Abstract
Background
Cognitive processing speed is important for performing everyday activities in persons with mild cognitive impairment (MCI). However, its role in daily function has not been examined while simultaneously accounting for contributions of Alzheimer’s disease (AD) risk biomarkers. We examine the relationships of processing speed and genetic and neuroimaging biomarkers to composites of daily function, mobility, and driving.
Method
We used baseline data from 103 participants on the MCI/mild dementia spectrum from the Applying Programs to Preserve Skills trial. Linear regression models examined relationships of processing speed, structural magnetic resonance imaging (MRI), and genetic risk alleles for AD to composites of performance-based instrumental activities of daily living (IADLs), community mobility, and on-road driving evaluations.
Results
In multivariable models, processing speed and the brain MRI neurodegeneration biomarker Spatial Pattern of Abnormality for Recognition of Early Alzheimer’s disease (SPARE-AD) were significantly associated with functional and mobility composite performance. Better processing speed and younger age were associated with on-road driving ratings. Genetic risk markers, left hippocampal atrophy, and white matter lesion volumes were not significant correlates of these abilities. Processing speed had a strong positive association with IADL function (p < .001), mobility (p < .001), and driving (p = .002).
Conclusions
Cognitive processing speed is strongly and consistently associated with critical daily functions in persons with MCI in models including genetic and neuroimaging biomarkers of AD risk. SPARE-AD scores also significantly correlate with IADL performance and mobility. Results highlight the central role of processing speed in everyday task performance among persons with MCI/mild dementia.
Keywords: Driving, Financial capacity, MCI, Processing speed, SPARE-AD
A growing body of evidence demonstrates that persons with mild cognitive impairment (MCI) have difficulties or declines in performing everyday tasks such as handling finances, driving, and other instrumental activities of daily living (IADLs) (1–4). While it has been established in cognitively normal older adults that speed in processing visual stimuli is key for the timely and successful performance of these activities (5–7), the link between cognitive processing speed and IADL function has been examined less extensively in persons on the continuum of MCI due to Alzheimer’s disease (AD) (8,9), in whom episodic memory is generally the hallmark deficit (10). In particular, the relationship of processing speed to everyday abilities in MCI and very mild dementia has not been examined while accounting for potential contributions of genetic and neuroimaging biomarkers that represent risk for AD and vascular cognitive impairment. We can gain additional insight by including the vantage points offered by this broader perspective.
If cognitive processing speed remains significantly related to everyday abilities in MCI and early dementia while accounting for AD risk biomarkers, maintaining or improving processing speed might extend the period of relatively preserved function in these patient groups. While not the focus of the present analyses, a long-term aim of this research is to identify genetic, neuroimaging, and cognitive characteristics of persons who are able to benefit from training interventions targeting processing speed, a modifiable cognitive construct, with corresponding gains in daily functioning.
The aim of the present research was to examine, among persons on the clinical continuum of MCI, the interrelationships of genetic risk factors, structural magnetic resonance imaging (MRI) markers of AD-related neurodegeneration, and processing speed as they relate to driving skills, IADL function, and community mobility. Specifically, we perform cross-sectional analyses, using data obtained at baseline from a longitudinal study, to first examine the associations of genetic, neuroimaging, and sociodemographic characteristics with concurrent performance on a composite measure of cognitive processing speed. We then examine the association of processing speed performance, as well as specific genetic and neuroimaging characteristics, with composite measures of IADL function, on-road driving performance, and community mobility.
Method
Participants with diagnoses along the spectrum of MCI due to AD were drawn from the Applying Programs to Preserve Skills (APPS) study at the University of Alabama at Birmingham (UAB). Briefly, the APPS study is a randomized controlled trial examining functional outcomes facilitated by a computer-based processing speed intervention in persons with MCI, relative to a closely matched active control training condition. The study was conducted from August 2014 to October 2019; each participant was involved in study assessments and cognitive interventions for 2 years. The prespecified primary outcomes over a 2-year period are comparisons between study arms in pre- to post-training changes in the performance of a functional activity composite (everyday IADLs), an on-road driving evaluation, and a mobility composite. The present study examines cross-sectional, pre-randomization, and pre-training relationships among the domains of interest. We will examine in a future report the primary outcomes as well as the relative contributions of biologic and other key factors in identifying who is able to benefit from cognitive training, how much, for how long, and in what domains of function.
Participants
Potential participants with a working diagnosis of MCI were recruited to the APPS study from the UAB Memory Disorders (n = 119) and Neuropsychology Clinics (n = 19), the UAB Geriatrics Clinic (n = 51), and a local private neuropsychology practice (n = 3).
After obtaining patients’ permission, the recruiting sites referred potential participants to the study. The APPS program manager then administered an eligibility screening by telephone. The initial inclusion criterion for study screening was a clinic referral with a working diagnosis of MCI during the previous year. Initial exclusion criteria included neurological illness, clinical stroke, traumatic brain injury, brain tumor, and psychiatric conditions that might reasonably account for cognitive difficulties. Persons who were using medications for memory loss, were not current drivers, or had contraindications for MRI were not excluded from participation.
Of 192 individuals referred for study consideration, 1 died before screening; 4 were ineligible due to screening exclusion criteria of stroke, brain injury, or Parkinson’s disease; and 56 declined to participate due to distance, poor health, time constraints, or lack of interest. Of the remaining 131 who enrolled at the first of 2 baseline study visits (described below), 22 were determined by an adjudication panel to be ineligible to continue. Reasons for ineligibility included absence of current cognitive impairment or evidence of decline, advanced AD, and psychiatric disturbance. Another 6 participants dropped out of the study between the first and second baseline visits. The remaining 103 participants completed both baseline visits and met the criterion of current adjudicated diagnoses of MCI (n = 90) or very mild probable dementia (n = 13) in which AD was presumed to be the major or a contributing etiology. Within the MCI classification, 30 participants were classified as amnestic single domain MCI, 45 amnestic multidomain MCI, 3 nonamnestic single domain MCI, 1 nonamnestic multidomain MCI, 2 cognitive impairment—cannot classify, and 9 possible MCI (defined as testing results and history suggestive of cognitive decline without frank impairment within any domain). These 103 participants with MCI and mild dementia are the focus of our analyses. Of these, 99 were current drivers, and a subset of 86 were eligible for MRI.
Procedures
Baseline testing was administered to enrollees at 2 in-person visits scheduled 2–6 weeks apart. At the first visit, participants provided written informed consent after an extensive consent interview which included screening with the Competency Assessment Checklist for Research Informed Consent (© Daniel Marson, JD, PhD, University of Alabama at Birmingham 2002; MCI Study version 1 July 2004). A predoctoral student in clinical psychology administered vision screening, a comprehensive neuropsychological test battery, a depression scale, and self-reported medical history and functional status questionnaires; this student also recorded behavioral observations. Meanwhile, an informant (most often a spouse or adult child) was interviewed by the program manager using the Functional Assessment Questionnaire (FAQ) (11). For each participant, results of these assessments were randomly assigned to 2 members of the adjudication panel for diagnosis classification. The adjudication panel was composed of selected study investigators, including neuropsychologists, geropsychologists, and a neurologist, to determine current (baseline) diagnoses using clinical procedures consistent with the National Institute on Aging and Alzheimer’s Association (NIA-AA) criteria for MCI due to AD (12,13), with the exception that biomarkers were not considered. Split decisions between the 2 independent reviewers were decided by a majority of the full panel at meetings held twice monthly as needed.
Participants who were eligible to continue based on their adjudicated classification completed a second baseline visit within 6 weeks. At this visit, all participants completed objective measures of functional abilities and saliva sampling for genetic risk markers. Licensed drivers completed an on-road driving evaluation (n = 99), and participants without contraindications to MRI completed a structural MRI (n = 86).
Participants received compensation for the baseline visits and additional compensation for undergoing the MRI. The UAB Institutional Review Board approved all study procedures.
Neurocognitive measures
The neuropsychological evaluation included assessment of global cognition (Dementia Rating Scale-2 [DRS-2]) (14), as well as multiple tests of the following cognitive domains: attention, working memory and processing speed, verbal and visual memory, executive function, verbal fluency, confrontation naming, and visuospatial skills. Further details are available elsewhere (15). The neuropsychological measures used by the study adjudicators as part of their diagnostic classifications are displayed in Supplementary Table 1.
Self-reported measures
Self-reported measures of medical history, concomitant medications, health habits, and depressive symptoms were also administered during this evaluation using a comorbidity checklist, the Center for Epidemiologic Studies Depression Scale (CES-D) (16), and a Health-Related Quality of Life Scale (17).
Measures of everyday function
Performance-based measures of everyday function were administered at the second baseline study visit. Results of these evaluations were not accessible to the study adjudicators.
Instrumental activities of daily living
The Timed Instrumental Activities of Daily Living (TIADL) (18) uses performance-based tasks to evaluate speed and accuracy within 5 domains: telephone use, nutrition identification, basic financial abilities, grocery shopping, and medication management. Time to complete tasks within each domain is recorded, within a time limit ranging from 2 to 3 minutes, depending on the task. The 5 domains are assessed with a total of 8 tasks. For the 2 domains consisting of more than one task, the task-level completion times within domain were averaged. Thus, we calculated 5 domain-level task completion times and then averaged these. The resulting scores were converted to Z-scores using published completion-time norms for cognitively normal older adults (19). Test–retest reliability for the TIADL among older adults over 8 weeks is good (Pearson correlation coefficient = .85) (18).
Financial capacity
The Financial Capacity Instrument-Short Form (FCI-SF) (20) was derived from the Financial Capacity Instrument (FCI long form) (21). The short form consists of 37 items that assess 5 domains of financial ability: Mental Calculation, Financial Conceptual Knowledge, Simple Checkbook/Register Task, Complex Checkbook/Register Task, and Using a Bank Statement. The FCI-SF has both performance and processing speed dimensions, and generates both a total performance score (accuracy) and a separate total processing speed score. In the present study, we used the total performance score as our primary FCI-SF outcome variable. We selected this score because in addition to reflecting accuracy, certain items in the performance index have strict time limits which, if exceeded, contribute to the performance score. On the other hand, the processing speed index is independent of accuracy (ie, one can complete an item quite quickly but inaccurately). The FCI-SF total performance score ranges from 0 to 74 (higher is better). The FCI-SF has excellent interrater reliability (96% exact agreement) and internal reliability (a = .90) (20).
Driving-related competencies
The Useful Field of View (UFOV) is a computer- administered measure of the amount of information one can process in a brief glance without moving the eyes or head (22). The UFOV evaluates cognitive processing speed with 4 increasingly difficult subtests of simple attention, divided attention, selective attention, and dual object discrimination within a selective attention paradigm. Each subtest includes multiple trials at varying display durations ranging from 16.67 to 500 ms. The program’s algorithm records the briefest display duration at which the participant correctly identifies the targeted response in 75% of trials within a given subtest. The total score is the sum of the 4 subtest scores, with higher scores indicating slower processing speed (23). Motor response time is not evaluated and does not contribute to the score. The test–retest reliability of the touch-screen version of the UFOV used in this study is .735 (22). Performance on the measure is related to at-fault motor vehicle crash involvement based on state crash records in both retrospective and prospective studies (24).
Mobility
Life space
The UAB Life Space Assessment measures one’s mobility in the community, ranging in distance from one’s bedroom to out of town, irrespective of the mode of transportation (25). This self-report assessment queries one’s range of life space over the past 4 weeks, as well as the frequency and independence of mobility. Scores range from 0 to 120, with higher scores reflecting greater mobility. The intraclass correlation coefficient for test–retest reliability is .96 (25).
Road sign reactions
The Road Sign Test (26) is a computer-administered test measuring reaction time to varying traffic signs. After orientation to the computer mouse and practice with the task, the participant either clicks or moves the mouse to the right or left, depending on the given sign and its task demand. The mean reaction time for sets of Road Sign Test trials (3 stimuli and 6 stimuli) were averaged for each participant. Higher scores indicate slower reaction time. The internal consistency between 3- and 6-stimuli trials was high in prior research (Cronbach’s α = .92) (27). This measure was a component of the “Everyday Speed” composite outcome in the Advanced Cognitive Training for Independent and Vital Elderly study (26), in which test–retest reliability was reported to be modest (r = .56).
On-road driving assessment
The on-road driving evaluation was conducted on a standardized route through urban and suburban settings in Birmingham, Alabama (28,29). Participants who were current drivers with a valid driver’s license were given a 45–60 minutes on-road driving evaluation in fair weather conditions in collaboration with the UAB Driving Assessment Clinic. Participants drove an instrumented vehicle with dual brakes under the guidance of a Certified Driving Rehabilitation Specialist (CDRS) who is also a licensed occupational therapist, as well as a back seat rater (Driving Assessment Clinic director and study investigator Cynthia Owsley, PhD). The CDRS rode in the front passenger seat and was responsible for giving directions and making verbal and/or physical interventions as needed. At the end of the evaluation, the CDRS provided summary judgments of driving skills, including an assessment of Global Driving Performance, which is made using a 5-point Likert scale: 5 = optimal, 4 = few minor flaws/satisfactory, 3 = unsatisfactory but not unsafe, 2 = unsafe, and 1 = drive terminated. The back seat rater made detailed ratings in real time at 50 predetermined points along the route. The backseat rater also independently rated the participant’s Global Driving Performance using the same 5-point Likert scale. Both raters were masked to the cognitive performance and specific diagnoses of participants. Prior research in which 2 backseat raters coded global performance using this scale obtained an interrater correlation of .96 (30). In the present study, the global rating scores of the CDRS and back seat coder had acceptable interrater agreement (kappa = .86, p < .001). Exact agreement was not expected due to the different physical perspectives and responsibilities of the front and back seat coders. The 2 ratings were averaged to form each participant’s Global Driving Performance score.
Genetic risk markers
Participants’ saliva (~2 mL) was collected at the second baseline study visit using Oragene DNA collection kits for DNA extraction with the prepIT-L2P reagent (DNA Genotek, Ottawa, Ontario, Canada). DNA samples were stored at −80°C. Genotyping was performed with the ABI 7500 Fast Real-Time PCR System using TaqMan 5’ nuclease on demand assays (Applied Biosystems).
A priori, we selected 5 of the most promising genetic markers associated with risk of late onset AD, each of which may exert an effect on AD expression through more than one biological pathway, including cholesterol metabolism, endocytosis, inflammation and immune function, and genetic networks such as lipoprotein–inflammation interactions. These markers are APOE, CLU, SORL1, PICALM, and TNF α. Polymorphisms robustly associated with risk for AD and cognitive function were evaluated. These are: APOE (e4 allele carrier status), CLU (SNP rs11136000 T allele), PICALM (SNP rs3851179; SNPrs541458), TNF-α-rs1800629, and 6 SORL1 SNPs (rs668387, rs689021, rs641120, rs3824968, rs2282649, and rs1010159). APOE e4 carrier status (positive or negative) was evaluated separately from the other SNP markers. The other markers were combined to form a polygenic risk score, calculated as the product of the odds ratios for AD for the 10 SNPs derived from www.alzgene.org. The ancestral gene was used as the reference for the polygenic risk score.
Neuroimaging
Structural MRI scans were performed at UAB’s Philips 3 Tesla Cardiovascular MRI facility during the second baseline study visit. The selected MRI sequences are an exact subset of those conducted at this facility among hundreds of participants in previous multisite research studies (eg, 30). The specific protocol used in the current study is described in detail elsewhere (31). Briefly, it included 1 mm isotropic 3D T1, T2, and FLAIR imaging. The prespecified neuroanatomical markers of interest in the present analysis were left hippocampal volume relative to total intracranial volume (as an inverse index of hippocampal atrophy) (32), white matter lesions as a marker of cerebrovascular disease (abnormal white matter relative to total white matter volume) (33), and a measure of AD-related atrophy called the Spatial Pattern of Abnormality for Recognition of Early Alzheimer’s disease (SPARE-AD) algorithms (34,35). SPARE-AD scores represent similarity to either AD brains or cognitively normal brains. The more positive the score, the more AD-like the brain. Scores >0 are more AD-like than normal. Scan time for each participant was approximately 30 minutes. Within 24 hours of the scan, the image was deidentified and transmitted from UAB to the MRI Reading Center at the University of Pennsylvania (U Penn) using a DICOM image transfer software program (TRIAD) supported by the American College of Radiology Imaging Network (ACRIN) in Philadelphia, Pennsylvania.
Composite scores
The test scores selected to create composites of processing speed, function, and mobility were specified by the investigators prior to initiation of this research. To evaluate the domains of interest, we calculated Z-scores for each participant’s individual test scores using published age- and education-stratified norms and standard deviations (SDs) (Z = observed score minus normative population mean divided by the normative SD). All Z-scores for the individual tests were truncated to + or −3 SDs to minimize the influence of outlying scores. (None reached the threshold of +3 SDs or higher.) We then formed composite measures of processing speed, IADL function (ie, functional composite), and mobility in order to avoid test-specific findings and as a data reduction strategy to preserve power for detecting effects. The test scores used to form the composites in the present analyses were selected based on face validity and conventional and/or prior usage as exemplars of the domains of interest.
Processing Speed Composite
The full Processing Speed Composite (PSC) consists of Z-scores obtained on the following timed tests: the Coding subtest of the Wechsler Adult Intelligence Scale, fourth edition (WAIS-IV) (36); Semantic Fluency (animals) from the Consortium to Establish a Registry for Alzheimer’s Disease battery (37); the Benton Controlled Oral Word Association Test (COWA) (38); Trail Making, Part B (39); and the UFOV, subtest 2 (normalized and reverse coded). The UFOV subtest 2 (divided attention) was selected because among the subtests, subtest 2 has been administered alone as an exemplar of visual processing speed (eg, Friedman et al. (40 ) and Ball et al. (41)), performs best for discriminating between cognitively normal persons and persons with MCI (42), and is the subtest most highly correlated with the total UFOV score (23,41). The average (mean) of each participant’s Z-scores on these tests is the full PSC score. A partial PSC score that excludes the UFOV subtest 2 was calculated for models evaluating predictors of the Functional Composite (FC) to avoid overlap, as the FC includes the UFOV total score as an index of driving-related competencies.
Functional Composite
The FC is composed of participants’ scores on the Timed IADL, the FCI-SF, and the total UFOV. Specifically, we used the mean of: (i) TIADL Z-score (mean of 5 domain completion time Z-scores, reverse coded); (ii) FCI-SF total Z-score (total raw score converted to an age-adjusted scaled score, further adjusted for education, and then converted to Z-score); and (iii) UFOV Z-score (UFOV total score, normalized and reverse scored).
Mobility Composite
The Mobility Composite (MC) is the mean of: (i) the Life Space Z-score and (ii) the Road Signs Test Z-score (reverse coded so that higher is better).
Global driving assessment
For simplicity, the Global Driving Assessment for the current analysis is based solely on the on-road driving evaluations of overall performance. Global Driving Score is the raw score mean of the front seat and back seat coders’ ratings, ranging from 1 to 5. Higher scores indicate better driving skills.
Statistical Analyses
Characteristics for the study sample were described using means and SDs for continuous variables and frequencies (percentages) for categorical variables.
In regression analysis, 4 outcomes were examined: the PSC, the FC, the MC, and the Global Driving Score. Using linear regression analyses, we estimated the univariate associations between each composite variable and individual predictor variables including sociodemographics, genetics and neuroimaging. Multivariable linear regression analyses were then performed to examine the association between each outcome variable and predictor variables entered simultaneously. Because both the PSC and FC were generated from Z-scores derived from tests with published normative values for age and education, no further adjustments for age and education were made in those regression analyses. Models examining the MC and Global Driving Score were adjusted for age, gender, and education. Results were considered statistically significant at p <.05.
In a sensitivity analysis, we calculated correlations of the component tests within each composite to each other and to the relevant composite scores.
In post hoc analyses, we examined the association of SPARE-AD scores to left hippocampal volumes. We also compared the characteristics of persons who did and did not undergo MRI scans.
Statistical analyses were performed using version 3.5.1 of the R programming environment (https://www.R-project.org).
Results
Table 1 displays the characteristics of this sample of persons on the continuum of MCI to very early dementia due to AD. The sample was evenly distributed between men and women, consisted primarily of White participants, and was highly educated. The average depressive symptom score of 9.4 was below the CES-D threshold for elevated distress (ie, a cutpoint of 16 of 30). The FAQ mean of 4.2 indicates informants’ assessments of some difficulty or assistance needed by participants in performing IADLs. This FAQ score is below the threshold for probable dementia, which has been found to range from 6 to 9 depending on the study (11,43,44). The DRS-2 mean score of 130 (SD 7.98) corresponds to that reported in an independent study among persons with MCI (45). In that study, the DRS-2 mean score was 130.67 (SD 7.85), with a range of 124–136. The majority (58.3%) of our participants were carriers of the APOE4 allele, indicating increased risk for AD. Composite Z-scores ranged from −0.66 for processing speed to −1.53 for mobility, consistent with the expected range of values for persons with MCI.
Table 1.
Baseline Characteristics of the Participants
| Total N | n (%) | Mean (SD) | |
|---|---|---|---|
| Age (years) | 103 | 73.5 (7.37) | |
| Women (self-report) | 103 | 54 (52.4%) | |
| Race (self-report) | 103 | ||
| African American | 9 (8.74%) | ||
| Asian | 2 (1.94%) | ||
| White | 92 (89.3%) | ||
| Education (years) | 103 | 15.6 (2.73) | |
| CES-Da raw score | 103 | 9.42 (7.47) | |
| FAQb score | 99 | 4.16 (3.56) | |
| DRS-2c raw score | 101 | 130 (7.98) | |
| APOE4d positive | 103 | 60 (58.3%) | |
| SNPe score | 103 | 0.84 (0.15) | |
| LH-adj.f (%) | 86 | 25.6 (3.36) | |
| WMLg (%) | 86 | 0.89 (1.01) | |
| SPARE-ADh | 86 | 1.28 (1.69) | |
| Composite Z-scores | |||
| Processing Speed Composite—fulli | 103 | −0.71 (0.82) | |
| Processing Speed Composite—partialj | 103 | −0.66 (0.83) | |
| Functional Compositek | 103 | −0.89 (0.96) | |
| Mobility Compositel | 103 | −1.53 (1.19) | |
| Global Driving Scorem | 99 | 3.97 (0.96) |
Notes: aCES-D = Center for Epidemiologic Studies Depression Scale, range 0–60. Higher scores indicate more depressive symptoms. bFAQ = Functional Assessment Questionnaire, administered to informant, range 0–30. Higher scores indicate more difficulty, assistance, or dependence in everyday instrumental activities. cDRS-2 = Mattis Dementia Rating Scale v. 2, range 0–144. Higher scores indicate better cognitive function. dAPOE4 = apolipoprotein e4 allele. Positive cases are carriers of 1 or both e4 alleles. eSNP score = product of the odds ratios for Alzheimer’s disease for 10 SNPs from www.alzgene.org (mutant vs ancestral). The SNPs are CLU-rs11136000, PICALM1-rs3851179, PICALM1-rs541458, TNF-α-rs1800629, and 6 SORL1 SNPs (rs668387, rs689021, rs641120, rs3824968, rs2282649, and rs1010159). fLH-adj. = left hippocampal volume expressed as percentage of total intracranial volume. The parameter estimate for LH-adjusted has been multiplied by 1000 for the ease of reading. gWML = white matter lesion load. WML is the volume of abnormal white matter expressed as percentage of total white matter. hSPARE-AD = multivariate metric of global cerebral atrophy. iProcessing Speed Composite—full is the mean of: (i) WAIS-IV Coding Z-score, (ii) Category fluency Z-score, (iii) Phonemic fluency Z-score, (iv) Trails B Z-score (truncated at −3), and (v) UFOV subtest 2 (divided attention) Z-score (normalized, reversed, and truncated at −3). Higher scores represent better processing speed. jProcessing Speed Composite—partial is the mean of: (i) WAIS-IV Coding Z-score, (ii) Category fluency Z-score, (iii) Phonemic fluency Z-score, and (iv) Trails B Z-score (truncated at −3). Higher scores represent better processing speed. kFunctional Composite is the mean of: (i) TIADL Z-score (mean of 5 domain Z-scores for completion times, reversed and truncated at −3), (ii) FCI total Z-score (FCI total → age-adjusted scaled score → age/education-adjusted scaled score → Z-score), and (iii) UFOV Z-score (UFOV total score → normalized and reversed, truncated at −3). Higher scores represent better everyday instrumental activities of daily living function. lMobility Composite is the mean of: (i) Life Space Z-score and (ii) Road Signs Test Z-score. Higher scores represent better mobility and greater road awareness. mGlobal Driving Score is derived from the on-road driving assessment and is the mean of: front seat coder rating and back seat coder rating, range 1–5. Higher scores indicate better driving skills.
Both univariate and multivariable linear regression models were used to examine the associations among the variables of interest. The results are summarized in Table 2, which includes the parameter estimates (ie, unstandardized regression coefficients) and p-values. Supplementary Table 2 summarizes similar results using standardized regression coefficients. In univariate analyses, the PSC was negatively associated with the SPARE-AD score (p = .001), indicating that slower processing speed is associated with a higher level of AD-like whole brain neuroanatomical patterns. However, APOE4 carrier status, SNP score, white matter hyperintensity volume and left hippocampal volume were not individually associated with cognitive processing speed. Better PSC performance, in turn, was positively and strongly associated with better scores on the FC and MC (univariate parameter estimates of 0.78 and 0.66, respectively), as well as the Global Driving Score (parameter estimate of 0.45) (all p-values <.001). SPARE-AD scores were negatively associated with these domains of function, indicating that patterns of cerebral atrophy consistent with AD are associated with poorer IADL function, everyday mobility, and driving skills (all p-values <.05).
Table 2.
Linear Regression Models of Association Between Outcome Composite Scores and Predictors (unstandardized coefficients)
| Composite Score | N | Univariate Parameter Estimate | p-Value | n | Multivariable Parameter Estimate | p-Value | |
|---|---|---|---|---|---|---|---|
| Processing Speed Composite (PSC)—fulla | 86 | ||||||
| APOE4: yes b | 103 | −0.161 | .331 | −0.132 | .459 | ||
| SNP c | 103 | −0.510 | .365 | −0.320 | .589 | ||
| SPARE-AD d | 86 | −0.174 | .001* | −0.228 | <.001* | ||
| WML e | 86 | −2.146 | .814 | 2.494 | .777 | ||
| LH-adj. f | 86 | −0.041 | .880 | −0.593 | .052 | ||
| Functional Compositeg | 86 | ||||||
| PSC-partial h | 103 | 0.777 | <.001* | 0.663 | <.001* | ||
| APOE4: yes | 103 | −0.138 | .475 | 0.046 | .768 | ||
| SNP | 103 | −0.609 | .353 | −0.292 | .573 | ||
| SPARE-AD | 86 | −0.261 | <.001* | −0.171 | .003* | ||
| WML | 86 | −6.418 | .542 | −1.636 | .832 | ||
| LH-adj. | 86 | 0.129 | .682 | −0.138 | .612 | ||
| Mobility Compositei | 86 | ||||||
| PSC-full | 103 | 0.662 | <.001* | 0.572 | <.001* | ||
| APOE4: yes | 103 | −0.261 | .285 | 0.179 | .453 | ||
| SNP | 103 | −0.581 | .485 | −0.135 | .865 | ||
| SPARE-AD | 86 | −0.296 | <.001* | −0.188 | .034* | ||
| WML | 86 | −5.655 | .699 | 0.168 | .989 | ||
| LH-adj. | 86 | 0.402 | .297 | 0.024 | .953 | ||
| Age, years | 0.007 | .690 | |||||
| Gender: men | 0.416 | .088 | |||||
| Education, years | 0.027 | .540 | |||||
| Global Driving Scorej | 83 | ||||||
| PSC-full | 99 | 0.447 | <.001* | 0.402 | .002* | ||
| APOE4: yes | 99 | −0.236 | .205 | −0.144 | .463 | ||
| SNP | 99 | −0.706 | .265 | −0.652 | .320 | ||
| SPARE-AD | 83 | −0.170 | .005* | −0.103 | .157 | ||
| WML | 83 | −1.382 | .905 | 2.796 | .791 | ||
| LH-adj. | 83 | 0.035 | .909 | −0.204 | .541 | ||
| Age, years | −0.035 | .022* | |||||
| Gender: men | 0.383 | .060 | |||||
| Education, years | 0.051 | .170 |
Notes: aPSC = Processing Speed Composite—full is the mean of: (i) WAIS-IV Coding Z-score, (ii) Category fluency Z-score, (iii) Phonemic fluency Z-score, (iv) Trails B Z-score (truncated at −3), and (v) UFOV subtest 2 Z-score (normalized, reversed, and truncated at −3). Higher scores represent better processing speed. bAPOE4 = apolipoprotein e4 allele, positive cases are carriers of 1 or both e4 alleles. cSNP score = product of the odds ratios for Alzheimer’s disease for 10 SNPs from www.alzgene.org (mutant vs ancestral). The SNPs are CLU-rs11136000, PICALM1-rs3851179, PICALM1-rs541458, TNF-a-rs1800629, and 6 SORL1 SNPs (rs668387, rs689021, rs641120, rs3824968, rs2282649, and rs1010159). dSPARE-AD = a multivariate metric of global cerebral atrophy. eWML = white matter lesion load. WML is the volume of abnormal white matter expressed as percentage of total white matter. fLH-adj. = left hippocampal volume expressed as percentage of total intracranial volume. The parameter estimate for LH-adj. has been multiplied by 1000 for the ease of reading. gFunctional Composite is the mean of: (i) TIADL Z-score (mean of 5 area Z-scores, truncated at −3), (ii) FCI total Z-score (FCI total → age-adjusted scaled score → age/education-adjusted scaled score → Z-score), and (iii) UFOV Z-score (UFOV total score → normalized and reversed, truncated at −3). Higher scores represent better everyday IADL function. hPSC = Processing Speed Composite, partial is the mean of: (i) WAIS-IV Coding Z-score, (ii) Category fluency Z-score, (iii) Phonemic fluency Z-score, and (iv) Trails B Z-score (truncated at −3). Higher scores represent better processing speed. iMobility Composite is the mean of: (i) Life Space Z-score and (ii) Road Signs Test Z-score. Higher scores represent wider mobility and greater road awareness. jGlobal Driving Score from the on-road driving assessment is the mean of front seat coder rating and back seat coder rating, range 1–5. Higher scores indicate better driving skills.
*p < .05 (bold font).
The results of multivariable regression models, also in Table 2, demonstrate that the same variables (SPARE-AD for the PSC; both Processing Speed and SPARE-AD for FC and MC) emerged as significant predictors, with parameter estimates of similar magnitude and significance (0.66, 0.57, and 0.40, respectively, for function, mobility, and driving), while controlling for the remaining predictor variables (PSC multivariable model R2 = .174, adj. R2 = .122; FC multivariable model R2 = .527, adj. R2 = .492; MC multivariable model R2 = .350, adj. R2 = .273). For the Global Driving Score, better Processing Speed and younger age were significant predictors of better driving ratings (multivariable model R2 = .360, adj. R2 = .281). APOE4 carrier status, SNP scores, adjusted left hippocampal volume scores, and white matter lesion volumes were not significant predictors in these models, and SPARE-AD scores also were not a significant predictor of global driving after multivariable adjustment.
Results of the sensitivity analysis examining correlations among the component tests of each composite and their relationships to the relevant composite scores are displayed in Table 3.
Table 3.
Correlation Among Composite Scores and Individual Component Scores
| Processing Speed Composite (PSC)—full | |||||
|---|---|---|---|---|---|
| PSC | Digit Symbol | Animal Fluency | COWA | Trails B | |
| Coding | .60**** | ||||
| Category fluency | .64**** | .29** | |||
| Phonemic fluency | .68**** | .29** | .40**** | ||
| Trails B | .76**** | .42**** | .29** | .32** | |
| UFOV subtest 2 | .72**** | .25* | .32*** | .31** | .47**** |
| Processing Speed Composite (PSC)—partial | |||||
| PSC—Partial | Digital Symbol | Animal Fluency | COWA | ||
| Coding | .65**** | ||||
| Category fluency | .67**** | .29** | |||
| Phonemic fluency | .73**** | .29** | .40**** | ||
| Trails B | .76**** | .42**** | .29** | .32** | |
| Functional Composite | |||||
| Functional Composite | TIADL | FCI Total | |||
| TIADL | .80**** | ||||
| FCI-SF total | .82**** | .58**** | |||
| UFOV | .83**** | .45**** | .48**** | ||
| Mobility Composite | |||||
| Mobility Composite | Life Space | ||||
| Life Space | .47**** | ||||
| Road Signs | .91**** | .05 |
Notes: FCI-SF = Financial Capacity Instrument-Short Form; UFOV = Useful Field of View.
*p < .05. **p < .01. ***p < .001. ****p < .0001.
Post hoc analyses revealed a modest but statistically significant relationship between SPARE-AD scores and left hippocampal volume (R2 = .21, p < .001). Comparison of the characteristics of participants who did and did not undergo MRI revealed no differences in any characteristic (all p-values >.05, data not shown).
Discussion
The main finding of this research is that cognitive processing speed is strongly associated with everyday function in older adults with MCI and very early AD, independent of genetic risks and neuroimaging biomarkers of AD. This finding was consistent across composites of IADL performance and community mobility, as well as on-road driving. The degree to which one’s brain structures conform to atrophic patterns consistent with AD, as measured by SPARE-AD scores from structural MRI images, is also a significant independent correlate of everyday IADL function and mobility. In this study, genetic risk markers, white matter lesion volume, and left hippocampal atrophy were not significant predictors of function.
The magnitude of the associations of processing speed to IADL function, mobility, and global driving ratings was large (parameter estimates of .66, .57, and .40 respectively). These parameters indicate that each unit of standardized processing speed (ie, 1 SD) is associated with a large incremental difference in performance within these domains. For example, 1 SD increment or decrement in processing speed is independently associated with a 0.66 SD unit difference in composite IADL function, denoting a strong and clinically meaningful relationship. A prior study investigating the relationship of global cognition to everyday function in persons with amnestic MCI used a dementia screening instrument, the DRS-2, and informant-reported function using the Everyday Cognition (Ecog) assessment rather than performance-based assessments (46). In that study, the DRS-2 total score accounted for 19% of the variance in informant-reported function. Our findings suggest that cognitive processing speed is a particularly sensitive marker of functional status, as it is the primary predictor within a multivariable model that accounts for 49% of the variance in objectively measured function. In comparison to full neuropsychological batteries and a global dementia screener such as the DRS-2, processing speed is relatively simple to evaluate using select brief and precise instruments. Our findings suggest that processing speed is an even stronger correlate of everyday abilities than structural MRI patterns consistent with AD, and certainly it is more accessible and cost-effective to measure. While SPARE-AD is relatively specific to AD-related neurodegeneration, clinical constructs like processing speed better capture the combined effects of neurodegeneration, comorbidities that contribute to cognitive decline, and cognitive reserve—each of which is important for functional outcomes.
Our own previous research in a different sample of persons with MCI and cognitively normal controls found that speed of IADL task performance was also compromised in persons with MCI relative to controls (47). In that research, the disparity in speed of TIADL performance between participants with and without MCI grew larger over a 5-year period, with MCI participants displaying significantly steeper rates of decline. We also found that the rate of decline for accuracy on financial tasks was steeper over time among the MCI participants. Furthermore, the disparity between speed and accuracy (speed sacrificed for accuracy, and vice versa) was larger for participants with MCI than for controls. The overall results reflected steeper rates of decline, with a magnified trade-off between speed and accuracy, in those with MCI relative to cognitively normal controls (47).
We found a modest but statistically significant relationship between SPARE-AD scores and left hippocampal volume. This is not unexpected for the correlation between a single brain region (left hippocampus) and a measure of global brain changes (SPARE-AD). Hippocampal atrophy would be expected to manifest primarily as memory deficits (eg, Vyhnalek et al. (48)). As a more global measure of brain changes, SPARE-AD scores encompass additional domains such as visuospatial skills, executive function, and processing speed that are affected by AD and have been shown to play a greater role in functional deficits among adults with amnestic MCI (49). Thus, it is not surprising that SPARE-AD scores predominate among the biomarkers in our models of everyday function.
Importantly, cognitively demanding activities such as handling finances and driving are key to independence in the vast majority of the U.S. population, and diminished capacity for successful performance of these activities can expose one to devastating consequences. Specifically, poor financial abilities and judgment are arguably the area of daily competency rendering one most vulnerable to exploitation and fraud, while poor driving skills render one susceptible to risk for injury, personal liability, and mortality. Thus, monitoring processing speed—the cognitive skill heavily engaged in these activities—can provide an indication of vulnerability to potential difficulties. Research has suggested that durable improvement of processing speed in persons with MCI is possible (50,51). Whether this improvement transfers to improved performance of key daily activities in persons with MCI has not yet been established.
Our study has limitations as well as strengths. Our sample was highly educated, predominantly white, and clinically referred rather than community-based. These characteristics may limit the generalizability of our findings. However, we did allow for comorbid vascular disease and risk factors, and we did not require MRI eligibility; thus, our sample was more broadly representative than is common in clinic-based AD research with narrow eligibility criteria. In addition, although we chose to examine only selected genetic markers with moderate to strong evidence of association with AD, the size of our sample likely provided inadequate power for detecting effects of individual SNPs on processing speed and function. In addition, we did not conduct tests of mediation or moderation for the variables examined in this study. Future work may include more detailed examination of these relationships. This research has a number of strengths. We enrolled persons with a working diagnosis of MCI from within the past year, and then we independently adjudicated the current diagnostic classification of each case, using a panel of experts. This panel used standard diagnostic criteria but did not have access to information obtained regarding genetics, neuroimaging, UFOV scores, on-road driving, performance-based function, or everyday mobility. Thus, these variables are distinct from diagnostic classifications made at baseline (and annually thereafter). Our use of objective, performance-based measures for IADL function and driving, and a combination of performance and self-report evaluating everyday mobility, is another strength of this research, as both self-report and informant-report are subject to potential reporting biases. We used multimodal assessments of cognition, of AD risk biomarkers including both structural MRI markers and saliva-based genetic assays, and of everyday function. Concurrent measurements across this array of variables in a relatively large sample of persons along the spectrum of MCI are another strength of this research.
In persons with generally normal cognition, processing speed is a pliable cognitive domain that is amenable to specific training protocols (28) which in recent years have been adapted for commercial use (Brain HQ, ©2020 Posit Science). Whether meaningful gains can be achieved by persons with MCI, and whether such gains will maintain function in cognitively demanding everyday activities, driving skills and safety, and general mobility are the questions underlying this program of research. Answers to these questions will be the focus of a future report.
Supplementary Material
Funding
This work was supported by the National Institute on Aging at the National Institutes of Health (grant numbers R01 AG045154 to V.G.W. and P30 AG022838).
Conflict of Interest
D.C.M. is the inventor of the FCI-SF financial capacity assessment measure which is owned by the UAB Research Foundation. D.C.M. was a funded co-investigator on the APPS research project but his grant role was unrelated to the FCI-SF. He currently receives grant funding and is a paid consultant in connection with other NIH grants using the FCI-SF. He previously was a paid consultant and received royalties from Janssen in connection to the FCI-SF. All other authors report no conflicts of interest.
Author Contributions
V.G.W., R.N.B., M.C., G.D., G.E., D.S.G., R.C.P.G., C.L.L.-G., D.C.M., I.M.N., C.O., R.T.P., K.A.S., and R.E.K. made substantial contributions to the conception or design of the work; V.G.W., T.B., C.B., L.D., C.L.L.-G., O.A.M., M.C.M., I.N.M., C.O., J.P., R.T.P., G.P., and K.A.S. made substantial contributions to the acquisition of the data; T.B., Y.Z., and R.E.K. made substantial contributions to the analysis of the data; V.G.W., T.B., Y.Z., R.N.B., M.C., G.E., D.S.G., R.C.P.G., C.L.L.-G., O.A.M., D.C.M., I.M.N., C.O., J.P., K.A.S., and R.E.K. made substantial contributions to the interpretation of the data; V.G.W. drafted the work. All authors revised the work critically for important intellectual content, gave final approval of the version to be published, and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
References
- 1.Tabert MH, Albert SM, Borukhova-Milov L, et al. . Functional deficits in patients with mild cognitive impairment: prediction of AD. Neurology. 2002;58:758–764. doi: 10.1212/wnl.58.5.758 [DOI] [PubMed] [Google Scholar]
- 2.Tuokko H, Morris C, Ebert P. Mild cognitive impairment and everyday functioning in older adults. Neurocase. 2005;11:40–47. doi: 10.1080/13554790490896802 [DOI] [PubMed] [Google Scholar]
- 3.Ueda M, Takayama Y, Koyama Y, Osada H. A detection of very mild Alzheimer’s disease and MCI by memory complaints and deterioration of IADL in a questionnaire for a memory clinic. No to Shinkei. 2006;58:865–871. [PubMed] [Google Scholar]
- 4.Wadley VG, Crowe M, Marsiske M, et al. . Changes in everyday function in individuals with psychometrically defined mild cognitive impairment in the Advanced Cognitive Training for Independent and Vital Elderly Study. J Am Geriatr Soc. 2007;55:1192–1198. doi: 10.1111/j.1532-5415.2007.01245.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Aartsen MJ, Smits CH, van Tilburg T, Knipscheer KC, Deeg DJ. Activity in older adults: cause or consequence of cognitive functioning? A longitudinal study on everyday activities and cognitive performance in older adults. J Gerontol B Psychol Sci Soc Sci. 2002;57:P153–P162. doi: 10.1093/geronb/57.2.p153 [DOI] [PubMed] [Google Scholar]
- 6.Brigman S, Cherry KE. Age and skilled performance: contributions of working memory and processing speed. Brain Cogn. 2002;50:242–256. doi: 10.1016/s0278-2626(02)00510-9 [DOI] [PubMed] [Google Scholar]
- 7.Owsley C, McGwin G Jr. Association between visual attention and mobility in older adults. J Am Geriatr Soc. 2004;52:1901–1906. doi: 10.1111/j.1532-5415.2004.52516.x [DOI] [PubMed] [Google Scholar]
- 8.Bublak P, Redel P, Sorg C, et al. . Staged decline of visual processing capacity in mild cognitive impairment and Alzheimer’s disease. Neurobiol Aging. 2011;32:1219–1230. doi: 10.1016/j.neurobiolaging.2009.07.012 [DOI] [PubMed] [Google Scholar]
- 9.Fabrigoule C, Rouch I, Taberly A, et al. . Cognitive process in preclinical phase of dementia. Brain. 1998;121:135–141. doi: 10.1093/brain/121.1.135 [DOI] [PubMed] [Google Scholar]
- 10.Chehrehnegar N, Nejati V, Shati M, et al. . Early detection of cognitive disturbances in mild cognitive impairment: a systematic review of observational studies. Psychogeriatrics. 2020;20:212–228. doi: 10.1111/psyg.12484 [DOI] [PubMed] [Google Scholar]
- 11.Pfeffer RI, Kurosaki TT, Harrah CH Jr, Chance JM, Filos S. Measurement of functional activities in older adults in the community. J Gerontol. 1982;37:323–329. doi: 10.1093/geronj/37.3.323 [DOI] [PubMed] [Google Scholar]
- 12.Winblad B, Palmer K, Kivipelto M, et al. . Mild cognitive impairment—beyond controversies, towards a consensus: report of the International Working Group on Mild Cognitive Impairment. J Intern Med. 2004;256:240–246. doi: 10.1111/j.1365-2796.2004.01380.x [DOI] [PubMed] [Google Scholar]
- 13.Petersen RC, Aisen P, Boeve BF, et al. . Mild cognitive impairment due to Alzheimer disease in the community. Ann Neurol. 2013;74:199–208. doi: 10.1002/ana.23931 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jurica PJ, Leitten CL, Mattis S. DRS-2 Dementia Rating Scale-2 Professional Manual. Psychological Assessment Resources; 2001. https://www.scirp.org/reference/ReferencesPapers.aspx?ReferenceID=1535467. Accessed April 13, 2020. [Google Scholar]
- 15.Steward KA, Bull TP, Kennedy R, Crowe M, Wadley VG. Neuropsychological correlates of anosognosia for objective functional difficulties in older adults on the mild cognitive impairment spectrum. Arch Clin Neuropsychol. 2020;35:365–376. doi: 10.1093/arclin/acz065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Radloff LS. The CES-D Scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385–401. doi: 10.1177/014662167700100306 [DOI] [Google Scholar]
- 17.Kazis LE, Miller DR, Skinner KM, et al. . Applications of methodologies of the Veterans Health Study in the VAHealth Care System: conclusions and summary. J Ambul Care Manage. 2006;29:182–188. doi: 10.1097/00004479-200604000-00011 [DOI] [PubMed] [Google Scholar]
- 18.Owsley C, Sloane M, McGwin G Jr, Ball K. Timed instrumental activities of daily living tasks: relationship to cognitive function and everyday performance assessments in older adults. Gerontology. 2002;48:254–265. doi: 10.1159/000058360 [DOI] [PubMed] [Google Scholar]
- 19.Wadley VG, Okonkwo O, Crowe M, Ross-Meadows LA. Mild cognitive impairment and everyday function: evidence of reduced speed in performing instrumental activities of daily living. Am J Geriatr Psychiatry. 2008;16:416–424. doi: 10.1097/JGP.0b013e31816b7303 [DOI] [PubMed] [Google Scholar]
- 20.Gerstenecker A, Eakin A, Triebel K, et al. . Age and education corrected older adult normative data for a short form version of the Financial Capacity Instrument. Psychol Assess. 2016;28:737–749. doi: 10.1037/pas0000159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Marson DC, Sawrie SM, Snyder S, et al. . Assessing financial capacity in patients with Alzheimer disease: a conceptual model and prototype instrument. Arch Neurol. 2000;57:877–884. doi: 10.1001/archneur.57.6.877 [DOI] [PubMed] [Google Scholar]
- 22.Edwards JD, Vance DE, Wadley VG, Cissell GM, Roenker DL, Ball KK. Reliability and validity of Useful Field of View test scores as administered by personal computer. J Clin Exp Neuropsychol. 2005;27:529–543. doi: 10.1080/13803390490515432 [DOI] [PubMed] [Google Scholar]
- 23.Edwards JD, Ross LA, Wadley VG, et al. . The Useful Field of View test: normative data for older adults. Arch Clin Neuropsychol. 2006;21:275–286. doi: 10.1016/j.acn.2006.03.001 [DOI] [PubMed] [Google Scholar]
- 24.Owsley C, Ball K, McGwin G Jr, et al. . Visual processing impairment and risk of motor vehicle crash among older adults. J Am Med Assoc. 1998;279:1083–1088. doi: 10.1001/jama.279.14.1083 [DOI] [PubMed] [Google Scholar]
- 25.Baker PS, Bodner EV, Allman RM. Measuring life-space mobility in community-dwelling older adults. J Am Geriatr Soc. 2003;51:1610–1614. doi: 10.1046/j.1532-5415.2003.51512.x [DOI] [PubMed] [Google Scholar]
- 26.Ball K, Berch DB, Helmers KF, et al. ; Advanced Cognitive Training for Independent and Vital Elderly Study Group . Effects of cognitive training interventions with older adults: a randomized controlled trial. J Am Med Assoc. 2002;288:2271–2281. doi: 10.1001/jama.288.18.2271 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fazeli PL, Ross LA, Vance DE, Ball K. The relationship between computer experience and computerized cognitive test performance among older adults. J Gerontol B Psychol Sci Soc Sci. 2013;68:337–346. doi: 10.1093/geronb/gbs071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wadley VG, Okonkwo O, Crowe M, et al. . Mild cognitive impairment and everyday function: an investigation of driving performance. J Geriatr Psychiatry Neurol. 2009;22:87–94. doi: 10.1177/0891988708328215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wood JM, McGwin G Jr, Elgin J, et al. . On-road driving performance by persons with hemianopia and quadrantanopia. Invest Ophthalmol Vis Sci. 2009;50:577–585. doi: 10.1167/iovs.08-2506 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.The SPRINT MIND Investigators for the SPRINT Research Group, Nasrallah IM, Pajewski NM, Auchus AP, et al. . Association of intensive vs standard blood pressure control with cerebral white matter lesions. J Am Med Assoc. 2019;322:524–534. doi: 10.1001/jama.2019.10551 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Steward KA, Kennedy R, Erus G, Nasrallah IM, Wadley VG. Poor awareness of IADL deficits is associated with reduced regional brain volume in older adults with cognitive impairment. Neuropsychologia. 2019;129:372–378. doi: 10.1016/j.neuropsychologia.2019.04.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Doshi J, Erus G, Ou Y, et al. ; Alzheimer’s Neuroimaging Initiative . MUSE: MUlti-atlas region segmentation utilizing ensembles of registration algorithms and parameters, and locally optimal atlas selection. Neuroimage. 2016;127:186–195. doi: 10.1016/j.neuroimage.2015.11.073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Zacharaki EI, Kanterakis S, Bryan RN, Davatzikos C. Measuring brain lesion progression with a supervised tissue classification system. Med Image Comput Comput Assist Interv. 2008;11:620–627. doi: 10.1007/978-3-540-85988-8_74 [DOI] [PubMed] [Google Scholar]
- 34.Fan Y, Batmanghelich N, Clark CM, Davatzikos C; Alzheimer’s Disease Neuroimaging Initiative . Spatial patterns of brain atrophy in MCI patients, identified via high-dimensional pattern classification, predict subsequent cognitive decline. Neuroimage. 2008;39:1731–1743. doi: 10.1016/j.neuroimage.2007.10.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Davatzikos C, Xu F, An Y, Fan Y, Resnick SM. Longitudinal progression of Alzheimer’s-like patterns of atrophy in normal older adults: the SPARE-AD index. Brain. 2009;132:2026–2035. doi: 10.1093/brain/awp091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wechsler D. Wechsler Adult Intelligence Scale Fourth Edition. Pearson Education, Inc.; 2008. [Google Scholar]
- 37.Morris JC, Heyman A, Mohs RCet al. . The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology. 1989;39:1159–1165. doi: 10.1212/WNL.39.9.1159 [DOI] [PubMed] [Google Scholar]
- 38.Ruff RM, Light RH, Parker SB, Levin HS. Benton Controlled Oral Word Association Test: reliability and updated norms. Arch Clin Neuropsychol. 1996;11:329–338. [PubMed] [Google Scholar]
- 39.Reitan RM, Wolfson D.. The Halstead–Reitan Neuropsychological Test Battery: Therapy and Clinical Interpretation. Neuropsychological Press; 1985. [Google Scholar]
- 40.Friedman C, McGwin G Jr, Ball KK, Owsley C. Association between higher order visual processing abilities and a history of motor vehicle collision involvement by drivers ages 70 and over. Invest Ophthalmol Vis Sci. 2013:54:778–782. doi: 10.1167/iovs.12-11249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Ball KK, Roenker DL, Wadley VG, et al. . Can high-risk older drivers be identified through performance-based measures in a Department of Motor Vehicles setting? J Am Geriatr Soc. 2006;54:77–84. doi: 10.1111/j.1532-5415.2005.00568.x [DOI] [PubMed] [Google Scholar]
- 42.Okonkwo OC, Wadley VG, Ball K, Vance DE, Crowe M. Dissociations in visual attention deficits among persons with mild cognitive impairment. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn. 2008;15:492–505. doi: 10.1080/13825580701844414 [DOI] [PubMed] [Google Scholar]
- 43.Cruz-Orduña I, Bellón JM, Torrero P, et al. . Detecting MCI and dementia in primary care: effectiveness of the MMS, the FAQ and the IQCODE [corrected]. Fam Pract. 2012;29:401–406. doi: 10.1093/fampra/cmr114 [DOI] [PubMed] [Google Scholar]
- 44.Teng E, Becker BW, Woo E, Knopman DS, Cummings JL, Lu PH. Utility of the Functional Activities Questionnaire for distinguishing mild cognitive impairment from very mild Alzheimer disease. Alzheimer Dis Assoc Disord. 2010;24:348–353. doi: 10.1097/WAD.0b013e3181e2fc84 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Springate BA, Tremont G, Papandonatos G, Ott BR. Screening for mild cognitive impairment using the Dementia Rating Scale-2. J Geriatr Psychiatry Neurol. 2014;27:139–144. doi: 10.1177/0891988714522700 [DOI] [PubMed] [Google Scholar]
- 46.Greenaway MC, Duncan NL, Hanna S, Smith GE. Predicting functional ability in mild cognitive impairment with the Dementia Rating Scale-2. Int Psychogeriatr. 2012;24:987–993. doi: 10.1017/S1041610211002717 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Lassen-Greene CL, Steward K, Okonkwo O, et al. . Mild cognitive impairment and changes in everyday function over time: the importance of evaluating both speed and accuracy. J Geriatr Psychiatry Neurol. 2017;30:220–227. doi: 10.1177/0891988717711807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Vyhnalek M, Nikolai T, Andel R, et al. . Neuropsychological correlates of hippocampal atrophy in memory testing in nondemented older adults. J Alzheimers Dis. 2014;42(suppl. 3):S81–S90. doi: 10.3233/JAD-132642 [DOI] [PubMed] [Google Scholar]
- 49.Putcha D, Tremont G. Predictors of independence in instrumental activities of daily living: Amnestic versus nonamnestic MCI. J Clin Exp Neuropsychol. 2016;38:991–1004. doi: 10.1080/13803395.2016.1181716 [DOI] [PubMed] [Google Scholar]
- 50.Valdes EG, O’Connor ML, Edwards JD. The effects of cognitive speed of processing training among older adults with psychometrically defined mild cognitive impairment. Curr Alzheimer Res. 2012;9:999–1009. doi: 10.2174/156720512803568984 [DOI] [PubMed] [Google Scholar]
- 51.Li H, Li J, Li N, Li B, Wang P, Zhou T. Cognitive intervention for persons with mild cognitive impairment: a meta-analysis. Ageing Res. Rev. 2011;10:285–296. doi: 10.1016/j.arr.2010.11.003 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.