SYNERGIC TRIAL (SYNchronizing Exercises, Remedies in Gait and Cognition) a multi-Centre randomized controlled double blind trial to improve gait and cognition in mild cognitive impairment

The target recruitment is 200 older adults aged 60 to 85 years old with MCI who meet inclusion and exclusion criteria as outlined here (London target recruitment: 40 participants, Waterloo-University of Waterloo target recruitment: 20, Waterloo-Wilfred Laurier University target recruitment: 20, Montreal target recruitment: 50, and Vancouver-University of British Columbia target recruitment: 70). Although age, sex, and education will be included as covariates in the analytical models, recruitment will not be stratified.

Participants must meet each of the following criteria for enrolment into the study:

Participants who exhibit any of the following conditions are to be excluded from the study:

Sample size calculation is based on changes in our primary outcome: global cognitive function measured using the Alzheimer Disease Assessment Scale Cognitive 13 and the plus modality (ADAS-Cog 13 and plus). Changes in the ADAS-Cog test has been used as primary outcome in pharmacological clinical trials in dementia [27] and in 2 previous studies testing the effect of exercises on cognition in MCI participants [28, 29]. A total of 200 participants, with 160 in the combined exercise (pooled from the first 4 arms in one group) and 40 participant in the BAT control group, would have 80% power to detect an effect size (mean difference divided by the standard deviation) of 0.5 at 2-sided, 5% significance level. In addition, two-group comparisons with 80 participants per group would have 88% power to detect an effect size of 0.5 at the 5% significance level. The effect size of 0.5 was selected based on previous studies showing an effect size of exercise on ADAS-Cog ranging from 0.6 to 0.3 in MCI populations. [28‐30]. The sample size estimations may be regarded as conservative, as the final data will be analyzed using analysis of covariance (ANCOVA) to adjust for baseline ADAS-Cog scores and other important patient characteristics.

Changes in cognitive function will be assessed using the ADAS-Cog in two modalities, the 13 items modality (ADAS-Cog 13) and the plus modality (ADAS-Cog plus, Table 1) [25]. Improvement in either modality is considered evidence of efficacy. The ADAS-Cog 13 is a scale that consists of 13 brief cognitive tests assessing memory, language, attention, concentration and praxis. Scores range from 0 to 84, with higher scores indicating higher severity of cognitive impairment [25]. The ADAS-Cog has been used a primary outcome measure in numerous trials with MCI and Alzheimer’s Disease (AD) [28, 29]. The ADAS-Cog plus has marked advantages over the ADAS-Cog 13 as an outcome measure in MCI populations since it incorporates items concerning executive function [25]. The following tests are adding to the ADAS-Cog 13 to comprise the plus modality: Trail-Making Test (TMT) A & B, the WAIS-R Digit Symbol Substitution Test (DSST), the Digit Span forward & backward, and Category Fluency. In brief, the TMT (A & B) is a two part test, which assesses attention, speed, and mental flexibility and has been widely used clinically for assessing deficits in attention and executive functioning [31]. Trails A, where participants connect numbers in ascending order is truncated at 3 min and Trails B, where participants connect numbers and letters in ascending and alternating order is truncated at 5 min. Psychomotor speed is assessed with the DSST [32], which evaluates the speed with which participants copy arbitrary symbols associated with corresponding digits, by referring to a number-symbol key at the top of the page. The Digit Span test is an auditory attention task, where participants are asked to recall a series of numbers forward and backward. For category fluency, a measure of speed and flexibility of verbal thought, participants are asked to name as many items as possible in a specified category (vegetables & animals); unique responses during the first minute in each category are counted.

We expect that our interventions, over 20 weeks, will show improvement and/or less decline in cognition as measured by the ADAS-Cog (13 and plus modalities) at month 6 and at month 12. Significant changes in either time points will be considered preliminary evidence of efficacy. Additionally, reducing the proportion of participants, expressed as percentages per allocated group, with abnormal ADAS-Cog scores (dichotomous variable, cut-off score to be determined with ADAS-Cog literature) after intervention will be considered evidence of efficacy. Furthermore, we expect a significant statistical improvement in the transformed score of ADAS-Cog plus using an algorithm validated by Crane et al. [33].

Secondary outcomes include neuropsychological assessments, gait and mobility outcomes (including incidence of fall), neuroimaging, and blood biomarkers.

Secondary cognitive outcomes will include MoCA test for global cognition, the recall list from the ADAS-Cog to evaluate verbal semantic memory, TMT A and B, DSST, Digit Span Test (Forward and Backwards), Boston Naming Test, Verbal fluency (animals and vegetables), and Colour Word Interference Test (Table 1). We expect improvement in participants of the active intervention in these cognitive tests, with the larger improvement in the multimodal intervention.

Gait velocity will be assessed as the time taken to walk 6 m using an electronic walkway system (ProtoKinetic® and/or GAITRite® Systems, Inc.). Gait variability of spatial and temporal gait variables (stride time, stride length, double support time and step width) will be calculated using the coefficient of variation (CV = (standard deviation / mean) × 100). The CV is a standardized measure of variability allowing comparison of gait variables measured in different units, having different means and range of values. Gait walks will be performed 3 times under single-task conditions, and one time under each of the three dual-task conditions (described below), and one time as fast gait. Dual-tasking assessments will permit calculation of dual-task cost for all gait variables of interest [34, 35]. Specifically, we expect participants in the combined exercise intervention to increase their gait velocity, decrease their gait variability and reduce their dual-task cost for the gait variables of interest.

Gait performance will be recorded using electronic walkway systems which automatically determine spatiotemporal gait parameters from imbedded sensors activated by foot pressure [36]. The gait mat will be located in a well-lit room with start and end points marked on the floor 1 m from either end of the mat. Participants will perform three main tasks: 1) preferred walking speed, 2) dual-task walking (counting backwards by 1’s, subtracting 7’s, and naming animals out loud while walking) and 3) fast walking. In all walks, participants will start 1 m before the beginning of the 6-m walkway and continue to travel 1 m past the end of the walkway. This procedure is in place to ensure steady-state walking and to minimize any effects of acceleration and de-acceleration during the course of the walk [37, 38]. The dual-task conditions selected are based on previous research which demonstrated that counting backwards requires both working memory and attention [39] and naming animals is related to verbal fluency, which relies on semantic memory [39, 40]. The evaluator will record any counting errors during walking so that it can be compared with the same mental tasks while seated. The seated assessments will be timed at 10 s and will be performed in the beginning of all cognitive assessments (before ADAS-Cog) to prevent practice effects in dual-task gait performance. Reliability has been previously established for this protocol in people with MCI [41] and an instructive video can be found at the “www.​gaitandbrain.​com/​resources” as the “Guidelines for Gait Assessments in CCNA”.

A fall is defined as ‘unintentionally coming to rest on the ground, floor, or other lower level and not due to a seizure, syncope, or an acute stroke’ [42]. Events caused by overwhelming environmental hazards (e.g., being struck by a moving object) are not considered a fall. Recurrent falls are defined as ‘two or more events in a 12-month period’. Falls will be recorded throughout the 12-month trial, participants will be provided with a falls calendars, on which they will record any falls that have occurred, and they will be asked to bring them monthly to the training sessions to review with a research staff member. After the completion of the intervention, participants will be contacted at month 9 and 12 to report incidental falls. We expect participants in the combined intervention to reduce the number of falls compared to the sham intervention.

To further evaluate mobility, participants will be performing the Short Physical Performance Battery (SPPB), and the 6 min walk test (6MWT). We expect participants in the active intervention to present the larger improvements.

Upon completion of the screening and baseline assessments, participants will be randomized and allocated to one of the 5 study arm. The randomization sequence of the participants will be generated centrally by the research pharmacist of the sponsor site using a central, web-based randomization service (www.​randomizer.​org) and will be specific for each site. The research pharmacy will assign the investigational product (vitamin D/matching placebo, as “kits”) in compliance with the randomization list(s) and allocate a sequential randomization number to each participant. A block randomization by 5 will be applied to ensure an appropriate balance of the characteristics of participants between each arm. Permuted blocks will be employed to ensure balance over time. After the baseline assessment, research personnel not involved in measurement or intervention will access the randomization list(s) to determine the arm allocation and institute the corresponding study interventions.

In order to minimize a source of bias, this is a double-blinded study. Research personnel performing the outcome assessments at 6 and 12 month follow-ups will be blinded to group allocation. Participants will be blinded to the “active” intervention and study hypotheses. If it was medically necessary to un-blind a participant, the central research pharmacy will be contacted by the principal investigator to obtain the code.

The combined exercise intervention will be held in appropriate fitness facilities and take place between Monday and Friday, ensuring that it is not on three consecutive days. Staff trained and certified in exercise training will supervise all sessions. Difficulty of aerobic and resistance exercise will be tailored to their individual functioning level, with constant monitoring by the trainers.

Participants will start with a 10 min warm up, which includes: marching in place with arm swings, bum kicks, dynamic hamstring stretching, hula hoop circles, shoulder circles, arm reaches, torso twists, ankle circles, dynamic calf stretching, side stepping with wrist circles, split-step knee bends, and quarter squats. The RT portion will occur next and includes 5 exercises: leg press & leg flexion (lower body), chest press, lat pull, and seated row (upper body). These exercises are performed alternating between lower and upper body exercises, with the number of repetitions, sets, and rest being modified every 4 weeks to increase muscular strength. RT will progress from loads and volumes appropriate for endurance (first 8 weeks) to those appropriate for maximal strength gains (last 12 weeks). Training volume, intensity, and rest are standardized and described in Additional file 1: Table S1. Participants should reach exhaustion at the last prescribed repetition of the last set. Training prescription for all exercises follows American College of Sport Medicine guidelines for strength development in older adults.

Following the RT portion of each session, participants will complete 2 sets of 10 min of aerobic exercise. The mode of exercise can vary and may include walking, a variety of ergometers (treadmill, elliptical machines, cycling ergometers, rowing machines, etc.), and other forms of free aerobic exercises, as long as the cognitive load is minimized. Intensity will be monitored using a Borg scale (0–10) with a target of 5–6 for weeks 1–8, 6–7 for weeks 9–16, and 7–8 for the last 4 weeks (Additional file 2: Table S2).

Each session will end with a ten-minute cool down, which will consist of the following stretches (each held for 20–30 s); quadriceps stretch, hamstring stretch, calf stretch, 2 hip stretches, static torso rotation, seated side bend, back and shoulder stretch, chest stretch, triceps stretch, neck stretch.

Participants assigned to the BAT control exercise condition will take part in balance and toning exercises in groups of up to eight participants, supervised by a trainer [43‐48]. The exercises will be devoted to improving muscle tone and flexibility, without improving strength, and cardiorespiratory capacity. Resistant load and number of repetitions will not progress across exercise sessions, unless participants were unable to complete required repetitions at the beginning of the intervention. The session will start with the same 10-min warm-up completed in the combined AE and RT group. This will be followed by 40 min of a variety of balance and toning exercises that will target the entire body. The sessions will include functional training (wall squats, standing calf raises, standing leg abduction/adduction, standing ball walk up & down, seated back row with resistance tube, shoulder retractions, wall push-ups, toe walking, heel walking, quarter squats, gluteus kickback holds, standing ball twists, shoulder circuits, yoga ball chest press, yoga ball shoulder abduction, yoga ball leg adduction squeeze), balance training (standing leg circumduction, tandem stance, single leg stance, tandem stance ball relay pass, partner ball pass in tandem stance, tandem forward & backward walking, toe taps on bench), agility training (4 step zig zag in place, 4 step zig zag walking, line and cone drills), and core training (core contractions seated on exercise ball, exercise ball seated ball pass, seated exercise ball marching, modified & full bug on floor). Exercises will be cycled in and out every 3 weeks to maintain variety and decrease likelihood of progression. The session will end with the same 10 min cool down stretches as for the exercise group.

The cognitive training intervention will involve tablet based (iPad®, held on a stand on a table) dual-task training that requires participants to maintain and prepare for many response alternatives (working memory) and to share attention between two concurrent tasks (divided attention). Difficulty of cognitive training is tailored to their individual functioning level. The training uses a custom-written program developed for neuro-rehabilitation and has been used in previous research trials for cognitive [39, 40] and mobility outcomes [12].

Cognitive training will take place in groups of up to eight participants before the exercise training session. Cognitive training will be 30 min, at maximum, and each participant will complete the cognitive training at an individual desk space, in a quiet room, wearing headphones. Each session, participants will perform a two different visuo-motor tasks, each with their respective sets of visual stimuli (e.g., letters, numbers, animals, vehicles, fruits, celestial bodies) and respective hand-button correspondences (i.e., buttons that are to be tapped with the thumbs on either the right or the left side of the tablet). Participants are instructed to perform as fast as possible, while maintaining accuracy. Tasks will be performed both separately and concurrently so that task-set cost and dual-task cost can be isolated. At each session, task combination for the sets of stimuli will change (from a total 18 combinations). Training will also include online feedback as well as histogram of daily performances to encourage improvement.

The cognitive training control group sessions will last a maximum of 30 min to align with the same time frame as the cognitive training group. Participants will alternate between 2 different tasks (touristic searching using internet and video watching) completed using the same tablet (iPad®) as in the cognitive training group. Sessions will be held in groups of up to eight participants, and in the first session they will receive a short introductory lesson on the use of the tablet and how to navigate the internet. For the touristic searching using internet, participants will be required to find 3 hotels, 3 touristic places, and 3 restaurants of their own preference in a city assigned by the instructor (a new city will be selected each session). They will also need to include the respective addresses of those places on their log sheet. For the video watching task, participants will watch a National Geographic video on YouTube selected by the instructor with a different video selected each session. They will watch the video for 20 min and during the remaining 10 min they will answer the following questions on their log sheet; 1) what is the video about; 2) what is the most important information in your opinion; 3) create a question based on the video and answer your own question. Despite completion of the tasks, participants will be stopped at 30 min.

Participants will receive vitamin D supplementation (1 tablet of 10,000 IU of vitamin D₃) three times per week in order to reach a weekly cumulative dose of 30,000 IU per week (equivalent of 4258 IU daily). The vitamin D capsules will be provided by the research coordinator in vials at the first session and every 4 weeks for a total of five vials during the training period. Vials will be returned by the participants to the research coordinator at the end of each 4 week block. Rationale and bio-safety of the dose: The dose of 10,000 IU/day is currently approved by Health Canada as a supplementation for older adults. Heaney et al. have administered doses of 30,000 IU/day vitamin D₃ to adult men for 5 months, with no significant changes in serum calcium concentrations or adverse reactions [49‐51]. A comprehensive review of toxic effects of vitamin D found that the lowest level at which an adverse effect was observed was a serum calcidiol concentration of 200 nmol/l, corresponding to a daily intake of 40,000 IU. Therefore, weekly doses of 30,000 IU have a 9-fold weekly margin of safety of the established safe dose.

Participants allocated to placebo vitamin D will receive a placebo capsule that will perfectly match the vitamin D tablets. They will take one tablet three times per week, to match what the active vitamin D participants will be taking. The vials will be dispensed on a 4 week basis in the same manner as the vitamin D tablets.

Descriptive statistics for demographic and baseline characteristics will be provided with means and standard deviations, or medians and the interquartile range where appropriate, for continuous characteristics, and frequencies and percentages for categorical variables. Analysis will be conducted as intention-to-treat (ITT) and as per-protocol analysis (PPA). Observation of a statistically significant difference in the primary outcome at post-intervention time point will be considered preliminary evidence of efficacy. Contrasts from linear mixed models containing the variable for intervention arm, the categorical variable for time points, and their interaction term will be used for comparing continuous outcomes such as ADAS-Cog scores, adjusting for baseline outcome values and patient characteristics including age, sex, and education. This approach can readily handle three time points (baseline, 6 months and 12 months) and thus may be regarded as a generalization of the ANCOVA. Similar analyses for binary outcomes such as reduction of prevalence rate of abnormal ADAS-Cog plus score will be performed with generalized linear mixed effect models. Specific contrasts from the mixed-effects models will be constructed to test the following hypotheses. The first hypothesis (exercise intervention) will be evaluated by collapsing arms 1–4 (exercise intervention, n = 160) and comparing them with arm 5 (no exercise intervention, n = 40). For the second hypothesis (cognitive intervention), arms 1–2 (cognitive intervention, n = 80) and arms 3–4 (no cognitive intervention, n = 80) will be collapsed and compared. For the third hypothesis (vitamin D intervention), we will collapse arms 1 and 3 (vitamin D intervention, n = 80) and compare them with collapsed arms 2 and 4 (no vitamin D, n = 80). For the fourth hypothesis (synergic effect), arm 1 (all active interventions, n = 40) will be compared with arm 5 (all control interventions, n = 40).

For the secondary outcomes, 6 month changes in gait performance and secondary cognitive outcomes of interest will be performed using one way analyses of variance. For group comparisons that are statistically significant, pairwise comparisons will be made using Tukey’s multiple comparisons tests. Secondary analyses will be also performed according to an intention-to-treat principle and per-protocol analysis. An economic analysis will also be performed.

Preliminary analysis will be performed after finishing recruitment to ascertain descriptive characteristics at baseline assessment. Interim efficacy analyses will be performed when recruitment is reaching 50% of target sample and final efficacy analysis will be performed at the end of the trial since no safety issues are anticipated in this study.

Adverse events will be recorded for subjects starting at the time of signing the Informed Consent until their discontinuation of the study. All adverse events will be recorded on an ongoing paper log, which will be reviewed by each site leader and forwarded to the team leader at London site, who will determine both the intensity of the event and the relationship of the event to study procedures, and monitored until resolved. Based on the outcome it will be up to the team leader to medically determine if it is justified for the participant to continue in the study or terminate their participation. A serious adverse event is defined as any incident that is unexpected, and related or possibly related to participation in the research study.

Higher doses of vitamin D may lead to hypervitaminosis D manifested by hypercalcemia and its sequelae. Treatment of acute or chronic intoxication includes withdrawal of vitamin D3 and any calcium supplements, maintenance of a low calcium diet and if needed corticosteroids or calciuric diuretics, such as furosemide and ethacrymic acid to decrease serum calcium concentrations. All participants will be monitored by site physicians and trained staff, and should any adverse events arise, participants will have access to a phone number and e-mail address provided on the copy of their signed consent form. They will be informed to use these contacts to notify the Principal Investigator about concerns that may be associated with treatment received in each site.