Evaluating Health Outcomes of AI-Based Fitness Wearables and App Programs in Older Adults Living Alone With Cognitive Decline

The overarching goal of our research is to develop personalized and accessible healthy aging lifestyle interventions aimed at promoting physical activity (PA) and improving health among community-dwelling older adults living alone with cognitive decline (LACD). To achieve this goal, the purpose of this project is to determine whether wearable and app-based mHealth intervention component(s) will contribute to increased PA and improved health outcomes in older adults LACD. Our specific aims are to: identify and evaluate mHealth intervention components that practically and significantly contribute to enhanced mechanistic outcomes (e.g., self-efficacy, outcome expectations) and increased PA (primary outcome) in older adults LACD over a 6-month period; determine the optimal combinations of intervention components for future efficacy testing; elucidate the mechanism of behavioral change (MoBC) and potential outcomes of these intervention components, namely, the mediating effects of MoBC variables (e.g., self-efficacy, outcome expectations) on the relationship between intervention components and change in PA. The first two aims are primary and fully-powered. The third aim is exploratory. The aims will support a refined, data-driven intervention design for a subsequent larger trial.

Study Overview

• Status: Recruiting
• Conditions: Physical Activity; Older Adults; Cognitive Decline; Wearables; Physical Inactivity; Older Adults With Cognitive Decline; AI-Based Fitness
• Intervention / Treatment: Other: Fitness app for self-efficacy; Other: Social network via app for social support; Other: Health education app targeting outcome expectations

Detailed Description

Mobile health (mHealth) is a promising approach to improving health behaviors, defined as "health services and information delivered or enhanced through the Internet and related technologies." It includes disease prevention and management tools, remote interventions, personalized health monitoring, and mobile healthcare data access. With widespread technology adoption, researchers increasingly use wearable devices and apps to enhance health outcomes by promoting PA and reducing sedentary behavior. Wearable devices and fitness apps are now widely integrated into PA intervention programs, helping individuals adopt more active lifestyles. These tools track steps, activity duration, and progress, providing real-time feedback, goal-setting, and social integration to enhance motivation and behavior regulation. Notably, 21% of U.S. adults regularly use smartwatches or fitness trackers, making them feasible for PA interventions in older adults. RCTs have shown their positive effects on PA, QoL, and psychosocial well-being in older adults though some studies reported modest improvements. Recent advancements in data science and AI-driven mHealth interventions enable scalable, personalized exercise prescriptions. Personalized approaches, particularly those enhancing self-efficacy, yield better outcomes than generalized interventions. However, few studies have leveraged fitness wearables and apps for older adult LACD. This trial addresses this major weakness by implementing an AI-driven mHealth intervention for tailored precision health programs in older adult LACD.

• Study Type: Interventional
• Enrollment (Estimated): 64
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Zan Gao, PhD; Phone Number: 865-974-7971; Email: zan@utk.edu

Study Locations

• United States
  • Tennessee
    • Knoxville, Tennessee, United States, 37920
      • Not yet recruiting
      • University of Tennessee
      • Principal Investigator: Zan Gao, PhD
      • Contact: Kinesiology, Recreation, and Sport Studies; Phone Number: (865) 974-3340; Email: krss@utk.edu
      • Sub-Investigator: Danielle Ostendorf, PhD
      • Sub-Investigator: Xiaopeng Zhao, PhD
      • Sub-Investigator: Jeffrey Labban, PhD
    • Knoxville, Tennessee, United States, 37996
      • Recruiting
      • The University of Tennessee, Knoxville. Health, Recreation, and Physical Education Building

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

• Participant must be at least 65 years of age older
• Participant must be living alone in the U.S. for the next 6 months
• Participant must have report mild cognitive decline [We will use a short self-report AD8 measure of cognitive concerns. Those scoring positive on the AD8 (≥2) will qualify as mild cognitive decline];
• Participant must own an Android/Apple smartphone
• Participant must have access to internet or Wi-Fi access
• Participant must be capable of engaging in some PA as determined by the PA Readiness Questionnaire or physician approval
• Participant must currently participate in weekly moderate-to-vigorous PA (MVPA) or less than 150 minutes
• Participant must have basic English communication skills.

Exclusion Criteria:

• Foreign residents or visitors

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Prevention
• Allocation: Randomized
• Interventional Model: Factorial Assignment
• Masking: Double

Number of Arms

8

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Access to all applications; Condition 1: Participant are provided with the prescription application (application1), social application (application 2), and health tips application (application 3).	Other: Fitness app for self-efficacy; AI-driven personalized exercise prescription via a fitness app. This targets self-efficacy.; Other: Social network via app for social support; Participants will be provided access to a social network via app. This targets social support.; Other: Health education app targeting outcome expectations; Participants are provided with an app-based health education. This targets outcome expectations.
Experimental: Access to application 1 & 2; Condition 2: Participant are provided with the prescription application, social application, but they aren't provided with the health tips application.	Other: Fitness app for self-efficacy; AI-driven personalized exercise prescription via a fitness app. This targets self-efficacy.; Other: Social network via app for social support; Participants will be provided access to a social network via app. This targets social support.
Experimental: Access to application 1 & 3; Condition 3: Participant are provided with the prescription application, and they aren't provided with the social application, but they are provided with the health tips application.	Other: Fitness app for self-efficacy; AI-driven personalized exercise prescription via a fitness app. This targets self-efficacy.; Other: Health education app targeting outcome expectations; Participants are provided with an app-based health education. This targets outcome expectations.
Experimental: Access to application 1 only; Condition 4: Participant are provided with the prescription application, but aren't provided with the social application, and the health tips application.	Other: Fitness app for self-efficacy; AI-driven personalized exercise prescription via a fitness app. This targets self-efficacy.
Experimental: Access to application 2 & 3; Condition 5: Participant are not provided with the prescription application, but they are provided with the social application, and the health tips application.	Other: Social network via app for social support; Participants will be provided access to a social network via app. This targets social support.; Other: Health education app targeting outcome expectations; Participants are provided with an app-based health education. This targets outcome expectations.
Experimental: Access to application 2 only; Condition 6: Participant are not provided with the prescription application, but they are provided with the social application, and they aren't provided with the health tips application.	Other: Social network via app for social support; Participants will be provided access to a social network via app. This targets social support.
Experimental: Access to application 3 only; Condition 7: Participant are not provided with the prescription application, or the social application, but they are provided with the health tips application.	Other: Health education app targeting outcome expectations; Participants are provided with an app-based health education. This targets outcome expectations.
No Intervention: No access to any application; Condition 8: Participant are not provided with the prescription application, or the social application, or with the health tips application.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Fitbit MVPA	Fitbit Inspire 3 Tracker will be used to assess participants' MVPA (active time which includes both fairly active time and very active time). Fairly active; duration associated with light intensity activities, i.e., walking, light cycling, housework (~3-6 METs). Very active; duration associated with high intensity activities, i.e., running, aerobic workouts (>6 METs).	Baseline (i.e., pre-intervention), 3 months (mid-point), and 6 months (end-point).
Physical Activity	We will use the Physical Activity Scale for the Elderly to assess PA. Higher scores means more physical activity. (Low activity: <100; Moderate: 100-250; High: >250)	Baseline (i.e., pre-intervention), 3 months (mid-point), and 6 months (end-point).
Mechanism of behavior change (MoBC) variables	Psychometrically validated questionnaires will be used to assess beliefs: self-efficacy, social support, and outcome expectations. Self-efficacy; low score indicates low confidence in ability to perform behavior, high score indicates strong confidence. Social support; low score indicates poor support from family or friends, high score indicates great support. Outcome expectation; low score indicates the belief that behavior won't help, high score indicates the belief that the behavior will lead to positive outcome.	Baseline (i.e., pre-intervention), 3 months (mid-point), and 6 months (end-point).

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Quality of life (QoL)	We will select the brief Older People's Quality of Life questionnaire to assess quality of life (OPQOL-BRIEF). Minimum: 13 (very poor quality of life) Maximum: 65 (excellent quality of life)	Baseline (i.e., pre-intervention), 3 months (mid-point), and 6 months (end-point).
Psychosocial wellbeing	We will assess wellbeing using the World Health Organization Well-Being Index (WHO-5) questionnaire. 0 -12 - Low well-being; possible depression (screen positive) 13 - 19 - Moderate well-being 20 - 25 - High well-being	Baseline (i.e., pre-intervention), 3 months (mid-point), and 6 months (end-point).
Cognition	To assess older adults' cognition (thinking and memory), we will use the 5-minute Montreal Cognitive Assessment (MoCA). 26 - 30: Normal; Indicates health cognitive functioning for most adults. 18 - 25: Mild Cognitive Impairment (MCI); Shows noticeable memory or thinking issues that do not severely disrupts daily life. 10 - 17: Moderate Cognitive Impairment; Suggests a heightened progression towards early-to-mid stage dementia. Less than 10: Severe Cognitive Impairment.	Baseline (i.e., pre-intervention), 3 months (mid-point), and 6 months (end-point).

Collaborators and Investigators

• Sponsor: The University of Tennessee, Knoxville
• Collaborators: National Institute on Aging (NIA); Arizona State University; Oregon Research Institute

Study record dates

Study Major Dates

• Study Start (Actual): June 8, 2026
• Primary Completion (Estimated): June 8, 2028
• Study Completion (Estimated): June 8, 2028

Study Registration Dates

• First Submitted: September 19, 2025
• First Submitted That Met QC Criteria: September 27, 2025
• First Posted (Actual): October 6, 2025

Study Record Updates

• Last Update Posted (Actual): June 15, 2026
• Last Update Submitted That Met QC Criteria: June 10, 2026
• Last Verified: June 1, 2026

More Information

Terms related to this study

• Keywords: Physical Activity; Cognitive decline; Older adults; Physical Inactivity; AI-Based Fitness; Living alone with cognitive decline; LACD; U.S Older adults
• Additional Relevant MeSH Terms: Mental Disorders; Neurocognitive Disorders; Cognition Disorders; Behavior; Cognitive Dysfunction; Motor Activity; Sedentary Behavior
• Other Study ID Numbers: STUDY00000016

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No