Exercise and Learning and Memory in Multiple Sclerosis

The proposed study involves a single-blind, Phase I randomized controlled trial (RCT) on the effects of supervised treadmill walking exercise training on learning and memory, hippocampal structure/function, and cardiorespiratory fitness outcomes in fully-ambulatory (i.e., EDSS ≤ 4.0) persons with multiple sclerosis (MS) who have impairment in learning new information. This study will compare the effects of 3 months of supervised, progressive (both intensity and duration) chronic treadmill walking exercise (designed based on pilot work [Sandroff et al., 2015; 2016a; 2016b] and American College of Sports Medicine (ACSM) guidelines) with those of an active control condition (i.e., stretching-and-toning activities) on learning and memory, hippocampal volume and resting-state functional connectivity, and cardiorespiratory fitness. Learning and memory represents the primary outcome, whereas hippocampal volume and resting-state functional connectivity and cardiorespiratory fitness represent the secondary outcomes. Such an investigation will provide the first Class I evidence for the effects of treadmill walking exercise training on those outcomes in MS.

During baseline, participants will first complete a battery of neuropsychological tests, followed by an incremental exercise test to exhaustion on a motor-driven treadmill to measure cardiorespiratory fitness, and will undergo a MRI scan. Those outcomes will be measured by assessors who are blinded to condition. Participants will undertake several neuropsychological tests addressing various aspects of learning and memory. These tests include the California Verbal Learning Test-2 [Delis et al., 2000] and Selective Reminding Task [Buschke \& Fuld, 1974] as measures of verbal learning and memory, and the Brief Visuospatial Memory Test-Revised [Benedict, 1997] and 10/36 Spatial Recall Test [Rao et al., 1984] as measures of visuospatial learning and memory. Cognitive processing speed will be measured using the Symbol Digit Modalities Test [Smith, 1982], verbal fluency will be measured using the Controlled Oral Word Association Test [Benton, 1994], and executive control (i.e., inhibitory control) will be measured using a modified flanker task [Eriksen \& Eriksen, 1974]. Cardiorespiratory fitness will be measured as peak oxygen consumption (VO2peak), using an incremental exercise test to exhaustion on a motor-driven treadmill and an open-circuit spirometry system (ParvoMedics True One 2400, Sandy, UT) for analyzing expired gases using a modified Balke protocol. This protocol is commonly used for measuring cardiorespiratory fitness in older adults [e.g. Colcombe et al., 2004] and persons with chronic stroke [e.g., Macko et al., 1997] and further is consistent with the ACSM guidelines for exercise testing of MS patients [ACSM, 2013]. Lastly, participants will undergo structural and resting-state functional connectivity magnetic resonance images of the hippocampus. The MR instrument that will be used is an FDA-approved Siemens 3 T clinical imager. Following baseline testing, participants will be randomly assigned to either the exercise training or active control condition using concealment. Participants further will blinded to condition (i.e., unaware that the treadmill walking exercise training condition represents the experimental condition and the stretching-and-toning condition represents the active control condition).

The exercise training will include 3-months of supervised, progressive light, moderate, and vigorous intensity treadmill walking exercise, and will be based on ACSM guidelines for exercise prescription [ACSM, 2013]. We selected treadmill walking exercise as the aerobic exercise modality based on pilot work that suggests that treadmill walking might exert larger beneficial effects on cognition compared with stationary cycling [Sandroff et al., 2015], and that such benefits occur for light, moderate, and vigorous intensity treadmill walking exercise [Sandroff et al., 2016]. The exercise training itself will take place 3 times per week over 3 months, led by trained exercise leaders, and will initially consist of 15 minutes of light-to-moderate intensity treadmill walking exercise and eventually progress to 40 minutes of vigorous intensity exercise by month 3. The stretching-and-toning activities will be included as a minimal exercise, active control condition. The control condition will be delivered using the same frequency and duration of the treadmill walking exercise condition. The stretching-and-toning activities will be based on a manual provided by the National Multiple Sclerosis Society, and stretching-and-toning activities will be progressively included over the course of the 3-month study period. Participants will further be asked to not undertake additional exercise (i.e., not join a gym and begin exercising) over the duration of the study. The cognitive, cardiorespiratory fitness, and MRI outcomes will be assessed again following the 3-month study period by assessors who are blinded to condition.

The data will be analyzed using mixed-factor ANOVA with time (pre- and post-) as a within-subjects factor and condition (exercise or control) as a between-subjects factor on learning and memory, hippocampal, and cardiorespiratory fitness outcomes (see Specific Aims/Hypotheses 1 and 2 in the Primary Study Objectives and Study Hypothesis sections). Effect sizes will be expressed as partial eta-squared (ηp2) and Cohen's d [Cohen, 1988]. Spearman's rho rank-order correlations (ρ) will be performed for the associations among changes in cardiorespiratory fitness, hippocampal structure/function, and learning and memory for those who underwent the treadmill walking exercise training condition (See Specific Aims/Hypotheses 3 and 4 in the Secondary Study Objectives and Study Hypothesis sections). The effect sizes from the interaction terms from the ANOVAs and correlations will serve as effect sizes for the subsequent power analyses required for a subsequent Phase II RCT.