JMIR Publications

ABSTRACT

Background:

With no cure currently available, lifestyle interventions aimed at slowing cognitive decline and delaying the onset of Alzheimer’s disease (AD) and related dementias are essential. However, existing cognitive training protocols often fall short in delaying dementia onset. We propose that interventions informed by the specific contributions of brain regions underlying cognition will yield greater efficacy. This pilot project focuses on spatial navigation (SN), as difficulties forming new and maintaining existing spatial memories represent a common and early symptom of AD, often leading to disorientation and loss of independence in daily life. Importantly, tau and amyloid-beta accumulation begins in brain regions critical for SN, including the medial temporal and posterior parietal cortices. Despite the relevance of SN impairments as an intervention target, few clinical trials have specifically addressed SN. Many existing protocols rely on desktop-based training, which deprives learners of movement-related sensory and kinematic information inherent to active navigation. Similarly, conventional neuroimaging methods require participants to remain immobile, limiting ecological validity.

Objective:

Our objective is to evaluate the feasibility of a fully immersive virtual-reality (VR) maze as a training paradigm, enabling participants to navigate mazes while walking in open space, thereby overcoming the aforementioned limitations.

Methods:

We employ Mobile Brain/Body Imaging (MoBI) to synchronously record whole-body movement and electroencephalography (EEG), enabling source-level analysis of brain activity during active navigation. The aims of this pilot project are to: (1) determine the feasibility of a VR-SN maze training protocol, and (2) explore training-induced neuroplasticity in older adults engaging in VR-SN exercises. We plan to enroll 25 older adults at risk of dementia, who will participate in sixteen 50-minute sessions over a 4-month period. VR mazes are designed to elicit distinct navigational strategies—allocentric and egocentric—during specific phases of maze learning (Stand/Encode vs. Walk/Navigate). Allocentric and egocentric strategies are known to rely primarily on the medial temporal and posterior parietal cortices, respectively. By integrating VR-SN maze training with source-localized, millisecond-precision brain measurements during active navigation, we will be able to identify, dissociate, and track brain dynamics underlying these strategies and assess their relationship with improvements in SN performance. The primary behavioral outcome will be navigation time. Neurophysiological outcome measures will include region-specific modulations in theta (3–7 Hz) and alpha (8–12 Hz) power.

Results:

Preliminary VR-MoBI findings indicate a significant increase in theta power as participants approached maze intersections, compared to a control condition in which navigation instructions were provided.

Conclusions:

This pilot study will provide the foundation for a future randomized clinical trial designed to evaluate the efficacy of VR-SN training for improving navigational abilities and, ultimately, delaying cognitive decline in older adults at risk for dementia.