Multi-Sensory Training for Balance, EMG, Sensory Deficits in DPN

Introduction Diabetes mellitus is a group of chronic metabolic disorders characterized by persistent hyperglycemia due to defects in insulin secretion, insulin action, or both(1).

With the world's population aged over 60 projected to increase from 12% to 22% between 2015 and 2050, the burden of diabetes among the elderly is expected to rise substantially. In Iran, the prevalence of diabetes in individuals over 65 is estimated to reach 6.5 million by 2030, and approximately 53% of type 2 diabetic patients develop peripheral neuropathy(2, 3).

Diabetic peripheral neuropathy (DPN) is a progressive neurodegenerative disorder involving both peripheral nerve damage and maladaptive central sensory processing(4). This dual pathology impairs the integration of vestibular, visual, and proprioceptive inputs critical for postural stabilit (5). Older adults with DPN exhibit increased postural sway and higher fall incidence compared to their non-neuropathic counterparts(6, 7). Damage to large-diameter afferent fibers and microangiopathic nerve alterations further disrupt joint proprioception, resulting in impaired anticipatory and compensatory postural adjustments(8).

Given these challenges, targeted interventions are crucial to mitigate fall risk and enhance functional outcomes in this vulnerable population(9). In addition to sensory deficits, muscle weakness-particularly a 17% reduction in ankle dorsiflexor strength and a 14% decrease in plantarflexor strength-further compromises postural stability and increases fall risk among DPN patients(8). Recent evidence demonstrates that focused strength training and neuromuscular rehabilitation can significantly improve postural control and reduce fall incidence in this population(10).

Exercise-based interventions-including whole-body vibration training(11), vestibular exercises(12), lower-limb strength training(13), and ankle-strategy drills(14) - have demonstrated significant benefits for balance and neuromuscular function. However, most protocols target only a single sensory modality and may not adequately address the multisystem impairments characteristic of DPN(12, 15, 16). Emerging evidence indicates that combined multisensory or sensorimotor-gait training can improve proprioception, nerve function, and muscle activation in DPN patients (17). Notably, Ahmad et al. (2020) reported in a randomized controlled trial that an 8-week sensorimotor and gait training program significantly enhanced proprioception, nerve conduction velocity, and muscle activation-particularly in the medial gastrocnemius and multifidus-during postural tasks and gait in DPN patients(17). Similarly, Malwanage et al. (2024) found that a 12-week ankle-focused proprioceptive training program significantly improved ankle reaction strategies and postural stability (18). Despite these promising results, no study has specifically investigated the effects of a structured 12-week selected multisensory training on neuromuscular activation and sensory function in older women with DPN-a population at particularly high risk of falls. Targeting multisensory integration pathways may offer novel neurorehabilitative strategies to counteract CNS maladaptation in DPN.

However, no prior randomized controlled Trial has concurrently examined neuromuscular activation, balance, and cutaneous sensitivity following a multisensory training program in elderly women with DPN. Therefore, this study aims to investigate the effects of a 12-week selected multisensory training (SMST) intervention on EMG RMS amplitude of postural control muscles (semitendinosus, rectus femoris, lateral gastrocnemius, and tibialis anterior), along with performance on the stork balance test, Timed Up and Go (TUG) test, and two-point discrimination test in older women with diabetic peripheral neuropathy, compared to a control group. We hypothesize that SMST will significantly enhance neuromuscular activation patterns and balance performance, offering a comprehensive fall-prevention approach for this high-risk population.

Materials and Methods Study Design This randomized controlled trial employed pre-test and post-test assessments with a control group to evaluate the effects of a 12-week selected multisensory training (SMST) program on lower limb EMG activity, balance, and sensory function in older women with diabetic peripheral neuropathy (DPN). The study was conducted at Parsian Diabetes Clinic under the supervision of a board-certified neurologist. All data collection took place in a standardized laboratory environment to ensure measurement reliability. Participants received detailed information about the study objectives and procedures upon arrival, and written informed consent was obtained in compliance with ethical guidelines. Baseline demographic and clinical characteristics (including age, sex, diabetes duration, and neuropathy status) were recorded prior to testing. The study protocol received approval from the Research Ethics Committee of Damghan University (Approval ID: IR.DU.REC.a404.011) and adhered to the Declaration of Helsinki and relevant ethical standards.

Participants

In March 2025, thirty women aged ≥60 years with clinically diagnosed diabetic neuropathy (DN) from Mashhad were initially telephone-screened at 25 march 2025. nineteen eligible participants who met the inclusion criteria provided written informed consent for study participation (Figure 1):

1. Female sex;
2. Age ≥ 60 years;
3. Ability to stand unaided for at least 10 minutes;
4. No structural damage to the lower extremities;
5. Mini-Mental State Examination (MMSE) score ≥ 24;
6. Ability to walk 20 meters unassisted;
7. At least 10 years' history of diabetes;
8. Normal or corrected-to-normal vision. The severity of neuropathy among participants ranged from mild to moderate. All medications, including antidiabetic agents, antihypertensives, and analgesics, were monitored and documented under the supervision of a board-certified neurologist to assess potential impacts on study outcomes. Additionally, blood glucose levels were regularly monitored throughout the study to ensure adequate glycemic control. Participants were randomly allocated to either the SMST intervention group (n=10) or control group (n=9). The mean participant age was 62.21 ± 1.8 years. An independent researcher conducted randomization using a computer-generated sequence, with assignments concealed in sequentially numbered, opaque envelopes that were not reused after opening.

Intervention SMST Intervention

Participants in the SMST group completed two 40-minute supervised sessions weekly for 12 consecutive weeks in the University Rehabilitation Laboratory, conducted by a board-certified motor behavior specialist. The progressive-overload protocol increased exercise intensity by 5-10% every three weeks. Each session included:

1. Warm-Up (5-8 minutes):

• Dynamic stretching targeting lower-limb joints and muscles

1. Core Multisensory Training (25-30 minutes):

• Proprioceptive challenges: Static/dynamic balance exercises on progressively firmer foam surfaces (soft→medium→firm), performed with eyes open then closed
• Vestibular activation: Gait drills incorporating forward/backward/lateral walking with head turns, plus eyes-closed movements on unstable surfaces
• Visual tasks: Tracking moving visual targets during standing with controlled head movements

1. Cool-Down (5 minutes):

• Seated breathing exercises with eyes closed
• Gentle lower-limb stretching

Progression Protocol:

• Biweekly intensity adjustments based on performance logs
• Progression methods: Increased surface instability, added handheld weights (1-2 lbs), reduced base of support, and transitioned from eyes-open to eyes-closed conditions
• Individual modifications and session adherence were documented to maintain intervention fidelity and appropriate challenge levels This structured SMST protocol provides concurrent vestibular, visual, and proprioceptive stimulation designed to improve neuromuscular activation patterns, as quantified by SEMG RMS measurements of key lower-limb muscles.

Control Group To control for social interaction effects, the control group participated in twice-weekly 40-minute unstructured group discussions with peers and staff while maintaining their usual activity levels and avoiding any new exercise regimens.

Outcome Measures Mini-Mental State Examination (MMSE) Cognitive function was assessed using the Mini-Mental State Examination (MMSE), a widely used dementia screening tool developed in 1975. The MMSE has a maximum total score of 30 points.(19).

Lower limb EMG:

Root mean square (RMS) amplitude of surface electromyography (sEMG) signals is a widely accepted biomarker for quantifying neuromuscular activation and muscle fatigue, offering critical insight into muscle function in DPN patients(7).

In this study, a surface EMG system (2000 Hz) was used to collect surface EMG data from 4 muscles of the right limb: semitendinosus, rectus femoris, lateral gastrocnemius, and tibialis anterior. Skin preparation included shaving, abrasion, and cleaning with alcohol to reduce impedance. Surface EMG signals were recorded using bipolar Ag/agcl electrodes placed per SENIAM guidelines (10 mm electrode diameter, 20 mm inter-electrode distance) on the midline of each muscle belly, oriented parallel to fiber direction.

The intercepted EMG data were preprocessed using Matlab r2022a, including mean removal, bandpass filtering (4th order Butterworth filter, 10-400 Hz), full-wave rectification, and smoothing using a low-pass filter. The low-pass filter applied for smoothing was a 4th-order Butterworth filter with a cut-off frequency of 20 Hz. This filter was designed to effectively remove high-frequency noise while preserving the relevant muscle signal characteristics. The 4th-order Butterworth filter was chosen to ensure a flat passband and a smooth roll-off at the cut-off frequency(20).

For each muscle, the peak root mean square (RMS) value calculated across 5 to 25 s time-series represented the maximal EMG activity as reference (RMSPEAK). The average of the RMS values (250 ms, 50% overlapped) from this same time-series portion was also computed, representing the mean RMS activity during balance (RMSTASK). Both RMSPEAK and RMSTASK were averaged across 2 trials and muscle activation (in %) was computed, as follows(21).

• RMS EMG = [(RMSTASK/RMSPEAK)×100% ]

Timed Up and Go (TUG) Test The Timed Up and Go (TUG) test was employed to assess both dynamic balance and functional mobility. Participants performed the test by standing from a chair, walking 3 meters, turning, returning to the chair, and sitting down. Each participant completed two trials, with the average time (in seconds) recorded for analysis(22). This average time also served as the single-task walking time for dual-task assessments. The TUG demonstrated excellent reliability (ICC >0.90) for participants with moderate neuropathy and good reliability (ICC 0.70-0.90) for those with severe neuropathy(23).

Single-Leg Stance Test Static balance was assessed using the Single-Leg Stance test. Participants stood on one leg with eyes open, maintaining the lifted foot without contact with the supporting limb or ground. The test continued for a maximum of 30 seconds or until termination due to: 1- contact between legs, 2- foot touching the floor, 3- hopping/jumping movements, or 4- use of external support - indicating balance impairment(24, 25). Research demonstrates that shorter single-leg stance times correlate with peripheral and cardiac autonomic nerve dysfunction as well as clinical neuropathy in type 2 diabetes patients, independent of age.(26).

Two-Point Discrimination (2PD) Test:

Sensory function was assessed using the Two-Point Discrimination (2PD) test. The 2PD threshold represents the minimum distance at which two simultaneous mechanical stimuli are perceived as distinct points. Testing was performed using a [brand/model name] two-point discriminator applied to the plantar surface of the foot. With eyes closed, participants verbally reported "one," "two," or "I can't tell" after each stimulus application (27).

Data Analysis All analyses adhered to the intention-to-treat principle, focusing exclusively on EMG RMS data. Between-group and longitudinal comparisons were conducted using repeated-measures ANOVA (α = 0.05) in SPSS version 26 (IBM Corp., Armonk, NY, USA).