Feasibility and validation of a web-based platform for the self-administered patient collection of demographics, health status, anxiety, depression, and cognition in community dwelling elderly

The coronavirus disease pandemic has brought a new urgency for the development and deployment of web-based applications which complement, and offer alternatives to, traditional one-on-one consultations and pencil-and-paper (PaP) based assessments that currently dominate clinical research. We have recently developed a web-based application that can be used for the self-administered collection of patient demographics, self-rated health, depression and anxiety, and cognition as part of a single platform. In this study we report the findings from a study with 155 cognitively healthy older adults who received established PaP versions, as well as our novel computerized measures of self-rated health, depression and anxiety, and cognition. Moderate to high correlations were observed between PaP and web- based measures of self-rated health (r = 0.77), depression and anxiety (r = 0.72), and preclinical Alzheimer’s disease cognitive composite (PACC) (r = .61). Test-retest correlations were variable with high correlations for a measure of processing speed and a measure of delayed episodic memory. Taken together, these data support the feasibility and validity of utilization of this novel web-based platform as a new alternative for collecting patient demographics and the assessment of self-rated health, depression and anxiety, and cognition in the elderly.

QOL-38. USE OF COMPUTERIZED NEUROPSYCHOLOGICAL MEASURES TO ASSESS COGNITIVE MORBIDITY IN SURVIVORS OF CHILDHOOD BRAIN TUMORS

Treatment of central nervous system (CNS) tumors in pediatric populations is associated with significant cognitive morbidity. Documentation of neuropsychological deficits is vital to treatment and educational planning. We investigated the feasibility and utility of a computerized neuropsychological measure (NIH Toolbox Cognitive Battery) in differentiating individuals who received tumor treatment from healthy controls. Participants included pediatric CNS tumor survivors (N = 85; Mean Age = 13.47; SD = 4.76) at least 1-year post-completion of treatment and healthy sibling controls (N = 20; Mean Age = 10.2; SD = 3.21) who completed the NIH Toolbox. Ninety-eight percent of the participants enrolled completed the computerized tasks. The overall logistical regression model, with NIH Toolbox tests as predictors, was statistically significant [χ2 (7, N = 105) = 26.176; p < .001] and improved correct group classification from 81% to 82.9%. Picture Sequencing (β = -0.059; Wald = 6.942; p = .008) and Flanker (β = -0.083; Wald = 7.473; p = .006) were both statistically significant and negatively predictive of membership in the treatment group. For each 1 unit increase in standard score on measures of working memory and inhibition, odds of membership in the treatment group decreased by 6.2% and 8.7%, respectively. Consistent with the literature, worse performance on computerized measures of cognitive functioning mediated by executive functioning was correlated with a history of brain tumor treatment. Further investigation will focus on comparing computerized neuropsychological tools to traditional comprehensive neuropsychological evaluations and clarifying the trajectory of these deficits across recovery.

QOL-37. USE OF COMPUTERIZED NEUROPSYCHOLOGICAL MEASURES TO ASSESS COGNITIVE MORBIDITY IN CHILDREN UNDERGOING ACTIVE RADIATION THERAPY

Cognitive late effects of brain tumors and related treatments are well-established; however, limited information regarding changes in cognition during radiation therapy (RT) is available. Recent advances in computerized neuropsychological assessments for monitoring of acute and late treatment effects have been developed, though the feasibility of using these tools in a population undergoing active RT has limited empirical evidence. This study investigated performance of pediatric patients with brain tumors actively undergoing RT on the NIH Toolbox (N = 10; M age = 11.29 ± 3.35 years; 86% Caucasian; 86% female). Given significant individual variability, one-sample proportion tests were calculated to assess whether the proportion of patients with performances >1 standard deviation below the mean significantly differed from normative expectations. Of the 12 participants that were enrolled in the study, 10 completed the NIH Toolbox during active RT. Compared to normative expectations, a greater proportion of participants undergoing active RT exhibited deficits on measures of processing speed, working memory, and response inhibition (p=<.01). Differences between participants and normative expectations were not seen on measures of visual memory and vocabulary (p=>.05). Seventy-seven percent of recruited participants completed computerized assessment during active RT, suggesting reasonable feasibility within the small cohort recruited. Consistent with the literature regarding late effects of RT, performance on computerized measures of cognitive functioning mediated by processing speed and aspects of executive functioning were lower for patients undergoing active RT. Further investigation will focus on clarifying the trajectory of deficits across treatment course and comparing computerized measures to traditional neuropsychological measures.

A-020 Advancing Reliable Measurement in Alzheimer’s Disease and Cognitive Aging (ARMADA): Extension of the NIH Toolbox in Healthy Cognitive Aging 86 years and Beyond

Background Early detection of cognitive decline in older individuals can be accomplished with brief, computerized measures in diverse settings. The goal of the ARMADA study is to validate the NIH Toolbox for Assessment of Neurological and Behavioral Function (NIHTB) in the neurological aging spectrum from cognitively normal to dementia, and to extend the age range to allow for longitudinal assessment of the oldest individuals. Here, we present baseline data from general population individuals over 85 years of age without cognitive impairment compared with those 65 to 85. Method ARMADA is a multi-site study recruiting participants from established research cohorts. Participants included cognitively average-for-age individuals between the ages of 65–85 (n = 82) and another group 86 and older (n = 105). Baseline characteristics and preliminary data are presented for the Cognition, Emotion, Motor and Sensation modules of the NIHTB. Results Analysis of preliminary baseline data revealed that mean uncorrected standard scores of most cognitive, motor, and sensation measures differed among the groups in the expected direction, such that individuals 65–85 had better performance than individuals 86 and older. Conversely, group scores were similar on most emotion measures. Conclusions This analysis of preliminary baseline data from ARMADA provides initial evidence for the utility of the NIHTB in individuals over age 85 without cognitive impairment, who, as a group, scored lower than those from age 65–85. These results demonstrate the importance of establishing cohort-specific norms relevant to the oldest age group. Longitudinal follow-ups over two years will track cognitive, motor, emotion, and sensory functioning in this group.

Testing the Hybrid Battery Approach to Evaluating Sports-Related Concussion in the National Hockey League: A Factor Analytic Study

Purpose Traditional “paper and pencil” neuropsychological tests and computerized test batteries have been employed in the assessment of sports-related concussion (SRC). Each type of test has inherent strengths and weaknesses. The NHL has constructed a “hybrid” battery of tests that uses both traditional and computerized measures in the evaluation of its players suspected of SRC. The purpose of this paper is to use factor analytic techniques to examine the relationships among the measures that comprise this battery. Methods Post-injury neuropsychological test data using the hybrid battery were obtained from 343 NHL players following their first concussion while playing in the league. ImPACT was used as the computer test battery. The traditional battery included: Hopkins Verbal Learning Test, Brief Visuospatial Memory Test, Color Trails, PSU Cancellation, Symbol Digits Modalities Test (including incidental memory), and Verbal Fluency. Results Five factors were extracted explaining 64.55% of the variance. The factor labels and the measures principally loading on each factor were as follows: Factor 1, Verbal Learning and Memory (HVLT Total and Delayed Recall); Factor 2, Processing Speed (Color Trails A/B, PSU Cancellation, and SDMT-Total); Factor 3, Visual Memory (BVMT- Total and Delayed); Factor 4, Cued Memory (ImPACT-Visual and Verbal Memory Composites); and Factor 5, Reaction Time (ImPACT Reaction Time and Visual Motor Speed composites). Conclusion These data underscore the unique contributions of traditional and ImPACT neuropsychological measures to the evaluation SRC in a sample of professional hockey players. It appears that both approaches measure different aspects of cognitive functioning. The next logical step is to use these data in evaluating the diagnostic utility of these measures as part of a combined battery.

Associations of Lipophilic Micronutrients with Physical and Cognitive Fitness in Persons with Mild Cognitive Impairment

Age-associated cognitive impairment in general and dementia in particular are a global concern. Preventive lifestyle strategies are highly used but there is a lack of information on the reciprocal relationships between nutrition biomarkers and measures of both cognitive and physical performance. To fill this gap of knowledge, the relationship between plasma levels of the robust nutrition- and antioxidant defense-related biomarkers carotenoid and tocopherols and both indicators of cognitive and physical performance was investigated in a group of persons with mild cognitive impairment participating in the NeuroExercise Study at the German Sport University in Cologne, Germany. In 56 participants with full dataset, significant correlations independently of fruit and vegetable intake were found between plasma levels of β-cryptoxanthin and Timed Up&Go test (p < 0.05), γ-tocopherol and number of daily steps (p < 0.01), as well as between four out of six measured carotenoids—lutein; zeaxanthin; β-cryptoxanthin and β-carotene—and the computerized CogState International Shopping List subtest (p < 0.01). In light of the increasing attention towards the nutritional cognitive neuroscience of carotenoids, computerized measures of cognitive performance might be further implemented in future studies investigating the effects of lifestyle interventions against cognitive and physical impairment.