Aging of Path Integration Ability and Its Neural Mechanisms

Faculty of Psychology, Southwest University; Key Laboratory of Cognition and Personality (Ministry of Education), Southwest University, Chongqing 400715, China

Path integration refers to the process by which individuals continuously integrate sensory cues during movement to track their spatial position. Against the backdrop of an increasingly aging population, a current research hotspot is whether the behavioral performance and neural characteristics of path integration can serve as predictors for early neurodegenerative diseases.

Studies conducted in both real-world and virtual reality environments indicate that path integration ability progressively declines during the transition from normal aging to pathological aging. This ability relies on the collaboration between grid cells and other spatial cells. The structural and functional deterioration of key nodes, such as the hippocampus and the entorhinal cortex, constitutes the neural basis for the decline of path integration in the aging process.

This article aims to provide a deep understanding of the behavioral differences in path integration decline during aging and its specific neural mechanisms. Furthermore, it seeks to provide a solid theoretical foundation for the development of assessment and diagnostic systems for aging, as well as targeted interventions.

Keywords

Path integration; Spatial navigation; Aging; Alzheimer's disease; Neural mechanisms

1 Introduction

According to the latest data from the National Bureau of Statistics, China's elderly population has reached 280 million, signaling that the country is facing severe challenges associated with an aging society. Along with the progression of deep aging, the incidence and mortality rates of age-related diseases have shown a significant upward trend, with Alzheimer's disease (AD) being a primary representative \cite{Rostagno2022}. Normal or pathological aging typically leads to a decline in cognitive functions, such as slowed information processing speed and memory loss, a phenomenon defined as cognitive aging \cite{Loaiza2024}.

Traditional research on cognitive aging has focused on fundamental cognitive abilities such as memory and attention \cite{Puthusseryppady2024, Stark2017}. However, recent studies have found that spatial navigation ability appears to be more sensitive to aging and holds promise as an early diagnostic marker for age-related cognitive impairment \cite{Ekstrom2023, Migo2023}. Spatial navigation is a critical cognitive function through which organisms perform spatial localization by integrating self-motion cues with environmental reference cues \cite{Ekstrom2023}. Path Integration (PI) is one of the core components of spatial navigation; it is the function by which an individual continuously updates and tracks their position in the environment by integrating real-time self-motion cues from the vestibular system, proprioceptors, and the visual system \cite{Loomis1993, Segen2024}.

An increasing body of literature suggests that PI is highly susceptible to both normal and pathological aging, exhibiting early and significant decline \cite{Koike2024}. This decline may stem from a specific response to the degeneration of the entorhinal-hippocampal circuit. Consequently, PI is considered a potential biomarker for early-stage neurodegenerative diseases \cite{Bierbrauer2023}. This study aims to explore the manifestations of PI decline and its specific neural characteristics during normal and abnormal aging processes. The analysis is organized into four areas: (1) analyzing differences in PI performance among healthy elderly individuals, high-risk populations, and clinical patients; (2) resolving neural mechanisms at microscopic and macroscopic levels; (3) synthesizing neural drivers of decline; and (4) reviewing limitations and future directions.

2 Decline of Path Integration Behavior in Normal and Abnormal Aging

Spatial navigation decline in healthy older adults, high-risk populations, and patients with Mild Cognitive Impairment (MCI) has been validated through various experimental paradigms. The Triangle Completion Task (TCT) serves as a core assessment tool for PI ability. In the TCT, participants traverse two target points while blindfolded and must rely on internal senses to return to the starting point, forming a closed triangular path \cite{Loomis1993}. Primary behavioral indices include distance error and angular error \cite{Colmant2025, Loomis1993, Newton2024}.

[FIGURE:1]

With the advancement of Virtual Reality (VR) technology, more interactive versions of the TCT have emerged, such as the "Apple Picking" paradigm and the Virtual Supermarket Task (VST). These allow for the isolation of specific cues, such as optic flow or boundary information \cite{Apple2024, Bierbrauer2020, Coughlan2020}.

2.1 Normal Aging and Path Integration

The decline of PI ability during normal aging is particularly evident in distance estimation deficits under single-cue conditions. Research has found that whether walking blindfolded or watching a VR environment while seated, older adults exhibit significantly higher distance errors compared to younger adults \cite{Stangl2018}. However, when multisensory cues are combined with rich environmental information (like landmarks), the PI decline associated with normal aging can be mitigated through compensatory mechanisms \cite{Bates2014}. Computational modeling suggests that distance errors in normal aging originate primarily at the velocity signal input stage rather than the integration process itself \cite{Stangl2020}.

2.2 Pathological Aging and High-Risk Populations

PI decline in populations at risk for AD differs from normal aging. Individuals carrying the $APOE$ $\epsilon4$ allele exhibit higher distance and angular errors in optic-flow-only tasks \cite{Bierbrauer2020}. Behavioral analysis suggests that PI deficits in at-risk populations are primarily driven by angular estimation errors, such as over-turning \cite{Newton2024}. Subjective Cognitive Decline (SCD) is also associated with higher distance errors, which modeling attributes to "memory leakage"—the decay rate of internal spatial representations \cite{Segen2025}.

As the disease progresses to preclinical AD (amyloid-beta positivity), individuals exhibit significantly higher angular errors under optic flow conditions, which correlate positively with medial temporal lobe (MTL) tau levels \cite{Colmant2025}. MCI patients exhibit significantly larger errors than healthy older adults and fail to improve even when landmark cues are provided \cite{Colmant2025, Howett2019}. This suggests that compensatory mechanisms fail in the later stages of pathological aging.

3 Neural Mechanisms of Path Integration

Path integration relies on the synergistic action of a multi-level neural system. At the cellular level, grid cells work with place cells, head-direction cells, and speed cells. At the network level, regions such as the posterior cingulate cortex (PCC), retrosplenial cortex (RSC), and hippocampus support spatial localization.

3.1 Neuronal Level Information Encoding

Grid cells, primarily located in the medial entorhinal cortex (MEC), provide a uniform, metric coordinate system for space \cite{Moser2005, Dong2024}. Their firing remains stable in darkness, relying on self-motion information. Hippocampal place cells fire at specific locations and achieve precise self-localization by integrating input from grid cells and environmental boundaries \cite{Hafting2005, Solstad2008}. Head direction (HD) cells encode spatial orientation based on vestibular and proprioceptive inputs \cite{Taube2007}, while speed cells encode movement velocity, providing the mathematical basis for distance estimation \cite{Dannenberg2019}.

3.2 Brain Region Level Integration

The vestibular system serves as the critical sensory foundation for self-motion data \cite{Donaldson2019}. This information is transmitted to the PCC and RSC. The PCC integrates vestibular information to support spatial updating \cite{Schindler2018}, while the RSC constructs dynamic spatial representations and tracks Euclidean distance \cite{Chrastil2015}.

The hippocampus serves a core computational function, potentially correcting accumulated errors in grid cells through feedback mechanisms \cite{Burgess2007}. The medial septum (MS) acts as the pacemaker for the limbic system's $\theta$ rhythm, which is crucial for grid cell activity and spatial memory \cite{Maidenbaum2018}. Finally, the medial prefrontal cortex (mPFC) is responsible for maintaining working memory and goal representations during navigation \cite{Arnold2014, Chrastil2017}.

4 Specific Neural Characteristics of Path Integration Aging

The decline of PI is closely related to the susceptibility of core brain regions to aging. In normal aging, impaired grid representation resulting from the deterioration of the entorhinal-hippocampal circuit is a key neural characteristic. Stronger grid representations in older adults are predictive of smaller navigation errors \cite{Stangl2018}.

In populations at risk for AD, grid cell dysfunction precedes structural atrophy. Although EC and hippocampal volumes may not differ significantly from healthy individuals initially, grid representation strength in the posteromedial EC is significantly weakened \cite{Newton2024}. At-risk populations may over-rely on less efficient head-direction encoding strategies. The RSC appears to be a key brain region for compensation in these groups \cite{Bierbrauer2020}.

In pathological aging (AD), the triple cascade of amyloid-beta deposition, tau-mediated neurofibrillary tangles, and progressive atrophy leads to severe impairment. Tau levels in the MTL are specifically correlated with angular error, suggesting that tau selectively disrupts direction encoding \cite{Colmant2025}.

5 Summary and Perspectives

Path integration is a sensitive indicator of cognitive decline. Normal aging is characterized by distance estimation deficits, while pathological aging is marked by early angular errors and the eventual failure of compensatory mechanisms. Future research should focus on:

1. Ecological Validity: Developing dynamic PI tasks that integrate multisensory cues and reduce the impact of technology unfamiliarity in the elderly.
2. Influencing Factors: Analyzing how gender, stress, and lifestyle modulate the aging of PI.
3. Neural Circuits: Moving beyond individual brain regions to explore the collaboration within neural networks.
4. Interventions: Utilizing deep learning for risk prediction and exploring neuromodulation (like Temporal Interference stimulation) and VR-based training to mitigate decline.

While predicting risk through PI is promising, ethical considerations regarding privacy and the psychological impact of early diagnosis must be addressed.