Association of Performance on the Financial Capacity Instrument–Short Form With Brain Amyloid Load and Cortical Thickness in Older Adults

Abstract

Background and Objectives

To investigate the association of the Financial Capacity Instrument–Short Form (FCI-SF) performance and timing total scores with brain β-amyloid and cortical thickness in cognitively unimpaired (CU) (at baseline) older adults.

Methods

A total of 309 participants (aged 70 years or older) of the Mayo Clinic Study of Aging underwent ¹¹C-Pittsburgh compound B PET amyloid imaging and MRI, and completed the FCI-SF. Abnormal amyloid PET was defined as standardized uptake value ratio ≥1.48 in an Alzheimer disease (AD)-related region of interest and reduced AD signature cortical thickness as ≤2.68 mm (neurodegeneration). A cohort of 218 (of the 309) participants had follow-up visits (every 15 months) with FCI-SF data for longitudinal analysis (number of visits including baseline, median [range]: 2 [2–4]). In the analysis, we used linear regression and mixed-effects models adjusted for age, sex, education, apolipoprotein E ε4 allele status, global cognitive z score, and previous FCI-SF testing.

Results

Participants' mean age (SD) was 80.2 (4.8) years (56.3% male individuals). In cross-sectional analysis, abnormal amyloid PET (vs normal) was associated with a lower FCI-SF total score and slower total composite time. In longitudinal analysis, FCI-SF total score declined faster (difference in annualized rate of change, beta coefficient [β] [95% confidence interval (CI)] = −1.123 [−2.086 to −0.161]) and FCI-SF total composite time increased faster (difference in annualized rate of change, β [95% CI] = 16.274 [5.951 to 26.597]) for participants with neurodegeneration at baseline (vs those without). Participants who exhibited both abnormal amyloid PET and neurodegeneration at baseline had a greater increase in total composite time when compared with the group without abnormal amyloid and without neurodegeneration (difference in annualized rate of change, β [95% CI] = 16.750 [3.193 to 30.307]).

Discussion

Performance and processing speed on the FCI-SF were associated with imaging biomarkers of AD pathophysiology in CU (at baseline) older adults. Higher burdens of imaging biomarkers were associated with longitudinal worsening on FCI-SF performance. Additional research is needed to delineate further these associations and their predictive utility at the individual person level.

Pathophysiologic changes of Alzheimer disease (AD) begin decades before clinical symptoms.^1-4 Such changes include low levels of CSF β-amyloid (Aβ) and elevated amyloid PET imaging biomarkers using radiotracers such as ¹¹C-Pittsburgh compound B (PiB).^2,4-8 With these biomarkers, the underlying pathophysiology of AD can be inferred in people who are asymptomatic.

Financial capacity is a complex instrumental activity of daily living (IADL) that shows early impairment in mild cognitive impairment (MCI)^9-11 and mild AD dementia.^12,13 In addition, in cognitively normal older adults, higher brain amyloid burden (as assessed by ¹⁸F-florbetapir PET) was associated with longer task completion time in the Financial Capacity Instrument-Short Form (FCI-SF).¹² We used a brief, direct performance measure of financial skills (i.e., FCI-SF)¹⁴ and examined the association of brain amyloid burden, and AD signature cortical thickness with FCI-SF performance and timing variables, in cognitively unimpaired (CU) (at baseline) older adults (aged 70 years or older) enrolled in the Mayo Clinic Study of Aging (MCSA). Our goal was to determine whether amyloid and neurodegeneration imaging biomarkers were associated with financial capacity and their change over time in older CU (at baseline) persons.

Methods

Study Population

The MCSA is a prospective population-based cohort, initiated in 2004 in Olmsted County (MN), to investigate the prevalence, incidence, biomarkers, and risk factors of cognitive decline, MCI, and dementia; the study setting, design, sampling, recruitment, and diagnostic procedures have previously been described in detail.^15,16

There were 1,460 MCSA participants (including 1,260 CU participants), aged 70 years or older, consecutively recruited between July 2011 and June 2015, who completed the FCI-SF financial capacity assessment in 1 or more visits. Of the 1,260 CU participants, 309 (24.5%) had a concurrent PiB-PET scan, MRI, neuropsychologic testing, and financial capacity assessment available for cross-sectional analysis. Two hundred eighteen of the 309 CU participants with available FCI-SF, PiB PET scan, and MRI at study baseline visit also had 1 or more follow-up visits (every 15 months) with FCI-SF data available for longitudinal analysis (number of visits including baseline, median [range]: 2 [2–4]). No significant differences were detected in the characteristics of those with (n = 218) vs without follow-up FCI-SF assessment (n = 91) after imaging.

Evaluation of Participants and Diagnostic Assessment

All participants were evaluated at baseline and every 15 months by a nurse or study coordinator and a physician and underwent neuropsychologic testing administered by a trained psychometrist supervised by a board-certified neuropsychologist. A nurse or study coordinator collected sociodemographic factors and asked questions on memory and administered the Clinical Dementia Rating scale¹⁷ and the Functional Activities Questionnaire to an informant.¹⁸ A physician reviewed the participant's medical history, performed a neurologic examination, and administered the Short Test of Mental Status.¹⁹ Nine neuropsychologic tests were administered to assess cognitive performance in 4 domains: (1) memory, (2) attention/executive function, (3) language, and (4) visuospatial skills.^15,16 At each MCSA visit (at baseline and follow-up), a diagnosis of MCI, dementia, or normal cognition was made by an expert consensus panel of nurses/study coordinators, physicians, and a neuropsychologist after reviewing all the information for each participant. Individuals who performed in the normative range and did not meet criteria for MCI²⁰ or dementia²¹ were classified as CU.

Standard Protocol Approvals, Registrations, and Patient Consents

Study approval was obtained from the institutional review boards of the Mayo Clinic and Olmsted Medical Center in Rochester, MN. Written informed consent was obtained from all participants or, in the case of subjects with cognitive impairment sufficient to interfere with capacity, from a legally authorized representative.

Measures

PET

Details are presented in previous reports.^22,23 In summary, amyloid PET imaging was performed with ¹¹C-PIB,⁵ and computer tomography was obtained for attenuation correction. Late-uptake amyloid PET images were acquired from 40 to 60 minutes after injection. PET images were analyzed with an in-house fully automated image processing pipeline,²⁴ where image voxel values are extracted from automatically labeled regions of interest (ROIs) propagated from an MRI template.²² A global amyloid load (standardized uptake value ratio [SUVR]) was formed from the voxel number–weighted average of the median uptake in the prefrontal, orbitofrontal, parietal, temporal, anterior, and posterior cingulate and precuneus ROIs, normalized to the cerebellar crus gray median. Both SUVR units and centiloid units were used to express amyloid PET values; the SUVR to centiloid conversion was performed as previously suggested.²⁵ The cut point to designate abnormal amyloid PET SUVR was the value of 1.48 (centiloid 22),²⁵ beyond which rates of amyloid PET reliably increased. An SUVR value ≥1.48 would assign participants as showing abnormal amyloid PET (A+).^22,26 The first ¹¹C-PiB-PET was used as the baseline in the current analyses.

MRI

All MRI images were acquired on 3 T GE MRI (GE Medical Systems, Milwaukee, WI). The neurodegeneration MRI measure was a FreeSurfer (version 5.3)-derived AD signature meta-ROI formed of the surface area–weighted average of the mean cortical thickness in the entorhinal cortex, fusiform, inferior temporal, and middle temporal gyri.²² Neurodegeneration (N+; reduced AD signature cortical thickness) was defined as ≤2.68 mm.^22,26 In our analysis, we also used the cortical thickness measurements of the parietal and frontal lobes (i.e., loci where processing of financial knowledge is believed to occur).

Neuroimaging Biomarker Combinations

We defined 4 neuroimaging biomarker categories for the study based on abnormal amyloid PET burden (A+/A−) and neurodegeneration, i.e., reduced AD signature cortical thickness (N+/N−): A−N−, A+N−, A−N+, and A+N+.

Financial Capacity Measure

The FCI-SF¹⁴ is a brief, direct performance measure of everyday financial skills and financial processing speed. It was administered by trained psychometrists who scored the FCI-SF items following well-operationalized criteria.^11,27 The FCI-SF measures traditional financial skills and constructs that are well known and meaningful to older adults. The FCI-SF comprises 37 performance items that evaluate monetary calculation, financial concepts, use of a checkbook/register, and use of a bank statement by each participant. The resulting domain scores (i.e., mental calculation, financial conceptual knowledge, single and complex checkbook/register, and bank statement management) sum to an overall performance score (total score) that ranges from 0 to 74 points (higher scores indicate better performance). In addition, FCI-SF considers the time to complete specific tasks (i.e., medical deductible problem, income tax problem, and single and complex checkbook/register problem) resulting into a composite time score for the checkbook tasks and an overall composite time for all timed tasks.¹⁴ The overall timing score for completion of timed tasks (total composite time) has a maximum time of 720 seconds (12 minutes). The FCI-SF also addresses a participant's experience with financial tasks through unscored questions. Participants were not excluded based on level of previous financial experience. In this analysis, we used the FCI-SF total score and total composite time.

Covariates

The chronic disease burden at the study baseline was assessed from a modified Charlson comorbidity index (Charlson index)²⁸ score based on electronic diagnosis codes (Berkson, Hospital International Classification of Diseases Adapted, International Classification of Diseases [ICD]-9, ICD-10) using the Rochester Epidemiology Project's resources. Diagnosis codes included myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, chronic pulmonary disease, peptic ulcer disease, mild liver disease, diabetes, diabetes with organ damage, hemiplegia, moderate/severe renal disease, moderate/severe liver disease, metastatic solid tumor, AIDS, rheumatologic disease, and other cancer; dementia was not included in the score (although included in the Charlson index)²⁸ because MCSA assesses cognitive status (all participants were CU at baseline) and does not base a dementia diagnosis on electronic diagnosis codes. Apolipoprotein E (APOE) ε4 status was determined from a blood draw at MCSA baseline assessment.

Statistical Methods

Participant characteristics by the presence or absence of abnormal brain amyloid or neurodegeneration were compared using the Wilcoxon rank sum test or χ² test, as appropriate. For analytical purposes, the raw scores for neuropsychologic tests in each cognitive domain were z scored and averaged to create domain-specific cognitive z scores; a global z score for overall cognitive performance was created by averaging the 4 domain z scores.

We investigated whether neuroimaging biomarkers were associated with FCI-SF total score and total composite time in cross-sectional and longitudinal analyses. We examined the cross-sectional associations of baseline neuroimaging biomarkers (exposure) and FCI-SF measures (outcomes) with linear regression models (adjusted for age, sex, previous FCI-SF testing [e.g., test naive or not], years of education [model 1], and global cognitive z score [model 2]). In models related to N+, model 2 included APOE ε4 allele status as well. Estimates are presented as beta coefficients (β) and 95% confidence intervals (CIs). We modeled PiB SUVR using the log(10) transformation because amyloid PET SUVR has an approximately constant coefficient of variation across its range, implying that models of log-transformed amyloid PET SUVR will be most efficient in regression.

We also examined the longitudinal associations of baseline neuroimaging biomarkers (exposure) and change in FCI-SF items (outcome) during follow-up using mixed-effects models with random subject-specific intercepts and slopes. Models were adjusted for age, sex, previous FCI-SF testing and years of education (model 1), and additionally for time-varying global cognitive z score (model 2). In models examining cortical thickness, model 2 included APOE ε4 allele status, as well. Estimates are presented as β (95% CI). Participants with 1 data point were included in the mixed-effects models to help with intercept estimation (baseline).

Interpretation of coefficients for time × covariate product terms in our models by 1-unit increase/decrease is not necessarily meaningful given the scales of our predictors. Thus, in all models with continuous measures of global amyloid PET and cortical thickness, we showed measures of associations β (95% CIs) for a 20% increase in baseline global amyloid PET and 0.2 mm decrease in cortical thickness, respectively. Potential effect modification by sex and education were examined using interaction terms in model 2. All analyses were considered statistically significant at a p value <0.05 and were performed using the SAS statistical software, version 9.4 (SAS Institute, Cary, NC).

Data Availability

The MCSA makes data available to qualified researchers on reasonable request.

Results

We first compared the participants of this study (having FCI-SF and imaging, n = 309), with the CU MCSA participants who had FCI-SF but no imaging studies (n = 951). CU participants with FCI-SF and imaging were more likely to be male individuals (56.3% vs 49.3%, p = 0.033); had lower mean Charlson index (mean [SD]: 3.6 [3.1] vs 4.2 [3.4], p = 0.003), higher mean global cognitive z score (mean [SD]: −0.1 [0.9] vs −0.2 [0.9], p = 0.036), and higher mean FCI-SF total score (mean [SD]: 63.5 [8.2] vs 61.9 [9.0], p = 0.005); and were more likely to have experience with preparing their income taxes (189 [61.2%] vs 503 [52.9%], p = 0.011). No mean difference was detected for the FCI-SF total composite time between the 2 groups.

Participants' Characteristics

Baseline characteristics of study participants are summarized in Table 1. Participants in the study had a mean age (SD) of 80.2 (4.8) years (56.3% male individuals and 25.9% APOE ε4 carriers). 97.4% of participants were White; in addition, 2 participants were Asian, 1 was Black or African American, 3 had more than 1 race reported, and 2 had unknown/not reported race. 99.4% of participants were not Hispanic or Latino. One hundred forty-one of the participants (46.5%) exhibited abnormal amyloid PET, and 199 (64.4%) showed neurodegeneration. Seventy-one (23%) participants exhibited neither abnormal amyloid PET nor neurodegeneration (i.e., A−N−), 39 (12.6%) showed only abnormal amyloid PET (i.e., A+N−), 97 (31.4%) exhibited only neurodegeneration (i.e., A−N+), and 102 (33%) exhibited both abnormal amyloid PET and neurodegeneration (i.e., A+N+). Twenty-four participants developed MCI during the study follow-up. Compared with participants with normal amyloid PET, participants with abnormal amyloid PET had a higher mean age, more frequently were APOE ε4 carriers, had more frequent neurodegeneration, had a lower mean FCI-SF total score, and had a higher mean FCI-SF total composite time (Table 1 and Figure 1). Participants with neurodegeneration (compared with those without) had a higher mean age, higher mean Charlson index, and more frequent abnormal amyloid PET. There were no differences in previous financial experience between participants with and without abnormal amyloid PET and participants with and without neurodegeneration.

Table 1.

Baseline Characteristics by Abnormal Amyloid PET and Neurodegeneration Status in Cognitively Unimpaired Participants

Figure 1. FCI-SF Total Score and Total Composite Time by Baseline Abnormal Amyloid PET (A+/A−), Neurodegeneration (N+/N−), and A/N Group Biomarkers.

FCI-SF = Financial Capacity Instrument–Short Form.

Of the 309 participants, 218 (70.6%) had 1 or more FCI-SF follow-up visits after baseline imaging (number of visits including baseline, median [range]: 2 [2–4]). The mean age (SD) of these participants was 80.3 (4.6) years, 57.8% were male individuals, and 27.1% had at least 1 APOE ε4 allele. Compared with participants with normal amyloid PET, participants with abnormal amyloid PET were older, had a lower mean FCI-SF total score, and were less likely to have experience using a medical insurance plan. Participants with neurodegeneration (compared with those without) had a higher mean age and a higher mean Charlson index.

Cross-sectional Analysis

Higher baseline amyloid PET was associated with lower FCI-SF total score, i.e., a 20% increase in amyloid PET SUVR was associated with a reduction in FCI-SF total score of 0.9 points. Abnormal amyloid PET (vs normal) was associated with a lower FCI-SF total score and slower total composite time.

Interaction terms for sex and education with amyloid PET and cortical thickness had p values greater than 0.05 for both models having FCI-SF total score and FCI-SF total composite time as outcomes, and no further analyses were pursued by sex or education.

Cross-sectional analysis related to neurodegeneration and the A/N biomarker groups did not reveal significant associations for the most part (Table 2). We also examined the analysis involving cortical thickness in the parietal and frontal lobes (i.e., loci where processing of financial knowledge is believed to occur). Cortical thickness (as a continuous measure) in the parietal lobes was not associated with FCI-SF total score or total composite time. Last, there was a marginal association between frontal lobe cortical thickness and FCI-SF total score (β [95% CI] = −5.697 [−11.897 to 0.502], in model 2).

Table 2.

Cross-sectional Associations Between Neuroimaging Biomarkers and Financial Capacity Instrument–Short Form in Cognitively Unimpaired Participants Aged 70 Years or Older in the Mayo Clinic Study of Aging

Longitudinal Analysis

There were no statistically appreciable differences in annualized change (from baseline) in the FCI-SF total score and total composite time between the group with abnormal amyloid PET and the group with normal amyloid PET (Table 3). FCI-SF total score declined faster for those with neurodegeneration (vs those without) (difference in annualized rate of change, β [95% CI] = −1.123 [−2.086 to −0.161]; Table 3, model 2). FCI-SF total composite time increased faster for those with neurodegeneration (vs those without) (difference in annualized rate of change, β [95% CI] = 16.274 [5.951 to 26.597]; Table 3, model 2). Participants who exhibited both abnormal amyloid PET and neurodegeneration (i.e., A+N+) had a greater increase in total composite time (i.e., showed more slowing on tasks) compared with the group without abnormal amyloid PET and without neurodegeneration (i.e., A−N−) (difference in annualized rate of change, β [95% CI] = 16.750 [3.193 to 30.307]; Table 4, model 2). Of interest, the A−N+ and the A+N+ groups showed slower task composite times when compared with the A−N− and A+N− groups, respectively, tentatively suggesting that N+ might be the driver in these associations.

Table 3.

Longitudinal Change in Financial Capacity Instrument–Short Form Variables by Neuroimaging Biomarkers at Baseline in Cognitively Unimpaired, at Baseline, Participants

Table 4.

Longitudinal Change in Functional Performance by AN Biomarkers at Baseline in Cognitively Unimpaired Participants

Discussion

We examined the associations of brain amyloid burden and AD signature cortical thickness with FCI-SF performance and timing variables in CU (at baseline) older adults. A higher baseline amyloid PET was associated with a lower performance on financial tasks and abnormal amyloid PET (vs normal) was associated with both worse performance on financial tasks and slower time to complete them. In longitudinal analysis, participants with neurodegeneration (i.e., reduced AD signature cortical thickness) (vs without) at baseline exhibited a faster decline in FCI-SF total score and a greater increase in FCI-SF total composite time. Participants with neurodegeneration at baseline regardless of amyloid burden (i.e., with A+N+ or A−N+) had a greater increase in total composite time (vs A−N−). Although cross-sectional and longitudinal findings are not amenable to an easy interpretation, they could suggest a neurobiologic association at a group level between financial cognition (FCI-SF performance and processing speed) and AD pathophysiology in CU (at baseline) older persons. Such an association is consistent with the known sensitivity and vulnerability of financial skills to AD-related cognitive decline in older persons with MCI and dementia. However, further research is needed to clarify these associations with financial cognition in CU older adults and their potential predictive value at the individual level.

Our cross-sectional findings related to abnormal amyloid PET are consistent with a previous analysis of a smaller MCSA sample.²⁹ Research using the Alzheimer Disease Neuroimaging Initiative-3 multisite sample¹² also showed that a higher amyloid PET SUVR was cross-sectionally associated with lower FCI-SF total score in a sample of participants who were cognitively normal, or had MCI or AD; in addition, FCI-SF total composite time was inversely associated with Aβ SUVR load in cognitively normal participants.¹² Previous studies^30-32 have also shown an association between IADL limitations and abnormal amyloid burden in persons without dementia.

We did not find an association between FCI-SF performance and completion time and neurodegeneration cross-sectionally, which is not in accord with previous reports indicating cross-sectional associations between lower IADL abilities and neurodegeneration across the cognitive spectrum.^31,33,34 Additional research has suggested that medial frontal cortex volume was associated with financial capacity in patients with mild AD,³⁵ and angular gyrus volumes were associated with FCI-SF score in participants with amnestic MCI.³⁶ Our analysis related to parietal and frontal lobe cortical thickness did not reveal significant associations. A potential explanation for null results is that our sample consisted of only CU (at baseline) participants (with 24 participants progressing to MCI during study follow-up).

Although our cross-sectional findings seem to be inconsistent with the premise that neurodegeneration rather than amyloid is more proximal to cognitive symptoms,² longitudinal findings agree with this premise. The participants with neurodegeneration at baseline exhibited a greater annualized increase in the FCI-SF total composite time and a greater decrease in the total score. The longitudinal analysis findings agree with previous studies suggesting that baseline AD biomarkers are associated with worsening IADL abilities, but none of these studies separately examined financial abilities.^33,37,38

We presented analysis models with and without adjustment for a measure of global cognitive performance (i.e., global cognitive z score) to understand and separate cognition from financial cognition/IADL function (performance and processing speed time), although its inclusion in the model did not change estimates appreciably.

The FCI-SF focuses on the cognitively complex and functionally important IADL of financial cognition, which has practical everyday and ecological validity for patients, families, clinicians, and researchers.³⁹ The FCI-SF group performance and processing speed differences we observed represent subtle but measurable functional difficulties that may be harbingers of subsequent overt functional decline and impairment⁹ that emerges clinically in late MCI and mild AD dementia.¹⁰ Figure 2 presents a hypothetical model on potential pathways linking Aβ, neurodegeneration, and financial cognition/IADL function. In brief, Figure 2 suggests that amyloid burden is associated with the level of neurodegeneration in the brain, which is associated with declining general cognition and declining neurobehavioral function (e.g., mood state, impulsivity, and interpersonal skills). Declining general cognition and neurobehavioral function are associated with declining financial cognition and performance in everyday life (impaired financial judgment, decisions, and actions), and declining financial cognition and performance are the basis of declining financial capacity/IADL function. Financial cognition is a set of specific abilities that are a product of multiple processes, particularly existing general cognitive abilities and behavioral abilities. Thus, declining financial IADL function is the end product of these multilevel declines. However, research on these pathways involving financial capacity is presently limited. Additional larger longitudinal studies with longer follow-up are necessary to understand the significance of these associations and their predictive utility at the individual person level.

Figure 2. Hypothetical Model on the Pathways Linking Financial Capacity/IADL Function, Amyloid Burden, and Neurodegeneration.

IADL = instrumental activity of daily living.

Longitudinal analyses examining the association of neuroimaging biomarkers and change in FCI-SF performance/time are currently lacking, and additional larger studies with a longer follow-up are needed. Although the FCI-SF is a measure of financial skills designed for older adults and measures traditional financial skills and constructs that are well known and meaningful to older adults, we acknowledge that the study did not account for other factors that could be related with financial skills and experience (e.g., occupational attainment and socioeconomic status).¹⁴ In addition, we examined the association of FCI-SF with cortical thickness measures, specifically in AD signature cortical regions. However, non-AD processes could have contributed to neurodegeneration (e.g., cerebrovascular disease, non-AD tauopathies, or synucleinopathies)⁴⁰ or confound the associations; participants with conditions due to non-AD pathology (e.g., participants with a history of stroke or Parkinson disease) were not excluded from the study.

The strengths of the study are the concurrent financial capacity measure of FCI-SF, which is a brief, psychometrically precise direct performance measure of financial cognition (everyday financial skills and processing speed), with baseline neuroimaging measures of amyloid PET and AD signature cortical thickness, and the availability of longitudinal FCI-SF data after imaging.

Study findings should be considered in light of potential limitations. Study participants were older adults (mean age [SD]: 80.2 [4.8] years) and exhibited a lower burden of medical conditions (lower Charlson index), higher global cognitive performance, and FCI-SF total score compared with participants with FCI-SF data available but without imaging. Thus, the study might have selected a healthier group of older participants, which could have biased the association between neuroimaging biomarkers and FCI-SF variables (probably toward the null, i.e., underestimating the associations). Study participants were 70 years old or older, 97.4% of the study participants were White, and 99.4% were not Hispanic or Latino. Thus, an investigation of these associations in younger, more diverse populations would be desirable to assist in generalizability of the findings. We dichotomized neuroimaging biomarker measures that represent an underlying continuous process, and classification errors especially close to the cut points are possible; thus, we also provided estimations related to the continuous neuroimaging measures. Differences were subtle yet measurable; however, chance findings cannot be precluded, and a longer follow-up period would be desirable in future studies. Last, the study does not include tau PET imaging information, which was implemented toward the end of the FCI-SF administration period.

Appendix. Authors

Study Funding

The study was supported by the NIH (U01 AG006786, P50 AG016574, R01AG057708, R01 AG011378, R01 AG021927, R01 AG041851, and R01 NS097495), the Alexander Family Alzheimer's Disease Research Professorship of the Mayo Clinic, the Alice Weiner Postdoctoral Research Fellowship in Alzheimer's Disease Research, Mayo Foundation for Medical Education and Research, the Liston Award, the Schuler Foundation, and the Alice Weiner Postdoctoral Research Fellow in Alzheimer's Disease Research and was made possible by the Rochester Epidemiology Project (R01 AG034676).

Disclosure

M. Vassilaki has received research funding from F. Hoffmann-La Roche Ltd and Biogen in the past and consults for F. Hoffmann-La Roche Ltd; she, receives research funding from NIH and St. Anne’s University Hospital Brno, International Clinical Research Center, Czech Republic/EU, and has equity ownership in Abbott Laboratories, Johnson and Johnson, Medtronic and Amgen. J.A. Aakre reports no disclosures. W.K. Kremers receives research funding from Department of Defense, NIH, Astra Zeneca, Biogen, and Roche. M.M. Mielke consults for Brain Protection Company and Biogen and receives research funding from the NIH/NIA. Y.E. Geda receives funding from NIH and Roche and previously served on Lundbeck Advisory Board. M.M. Machulda receives NIH funding. D.S. Knopman serves on a Data Safety Monitoring Board for the DIAN study; is an investigator in clinical trials sponsored by Lilly Pharmaceuticals, Biogen, and the Alzheimer's Treatment and Research Institute at USC; and receives research support from the NIH. P. Vemuri receives NIH funding. V.J. Lowe serves on scientific advisory boards for Bayer Schering Pharma, Piramal Life Sciences, Merck Research, and AVID Radiopharmaceuticals and receives research support from GE Healthcare, Siemens Molecular Imaging, AVID Radiopharmaceuticals, and the NIH (NIA and NCI). C.R. Jack, Jr., serves on scientific advisory board for Eli Lilly & Company; receives research support from the NIH/NIA, and the Alexander Family Alzheimer's Disease Research Professorship of the Mayo Foundation; and holds stock in Johnson & Johnson. E.D. Roberson serves on scientific advisory boards for Biogen, AGTC, and AVROBIO and receives funding from NIH, Bluefield Project, and Alzheimer's Drug Discovery Foundation. A. Gerstenecker receives NIH funding. R.C. Martin receives funding from NSF and NIH. R.E. Kennedy receives NIH funding. D.C. Marson receives NIH funding; is the inventor of the FCI-SF and the UAB Research Foundation (UABRF); owns the FCI-SF through copyright and trademark (FCAP); has previously received royalty and consulting income from UABRF licensed use and sale of the FCI-SF; and is currently a consultant on an unaffiliated NIH grant using the FCI-SF. R.C. Petersen is a consultant for Roche, Inc., Biogen, Inc., Merck, Inc., Eli Lilly and Company, and Genentech, Inc.; receives publishing royalties from Mild Cognitive Impairment (Oxford University Press, 2003), and receives research support from the National Institute of Health. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.

Publication History

Received by Neurology: Clinical Practice June 4, 2021. Accepted in final form January 18, 2022. Submitted and externally peer reviewed. The handling editor was Richard Barbano, MD, PhD, FAAN.