national sample of community-dwelling older adults and the inclusion of prescription drugs. There are several weaknesses to the study that should be considered. First, the data was collected prior to the black-box warning associating antipsychotic use with cerebrovascular events and death, which may have modified prescribing practices over time. However, the study’s findings remain relevant as no single therapy is FDA approved for neuropsychiatric symptoms in persons with dementia and a dearth of research is available on psychotropic medication use in this population. Second, interviewers collected prescription medication use and this may not reflect what the patient is actually taking in the home. Third, the participation rate of 56% may impact the generalizability of the findings; it is comparable with other epidemiologic studies of older adults. Finally, the current study is cross-sectional in nature and cannot help determine if psychotropic medication use is increasing or decreasing over time for community-dwelling persons with dementia. At the same time, we are unable to establish directionality or causality among observed associations.
We found psychotropic medication use is common among community-dwelling older adults with dementia. Antipsychotic use may not only benefit patients, but also positively impact a caregiver, which adds additional complexity to conversations discussing the risk-benefit ratio of this medication class. Evidence-based comparative effectiveness research for the management of neuropsychiatric symptoms incorporating patient and caregiver outcomes is sorely needed.
Dr. Rhee received a Ken and Ruth Davee award from Department of Psychiatry and Behavioral Sciences of Feinberg School of Medicine at Northwestern University. Dr. Shega received a career development award from the National Institute on Aging (K23AG029815).
Sponsor’s Role:. Funding agencies had no role in the design and conduct of the study; data collection, management, analysis, and interpretation of the data; and preparation, review or approval of the manuscript.
Primary findings were presented at the Annual Meeting of ISPOR 13th Annual European Congress, November 2010 at the Prague Congress Centre in Prague, Czech Republic
Conflict of Interest
Author Contributions: Dr. Rhee: concept, design, data acquisition, analysis, interpretation of data, and preparation of manuscript. Dr. Chang: analysis, interpretation of data, and input for statistical methods Dr. Csernansky: interpretation of data and preparation of manuscript. Dr. Shega: concept, design, interpretation of data, and preparation of manuscript.