Design
The ACT Trial was a Phase II and Stage II, single-blinded randomized controlled trial with three geographically diverse sites (Midwest, Southwest, and East Coast) and used a 2 × 2 factorial design11. Participants were randomized to one of four arms with equal allocation for six months (ACT, cycling only, cognitive training only, or attention control within site) and then followed for 12 months. Randomization was performed by a study statistician and was stratified by site (University of Minnesota [UMN] and University of Rochester [UR]) and age (< 75 and ≥ 75 years), using random permuted blocks of 4 and 8 participants. All participants signed an informed consent form to participate in the study11. All outcome assessors and study investigators were blinded. Data collectors were not involved in the delivery of the intervention and did not participate in trainings for intervention implementation, while study participants were instructed not to reveal details of their intervention during outcome visits. The study was approved by the Institutional Review Board (IRB) at each site: IRB# STUDY00001135 on 8/23/2017 at UMN, STUDY00001484 on 2/19/2018 at UR, and WCG IRB #1305160 & IRB#STUDY00013092 on 04/21/2021 at, and Arizona State University [ASU]). The ACT Trial was conducted in accordance with the Declaration of Helsinki. This trial was registered at Clinicaltrials.gov (NCT03313895) on 10/18/2017; details of the trial protocol were previously published11.
Setting
All screenings and data collections (4/1/2018-7/17/2024) were performed on each university campus. Prior to the pandemic in March 2020, all intervention sessions were supervised in-person at YMCA gyms or senior community centers. During the pandemic (March 2020-November 2020), interventions were delivered via video, using a synchronous, audiovisual telerehabilitation format (Zoom®). Post-pandemic, intervention sessions continued in-person, over telerehabilitation, or in a hybrid format.
Sample
Older adults 65 + years old who were English-speaking, community-dwelling, and did not have dementia were eligible for the study if they met criteria for a clinical diagnosis of aMCI according to 2011 diagnostic criteria. This criteria includes the following: (1) Montreal Cognitive Assessment (MoCA) scores 18–26, (2) memory deficits on the Rey Auditory Verbal Learning Test (RAVLT) criteria of at least 1 standard deviation below age- and/or education corrected population norms, (3) preserved activities of daily living (Activities of Daily Living-Prevention Instrument [ADL-PI] score < 30), and (4) no dementia11. Individuals were excluded who had psychiatric disorders (e.g., major depressive disorder), alcohol or chemical dependency, or neurologic disorders (e.g., Parkinson’s disease) that were likely causative of the aMCI. Additional exclusionary criteria included exercise contraindications and abnormal findings from the magnetic resonance imaging (MRI)11.
Recruitment
The ACT Trial used a comprehensive multi-channel recruitment approach, including presentations to community partners (e.g., YMCAs), fliers and brochures, websites, social media, advertisement, exhibits, registries, listservs, referrals, and other online tools. Fliers and brochures were placed in community centers, clinics, libraries, and fitness and rehabilitation facilities. Online tools included the use of eLetters that were sent to persons with a clinical aMCI diagnosis who had opted in to research contact within their electronic health record.
Implementation overview
Based on the team processes of MOBILIZE, we developed a study protocol and its accompanying standard operating procedure (SOP) for screening, intervention delivery, safety, and data collection for use by all sites. Staff were trained using these materials through the initial onboard training, semi-annual training, and quarterly booster training. In addition, weekly interventionist meetings and monthly cross-site team meetings focused on impromptu training to ensure compliance. For this paper, team processes that function in multiple outcomes of MOBILIZE are discussed according to the logical sequence of the study procedure for simplicity. For example, two team processes for safety that started during screening (obtain medical clearance and assess safety) were discussed under “Screening.”
Screening (minimize time in screening phase)
A thorough but efficient and timely screening process is important to attenuate change in interest regarding study participation. Five team processes for screening were implemented (Fig. 1) to balance a thorough but efficient process.
Know the person (screening 1–4 [S1-S4])
Participants were screened for eligibility and exercise safety using a four-step procedure: a phone interview (screening 1 [S1]), an in-person interview with consenting (screening 2 [S2]), primary-care provider clearance, and cardiopulmonary exercise test (CPET) (screening 3 [S3]) and MRI (screening 4 [S4]). During the 20–30 min S1, participants provided answers to questions directed at health history related to exercise and MRI risk and history of cognitive impairment.
Assess and meet capacity (S2-S3)
During the S2 phase (1.5–2 h), informed consent was obtained. Additionally, study staff administered instruments to (a) assess clinical diagnosis of aMCI, (b) conduct a focused physical assessment, and (c) complete MRI safety checklist. Staff interacted with potential participants based on a capacity assessment in the previous step(s) and identified capacity barriers to address. Staff communicated at a person’s communication level and arranged transportation for study activities as needed. During the S2 phase, study staff ensured participants understood the time demands of the study, addressed any questions or concerns the participant had regarding the study and aMCI, and inquired about previous experiences with research and lifestyle interventions. Participants gave verbal consent at the beginning of each encounter and formal written consent at 12 and 18 months.
Build rapport
Rapport was built from the first encounter and solidified over the screening steps as staff got to know the potential participants over time. Flexibility was given to accommodate participant’s preference for intervention sessions and locations, and staff worked with participants’ schedules and needs for vacations and time off.
Obtain medical clearance
Following the S2 and prior to scheduling the CPET, clearance for study participation from each participants’ primary-care provider was sought. Faxes were sent, which contained a study synopsis and a brief checklist regarding ability to participate in a supervised exercise program and a 3T MRI. Follow-up calls were made to ensure receipt of faxes and on a weekly basis for status update as needed.
Assess safety
During the S2 and S3 phases, study staff administered a comprehensive health-history questionnaire that focused on diagnoses and symptoms pertaining to cardiovascular, renal, metabolic, and pulmonary health. Summaries from the questionnaire were subsequently generated for the exercise physiologist to evaluate for potential contraindications to exercise prior to scheduling the CPET. The medically supervised S3 CPETs with 12-lead ECG were conducted as previously discussed23, with the first goal of excluding participants with latent cardiac ischemia, serious arrhythmia, or other exercise contraindications. Stage measurements of heart rate and blood pressure allowed for comprehensive monitoring of hemodynamic response to exercise stimuli. Results from the CPET also provided information (symptoms to watch for) for the study interventionists when later creating the individualized exercise prescriptions. For instance, the study staff informed the study exercise physiologist orthopedic concerns in S2 and/or joint discomfort in S3 to receive instructions on how to modify the exercise prescription and delivery to minimize adverse symptoms during intervention sessions. The CPET allowed for study investigators ability to analyze baseline aerobic fitness and limitations to cycling performance, providing information about the participants’ abilities to successfully participate in the intervention phase. Screening MRIs (S4) were implemented to search for any clinically significant brain abnormalities that would interfere with a research definition of aMCI, including normal pressure hydrocephalus, brain tumor, subdural hematoma, significant post-traumatic encephalomalacia, or one or more large hemispheric infarctions.
Intervention adherence (maximize attendance and session dose adherence)
Following randomization, participants completed 1 of 4 assigned activities (ACT, cycling only, speed of processing [SOP] training only, or attention control). Each participant was encouraged to attend three supervised sessions weekly for 6 months (72 total sessions). Because poor adherence to intervention can affect efficacy-related outcomes, team processes for exercise delivery (Fig. 1) were implemented and operationalized as follows.
Individualized intervention dose
Briefly, cycling only was performed on recumbent cycle ergometers at moderate intensity (heart rate reserve [HRR] 50–60% or rating of perceived exertion [RPE] 11–12) with a session length of 30 min. Exercise was gradually progressed to achieve moderate-vigorous intensity (HRR 65–75% or RPE 13–15) for 50 min by week 8. Session lengths did not include the 5-minute warm-up and 5-minute cool-down. SOP training was conducted with computerized BrainHQ, which adjusts based on a participant’s performance. The session duration of SOP training started at 50 min per session and was progressively reduced to 30 min by week 8. ACT sessions were 80 min in duration, excluding the 5-minute warm-up and 5-minute cool-down. Cycling was completed first, followed by the SOP activity, with intensity and duration progressions mimicking the cycling and SOP training groups. Attention control followed a similar format to ACT but included low intensity (HRR < 30% or RPE ≤ 9) range of motion and stretching exercise followed by mental leisure activities (e.g., Sudoku).
Monitor responses
Interventionists monitored heart rates, RPE talk ability, blood pressure, and signs and symptoms continuously during each session. De-identified session reports and summarized session duration, HRR and RPE achieved (the latter two for ACT, cycling, and stretching control groups only) were reviewed weekly by the study exercise physiologist during weekly intervention meetings.
Adjust dose (as needed)
When participants were unable to achieve minimally prescribed heart rate or RPEs, discussions ensued to promote greater dose adherence. Factors that may have been limiting these adherence indicators could have included beta-blocker use (heart rate) or inadequate understanding of the RPE scale. In the former, advice included how to use the RPE and talk test as the primary indicators of intensity. For the latter, advice included using markers 11, 13, and 15 only (ACT or cycling) so that the participant could gain better understanding of light, moderate, and hard on the scale for simplicity and training purposes. For individuals who did not have medication or subjective reasons for reduced dose (intensity) adherence, other strategies were implemented on a case-by-case basis.
Set up reminders
Day and time of the next intervention session was confirmed prior to adjourning the current session. Staff provided participants a reminder text the night before for morning sessions or in the morning for afternoon sessions.
Make up for absences
When a session was cancelled, interventionists attempted to schedule the make-up session within the same week if possible or in the subsequent week. Only one make up session could be scheduled per week in an attempt to not exceed four sessions/week.
Review progress
Intervention protocol adherence was monitored during weekly Zoom meetings held by the study exercise physiologist and each of the study interventionists from the three sites. Meetings focused on review of attendance, adherence to intervention prescription, tolerance to intervention dose and adverse symptoms. Attendance to interventions were reviewed for each participant and individual factors that were limiting attendance were discussed. Strategies for overcoming participant barriers to attending sessions were discussed for implementation thereafter.
Safety (minimize interventional/procedural-related AEs)
Team processes for preventing AEs to increase minimize occurrence of interventional/procedural-related AEs (Fig. 1) were operationalized as follows:
Educate about signs and symptoms
Prior to the CPET and prior to the first intervention session (ACT, cycling, and attention control groups), participants were educated on adverse symptoms of exercise pertaining to cardiovascular, pulmonary, and musculoskeletal systems. Participants were educated through the reading of a brief standardized script and were offered time for questions. In addition, prior to each intervention session, study interventionist interviewed participants regarding residual symptoms from the previous session, evaluating tolerability to the intervention and making any necessary adjustments.
Assess and document adverse events
In the event of an AE, protocols were set for when to defer sessions, adjust the prescription of sessions, terminate sessions, or refer the participant to their primary care provider for further evaluation. When needed, interventionists contacted the study investigators for further guidance in evaluation. Following any AE, summaries were written by the interventionist and shared with the study investigators for grading and evaluation.
Communicate with healthcare provider
In the event that the AE warranted further evaluation from a participant’s primary care provider, summaries of the AE were shared with the participant and faxed to the primary care provider. Participants were then required to obtain “re-clearance” from their primary care provider prior to continuing interventions or physical assessments in the study. Efforts were made by study staff to help facilitate the process.
Measures for implementation outcomes
Screening
Screening outcomes were operationalized to capture the team processes: (1) durations in days between each screening step (S1-S4), (2) duration in days from enrollment to the first intervention session, and (3) duration in days from S1 to the first intervention session.
Intervention adherence
Intervention adherence outcomes were operationalized as (1) session attendance defined as the total number of sessions attended out of 72 possible sessions and (2) Session dose adherence was calculated by dividing the number of sessions that achieved prescribed session dose/total number of sessions completed. For exercise sessions, adherence to 100% exercise dose means that participants exercised above the minimum target HR or RPE for the prescribed duration for attended sessions when cycling or exercised below the maximum target HR or PRE for the prescribed duration when stretching for attended sessions. For SOP/MLA, adherence to 100% cognitive session dose means that all attended sessions met the prescribed durations for attended sessions.
Safety
Safety outcomes were operationalized as the type, number, and severity of AEs. Type of AEs were dichotomized as (1) non-study-related: either “clearly not related to the study” or “doubtfully related to the study” and (2) study-related (intervention or procedure-related): “possibly,” “likely,” or “clearly” related to the study. In the ACT Trial11, each study-related AE was rated for severity and classification. The severity was rated with a 4-point Likert scale: 1 = minor (no treatment required), 2 = moderate (resolved with treatment), 3 = serious (resulted in inability to carry on normal activities or ongoing medical treatment was still required), and 4 = life threatening or fatal. Classifications of study-related AEs included (1) fall, (2) cardiac, (3) musculoskeletal, (4) gastrointestinal, and (5) other. All AEs were assessed and graded by the interventionist, the exercise physiologist, and a principal investigator with consensus.
Statistical analysis
Continuous variables were quantified with means (standard deviation) and median (interquartile range) and categorical variables with frequency (percent). Normality and homogeneity of variance were assessed by the Shapiro Wilk test and Levene’s test, respectively. Implementation consistency across sites was assessed by analysis of variance (ANOVA) or Kruskal Wallis test to satisfy test assumptions. Collective sex and race-based differences in outcomes were assessed using independent samples t-test or Mann Whitney U test. Specifically, days spent in screening phases and intervention length were non-normally distributed. Site-based differences were assessed by the Kruskal Wallis test, while Mann Whitney U test was used for assessing collective sex and race-based differences. Chi Squared test and Fisher’s exact tests were used to analyze AEs and sex and racial effects on AEs for the entire sample. Alpha was set at p = 0.05. All data were analyzed using SPSS version 28.0 (IBM Corp.; Armonk, NY).