The Manual Therapy and Strengthening for the Hip (MASH) Trial: Protocol for a Multisite Randomized Trial of a Subgroup of Older Adults With Chronic Back and Hip Pain

Abstract

Objective

Chronic low back pain (CLBP) is a disabling and costly condition for older adults that is difficult to properly classify and treat. In a cohort study, a subgroup of older adults with CLBP who had elevated hip pain and hip muscle weakness was identified; this subgroup differentiated itself by being at higher risk for future mobility decline. The primary purpose of this clinical trial is to evaluate whether a hip-focused low back pain (LBP) treatment provides better disability and physical performance outcomes for this at-risk group compared with a spine-focused LBP treatment.

Methods

This study is a multisite, single-blinded, randomized controlled, parallel arm, Phase II trial conducted across 3 clinical research sites. A total of 180 people aged between 60 and 85 years with CLBP and hip pain are being recruited. Participants undergo a comprehensive baseline assessment and are randomized into 1 of 2 intervention arms: hip-focused or spine-focused. They are treated twice weekly by a licensed physical therapist for 8 weeks and undergo follow-up assessments at 8 weeks and 6 months after randomization. Primary outcome measures include the Quebec Low Back Disability Scale and the 10-Meter Walk Test, which are measures of self-report and performance-based physical function, respectively.

Impact

This multicenter, randomized clinical trial will determine whether a hip-focused or spine-focused physical therapist intervention results in improved disability and physical performance for a subgroup of older adults with CLBP and hip pain who are at increased risk of mobility decline. This trial will help further the development of effective interventions for this subgroup of older adults with CLBP.

Introduction

Chronic low back pain (CLBP) is a prevalent, costly, and disabling problem in the older adult population; the rising prevalence suggests that current treatment approaches are not effective.^1–8 Several large, longitudinal studies have demonstrated that CLBP among older adults is independently associated with a steeper rate of decline in performance-based mobility function (ie, gait speed, chair rise performance, balance) compared with older adults without low back pain (LBP).^9–11 Because poor and/or declining mobility function in this population is predictive of increased future risk for mortality, institutionalization, and overall disability,^12–16 effective interventions are needed to improve function.

Experts in LBP have speculated that older adults with LBP do not belong to 1 homogeneous group but rather belong in subgroups that share similar clinical characteristics.^17,18 In younger cohorts, evidence exists that matching treatments to the limitations of an individual patient can improve physical therapy outcomes.^19–22 It stands to reason that older adults would similarly benefit from matching treatments to impairments; however, older adults are often excluded from clinical trials.²³ The National Institutes of Health (NIH) has recently highlighted the importance of including participants across the lifespan.²⁴ Tailored interventions, which align with recent calls for precision medicine initiatives,²⁵ may be particularly important in older adults with painful conditions, where heterogeneity is particularly prevalent.^26,27 Thus, specific classification systems must be developed for this population.

In a previous cohort study involving older adults with CLBP,²⁸ we identified a subgroup with a combination of elevated hip pain and global hip muscle weakness. This subgroup was at risk for markedly worse CLBP, functional outcomes, and self-efficacy over the course of 12 months, putting them at risk for future mobility decline and institutionalization. Compared with older adults without pain, those with CLBP had a greater prevalence of clinical hip symptoms associated with hip osteoarthritis, which were associated with worse health-related quality of life.²⁹ The totality of these results prompted us to develop the current trial, which focuses on evaluating whether a hip-focused LBP treatment will lead to reduced disability and improved physical function compared with a spine-focused LBP treatment. Our primary hypothesis is that members of the at-risk subgroup who receive hip-focused LBP treatment will have decreased pain and better functional outcomes than those who receive spine-focused LBP treatment.

Methods

Design

The Manual Therapy and Strengthening for the Hip in Older Adults With Chronic Low Back Pain Trial (MASH) is a multi-site, single-blind, parallel arms, randomized controlled, Phase II trial and adheres to the SPIRIT guidelines (Figure).³⁰

Figure.

Study overview.

Setting

Assessment and treatment sessions take place within clinical laboratories at the University of Delaware, the University of Pittsburgh, and Duke University. The home exercise program (HEP) portion is completed without supervision at each participant’s residence.

Participants

A total of 180 older adults aged 60 to 85 years are being recruited across the aforementioned 3 sites. We are utilizing newspaper advertisements, on-site visits to local senior centers, research registries, referrals from nearby rheumatology and geriatrics clinics, and electronic medical record invitations to recruit participants meeting specific inclusion criteria (Tab. 1). Two of the inclusion criteria are used to classify participants into the at-risk subgroup: hip internal rotation (IR) strength <0.26 (taken isometrically with a handheld dynamometer and normalized to body weight), and Hip Disability Osteoarthritis Outcome Score (HOOS) >5 on items P4–P8. These 5 items from the HOOS pain subscale ascertain the amount of hip pain a person has experienced in the last week during a specified activity; individual items are scored from 0 = no hip pain to 4 = extreme hip pain, with greater summed scores indicating greater severity. Using data from previous work,²⁸ we developed a simple decision tree using regression and receiver operating characteristic curve analyses and found that if individuals met the criteria for hip IR strength and HOOS item scores, they were placed accurately into the group with elevated hip pain and global muscle weakness (ie, our study population of focus) 81% of the time. To maximize safety, we adhere to specific exclusion criteria as well (Tab. 1).

Table 1.

Inclusion/Exclusion Criteria

Inclusion criteria	Low back pain >3 mo for at least half of the days in last 6 mo50
	Moderate low back pain intensity (>3 on 0–10)
	Normalized isometric hip internal rotation strength <0.26a
	Hip Disability and Osteoarthritis Outcome Score >5 on pain-related items P4–P8a
Exclusion criteria	Previous hip fracture with repair
	Hip fracture within last 15 y without repair
	Total hip replacement
	Known spinal pathology other than spinal stenosis and/or osteoarthritis
	Severely impaired mobility (ie, requires use of a wheelchair)
	Folstein Mini-Mental State Examination Score <2451
	Severe visual or hearing impairment
	Red flags such as fever, significant unintentional weight loss >10 pounds, pain that awakens or keeps one awake at night, trauma that preceded the onset of pain, or signs and symptoms of cauda equina
	Significant pain in legs greater than the back
	Acute illness (eg, COVID-19)
	Inability to participate in the study for the full 6 mo (eg, moving residences)
	Receipt of manual or exercise therapy for low back or hip within the last 3 mo

^a Criteria defined from decision tree analysis.

Enrollment and Assessments

All potential participants complete an initial phone screen. Individuals deemed preliminarily eligible are invited on-site to participate in an in-person screening, during which they complete the informed consent and answer questionnaires related to demographics. They have their height and weight taken and undergo hip IR strength testing; this information is used to determine final eligibility.

Eligible participants who provide informed consent undergo 3 assessments: (1) baseline (completed at the time of the in-person screening), (2) 8 weeks postrandomization (postintervention), and (3) 6 months postrandomization. Each comprehensive assessment is performed by a licensed physical therapist and includes self-report questionnaires, range of motion and strength measurements, and physical performance tests. Gerontological evidence indicates self-report and performance-based measures of function offer complementary information; therefore, both types should be used in the comprehensive assessment of physical function.^13,31–34 For this trial, primary outcome measures are the Quebec Low Back Pain-Related Disability Scale (QBPD)³⁵ and the 10-Meter Walk Test (10MWT).^36,37 The QBPD is a disease-specific self-reported measure of physical function that has excellent reliability and good construct validity compared with other pain and disability measures among older adults.³⁵ The 10MWT is a reliable and valid measure of gait speed, which is predictive of disability and mortality in older adults.^12–16 Secondary outcome measures were selected to capture constructs such as self-efficacy and pain perception (Tab. 2).

Table 2.

Outcome Measures^a

Test	Domain	Description	Scoring
QBPD35,b	Low back pain-related disability	Questionnaire: 20 items related to daily activities, which are grouped into 6 categories	Higher scores = greater disability
10MWTb,36,37	Gait speed	Participants walk along a linear pathway at their “usual pace” and “as quickly as possible”	Higher gait speeds = better outcomes
PHQ-952,53	Depressive symptoms	Questionnaire: 9 items related to depressive symptoms	Scores ≥10/27 indicative of major depression
LOBACS54	Self-efficacy	Questionnaire: 15 items scored according to participant’s level of confidence in performing activities	Higher scores = greater self-efficacy
PCS55,56	Pain perception	Questionnaire: 13 items scored as 3 subscales to evaluate constructs of rumination, magnification, and helplessness	Higher scores = increased catastrophizing
Movement-evoked pain57	Pain provocation with activity	Participants reported pain intensity before, during, and after 6MWT and 30 Second Chair Stand Tests	Smaller increases in pain = lower pain provocation with activity
Quantitative sensory testing58–60	Pain pressure threshold	Algometer used to measure pressure pain sensitivity at 4 sites bilaterally: upper trapezius, posterior superior iliac supine, greater trochanter, and tibialis anterior	Higher values = higher pain pressure threshold
Hip strength61,62	Strength	Hip strength (abduction, extension, internal and external rotation, flexion) taken with a hand-held dynamometer, normalized to body weight	Higher forces = greater strength
6MWT63	Functional mobility	Participants walk around a predetermined course trying to cover as much ground as possible in 6 min	Longer distances = better outcome
30-Second Chair Stand Test64	Functional mobility	Participants perform as many sit-to-stands as possible in 30 sec with arms folded across chest	More sit-to-stands = better functional mobility
HOOS65	Hip-related disability	Questionnaire: 40 items across 5 domains (pain, symptoms, activities of daily living, sport and recreation function, hip-related quality of life)	Higher scores = better hip outcomes
PROMIS-2966–68	Health-related quality of life	Questionnaire: 29 items across 8 domains (physical function, anxiety, depression, fatigue, sleep disturbance, ability to participate in social roles and activities, pain interference, and pain intensity)	Higher scores = greater presence of each outcome

^a 6MWT = 6-Minute Walk Test; 10MWT = 10 Meter Walk Test; HOOS = Hip Disability and Osteoarthritis Outcome Score; LOBACS = Low Back Activity Confidence; PCS = Pain Catastrophizing Scale; PHQ-9 = Patient Health Questionnaire-9 item; PROMIS-29 = Patient-Reported Outcomes Measurement Information System; QBPD = Quebec Back Pain Disability Scale.

^b Primary outcome measure.

Modifications Related to COVID-19

Early in study recruitment, the COVID-19 pandemic forced changes to our assessment protocol. Modifying trials to reduce in-person exposure is consistent with current clinical trial recommendations.^38,39 We sought to reduce the on-site time and accomplished this by having participants fill out questionnaires remotely.

Randomization and Blinding

The study biostatistician uses statistical software to generate a randomization plan using permuted block with random block sizes to ensure a roughly equal number of participants is assigned to each arm by the end of the study in a 1:1 allocation. Randomization is stratified by site, as well as by sex, due to the reported sex differences in pain conditions among older adults.²⁰ Assignments generated by the statistician are uploaded into the electronic data capture (EDC) system. The site study coordinator retrieves the assignment once a participant is deemed fully eligible and provides this information to the treating therapists (interventionists), thereby concealing allocation. Therapists performing the assessments (assessors) are masked to group assignment. Interventionists are masked to participants’ outcome data. Because treatment assignment cannot be withheld from the participants, both assessors and interventionists ask the participants to avoid discussion of any aspects of treatment with the assessors and other research participants. Interventionists are not assigned to a particular intervention arm and may treat participants in both groups.

Interventions

Eligible participants are randomized into 1 of 2 intervention arms: hip-focused or spine-focused treatment. Participants receive treatment from a licensed physical therapist twice weekly for 8 weeks. All interventionists participate in at least 4 hours of one-on-one training, during which they practice hands-on techniques under the supervision of a therapist trained in the MASH trial treatments; assessments of treatment fidelity occur regularly. Interventionists give participants an HEP that includes a pictorial exercise handout and an exercise log, and ask them to perform it twice weekly on non-therapy days. Although education related to the HEP is primarily focused on correct performance, interventionists also discuss several principles: the promotion and understanding of the anatomical and structural strength of the human spine; the use of active coping strategies, such as exercise; early resumption of daily, vocational, and social activities, even when pain is present; and the importance of improved activity levels, not just pain relief.⁴⁰ All exercises are progressed based on meeting a quota criterion (eg, 3 sets of 10 repetitions). Interventions are displayed in Table 3. Full manual therapy protocols, exercise protocols, and HEP pictorial handouts for both groups are included in the Supplementary Material. Our intervention development was informed by both our pilot work and current evidence.

Table 3.

Summary of Intervention Components

	Component	Description
Hip-focused
On-site (45 min)	Hip manual therapy	Four hip mobilization techniques (bilateral) • Long-axis distraction (Grade III sustained) plusmanipulation • Anterior–posterior oscillatory mobilizations (GradeIII/IV) • Lateral femoral glide (Grade III sustained) with hipinternal rotation oscillatory mobilizations (GradeIII/IV) • Posterior–anterior oscillatory mobilizations (GradeIII/IV) Manual stretches (bilateral) • Hamstrings • Hip flexors
	Functional hip exercises	Two phases • Visits 1–8: partial wall squats, hip abduction withelastic band, • forward step-ups •Visits 9–16: wall squats, side-stepping with band, lateral step-ups
Home exercise program education (15 min)	Hip strengthening exercises with elastic band	Hip abduction Hip extension Hip internal rotation Hip external rotation
	Trunk muscle training exercises	Bracing Anterior trunk (eg, curl-ups) Posterior trunk (eg, alternating arm lifts in quadruped) Lateral trunk (eg, side bridges)
Spine-focused
On-site (45 min)	Lumbar manual therapy	Central posterior–anterior oscillatory mobilizations (Grade I–II) to L1-L5 levels Effleurage to thoracolumbar area
	Trunk muscle training exercises	Bracing Anterior trunk (eg, curl-ups) Posterior trunk (eg, alternating arm lifts in quadruped)
	Stationary cycling	Stationary cycle without resistance for submaximal intensity
Home exercise program education (15 min)	Lumbar flexibility exercises	Generalized stretches to enhance lumbar mobility
	Trunk muscle training exercises	Same as above but lateral trunk exercises omitted

Hip-Focused

The hip-focused LBP treatment consists of 2 parts: an on-site session and an HEP. During the on-site session, participants receive manual therapy to both hips and participate in supervised functional hip exercises. Manual therapy to the hip includes joint mobilizations and stretching. Joint mobilizations include long-axis distraction (with a manipulation at the end of the 30 seconds) as well as oscillatory grade III to IV (3 × 30 seconds) joint mobilizations in 3 other directions: anterior–posterior, lateral glide with IR, and posterior–anterior. Hip joint mobilizations were adapted from prior work, where they were shown to be effective in improving various clinical outcomes among individuals with a primary diagnosis of hip osteoarthritis.^41,42 Stretches include supine hamstring stretching and either side-lying or supine hip flexor stretching. Functional hip exercises focus on strength, endurance, and control of the hip muscles while completing functional tasks (ie, squatting and stepping). They are performed with therapist supervision for safety. The last 15 minutes of each on-site treatment session are dedicated to moist heat pack application and education regarding the delivery of the HEP: therapists demonstrate all exercises, ensure that participants use proper form, and increase the dosage on HEP logs according to each participant’s progress.

The hip-focused HEP consists of 3 parts: hip stretching, hip strengthening exercises, and trunk muscle training (TMT) exercises. The strengthening exercises, which were adapted from prior work,⁴³ focus on the strengthening of various hip muscles using resistance bands. The TMT exercises, which target muscles in the anterior, posterior, and lateral trunk, are used to help improve neuromuscular control and dynamic spinal stability. TMT exercises were adapted from prior work in which they were shown to improve various clinical outcomes among older adults with CLBP.⁴⁴

Spine-Focused

The spine-focused LBP treatment consists of 2 parts: an on-site session and an HEP. During the on-site session, participants receive manual therapy to the lumbar spine and participate in TMT exercises and stationary cycling without resistance. Manual therapy includes oscillatory grade I to II central posterior–anterior joint mobilizations to lumbar levels L1 to L5 for pain relief as well as light soft tissue effleurage massage to the thoracolumbar paraspinal muscles. The TMT exercises performed in this group target muscles in the anterior and posterior trunk; lateral trunk exercises are omitted to avoid inadvertently targeting the hip abductor muscles. Participants perform stationary cycling without resistance, because clinical practice guidelines highlight strong evidence supporting the effectiveness of low-intensity submaximal fitness and endurance exercises for adults with CLBP.⁴⁰ As in the hip-focused group, the last 15 minutes of each session are dedicated to education regarding the delivery of the HEP. Once the HEP education is completed for each session, if there is time remaining, therapists may provide participants with moist heat to the lumbar spine to equalize face-to-face time between intervention groups.

The spine-focused HEP consists of lumbar flexibility exercises and the same TMT performed during the on-site visit. The lumbar flexibility exercises are matched to each participant’s preference and are varied throughout the treatment sessions.

Modifications Related to COVID-19

Due to COVID-19, aspects of the interventions were modified to reduce the overall time on-site. We moved the hip strengthening and TMT exercises (hip-focused group) and lumbar flexibility exercises (spine-focused group) from the on-site sessions to the HEP format.

Adherence

Participants are encouraged to adhere to the intervention arm to which they have been assigned; we define intervention adherence based on the number of in-person and HEP completed sessions (Suppl. Appendix). During the last 15 minutes of each treatment session for all participants, interventionists discuss updates to HEPs while the participants are receiving their moist hot pack application. Participants receive positive feedback and reinforcement if they adhere to their HEP and return their exercise logs. Participants receive treatment free of charge, and receive monetary compensation for follow-up assessments. Participants receive follow-up phone calls at the 3-, 4-, and 5-month post-randomization time points to promote retention and to monitor for adverse events (AEs).

Adverse Events

All study staff are responsible for monitoring for the occurrence of AEs, which are operationalized as any unfavorable or untoward occurrence to participants during their involvement in the clinical trial, including any abnormal sign, symptom, or disease that may or may not be related to research participation. Study staff catalogue new signs, symptoms, or diseases within Possible Adverse Event Report forms. Site investigators adjudicate Possible Adverse Event Reports and report any serious AEs or unanticipated problems to site and sponsor regulatory officials. We utilize the Common Terminology Criteria for Adverse Events classification and NIH common data elements (eg, severity) in recording AE classifications.⁴⁵

Process Evaluation

Employing a multisite trial design allows for exploration of the generalizability of the approach and findings but requires a focused approach to ensure that internal validity is maximized. We employ a rigorous assessment and treatment fidelity approach across sites to promote uniform and high-quality data acquisition: we have implemented a quality control program that includes the use of detailed operations manuals, therapist training, and periodic data review by the principal investigator. Protocols describe each testing or treatment procedure and contain an exemplar script. We have regular training workshops for all study staff and monthly meetings for both assessors and interventionists. The study investigators meet once monthly to discuss AEs and address any issues that arose over the prior month.

We have implemented an internal process review system (Tab. 4) designed to promote adherence to the study design and provide excellent participant care. Specifically related to treatment fidelity, we utilize processes that increase the likelihood of consistent treatment delivery across sites: automated chart review, manual chart review, and on-site observation of treatment sessions. Compliance issues are mitigated by on-site targeted retraining.

Table 4.

Internal Review Processes

Process	Type	Reviewer	Description
Automated	Chart	Web-based application	Significant deviations from standardized protocols (eg, omission of treatment components) automatically flagged in real-time by electronic data capture system
Manual	Chart	Study staff member	≥25% of charted treatment sessions reviewed for fidelity to treatment prescription and progression in mo 1–6; ≥10% of treatment sessions every year thereafter
On-site	Live	Study staff member	≥5 Audits per interventionist in first year observed for fidelity to treatment prescription and progression; ≥2 evaluations per year per interventionist thereafter

Data Analysis Plan

All analyses will be performed using intention-to-treat methodology. We will use the QBPD score and gait speed result to test our primary hypotheses regarding changes in disability and physical function both in the short term (immediately posttreatment) and longer term (6 months postrandomization). We will use constrained baseline linear mixed models with time (baseline, 8 weeks, and 6 months) and the group by time interaction as fixed effects while controlling for repeated measures using an unstructured correlation matrix between time points. We will treat the effect of time as nominal. We will also control for site and sex, because these are stratification factors in the randomization. If the interaction effect is significant, the simple main effect of group will be tested at each follow-up time point. If model normality and homoscedasticity assumptions are violated, we will either consider transformations that can be applied but maintain interpretability or use robust standard errors. We will test the secondary outcome measures using the same statistical models to identify which outcomes should be considered in future trials with a more conservative alpha.

The sample size for the trial was determined using the GLIMMPSE framework assuming the group by time interactions were significant for disability and physical function.⁴⁶ With alpha = .05 and 80% power, we will be able to detect a moderate effect, d = .45, for a total sample of 150 or 75 per treatment group. This assumes a correlation of repeated measures of 0.5. This effect size translates to a between-group difference of 6.8 points in QBPD scores and 0.095 m/s in gait speed over time between groups. For our final sample size of 150, a total of 180 participants will be enrolled, accounting for a dropout rate of 17% based on our previous trial of older adults with CLBP.⁴⁴ Sensitivity analyses using logistic regression models will be performed to identify predictors of missed assessment visits using baseline characteristics; if significant predictors are found, they will be included in a secondary analysis of the primary outcomes. If their inclusion results in estimates that differ from the initial analysis, these results will be reported because they would be unbiased if data are missing at random with respect to baseline characteristics.⁴⁷ The mixed models yield accurate parameter estimates if data are missing at random and have been shown to perform as well as multiple imputation given the same assumption of the missing data mechanism.⁴⁸ We will conduct several sensitivity analyses assuming non-ignorable missingness with differential imputation of poor change scores and pattern mixture models.⁴⁹ We will compare the results from these analyses to our primary analyses with all observed data to assess the robustness of our findings.

Ethics

This study was approved by the institutional review boards at each of the participating study sites; each site provides continuing reviews per their site requirements. Oversight of the entire trial is provided by an independent data and safety monitoring board. At each site, participants complete the informed consent process and are assigned an identification code. Data are then entered into the EDC, which is password protected and securely stored. Enrollment for the trial began in November 2019 and will continue through 2022; trial recruitment and enrollment were suspended between March 2020 and August 2020 due to site shutdowns related to COVID-19 but were reinitiated as each site reopened.

Role of the Funding Source

The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Discussion

The development of safe, alternative treatments that lessen the burden of CLBP and promote the independence of community-dwelling older adults is essential. Although experts agree that tailored approaches to treatment are necessary, there is a scarcity of evidence relating to older adults. Our study is unique in that by identifying a subgroup with hip impairments, we can determine the effectiveness of a tailored versus standard approach and address key initiatives set forth by the NIH.

This trial has limitations to consider when results are reported. We specifically exclude people with total hip replacements as well as certain individuals with a history of hip fracture. This decision was made with participant safety in mind, because hip mobilization is contraindicated in many of those individuals; however, we recognize that this may exclude participants who would fall into our subgroup of interest. Further, this is a single-blind trial, and there is no way to mask treatment allocation from the interventionists or participants. This leaves open the possibility of a participant inadvertently revealing allocation; we have attempted to mitigate this by consistently reminding participants to discuss treatments with only interventionists and not assessors or other participants. Another potential area of bias may arise if interventionists believe that 1 treatment arm is more effective than the other; we have attempted to mitigate this by incorporating the concept of equipoise (ie, the understanding that we do not yet know which treatment arm is indeed more effective) into the therapist training modules, because both interventions target impairments commonly experienced in this subgroup. The concept of equipoise is important in this trial due to the potential misperception that 1 randomly assigned treatment group was favored by the investigator team. Accordingly, we have created a protocol that ensures both treatment groups are receiving similar proportions of active exercise/intervention and both groups received the same type of evidence-informed information regarding exercise and CLBP.⁴⁰

Our trial has several strengths. It is a multisite study involving 3 sites with access to a broad demographic of older adults. We used specific inclusion criteria to identify members of a subgroup with increased risk of mobility decline. We incorporated tablet technology and an EDC to streamline data collection and reduce operator error. We incorporated internal review processes to promote delivery of standardized interventions and provide treatments with excellence. Importantly, the treatments used in this trial require no specialized equipment or advanced training and should translate well into clinical practice.

Author Contributions

Concept/idea/research design: P.C. Coyle, J.M. Sions, C.G. Patterson, R.T. Pohlig, C.B. Simon, S.Z. George, G.E. Hicks

Writing: J.M. Pugliese, P.C. Coyle, P.J. Knox, J.M. Sions, C.G. Patterson, R.T. Pohlig, C.B. Simon, D.K. Weiner, S.Z. George, S. Piva, G.E. Hicks

Data collection: J.M. Pugliese, P.C. Coyle, P.J. Knox, J.M. Sions, C.B. Simon, G.E. Hicks

Data analysis: C.G. Patterson, R.T. Pohlig, G.E. Hicks

Project management: J.M. Pugliese, P.C. Coyle, P.J. Knox, J.M. Sions, S.Z. George, G.E. Hicks

Fund procurement: G.E. Hicks

Providing participants: C.B. Simon, S.Z. George, G.E. Hicks

Providing facilities/equipment: S. Piva, G.E. Hicks

Providing institutional liaisons: S. Piva, G.E. Hicks

Clerical/secretarial support: J.M. Pugliese, P.C. Coyle, J.M. Sions

Consultation (including review of manuscript before submitting): J.M. Pugliese, P.C. Coyle, P.J. Knox, J.M. Sions, C.B. Simon, D.K. Weiner, S.Z. George, S. Piva, G.E. Hicks

Funding

This trial is funded by the National Institutes of Health’s National Institute on Aging (grant no: 2 R01 AG041202).

Ethics Approval

This study was approved by the institutional review boards at each of the participating study sites; each site provides continuing reviews per their site requirements. Oversight of the entire trial is provided by an independent data and safety monitoring board.

Clinical Trial Registration

This trial is registered at ClinicalTrials.gov NCT04009837.

Disclosures

The authors completed the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported no conflicts of interest.