Executive functioning and processing speed as predictors of global cognitive decline in Alzheimer's disease

Abstract

Background

Better cognitive tools to predict disease progression in mild cognitive impairment (MCI) and Alzheimer's disease (AD) are needed.

Objective

In this prospective longitudinal cohort, we are testing if changes in the cognitive domains of executive functioning and processing speed can predict global cognitive decline.

Methods

We assessed patients with MCI, AD, and cognitively healthy controls (cHC) using NIH toolbox assessments for processing speed and executive functioning and overall cognitive decline by the Alzheimer's Disease Assessment Scale-Cognitive (ADAS-Cog).

Results

Among 184 participants over a median follow-up of 540 days, both between- and within-subjects variance in NIH toolbox and ADAS-Cog assessments increased from cHC to MCI to AD patients. Among patients with AD (n = 24), pattern comparison processing speed (PCPS) and dimensional change card sort tests (DCCS) declined at 3 and 6 months prior to global cognitive decline (p = 0.008 and 0.0012). A 5-point decrease in either PCPS or DCCS increased risk of global cognitive decline (HR 1.32 (1.08–1.60) and 1.62 (1.16–2.26)).

Conclusions

Testing for cognitive domains of processing speed and executive functioning may predict subsequent global cognitive.

Keywords

Alzheimer's disease Alzheimer's disease assessment scale-cognitive cognitive testing mild cognitive impairment NIH toolbox

Introduction

Today there are nearly 6.9 million older Americans living with Alzheimer's disease (AD).¹ AD is a progressive neurodegenerative disorder; however, there is much variability in the observed rates of cognitive decline across the spectrum of AD.^2–4 The average duration of the disease varies between 4 and 8 years, with some upwards of 20 years.⁵ There are known demographic and clinical characteristics as well as radiological and genetic features that associate with the rate of cognitive decline.^6,7 More recently both central and peripheral biomarkers have been identified as predicting long-term cognitive decline.^8–10 However, these risk factors do not provide short-term predictions of AD disease trajectory. The ability to predict the timing of when a patient will become completely dependent on others would be a powerful tool, enabling patients and families to optimize care and, with newer disease-modifying therapeutics, possibly optimize timing of therapeutic interventions.

The sequence of deficits affecting different cognitive domains in AD commonly starts with memory however an individual's impairment in executive functioning and processing speed exist in the early stages of disease.^11,12 Early pathological changes in AD that involve the medial temporal lobe affect episodic memory¹³ however, decline in AD may be linked to executive functioning and processing speed. Executive functioning defines the higher-level cognitive skills used to control other cognitive abilities^14,15 while processing speed is the time it takes to execute multi-step information processing in a mental tasks.¹⁶ Both executive functioning and processing speed support other cognitive processes^17,18 and may serve as a useful cognitive markers for the early trajectory of AD symptoms, being potentially predictive of a more rapid decline.^19,20 Thus, both of these domains may serve as predictors of an inflection point in AD symptoms, when the individual might begin a period of more rapid global cognitive decline, losing function and freedom.

In current clinical practice, there are no established guidelines or tools for monitoring cognitive function after diagnosis for the sole purpose of predicting decline. We reasoned that given that deficits in executive function and processing speed often precede more severe cognitive deficits, standardized testing of these two domains may hold value in clinical practice. While not specifically developed for older populations, the U.S. National Institutes of Health Toolbox for the Assessment of Neurological and Behavioral Function (NIH toolbox) includes a Cognition Battery (CB) that includes brief, comprehensive tests for these functions.²¹ This CB contains a Dimensional Change Card Sort (DCCS) test that is a sensitive and reliable measure of executive functioning and cognitive flexibility,²² and a Pattern Comparison Processing Speed Test (PCPS), which evaluates processing speed.²³ These NIH toolbox assessments are easy to administer and offer standardized, non-invasive measures which can be compared across studies.²¹

The aim of this interim analysis of the Gut-brain Alzheimer's disease Inflammation and Neurocognitive Study (GAINS) cohort was to evaluate early cognitive data as potential predictors of AD-associated global cognitive decline as measured by the modified Alzheimer's Disease Assessment Scale-Cognitive Subscale 13 (ADAS-Cog13). ADAS-Cog is a well-established standard for the assessment of cognitive function in patients with AD and is routinely used to help differentiate patients in research as having mild cognitive impairment (MCI), AD, or normal cognition²⁴ as well as to measure clinically relevant changes.^25,26 The ADAS-Cog is a comprehensive cognitive rating scale to assess cognitive functioning in AD.^27,28 The addition of executive functioning and functional ability items to create ADAS-Cog13 from the original 11 question set (ADAS-Cog11) has improved the test's sensitivity in milder disease,^29,30 but the test still lacks an assessment of processing speed and skews towards language and verbal memory tasks. The ADAS-Cog13 version we used has more recently been shown useful for disease progression modeling.³¹ Greater changes in ADAS-Cog and ADAS-Cog13 scores over time have been associated with AD versus MCI.³²

We followed participants in the GAINS cohort at 3-month intervals and assessed their NIH toolbox DCCS and PCPS and ADAS-Cog13 scores. In this longitudinal cohort, we are testing the hypothesis of can a 4-point decrease in the cognitive domains of DCCS or PCPS be predictive of global cognitive and functional decline as measured by the ADAS-Cog13.

Methods

Study setting and population

Older adults, ≥60 years of age, living independently, were recruited into the GAINS and included those diagnosed with AD, MCI, or had no cognitive issues, serving as healthy controls (cHC). GAINS subjects did in-person visits with cognitive testing every 90 days for upwards of 2 years. Subjects were included in this sub-group analysis if they had completed 4 study visits (270 days). This prospective cohort study was approved by the institutional review board at the University of Massachusetts Chan Medical School.

Data collection

We collected demographic measures at enrollment, which included age, sex, race, ethnicity, level of education, and past medical history. At enrollment and at each subsequent visit, we collected information on nutritional status, frailty, and any hospitalizations or changes in medication. We assessed nutritional status using the Mini Nutritional Assessment (MNA) tool, which combines anthropometric assessment (weight, height and weight loss), general assessment (lifestyle, medication, and mobility), dietary assessment (number of meals, food and fluid intake), and autonomy of eating self-assessment (self-perception of health and nutrition), into a score that ranges from 0 to 30 points.^33–35 Subjects were categorized as normal, at risk, or malnourished based on the MNA. Frailty was categorized according to the validated and widely-utilized Canadian Study of Health and Aging's (CSHA) 7-point Clinical Frailty Scale where a healthcare professional assessed the ability to perform daily activities and assigns a score from 1 (very fit) to 9 (terminally ill) based on their level of frailty.³⁶ Malnutrition and frailty we specifically collected from subjects because both are known to associate with cognitive decline.^37,38

During each visit, cognitive testing was performed. We used the modified Alzheimer's Disease Assessment Scale-Cognitive Subscale 13 (ADAS-Cog13). ADAS-Cog was developed to measure cognitive dysfunction in AD, but is now used for assessment of cognition in other dementia or pre-dementia populations, and is one of the most widely-used cognitive scales in clinical trials.^32,39 We also utilized the NIH toolbox to assess changes in cognition under the subdomains of processing speed, using the Pattern Comparison Processing Speed Test (PCPS), and executive functioning using the Dimensional Change Card Sort (DCCS).⁴⁰ The Toolbox PCPS provides a reliable measurement of complex processing speed over the lifespan that is sensitive to neurological insults.⁴¹

Longitudinally, we used a 4-point change in the ADAS-Cog13 score from the day 0 visit as clinically meaningful based off of prior clinical trials, either as an improvement (≤ −4 points) or as a decline ≥ + 4 points).^42–44 Selection of this change threshold in ADAS-Cog13 has recently been demonstrated in a systematic review investigating the level of multiple cognitive tests to detect minimal clinical importance especially in validation studies they presented.⁴⁵ Although differences exist in the literature from other investigations having used different thresholds of change (ranging 2–4), we chose a 4-point change for this investigation with the belief of it being a more rigorous choice than lower values. Those that remained within 4 points of their initial visit ADAS-Cog13 scores were categorized as stable, while those with improvement were categorized as improved, and declining as decline. We chose the first timepoint (or visit date) where a subject had a change in ADAS-Cog13 score of ≤−4 points as the point of decline and the previous visit as the 3 months prior, and the visit before that as 6 months prior to decline. All other previous timepoints were labeled as baseline for analysis.

Diagnostic criterion for AD and MCI

Determination of AD diagnosis was made by previous cognitive testing coupled with various neuroimaging techniques performed by the subjects’ own physician anywhere from 1 to 10 years prior to enrollment in this study. For MCI and cHC we used criteria including the Clinical Dementia Rating scale (CDR), a global rating tool based on a semi structured interview of a subject and caregiver on the clinical judgment of a clinician.^46,47 MCI criteria determined at enrollment was followed as in Ganguli et al.⁴⁸ included (1) impaired memory: Word List Delayed Recall score <1 SD above mean; (2) screening score indicative of normal mental status: Mini-Mental State Examination score 25+;⁴⁹ (3) normal daily functioning; (4) memory complaint: we asked the subject and family member caring for the subject if they perceived the subject as having a memory issue; and (5) either normal or very mild cognitive impairment: CDR score <1.⁴⁸ The cHC group had to have a CDR = 0 and no complaints of memory issues.

Polypharmacy, known to associate with dementia,⁵⁰ was included as a variable and defined using the commonly reported definition of five or more daily medications.^38,51 All survey data was collected by trained research staff or study physicians. Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Massachusetts Chan Medical School.⁵²

Data analysis

We used chi-square tests to compare categorical variables, and analysis of variance (ANOVA) for continuous variables between patient groups with different outcomes from their baseline data. In cases where there were different variances between groups (i.e., cognitive scores), we used Kruskal-Wallis test in place of ANOVA. The equality of standard deviations (variance), reported in Table 3, was tested using the sdtest package in Stata⁵³ to determine whether ANOVA or Kruskal-Wallis test was most appropriate.⁵⁴ Cox proportional hazards models were used to evaluate the NIH neuropsychological measures that predicted time to significant global cognitive decline, as defined by a greater than 4-point increase in ADAS-Cog13 testing, as described above. The hazard ratio indicates the change in risk per 5-unit change in the predictor. For instance, if the hazard ratio for PCPS is 1.57, each 5-point change in PCPS equates to an increases risk by 57%. Two sets of adjusted Cox models were completed each for PCPS and DCCS, where these neuropsychological measures were adjusted for age (in years), sex, education level, frailty, malnutrition, and polypharmacy. The software used for the analyses was Stata, Release 13.1 (StataCorp LLC, College Station, TX) and Prism Release 10 (GraphPad by Dotmatics, Ltd, United Kingdom). We used a p-value cutoff of 0.05 as statistically significant.

Results

Clinical and cognitive scores differed among GAINS cohort patient groups

We enrolled 243 older adults, and at the time of this analysis, 184 (75.7%) completed 4 study visits for a longitudinal length of 270 days and were eligible for the analysis. There were no significant differences among those subjects that completed this study compared to those that dropped. The average length of time in the study was 564 (sd 260) days. Of these 184 adults in the GAINS cohort, 131 (71.2%) were cHC while 24 (13.0%) had AD and 29 (15.8%) had MCI. This cohort was predominately white (95.1%) and non-Hispanic (91.8%). At the first visit there were 128 (69.6%) with CDR = 0, 43 (23.4%) with CDR = 0.5, 8 (4.4%) with CDR = 1, 3 (1.6%) with CDR = 2 and 2 (1.1%) with CDR = 3. Unsurprisingly, adults with AD and MCI were older, with higher frailty and malnutrition scores as well as taking more daily medications as defined by polypharmacy (Table 1). This is consistent with what is known about frailty and malnutrition in relation to AD and mild cognitive impairment.^37,55–57 Both frail and pre-frail older adults usually have poorer cognitive status,⁵⁸ and frailty has been linked to the extent of AD pathophysiology.⁵⁶ Malnutrition is also closely linked to decreased cognitive functioning.⁵⁵ Not surprisingly, mean score on the ADAS-Cog13 increased from cHC, to MCI and AD subjects, and NIH Toolbox module mean scores decreased from cHC, to MCI and AD subjects (Table 1). These results further validate the ADAS-Cog13 in responsiveness between cognitively healthy individuals and those with MCI versus AD. They also demonstrate the utility of the NIH toolbox cognitive modules in distinguishing these populations.

Table 1.

Clinical characteristics of the GAINS cohort at initial visit.

	HC	MCI	AD	p
Age *	70.2 (7.6)	75.4 (7.2)	74.3 (5.6)	<0.001
Male	39 (29.8)	15 (51.7)	11 (45.8)	0.042
Education level	6.0 (1.5)	5.8 (1.2)	5.7 (1.7)	0.43
Malnutrition Score*	1.2 (0.5)	1.3 (0.5)	1.8 (0.7)	<0.001
CFS *	2.1 (1.0)	2.6 (1.0)	3.8 (1.4)	<0.001
Polypharmacy	34 (26.0)	10 (34.5)	16 (66.7)	<0.001
BMI	27.7 (6.0)	26.7 (4.9)	26.1 (6.3)	0.44
ADAS-Cog13	8.9 (4.6)	19.4 (9.2)	36.8 (22.5)	<0.001
PCPS	93.5 (14.2)	83.6 (16.4)	66.2 (20.1)	<0.001
DCCS	101.2 (10.0)	96.8 (12.5)	84.0 (18.3)	<0.001

Data presented as n (%) unless marked with * then presented means (sd). CFS: Clinical Frailty Score; BMI: body mass index; ADAS-Cog13: Alzheimer's Disease Assessment Scale-Cognitive Subscale 13; PCPS: NIH Toolbox Pattern Comparison Processing Speed Test; DCCS: NIH Toolbox Dimensional Change Card Sort; GAINS: Gut-brain Alzheimer’s disease Inflammation and Neurocognitive Study.

Cognitive outcomes in AD was not correlated with initial cognitive scores

Among the GAINS cohort we noted 3 cognitive trajectory patterns on longitudinal ADAS-Cog13 testing. Using a change in score of ±4 in ADAS-Cog testing for 2 or more timepoints to categorize outcomes, 48 patients improved their scores over time (26.1%), 119 patients remained stable (64.7), and 17 patients experienced cognitive decline (9.2%). Of those with AD, 14 had cognitive decline (58.3%) while 8 remained stable (33.3%) and 2 improved (8.3%). Among AD patients, there were no significant differences among the 3 cognitive outcomes observed (Table 2). Importantly, there were no differences in the day 0 cognitive scores using the ADAS-Cog or NIH toolbox among the AD subjects by cognitive trajectory outcome. In the cHC and MCI patient groups, there were a combined n = 5 patients with decline (3.1%), who were older and had worse initial ADAS-Cog testing scores (Supplemental Table 1a and 1b).

Table 2.

Baseline characteristics of Alzheimer's disease patients by cognitive outcomes.

	Improve (2)	Stable (8)	Decline (14)	p
Age (years/SD)*	76.0 (4.2)	76.5 (4.3)	72.7 (6.2)	0.30
Male	0 (0.1)	4 (50.0)	7 (50.0)	0.40
Education level*	5.0 (2.8)	6.1 (1.7)	5.6 (1.7)	0.52
Malnutrition Score*	1.5 (0.7)	2.0 (0.9)	1.8 (0.6)	0.64
CFS*	3.0 (1.4)	3.8 (1.8)	3.9 (1.3)	0.71
Polypharmacy	1 (50.0)	6 (75.0)	9 (64.3)	0.77
BMI*	29.8 (4.9)	26.0 (5.8)	25.7 (7.0)	0.70
ADAS-Cog13*	31.3 (25.0)	38.8 (29.9)	38.3 (19.2)	0.86
PCPS *	59.0 (14.1)	74.3 (19.3)	60.7 (19.4)	0.31
DCCS*	79.0 (19.8)	94.2 (16.1)	80.5 (19.3)	0.39

Significant variance in cognitive testing exists between cognitively healthy older adults and those with MCI and AD

To further characterize the baseline characteristics of our study groups (cHC, MCI, AD), we examined the variance between individuals (between) as well as within-person variance (within) (Table 3) as standard deviations are not easy to interpret without a frame of reference.⁵⁹ Both the between- and within- variance increased from cHC to MCI to AD patients for ADAS-Cog13 testing, with only a slight increase from cHC to MCI (Table 3). For both the NIH toolbox assessment of PCPS and DCCS, we did not observe differences in within-group variance across groups. This is reflected in the increasing intraclass correlation coefficient as one goes from cHC to MCI to AD among all cognitive tests (Supplemental Table 2, all p < 0.01). The values in Table 3 indicate that there is considerable within-person variability in each of the measures of cognitive functioning. Moreover, the within-person to between-person variability ratios were somewhat larger for the cHC patients compared to those with MCI or AD mostly due to the near doubling of between variance as we move from HC to MCI to AD across all cognitive testing types. The variance for the cHC and MCI groups indicate that the variation for a given individual from one test occasion to the next is more than half as much as the variation from one person to the next. This was the opposite tendency for AD patients where variation for the individual was one quarter to one third of the variation from one person to the next across cognitive tests.

Table 3.

Variance among different cognitive group types.

	HC	MCI	AD
ADAS-Cog13
Mean (SD)	7.5 (4.5)	17.0 (10.4)	37.8 (23.3)
Between variance	3.9	10.7	23.4
Within variance	2.5	3.1	5.7
Within/Between	0.64	0.29	0.24
PCPS
Mean (SD)	99.6 (14.8)	89.9 (17.8)	70.6 (20.5)
Between variance	12.5	16.7	21.4
Within variance	8.1	8.0	8.1
Within/Between	0.65	0.50	0.38
DCCS
Mean (SD)	103.0 (8.8)	99.7 (10.8)	87.7 (17.1)
Between variance	7.2	10.1	19.4
Within variance	5.3	5.5	6.1
Within/Between	0.74	0.54	0.31
P-value	<0.01	<0.01	<0.01

p-values calculated using Levene's test. HC: healthy control; MCI: mild cognitive impairment; AD: Alzheimer's disease; ADAS-Cog13: Alzheimer's Disease Assessment Scale-Cognitive Subscale 13; PCPS: NIH Toolbox Pattern Comparison Processing Speed Test; DCCS: NIH Toolbox Dimensional Change Card Sort.

NIH toolbox assessments of executive functioning decline as early as 6 months prior to cognitive decline measured by ADAS-Cog13

In light of the differences in variances between tests among patients, we next sought to determine whether longitudinal patterns were correlated with cognitive testing trends, among the AD patients within the GAINS cohort. We first explored differences between those who experienced a decline in cognition versus those who did not (both stable and improved patients). We did not observe any differences in demographic or clinical characteristics, including in the medical diagnoses (i.e., hypertension and diabetes) or medications taken by the subjects during the study period (Supplemental Table 3). We did, however, notice a significant decline in executive functioning testing, in both NIH toolbox PCPS and DCCS, at the 3- and 6-month visits before a decline in cognition measured by the ADAS-Cog13 compared to all other timepoints where the ADAS-Cog13 testing was stable (Figure 1(A)). There were no significant differences in ADAS-Cog13 testing scores at these same time points. The differences from baseline score tests are visualized in Figure 1(B).

Figure 1.

Clinical cognitive score testing among Alzheimer's disease patients. (A) Average clinical scores for the Alzheimer's Disease Assessment Scale-Cognitive Subscale 13 (ADAS-Cog13) were lowest among Alzheimer's Disease with stable cognition (Stable) and then not significantly different at 6 months and 3 months prior to a point of cognitive decline (Decline). Scores for the NIH Toolbox assessment for Pattern Comparison Processing Speed Test (PCPS) and Dimensional Change Card Sort (DCCS) did significantly drop at both 6- and 3-month timepoints prior to ADAS-Cog13 decline. (B) Average changes in the inverse of the ADAS-Cog13 (black triangles), PCPS (blue circles) and DCCS (red boxes) compared to baseline demonstrate the change in both NIH toolbox assessments 6- and 3-months prior to a change in ADAS-Cog13.

We next performed univariate cox models predicting cognitive decline by ADAS-Cog for both NIH toolbox domains for PCPS and DCCS among AD patients. We used a 5-point decline in both NIH CB tests as predictor of general cognitive decline given the mean change at both 3 and 6 months prior in each test was about 5 points. Both in the crude and adjusted models, a 5-point change in PCPS or DCCS was associated with a 24% to 32% or 44% to 62% increased risk of cognitive decline respectively (Table 4). In the modeling, adjustments were made for age, sex, education, frailty, malnutrition and polypharmacy.

Table 4.

Analysis of NIH toolbox cognitive testing and risk of cognitive decline.

	Crude		Adjusted^a
	Hazard Ratio (95% CI)	p	Hazard Ratio (95% CI)	p
PCPS delta	1.24 (1.08–1.43)	0.002	1.32 (1.08–1.60)	0.006
DCCS delta	1.44 (1.16–1.77)	0.001	1.62 (1.16–2.26)	0.005

PCPS: NIH Toolbox Pattern Comparison Processing Speed Test; DCCS: NIH Toolbox Dimensional Change Card Sort; delta is for every 5-point decrease in score.

Adjusted for age, sex, education, frailty, malnutrition, and polypharmacy

Discussion

In this study we found that as early as 6 months prior to a global cognitive decline, there were significant decreases in processing speed and executive functioning among AD patients using the NIH toolbox CB assessments for PCPS and DCCS. There was also a noted increase in the between-person variability from cHC to MCI to AD subjects using either the ADAS-Cog or NIH toolbox modules; however, the within-group variability only increased between these groups in the ADAS-Cog assessments, with the greatest variability in among AD patients. A 5-point decrease in either PCPS and DCCS resulted in a greater than 30% increased risk of subsequent global cognitive decline. We propose here that these NIH toolbox assessments for executive functioning and processing speed may serve as a monitoring tool to predict which AD patients will go on to experience clinically significant cognitive decline.

Executive function comprises the higher-level cognitive skills used to control and coordinate other cognitive abilities and behaviors.^14,15 Deficits in executive functioning have been shown as one of the most useful cognitive markers for the early detection of AD,¹⁹ and declining performance can be detected 2–3 years before the diagnosis of AD.⁶⁰ In this investigation we used this marker in a different context, that being after AD diagnosis. Executive dysfunction occurs in all stages of AD and has been linked to functional decline in activities of daily living.^61,62 Executive functioning is also one of the cognitive domains in AD that shows the greatest decline and can indicate a faster disease progression.^63–65 The ADAS-Cog test does a poor job of testing for executive functioning. It was built based on the original ADAS-Cog,^55,66 and the ADAS-Cog13 or ADAS-Cog-Modified had delayed word recall or number cancellation added in order to improve the tests responsiveness, to cover the range of mild to moderate AD.³⁰ Others tried adding additional testing domains to the ADAS-Cog test to better cover executive functioning testing such as the ADAS-Cog-Exec⁶⁷ and ADAS-Cog-Plus.²⁶ These additional testing modules are not universally used, with the ADAS-Cog still considered the gold standard³² especially for assessing efficacy of AD treatments.

As we age, both executive functioning and processing speed decline, and this decline is linked to performance in learning and memory.⁶⁸ The processing speed theory states that the age-related decline in processing speed is the fundamental mechanism with which memory declines with normal aging.⁶⁹ However, executive functions and processing speed can differentially influence memory decline.⁶⁸ Processing speed is thought to serve as the foundation for other cognitive processes¹⁷ and it is associated with subsequent deficits in other cognitive domains such as working memory,⁷⁰ attention⁷¹ and memory.⁷² Disproportionate slowing of processing speed has been shown to be related to a faster decline in AD.²⁰ Our finding is consistent with a predictive role of processing speed change: the NIH toolbox CB Comparison Processing Speed Test to assess processing speed in GAINS participants was sensitive to cognitive change, with decreases in processing speed preceding global cognitive decline, with a 5-point drop in score increasing the risk of cognitive decline by greater than 30%, after adjustment. We also primarily believe that we did not see this association in the cHC or MCI groups due to the lower number of subjects in these group that experiences a decline. It is also possible that changes in PCPS and DCCS occur earlier before decline and we did not study subjects long enough to capture this time window.

Instead of expanding the ADAS-Cog to improve its sensitivity, it might be of greater benefit to predict cognitive decline by periodically test executive functioning with other standard measures. The NIH toolbox CB is a set of brief measures assessing cognitive domains including executive functioning studies.²¹ Subjects that are part of a clinical trial could self-administer these tests as an alternative to the ADAS-Cog, which takes upwards of 45 min to administer and needs to be done in-person by trained staff.³² Remote digital cognitive testing is now being shown to be an accurate method to detect cognitive impairment.^73–75 A robust tool with a 3- to 6-month predictive window would offer patients and families a reasonable timeframe to make practical care arrangements such as changing living situation, finding caretakers, or applying for elder care benefits and services.

This study does have limitations. First, the majority of the GAINS cohort was without AD, which influences the within-group variance in the AD group. Testing NIH toolbox assessments for their ability to predict cognitive decline in a larger longitudinal cohort would strengthen the findings. Additionally, the GAINS cohort was mostly white and non-Hispanic, limiting the generalizability of our data. These limitations are balanced by the frequent in-person model to assess both ADAS-Cog and the NIH toolbox CB domains. Finally, the AD group is heterogenous in the initial severity score of their AD ranging from CDR scores of 1 to 3. This leaves heterogeneity in initial disease stage as a possible influence the observed trajectories in repeat cognitive scoring.

Conclusion

Based on our findings, we would suggest that brief assessments of both executive functioning and processing speed may serve as a marker of subsequent global cognitive decline. This can be helpful clinically for planning the timing of treatments, interventions, or life management, especially if executive function and processing speed testing is potentially more rapid than a comprehensive cognitive testing session, and more feasibly performed frequently. Tools such as PCPS and DCCS could also help researchers as well as clinicians with monitoring efficacy of AD treatments as part of clinical trials of disease-modifying therapy.

Supplemental Material

sj-docx-1-alr-10.1177_25424823251363549 - Supplemental material for Executive functioning and processing speed as predictors of global cognitive decline in Alzheimer's disease

Supplemental material, sj-docx-1-alr-10.1177_25424823251363549 for Executive functioning and processing speed as predictors of global cognitive decline in Alzheimer's disease by John P Haran, AM Barrett, YuShuan Lai, Samuel N Odjidja, Protiva Dutta, Patrick M McGrath, Imane Samari, Lethycia Romeiro, Abigail Lopes, Vanni Bucci and Beth A McCormick in Journal of Alzheimer's Disease Reports

Footnotes

Acknowledgements

We would like to thank the administration and staff from the Clinical Research Center here at UMass Medical Center and the Center for Clinical and Translational Sciences at UMass Chan Medical School for clinical facilities that supported the GAINS cohort.

ORCID iD

John P Haran

Ethical considerations

This prospective cohort study was approved by the institutional review board at the University of Massachusetts Chan Medical School (IRB docket H00021745) and followed the ethical guidelines set by the Helsinki Declaration. Written informed consent was obtained from all participants or their legal representatives prior to their inclusion in the study.

Consent to participate

All participants provided written informed consent.

Consent for publication

The current data was taken from individual participants who gave informed consent to participate in GAINS with the understanding that results will be disseminated via presentation or publication.

Author contributions

John Haran: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Writing - original draft, Writing - review & editing.

AM Barrett: Investigation, Methodology.

Yushuan Lai: Data curation, Investigation, Visualization, Writing - review & editing.

Samuel Odjidja: Investigation, Project administration, Writing - review & editing.

Protiva Dutta: Investigation, Project administration, Writing - review & editing.

Patrick McGrath: Data curation, Investigation, Writing - review & editing.

Imane Samari: Investigation, Writing - review & editing.

Lethycia Romeiro: Investigation, Writing - review & editing.

Abigail Lopes: Investigation, Project administration, Writing - review & editing.

Vanni Bucci: Conceptualization, Investigation, Writing - review & editing.

Beth McCormick: Conceptualization, Investigation, Methodology, Writing - review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was designed and carried out at the University of Massachusetts Chan Medical School. JPH was supported by an Alzheimer’s Association Grant (2019-AARG-NTF-641955) and NIH grants from the National Institute on Aging (grant numbers: 2019-AARG-NTF-641955, R01AG067483-01).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Supplemental material

Supplemental material for this article is available online.

References

Rajan

Weuve

Barnes

, et al. Population estimate of people with clinical Alzheimer's disease and mild cognitive impairment in the United States (2020–2060). Alzheimers Dement 2021; 17: 1966–1975.

Schindler

Buckles

, et al. Predicting symptom onset in sporadic Alzheimer disease with amyloid PET. Neurology 2021; 97: e1823–e1834.

Koscik

Betthauser

Jonaitis

, et al. Amyloid duration is associated with preclinical cognitive decline and tau PET. Alzheimers Dement (Amst) 2020; 12: e12007.

Insel

Donohue

Berron

, et al. Time between milestone events in the Alzheimer's disease amyloid cascade. Neuroimage 2021; 227: 117676.

Masters

Bateman

Blennow

, et al. Alzheimer's disease. Nat Rev Dis Primers 2015; 1: 15056.

Adak

Illouz

Gorman

, et al. Predicting the rate of cognitive decline in aging and early Alzheimer disease. Neurology 2004; 63: 108–114.

Cosentino

Scarmeas

Helzner

, et al. APOE Epsilon 4 allele predicts faster cognitive decline in mild Alzheimer disease. Neurology 2008; 70: 1842–1849.

Steenland

Zhao

Goldstein

, et al. Biomarkers for predicting cognitive decline in those with normal cognition. J Alzheimers Dis 2014; 40: 587–594.

Gunes

Aizawa

Sugashi

, et al. Biomarkers for Alzheimer's disease in the current state: a narrative review. Int J Mol Sci 2022; 23: 4962.

10.

Wang

Sun

, et al. Biomarkers associated with the pathogenesis of Alzheimer's disease. Front Cell Neurosci 2023; 17: 1279046.

11.

Kim

Kang

, et al. Disproportionate decline of executive functions in early mild cognitive impairment, late mild cognitive impairment, and mild Alzheimer's disease. Dement Neurocogn Disord 2016; 15: 159–164.

12.

Henneges

Reed

Chen

, et al. Describing the sequence of cognitive decline in Alzheimer's disease patients: results from an observational study. J Alzheimers Dis 2016; 52: 1065–1080.

13.

Serrano-Pozo

Frosch

Masliah

, et al. Neuropathological alterations in Alzheimer disease. Cold Spring Harb Perspect Med 2011; 1: a006189.

14.

Nemeth

Chustz

. Chapter 6 - executive functions defined. In: Nemeth

Glozman

(eds) Evaluation and treatment of neuropsychologically compromised children. Cambridge, MA: Academic Press, 2020, pp.107–120.

15.

Guarino

Favieri

Boncompagni

, et al. Executive functions in Alzheimer disease: a systematic review. Front Aging Neurosci 2018; 10: 437.

16.

Ebaid

Crewther

MacCalman

, et al. Cognitive processing speed across the lifespan: beyond the influence of motor speed. Front Aging Neurosci 2017; 9: 62.

17.

Sliwinski

Buschke

. Cross-sectional and longitudinal relationships among age, cognition, and processing speed. Psychol Aging 1999; 14: 18–33.

18.

Rabinovici

Stephens

Possin

. Executive dysfunction. Continuum (Minneap Minn) 2015; 21: 646–659.

19.

Crowell

Luis

Vanderploeg

, et al. Memory patterns and executive functioning in mild cognitive impairment and Alzheimer's disease. Aging Neuropsychol Cogn 2002; 9: 288–297.

20.

Parikh

Hynan

Weiner

, et al. Single neuropsychological test scores associated with rate of cognitive decline in early Alzheimer disease. Clin Neuropsychol 2014; 28: 926–940.

21.

Weintraub

Dikmen

Heaton

, et al. The cognition battery of the NIH toolbox for assessment of neurological and behavioral function: validation in an adult sample. J Int Neuropsychol Soc 2014; 20: 567–578.

22.

Zelazo

Anderson

Richler

, et al. NIH Toolbox cognition battery (CB): validation of executive function measures in adults. J Int Neuropsychol Soc 2014; 20: 620–629.

23.

Carlozzi

Beaumont

Tulsky

, et al. The NIH toolbox pattern comparison processing speed test: normative data. Arch Clin Neuropsychol 2015; 30: 359–368.

24.

Warren

Reid

Whitfield

, et al. Cognitive and behavioral abnormalities in individuals with Alzheimer’s disease, mild cognitive impairment, and subjective memory complaints. Curr Psychol 2024; 43: 800–810.

25.

Schrag

Schott

. What is the clinically relevant change on the ADAS-Cog? J Neurol Neurosurg Psychiatry 2012; 83: 171–173.

26.

Karcher

Savelieva

, et al. Modelling decline in cognition to decline in function in Alzheimer's disease. Curr Alzheimer Res 2020; 17: 635–657.

27.

Rockwood

Fay

Gorman

, et al. The clinical meaningfulness of ADAS-Cog changes in Alzheimer's disease patients treated with donepezil in an open-label trial. BMC Neurol 2007; 7: 26.

28.

Wei

Chen

Lin

, et al. Normative data of Mini-mental state examination, Montreal cognitive assessment, and Alzheimer's disease assessment scale-cognitive subscale of community-dwelling older adults in Taiwan. Dement Geriatr Cogn Disord 2022; 51: 365–376.

29.

Skinner

Carvalho

Potter

, et al. The Alzheimer's disease assessment scale-cognitive-plus (ADAS-Cog-plus): an expansion of the ADAS-Cog to improve responsiveness in MCI. Brain Imaging Behav 2012; 6: 489–501.

30.

Mohs

Knopman

Petersen

, et al. Development of cognitive instruments for use in clinical trials of antidementia drugs: additions to the Alzheimer's disease assessment scale that broaden its scope. The Alzheimer's disease cooperative study. Alzheimer Dis Assoc Disord 1997; 11: S13–S21.

31.

Cho

Woo

Kim

, et al. Disease progression modelling from preclinical Alzheimer's disease (AD) to AD dementia. Sci Rep 2021; 11: 4168.

32.

Kueper

Speechley

Montero-Odasso

. The Alzheimer's disease assessment scale-cognitive subscale (ADAS-Cog): modifications and responsiveness in pre-dementia populations. A narrative review. J Alzheimers Dis 2018; 63: 423–444.

33.

Rubenstein

Harker

Salvà

, et al. Screening for undernutrition in geriatric practice: developing the short-form Mini-nutritional assessment (MNA-SF). J Gerontol A Biol Sci Med Sci 2001; 56: M366–M372.

34.

Saarela

Lindroos

Soini

, et al. Dentition, nutritional status and adequacy of dietary intake among older residents in assisted living facilities. Gerodontology 2016; 33: 225–232.

35.

Guigoz

. The Mini nutritional assessment (MNA) review of the literature–what does it tell us? J Nutr Health Aging 2006; 10: 485–487.

36.

Rockwood

Song

MacKnight

, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ 2005; 173: 489–495.

37.

Buchman

Boyle

Wilson

, et al. Frailty is associated with incident Alzheimer's disease and cognitive decline in the elderly. Psychosom Med 2007; 69: 483–489.

38.

Borda

Ayala Copete

Tovar-Rios

, et al. Association of malnutrition with functional and cognitive trajectories in people living with dementia: a five-year follow-up study. J Alzheimers Dis 2021; 79: 1713–1722.

39.

Connor

Sabbagh

. Administration and scoring variance on the ADAS-Cog. J Alzheimers Dis 2008; 15: 461–464.

40.

Gershon

Wagster

Hendrie

, et al. NIH Toolbox for assessment of neurological and behavioral function. Neurology 2013; 80: S2–S6.

41.

NIH Toolbox Cognitive Battery (NIHTB-CB): The NIHTB Pattern Comparison Processing Speed Test.

42.

Matthews

Korbey

Wilkinson

, et al. Donepezil in Alzheimer's disease: eighteen month results from Southampton Memory Clinic. Int J Geriatr Psychiatry 2000; 15: 713–720.

43.

Aisen

Schafer

Grundman

, et al. Effects of rofecoxib or naproxen vs placebo on Alzheimer disease progression: a randomized controlled trial. JAMA 2003; 289: 2819–2826.

44.

Le Bars

Kieser

Itil

. A 26-week analysis of a double-blind, placebo-controlled trial of the ginkgo biloba extract EGb 761 in dementia. Dement Geriatr Cogn Disord 2000; 11: 230–237.

45.

Muir

Hill

Black

, et al. Minimal clinically important difference in Alzheimer's disease: rapid review. Alzheimers Dement 2024; 20: 3352–3363.

46.

Morris

. Clinical dementia rating: a reliable and valid diagnostic and staging measure for dementia of the Alzheimer type. Int Psychogeriatr 1997; 9: 173–176. discussion 177–178.

47.

Morris

. The clinical dementia rating (CDR): current version and scoring rules. Neurology 1993; 43: 2412–2414.

48.

Ganguli

Dodge

Shen

, et al. Mild cognitive impairment, amnestic type: an epidemiologic study. Neurology 2004; 63: 115–121.

49.

Creavin

Wisniewski

Noel-Storr

, et al. Mini-Mental state examination (MMSE) for the detection of dementia in clinically unevaluated people aged 65 and over in community and primary care populations. Cochrane Database Syst Rev 2016; 2016: CD011145.

50.

Leelakanok

. Association between polypharmacy and dementia - a systematic review and metaanalysis. Aging Ment Health 2019; 23: 932–941.

51.

Masnoon

Shakib

Kalisch-Ellett

, et al. What is polypharmacy? A systematic review of definitions. BMC Geriatr 2017; 17: 230.

52.

Harris

Taylor

Thielke

, et al. Research electronic data capture (REDCap) - a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed 2009; 42: 377–381.

53.

sdtest — Variance-comparison tests. 2014.

54.

Ostertagová

Ostertag

Kováč

. Methodology and application of the Kruskal-Wallis test. Appl Mech Materials 611: 115–120.

55.

Guerin

Soto

Brocker

, et al. Nutritional status assessment during Alzheimer's disease: results after one year (the REAL French study group). J Nutr Health Aging 2005; 9: 81–84.

56.

Buchman

Schneider

Leurgans

, et al. Physical frailty in older persons is associated with Alzheimer disease pathology. Neurology 2008; 71: 499–504.

57.

Meijers

Schols

Halfens

. Malnutrition in care home residents with dementia. J Nutr Health Aging 2014; 18: 595–600.

58.

Boulos

Salameh

Barberger-Gateau

. Malnutrition and frailty in community dwelling older adults living in a rural setting. Clin Nutr 2016; 35: 138–143.

59.

Salthouse

. Implications of within-person variability in cognitive and neuropsychological functioning for the interpretation of change. Neuropsychology 2007; 21: 401–411.

60.

Grober

Hall

Lipton

, et al. Memory impairment, executive dysfunction, and intellectual decline in preclinical Alzheimer's disease. J Int Neuropsychol Soc 2008; 14: 266–278.

61.

Tekin

Fairbanks

O'Connor

, et al. Activities of daily living in Alzheimer's disease: neuropsychiatric, cognitive, and medical illness influences. Am J Geriatr Psychiatry 2001; 9: 81–86.

62.

Skurla

Rogers

Sunderland

. Direct assessment of activities of daily living in Alzheimer's disease. A controlled study. J Am Geriatr Soc 1988; 36: 97–103.

63.

Chen

Ratcliff

Belle

, et al. Patterns of cognitive decline in presymptomatic Alzheimer disease: a prospective community study. Arch Gen Psychiatry 2001; 58: 853–858.

64.

Tosto

Gasparini

Brickman

, et al. Neuropsychological predictors of rapidly progressive Alzheimer's disease. Acta Neurol Scand 2015; 132: 417–422.

65.

Zhao

Zhou

Ding

, et al. Cognitive decline in patients with Alzheimer's disease and its related factors in a memory clinic setting, Shanghai, China. PloS One 2014; 9: e95755.

66.

Rosen

Mohs

Davis

. A new rating scale for Alzheimer's disease. Am J Psychiatry 1984; 141: 1356–1364.

67.

Jacobs

Thomas

Salmon

, et al. Development of a novel cognitive composite outcome to assess therapeutic effects of exercise in the EXERT trial for adults with MCI: the ADAS-Cog-Exec. Alzheimers Dement (N Y) 2020; 6: e12059.

68.

Saikia

Tripathi

. Executive functions, processing speed, and memory performance: untangling the age-related effects. J Psychiatry Spectr 2024; 3: 12–19.

69.

Salthouse

. The processing-speed theory of adult age differences in cognition. Psychol Rev 1996; 103: 403–428.

70.

Chiaravalloti

Christodoulou

Demaree

, et al. Differentiating simple versus complex processing speed: influence on new learning and memory performance. J Clin Exp Neuropsychol 2003; 25: 489–501.

71.

Mayes

Calhoun

. Learning, attention, writing, and processing speed in typical children and children with ADHD, autism, anxiety, depression, and oppositional-defiant disorder. Child Neuropsychol 2007; 13: 469–493.

72.

Baudouin

Clarys

Vanneste

, et al. Executive functioning and processing speed in age-related differences in memory: contribution of a coding task. Brain Cogn 2009; 71: 240–245.

73.

Berron

Olsson

Andersson

, et al. Remote and unsupervised digital memory assessments can reliably detect cognitive impairment in Alzheimer's disease. Alzheimers Dement 2024; 20: 4775–4791.

74.

Boots

Frank

Fan

, et al. Continuous associations between remote self-administered cognitive measures and imaging biomarkers of Alzheimer’s disease. J Prev Alzheimers Dis 2024; 11: 1467–1479.

75.

Butler

Watermeyer

Matterson

, et al. The development and validation of a digital biomarker for remote assessment of Alzheimer's diseases risk. Digit Health 2024; 10: 20552076241228416.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.00 MB