Abstract
Commercial wearable technologies are becoming increasingly prevalent and offer the potential for monitoring physiological changes over time, particularly in older adults with Mild Cognitive Impairment (MCI). While literature suggests the utility of such devices, little evidence exists of their successful deployment among people with MCI. This feasibility study aimed to establish research guidelines, optimize participant training, and identify data collection challenges when collecting longitudinal physiological data using the Oura Ring. Our results indicated that adequate researcher support was crucial for study success and participant engagement. This involves including care partners of participants with MCI and providing appropriate technological training to participants. Care partner support reduced device troubleshooting instances and emotionally encouraged participants with MCI. Lastly, maintaining regular researcher-participant communication fostered rapport and ensured ongoing participant consent and data quality. This research lays the groundwork for future health monitoring studies by emphasizing the importance of tailored support, communication, and researcher involvement.
Keywords
Introduction
Commercially available wearable technologies are becoming more advanced and increasingly ubiquitous in society. These advancements have led to increased interest in employing these technologies in a research context. The non-intrusive nature of wearable technologies could improve researchers’ ability to monitor the physiological changes of people with Mild Cognitive Impairment (MCI) over time (Ghorbani et al., 2022). Despite the potential benefits that older adults, both with and without MCI, might gain from integrating these technologies into their lives to monitor their health, there is little evidence of them being successfully deployed (Holthe et al., 2018).
This paper covers a sub-study of a larger, longitudinal study aimed at discovering correlations between physiological markers and cognitive changes in older adults with Mild Cognitive Impairment (MCI) that can lead to targeted interventions that might slow the progression of MCI to Alzheimer’s disease (AD). This feasibility sub-study was conducted to determine whether this population would successfully adopt and use a commercially available wearable health monitoring device in daily life and as research study participants. Additionally, this study aimed to determine the best practices for conducting this type of research by identifying some of the main challenges this user group experiences when adopting and using wearable health monitoring devices. Further, this study aimed to determine the resources needed for study infrastructure, participant training requirements, and potential challenges during data collection.
The commercially available technology used in this study was the Oura ™ Ring Gen3, a wearable smart ring that collects sleep, activity, and other wellness measures through three small sensors inside the band (Ōura Health Oy, 2024). The ring uses Bluetooth to connect to a smartphone or tablet application, where the data is presented to the user in a series of graphs, scores, and trends. The decision to use the Oura Ring arose from discussions with subject-matter experts, who indicated that the unobtrusive device, with its relatively long battery life, would be easier to implement for older adults with MCI. Additionally, Oura provides a data-focused application programing interface (API) known as Oura Teams (Ōura Health Oy, 2023). This platform allowed the research team to access and monitor participant data.
The goals of the sub-study are to better understand the following:
1. What factors influence participant satisfaction and engagement with the Oura Ring and its application?
2. How can researchers best support participants, and how can study design be improved?
Method
Approach
This study used an “action research” approach. Action research is an applied research framework that allows for flexible research methods and new strategies to achieve a goal (Antonellis & Berry, 2017). This approach enables researchers to identify issues that hinder data collection as they arise during the study. It places particular emphasis on continuous communication between researchers and participants. There is also a high level of participant involvement in determining what content is needed for training, how frequently participants meet with researchers, and how the study protocol can be improved. This methodology allows researchers to optimize the participant experience, thereby increasing the quality and quantity of data collected. Action research has been used successfully in contexts similar to the current study and in clinical trials (McCormack, 2015). The application of this method in our research was reflected during the monthly check-ins and troubleshooting meetings with participants.
Sample
Participants aged 50 and older were recruited from the Emory Cognitive Empowerment Program (Emory University, 2024). They included program members with a clinical MCI diagnosis and spousal care partners. Recruitment occurred between April and December of 2023. A total of 36 participants (24 members and 12 care partners) enrolled in the study. The average age of participants was 74.75 (SD = 7.95, min = 55, max = 85). Of the participants, 21 identified as male, and 15 as female. The relationship status of participants was married (n = 33), partnered (n = 2), or single (n = 1). Three participants identified as LGBTQ+. Participants also identified as White or Caucasian (n = 18), Black or African American (n = 15), Eastern Indian (n = 1), Caucasian and Lebanese (n = 1), or preferred not to say (n = 1). Of the 36 participants, five withdrew from the study before 6 months of enrollment due to declining health (n = 2) and lack of interest in participating (n = 3). When this paper was written, the average duration of time in the study among all participants was 241 days (SD = 102, min = 34, max = 403 days).
Data Collection and Analysis
Before starting the study, participants were given a set of plastic rings corresponding to the available sizes of the Oura Ring device (sizes 6–13). Participants were instructed to try different ring sizes on various fingers for 1 week. This was done to account for changes in finger size caused by swelling, medication, arthritis, or other reasons unique to older adults. Once an appropriate size was found, researchers provided participants with an Oura Ring device that fit them comfortably. At this time, researchers helped each participant download the Oura Ring application to their smart device and complete the setup.
Participants were instructed to synchronize their Oura rings daily, charge the device at least once every 4 days, and wear the ring throughout the day and night. To synchronize the Oura Ring, the participants were required to open the Oura Ring application on their smart device. The Oura Ring transfers the data collected via Bluetooth to the smart device. These data were transferred wirelessly over the internet to the research platform, Oura Teams, for download by the researchers.
Along with the physiological data collected by the Oura ring, monthly 15-min phone interviews were conducted with each participant. Participants were asked about their experience with the ring over the last month and any troubles they might have experienced. If applicable, they were also asked how much and what type of help or information participants with MCI received from their care partner when using the Oura Ring and the application. These data were thematically analyzed using a grounded theory approach. We followed the steps outlined in Braun and Clarke (2006) by familiarizing ourselves with the data, open-coding the comments inductively to note salient patterns, and categorizing the patterns into collapsed and meaningful themes via a codebook. Additional quantitative analyses were conducted using the physiological data to determine the number of days that data collection was missed per participant. Lastly, the frequency and type of troubleshooting instances that researchers encountered are reported. Based on these results, recommendations for researchers are provided.
Thematic Analysis Results
This analysis identified three themes. Themes one (Physical Usability and Device Comfort) and two (User Engagement) pertain to research question one, and theme three (Technical Support and Training) pertains to research question two.
Theme 1: Physical Usability and Device Comfort
This theme encompasses the design and comfort of wearing the device. It also includes considerations related to how to use and maintain the device. Several subthemes were identified.
Comfort and Fit
Arthritis and swelling were a primary concern for participants when using this device. Despite using the size kit before starting the study, participants often experienced swelling in their fingers, impacting their comfort when taking it on and off or wearing it throughout the day. “I’ve lost weight and my fingers swell more.”
Theme 2: User Engagement and Experience With the Application
This theme focuses on the participant’s interactions with the device and the application, capturing positive and negative experiences. It also includes the benefits and extent to which users engage with the application and their interest in learning more about their health data.
Positive Experiences
Participants communicated numerous positive experiences using the application, including improved and more satisfying sleep quality. Some participants also emphasized that they were self-sufficient when using the application and had developed a routine around it. Participants also noted that the information provided by the application was helpful and that they enjoyed learning from the insights provided. “It encourages me to get better sleep, so I am getting better sleep”.
Usage Difficulties
Participants expressed that remembering how frequently to charge and synchronize the ring involved a learning curve. Additionally, they expressed difficulty navigating the application and finding certain features.
“There is a learning curve with reading the Oura Ring information.” “Sleep data in Oura shows different results than Fitbit.”
Skepticism
Some participants expressed skepticism about the quality and accuracy of the data that the device collected. This was either based on introspective reflection on what the application was telling them or by comparing the results of this device to those of other devices.
Theme 3: Technical Support and Training
This theme encompasses the issues participants experienced and the support systems in place to help them. The theme also includes reliance on care partners for support and requests for various training modalities.
Role of the Care Partner
Care partner involvement was crucial for many participants with MCI. Care partners were key in helping their loved ones remember how to charge, synchronize, and interpret the Oura Ring data. Some care partners took a more supervisory role, providing only emotional support and encouragement to the participant with MCI. In contrast, others were hands-off, allowing the participant with MCI to manage their device completely. While less common, some participants with MCI indicated they did not engage with the Oura Ring application at all. In these cases, care partners managed to charge and to synchronize the ring. They also used the mobile application to view their partner’s health data and trends. “She [care partner] does the syncing. Helps me read the graphs. We talk about it in the morning and help each other understand.”
Training and Research Support
Participants requested more training opportunities than what was originally provided in the study. Requests included different training modalities, such as small group classes, 1-1 training, or virtual classes. Additionally, participants indicated a desire to have recurring or repeated “refresher” classes. They also requested immediate access to researchers via a single point of contact.
“I would like to have had more in-person trainings at the start to walk through the app.”
Synchronization compliance
Due to the different enrollment durations for participants, these results are reported in terms of the percentage of days where at least one physiological measure was missing. All participants’ average percentage of missing data was 15.59% (SD = 19.29%, median = 7.78%, max = 73.95%, min = 0%). A post-hoc Welsh’s-T test indicated that there was a significant difference in the percentage of missing data among participants with MCI and Care Partners, t(34) = 8.12, p < .001, with participants with MCI having missed 20.58% while Care Partners missed an average of 3.18% of data collection.
Participants were contacted if data was not received after 4 days. Researchers helped the participants resume using their devices and recorded the reasons for the missing data. The most frequent causes included instances where participants forgot to synchronize their device (22.06%, n = 15), were traveling (14.71%, n = 10), forgot the procedure to synchronize (13.24%, n = 9), forgot to charge their device (7.35%, n = 5), or lost the device or charger (7.35%, n = 5). Issues that were less common but still contributed to periods of missing data included accidentally logging out of the application (5.88%, n = 4), turning off features (2.94%, n = 2), accidentally switching devices with another participant (2.94%, n = 2), substance covering the sensors (2.94%, n = 2), unplanned hospitalizations (2.94%, n = 2), cellphone/tablet issue (2.94%, n = 2), damaged device (1.47%, n = 1), and the device requiring an update (1.47%, n = 1).
Participants with a spousal care partner were asked how much their care partners helped them understand the various scores and graphs presented on the Oura application. Similarly, care partners were asked how much they helped their spouse with MCI understand the various scores and graphs. This was measured on a five-point Likert-style scale: “none at all” = 1, “a little” = 2, “a moderate amount” = 3, “a lot” = 4, and “a great deal” = 5. The average score for participants with MCI was M = 2.51 (SD = 1.37), while the care partners’ average score was M = 2.17 (SD = 0.75). A post hoc Mann-Whitney U test was conducted to determine if there was a significant difference between the perceived amount of help received by participants with MCI and the perceived amount of help given by care partners. The results indicated no significant difference between members’ reported amount of help received and care partners’ reported amount of help given (U = 91.5, p = .72).
Discussion
The completion of this feasibility study uncovered critical insights. Key guidelines for engaging older adults with and without MCI in research using commercially available wearable health devices were identified. Regarding the first research question, it became apparent that participant satisfaction with Oura Ring usage depended on physical and psychological comfort. As found in the qualitative results, it was common for some participants to report physical discomfort associated with finger swelling and arthritis. Therefore, having replacement devices in different sizes, providing strategies to switch fingers, and implementing ring size adjusters were essential to help improve physical comfort. Additionally, participants had a more enjoyable experience when they understood and could easily interact with the Oura Ring and application. Participants expressed satisfaction and reassurance from receiving positive feedback and seeing their health information within the Oura Ring application.
From these results, we inferred psychological comfort from participant responses aligning with the literature. According to Mitra et al. (1999), psychological comfort includes understanding and emotional ease. Participants demonstrated understanding through their ability to utilize the Oura Ring successfully. Emotional ease was expressed by reporting satisfaction with the Oura Ring application and improved sleep habits.
Regarding our second research question, we discovered three major areas where researchers’ support was essential to the study’s success and the participants’ continued involvement. These areas might be considered best practices for the development of future studies that involve deploying commercially wearable health devices to people with MCI. First, researchers must effectively train participants to use the technology. This includes assisting participants with the device’s initial setup, demonstrating proper use, and providing curated materials to support the learning process. When training, researchers benefited from adopting a flexible approach and considering the unique needs of each participant.
The findings of this study are consistent with past work on training for people with cognitive disabilities. For example, a 2011 International Labor Office working paper on integrating people with cognitive disabilities into the workplace recommended that training programs be tailored to each trainee’s specific needs (Parmenter, 2011). Similarly, Buzzi et al. (2019) highlighted the importance of adjusting the difficulty of training programs based on the individual responses of trainees. This flexibility supports using an action research framework, where programs and methodologies can be continually iterated based on the participant’s needs.
Second, researchers should consider participant needs when developing the research protocol. This includes identifying the most comfortable frequency and modality of communication for them, such as a phone call versus an in-person appointment. During this time, researchers addressed technological hurdles, such as the nature of the issue, and contextual aspects, such as the type of device and operating system. Researchers’ familiarity with wearable technology and smartphone/tablet usage was crucial for understanding, promptly resolving, and preventing missing data without overwhelming the participant. Ultimately, ongoing interaction allowed participants to pose questions or concerns that arose during the study and helped the research team build rapport with them. Regular communication also helped maintain data quality by allowing researchers to easily resolve issues while ensuring consent was maintained throughout the study.
Lastly, past research shows that programs incorporating participant interactivity and social support elements lead to increased engagement and motivation for people with cognitive impairments (Leung et al., 2015). Therefore, our final recommendation is to involve care partners when possible. Care partners played a crucial role as valuable resources for participants. The results of this feasibility sub-study indicate that care partners assisted participants with MCI in troubleshooting and device use while offering emotional support and encouragement.
Implications for Future Research
A promising area of future research is investigating the long-term effects of highly personalized training programs on adopting and using commercially available wearable technologies. Additionally, research could be conducted to assess which training modality could best facilitate learning and retention (e.g., in-person, virtual, individual, or group classes). Next, researchers could investigate the effectiveness of different communication modalities (e.g., video calls, phone calls, and in-person visits) in supporting older adults and older adults with MCI in longitudinal studies. Lastly, researchers could delve into the specific dynamics between people with MCI and their care partners, how support needs change and evolve, and how researchers can help care partners support their partner MCI while enrolled in longitudinal studies.
Conclusion
Commercially available wearable health devices have the potential to enhance the health and overall quality of life of older adults with MCI by providing insights into their health (Holthe et al., 2018). This feasibility study offered essential insights into the experiences of older adults using the Oura Ring, emphasizing the need for physical and psychological comfort to ensure participant satisfaction and engagement. Likewise, we identified three key guidelines for future research: effective training, involving care partners, and maintaining regular communication with participants. These elements facilitate successful device use, increase participant engagement, and improve data quality. Ultimately, our findings contribute to Human Factors literature by emphasizing the use of tailored materials, flexible research methodologies, and robust participant communication strategies when conducting research studies that utilize commercially available wearable technologies in a sample of older adults with and without MCI.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research is funded by the Charlie and Harriet Shaffer Cognitive Empowerment Program (CEP), a research partnership between Emory Brain Health and the Georgia Institute of Technology, through the generosity of the James M. Cox Foundation.
