BackgroundCognitive decline in the aging population presents an unprecedented challenge worldwide. Recent research has shown the potential of cognitive training programs to mitigate cognitive decline. However, these interventions require sustained adherence to be effective, which can be challenging. ObjectiveIn this study, we aim to enhance the accuracy of predicting adherence patterns in cognitive training programs for older adults, with the goal of developing personalized support systems that promote adherence and improve cognitive outcomes. MethodsA major challenge in developing deep neural networks for predicting adherence patterns is the limited availability of individual participants’ training data. Although domain adaptation techniques can address this issue by leveraging training data from other clinical studies, our research considers a more practical scenario where the use of such data from other studies is restricted due to privacy and confidentiality concerns. Therefore, we used source-free domain adaptation (SFDA), which uses models trained on other cognitive studies without requiring access to the corresponding datasets. To the best of our knowledge, this is the first effort to use SFDA to predict older adults’ daily adherence to cognitive training programs. ResultsUsing data from 3 previously conducted cognitive training intervention studies, our results demonstrated the efficacy of deep learning models combined with SFDA to accurately predict adherence lapses while addressing data privacy concerns. ConclusionsOur findings indicate that deep learning and SFDA techniques can be useful in the development of adherence support systems for computerized cognitive training, aimed at improving the health and well-being of older adults.

Caregivers play a critical role in supporting older adults, yet they often face significant emotional, physical, and logistical challenges. This symposium presents innovative research on technology-based interventions designed to empower caregivers and optimize care delivery. Dr. Sara Czaja will provide an overview of digital solutions for family caregivers, discussing mobile applications, web-based platforms, and sensing systems and share findings on the feasibility, acceptability, and efficacy of these tools in improving caregiver outcomes. Taekyung Kim MS will present research on the adoption of AI-assisted care technologies among formal care providers, highlighting key factors such as social influence and effort expectancy that shape technology acceptance and use in real-world caregiving environments. Dr. Felipe Jain will discuss findings from an 8-week randomized controlled trial examining the impact of mentalizing imagery therapy (MIT) delivered via a mobile app on daily happiness and stress in family caregivers of people living with dementia. Dr. Xin Yao Lin will share results from a scoping review of app-based interventions for caregivers, identifying common design features, benefits, and research gaps that must be addressed to improve intervention effectiveness. Finally, Dr. Katherine Carroll Britt will present qualitative findings from caregivers and dementia care experts to inform the development of an AI chatbot designed to support cognitive care planning for individuals living with dementia and their care partners. By showcasing these diverse approaches, this symposium will explore the potential of digital innovations to create smarter, more connected caregiving solutions that enhance well-being and care quality for both caregivers and older adults.

Healthy aging requires acquiring new functional skills for adaptation in a dynamic environment. Cognitive interventions with older adults have largely focused on improving broad cognitive abilities, aiming for transfer to functional effects. By contrast, interventions focusing directly on acquiring new functional skills can address current real-world issues, including the need for reskilling and reducing the digital divide, especially for underserved communities. In doing so, we may better understand how aspects of age-related learning and cognitive and functional decline may be due to suboptimal learning circumstances rather than senescence. In this opinion paper, we highlight key aspects for designing long-lasting, real-world interventions to improve functional skills, and potentially transfer to cognitive effects, for older adults. This approach could help build more inclusive theories of cognitive aging, while progressing the field toward developing more effective and useful interventions.

Advancements in monitoring technologies offer new opportunities to enhance the health, independence, and well-being of older adults. This symposium presents innovative, non-invasive solutions for the early detection of functional decline, fall prevention, and support for aging in place. Dr. Megan Huisingh-Scheetz will discuss how hip accelerometry and machine learning models can detect and forecast frailty decline in older adults, enabling proactive interventions. Dr. Jeannette Mahoney will introduce CatchU... Before You Fall, a multisensory digital health tool designed to predict fall risk—particularly in individuals with preclinical Alzheimer’s disease—while promoting independence and facilitating provider-initiated falls counseling. Dr. Clara Berridge will explore the Let’s Talk Tech (LTT) decision aid, a web-based tool that helps individuals with memory loss and their caregivers make informed decisions about digital health technologies, highlighting its potential value to both family members and providers by supporting the sharing of technology preferences beyond the care dyad. Dr. Elinor Schoenfeld will present on the development and deployment of contactless sensors for Monitoring Vital Signs and Movement, advancing real-time health monitoring to support aging in place. Finally, Dr. Jane Chung will introduce a Wi-Fi Sensing-Based Deep Learning Solution that recognizes daily activities in older adults, offering insights into functional decline and cognitive health. By showcasing these cutting-edge approaches, this symposium highlights the transformative potential of monitoring technologies in aging research, emphasizing user-centered and scalable solutions to support older adults’ health and independence.

PURPOSE:Older adults with a cognitive impairment may be challenged by the demands associated with technology systems used to support everyday activities. We investigated technology attitudes, proficiency, and usage across the domains of health, social, transportation, leisure, and domestic activities among older adults with mild cognitive impairment, traumatic brain injury, and post-stroke cognitive impairment. We examined whether age, gender, health, cognition, or technology attitudes (comfort, interest, efficacy) predicted technology proficiency and usage patterns. All variables were measured with validated questionnaires. MATERIALS AND METHODS:Participants (N = 163, age range 60-93) were part of the Everyday Needs Assessment for Cognitive Tasks (ENACT) study. RESULTS:The participants were largely proficient in using technologies and had generally positive attitudes toward technology. Usage patterns varied across domains, with participants most engaged in technology uses to support social and domestic activities, followed by health activities. Technology was used least frequently to support transportation and leisure activities. Findings highlighted the complex interplay of demographic factors, cognition, and attitudes towards technologies in shaping older adults with cognitive impairments' adoption and use of technology across the various domains. Older age was associated with lower technology use, whereas positive attitudes towards technology (interest, comfort, and efficacy) were associated with higher use. Gender differences were evident in social, domestic, and leisure technology domains. CONCLUSION:These findings underscore that older adults with a cognitive impairment can use and are receptive towards technology. The findings provide valuable insights for tailoring interventions to meet the needs and preferences of older adults with a cognitive impairment.

As telehealth continues to evolve, it holds immense potential to support the health and well-being of older adults. However, barriers related to digital literacy, accessibility, and user preferences must be addressed to ensure equitable access and effectiveness. This symposium brings together experts to examine the challenges and opportunities of telehealth in diverse aging populations, with a focus on digital inclusion, usability, and the future of remote healthcare solutions. Dr. Sohyun Kim will explore the role of video chat interventions in Long-Term Services and Supports (LTSS) settings, highlighting their potential to enhance social engagement and family relationships for persons living with dementia while addressing technological and logistical barriers. Sungjae Hong, MS, MPH, will present an exploratory study on telehealth adoption among Korean older adult immigrants, revealing key factors influencing digital healthcare use and the role of healthcare providers in shaping adoption. Dr. Paul P. Freddolino will introduce a comprehensive cost-benefit model assessing the impact of telehealth awareness and digital literacy training for direct care workers, examining the economic, social, and functional value of these initiatives for multiple stakeholders. Dr. Renato Ferreira Leitao Azevedo will discuss findings from an online feasibility study investigating older adults’ preferences for different telehealth communication mediums, evaluating their effectiveness in enhancing comprehension, emotional engagement, and risk perception. By addressing critical issues in telehealth adoption and design, this symposium will provide valuable insights into optimizing remote care for aging populations.

The likelihood of developing a cognitive impairment (CI) increases with age. Older adults with CI experience difficulties performing a range of everyday activities and are at risk for social isolation. Technology-based interventions have the potential to improve everyday functioning. Artificial intelligence capabilities can be harnessed to tailor solutions to the unique and changing needs of older adults with CI. Most prior efforts focused on cognitive training or rehabilitation but have not included other aspects of functioning such as everyday activities or social engagement. Our goal is to develop and evaluate an innovative intelligent adaptive software system to support personal activities and reinforce cognition (SPARC). The system will be designed to adapt to the needs and abilities of the user with CI, following the CREATE model of user-centered design. We started with interviews with subject matter experts to assess older adults’ needs for activity support in their homes and identify potential technology solutions. We interviewed both clinicians and technical experts to obtain guidance for the initial system design. In this presentation, we will present results of focus group interviews with older adults who have CI wherein we present prototype images of SPARC and explore their perceptions of usefulness, ease of use, preferences for instructional support, and ideas for specific content that would be of interest to them. This multi-faceted needs assessment approach will guide design of SPARC to support everyday activities, primarily in the home environment. Our findings inform design of other technologies and interventions to support older adults with CI.

Existing and emerging technologies offer powerful tools to foster social connection and enhance well-being among older adults. From digital communication platforms to personalized digital health tools, these innovations can help reduce loneliness, support mental health, and promote meaningful engagement. As the older population grows and diversifies, leveraging technology to meet their social and emotional needs has become both an opportunity and a necessity. This symposium brings together researchers advancing our understanding of how technology can support connection and quality of life in later life. Dr. Kalon Sou will present findings from a study evaluating older adults’ experiences with an age-friendly, AI-powered chatbot, identifying key factors—such as empathy, sociability, and agency—that shape willingness to use social chatbots. Nicole Memmer will explore how different types of Internet use, especially for social purposes, relate to older adults’ offline participation, with digital skills and local context influencing the benefits gained. Dr. Paul P. Freddolino will highlight a multi-year initiative to increase digital health literacy among hard-to-reach older adults using community-based strategies, emphasizing the role of trust and early outcomes related to technology engagement. Dr. Ines Simbrig will examine how aging-in-place technologies affect older adults’ perceptions of age-related change, revealing a complex trade-off between feelings of safety and potentially negative shifts in self-perception. Finally, Dr. Yong Kyung Choi will introduce the Adult Wellbeing CheckUp, a web-based platform that offers multi-domain assessments and personalized recommendations to help older adults and care partners improve health, engagement, and quality of life.

Older adults are increasingly using the Internet, laptops, smartphones, and other digital devices to access healthcare services and information. Additionally, wearable technologies offer the potential to continuously monitor health data, which could be used to detect and address age-related health issues. However, the adoption of these technologies remains challenging due to difficulties in using the technology and concerns about privacy. In this study, we build upon the results of Fowe and Boot’s (2022) survey study, which examined older adults’ attitudes toward the use of wearable and mobile technologies for predicting cognitive decline, supporting healthy behaviors, and collecting self-reported health data. The current study aimed to qualitatively explore the barriers and facilitators of wearable technology use. Data were collected through four focus group sessions with older adults. Thematic analysis revealed five main themes: (1) privacy concerns, (2) digital “nagging,” (3) accuracy and reliability concerns, (4) proficiency challenges, and (5) a strong interest in monitoring personal health. These findings underscore the importance of developing resources that enhance older adults’ skills, confidence, and self-efficacy in using wearable technologies safely and effectively.

Older adults represent a remarkably heterogeneous population that is diverse in many dimensions, including cognition. Cognitive diversity includes normative age-related changes in cognition, mild cognitive impairment (MCI), and more severe impairments such as dementia. The causes and nature of these cognitive impairments (CIs) are varied, extending beyond MCI and Alzheimer’s disease and related dementias to encompass stroke, Parkinson’s disease, and traumatic brain injury. When developing cognitive support strategies, it is crucial to account for the wide range of cognitive diversity and the differing cognitive patterns and trajectories within older adult populations. In this regard, intersectionality is an important consideration given the increased diversity of the older adult population and the existence of health disparities among underrepresented groups. This paper focuses on demonstrating how technology applications can provide cognitive support to older adults across the continuum of cognition. Examples of ongoing research in this area are provided from the authors’ Enhancing Neurocognitive Health, Abilities, Networks, & Community Engagement (ENHANCE) Center, which focuses on developing and evaluating technology support solutions for older adults with a CI, and the Center for Research and Education on Aging and Technology Enhancement (CREATE), which focuses on older adults and their interactions with technology systems and ensuring that the benefits of technology can be realized by older adults. We describe our research framework, based on a user-centered design approach, that emphasizes user characteristics, environmental contexts, and the importance of including diverse user groups in design activities.

As of March 2024, the NIH annual budget for Alzheimer’s disease and related dementias (ADRD) research is set to nearly $3.8 billion, after a $100 million increase (Alzheimer’s Association, 2025). This significant increase in funding for cognitive aging research is focused on identifying the causes of age-related cognitive decline and advancing methods and technologies to improve cognitive function in older adults. Whether grant- or industry-funded, clinical trial or observational, two of the most challenging aspects of research are recruitment and retention, which are integral to maintaining validity and generalizability of study results. Computerized cognitive training is one of the most promising non-pharmaceutical interventions for ADRD prevention, so understanding participants’ reasons for engaging in this type of research is crucial to the next stages of implementation research. This symposium highlights research teams who have assessed participant motivation to engage in computerized cognitive training in order understand how motivations for research engagement differ between underrepresented groups (paper 1), how motivation relates to both participant enrollment and drop-out (paper 2), and how to use participant motivation to develop tailored supportive communications in randomized controlled trials to encourage adherence and retention (paper 3). The symposium presents refinement in research examining participant motivation with the studies representing novel contributions given their sample size (paper 1; n = 5,721), link to both a theoretical framework and information on participant withdrawal (paper 2), and a manner in which participant motivation can be strategically applied to increase study retention (paper 3).

The Adherence Promotion with Person-centered Technology (APPT) trial addresses adherence challenges in cognitive assessment and training by leveraging smart, tailored technologies. Early detection of cognitive decline is critical for advancing scientific understanding, improving clinical trial recruitment, and enabling early intervention. However, adherence to frequent at-home testing and behavioral interventions remains a significant barrier. Designing an effective adherence support system requires understanding the motivations that drive older adults to participate in research and engage with interventions. To initially explore these motivations, we surveyed 472 older adults about their reasons for participating in research. brain health advocates, research helpers, fun seekers, and multiple motivation enthusiasts. Individual differences—including age, employment status, and cognitive difficulties—shaped motivation profiles, informing the development of tailored engagement strategies. Building on this foundation, we then pilot-tested a just-in-time adherence support system using personalized text messages to encourage engagement in a 10-day cognitive training study. Forty-three older adults received messages aligned with their stated participation motivations. Matched messages were rated as significantly more effective in promoting adherence than mismatched ones. Participants preferred messages that were personalized and formal, reinforcing the importance of message tailoring. With these insights, the first randomized controlled trial (N = 199) evaluating the APPT system for home-based cognitive training has been completed, and a second trial, examining home-based assessment, is now underway. Findings will inform future refinements of dynamic, machine-learning-driven adherence support, with broad applications for cognitive training, telehealth, and other health-related interventions.

This cross-sectional study explores the reliability and validity of a newly developed 15-item Medicare proficiency questionnaire (MPQ) across a mixed group of participants enrolled and unenrolled in Medicare. The MPQ was designed to assess beneficiary knowledge across a variety of Medicare topics and was developed by combining questions selected from the 2003 Medicare Current Beneficiary Survey and updating it with researcher-generated Medicare Part D questions. During the month of February in 2024, participants enrolled and unenrolled in Medicare, which were recruited on Prolific, completed online surveys which included the MPQ and demographic questions. We found that the MPQ has adequate internal consistency reliability with Cronbach's alpha = .73 across participants enrolled and unenrolled in Medicare as well as adequate validity, as demonstrated by positive relationships of MPQ scores to education level and to enrollment status (enrollees scoring higher).

Large language models (LLMs) have strong potential as decision aids for older adults seeking information related to their health, well-being, and independence. Langston et al. (2025) found that off-the-shelf LLMs, such as Bard (now Gemini) and ChatGPT, provide quick and generally accurate responses to questions relevant to older adults. However, these responses are often long and complex, which may create difficulties due to age-related changes in information processing, particularly for those with mild cognitive impairment (MCI). We tested whether simple prompt modifications, such as stating that the user is an older adult, has MCI, or is seeking a simple explanation, could improve the readability of responses. Three independent raters evaluated a subset responses by ChatGPT4o to questions from Langston et al. (2025), scoring responses on readability metrics. Prompts indicating the user had MCI resulted in slightly simpler language (1.75 vs. 1.80 syllables per word, p < .05), but had little effect on other measures. In contrast, prompts that explicitly requested simple answers led to significantly improved readability, including shorter responses (209 vs. 279 words, p < .05) and fewer syllables per word (1.59 vs. 1.80, p < .001). Sentence length increased (10.87 vs. 9.15 words, p < .001), perhaps because responses contained fewer short filler sentences, more information-dense phrasing, and greater context. These results suggest that prompt engineering can improve LLM usability for older adults, but requests must be direct. LLMs do not reliably adapt based only on descriptions of the user.

Home-based cognitive training programs delivered via computers and tablets hold promise as cost-effective, population-level interventions to prevent or mitigate age-related cognitive decline. However, adherence to such programs is often low. Using message-tailoring techniques and an adaptive algorithm, we developed a person-centered reminder smart system that delivers motivational messages at times when participants are predicted to be available for training activities. This paper presents the background, study design, methodology, and baseline data for a randomized controlled trial examining the system's efficacy in supporting adherence to cognitive training. A total of 199 cognitively normal, community-dwelling older adults aged 62 to 88 were randomly assigned (1:1) to either the smart reminder or the control condition. Participants were instructed to engage in training activities for 30 min per day, five days per week, for 18 consecutive weeks. Those in the smart reminder condition received personalized messages, delivered at optimized times, and with targeted content, while those in the control condition received generic messages at a fixed time. Adherence rate will be the primary outcome measure, calculated and compared across conditions. Findings from this study will have implications not only for adherence support in cognitive training but also for broader applications of technology-mediated smart reminder systems, including physical exercise, nutrition, medication management, telehealth, and social connectivity. By enhancing intervention engagement, these systems have the potential to improve the health and well-being of older adults on a large scale.

In 2022, ∼ 99% of American adults age 65+ yr were enrolled in Medicare, a complex, difficult to use insurance system. We describe a multi-pronged approach to assessing the potential value of AI to support Medicare decision-making processes that includes querying subject matter experts who provide Medicare advice, assessing Medicare users’ knowledge, preferences, and abilities through interviews and while interacting with knowledge sources, and evaluating existing AI tools for accuracy, reliability, and conciseness. In this presentation we present findings about the accuracy and reliability of digital assistants in answering Medicare questions, a new tool for assessing Medicare knowledge, and a study to assess preferences and performance with Medicare information sources. Generative AI (ChatGPT, Bard) were highly accurate (>90%) and reliable as well as superior to the average Medicare beneficiary, and much superior to digital home assistants (Alexa and Google Home Assistant). The Medicare Proficiency Questionnaire proved to be a short, reliable, and valid scale of older adults’ Medicare knowledge, differentiating between Medicare enrollees and non-enrollees. Prior access to the Medicare website and the Medicare and You Handbook mediated the relationship between knowledge and education level as well as knowledge and Medicare enrollment status, suggesting that understanding Medicare information sources may be critical for designing AI decision support tools and training. We also describe early results from an observational study with Medicare-enrolled and unenrolled older adults examining preferences for, and performance with, the Medicare.gov website, the Medicare and You Handbook, and Gemini (a Large Language Model).

Smartwatches have potential to provide support for prospective memory (PM), the ability to remember and carry out an intention in the future. How older adults (OAs), particularly those with cognitive impairment (CI), might interact with smartwatches is undetermined. This study aimed to understand the usability of smartwatches among cognitively diverse OAs and the potential for smartwatches to serve as reminder aids for this population. Participants were 58 OAs (age 60+) with and without CI. After using a smartwatch as a reminder aid for 10 days, participants gave low usability ratings overall, and the smartwatches did not aid performance of a daily PM task. Perceived usefulness of the smartwatches was associated with subjective memory, suggesting that perceived memory challenges may play an important role in smartwatch adoption. Results can inform the development of future efficacy tests and interventions involving smartwatches.

Social and cognitive engagement are critical elements of successful aging. Yet many older adults, especially those who live alone or have a chronic condition or disability, lack social connectivity, experience loneliness, and lack opportunities for cognitive engagement. Virtual Reality (VR) applications can play a key role in enhancing social and cognitive engagement and decreasing loneliness among older adults. VR provides an immersive and interactive experience, which enables users to feel “present” in an environment, and have the ability to interact solely or with others in a virtual space or activity (e.g., museum, gardening). The presentation will focus on data from a cross-site study that examined perceptions of value, enjoyments and usability (comfort, safety) of VR applications as well as preferences for VR apps among a sample of 48 of community dwelling older adults aged 65 years and older. Sixteen pre-selected apps that focused on social, cognitive and productive engagement were examined. Each participants received a brief training on the use of VR and were exposed to eight apps either interactively or in a video format (apps were counter-balanced across participants). They rated each app after exposure, completed the Attitudes Toward VR Scale, A Virtual Reality Activity Inventory, and an Opinion Interview. The findings indicated that the older adults rated the experience of VR as valuable and highly enjoyable. There were also low ratings of discomfort and high ratings of usability. Most participants indicated that they would like to engage with VR in the future.

Assistive technology (AT) holds great potential to increase independence for older adults, specifically persons with mild cognitive impairment (PwMCI). MCI affects approximately 20% of individuals aged 50 and above. Despite this prevalence, the existing literature does not include robust information on their attitudes, perceived needs, or desires regarding AT. We conducted six focus groups across two sites (University of Illinois Urbana-Champaign and Weill Cornell Medicine). There were 20 total participants: 6 male, 14 female, aged 60-85, with a TICS-M score between 22-37, indicative of some level of cognitive impairment. The goal of the group discussions was to develop recommendations for an AT system, called SPARC (Supporting Personal Activities and Reinforcing Cognition). The SPARC system offers features designed to aid PwMCI in completing complex activities of daily living, support their cognitive health, increase social engagement, and provide educational opportunities. Discussions focused on content, ease of use, relevance, and integration with other tools. Participants responded favorably to the “all-in-one” nature of the system. They showed some hesitation about potential effort and determination needed to adopt the system. The activity support and education features were well-received. Participants discussed potential barriers to adoption for the social and reminiscence features, most common being relevance to them, ease of use, privacy concerns, and integration with current approaches. These findings underscore the importance of assessing PwMCI’s attitudes and opinions during the early stages of the design process, which will impact the adoption and success of the SPARC or other systems intended to support their needs.

The rapid advancement of artificial intelligence (AI) is reshaping society, offering unprecedented opportunities for innovation. In aging and gerontological care, AI holds transformative potential to enhance research, improve clinical decision-making, and support the health and well-being of older adults. This symposium highlights cutting-edge work applying AI to challenges in aging populations. Dr. Ian B. Kwok will introduce Clinical SmartReporter™, an AI-powered tool designed to enhance serious illness communication through real-time, patient-facing transcripts of family meetings. Yijung Kim will present an analysis of social determinants of health among homebound older adults, demonstrating how large language models like GPT-3.5 can support—but not fully automate—social needs identification. Dr. Amy M. Schuster will share survey findings showing that older adults’ attitudes toward AI significantly influence their likelihood of using it in healthcare, with implications for improving acceptance through targeted interventions. Taekyung Kim will discuss age-related stereotypes in GPT-4o-generated content, revealing lower perceived competence in descriptions of older adults and calling for strategies to reduce ageism in AI systems. Dr. Bo Xie will present a multi-agent large language model framework for streamlining systematic reviews of AI interventions for older adults, showing how hybrid human–AI collaboration can accelerate evidence synthesis. Collectively, these talks explore how AI can be thoughtfully integrated into aging research and care to promote equity, effectiveness, and inclusion. Technology and Aging Interest Group Sponsored Symposium