This randomized control trial (trial registration: ClinicalTrials.gov NCT04969094) will examine the impact of 12 weeks (3 times per week) of NMES + an MMBI versus an MMBI alone on fall risk, balance, mobility, and hip abductor composition and strength in 80 (40/group) older veterans at risk for falls. Following completion of the 12-week outpatient center-based intervention, individuals will be provided with a written home exercise program, and we will reexamine subjects at 6 months and 1 year post intervention to assess retention of changes in hip abductor function, mobility, and fall risk. We will enroll veterans aged 55 years and older who are currently at risk for a fall (defined as a four-square step test [FSST] time >12 seconds [24], or a history of a previous fall in the last year, or a reported fear of falling). This study has 6 phases, completed over ~15 months per person. Figure 1 details the study design, and Figure 2 shows the study timeline. The study protocol follows the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) guidelines [25]. Any protocol deviations, confidentiality breaches, or reportable adverse events will be reported to the respective institutional review board (IRB) and data safety monitoring board in accordance with local policies. In addition, the data safety monitoring board will conduct at least annual reviews of study materials and collected data to ensure integrity, security, and quality control.

All testing and interventions will take place in a Veterans Affairs (VA)-based outpatient geriatrics exercise and rehabilitation center.

An a priori power analysis was conducted for sample size estimation based on changes in the FSST and hip abductor strength in our nonrandomized pilot study. We estimate that 18-22 participants per group will be required to detect between group changes (effect size of 0.91-1.06; α=.05; β=.80) after 3 months of intervention for our primary outcomes of FSST and hip abductor strength. We aim to have 30 participants in each group complete the 3-month intervention and the 6- and 12-month retention testing phases. Accounting for an up to 25% dropout rate from enrollment through 12-month retention testing, we will randomize 40 individuals per group.

Participants will be excluded if they have any of the following characteristics that impair their ability to safely participate in exercise or the NMES intervention: (1) high cardiovascular risk (such as poorly controlled hypertension of >160/100 mm Hg, symptomatic angina at rest or with exercise, or syncope in the last year without known resolution of cause), (2) require home oxygen, (3) contraindications to resistance training such as active proliferative retinopathy, (4) contraindications for NMES such as any implantable cardiac device, (5) dementia, and (6) other medical conditions that preclude patient participation in this study as per medical judgment of study team.

Recruitment will focus on enrolling participants who are veterans. Participants will be recruited using the Department of Veterans Affairs Informatics and Computing Infrastructure and will be recruited from the Baltimore metropolitan area through media advertisement. Interested participants will be pre-screened through a standardized phone questionnaire. Individuals who pass the prescreening will be invited to our facility. After providing written informed consent and HIPAA (Health Insurance Portability and Accountability Act) authorization, they will undergo a history and physical examination conducted by the study’s medical personnel to ensure they (1) meet all inclusion and exclusion criteria and (2) can safely participate in the intervention.

After baseline testing, participants will be randomized 1:1 to either NMES + MMBI versus MMBI alone using computer-generated random numbers by the study biostatistician. We will use block randomization to stratify participants by age (55-64 years and aged 65 years and older). To address potential concerns of age after the completion of the study, we will also adjust any analysis with age as an independent factor. Group assignments will be placed in sealed envelopes. Upon recruitment and completion of baseline testing, the research coordinator will open the next envelope and record the group assignment. Only the biostatistician will have access to and knowledge of the randomization list.

Outcome measures will be assessed at baseline, following the 12-week intervention, 6 months after completion of the intervention (~9 months from baseline testing), and 12 months after completion of the intervention (~15 months from baseline testing). Testing will generally be conducted over 3 visits. Table 3 shows the details of the testing process.

For our primary and secondary outcomes, we will use repeated measures ANOVA (Statistical Analysis Software [SAS] MIXED Procedure) to compare values of our outcome variables at baseline, following the 12 weeks of intervention, and 6 and 12 months following completion of exercise: outcome=group + time + group × time. If there is an imbalance between the groups at baseline, the out-of-balance variable will be added as an additional independent variable to the model. We will use Corrected Akaike Information Criterion, a variation of Akaike’s Information Criteria, to select the correlation structure (unstructured, compound symmetry, autoregressive lag 1) that best accounts for the serial autocorrelation of repeated measures from the same subject. We will use per-planned contrasts (implemented using estimate statements) to compare the changes (3 months minus baseline; and 6 and 12 months minus baseline) in our 2 groups. We will use Poisson regression (SAS GENMOD Procedure) to compare the number of falls in our two groups over the course of the 1-year follow-up. Our analyses will be 2-tailed and will be performed following the principle of intention-to-treat. Missing data will be imputed using multiple imputation, which provides unbiased estimates of whether the pattern of missing data is missing completely at random or at random. If the pattern of missing data is nonignorable missing, accounting for missing values is complex and there is no fully satisfactory method for preventing bias other than modeling the mechanism that resulted in missing values which most often cannot practically be done, and in this case, multiple imputations is no worse than the usual approach of dropping subjects with missing data. Before accepting the results of the model, we will make sure the results conform to the assumptions of the model (eg, reviewing residual plots), and we will make sure that no observation has undue influence on the estimates produced by the model (eg, Cook D). We will conduct exploratory analyses to examine relationships among variables and their changes in response to the intervention. We will examine, in separate models, the association of hip abductor muscle properties including (1) peak MVIC as determined with the biodex, (2) cross-sectional area, (3) composition (% of intramuscular fat determined from CT), and (4) activation during gait as determined by electromyography as predictors of 5 outcome measures (1) FSST, (2) mini-BESTest, (3) functional gait assessment, (4) modified physical performance test, and (5) stride time variability. We will test the associations (1) at baseline and (2) we will examine the association between the change in the predictor variables and the change in the outcome variables. In addition, we will examine the association between the changes in baseline variables and fall rate, including the fall rate assessed during 1 year of follow-up post intervention, and the number of fall-free weeks. Finally, we will assess the association between the changes in predictor variables during the intervention and subsequent fall rates. This will provide a wealth of information on the interrelationship between hip abductors, balance, mobility, and fall risk.

Participants will be assigned a unique study ID stored in a password-protected database on a VA server that has a level and scope of security that equals or exceeds that established by the HIPAA Security Rules. Data collected during the study that is not captured electronically will be entered and stored in the VA REDCap (Research Electronic Data Capture; Vanderbilt University) system. Participants’ charts will be held in a locked room inside a locked cabinet. Access to individual participant numbers or personal identifiers will be limited to research team members who need these data to perform their roles in this study.

While we anticipate that there may be some risks associated with participating in this research program, we have attempted to minimize them. Subjects will undergo procedures that involve a mild-to-moderate degree of risk, including exercise and balance training and testing, strength assessments, and muscle composition assessments, including DXA and CT scans. Completion of exercise and balance training and testing is associated with the risk of falls and cardiovascular complications such as chest pain, heart attack, or sudden death, and complications related to stress and strains of muscles or twisted ankles. The risk of falling is comparable to the participants’ everyday lives, as the training is designed to use elements of everyday activities. We anticipate that some individuals may fall during the intervention, but the risk of falls and injury will be minimized by the supervision of all activities by trained study personnel. Furthermore, the risk of cardiovascular complications is minimal. The American Heart Association consensus statement on exercise standards estimates that the acute risk of sudden cardiac arrest during exercise training in participants with known cardiac disease is approximately 1 event per 60,000 hours of aerobic exercise [34]. The risk of exercise training is greater at higher exercise intensities. This risk is offset by prescreening participants with medical evaluations before exercise. There is also some risk for muscle soreness with the use of strength training as well as NMES. This will be minimized through the delivery of the intervention by trained study staff and proper education with all activities.

The risks of muscle composition assessments with DXA and CT scans include the risk of radiation exposure. While any dose of radiation could be potentially harmful, the radiation-related risks associated with the measurement of muscle composition are well within the established dosimetry of radiation guidelines for research participants, and radiation safety protocols will be maintained for all participants.

This study has been approved by the University of Maryland School of Medicine IRB (HP-00110719) and the Baltimore VA Medical Center Research and Development Committee in accordance with the ethical standards of responsible committees on human experimentation and with the Helsinki Declaration. Informed consent will be obtained from each participant before any study procedures, and they will be allowed to opt out at any point. Any adverse, notable, or unexpected events will be reported to an independent, internal Data Safety Monitoring Board. The University of Maryland IRB and the VA Maryland Health Care System Research and Development Committees require a data and safety monitoring plan for all studies classified as greater than minimal risk. This study meets the regulatory criteria for a greater than minimal risk study. As it is supported by the Baltimore VA Medical Center Geriatric Research and Education Clinical Center, per internal policies, it will undergo biannual review by the Baltimore VA Medical Center Geriatric Research and Education Clinical Center’s internal data safety monitoring board. The board's composition and standard operating procedures have been approved by the University of Maryland IRB. Its primary focus is safety, particularly monitoring expected and unexpected adverse events directly attributable to study participation (ie, falls on the treadmill).

Results from this study will be available to health care professionals and the public through the National Library of Medicine PubMed Central website within 1 year after the date of publication. In addition, study findings will also be presented at scientific meetings. The study investigators will be solely responsible for drafting all publications and will not engage professional writers.

This study was funded on January 1, 2022, with a data collection period from April 1, 2022, to December 31, 2026. As of March 2025, we have screened 100 potential participants and excluded 38. Of the 61 participants enrolled to date, 21 have completed the 12-month follow-up, 32 have completed the 6-month follow-up, and 41 have completed the posttesting. A total of 4 participants are currently in the intervention phase; 1 has just completed the baseline testing, while 15 have been dropped from the study (Figure 1).

Based on our preliminary data showing that the addition of NMES to the MMBI program results in greater improvements in the FSST and hip abduction strength, we anticipate that those in the NMES + MMBI group will make significantly greater improvements in all measures during the 12-week intervention compared with the MMBI-only group. As noted above, we anticipate that a small number of individuals will make no improvement following the intervention (nonresponders). We expect that we will have fewer nonresponders in the NMES + MMBI group compared with the matched MMBI group due to the superimposed muscle activation provided by the addition of NMES. We further hypothesize that the nonresponders in the MMBI-only group will exhibit increased IMAT and lower torque of their hip abductors at baseline, as the presence of IMAT is thought to contribute to impaired muscle activation [15,16,21]. While we anticipate that NMES will primarily maximize changes in the gluteus medius and gluteus minimus (the primary hip abductors), we will also examine how changes in these muscles, along with the tensor fasciae lata, contribute to changes in fall risk and strength. If changes in the TFL are advantageous, this information can be used to design a future trial targeting the use of the TFL to compensate for maladaptive changes in the primary hip abductors. We also anticipate that the NMES + MMBI group will demonstrate greater retention of the positive changes from baseline at both the 6- and 12-month timepoints following the cessation of the program and fewer falls due to improvements in the muscle quality of the primary hip abductors.

We anticipate that this VA Merit award will establish NMES as an efficacious therapy that can be used in conjunction with hip abductor strengthening and balance training programs for fall prevention. This comprehensive and augmented training approach has the potential to change existing treatment paradigms for the management of falls in older adults. NMES is a low-cost intervention that can be easily disseminated with the aim of reducing disability and health care costs related to the consequences of falls in older adults. While other studies have investigated the use of NMES in older adults, the majority have focused on its application to the more distal muscles of the lower extremities, such as the quadriceps or the plantar flexors [35]. The outcome measures in these small (often less than 20 people per intervention) studies [26,36,37] have been strength or static balance measures, but do not examine changes to balance and fall risk as they relate to the hip abductor alterations. Instead, NMES was frequently limited to unilateral application to measure changes in gait mechanics.[26] Also, these studies did not target older adults [36,38,39]. Our study will be among the first, large RCTs to examine the combined use of NMES on the hip abductors in conjunction with a multimodality balance intervention designed to improve balance and mobility and thereby reduce falls in older veterans.

Our study will use CT imaging to evaluate changes in muscle cross-sectional area and quality (ie, composition) posttraining. The use of CT imaging will allow us to examine changes in both the primary (gluteus minimus and gluteus medius) as well as the secondary (tensor fascia lata) hip abductors. Imaging the primary and secondary hip abductors will allow us to examine the changes and relative contributions of changes in both the primary and secondary hip abductors to mobility and fall risk in older veterans.

Understanding the contributions of the hip abductors to balance and mobility and finding targeted means to improve their function opens important new treatment options for older adults who are at a high fall risk. This study will address an important knowledge gap about the relative contribution of the primary and secondary hip abductors to dysmobility and fall risk. In previous studies examining the effects of resistance training on balance, less than 30% included any resistance training of the hip abductors [40] and no studies examined the hip abductors as a primary outcome measure, or used muscle imaging techniques such as CT scans to examine changes in both the primary and secondary hip abductors. Given our previous work demonstrating the importance of the hip abductors to balance recovery [11] and stepping [13], our study is a novel and important outcome in this population of older veterans. Finally, following the changes in both muscle composition and fall risk at 6 and 12 months after the end of training will allow us to evaluate the durability of the intervention to convey meaningful long-term benefits of the program.

Despite the numerous strengths of this study, it has several potential limitations. Our target population includes adults aged 55 years and older who are at risk of falls. This may confound our findings’ generalizability to other studies, as it is a younger age than many fall studies that traditionally only recruit older adults aged 65 years and older. We believe this traditional approach is suboptimal as falls are not just a problem of advanced age, as demonstrated by the earlier sharp increase in the prevalence of falls starting in middle age [41]. Excluding individuals 65 years and younger who are at high risk for falling (those who have previously had a fall, or those who demonstrate balance abnormalities) may prevent the opportunity for early intervention in an “at-risk” population [41]. Therefore, we intentionally decided to include a younger population in our study than the one traditionally used.

Another potential limitation of this study is the anticipated predominance of male participants. So far, we have enrolled 61 participants, consisting of 50 (82%) men and 11 (18%) women, which differs significantly from the general population, where 44.6% of individuals aged 65 years and older are male and 55.4% are female [42]. However, census data show that 96% of veterans in this age group are male [1], suggesting that our cohort may more accurately represent the veteran population. The inclusion of adults aged 55 years and older may explain the higher proportion of female participants in our study.

Our study does not include any “sham” treatment for NMES. The inclusion of a sham treatment with NMES on the hip abductors was considered. However, even low currents may have a confounding effect, and we therefore chose not to include a sham treatment. Finally, there are multiple other outcomes that could be evaluated, such as biomechanics. We have purposefully chosen our primary outcomes for this grant to be clinically meaningful, with an interest in hastening translation to patient care. Though we hope this study will inspire future biomechanistic studies, the inclusion of such measures is outside the scope of this project. Should NMES indeed enhance the beneficial effects of MMBI, there will be a strong rationale to examine the underlying biomechanical changes further.

Our analysis of the responders and nonresponders will provide important information for future studies, as it may help us to identify a subpopulation of individuals that may benefit most from the application of NMES. Our collection of the NMES stimulation intensity (mA) used and the number of sessions attended will enable us to evaluate a dose response of NMES and changes as our primary outcomes. This information could be used to design a future trial that examines these variables to optimize the application. As previously described, we will also examine the contribution of changes in both the gluteus medius and gluteus minimus and the tensor fasciae lata to changes in fall risk and strength. This may lead to future trials targeting the TFL to compensate for maladaptive changes of the primary hip abductors if our analysis reflects that it is indeed advantageous. Thus, the proposed research from this study will not only provide valuable information about the use of NMES to improve hip abductors and reduce fall risk but also provide multiple future research directions.

This clinical trial seeks to provide evidence from the first large, randomized control trial to test the hypothesis that NMES applied to the primary hip abductors, in addition to a multimodality balance intervention, is an efficacious adjunct to traditional rehabilitation to improve balance, mobility, and function in older veterans at risk for falls. If proven effective, this low-cost adjunct to resistance exercise can be easily transitioned into clinical practice to better reduce disability and defer health care costs related to falls in older adults.