KBruce Newbold (p), Manaf Zargoush, Somayeh Ghazalbash

Understanding the potential risks associated with residential mobility, the local movement of individuals and households between residences, is essential for effective public policy and planning. This study explores the use of machine learning (ML) techniques to predict residential mobility among mid- and late- life Canadians three years in advance. To enhance the interpretability of our predictive models, we applied various explainable artificial intelligence (XAI) techniques to identify the primary independent drivers of mobility risk at both global (population-specific) and local (individual-specific) levels. Our approach achieved robust predictive discrimination (AUROC = 0.68) while offering transparent, data-driven insights into the complex decision-making processes underlying relocation. The local and global XAI techniques offered superior explanatory power, indicating that although regional disparities exist, inter-provincial differences in mobility are almost compositional, driven by variations in individual characteristics rather than the province alone. Individual-level analysis further showed that the impact of provincial context can change from a risk factor to a protective factor, or vice versa, depending on an individual’s specific profile.
The analyses also uncovered “silent high-risk profiles”, where mobility is driven by the cumulative psychosocial vulnerabilities, such as social isolation, lack of community engagement, or depressive symptoms, even when individuals report globally protective structural stability factors, such as satisfaction with home and life. Our age-stratified analysis indicates that mobility among the younger-old is primarily influenced by housing and geographic context, along with mental health and socio-demographic factors. In contrast, mobility among the older-old adults is characterized by increasing care needs, indicators of frailty, functional limitations, and levels of social capital and community engagement. By integrating predictive analytics with interpretable ML, this study proposes a practical framework for identifying individuals at heightened risk of relocation, thereby promoting more proactive, data-driven, and person-centred
approaches to regional planning.

Dr. Tiziano Pavanini (p), Dr. Federica Maria Rossi

The progressive ageing of the global population has led to a growing interest in the mobility and transport needs of older adults within academic research (de Magalhães, 2025). Since mobility and social interaction are key determinants of psychological well-being in later life, transport planning should increasingly focus on services and environments that align with elderly people’s travel behaviours and capabilities (Mariotti et al., 2021). Empirical studies confirm that maintaining an active lifestyle is associated with higher levels of well-being among older adults compared to more sedentary individuals (Thaithatkul et al., 2022). A substantial body of literature has investigated the factors shaping elderly users’ satisfaction with local public transport (LPT). Service-related aspects such as punctuality, driver behaviour, complaint management, and the provision of audio information at stops have been identified as particularly relevant (Lan et al., 2022). In addition, several studies underline the importance of the surrounding walking environment, highlighting barriers such as poor sidewalk conditions (Močnik et al., 2022), the presence of roadworks along access routes to stops (Sun and Lau, 2021), and insufficient last-mile connectivity between homes and public transport services (Jahangir et al., 2022). Despite these contributions, the emotional and psychological dimensions of public transport use among older adults remain largely overlooked. This paper seeks to address this gap by exploring the relationship between LPT, mobility constraints, and life satisfaction. To this end, a questionnaire survey was administered to elderly residents living in four selected neighbourhoods in Milan and Genoa, with the objective of identifying perceived obstacles to public transport use and assessing their impact on overall well-being. The analysis combines descriptive statistics with non-linear econometric models. The expected results aim to provide policymakers and transport authorities with more detailed and targeted evidence, encouraging a shift away from aggregate approaches to public transport planning. Recognising the heterogeneity of mobility needs across age groups is particularly crucial for older adults, who require improved access to LPT and urban services in order to support active ageing and social inclusion.

Ms Youhyun Kim (p), Prof Donghyun Kim

South Korea is experiencing one of the most rapid demographic transitions globally, characterized by accelerated population aging and persistently low fertility rates. As a result, the proportion of older adults is increasing sharply, particularly in rural and shrinking regions that are transitioning into super-aged societies. Alongside these demographic shifts, a growing body of research has documented elevated risks of depression, social isolation, and poor mental health outcomes among older populations. At the same time, strong capital-area concentration has intensified spatial inequalities in access to services, infrastructure, and environmental resources, further shaping regional health disparities. Within this context, nature environments have been consistently shown to produce positive mental health outcomes. Importantly, the concept of the equigenesis effect suggests that such benefits may be disproportionately concentrated among socioeconomically vulnerable populations, thereby reducing health inequalities rather than simply improving average health levels.
This study aims to examine whether the equigenesis effect of nature environments on mental health operates differently between super-aged regions and non-super-aged regions in South Korea. Specifically, we investigate whether nature environments mitigate mental health inequalities among older adults through distinct mechanisms depending on regional demographic and spatial contexts. Using nationally representative regional data, we apply quantile regression models to assess how nature environment exposure influences different points of the mental health distribution, with a focus on lower quantiles representing more vulnerable populations.
We hypothesize that in super-aged regions, nature environments primarily reduce mental health inequality by enhancing social connectedness and compensating for service accessibility limitations, whereas in non-super-aged or metropolitan regions, their effects may operate more through environmental stress reduction pathways such as heat mitigation, air quality improvement, and opportunities for physical activity. We expect to find stronger equigenesis effects at lower mental health quantiles in super-aged regions, indicating greater benefits among the most vulnerable older adults. By integrating demographic transition, spatial inequality, and environmental health perspectives, this study provides policy-relevant evidence for designing targeted nature-based interventions to reduce mental health inequalities in rapidly aging societies.

Associate Prof. Yuji Takeguchi, Dr. Ayako Suzuki, Prof. Soushi Suzuki (p)

Phrases “increase in unmarried rate” and “high mortality society” have got attention in recent years in Japan. These trends mean that the number of older individuals living in single-person households will grow and will be increased vulnerability from the viewpoint of lose means of transportation in the future. These situations indicate that Japan is moving from a “super-aging society” to a “super-aging and solitary society”. Even under such difficult circumstances, it is necessary to implement sustainable regional development. Based on these backgrounds, this study especially focused on the issue of "means of transportation" which is already one of a major issue in the super-aging society among the issues predicted to arise in a super-aging and solitary society. This study clarifies the issues of “super-aging and solitary society” based on the analysis that subjective well-being (SWB) as an objective variable and focused on household structure and transportation attributes. The subject of this study set Abashiri city in Hokkaido prefecture where is a region with challenges related to aging and population decline. This study used datasets based on questionnaire survey for Abashiri city inhabitants in 2022. The analysis results revealed that the SWB of older individuals living in single-person households is statistically significantly lower than couple household and household with children. The analysis result also revealed that the SWB of lose means of transportation household is statistically significantly lower than other households. Moreover, this study utilized quantification theory type I to set SWB as the objective variable and to analyzed the category scores of household composition and transportation attributes. Based on these category scores, this study carried out a simulation analysis on SWB. In addition, this study conducted a future simulation analysis of total amount of SWB in Abashiri city based on the category scores as well as future estimated results of the lifetime unmarried rate and mortality rate. The simulation analysis result revealed that decline SWB overall Abashiri city caused by an increase in older individuals living in single-person households with poor means of transportation. On the other hand, it was suggested that in cases of public transportation and other services were increased, it may be possible to maintain or increase the SWB of overall Abashiri city.

Ms Chaeyeon Lee (p), Professor Donghyun Kim

As population ageing intensifies, age-friendly cities and Ageing-in-Place (AIP) have become central to regional policy and urban planning. Yet treating AIP as a simple goal of preventing migration risks overlooks how older adults’ migration and non-migration reflect needs, choices, constraints, and place-based differences in social connectedness, service accessibility, and heat risk. This study classifies 229 si-gun-gu in South Korea using the composition of older adults’ migration types as a core structural indicator and identifies how regional functions evolve over time. Using the 20 percent sample of the Population and Housing Census, we compute, for each si-gun-gu, the shares of older adults by five migration types derived from residential histories across four time points: non-migration, initial migration, return to birthplace, return to a non-birthplace previously lived in, and onward migration. We integrate these shares with three policy-relevant context dimensions: Economic, Environmental and Built (including heatwave days and tropical-night days), and Social and Service. A Self-Organizing Map (SOM) is employed to classify multi-dimensional regional profiles and trace cluster transitions as regional trajectories. To support the interpretation of current context combinations, we additionally construct a 2020 RGB indicator that maps the three dimensions to R, G, and B channels. Results show five stable regional types and four representative trajectories: Stable AIP Lock-in, Urban Absorber with Initial Migration, Return Migration Reinforcement, and Circular or Transient Gateway. The types differ not only in migration-type composition but also in how heat exposure and service foundations co-occur. The RGB profiles highlight, for example, that Stable AIP Lock-in regions tend to combine relatively weak Social and Service foundations with relatively high Environmental and Built contributions, whereas Urban Absorbers exhibit strong Economic and Social and Service capacity, with heat exposure often coexisting. These findings imply differentiated policy priorities: integrated neighborhood-based interventions for lock-in regions, proactive heat adaptation and service capacity for absorbers, strengthened re-settlement absorptive capacity for return regions, and continuity-oriented service design for gateway regions.