Optimizing Detection and Prediction of Cognitive Function in Multiple Sclerosis With Ambulatory Cognitive Tests: Protocol for the Longitudinal Observational CogDetect-MS Study

Abstract

Background: Cognitive dysfunction is a common problem in multiple sclerosis (MS). Progress toward understanding and treating cognitive dysfunction is thwarted by the limitations of traditional cognitive tests, which demonstrate poor sensitivity and ecological validity. Ambulatory methods of assessing cognitive function in the lived environment may improve the detection of subtle changes in cognitive function and the identification of predictors of cognitive changes and downstream effects of cognitive change on other functional domains.

Objective: This paper describes the study design and protocol for the Optimizing Detection and Prediction of Cognitive Function in Multiple Sclerosis (CogDetect-MS) study, a 2-year longitudinal observational study designed to examine short- and long-term changes in cognition, predictors of cognitive change, and effects of cognitive change on social and physical function in MS.

Methods: Participants-ambulatory adults with medically documented MS-are assessed over the course of 2 years on an annual basis (3 assessments: T1, T2, and T3). A comprehensive survey battery, in-laboratory cognitive and physical performance tests, and 14 days of ambulatory data collection are completed at each annual assessment. The 14-day ambulatory data collection includes continuous wrist-worn accelerometry (to measure daytime activity and sleep); ecological momentary assessments (real-time self-report) of somatic symptoms, mood, and contextual factors; and 2 brief, validated cognitive tests, administered by smartphone app 4 times per day. Our aim was to recruit 250 participants. To ensure standard test protocol administration, all examiners passed a rigorous examiner certification process. Planned analyses include (1) nonparametric 2-tailed t tests to compare in-person to ambulatory cognitive test scores; (2) mixed effects models to examine cognitive changes over time; (3) mixed effects multilevel models to evaluate whether ambulatory measures of physical activity, sleep, fatigue, pain, mood, and stress predict changes in objective or subjective measures of cognitive functioning; and (4) mixed effects multilevel models to examine whether ambulatory measures of cognitive functioning predict social and physical functioning over short (within-day) and long (over years) time frames.

Results: The study was funded in August 2021 and approved by the University of Michigan Medical Institutional Review Board on January 27, 2022. A total of 274 adults with MS (first participant enrolled on May 12, 2022) have been recruited and provided T1 data. Follow-up data collection will continue through March 2026.

Conclusions: Results from the CogDetect-MS study will shed new light on the temporal dynamics of cognitive function, somatic and mood symptoms, sleep, physical activity, and physical and social function. These insights have the potential to improve our understanding of changes in cognitive function in MS and enable us to generate new interventions to maintain or improve cognitive function in those with MS.

Trial registration: ClinicalTrials.gov NCT05252195; https://clinicaltrials.gov/study/NCT05252195.

International registered report identifier (irrid): DERR1-10.2196/59876.

Keywords: ambulatory assessment; cognitive assessment; cognitive function; ecological momentary assessment; longitudinal data collection; mobile phone; multiple sclerosis; neuropsychology; observational study; physical activity; physical function; smartphone app; social function; wrist-worn accelerometry.

Figure 1

(A) The Wear-IT app landing page, (B) morning survey landing page, (C) self-reported wake-up time, (D) an item from the perceived cognitive function scale, (E) an item from the depressed mood scale, and (F) the survey of substance consumption.

Figure 2

The ActiGraph wGT3X-BT triaxial accelerometer.

Figure 3

(A) Symbol Search landing page, (B and C) test instructions, and examples of (D) a “nonlure trial” (where neither symbol in the incorrect pair on the bottom appears in the pairs above) and (E) a “lure trial” (where one of the symbols in the incorrect pair on the bottom appears in a pair above).

Figure 4

(A) Dot Memory Test landing page, (B) instructions, (C) E’s and F’s distraction phase, and (D) response page.