This study investigates the spatial navigation challenges faced by visually impaired older adults in complex transportation hubs, using eye-tracking and mixed-reality simulations. Findings demonstrate that age-related neurocognitive declines necessitate evidence-based universal design interventions, including enhanced visual contrast and simplified architectural geometries, to preserve independent mobility.
Research Article
Research on Inclusive Wayfinding Systems in Transportation Hubs for Visually Impaired Older Adults
In This Article
Summary
Abstract
As global demographics shift toward an increasingly aged society, the preservation of independent mobility and spatial navigation has emerged as a critical public health priority. This study investigates the spatial exploration behaviors of visually impaired older adults in complex airport terminals, utilizing real-world wayfinding tasks coupled with integrated eye-tracking technology. Through a comparative analysis between a visually impaired elderly cohort and a younger sighted control group, this research quantifies navigational effort via gaze angular velocity (AV), wayfinding duration, and gaze distribution heatmaps. Experimental results indicate that visually impaired older adults exhibit significantly higher gaze AV across all navigational sections, a phenomenon characterized here as a compensatory behavioral response to declining visuospatial function. However, this compensatory mechanism demonstrates a "saturation point" at structurally complex intersections, such as those with acute connection angles, where elderly participants fail to further scale their exploratory efforts. The analysis further reveals that dynamic traffic entities, specifically pedestrians, create an "attentional lock-in" effect that disproportionately suppresses peripheral awareness in older navigators. Correlating these behavioral markers with clinical indicators, including Mean Deviation (MD), Pattern Standard Deviation (PSD), and Useful Field of View (UFOV), this article establishes a logical framework for evidence-based inclusive design. Recommendations focus on geometric simplification of transit nodes, multisensory redundancy, and the integration of adaptive assistance technologies to mitigate the "environmental press" on the aging brain.
Introduction
The global transition toward an increasingly aged society represents a profound multi-disciplinary challenge for urban infrastructure, public health, and architectural engineering1. According to the World Health Organization (WHO), the population of individuals aged 65 and older is mathematically projected to reach 1.6 billion globally by 20502. Progressing in parallel to this demographic shift is a notable increase in age-related sensory and cognitive impairments3. The Lancet Global Health Commission indicates that approximately 1.1 billion people were living with untreated vision loss in 2020, a figure expected to escalate to 1.8 billion by 20504. In fact, epidemiological statistics suggest that one person in the world becomes functionally blind every 5 s5. Despite these significant demographic trends, visually impaired seniors remain structurally marginalized. It is crucial to acknowledge that "visually impaired older adults" do not constitute a homogeneous group; rather, they represent a highly diverse population characterized by intersecting variables, including varying degrees of sensory loss progression, baseline cognitive reserves, concurrent physical mobility constraints, and differing socioeconomic support systems6.
To systematically analyze the interaction between aging navigators and complex built environments, it is essential to anchor the discourse in M. Powell Lawton's seminal "Environmental Docility Hypothesis," originally conceptualized in 19687. This ecological model of aging posits a dynamic interaction between personal competence (biological health, sensory capacity, and cognition) and environmental press (the demands placed upon the individual by their surroundings)8. The fundamental axiom is that as an individual's internal competence decreases, the characteristics of their physical environment explain a significantly larger share of their behavioral outcomes9. In complex spatial hubs like airports, existing wayfinding systems impose a disproportionately high "navigational tax," rendering independent mobility almost impossible and causing direct "environmental exclusion"10,11.
Spatial navigation is a complex cognitive function coordinated by the hippocampal-entorhinal circuit, frequently described as the brain's internal Global Positioning System (GPS)12. Landmark research in Nature and Science has identified specialized neurons, such as grid cells and place cells, which allow the construction of internal cognitive maps13. However, as the brain ages, these neural representations become increasingly unstable14. High-resolution functional magnetic resonance imaging (fMRI) studies demonstrate that older adults exhibit lower neural pattern similarity when retrieving spatial information15,16. For visually impaired older passengers, a significant majority of whom rely on static signage in unfamiliar terminals, inadequate wayfinding systems constitute a substantial barrier to independent mobility and social participation. In the long term, such spatial and environmental exclusion directly precipitates severe real-world mobility outcomes, including chronic social isolation, a forced transition to a sedentary lifestyle, and a consequent acceleration of both physical frailty and cognitive decline.
Furthermore, older adults exhibit a "dedifferentiation" of neural patterns, meaning their neural representations of different spatial environments become less distinct, leading to profound difficulties in spatial distance discrimination and allocentric orientation (i.e., the cognitive ability to map and understand the relationship between objects in space independently of one's own physical position)17,18.
Traditional wayfinding assessments primarily rely on post-task questionnaires or observational protocols, which often fail to capture the real-time cognitive workload and transient attentional shifts of users19. Furthermore, conventional studies typically evaluate aging and visual impairment in isolation, utilizing standardized clinical tests that do not reflect the dynamic, multisensory demands of large-scale transit hubs. To bridge this methodological gap, this study proposes an integrated approach combining real-world eye-tracking with mixed-reality (MR) visual impairment simulations. The primary objective of this study is to quantitatively isolate and evaluate the specific navigational burdens placed on older adults with visual impairments in a complex airport terminal. By deploying this dual-modal methodology, we aim to uncover the precise visual and architectural variables that precipitate wayfinding failure, thereby translating behavioral deficits into evidence-based, inclusive design mandates.
Protocol
All experimental procedures involving human participants were conducted in accordance with the Declaration of Helsinki. The study protocol was reviewed and approved by the Tsinghua University Science and Technology Ethics Committee (Medicine) (Project No. THU01-20250087). Written informed consent was obtained from all participants prior to their inclusion in the study. The consumables and the equipment used are listed in the Table of Materials.
Dual-modal experimental framework
The methodology utilized a dual-modal verification system to capture the intersection of reality and impairment. In the first phase, participants wore a wearable eye-tracking system with a 50 Hz sampling frequency. This device captured high-resolution gaze dynamics, including saccadic frequency and fixation duration, as participants independently navigated a real-world terminal route from entrance to gate. In the second phase, head-mounted mixed-reality (MR) headsets were employed to simulate pathological vision loss. Using real-time OpenGL shaders, the system superimposed visual field defects onto participants' views of the terminal environment, replicating the characteristics of Macular Degeneration (central scotoma), Glaucoma (peripheral loss), Diabetic Retinopathy (patchy vision), and Retinitis Pigmentosa (tunnel vision). To identify the specific variables driving navigational failure, a standard L9 orthogonal array design was implemented. This statistical design systematically evaluated four independent variables, each assigned three discrete levels: information density (low, medium, high), coding format (text-only, icon-only, hybrid), environmental interference (low, moderate, severe), and time pressure (none, moderate, high). These parameters generated nine symmetrical experimental combinations. In addition to one unweighted baseline control scenario, this yielded a total of 10 experimental conditions. To mitigate sequential learning effects and carryover bias, the presentation order of these 10 scenarios was strictly randomized for each participant using a Latin square strategy (Supplementary Table 1).
Navigation tasks and visual impairment simulations
The study deployed a dual-tiered methodological apparatus. In the first phase, participants performed a sequence of standardized, real-world wayfinding tasks (Entrance → Security → Gate, spanning an approximate predefined route of 450 m) within a functioning large-scale airport terminal. Participants were instructed to navigate at their natural walking pace without requesting external assistance. All experimental sessions were conducted under standardized diurnal ambient lighting conditions (approximately 300–500 lux). Each session lasted approximately 30 min and included a mandatory 5 min adaptation period to allow participants to acclimate to the headset's weight and visual displays prior to data recording. Continuous gaze data were recorded at a 50 Hz sampling frequency. Data preprocessing and operational analysis were executed utilizing dedicated eye-tracking analysis software. During this phase, participants wore the aforementioned eye-tracking glasses. Prior to navigation, the system was individually calibrated for each participant using a standard single-point calibration procedure via its companion controller application. Experimental trials commenced only after the software generated a successful calibration confirmation, ensuring a spatial mapping accuracy threshold of <0.5 degrees of visual angle. Raw gaze vectors were smoothed via a median filter (window size = 3 samples) to isolate high-frequency signal noise. Fixations were operationally defined using a velocity-threshold identification (I-VT) algorithm, with the velocity threshold parameterized at 30 deg/s and a minimum fixation duration of 60 ms. Gaze-filtering parameters excluded periods of signal loss exceeding 100 ms, resulting from blinks or hardware artifacts. In the second phase, to strictly isolate the behavioral impact of visual deficits from general age-related cognitive decline, only the younger sighted control group utilized the MR headsets to simulate specific profiles of pathological vision loss. By executing custom-programmed real-time OpenGL shaders (developed within a standard cross-platform game engine), the MR system superimposed visual filters over the physical environment. Before each simulation block, the research team systematically verified shader activation and accurate visual field occlusion against a standardized physical calibration grid as an essential intermediate checkpoint. To ensure replicability, the simulation parameters were anchored in established clinical ophthalmological models20: Macular Degeneration was simulated via a central absolute scotoma occluding the central 10° of the visual field; Glaucoma and Retinitis Pigmentosa were simulated through varying degrees of peripheral occlusion (tunnel vision restricted to a 20° and 10° central aperture, respectively); and Diabetic Retinopathy was replicated using randomized patchy scotomas with a uniform 40% reduction in contrast sensitivity.
Methodological troubleshooting and quality control
To ensure the reliability of the dual-modal protocol, rigorous quality control measures were implemented. A critical step for eye-tracking fidelity was maintaining a spatial mapping accuracy threshold of <0.5 degrees; calibration drift, a common challenge during extended, real-world walking tasks, was addressed by scheduling mandatory recalibration checkpoints every 10 min. For the mixed-reality simulations, maintaining stable visual field occlusion despite participant head movement posed a significant technical challenge. This was mitigated by using the MR headset's spatial anchoring capabilities to rigidly lock the shaders to the user's interpupillary distance (IPD), and by systematically verifying occlusion alignment against a physical calibration grid prior to every trial.
Calculating average angular velocity
The core quantitative metric extracted was the Average Angular Velocity (AV) of gaze, serving as an operational indicator of spatial scanning intensity that potentially reflects both compensatory exploratory effort and underlying navigational uncertainty. AV was calculated as the magnitude of the angular displacement rate of the gaze vector over discrete sampling intervals (Δt = 20 ms, mathematically corresponding to the 50 Hz hardware sampling rate). The instantaneous angular velocity, denoted as ω(t), was computed using the filtered two-dimensional gaze coordinates (x, y) on the projected visual plane:
(3)
Results
Quantitative analysis of the eye-tracking data revealed a significant efficiency disparity between the primary cohorts, highlighting areas where contemporary terminal design may inadequately accommodate neurocognitive aging. Quantitative analysis revealed significant disparities in navigational efficiency between the two cohorts. The visually impaired elderly group required substantially more time to complete the security-to-gate sequence compared to the younger sighted control group (1092.9 s ± 134.5 s vs. 467.3 s ± 42.1 s; t(60) = 24.62, p < 0.001). Furthermore, error rates, defined as proceeding down an incorrect corridor requiring physical course correction, were significantly elevated in the visually impaired elderly cohort (22.8% ± 5.4% vs. 5.2% ± 2.1%; t(60) = 16.85, p < 0.001). Detailed demographic characteristics and baseline navigational metrics across groups, along with their statistical comparisons, are summarized in Table 1.
Eye-tracking data and trajectory analysis
The raw gaze dynamics, as recorded by the eye-tracking system, provided insights into visual search dynamics and modulated cognitive workload and internal disorientation of the elderly group. Figure 1 displays the raw eye-tracking data and the resulting spatial exploration trajectories for both groups during the decision-making phase at a major corridor junction.
The Average Angular Velocity (AV) metrics served as a primary indicator of cognitive workload and visual search effort. The visually impaired elderly group exhibited a significantly higher mean gaze AV compared to the sighted control group (129.0 deg/s ± 14.8 deg/s vs. 75.1 deg/s ± 9.3 deg/s; t(60) = 17.24, p < 0.001). This elevated spatial sampling rate provides quantitative evidence of the increased informational demand placed on participants with structural visual field deficits. Heatmap generation further illuminated these divergences. Sighted controls demonstrated highly efficient, "top-down" visual processing, clustering fixations tightly on high-level directional signage. Conversely, visually impaired participants demonstrated erratic, "bottom-up" gaze distributions. They frequently fixated on non-navigational structural features, such as high-glare floor reflections or the movement of dynamic entities, while entirely overlooking critical overhead wayfinding cues.
Heatmap visualization of spatial attention
Heatmap generation further illuminated the fundamental divergence in attentional allocation between the two cohorts. Figure 2 and Figure 3 provide a comparative look at gaze distribution at the entrance to the security screening area.
Pathological perspective simulation analysis
The deployment of mixed-reality shaders demonstrated how specific ophthalmological conditions disrupt the wayfinding process (Table 2). Participants simulating Macular Degeneration (AMD) experienced a central scotoma, obliterating foveal vision. Consequently, they exhibited a 45% increase in high-velocity head rotation (yaw) to utilize eccentric viewing, forcing the low-resolution peripheral retina to interpret distant text, which frequently resulted in misidentification. Under the Glaucoma simulation (tunnel vision), participants entirely missed 65% of signs mounted below 1.4 meters because their narrow visual aperture was strictly locked onto the immediate floor to prevent tripping. Diabetic Retinopathy severely disrupted the gestalt recognition of symbols due to patchy scotomas, causing high gaze dispersion. Finally, Retinitis Pigmentosa (RP) significantly impaired spatial updating; extreme peripheral loss restricted the continuous optical flow associated with path integration, leading to a complete "memory leak" regarding environmental boundaries whenever participants maneuvered around dynamic entities.
DATA AVAILABILITY:
The datasets generated and analyzed during the current study have been deposited in Zenodo and are available at: https://zenodo.org/records/20772473.
Figure 1: Eye movement experiment data records and experimental analysis. Please click here to view a larger version of this figure.
Figure 2: Heatmap visualization of spatial attention. Gaze distribution of the sighted control group at the entrance to the security screening area. Please click here to view a larger version of this figure.
Figure 3: Heatmap visualization of spatial attention. Gaze distribution of the visually impaired elderly group at the entrance to the security screening area. Please click here to view a larger version of this figure.
| Participant Group | Mean Duration (min:sec) | Avg. Gaze AV (deg/s) | Error Rate (%) |
| Sighted (Control) | 07:47.3 | 75.1 | 5.2 |
| Visually Impaired (Elderly) | 18:12.9 | 129 | 22.8 |
Table 1: Navigational metrics and subjective workload by cohort.
| Simulation Type | Key Visual Characteristic | Behavioral Impact |
| AMD | Central scotoma (black spot) | Failure to recognize primary gate text; high head yaw frequency. |
| Glaucoma | Peripheral loss (tunnel vision) | Consistently missing auxiliary facility signage (e.g., elevators). |
| Diabetic Retinopathy | Patchy, irregular scotomas | Disruption of character recognition; high gaze dispersion. |
| Retinitis Pigmentosa | Extreme peripheral loss + blur | Loss of spatial context; inability to recover after distraction. |
Table 2: Pathological perspective simulation analysis.
Supplementary Table 1: L9 Orthogonal array design for wayfinding simulation scenarios.Please click here to download this file.
Discussion
The transition from raw empirical data to actionable architectural intelligence requires rigorous deductive reasoning. To maintain scientific transparency, the following discussion explicitly delineates between direct behavioral observations measured in our eye-tracking and mixed-reality paradigms, and theoretical extrapolations based on established neurocognitive literature. The primary theoretical discovery of this research is the behavioral manifestation of the "Compensation Ceiling" effect. To understand the elevated AV scores in the elderly cohort, this study applies the Compensation-Related Utilization of Neural Circuits Hypothesis (CRUNCH) developed by Reuter-Lorenz. The CRUNCH model posits that the aging brain attempts to counteract declining neural efficiency by over-recruiting cortical resources at lower levels of task demand. In airport wayfinding, this likely manifests as heightened oculomotor activity; visually impaired older subjects exhibited elevated spatial sampling rates (driving AV higher), which may reflect a combination of active compensatory scanning and increased visual uncertainty when attempting to synthesize a coherent map from a degraded sensory environment. However, this compensatory mechanism hits a strict functional plateau. In corridors with acute-angle intersections, younger cohorts scaled their exploration upward to meet the structural demand, whereas the elderly group failed to do so. This drop in adaptability implies a cognitive breakdown; when the environment's complexity exceeds the brain's remaining "reserve," spatial disorientation occurs21.
This suggests that aging navigators may struggle to construct stable structural anchors in complex hubs, potentially leading them to rely on more dispersed and less efficient visual search strategies. Consequently, the elevated gaze AV observed in the visually impaired elderly group suggests an intensified, yet potentially fragmented, visual search strategy to acquire environmental cues, which may accelerate cognitive fatigue and contribute to subsequent decision errors22.
Furthermore, the data revealed a phenomenon of "Attentional Lock-in." Tracking dynamic pedestrians consumed the limited divided-attention capacity of the elderly (quantified by lower UFOV scores), causing them to miss peripheral signage. This is explained by the concept of "foveal load" in visual psychology (the cognitive demand placed on central, high-resolution vision). High cognitive foveal load has been associated with a functional constriction of the peripheral visual field (often referred to as cognitive "tunnel vision," where individuals fail to perceive peripheral spatial cues when central attention is heavily taxed). This dynamic may contribute to a "Search Paradox": as older navigators engage in resource-intensive sampling (higher AV) to compensate for central vision loss, their effective rate of holistic information acquisition appears to decrease. They are looking harder, but seeing less23. It is critical to note, however, that while elevated AV is interpreted here primarily within the framework of compensatory exploration, this behavioral metric inherently confounds multiple underlying states, including localized environmental unfamiliarity, search inefficiency, and cognitive distraction.
While the integration of mobile eye-tracking and MR simulations provides high ecological validity, several methodological limitations must be explicitly acknowledged. Technically, the physical weight of the MR headsets and their constrained field of view (FOV) may introduce artificial biomechanical movement constraints, potentially altering natural gait. Experimentally, simulating pathological visual impairment in healthy young adults assesses acute sensory deprivation, which does not fully replicate the long-term neurocognitive adaptations and compensatory strategies developed by actual visually impaired individuals over the years. Interpretively, as previously noted, AV remains a composite behavioral metric that cannot definitively isolate active compensatory scanning from baseline cognitive disorientation without concurrent subjective self-reporting. Regarding practical implementation and scalability, the extensive calibration requirements and high hardware costs currently limit the rapid deployment of this protocol across larger, more diverse architectural layouts. Future research should prioritize optimizing the reproducibility of these MR paradigms through standardized, open-source shader libraries and the adoption of lighter-weight spatial computing hardware.
These behavioral findings align with existing theoretical frameworks of spatial disorientation. When peripheral vision is eliminated (as in RP simulations), participants lose the continuous optical flow traditionally associated with anchoring internal spatial maps, which may contribute to the observed navigational errors and spatial memory loss. The navigator cannot subconsciously calculate their trajectory and must rely entirely on immediate, redundant visual cues. Therefore, the architectural environment must be systematically simplified to stay below the user's CRUNCH point, acting as an external cognitive prosthesis23.
Based on the empirical findings of this study (e.g., the specific impact of acute-angle intersections and high visual scanning burdens) and established ergonomic literature, this study proposes an inclusive framework. The following recommendations explicitly differentiate between our direct experimental outcomes and supplementary literature-supported guidelines. This framework aims to reduce the "navigational tax" imposed on fragile cognitive systems through standardized environmental interventions.
While the heatmap data highlights the severe negative impact of central scotomas on text recognition, existing accessibility literature suggests that spatial designers should implement verified contrast standards to mitigate this effect. For instance, established guidelines recommend enforcing a Light Reflectance Value (LRV) difference of ≥70% between navigational text and its background. This specific threshold ensures that symbols can be detected by rod cells in the peripheral retina without requiring foveal fixation, which is often obstructed in elderly populations. Furthermore, character height (H) must be scaled according to the viewing distance (D) using the formula H ≥D/100, with a minimum character size of 200 mm for primary directional signs to account for reduced visual acuity24.
To counteract the "Search Paradox" and ensure information is accessible under high cognitive load, critical junction points should introduce dynamic or animated graphics. The human visual system's motion-detection pathways remain highly preserved even when static object recognition fails, allowing a "flickering" arrow or high-saliency symbol to trigger an orienting response24. Additionally, informational signage should be simplified to contain no more than 3 to 4 units of information per panel to prevent the "choice overload" that destabilizes mental maps in aged navigators25.
To address the "Attentional Lock-in" caused by foveal load and the physical constraints of scanning overhead gantries, transportation hubs must mandate a "Dual-Height" wayfinding architecture. Primary 3 m overhead signs must be supplemented with auxiliary identifiers mounted at the 1.4 m to 1.6 m eye-level. This vertical zoning intersects the natural downward-resting gaze vector of an elderly pedestrian, ensuring information is absorbed passively without demanding high-cost saccadic vertical scanning. This redundancy is particularly critical near stairs and elevators, where the visual focus is primarily occupied by ground-level obstacles and physical balance.
Furthermore, to keep cognitive demands below the CRUNCH compensation ceiling, fundamental floorplan geometries must be standardized to right-angle (orthogonal) intersections. This approach eradicates the severe wide-angle scanning burden imposed by acute junctions, which this study identified as a primary failure point for elderly navigators. Standardizing nodes to a symmetric tree-branch structure rather than complex circular layouts fosters environmental legibility and reduces the frequency of "stops" required for spatial re-orientation. Designers should also prioritize "Geometric Polarization", the use of high-contrast pavement markings and large-scale architectural walls, to provide the hippocampal system with stable structural anchors26.
Finally, to compensate for failing medial entorhinal cortex (MEC) grid-cell networks and prevent spatial "memory leaks," wayfinding must transition from discrete checkpoints to continuous, multisensory guidance vectors. Navigational systems should establish an unbroken "wayfinding chain" where each decision point is logically linked to the next. The integration of Augmented Reality (AR) path projection, projecting high-contrast navigational arrows directly onto the physical floor via MR head-mounted displays, dramatically reduces the spatial scanning burden (AV) by providing a persistent egocentric cue27.
For passengers with severe visual impairment, these visual cues must be coupled with acoustic and tactile redundancy. Acoustic beacons utilizing directional sound or chimes at elevator banks can announce location arrivals, effectively bypassing damaged visual cortices28. High-contrast Tactile Walking Surface Indicators (TWSI) should be deployed at primary nodes to provide uninterrupted physical grounding, supporting "path integration" through proprioceptive and vestibular feedback. By aligning environmental "press" with the remaining functional "competence" of the elderly, these mandates fulfill the Universal Design vision of a hub that moves before the people do.
This study highlights the substantial navigational burden placed on visually impaired older adults in complex transportation hubs. By utilizing mixed-reality simulations and real-world eye-tracking, we demonstrated that spatial disorientation in this population primarily manifests as more dispersed and less efficient visual search patterns (elevated AV), rather than merely prolonged navigation times. Furthermore, the current empirical data emphasizes the critical need for clearer, geometrically simpler wayfinding systems, specifically, the implementation of orthogonal intersections and dual-height signage architectures. Future design interventions should prioritize these data-driven geometric and visual simplifications to effectively support the independent mobility of aging populations, avoiding over-reliance on complex theoretical cognitive reserves.
Disclosures
The authors have nothing to disclose.
Acknowledgements
The authors would like to express their sincere gratitude for the support that made this study possible. This research was funded by the 7th Batch of the 2025 Ministry of Education of the PRC Industry-University Collaborative Education Program, specifically the Kingfar-CES Human Factors and Ergonomics Program under Program No. 20251109. Appreciation is also extended to the Xiamen Academy of Arts and Design at Fuzhou University and Xiamen Rekey Medical Technology Co., Ltd. for providing the collaborative environment necessary to conduct our neurocognitive and spatial navigation simulations.
Materials
| Name | Company | Catalog Number | Comments |
|---|---|---|---|
| Airport terminal wayfinding scenarios | Self-developed experimental setup | N/A (self-developed) | Standardized navigation task design for behavioral assessment |
| Microsoft HoloLens | Microsoft, USA | www.microsoft.com/hololens | Simulation of pathological visual impairment in mixed-reality conditions |
| OpenGL shaders | Custom programmed | www.opengl.org | Real-time rendering of simulated visual field defects |
| Tobii Pro Glasses 3 | Tobii Pro, Sweden | www.tobii.com/products/eye-trackers/wearables/tobii-pro-glasses-3 | Real-world eye-tracking data acquisition during wayfinding tasks |
| Tobii Pro Lab | Tobii Pro, Sweden | www.tobii.com/products/software/applications-and-tools/tobii-pro-lab | Recording and analysis of fixation, saccade, and gaze trajectory data |
References
- Beard JR, et al. The World report on ageing and health: a policy framework for healthy ageing. Lancet. 2016;387(10033):2145–2154.
- Vollset SE, et al. Fertility, mortality, migration, and population scenarios for 195 countries and territories from 2017 to 2100: a forecasting analysis for the Global Burden of Disease Study. Lancet. 2020;396(10258):1285–1306.
- Swenor BK, Lee MJ, Tian J, Varadaraj V, Bandeen-Roche K. Visual impairment and frailty in older adults. JAMA Ophthalmol. 2020;138(1):11–19.
- Burton MJ, et al. The Lancet Global Health Commission on Global Eye Health: vision beyond 2020. Lancet Glob Health. 2021;9(4):e489–e551.
- Bourne RRA, et al. Causes of blindness and vision impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the Right to Sight: an analysis for the Global Burden of Disease Study. Lancet Glob Health. 2021;9(2):e144–e160.
- Flaxman SR, et al. Global causes of blindness and distance vision impairment 1990-2020: a systematic review and meta-analysis. Lancet Glob Health. 2017;5(12):e1221–e1234.
- Wahl HW, Gerstorf D. Person-environment-behavior research: where we have been, where we are, and where we could be heading. Gerontologist. 2020;60(1):108–112.
- Chaudhury H, Oswald F. Advancing understanding of person-environment interaction in later life: a commentary. J Aging Stud. 2019;51:100808.
- Iwarsson S, Horstmann V, Carlsson G, Oswald F, Wahl HW. Importance of the home environment for healthy aging: a conceptual framework. Gerontologist. 2023;63(3):442–452.
- Jia G, Guo C, Han X, Zhu W, Ji M. Wayfinding behaviors of elderly passengers in airport terminals: an empirical study based on eye-tracking technology. Appl Ergon. 2022;106:103901.
- Chang YC, Chen CF. Meeting the needs of disabled air passengers: factors that facilitate help from airlines and airports. Tour Manage. 2017;63:161–171.
- Lester AW, Moffat SD, Wiener JM, Barnes CA, Wolbers T. The aging navigational system. Neuron. 2017;95(5):1019–1035.
- Rowland DC, Roudi Y, Moser MB, Moser EI. Ten years of grid cells. Annu Rev Neurosci. 2016;39:19–40.
- Stangl M, et al. Compromised grid-cell-like representations in old age as a key mechanism to explain age-related navigational deficits. Curr Biol. 2018;28(7):1108–1115.
- Zheng L, Gao Z, McAvan AS, Isham EA, Ekstrom AD. Hippocampal contributions to novel spatial learning are both age-related and age-invariant. Proc Natl Acad Sci U S A. 2024;121(3):e2307884120.
- Berron D, et al. Age-related functional changes in domain-specific pathways of the hippocampus. Proc Natl Acad Sci U S A. 2018;115(11):E2595–E2604.
- Koen JD, Rugg MD. Neural dedifferentiation in the aging brain. Trends Cogn Sci. 2019;23(7):547–559.
- Colombo D, et al. Egocentric and allocentric spatial reference frames in aging: a systematic review. Neurosci Biobehav Rev. 2017;80:605–621.
- Wang L, Wang X. Design mechanisms of airport visual guidance systems on passenger wayfinding performance: evidence from causal machine learning and a moderated mediation approach. J Air Transp Manage. 2024;115:102521.
- Churchill A, Dada E, de Barros AG, Wirasinghe SC. Quantifying and validating measures of airport terminal wayfinding. Journal of Air Transport Management. 2008;14(3):151–158.
- Jones PR, Somoskeöy T, Chow-Wing-Bom H, Crabb DP. Seeing other perspectives: evaluating the use of virtual and augmented reality to simulate visual impairments (OpenVisSim). npj Digital Medicine. 2020;3(1):32.
- Cabeza R, et al. Maintenance, reserve and compensation: the cognitive neuroscience of healthy ageing. Nat Rev Neurosci. 2018;19(11):701–710.
- Ringer RV, Throneburg Z, Johnson AP, Kramer AF, Loschky LC. Impairing the useful field of view in natural scenes: tunnel vision versus general interference. J Vis. 2016;16(2):1167.
- Legge GE, Chung STL. Low vision and reading: what makes a letter legible? Vision Res. 2016;129:117–130.
- Theeuwes J. Goal-driven, stimulus-driven, and history-driven selection. Curr Opin Psychol. 2019;29:97–101.
- Cowan N. The many faces of working memory and short-term storage. Psychon Bull Rev. 2017;24(4):1158–1170.
- Spiers HJ, Gilbert SJ. Solving the detour problem in navigation: a model of prefrontal and hippocampal interactions. Front Hum Neurosci. 2015;9:125.
- Rehman U, Cao S. Augmented-reality-based indoor navigation: a comparative analysis of handheld devices versus Google Glass. IEEE Trans Hum-Mach Syst. 2017;47(1):140–151.