Towards socially-just and technically robust data infrastructure for AI applications in the field of dementia prevention and health promotion

Dementia is an escalating global public health challenge, which has significant implications for people living with dementia, their families, communities, and health systems (WHO, 2017). In both Korea and Sweden, the prevalence of dementia among individuals aged 65 and older is rising (Shon & Yoon, 2021; OECD/European Observatory on Health Systems and Policies, 2019), raising urgent concerns about the sustainability of care systems and the financial burden on societies. While much attention has been given to medical and individual risk factors, growing evidence highlights the critical role of environmental determinants in dementia risk. For example, living near major roadways has been linked to a roughly 10% higher risk of developing dementia, while exposure to fine particulate matter from vehicle emissions and industrial pollution has been associated with a 9% increase in risk (Da et al., 2024). These findings underscore the urgent need to rethink and transform our living environments to be more dementia friendly. Addressing these environmental risks is not only essential for promoting
health but also carries significant implications for urban planning, public health policy, and the creation of more sustainable and age-inclusive communities.

Emerging evidence highlights the qualitative and quantitative impact of social and physical environmental factors on cognitive health. For instance, Röhr et al. (2022) identified several environmental factors perceived to mitigate dementia risk, including (i) social
participation and inclusion (e.g., intergenerational connections, neighborhood assistance, and digital literacy), (ii) proximity and accessibility (e.g., access to healthcare, mobility, and affordable services), and (iii) local recreation and well-being (e.g., urban greenery, climate
change adaptation, and outdoor physical activity). Additionally, growing quantitative research links the built environment features to dementia prevention strategies. Despite growing recognition of the role of environmental factors in dementia risk, dementia prevention research remains fragmented and faces significant challenges. While various interventions have been developed to reduce dementia risk, most focus either on individual-level lifestyle changes—such as smoking cessation, physical activity, and diet—or on environmental modifications, without integrating both dimensions into a cohesive, person-environment fit approach. Existing environmental interventions tend to focus on single domains rather than adopting a multidimensional approach (Mangialasche et al., 2012). Moreover, few interventions take a life-course approach to dementia prevention, due to challenges in long-term follow-up and monitoring (Killin et al., 2016). These gaps underscore an urgent need for integrative interventions that combine individual and environmental factors to create supportive, dementia-friendly environments for individuals across the life course.

Artificial intelligence (AI) has the potential to revolutionize dementia prevention by addressing the current gaps in research and intervention design. Unlike traditional approaches that focus either on individual lifestyle changes or environmental modifications in isolation, AI
can facilitate a more integrative, personalized, and scalable strategy by leveraging data-driven insights.

Our activities aim to
• Synthesize the best available evidence on opportunities, challenges, and barriers in utilizing data for AI applications in dementia prevention and health promotion, and identify critical knowledge gaps which could inform future research grant applications.
• Discuss the critical issues related to data with stakeholders based in Sweden and Korea, which could support long-term partnership.
• Share the knowledge gained and increase the visibility of our group work.