VR-CogMoBal Training for Reducing Falls Among Older Adults With Mild Cognitive Impairment

Virtual-reality Based Cognitive-motor-balance (VR-CogMoBal) Training: Examining Behavioral and Neuromarkers for Fall-risk Reduction in Older Adults With Mild Cognitive Impairment

Older adults often display gait instability, impaired balance control and cognitive decline that lead to falls and fall risks. Approximately 60% of the elderly people with cognitive deficits experience a detrimental fall each year. Such motor and cognitive impairments further decreases physical activity levels in this population leading to restricted community integration, social behavior, depression and long-term disability. With the help of computer technology, studies have employed virtual-reality based interventions to address the above-mentioned concerns including sensori-motor, balance control and cognitive impairments. Previous studies have demonstrated promising results on improving the behavioral outcomes, and have identified such interventions have the potential to improve the underlying neurophysiological outcomes as well. While VR based training studies have demonstrated remarkable improvement in the balance control and gait parameters, physical activity levels and fall risk reduction, the gains on cognitive function is less pronounced. There is little evidence that VR-based training can explicitly address the higher executive cognitive domains associated with balance control and falls. Further, the effect of VR-based training on balance control and cognitive function is unknown among the older adults with mild cognitive impairment. Therefore, to address the cognitive domains explicitly, the current study aims to test the applicability of Wii-Fit Nintendo along with an additional cognitive load delivered via VR-based cognitive-motor training paradigm (VR-CogMoBal) in older adults with mild cognitive impairment. Lastly, the study also aims to identify the effect of such training on the underlying behavioral and neural outcomes. The behavioral outcomes will be assessed via performance on dual-tasking and clinical measures in the laboratory. The underlying neural outcomes will be assessed via fMRI outcomes. In order to determine the generalizing training effect at community level, a pilot sub-study to determine the physical activity levels post 4 weeks of training will also be conducted.

Older adults suffer from mild cognitive impairments with a prevalence rate of 3% to 22% and an incidence rate of 1% to 6% per year in the United States. Along with age associated locomotor-balance impairments, such cognitive decline among the elderly is known to increase the risk of falls, reduced physical activity and community integration, thus contributing to long-term disability. Daily living activities comprises of several concurrent motor and cognitive performances (dual-tasking) such as shopping in a supermarket, that requires higher executive cognitive functions and intact locomotor-balance control abilities. Falls during dual-tasking occur mostly due to the interference caused, i.e. during dual-task performances, either one or both task (motor or cognitive) performance is deteriorated that is known as cognitive-motor interference. Given that dual-task performances decline due to age-associated factors, daily living activities are highly challenging and difficult to perform for older adults with mild cognitive impairment. Although there are several conventional methods that incorporate locomotor-balance training, the nature of such interventions does not result in pronounced cognitive gains. Additionally, these interventions lack multi-sensory feedback, and due to the monotonous and repeated task practice of exercises characteristic, individuals do not seem to adhere to therapy leading to less compliance and decreased motivation to exercise training. In order to overcome such barriers, alternate form of therapy with the help of Virtual-reality devices, especially off the shelf commercially available exercise platforms emerged for training purposes. Although there is evidence that VR based training improves locomotor-balance control and is known to implicitly address cognitive functions, there is no knowledge that such VR based training can explicitly address higher executive cognitive functions. Therefore, based on preliminary studies tested the efficacy of cognitive training along with exergaming delivered via the commercially available off the shelf device- Wii-fit Nintendo and demonstrated promising results in improving balance control and cognitive function among the individuals with Chronic Stroke. The study resulted in decreased cognitive-motor interference during dual-task performance thereby exhibiting an improved performance on both cognitive and balance control function. Currently, there is lack of knowledge in determining specific interventions for improving dual-task performances among the older adults with MCI. Given that mild cognitive impaired older adults suffer from both motor and cognitive impairments, there is a need for testing the feasibility of a similar intervention among them and determine the change in the underlying neural biomarkers. Aim 1: The study is designed to test the feasibility (tolerability, compliance and effectiveness) of VR-CogMoBal training to improve physical function and reduce fall-risk in community-dwelling older adults with mild cognitive deficits by lowering cognitive-motor interference and dual task costs. Hypothesis: Participants will tolerate the training paradigm and will demonstrate significant improvements in balance, gait, cardiovascular and cognitive performance under dual-task conditions. Aim 2: To examine if the (VR-CogMoBal) will lead to higher cognitive function post-intervention. Hypothesis: Post-training compared to pre-training participants will show significantly greater global cognitive function, executive and working memory and decreased cognitive-motor load. Aim 3: To examine effect of VR-CogMoBal Training on changes in structural and functional connectivity within the cognitive-motor areas in the brain. Hypothesis: Post-training compared to pre-training, participants will show increased structural and functional connectivity at rest in the default mode network (memory consolidation, self-referential memory), fronto-parietal and supplementary motor areas (motor planning and execution, attention). Pilot Sub-study aim: The study aims to monitor the change in the number of steps taken a day before undergoing VR training and 4 weeks post-training.

Dual-task training, Virtual-reality exergaming, Cognitive-motor-balance training, Mild cognitive impairment, Older adults

Virtual reality based cognitive-motor balance (VR-CogMoBal) training will be delivered using the commercially available Wii-Fit Nintendo in conjunction with cognitive training. All participants will undergo 12 sessions of training in a tapering manner for four weeks with 90 minutes of training per session, i.e., 5 sessions for the first week, 3 sessions for the second week, and 2 sessions for the third and fourth week. Each session will be divided into 3 sub-sessions, where each sub-session will consist of playing 4 games in conjunction with cognitive task. All the games will be performed using a Wii-Fit balance board in front of a TV screen.

The participant will play four of the six balance board games Table tilt, Tightrope, Soccer, Balance bubble, Light Run and Basic Step (each game is max 1.5 minutes). Each game will be superimposed with any 3 of the 6 cognitive tasks (word list generation consisting of verbal fluency (VF) and category fluency (CF), digit recall (DR), analogies (AN), mental arithmetic (MA), repeated letter (RL). The cognitive tasks will be randomized making sure that all the cognitive tasks are played with all the games. The cognitive and balance board game scores will be noted on the scoring sheet for each session. A total 10 minutes rest interval between every sub-session will be mandatorily provided.

It is the average speed of center of gravity movement during intentional movement measured in degrees per second under single and dual-task conditions. Higher values indicate better performance.

It is the magnitude of a self-initiated movement (i.e., how far he/she wills to reach a target) without taking a step or losing balance measured in percentage under single and dual-task conditions. Higher values indicate better performance.

It is the actual magnitude of a self-initiated movement (i.e., how far did he/she actually reach a target) without taking a step or losing balance measured in percentage under single and dual-task conditions. Higher values indicate better performance.

It is the quality of a self-initiated movement (i.e., amount of movement actually exhibited towards the target to the amount of extraneous movement away from the target) measured in percentage under single and dual-task conditions. Higher values indicate better performance.

Reactive balance control will be examined via the stance perturbation test under single and dual-task conditions (simultaneous performance of Letter number sequencing task or auditory stroop task). Postural stability can be defined as simultaneous control of center of mass (COM) position and velocity during slip-like or trip like perturbation relative to the rear edge of base of support (rear heel). The position is normalized with the individual's foot length, and velocity by square root of gravitational acceleration and individual's body height. Larger values indicate greater stability.

The total time taken to complete the 4 meters will be noted. Speed will then be determined by using the formula distance (4 meters) covered by time taken to complete the test. Higher speed indicate better performance.

Spatial and temporal gait parameters like Step length, cadence and stride length will be determined during single and dual-task walking performance via the GaitRite mat. Higher values for step length and stride length, and lower cadence indicates better performance.

This is an oral trail making test which includes listing alternate letter and number from the cue given in sequence. This test will be performed under single and dual-task conditions. Accuracy (number of correct responses out of the total responses) of letter number sequencing will be calculated. Higher accuracy indicates better performance.

This test involves responding to the pitch (high or low) of the words "High" or "Low". This test will be performed under single and dual-task conditions. Accuracy (number of correct responses out of the total responses) of Auditory stroop will be calculated. Higher accuracy indicates better performance. Higher values indicate better performance.

Dual-task motor and cognitive cost will be calculated using the formula- [(Dual-task performance- Single Task performance)/Single task performance]. This will be calculated for dual-task performance during intentional postural sway, reactive balance control and gait conditions. Lower cost indicates better performance.

Interference in the reaction time via visual stroop task where the individual is expected to respond to the color in which the word is printed and not read the word. Performance will be identified via time taken to complete the test. Lower time indicate better performance.

Language fluency via verbal and category task will be administered. Performance will be identified via the total number of appropriate words responded on each of the task. Higher values indicate better performance.

The individual is asked to hit a key on the number keypad when a cue appears on the screen. Performance will be identified with time taken to hit the key after the individual sees the cue in seconds. Lower time indicate better performance.

Paired associated learning via grid task will be administered. Accuracy (number of correct responses out of the total responses) will be represented in percentage. Higher value indicate better performance.

Spatial working memory via unveil the star task will be administered. Performance will be identified via the total time (in seconds) to complete the task. Lower time indicate better performance.

List Sorting Memory test to evaluate working memory. This test requires the participant to recall and sequentially list the visually and orally presented stimuli. The accuracy of the participants response is computer generated. Higher value indicate better performance.

Picture sequence memory test will assess episodic memory of the individual. The number of adjacent pairs of pictures placed correctly will score a point. The scores are computer generated. Higher value indicate better performance.

Flanker inhibitory control and attention test is used to evaluate the participants ability to inhibit the attention to the stimulus flanking it and focus on a particular stimulus. Accuracy of the responses are recorded and the scores are computer generated. Higher value indicate better performance.

Dimensional Change card sort assesses cognitive flexibility and attention. Participants are asked to match a series of bivalent pictures either by colors or shapes accordingly. Responses are computer recorded and accuracy scores are computer generated. Higher value indicate better performance.

Pattern comparison processing speed test is used to evaluate the processing speed. The participants are expected to respond whether the two pictures side-by side are same or not the same. Accuracy will be recorded by the computer and scores are computer generated. Higher value indicate better performance.

Image acquisition will be performed in a 3T and 1.5T Magnetic resonance scanner (MR 750, GE healthcare, Milwaukee). Fractional anisotropy (FA) is a scalar value ranging from 0-1 and change from pre- to post-training will compared to determine the structural and functional connectivity. Increase in FA values post-training will indicate positive results of training.

Assess static and dynamic balance. This scale consists of the participant transferring from one chair to another, reaching forward, stepping up and down from a stepping stool, standing with eyes closed and open, one leg standing. It is a 14-item scale with each item score ranging from 0-4. Performance on the scale will be calculated on a total of 56. Less than 45 will indicate greater risk of falling.

Questionnaires such as Physical Activity Scale for elderly and activity specific balance confidence scale will be self-reported by the participant. Activity specific balance confidence scale consists of 16 items, and each item score ranges from 0-100. The total score with 0 confidence indicates no confidence and 100 with complete confidence.

Inclusion Criteria: MOCA less than 26 out of 30 Bone density with a T-score ≥ -2.5 Can understand and communicate in English Ability to stand for at least 5 minutes without an assistive device (length of a Wii Fit game) Exclusion Criteria: any acute or chronic neurological (Stroke, Parkinson's disease, Alzheimer's disease), cardiopulmonary, musculoskeletal, or systemic diagnosis recent major surgery (< 6 months) or hospitalization (< 3 months) Use of any sedative drugs HR > 85% of age-predicted maximal heart rate (HRmax) (HRmax = 220 - age) systolic blood pressure (SBP) > 165 mmHg and/or diastolic blood pressure (DBP) > 110 mmHg during resting), and/or oxygen saturation (measured by pulse oximeter) during resting < 90% Specific to MRI: Self-reported presence of pacemaker, metal implants, and/or Claustrophobia