- Email: [email protected]
Declining Financial Capacity in Patients With Mild Alzheimer Disease: A One-Year Longitudinal Study
Declining Financial Capacity in Patients With Mild Alzheimer Disease: A One-Year Longitudinal Study Roy Martin, Ph.D., H. Randall Griffith, Ph.D., Katherine Belue, Lindy Harrell, M.D., Ph.D., Edward Zamrini, M.D., Britt Anderson, M.D., Alfred Bartolucci, Ph.D., Daniel Marson, J.D., Ph.D.
Objective: The objective of this study was to investigate change over time in financial abilities in patients with mild Alzheimer disease (AD). Methods: The authors conducted a prospective 1-year longitudinal study at a large southern U.S. metropolitan-area medical school university. Participants included healthy older adults (N⫽63) and patients with mild AD (N⫽55). The authors conducted a standardized performance measure of financial capacity. Performance was assessed on 18 financial tasks, nine domains of financial activity, and overall financial capacity. Capacity outcomes classifications (capable, marginally capable, or incapable) for domains and overall performance were made using cut scores referenced to comparison group performance. Results: At baseline, patients with mild AD performed significantly below healthy older adults on 16 of 18 tasks, on all nine domains, and on overall financial capacity. At one-year follow up, comparison group performance was stable on all variables. In contrast, patients with mild AD showed substantial declines in overall financial capacity, on eight of nine domains, and on 12 of 18 tasks. Similarly, the proportion of the mild AD group classified as marginally capable and incapable increased substantially over one year for the two overall scores and for five financial domains. Conclusions: Financial capacity is already substantially impaired in patients with mild AD at baseline and undergoes rapid additional decline over one year. Relative to the comparison group, overall financial capacity performance in the AD group declined 10%, from approximately 80% of the comparison group performance at baseline to 70% at follow up. Financial skills showed differential rates of decline on both simple and complex tasks. Of clinical and public policy interest was the declining judgment of patients with mild AD regarding simple fraud schemes. The study supports the importance of prompt financial supervision and planning for patients newly diagnosed with AD. (Am J Geriatr Psychiatry 2008; 16:209–219) Key Words: dementia, financial capacity, Alzheimer’s disease, longitudinal research
P
rogressive cognitive and functional decline are hallmark clinical features of Alzheimer disease
(AD).1,2 Longitudinal investigations of cognitive change have provided important insights into the
Received April 25, 2007; revised June 26, 2007; accepted July 18, 2007. From the Department of Neurology and Alzheimer’s Disease Research Center (RM, HRG, KB, LH, DM) and the Department of Biostatistics (AB), University of Alabama at Birmingham, Birmingham, AL; VA Medical Center (LH), Birmingham, AL; the Department of Neurology (EZ), University of Utah Health Sciences, Salt Lake City, UT; and Brown University (BA), Providence, RI. Send correspondence and reprint requests to Dr. Daniel Marson, Department of Neurology, SC 650, University of Alabama at Birmingham, Birmingham, AL 35294-0017. e-mail: [email protected] © 2008 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 16:3, March 2008
209
Declining Financial Capacity in AD natural history of the disease.3,4 In contrast, few studies exist investigating changes in independent activities of daily living.5 This is surprising given the importance of independent activity of daily living change to dementia diagnosis1,6 and to clinician management of patients with dementia.7 Although cognitive and functional decline in dementia are by definition related, a single cognitive measure or domain may not be associated with observed levels of functional disability.8,9 In addition, it is loss of functional abilities rather than cognitive decline per se that causes patient disability and burden for family members and caregivers.10 Thus, understanding the patterns, timeframes, and rates of functional decline over time in AD has both scientific and clinical importance. Of particular interest to our group has been impairment and eventual loss of financial abilities in dementia.2,10 Financial capacity is a complex, multidimensional independent activities of daily living, which has critical relationships to both functional independence and personal autonomy in older adults.11,12 Cross-sectional studies have shown that financial capacity is highly sensitive to AD across its continuum—from mild cognitive impairment13 to mild and moderate AD.2,10,14 Thus, financial capacity is a valuable prism through which to study and understand functional decline in AD. In this study, we used a direct performance measure to investigate change in financial abilities over a oneyear period in patients with mild AD. We hypothesized that 1) financial capacity would show significant decline over one year and 2) that decline would occur primarily on complex rather than simple tasks.
METHODS Conceptual Model of Financial Capacity Financial capacity involves a broad range of declarative, procedural, and judgment-based knowledge and skills.2,10 Our conceptual model views financial capacity at three levels: specific financial abilities (tasks); broader areas of financial activity (domains) that have clinical relevance for independent functioning (e.g., checkbook management; judgment concerning potential fraud); and overall financial capacity (global). Tasks were defined as
210
simple or complex a priori based on their novelty and demand characteristics. Development of this model is discussed in detail elsewhere.2,10,14 A schematic of the model is presented in Table 1. Financial Capacity Instrument The Financial Capacity Instrument (FCI-9)2,13 is a standardized instrument developed by our group that is based on the previously described conceptual model and that directly assesses subjects’ financial abilities. The FCI consists of standardized tasks of varying difficulty level organized into domains (Table 1). The current version assesses 18 tasks within nine domains and has two overall scores.13 The 18 tasks are included in Domains 1–7, whereas Domains 8 and 9 are analyzed at the domain level only. The FCI uses detailed, intervallevel quantitative scoring to assess subject performance on tasks. Task scores are then summed to obtain domain and overall financial capacity scores. The FCI has demonstrated good to excellent internal, test–retest, and interrater reliabilities (25). Participants Participants consisted of 63 healthy older adults and 55 patients with mild AD. Our comparison group was cognitively and neurologically intact, healthy older adults who were community-dwelling and recruited into the University of Alabama at Birmingham Alzheimer’s Disease Research Center. Each subject was clinically evaluated by a neurologist (LEH, BA, or EYZ) to ensure the absence of medical and psychiatric conditions that could compromise cognition. Subjects were characterized as cognitively normal after a diagnostic consensus conference involving the neurologists and a neuropsychologist (DCM). All of our older adult comparison group members received a Clinical Dementia Rating15 staging of 0.0. Patients with probable AD were communitydwelling individuals recruited into the Alzheimer’s Disease Research Center whose dementia was well characterized based on neurologic, neuropsychologic, and radiologic procedures. Diagnosis of probable AD was made in the diagnostic consensus conference and was based on National Institute of Neurological and Communicative Diseases and Stroke–Alzheimer’s Disease and Related Disorders
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al.
TABLE 1.
Revised Conceptual Model of Financial Capacity: 18 Tasks, Nine Domains, and Overall Capacity Task Description
Domain 1: Basic monetary skills Task 1a: Naming coins/currency Task 1b: Coin/currency relationships Task 1c: Counting coins/currency Domain 2: Financial conceptual knowledge Task 2a: Define financial concepts Task 2b: Apply financial concepts Domain 3: Cash transactions Task 3a: One-item grocery purchase Task 3b: Three-item grocery purchase Task 3c: Change/vending machine Task 3d: Tipping Domain 4: Checkbook management Task 4a: Understand checkbook Task 4b: Use checkbook/register Domain 5: Bank statement management Task 5a: Understand bank statement Task 5b: Use bank statement Domain 6: Financial judgment Task 6a: Detect mail fraud risk Task 6b: Detect telephone fraud risk Domain 7: Bill payment Task 7a: Understand bills Task 7b: Prioritize bills Task 7c: Prepare bills for mailing Domain 8: Knowledge of assets/estatea Domain 9: Investment decision-making Overall financial capacity a
Difficulty
Identify specific coins and currency Indicate monetary values of coins/currency Accurately count arrays of coins and currency
Simple Simple Simple
Define simple financial concepts Practical applications/computation using concepts
Complex Complex
Conduct one-item transaction; verify change Conduct three-item transaction; verify change Obtain change for vending machine; verify charge Understand tipping convention; calculate tips
Simple Complex Complex Complex
Identify/explain parts of checkbook and register Conduct simple transaction and pay by check
Simple Complex
Identify/explain parts of a bank statement Identify specific transactions on bank statement
Complex Complex
Detect/explain risks in mail fraud solicitation Detect/explain risks in telephone fraud solicitation
Simple Simple
Explain meaning and purpose of bills Identify overdue utility bill Prepare bills, checks, envelopes for mailing Indicate personal assets and estate arrangements Understand investment options; determine returns; make and explain decision Overall functioning across tasks and domains
Simple Simple Complex Simple Complex Complex
Requires corroboration by informant.
Association criteria.1 All patients with probable AD in this study were classified as having mild dementia based on a consensus conference agreement and Clinical Dementia Rating staging15 of either 0.5 or 1.0 (Table 2). Informed consent was obtained from all comparison group participants and from patients with mild AD and their participant caregivers in accordance with this Institutional Review Board-approved research.
Data Exclusion: Prior/Premorbid Financial Experience and Missing Data Because individual financial experience can differ across subjects,2 we accounted for lack of prior financial skills and experience in our participants. The Prior/ Premorbid Financial Capacity Form,2 a rating measure identifying a subject’s prior (comparison group) or premorbid (patient with AD) experience across the 18 tasks and nine domains of the financial capacity model,
Am J Geriatr Psychiatry 16:3, March 2008
was administered to all study participants and to their informants (not all comparison group participants had an informant available). Premorbid/prior experience for each task and domain was rated in one of three categories: “could do without help,” “could do but needed help,” or “could not do even with help.” Any participant who either self-reported having no prior experience or ability on a domain or task or whose informant (i.e., spouse, family member) reported that the participant was previously incapable of handling a domain or task was excluded from the analysis for that specific domain or task, although data for that domain were collected. In the case of discrepant data (i.e., a caregiver rating a participant as incapable but the participant endorsing capacity for the same domain or task), the caregiver’s judgment was used as the decision rule. The data exclusion procedures resulted in the exclusion of two persons from the comparison group for Domain 5 analysis as a result of having no prior
211
Declining Financial Capacity in AD
TABLE 2.
Demographic Characteristics of the Comparison Group and Patients With Mild AD Comparison Group (n ⴝ 63) (standard deviation)
Age (years) Gender (M/F) Race White Black Asian American Education Mini-Mental State Examination Dementia Rating Scale Global Deterioration Scalea 1 2 3 4 5 Geriatric Depression Scale
Patients With Mild AD (n ⴝ 55) (standard deviation)
p
66.3 (7.6) 20/43
70.6 (8.4) 31/24
0.005 0.007
53 9 1 14.4 (1.7) 29.3 (1.0)
50 5 0 13.1 (2.3) 24.5 (3.1)
0.26
136.7 (4.8)
116.2 (10.5)
0.001
38 24 0 0 0
0 3 21 21 9
0.001
7.4 (5.1)
0.004
4.1 (3.7)
0.001 0.001
⫽ 105.3, p ⬍0.001. AD: Alzheimer disease.
a 2
experience. In the mild AD group, the following exclusions or actual missing data at one-year follow up were identified: Domain 3 (N⫽1, missing data), Domain 4 (N⫽2, missing data), Domain 5 (N⫽6, excluded no prior experience), Domain 7 (N⫽7, four excluded and three missing data), Domain 8 (N⫽6, missing data), and Domain 9 (N⫽13, 11 excluded and two missing data). As expected, Domain 9 had the most data points excluded insofar as investment decision-making is a specialized, socioeconomic status-related skill, which a minority of our patient sample lacked. This outcome was very similar to the data exclusion results of a prior study using the FCI.13
Statistical Analyses The comparison group and mild AD group performances on the demographic and mental status variables were compared at baseline using independentsamples t test. Chi-square statistics were used for the gender and ethnicity variables. Group comparisons for the FCI-9 tasks, domains, and two FCI total scores (Domains 1–7; Domains 1– 8) were compared at baseline using analysis of covariance (ANCOVA). Covariates of age, education, and gender were included in the analyses. Pairedsample t tests were used within groups to examine performance change over one year across the nine domains and the 18 tasks. Interaction effects were examined using repeated-measures ANCOVA with group as the between-subjects variable and time as the within-groups variable. Age, education, and gender were included as covariates for this set of analyses. Alpha level was set at 0.01 to correct for multiple comparisons. All statistical analyses were conducted using SPSS 10.0.16 Capacity Outcome Analysis. Similar to previous studies,2 each patient with mild AD also received a capacity outcome rating (capable, marginally capable, incapable) for each of the nine domains and the two overall total scores. These capacity outcomes were determined by using psychometric cut scores referenced to mean comparison group performance. A capable outcome rating was assigned to a patient with mild AD if the patient’s score fell at ⱕ1.5 standard deviations (SDs) below the comparison group mean score for that domain; a marginally capable outcome rating reflected a patient score falling between 1.5 SDs and 2.5 SDs below the comparison group mean; and an incapable outcome rating reflected a score ⬎2.5 SD below the comparison group mean. Nonparametric analysis (Wilcoxon test16) was conducted to examine the change over one year in the proportion of patients receiving each of the three outcome ratings.
Test Administration Procedures Participants were administered the FCI-9 at baseline and at a 12-month follow-up evaluation. All portions of the FCI-9 were administered in a single session. The FCI took approximately 40 to 50 minutes to administer for comparison group participants and 50 – 60 minutes for patients with mild AD. The FCI has a standardized administration protocol and detailed scoring system.2,10,13
212
RESULTS Demographic and Mental Status Variables Table 2 displays demographic and mental status variables for the two groups. The comparison group was younger than patients with mild AD (t⫽2.89,
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al. df ⫽116, p ⬍0.005) by an average of four years. The patients with mild AD had one year less education (t ⫽ ⫺3.4, df ⫽119, p ⬍0.001). Male-to-female proportion ratios were different between the two groups (2 ⫽ ⫺2.68, p ⬍0.007) with a higher male-to-female ratio present in the mild AD group. Blacks made up 15% of the overall sample size. No between-group differences were found in the racial distribution (2 ⫽ ⫺1.1, p⫽0.26). As expected, patients with mild AD had lower scores on the Mini-Mental State Examination17 (t ⫽ ⫺12.1, df ⫽113, p ⬍0.001) and the Mattis Dementia Rating Scale (t⫽⫺14.7, df⫽116, p ⬍0.001).18 Patients with mild AD reported more depressive symptoms on the Geriatric Depression Scale19 (t⫽3.9, df⫽ 114, p ⬍0.001). However, the mean depression score for the patients with mild AD was well below the clinical cutoff for depression. Changes in Task Level Performance Table 3 sets forth group performance for the 18 financial capacity tasks (across Domains 1–7) at both assessment visits. At baseline assessment, the comparison group performed better than the mild AD group on 16 of the 18 financial capacity tasks. Patients with mild AD performed equivalently with the comparison group at baseline on only two simple tasks: naming coins/currency and a one-item grocery store transaction. Age, education, and gender were included as covariates in this set of analyses. Education level was a significant contributor to nine of the 18 tasks (coins/currency relationships, understanding financial concepts, applying financial concepts, one-item purchase, multiitem transactions, tipping, understanding checkbook, using checkbook, understanding bank statement), whereas gender contributed to two of the 18 tasks (understanding checkbook, using checkbook). At follow up using the paired-samples t test, patients with mild AD demonstrated within-group 1-year declines on 12 of the 18 tasks. In contrast, the comparison group demonstrated no significant performance change across any of the 18 tasks relative to baseline. The group by time interaction, as measured by ANCOVA, was significant for five of 18 tasks representing steady decline for the patients with mild AD and no change for the comparison group. Across the 18 tasks, the covariates contributed to
Am J Geriatr Psychiatry 16:3, March 2008
only two of the 18 tasks. Age had a modest effect on the applying concepts task (F (1,113) ⫽4.6, p ⬍0.035), whereas gender had a modest effect on the understanding bank statements task (F (1,105)⫽4.1, p ⬍0.05). The percentage decline in raw scores from baseline to one-year follow up ranged from 3% (naming coins/currency) to 29% in the mild AD group (using bank statement, preparing bills for mailing).
Changes in Domain and Overall Capacity Performance Table 3 also displays group performance on the FCI-9 domains and two overall scores. At baseline, ANCOVA revealed that the comparison group performed better than patients with mild AD across all nine domains and both overall scores (Domains 1–7, Domains 1– 8). At baseline, covariates were significant contributing factors in seven of nine domains (education level—seven domains, gender— one domain). However, no covariate affected significance values for the between-groups analyses. At one-year follow up, patients with mild AD demonstrated declines on eight of the nine domains (only Domain 8 did not show a decline) and on both overall scores. In contrast, the comparison group demonstrated stability across all nine domains and both overall scores. Group-by-time interactions revealed a decline in the mild AD group for four of nine domains and both overall scores. The covariate age contributed to Domain 8, but no other covariate contributed to the other domains.
Financial Capacity Instrument-9 Variable Difficulty and Annual Change in the Mild Alzheimer Disease Group We examined the difficulty of FCI-9 variables for patients with mild AD at baseline and at one-year follow up (Table 4). A variable’s difficulty for patients with AD was estimated as the percentage of comparison group performance attained at that time point. For example, the difficulty level for patients with mild AD of task 7c (preparing bills for mailing) at baseline was estimated at 63% (mild AD patients’ baseline mean raw score of 16.0 divided by the comparison group’s baseline mean raw score of 25.5) (Table 3). Using the baseline and 1-year follow-up
213
Declining Financial Capacity in AD
TABLE 3.
Change in Financial Abilities Over One Year in the Comparison Group and in Patients With Mild AD Comparison Group
Domain 1: Basic monetary skills Naming coins/currency Coins/currency relationships Counting coins/money Domain 2: Financial concepts Understanding concepts Applying concepts Domain 3: Cash transactions One-item transaction Multi-item transaction Vending machine Tipping Domain 4: Checkbook management Understanding checkbook Using checkbook Domain 5: Bank statement management Understanding bank statement Using bank statement Domain 6: Financial judgment Mail fraud Telephone fraud Domain 7: Bill payment Understanding bills Identifying/prioritizing bills Preparing bills for mailing Domain 8: Knowledge assets/estate Domain 9: Investment decision-making Total score (Domains 1–7) Total score (Domains 1–8)
Score Range
Time 1 (n ⴝ 63)
Time 2 (n ⴝ 63)
0–48
46.0 (2.9)
0–8 0–28 0–12 0–40
Patients With AD pa
Time 1 (n ⴝ 55)
Time 2 (n ⴝ 55)
pb
Baseline
Interaction
46.2 (2.7)
0.62
41.6 (5.7)
37.5 (8.5)
0.001
.001
.001
7.9 (.34) 26.2 (2.7) 11.9 (.35) 36.0 (3.2)
7.9 (.27) 26.4 (2.5) 11.9 (.42) 35.3 (4.1)
0.41 0.63 0.64 0.12
7.7 (.48) 22.2 (5.3) 11.5 (1.2) 26.9 (6.2)
7.5 (.96) 19.2 (7.0) 10.9 (2.4) 23.9 (7.2)
0.08 0.007 0.001 0.001
.16 .001 .009 .001
.08 .001 .10 .03
0–15 0–25 0–30
13.5 (1.3) 22.5 (2.8) 27.4 (2.9)
13.0 (1.4) 22.4 (3.2) 27.7 (2.6)
0.03 0.74 0.42
10.5 (2.4) 16.5 (4.7) 21.8 (4.6)
10.0 (3.0) 13.9 (5.3) 18.8 (5.7)
0.20 0.001 0.007
.001 .001 .001
.42 .002 .001
0–6 0–7 0–9 0–8 0–54
5.9 (.51) 6.4 (1.4) 8.6 (1.0) 6.4 (1.6) 53.0 (1.5)
5.9 (.43) 6.4 (1.2) 8.8 (.76) 6.6 (1.5) 52.4 (2.3)
0.83 0.78 0.12 0.49 0.08
5.6 (.99) 4.8 (1.8) 6.4 (2.2) 5.0 (1.9) 42.2 (10.0)
5.1 (1.4) 3.7 (2.1) 5.8 (2.3) 4.2 (2.0) 39.1 (11.9)
0.004 0.001 0.09 0.001 0.007
.03 .001 .001 .001 .001
.03 .001 .02 .03 .04
0–24 0–30 0–38
23.6 (.71) 29.3 (1.4) 34.4 (3.1)
23.4 (.96) 28.9 (2.0) 33.8 (3.4)
0.28 0.14 0.10
20.8 (2.8) 21.3 (7.7) 24.5 (7.8)
19.4 (4.4) 19.7 (8.8) 19.2 (8.5)
0.006 0.03 0.001
.001 .001 .001
.03 .13 .001
0–18 0–20 0–26
15.5 (2.0) 18.9 (1.7) 24.9 (2.6)
15.1 (2.1) 18.7 (1.8) 24.7 (2.4)
0.18 0.13 0.35
11.5 (3.2) 13.0 (5.3) 20.9 (5.9)
10.0 (3.9) 9.2 (5.3) 17.9 (7.0)
0.001 0.001 0.001
.001 .001 .001
.004 .001 .02
0–8 0–18 0–46 0–6 0–13 0–27 0–20
7.8 (.78) 17.1 (2.0) 43.8 (4.1) 5.7 (.74) 12.6 (.69) 25.5 (3.7) 18.0 (2.1)
7.8 (.73) 16.9 (2.1) 43.0 (4.8) 5.5 (1.0) 12.4 (.93) 24.9 (4.9) 18.6 (1.4)
0.78 0.34 0.20 0.34 0.24 0.37 0.12
6.5 (2.2) 14.4 (4.3) 31.6 (9.6) 4.6 (1.5) 11.0 (1.6) 16.0 (8.4) 15.8 (2.9)
5.4 (2.7) 12.5 (5.3) 27.6 (11.6) 4.1 (1.8) 10.3 (2.4) 13.0 (10.0) 15.9 (2.9)
0.001 0.006 0.003 0.11 0.06 0.01 0.72
.001 .001 .001 .001 .001 .001 .009
.02 .13 .06 .63 .24 .11 .22
0–17
13.3 (2.8)
13.3 (2.5)
0.89
9.6 (3.2)
7.9 (3.4)
0.001
.001
.005
0–282 0–302
266.1 (12.4) 285.0 (9.7)
263.4 (13.9) 281.5 (13.5)
0.04 0.06
209.4 (37.4) 230.5 (34.8)
183.2 (50.8) 203.4 (45.5)
0.001 0.001
.001 .001
.001 .001
Note: Not all domains had all data points for patients with mild AD or the comparison group as a result of missing data at year 1 evaluation or data exclusion based on no prior/premorbid experience for that financial ability. Comparison group sample size equal to 63 at year 1 evaluation except for Domain 5 (n ⫽ 61), Domain 8 (n ⫽ 42), and Domain 9 (n ⫽ 58). Mild AD sample size equal to 55 at year one evaluation except for Domain 5 (n ⫽ 49), Domain 7 (n ⫽ 48), Domain 8 (n ⫽ 49), and Domain 9 (n ⫽ 42). a No significant t values (p ⬎0.01) for comparison group difference across baseline and one-year follow-up assessments. Paired-samples t test was the statistic used. Degrees of freedom were as follows: Domains 1– 4, 6, 7 (62), Domain 5 (60), Domain 8 (41), and Domain 9 (57). b All significant t values ⬎2.6 (p ⬍0.01) for mild AD group difference across baseline and one-year follow-up assessments. Paired-samples t test was the statistic used. Degrees of freedom as follows: Domain 1– 4 (54), Domain 5 (48), Domain 6 (54), Domain 7 (47), Domain 8 (48), and Domain 9 (41). Baseline ⫽ all significant F values ⬎4.47 (p ⬍0.01) for comparison group and mild AD group differences at baseline assessment. Total degrees of freedom (df) for tasks on Domains 1– 4 and 6 ⫽ 118; df for tasks on Domains 5 ⫽ 110, df for tasks on Domain 7 ⫽ 114, df for tasks on Domain 8 ⫽ 101, and df for tasks on Domain 9 ⫽ 102. Analysis of covariance was the statistic used. Interaction ⫽ all significant F values ⬎8.8 (p ⬍0.01) for the group-by-time interaction statistic. Total df for tasks on Domains 1– 4 and 6 ⫽ 118; df for tasks on Domain 5 ⫽ 110, df for tasks on Domain 7 ⫽ 114, df for tasks on Domain 8 ⫽ 101, and df for tasks on Domain 9 ⫽ 102. A repeated-measures analysis of covariance was used. AD: Alzheimer disease.
difficulty values, a percentage annual decline value was then derived, which represented the difference between the two difficulty values. In the example of
214
task 7c, this annual decline for patients with mild AD was ⫺11% (52% (one-year follow-up difficulty) to 63% (baseline difficulty).
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al.
TABLE 4.
FCI Variable Difficulty and Annual Percentage Change Within the Mild AD Group (n ⴝ 55)
Variables Task 7c: Preparing bills for mailing Task 5b: Using bank statement Task 2b: Applying concepts Task 4b: Using checkbook/register Task 3c: Vending machine Task 5a: Understanding bank statement Task 3b: Multi-item transaction Task 2a: Understanding concepts Task 3d: Tipping Task 7a: Understanding bills Task 6a: Mail fraud Task 6b: Telephone fraud Task 1b: Coin/currency relationships Task 7b: Prioritizing bills Task 4a: Understanding checkbook/register Task 3a: One-item transaction Task 1c: Counting coins/currency Task 1a: Naming coins/currency Domain 7: Bill payment Domain 5: Bank statement management Domain 9: Investment decision-making Domain 2: Financial concepts Domain 3: Cash transactions Domain 4: Checkbook management Domain 6: Financial judgment Domain 8: Knowledge assets/estate Domain 1: Basic monetary skills Total score (Domains 1–7) Total score (Domains 1–8)
Baselinea
One- Yearb Follow Up
Annual Percentage Changec
63 69 73 73 74 74 75 78 78 80 83 84 85 87 88 95 97 97 70 71 72 75 80 80 84 88 90 79 81
⬎52 49 62 68 66 66 58 77 64 76 69 74 73 83 83 86 87 95 64 57 59 68 68 75 72 85 82 69 72
–11 –20 –11 –05 –08 –08 –17 –01 –14 –04 –14 –10 –12 –04 –05 –09 –10 –02 –06 –14 –13 –07 –12 –05 –12 –03 –08 –10 –09
a Difficulty value is the percentage of comparison group performance attained by the mild AD group at baseline. Variables are listed in descending order of difficulty. b Difficulty value is the percentage of comparison group performance attained by the mild AD group at one-year follow up. c Change value is the decline in percentage of comparison group performance attained by the mild AD group from baseline to one-year follow up. Change values ⬎⫺10% are bolded. FCI: Financial Capacity Instrument; AD: Alzheimer disease.
At the overall capacity level, the mild AD group achieved approximately 80% of comparison group performance at baseline. Over one year, the AD group declined a further 9%–10% relative to the comparison group performance. At the domain level, difficulty values differed substantially, ranging from a high of 70% for bill payment to a low of 90% for basic monetary skills. One-year changes were greatest for bank statement management (⫺14%), investment decision (⫺13%), cash transactions (⫺12%), and financial judgment (⫺12%). Knowledge of assets/estate arrangements showed the least change (⫺3%). Similarly, at the task level, there was variability in difficulty values at baseline, ranging from a high of 63% for preparing bills to a low of 97% for naming coins/currency and also for counting coins/ currency. One-year changes were greatest for using
Am J Geriatr Psychiatry 16:3, March 2008
bank statement (⫺20%) and a multi-item grocery store purchase (⫺17%) and least for understanding concepts (⫺1%) and naming coins/currency (⫺2%). Relationship of Capacity Performance to Dementia and Geriatric Depression Scale Score To assess the relationship between dementia severity and FCI-9 performance, we correlated Mini-Mental Status Examination scores with the FCI-9 overall score (Domains 1–7) at both baseline and one-year follow-up assessments. Dementia severity at baseline was strongly related to baseline score (r ⫽ 0.71, df⫽ 45, p ⬍0.001) and also to follow-up score (r⫽0.73, df⫽ 45, p ⬍0.001). In contrast, the mild AD patients’ level of depression (Geriatric Depression Scale score) was not related to FCI-9 overall score at either base-
215
Declining Financial Capacity in AD line (r⫽⫺0.05, df⫽45, p⫽0.74) or one-year follow up (r⫽⫺.09, df⫽45, p⫽0.55). It was noted that 14 patients with mild AD had Geriatric Depression Scale scores over 10 with a maximum score of 20 in one patient. Changes in the Proportion of Capacity Outcome Classifications for Patients With Mild Alzheimer’s Disease Over One Year Table 5 presents capacity outcome classifications for the mild AD group at baseline and one-year follow up on the FCI-9 domains and two overall scores. These capacity outcome classifications should be interpreted cautiously, because they have been derived experimentally through application of psychometric cut scores and are not equivalent to clini-
TABLE 5.
cally or legally determined competency outcomes. However, the outcomes serve as a useful additional perspective for understanding the study results. At baseline, substantial variability in domain outcomes existed in the mild AD group (Table 5). For example, 67% of the AD group was assigned a “capable” outcome for knowledge of assets/estate, whereas only 21% of patients were deemed capable on checkbook management. More than 40% of patients with AD at baseline were found to be capable on the domains of basic monetary skills, cash transactions, financial judgment, and investment decision-making. At 1-year follow up, a clear shift in the distribution of capacity outcomes occurred in the mild AD group with many patients previously classified as capable
FCI-9 Capacity Outcome Classifications at Baseline and One-Year Follow Up for Patients With Mild AD
Domain 1: Basic monetary skills Baseline One-year follow up Domain 2: Financial concepts Baseline One-year follow up Domain 3: Cash transactions Baseline One-year follow up Domain 4: Checkbook management Baseline One-year follow up Domain 5: Bank statement management Baseline One-year follow up Domain 6: Financial judgment Baseline One-year follow up Domain 7: Bill payment Baseline One-year follow up Domain 8: Knowledge assets/estate Baseline One-year follow up Domain 9: Investment decision-making Baseline One-year follow up Total score (Domains 1–7) Baseline One-year follow up Total score (Domains 1–8) Baseline One-year follow up
Capable
Marginally Capable
Incapable
z
pa
64% (35/55) 33% (18/55)
11% (6/55) 14% (8/55)
25% (14/55) 53% (29/55)
⫺4.23
0.001
26% (14/55) 27% (15/55)
27% (15/55) 18% (10/55)
47% (26/55) 55% (30/55)
⫺0.43
0.67
41% (22/54) 22% (12/54)
17% (9/54) 15% (8/54)
42% (23/54) 63% (34/54)
⫺2.53
0.001
21% (11/53) 25% (13/53)
7% (4/53) 9% (5/53)
72% (38/53) 66% (35/53)
⫺1.07
0.28
33% (16/49) 12% (6/49)
16% (8/49) 18% (9/49)
51% (25/49) 69% (34/49)
⫺3.48
0.001
66% (36/55) 44% (24/55)
11% (6/55) 9% (5/55)
23% (13/55) 47% (26/55)
⫺3.43
0.001
27% (13/48) 33% (16/48)
13% (6/48) 8% (4/48)
60% (29/48) 58% (28/48)
⫺0.72
0.47
67% (33/49) 41% (20/49)
14% (7/49) 26% (13/49)
18% (9/49) 33% (16/49)
⫺1.68
0.09
48% (20/42) 26% (11/42)
31% (13/42) 21% (9/42)
21% (9/42) 52% (22/42)
⫺3.51
0.001
13% (6/47) 9% (4/47)
15% (7/47) 9% (4/47)
72% (34/47) 82% (39/47)
⫺2.11
0.04
9% (4/44) 9% (4/44)
11% (5/44) 5% (2/44)
80% (35/44) 86% (38/44)
⫺1.13
0.26
Significance of difference (p ⬍0.05) between baseline and 1-year follow-up outcomes using Wilcoxon test. FCI-9: Financial Capacity Instrument-9; AD: Alzheimer disease.
a
216
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al. (i.e., FCI performance equal to or greater than 1.5 SD below the comparison group mean score) now being classified as marginally capable or incapable. On five of the nine domains (basic monetary skills, cash transactions, bank statement management, financial judgment, and investment decision-making), a larger proportion of patients with mild AD at follow up were classified as marginally capable or incapable compared with baseline. Interestingly, although at baseline only 36% of patients with mild AD were classified as either marginally capable or incapable on Domain 1 (basic monetary skills), after one year, the proportion of such patients had increased to 67%. The largest increases over one year in “‘incapable” outcomes occurred for the domains of basic monetary skills (⫹28%) and financial judgment (⫹24%).
CONCLUSIONS Functional decline is a defining aspect of AD and has been clearly documented in the literature.5,20 Although one study has shown declines over time on isolated financial skills such as counting currency and check writing,5 we do not yet understand the natural history of financial capacity loss in dementia. The present study seeks to build this understanding by examining longitudinal change in a wide range of financial abilities in patients with AD. Our findings indicated that patients with mild AD early on have a broad range of impaired financial skills and that these skills show relatively rapid further decline over a one-year period. Declines emerged on both simple and complex tasks on almost all domains and on both overall scores. From a clinical standpoint, these longitudinal findings underscore the importance of patients with mild AD and their families timely pursuing financial planning and transfer of financial responsibilities at the time of diagnosis. The longitudinal results demonstrated near global one-year decline in financial skills in the AD group. Group-by-time interaction effects were pronounced for both overall scores, indicative of the significant decline in overall financial capacity. At the task and domain level, interactions involved both elementary financial skills (basic monetary skills, cash transac-
Am J Geriatr Psychiatry 16:3, March 2008
tions, financial judgment) and complex skills (bank statement management, investment decision-making). However, interaction effects were not found across all tasks and domains. For example, some complex financial tasks did not show an interaction effect (understanding concepts, using checkbook/ register, preparing bills for mailing). This finding may reflect several factors, including existing level of baseline impairment in the mild AD group, comparison group performance, and large within-AD-group SDs. The lack of interaction effects for some financial tasks is important, because it suggests different trajectories of change across financial skills in AD over time. The FCI variable difficulty data supported this hypothesis. A review of task level data revealed differential rates of change in the mild AD group (Table 4). The greatest one-year declines (⬎⫺15%) occurred in complex tasks of using a bank statement and carrying out a multi-item grocery purchase (involving mental computation). However, one complex task (understanding financial concepts) showed relative stability (⬍⫺5% decline) over one year. Some simple tasks demonstrated notable declines (ⱖ⫺10%) over one year (mail fraud, telephone fraud, counting coins/currency), suggesting that basic judgment and counting skills are eroding in mild AD. The judgment finding is particularly clinically salient given the vulnerability of cognitively impaired adults to financial scams and exploitation. Other simple skills remained relatively stable over the same period (understanding bills, prioritizing bills, naming coins/currency). The variable difficulty data also provide insight into overall loss of financial skills in mild AD. At baseline, the mild AD group had already lost approximately 20% of the overall financial skills demonstrated by the comparison group. Over one year, a further ⫺10% change occurred in overall performance (down to 70% of comparison group performance) indicative of relatively rapid decline. Additional longitudinal studies across both mild and moderate stages and longer time periods will be needed to chart the full trajectory of financial capacity loss in AD. The capacity outcome data (Table 5) provided a different and more clinical perspective on declining financial abilities in the mild AD group. Caution should be exercised in reviewing these outcome data,
217
Declining Financial Capacity in AD because they represent applications of psychometric cutoff scores derived from the comparison group performance and are not clinical or legal determinations of financial competency. Capacity outcomes at baseline assessment demonstrated high levels of compromise (combination of “marginal” and “incapable” outcomes) in the mild AD group. For overall financial capacity (Domains 1–7), 87% of patients were classified as either marginally capable or incapable and only 13% were classified as capable. Baseline domain outcomes showed considerable variability, ranging from 64%– 67% capable (basic monetary skills, financial judgment, knowledge of assets/estate) to as low as 21% capable (checkbook management). This variability highlights the fact that although all patients met diagnostic criteria for mild AD, the patients themselves possessed a broad range of financial skills and performance. At one-year follow up, the distribution of marginally capable and incapable outcomes for patients with mild AD increased at both the overall and domain levels. For overall financial capacity, capacity compromise increased from 87%–91% with only 9% of patients with mild AD classified as capable at follow up. Significant outcome distribution changes and increased capacity compromise also occurred not only for complex domains of bank statement management and investment decision-making, but also simple domains of basic monetary skills, cash transactions, and financial judgment. The findings again indicate a loss of both complex and simple skills in mild AD over time. The findings are consistent with prior cross-sectional work that showed that 90% or more of patients with moderate AD were rated as incapable across all domains.2 We acknowledge several limitations of the present study, including the specificity of the findings to AD and related disorders. Patterns of impairment in financial skills may differ across other neurocognitive disorders of aging such as Parkinson disease, frontotemporal dementia, vascular dementia, and stroke. Second, the present study addressed only group-level change. We could not investigate intraindividual rates of change over time (i.e., growth curve analysis) as a result of the small sample sizes and the limited number of time points available. Analysis of longer follow-up periods will be necessary to establish rates of decline in financial capacity in patients with AD. Third, the study’s financial assessments, although using real-
218
world tasks, were conducted in a clinical laboratory setting and may not have always completely reflected actual performance in the home setting. Fourth, we acknowledge that the two samples were not matched for several demographic characteristics (i.e., age, education, age). However, we addressed these differences by using covariate analyses and found statistically significant findings even after controlling for the baseline group demographic differences. Finally, we acknowledge that our patient sample size was relatively small and that a small portion of participant data were excluded on methodological grounds, specifically, lack of prior experience with the particular financial task or domain. However, we believe that this data exclusion actually lead to a more conservative estimation of the between-group differences. Had we included these patients with mild AD in the analyses, we may well have found even larger between-group differences given the reasonable assumption that these patients with mild AD would have performed poorly on the financial tasks. The present study has clinical and policy implications. Impairment in financial skills occurs early in AD and proceeds relatively rapidly over time involving both complex abilities and also very basic judgment and monetary calculation skills.21 It is therefore important for patients and families to be proactive in the financial sphere on first receiving a diagnosis of mild AD. Finalizing trust and estate arrangements, delegating financial decision-making powers, planning for eventual financial incapacity, and providing increased supervision of existing financial activities are all aspects of such a proactive approach. In particular, careful supervision is warranted in light of the study’s findings of decline in financial judgment and susceptibility to simple fraud schemes. The authors hope that the study findings will heighten clinician and public awareness of the financial vulnerability of older adults with mild dementia. This study was supported by research grants (National Institutes of Health, National Institute on Aging 1R01 MH55247 and National Institute on Aging 1 R01 AG021927) (D.M., Principal Investigator) and by an Alzheimer’s Disease Research Center grant (National Institutes of Health, NIA 1P50 AG16582) (D.M., Principal Investigator). The authors appreciate the assistance of Anna Sicola in data collection, and Justin Huthwaite, Ph.D. in reviewing the manuscript.
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al. References 1. McKhann G, Drachman D, Folstein M, et al: Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA work group under the auspices of the Department of Health and Human Services Task Force on Alzheimer’s disease. Neurology 1984; 34:939 –944 2. Marson D, Sawrie S, Snyder S, et al: Assessing financial capacity in patients with Alzheimer’s disease: A conceptual model and prototype instrument. Arch Neurol 2000; 57:877–884 3. Fillenbaum G, Peterson B, Welsh-Bohmer K, et al: Progression of Alzheimer’s disease in white and black patients. Neurology 1998; 51:154 –158 4. Welsh K, Butters N, Mohs R, et al: The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD): Part V. A normative study of the neuropsychological battery. Neurology 1994; 44: 609 –614 5. Loewenstein D, Duara R, Rubert M, et al: Deterioration of functional capacities in Alzheimer’s disease after a one year period. International Psychogeriatrics 1995; 7:495–503 6. American Psychiatric Association : Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC, American Psychiatric Association, 1994 7. Loewenstein D, Amigo E, Duara R, et al: A new scale for the assessment of functional status in Alzheimer’s disease and related disorders. J Gerontol B Psychol Sci Soc Sci 1989; 44:114 –121 8. Royall D, Palmer R, Chiodo LK, et al: Executive control mediates memory’s association with change in functional status: The Freedom House Study. J Am Geriatr Soc 2005a; 53:11–17 9. Royall D, Palmer R, Chiodo L, et al: Normal rates of cognitive change in successful aging: The Freedom House Study. J Int Neuropsychol Soc 2005b; 11:899 –909 10. Marson D: Loss of financial capacity in dementia: Conceptual and
Am J Geriatr Psychiatry 16:3, March 2008
empirical approaches. Aging, Neuropsychology and Cognition 2001; 8:164 –181 11. Melton G, Petrila J, Poythress N, et al: Psychological Evaluations for the Courts. New York, Guilford, 1987 12. Moye J: Theoretical frameworks for competency in cognitively impaired elderly adults. Journal of Aging Studies 1996; 10:27–42 13. Griffith H, Belue K, Sicola A, et al: Impaired financial abilities in mild cognitive impairment: A direct assessment approach. Neurology 2003; 60:449 –457 14. Earnst K, Wadley V, Aldridge T, et al: Loss of financial capacity in Alzheimer’s disease: the role of working memory. Aging, Neuropsychology, and Cognition 2001; 8:109 –119 15. Morris J: The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 1993; 43:2412–2414 16. SPSS Inc. SPSS Base 10.0 Applications Guide. Chicago, SPSS Inc, 2000 17. Folstein M, Folstein S, McHugh P: Mini-Mental State: a practical guide for grading the cognitive state of the patient for the physician. J Psychiatr Res 1975; 12:189 –198 18. Mattis S: Dementia rating scale, in Geriatric Psychiatry. Edited by Bellack R, Karasu B. New York, Grune & Stratton, 1976 19. Yesavage J: Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res 1983; 17: 37–49 20. Brickman A, Riba A, Bell K, et al: Longitudinal assessment of patient dependence in Alzheimer’s disease. Arch Neurol 2002; 59:1304 –1308 21. Martin R, Annis S, Zebley L, et al: Loss of calculation abilities in patients with mild and moderate Alzheimer’s disease. Arch Neurol 2003; 60:1585–1589
219
Objective: The objective of this study was to investigate change over time in financial abilities in patients with mild Alzheimer disease (AD). Methods: The authors conducted a prospective 1-year longitudinal study at a large southern U.S. metropolitan-area medical school university. Participants included healthy older adults (N⫽63) and patients with mild AD (N⫽55). The authors conducted a standardized performance measure of financial capacity. Performance was assessed on 18 financial tasks, nine domains of financial activity, and overall financial capacity. Capacity outcomes classifications (capable, marginally capable, or incapable) for domains and overall performance were made using cut scores referenced to comparison group performance. Results: At baseline, patients with mild AD performed significantly below healthy older adults on 16 of 18 tasks, on all nine domains, and on overall financial capacity. At one-year follow up, comparison group performance was stable on all variables. In contrast, patients with mild AD showed substantial declines in overall financial capacity, on eight of nine domains, and on 12 of 18 tasks. Similarly, the proportion of the mild AD group classified as marginally capable and incapable increased substantially over one year for the two overall scores and for five financial domains. Conclusions: Financial capacity is already substantially impaired in patients with mild AD at baseline and undergoes rapid additional decline over one year. Relative to the comparison group, overall financial capacity performance in the AD group declined 10%, from approximately 80% of the comparison group performance at baseline to 70% at follow up. Financial skills showed differential rates of decline on both simple and complex tasks. Of clinical and public policy interest was the declining judgment of patients with mild AD regarding simple fraud schemes. The study supports the importance of prompt financial supervision and planning for patients newly diagnosed with AD. (Am J Geriatr Psychiatry 2008; 16:209–219) Key Words: dementia, financial capacity, Alzheimer’s disease, longitudinal research
P
rogressive cognitive and functional decline are hallmark clinical features of Alzheimer disease
(AD).1,2 Longitudinal investigations of cognitive change have provided important insights into the
Received April 25, 2007; revised June 26, 2007; accepted July 18, 2007. From the Department of Neurology and Alzheimer’s Disease Research Center (RM, HRG, KB, LH, DM) and the Department of Biostatistics (AB), University of Alabama at Birmingham, Birmingham, AL; VA Medical Center (LH), Birmingham, AL; the Department of Neurology (EZ), University of Utah Health Sciences, Salt Lake City, UT; and Brown University (BA), Providence, RI. Send correspondence and reprint requests to Dr. Daniel Marson, Department of Neurology, SC 650, University of Alabama at Birmingham, Birmingham, AL 35294-0017. e-mail: [email protected] © 2008 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 16:3, March 2008
209
Declining Financial Capacity in AD natural history of the disease.3,4 In contrast, few studies exist investigating changes in independent activities of daily living.5 This is surprising given the importance of independent activity of daily living change to dementia diagnosis1,6 and to clinician management of patients with dementia.7 Although cognitive and functional decline in dementia are by definition related, a single cognitive measure or domain may not be associated with observed levels of functional disability.8,9 In addition, it is loss of functional abilities rather than cognitive decline per se that causes patient disability and burden for family members and caregivers.10 Thus, understanding the patterns, timeframes, and rates of functional decline over time in AD has both scientific and clinical importance. Of particular interest to our group has been impairment and eventual loss of financial abilities in dementia.2,10 Financial capacity is a complex, multidimensional independent activities of daily living, which has critical relationships to both functional independence and personal autonomy in older adults.11,12 Cross-sectional studies have shown that financial capacity is highly sensitive to AD across its continuum—from mild cognitive impairment13 to mild and moderate AD.2,10,14 Thus, financial capacity is a valuable prism through which to study and understand functional decline in AD. In this study, we used a direct performance measure to investigate change in financial abilities over a oneyear period in patients with mild AD. We hypothesized that 1) financial capacity would show significant decline over one year and 2) that decline would occur primarily on complex rather than simple tasks.
METHODS Conceptual Model of Financial Capacity Financial capacity involves a broad range of declarative, procedural, and judgment-based knowledge and skills.2,10 Our conceptual model views financial capacity at three levels: specific financial abilities (tasks); broader areas of financial activity (domains) that have clinical relevance for independent functioning (e.g., checkbook management; judgment concerning potential fraud); and overall financial capacity (global). Tasks were defined as
210
simple or complex a priori based on their novelty and demand characteristics. Development of this model is discussed in detail elsewhere.2,10,14 A schematic of the model is presented in Table 1. Financial Capacity Instrument The Financial Capacity Instrument (FCI-9)2,13 is a standardized instrument developed by our group that is based on the previously described conceptual model and that directly assesses subjects’ financial abilities. The FCI consists of standardized tasks of varying difficulty level organized into domains (Table 1). The current version assesses 18 tasks within nine domains and has two overall scores.13 The 18 tasks are included in Domains 1–7, whereas Domains 8 and 9 are analyzed at the domain level only. The FCI uses detailed, intervallevel quantitative scoring to assess subject performance on tasks. Task scores are then summed to obtain domain and overall financial capacity scores. The FCI has demonstrated good to excellent internal, test–retest, and interrater reliabilities (25). Participants Participants consisted of 63 healthy older adults and 55 patients with mild AD. Our comparison group was cognitively and neurologically intact, healthy older adults who were community-dwelling and recruited into the University of Alabama at Birmingham Alzheimer’s Disease Research Center. Each subject was clinically evaluated by a neurologist (LEH, BA, or EYZ) to ensure the absence of medical and psychiatric conditions that could compromise cognition. Subjects were characterized as cognitively normal after a diagnostic consensus conference involving the neurologists and a neuropsychologist (DCM). All of our older adult comparison group members received a Clinical Dementia Rating15 staging of 0.0. Patients with probable AD were communitydwelling individuals recruited into the Alzheimer’s Disease Research Center whose dementia was well characterized based on neurologic, neuropsychologic, and radiologic procedures. Diagnosis of probable AD was made in the diagnostic consensus conference and was based on National Institute of Neurological and Communicative Diseases and Stroke–Alzheimer’s Disease and Related Disorders
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al.
TABLE 1.
Revised Conceptual Model of Financial Capacity: 18 Tasks, Nine Domains, and Overall Capacity Task Description
Domain 1: Basic monetary skills Task 1a: Naming coins/currency Task 1b: Coin/currency relationships Task 1c: Counting coins/currency Domain 2: Financial conceptual knowledge Task 2a: Define financial concepts Task 2b: Apply financial concepts Domain 3: Cash transactions Task 3a: One-item grocery purchase Task 3b: Three-item grocery purchase Task 3c: Change/vending machine Task 3d: Tipping Domain 4: Checkbook management Task 4a: Understand checkbook Task 4b: Use checkbook/register Domain 5: Bank statement management Task 5a: Understand bank statement Task 5b: Use bank statement Domain 6: Financial judgment Task 6a: Detect mail fraud risk Task 6b: Detect telephone fraud risk Domain 7: Bill payment Task 7a: Understand bills Task 7b: Prioritize bills Task 7c: Prepare bills for mailing Domain 8: Knowledge of assets/estatea Domain 9: Investment decision-making Overall financial capacity a
Difficulty
Identify specific coins and currency Indicate monetary values of coins/currency Accurately count arrays of coins and currency
Simple Simple Simple
Define simple financial concepts Practical applications/computation using concepts
Complex Complex
Conduct one-item transaction; verify change Conduct three-item transaction; verify change Obtain change for vending machine; verify charge Understand tipping convention; calculate tips
Simple Complex Complex Complex
Identify/explain parts of checkbook and register Conduct simple transaction and pay by check
Simple Complex
Identify/explain parts of a bank statement Identify specific transactions on bank statement
Complex Complex
Detect/explain risks in mail fraud solicitation Detect/explain risks in telephone fraud solicitation
Simple Simple
Explain meaning and purpose of bills Identify overdue utility bill Prepare bills, checks, envelopes for mailing Indicate personal assets and estate arrangements Understand investment options; determine returns; make and explain decision Overall functioning across tasks and domains
Simple Simple Complex Simple Complex Complex
Requires corroboration by informant.
Association criteria.1 All patients with probable AD in this study were classified as having mild dementia based on a consensus conference agreement and Clinical Dementia Rating staging15 of either 0.5 or 1.0 (Table 2). Informed consent was obtained from all comparison group participants and from patients with mild AD and their participant caregivers in accordance with this Institutional Review Board-approved research.
Data Exclusion: Prior/Premorbid Financial Experience and Missing Data Because individual financial experience can differ across subjects,2 we accounted for lack of prior financial skills and experience in our participants. The Prior/ Premorbid Financial Capacity Form,2 a rating measure identifying a subject’s prior (comparison group) or premorbid (patient with AD) experience across the 18 tasks and nine domains of the financial capacity model,
Am J Geriatr Psychiatry 16:3, March 2008
was administered to all study participants and to their informants (not all comparison group participants had an informant available). Premorbid/prior experience for each task and domain was rated in one of three categories: “could do without help,” “could do but needed help,” or “could not do even with help.” Any participant who either self-reported having no prior experience or ability on a domain or task or whose informant (i.e., spouse, family member) reported that the participant was previously incapable of handling a domain or task was excluded from the analysis for that specific domain or task, although data for that domain were collected. In the case of discrepant data (i.e., a caregiver rating a participant as incapable but the participant endorsing capacity for the same domain or task), the caregiver’s judgment was used as the decision rule. The data exclusion procedures resulted in the exclusion of two persons from the comparison group for Domain 5 analysis as a result of having no prior
211
Declining Financial Capacity in AD
TABLE 2.
Demographic Characteristics of the Comparison Group and Patients With Mild AD Comparison Group (n ⴝ 63) (standard deviation)
Age (years) Gender (M/F) Race White Black Asian American Education Mini-Mental State Examination Dementia Rating Scale Global Deterioration Scalea 1 2 3 4 5 Geriatric Depression Scale
Patients With Mild AD (n ⴝ 55) (standard deviation)
p
66.3 (7.6) 20/43
70.6 (8.4) 31/24
0.005 0.007
53 9 1 14.4 (1.7) 29.3 (1.0)
50 5 0 13.1 (2.3) 24.5 (3.1)
0.26
136.7 (4.8)
116.2 (10.5)
0.001
38 24 0 0 0
0 3 21 21 9
0.001
7.4 (5.1)
0.004
4.1 (3.7)
0.001 0.001
⫽ 105.3, p ⬍0.001. AD: Alzheimer disease.
a 2
experience. In the mild AD group, the following exclusions or actual missing data at one-year follow up were identified: Domain 3 (N⫽1, missing data), Domain 4 (N⫽2, missing data), Domain 5 (N⫽6, excluded no prior experience), Domain 7 (N⫽7, four excluded and three missing data), Domain 8 (N⫽6, missing data), and Domain 9 (N⫽13, 11 excluded and two missing data). As expected, Domain 9 had the most data points excluded insofar as investment decision-making is a specialized, socioeconomic status-related skill, which a minority of our patient sample lacked. This outcome was very similar to the data exclusion results of a prior study using the FCI.13
Statistical Analyses The comparison group and mild AD group performances on the demographic and mental status variables were compared at baseline using independentsamples t test. Chi-square statistics were used for the gender and ethnicity variables. Group comparisons for the FCI-9 tasks, domains, and two FCI total scores (Domains 1–7; Domains 1– 8) were compared at baseline using analysis of covariance (ANCOVA). Covariates of age, education, and gender were included in the analyses. Pairedsample t tests were used within groups to examine performance change over one year across the nine domains and the 18 tasks. Interaction effects were examined using repeated-measures ANCOVA with group as the between-subjects variable and time as the within-groups variable. Age, education, and gender were included as covariates for this set of analyses. Alpha level was set at 0.01 to correct for multiple comparisons. All statistical analyses were conducted using SPSS 10.0.16 Capacity Outcome Analysis. Similar to previous studies,2 each patient with mild AD also received a capacity outcome rating (capable, marginally capable, incapable) for each of the nine domains and the two overall total scores. These capacity outcomes were determined by using psychometric cut scores referenced to mean comparison group performance. A capable outcome rating was assigned to a patient with mild AD if the patient’s score fell at ⱕ1.5 standard deviations (SDs) below the comparison group mean score for that domain; a marginally capable outcome rating reflected a patient score falling between 1.5 SDs and 2.5 SDs below the comparison group mean; and an incapable outcome rating reflected a score ⬎2.5 SD below the comparison group mean. Nonparametric analysis (Wilcoxon test16) was conducted to examine the change over one year in the proportion of patients receiving each of the three outcome ratings.
Test Administration Procedures Participants were administered the FCI-9 at baseline and at a 12-month follow-up evaluation. All portions of the FCI-9 were administered in a single session. The FCI took approximately 40 to 50 minutes to administer for comparison group participants and 50 – 60 minutes for patients with mild AD. The FCI has a standardized administration protocol and detailed scoring system.2,10,13
212
RESULTS Demographic and Mental Status Variables Table 2 displays demographic and mental status variables for the two groups. The comparison group was younger than patients with mild AD (t⫽2.89,
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al. df ⫽116, p ⬍0.005) by an average of four years. The patients with mild AD had one year less education (t ⫽ ⫺3.4, df ⫽119, p ⬍0.001). Male-to-female proportion ratios were different between the two groups (2 ⫽ ⫺2.68, p ⬍0.007) with a higher male-to-female ratio present in the mild AD group. Blacks made up 15% of the overall sample size. No between-group differences were found in the racial distribution (2 ⫽ ⫺1.1, p⫽0.26). As expected, patients with mild AD had lower scores on the Mini-Mental State Examination17 (t ⫽ ⫺12.1, df ⫽113, p ⬍0.001) and the Mattis Dementia Rating Scale (t⫽⫺14.7, df⫽116, p ⬍0.001).18 Patients with mild AD reported more depressive symptoms on the Geriatric Depression Scale19 (t⫽3.9, df⫽ 114, p ⬍0.001). However, the mean depression score for the patients with mild AD was well below the clinical cutoff for depression. Changes in Task Level Performance Table 3 sets forth group performance for the 18 financial capacity tasks (across Domains 1–7) at both assessment visits. At baseline assessment, the comparison group performed better than the mild AD group on 16 of the 18 financial capacity tasks. Patients with mild AD performed equivalently with the comparison group at baseline on only two simple tasks: naming coins/currency and a one-item grocery store transaction. Age, education, and gender were included as covariates in this set of analyses. Education level was a significant contributor to nine of the 18 tasks (coins/currency relationships, understanding financial concepts, applying financial concepts, one-item purchase, multiitem transactions, tipping, understanding checkbook, using checkbook, understanding bank statement), whereas gender contributed to two of the 18 tasks (understanding checkbook, using checkbook). At follow up using the paired-samples t test, patients with mild AD demonstrated within-group 1-year declines on 12 of the 18 tasks. In contrast, the comparison group demonstrated no significant performance change across any of the 18 tasks relative to baseline. The group by time interaction, as measured by ANCOVA, was significant for five of 18 tasks representing steady decline for the patients with mild AD and no change for the comparison group. Across the 18 tasks, the covariates contributed to
Am J Geriatr Psychiatry 16:3, March 2008
only two of the 18 tasks. Age had a modest effect on the applying concepts task (F (1,113) ⫽4.6, p ⬍0.035), whereas gender had a modest effect on the understanding bank statements task (F (1,105)⫽4.1, p ⬍0.05). The percentage decline in raw scores from baseline to one-year follow up ranged from 3% (naming coins/currency) to 29% in the mild AD group (using bank statement, preparing bills for mailing).
Changes in Domain and Overall Capacity Performance Table 3 also displays group performance on the FCI-9 domains and two overall scores. At baseline, ANCOVA revealed that the comparison group performed better than patients with mild AD across all nine domains and both overall scores (Domains 1–7, Domains 1– 8). At baseline, covariates were significant contributing factors in seven of nine domains (education level—seven domains, gender— one domain). However, no covariate affected significance values for the between-groups analyses. At one-year follow up, patients with mild AD demonstrated declines on eight of the nine domains (only Domain 8 did not show a decline) and on both overall scores. In contrast, the comparison group demonstrated stability across all nine domains and both overall scores. Group-by-time interactions revealed a decline in the mild AD group for four of nine domains and both overall scores. The covariate age contributed to Domain 8, but no other covariate contributed to the other domains.
Financial Capacity Instrument-9 Variable Difficulty and Annual Change in the Mild Alzheimer Disease Group We examined the difficulty of FCI-9 variables for patients with mild AD at baseline and at one-year follow up (Table 4). A variable’s difficulty for patients with AD was estimated as the percentage of comparison group performance attained at that time point. For example, the difficulty level for patients with mild AD of task 7c (preparing bills for mailing) at baseline was estimated at 63% (mild AD patients’ baseline mean raw score of 16.0 divided by the comparison group’s baseline mean raw score of 25.5) (Table 3). Using the baseline and 1-year follow-up
213
Declining Financial Capacity in AD
TABLE 3.
Change in Financial Abilities Over One Year in the Comparison Group and in Patients With Mild AD Comparison Group
Domain 1: Basic monetary skills Naming coins/currency Coins/currency relationships Counting coins/money Domain 2: Financial concepts Understanding concepts Applying concepts Domain 3: Cash transactions One-item transaction Multi-item transaction Vending machine Tipping Domain 4: Checkbook management Understanding checkbook Using checkbook Domain 5: Bank statement management Understanding bank statement Using bank statement Domain 6: Financial judgment Mail fraud Telephone fraud Domain 7: Bill payment Understanding bills Identifying/prioritizing bills Preparing bills for mailing Domain 8: Knowledge assets/estate Domain 9: Investment decision-making Total score (Domains 1–7) Total score (Domains 1–8)
Score Range
Time 1 (n ⴝ 63)
Time 2 (n ⴝ 63)
0–48
46.0 (2.9)
0–8 0–28 0–12 0–40
Patients With AD pa
Time 1 (n ⴝ 55)
Time 2 (n ⴝ 55)
pb
Baseline
Interaction
46.2 (2.7)
0.62
41.6 (5.7)
37.5 (8.5)
0.001
.001
.001
7.9 (.34) 26.2 (2.7) 11.9 (.35) 36.0 (3.2)
7.9 (.27) 26.4 (2.5) 11.9 (.42) 35.3 (4.1)
0.41 0.63 0.64 0.12
7.7 (.48) 22.2 (5.3) 11.5 (1.2) 26.9 (6.2)
7.5 (.96) 19.2 (7.0) 10.9 (2.4) 23.9 (7.2)
0.08 0.007 0.001 0.001
.16 .001 .009 .001
.08 .001 .10 .03
0–15 0–25 0–30
13.5 (1.3) 22.5 (2.8) 27.4 (2.9)
13.0 (1.4) 22.4 (3.2) 27.7 (2.6)
0.03 0.74 0.42
10.5 (2.4) 16.5 (4.7) 21.8 (4.6)
10.0 (3.0) 13.9 (5.3) 18.8 (5.7)
0.20 0.001 0.007
.001 .001 .001
.42 .002 .001
0–6 0–7 0–9 0–8 0–54
5.9 (.51) 6.4 (1.4) 8.6 (1.0) 6.4 (1.6) 53.0 (1.5)
5.9 (.43) 6.4 (1.2) 8.8 (.76) 6.6 (1.5) 52.4 (2.3)
0.83 0.78 0.12 0.49 0.08
5.6 (.99) 4.8 (1.8) 6.4 (2.2) 5.0 (1.9) 42.2 (10.0)
5.1 (1.4) 3.7 (2.1) 5.8 (2.3) 4.2 (2.0) 39.1 (11.9)
0.004 0.001 0.09 0.001 0.007
.03 .001 .001 .001 .001
.03 .001 .02 .03 .04
0–24 0–30 0–38
23.6 (.71) 29.3 (1.4) 34.4 (3.1)
23.4 (.96) 28.9 (2.0) 33.8 (3.4)
0.28 0.14 0.10
20.8 (2.8) 21.3 (7.7) 24.5 (7.8)
19.4 (4.4) 19.7 (8.8) 19.2 (8.5)
0.006 0.03 0.001
.001 .001 .001
.03 .13 .001
0–18 0–20 0–26
15.5 (2.0) 18.9 (1.7) 24.9 (2.6)
15.1 (2.1) 18.7 (1.8) 24.7 (2.4)
0.18 0.13 0.35
11.5 (3.2) 13.0 (5.3) 20.9 (5.9)
10.0 (3.9) 9.2 (5.3) 17.9 (7.0)
0.001 0.001 0.001
.001 .001 .001
.004 .001 .02
0–8 0–18 0–46 0–6 0–13 0–27 0–20
7.8 (.78) 17.1 (2.0) 43.8 (4.1) 5.7 (.74) 12.6 (.69) 25.5 (3.7) 18.0 (2.1)
7.8 (.73) 16.9 (2.1) 43.0 (4.8) 5.5 (1.0) 12.4 (.93) 24.9 (4.9) 18.6 (1.4)
0.78 0.34 0.20 0.34 0.24 0.37 0.12
6.5 (2.2) 14.4 (4.3) 31.6 (9.6) 4.6 (1.5) 11.0 (1.6) 16.0 (8.4) 15.8 (2.9)
5.4 (2.7) 12.5 (5.3) 27.6 (11.6) 4.1 (1.8) 10.3 (2.4) 13.0 (10.0) 15.9 (2.9)
0.001 0.006 0.003 0.11 0.06 0.01 0.72
.001 .001 .001 .001 .001 .001 .009
.02 .13 .06 .63 .24 .11 .22
0–17
13.3 (2.8)
13.3 (2.5)
0.89
9.6 (3.2)
7.9 (3.4)
0.001
.001
.005
0–282 0–302
266.1 (12.4) 285.0 (9.7)
263.4 (13.9) 281.5 (13.5)
0.04 0.06
209.4 (37.4) 230.5 (34.8)
183.2 (50.8) 203.4 (45.5)
0.001 0.001
.001 .001
.001 .001
Note: Not all domains had all data points for patients with mild AD or the comparison group as a result of missing data at year 1 evaluation or data exclusion based on no prior/premorbid experience for that financial ability. Comparison group sample size equal to 63 at year 1 evaluation except for Domain 5 (n ⫽ 61), Domain 8 (n ⫽ 42), and Domain 9 (n ⫽ 58). Mild AD sample size equal to 55 at year one evaluation except for Domain 5 (n ⫽ 49), Domain 7 (n ⫽ 48), Domain 8 (n ⫽ 49), and Domain 9 (n ⫽ 42). a No significant t values (p ⬎0.01) for comparison group difference across baseline and one-year follow-up assessments. Paired-samples t test was the statistic used. Degrees of freedom were as follows: Domains 1– 4, 6, 7 (62), Domain 5 (60), Domain 8 (41), and Domain 9 (57). b All significant t values ⬎2.6 (p ⬍0.01) for mild AD group difference across baseline and one-year follow-up assessments. Paired-samples t test was the statistic used. Degrees of freedom as follows: Domain 1– 4 (54), Domain 5 (48), Domain 6 (54), Domain 7 (47), Domain 8 (48), and Domain 9 (41). Baseline ⫽ all significant F values ⬎4.47 (p ⬍0.01) for comparison group and mild AD group differences at baseline assessment. Total degrees of freedom (df) for tasks on Domains 1– 4 and 6 ⫽ 118; df for tasks on Domains 5 ⫽ 110, df for tasks on Domain 7 ⫽ 114, df for tasks on Domain 8 ⫽ 101, and df for tasks on Domain 9 ⫽ 102. Analysis of covariance was the statistic used. Interaction ⫽ all significant F values ⬎8.8 (p ⬍0.01) for the group-by-time interaction statistic. Total df for tasks on Domains 1– 4 and 6 ⫽ 118; df for tasks on Domain 5 ⫽ 110, df for tasks on Domain 7 ⫽ 114, df for tasks on Domain 8 ⫽ 101, and df for tasks on Domain 9 ⫽ 102. A repeated-measures analysis of covariance was used. AD: Alzheimer disease.
difficulty values, a percentage annual decline value was then derived, which represented the difference between the two difficulty values. In the example of
214
task 7c, this annual decline for patients with mild AD was ⫺11% (52% (one-year follow-up difficulty) to 63% (baseline difficulty).
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al.
TABLE 4.
FCI Variable Difficulty and Annual Percentage Change Within the Mild AD Group (n ⴝ 55)
Variables Task 7c: Preparing bills for mailing Task 5b: Using bank statement Task 2b: Applying concepts Task 4b: Using checkbook/register Task 3c: Vending machine Task 5a: Understanding bank statement Task 3b: Multi-item transaction Task 2a: Understanding concepts Task 3d: Tipping Task 7a: Understanding bills Task 6a: Mail fraud Task 6b: Telephone fraud Task 1b: Coin/currency relationships Task 7b: Prioritizing bills Task 4a: Understanding checkbook/register Task 3a: One-item transaction Task 1c: Counting coins/currency Task 1a: Naming coins/currency Domain 7: Bill payment Domain 5: Bank statement management Domain 9: Investment decision-making Domain 2: Financial concepts Domain 3: Cash transactions Domain 4: Checkbook management Domain 6: Financial judgment Domain 8: Knowledge assets/estate Domain 1: Basic monetary skills Total score (Domains 1–7) Total score (Domains 1–8)
Baselinea
One- Yearb Follow Up
Annual Percentage Changec
63 69 73 73 74 74 75 78 78 80 83 84 85 87 88 95 97 97 70 71 72 75 80 80 84 88 90 79 81
⬎52 49 62 68 66 66 58 77 64 76 69 74 73 83 83 86 87 95 64 57 59 68 68 75 72 85 82 69 72
–11 –20 –11 –05 –08 –08 –17 –01 –14 –04 –14 –10 –12 –04 –05 –09 –10 –02 –06 –14 –13 –07 –12 –05 –12 –03 –08 –10 –09
a Difficulty value is the percentage of comparison group performance attained by the mild AD group at baseline. Variables are listed in descending order of difficulty. b Difficulty value is the percentage of comparison group performance attained by the mild AD group at one-year follow up. c Change value is the decline in percentage of comparison group performance attained by the mild AD group from baseline to one-year follow up. Change values ⬎⫺10% are bolded. FCI: Financial Capacity Instrument; AD: Alzheimer disease.
At the overall capacity level, the mild AD group achieved approximately 80% of comparison group performance at baseline. Over one year, the AD group declined a further 9%–10% relative to the comparison group performance. At the domain level, difficulty values differed substantially, ranging from a high of 70% for bill payment to a low of 90% for basic monetary skills. One-year changes were greatest for bank statement management (⫺14%), investment decision (⫺13%), cash transactions (⫺12%), and financial judgment (⫺12%). Knowledge of assets/estate arrangements showed the least change (⫺3%). Similarly, at the task level, there was variability in difficulty values at baseline, ranging from a high of 63% for preparing bills to a low of 97% for naming coins/currency and also for counting coins/ currency. One-year changes were greatest for using
Am J Geriatr Psychiatry 16:3, March 2008
bank statement (⫺20%) and a multi-item grocery store purchase (⫺17%) and least for understanding concepts (⫺1%) and naming coins/currency (⫺2%). Relationship of Capacity Performance to Dementia and Geriatric Depression Scale Score To assess the relationship between dementia severity and FCI-9 performance, we correlated Mini-Mental Status Examination scores with the FCI-9 overall score (Domains 1–7) at both baseline and one-year follow-up assessments. Dementia severity at baseline was strongly related to baseline score (r ⫽ 0.71, df⫽ 45, p ⬍0.001) and also to follow-up score (r⫽0.73, df⫽ 45, p ⬍0.001). In contrast, the mild AD patients’ level of depression (Geriatric Depression Scale score) was not related to FCI-9 overall score at either base-
215
Declining Financial Capacity in AD line (r⫽⫺0.05, df⫽45, p⫽0.74) or one-year follow up (r⫽⫺.09, df⫽45, p⫽0.55). It was noted that 14 patients with mild AD had Geriatric Depression Scale scores over 10 with a maximum score of 20 in one patient. Changes in the Proportion of Capacity Outcome Classifications for Patients With Mild Alzheimer’s Disease Over One Year Table 5 presents capacity outcome classifications for the mild AD group at baseline and one-year follow up on the FCI-9 domains and two overall scores. These capacity outcome classifications should be interpreted cautiously, because they have been derived experimentally through application of psychometric cut scores and are not equivalent to clini-
TABLE 5.
cally or legally determined competency outcomes. However, the outcomes serve as a useful additional perspective for understanding the study results. At baseline, substantial variability in domain outcomes existed in the mild AD group (Table 5). For example, 67% of the AD group was assigned a “capable” outcome for knowledge of assets/estate, whereas only 21% of patients were deemed capable on checkbook management. More than 40% of patients with AD at baseline were found to be capable on the domains of basic monetary skills, cash transactions, financial judgment, and investment decision-making. At 1-year follow up, a clear shift in the distribution of capacity outcomes occurred in the mild AD group with many patients previously classified as capable
FCI-9 Capacity Outcome Classifications at Baseline and One-Year Follow Up for Patients With Mild AD
Domain 1: Basic monetary skills Baseline One-year follow up Domain 2: Financial concepts Baseline One-year follow up Domain 3: Cash transactions Baseline One-year follow up Domain 4: Checkbook management Baseline One-year follow up Domain 5: Bank statement management Baseline One-year follow up Domain 6: Financial judgment Baseline One-year follow up Domain 7: Bill payment Baseline One-year follow up Domain 8: Knowledge assets/estate Baseline One-year follow up Domain 9: Investment decision-making Baseline One-year follow up Total score (Domains 1–7) Baseline One-year follow up Total score (Domains 1–8) Baseline One-year follow up
Capable
Marginally Capable
Incapable
z
pa
64% (35/55) 33% (18/55)
11% (6/55) 14% (8/55)
25% (14/55) 53% (29/55)
⫺4.23
0.001
26% (14/55) 27% (15/55)
27% (15/55) 18% (10/55)
47% (26/55) 55% (30/55)
⫺0.43
0.67
41% (22/54) 22% (12/54)
17% (9/54) 15% (8/54)
42% (23/54) 63% (34/54)
⫺2.53
0.001
21% (11/53) 25% (13/53)
7% (4/53) 9% (5/53)
72% (38/53) 66% (35/53)
⫺1.07
0.28
33% (16/49) 12% (6/49)
16% (8/49) 18% (9/49)
51% (25/49) 69% (34/49)
⫺3.48
0.001
66% (36/55) 44% (24/55)
11% (6/55) 9% (5/55)
23% (13/55) 47% (26/55)
⫺3.43
0.001
27% (13/48) 33% (16/48)
13% (6/48) 8% (4/48)
60% (29/48) 58% (28/48)
⫺0.72
0.47
67% (33/49) 41% (20/49)
14% (7/49) 26% (13/49)
18% (9/49) 33% (16/49)
⫺1.68
0.09
48% (20/42) 26% (11/42)
31% (13/42) 21% (9/42)
21% (9/42) 52% (22/42)
⫺3.51
0.001
13% (6/47) 9% (4/47)
15% (7/47) 9% (4/47)
72% (34/47) 82% (39/47)
⫺2.11
0.04
9% (4/44) 9% (4/44)
11% (5/44) 5% (2/44)
80% (35/44) 86% (38/44)
⫺1.13
0.26
Significance of difference (p ⬍0.05) between baseline and 1-year follow-up outcomes using Wilcoxon test. FCI-9: Financial Capacity Instrument-9; AD: Alzheimer disease.
a
216
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al. (i.e., FCI performance equal to or greater than 1.5 SD below the comparison group mean score) now being classified as marginally capable or incapable. On five of the nine domains (basic monetary skills, cash transactions, bank statement management, financial judgment, and investment decision-making), a larger proportion of patients with mild AD at follow up were classified as marginally capable or incapable compared with baseline. Interestingly, although at baseline only 36% of patients with mild AD were classified as either marginally capable or incapable on Domain 1 (basic monetary skills), after one year, the proportion of such patients had increased to 67%. The largest increases over one year in “‘incapable” outcomes occurred for the domains of basic monetary skills (⫹28%) and financial judgment (⫹24%).
CONCLUSIONS Functional decline is a defining aspect of AD and has been clearly documented in the literature.5,20 Although one study has shown declines over time on isolated financial skills such as counting currency and check writing,5 we do not yet understand the natural history of financial capacity loss in dementia. The present study seeks to build this understanding by examining longitudinal change in a wide range of financial abilities in patients with AD. Our findings indicated that patients with mild AD early on have a broad range of impaired financial skills and that these skills show relatively rapid further decline over a one-year period. Declines emerged on both simple and complex tasks on almost all domains and on both overall scores. From a clinical standpoint, these longitudinal findings underscore the importance of patients with mild AD and their families timely pursuing financial planning and transfer of financial responsibilities at the time of diagnosis. The longitudinal results demonstrated near global one-year decline in financial skills in the AD group. Group-by-time interaction effects were pronounced for both overall scores, indicative of the significant decline in overall financial capacity. At the task and domain level, interactions involved both elementary financial skills (basic monetary skills, cash transac-
Am J Geriatr Psychiatry 16:3, March 2008
tions, financial judgment) and complex skills (bank statement management, investment decision-making). However, interaction effects were not found across all tasks and domains. For example, some complex financial tasks did not show an interaction effect (understanding concepts, using checkbook/ register, preparing bills for mailing). This finding may reflect several factors, including existing level of baseline impairment in the mild AD group, comparison group performance, and large within-AD-group SDs. The lack of interaction effects for some financial tasks is important, because it suggests different trajectories of change across financial skills in AD over time. The FCI variable difficulty data supported this hypothesis. A review of task level data revealed differential rates of change in the mild AD group (Table 4). The greatest one-year declines (⬎⫺15%) occurred in complex tasks of using a bank statement and carrying out a multi-item grocery purchase (involving mental computation). However, one complex task (understanding financial concepts) showed relative stability (⬍⫺5% decline) over one year. Some simple tasks demonstrated notable declines (ⱖ⫺10%) over one year (mail fraud, telephone fraud, counting coins/currency), suggesting that basic judgment and counting skills are eroding in mild AD. The judgment finding is particularly clinically salient given the vulnerability of cognitively impaired adults to financial scams and exploitation. Other simple skills remained relatively stable over the same period (understanding bills, prioritizing bills, naming coins/currency). The variable difficulty data also provide insight into overall loss of financial skills in mild AD. At baseline, the mild AD group had already lost approximately 20% of the overall financial skills demonstrated by the comparison group. Over one year, a further ⫺10% change occurred in overall performance (down to 70% of comparison group performance) indicative of relatively rapid decline. Additional longitudinal studies across both mild and moderate stages and longer time periods will be needed to chart the full trajectory of financial capacity loss in AD. The capacity outcome data (Table 5) provided a different and more clinical perspective on declining financial abilities in the mild AD group. Caution should be exercised in reviewing these outcome data,
217
Declining Financial Capacity in AD because they represent applications of psychometric cutoff scores derived from the comparison group performance and are not clinical or legal determinations of financial competency. Capacity outcomes at baseline assessment demonstrated high levels of compromise (combination of “marginal” and “incapable” outcomes) in the mild AD group. For overall financial capacity (Domains 1–7), 87% of patients were classified as either marginally capable or incapable and only 13% were classified as capable. Baseline domain outcomes showed considerable variability, ranging from 64%– 67% capable (basic monetary skills, financial judgment, knowledge of assets/estate) to as low as 21% capable (checkbook management). This variability highlights the fact that although all patients met diagnostic criteria for mild AD, the patients themselves possessed a broad range of financial skills and performance. At one-year follow up, the distribution of marginally capable and incapable outcomes for patients with mild AD increased at both the overall and domain levels. For overall financial capacity, capacity compromise increased from 87%–91% with only 9% of patients with mild AD classified as capable at follow up. Significant outcome distribution changes and increased capacity compromise also occurred not only for complex domains of bank statement management and investment decision-making, but also simple domains of basic monetary skills, cash transactions, and financial judgment. The findings again indicate a loss of both complex and simple skills in mild AD over time. The findings are consistent with prior cross-sectional work that showed that 90% or more of patients with moderate AD were rated as incapable across all domains.2 We acknowledge several limitations of the present study, including the specificity of the findings to AD and related disorders. Patterns of impairment in financial skills may differ across other neurocognitive disorders of aging such as Parkinson disease, frontotemporal dementia, vascular dementia, and stroke. Second, the present study addressed only group-level change. We could not investigate intraindividual rates of change over time (i.e., growth curve analysis) as a result of the small sample sizes and the limited number of time points available. Analysis of longer follow-up periods will be necessary to establish rates of decline in financial capacity in patients with AD. Third, the study’s financial assessments, although using real-
218
world tasks, were conducted in a clinical laboratory setting and may not have always completely reflected actual performance in the home setting. Fourth, we acknowledge that the two samples were not matched for several demographic characteristics (i.e., age, education, age). However, we addressed these differences by using covariate analyses and found statistically significant findings even after controlling for the baseline group demographic differences. Finally, we acknowledge that our patient sample size was relatively small and that a small portion of participant data were excluded on methodological grounds, specifically, lack of prior experience with the particular financial task or domain. However, we believe that this data exclusion actually lead to a more conservative estimation of the between-group differences. Had we included these patients with mild AD in the analyses, we may well have found even larger between-group differences given the reasonable assumption that these patients with mild AD would have performed poorly on the financial tasks. The present study has clinical and policy implications. Impairment in financial skills occurs early in AD and proceeds relatively rapidly over time involving both complex abilities and also very basic judgment and monetary calculation skills.21 It is therefore important for patients and families to be proactive in the financial sphere on first receiving a diagnosis of mild AD. Finalizing trust and estate arrangements, delegating financial decision-making powers, planning for eventual financial incapacity, and providing increased supervision of existing financial activities are all aspects of such a proactive approach. In particular, careful supervision is warranted in light of the study’s findings of decline in financial judgment and susceptibility to simple fraud schemes. The authors hope that the study findings will heighten clinician and public awareness of the financial vulnerability of older adults with mild dementia. This study was supported by research grants (National Institutes of Health, National Institute on Aging 1R01 MH55247 and National Institute on Aging 1 R01 AG021927) (D.M., Principal Investigator) and by an Alzheimer’s Disease Research Center grant (National Institutes of Health, NIA 1P50 AG16582) (D.M., Principal Investigator). The authors appreciate the assistance of Anna Sicola in data collection, and Justin Huthwaite, Ph.D. in reviewing the manuscript.
Am J Geriatr Psychiatry 16:3, March 2008
Martin et al. References 1. McKhann G, Drachman D, Folstein M, et al: Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA work group under the auspices of the Department of Health and Human Services Task Force on Alzheimer’s disease. Neurology 1984; 34:939 –944 2. Marson D, Sawrie S, Snyder S, et al: Assessing financial capacity in patients with Alzheimer’s disease: A conceptual model and prototype instrument. Arch Neurol 2000; 57:877–884 3. Fillenbaum G, Peterson B, Welsh-Bohmer K, et al: Progression of Alzheimer’s disease in white and black patients. Neurology 1998; 51:154 –158 4. Welsh K, Butters N, Mohs R, et al: The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD): Part V. A normative study of the neuropsychological battery. Neurology 1994; 44: 609 –614 5. Loewenstein D, Duara R, Rubert M, et al: Deterioration of functional capacities in Alzheimer’s disease after a one year period. International Psychogeriatrics 1995; 7:495–503 6. American Psychiatric Association : Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC, American Psychiatric Association, 1994 7. Loewenstein D, Amigo E, Duara R, et al: A new scale for the assessment of functional status in Alzheimer’s disease and related disorders. J Gerontol B Psychol Sci Soc Sci 1989; 44:114 –121 8. Royall D, Palmer R, Chiodo LK, et al: Executive control mediates memory’s association with change in functional status: The Freedom House Study. J Am Geriatr Soc 2005a; 53:11–17 9. Royall D, Palmer R, Chiodo L, et al: Normal rates of cognitive change in successful aging: The Freedom House Study. J Int Neuropsychol Soc 2005b; 11:899 –909 10. Marson D: Loss of financial capacity in dementia: Conceptual and
Am J Geriatr Psychiatry 16:3, March 2008
empirical approaches. Aging, Neuropsychology and Cognition 2001; 8:164 –181 11. Melton G, Petrila J, Poythress N, et al: Psychological Evaluations for the Courts. New York, Guilford, 1987 12. Moye J: Theoretical frameworks for competency in cognitively impaired elderly adults. Journal of Aging Studies 1996; 10:27–42 13. Griffith H, Belue K, Sicola A, et al: Impaired financial abilities in mild cognitive impairment: A direct assessment approach. Neurology 2003; 60:449 –457 14. Earnst K, Wadley V, Aldridge T, et al: Loss of financial capacity in Alzheimer’s disease: the role of working memory. Aging, Neuropsychology, and Cognition 2001; 8:109 –119 15. Morris J: The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 1993; 43:2412–2414 16. SPSS Inc. SPSS Base 10.0 Applications Guide. Chicago, SPSS Inc, 2000 17. Folstein M, Folstein S, McHugh P: Mini-Mental State: a practical guide for grading the cognitive state of the patient for the physician. J Psychiatr Res 1975; 12:189 –198 18. Mattis S: Dementia rating scale, in Geriatric Psychiatry. Edited by Bellack R, Karasu B. New York, Grune & Stratton, 1976 19. Yesavage J: Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res 1983; 17: 37–49 20. Brickman A, Riba A, Bell K, et al: Longitudinal assessment of patient dependence in Alzheimer’s disease. Arch Neurol 2002; 59:1304 –1308 21. Martin R, Annis S, Zebley L, et al: Loss of calculation abilities in patients with mild and moderate Alzheimer’s disease. Arch Neurol 2003; 60:1585–1589
219