Cardiorespiratory Fitness, Physical Activity
Conditions
Keywords
Artificial Intelligence, Digital Health, Male University Students, Samsung Health, Wearable Technology, Digital Coaching, Physical Activity, Cardiorespiratory Fitness
Brief summary
This study evaluated whether AI-supported digital coaching combined with Samsung Health-based activity monitoring could improve cardiorespiratory fitness and physical activity in male university students compared with standard Samsung Health monitoring. In this single-center randomized controlled trial, 312 male university students were assigned in a 1:1 ratio to approximately 10 weeks of either AI-supported digital coaching or standard activity monitoring. The AI-supported intervention provided adaptive daily step goals, activity coaching, progress feedback, and recovery recommendations based on recent activity, while the control group used standard monitoring without AI-supported coaching. The primary outcome was estimated maximal oxygen uptake (VO2max) derived from the Cooper 12-minute test. Physical activity was also evaluated using daily step counts collected through the study activity-monitoring workflow.
Detailed description
This was a single-center, parallel-group randomized controlled trial conducted at Transilvania University of Brașov, Romania. Participants meeting the study-defined semi-active criteria were randomized 1:1 to AI-supported digital coaching or standard Samsung Health-based activity monitoring. The experimental intervention combined activity monitoring with adaptive digital coaching. Coaching outputs included personalized daily step targets, activity coaching and goal adjustment, weekly progress summaries, and recovery or rest-day recommendations based on recent activity. Android participants received the configured intervention through the research Samsung Health environment. For participants using iOS devices, activity data were incorporated through the study workflow using Apple HealthKit where necessary, and equivalent coaching was delivered through a research-administered web dashboard. Participants in the control group used standard Samsung Health monitoring, including step counting, activity displays, and generic weekly reports, without the research-configured AI-supported coaching. No study-provided individualized exercise program was delivered to either group. Cardiorespiratory fitness was assessed using estimated VO2max derived from distance completed during the Cooper 12-minute test. Physical activity was monitored using daily step counts. The trial was designed to compare cardiorespiratory fitness and physical-activity trajectories between the randomized groups.
Interventions
BEHAVIORALAI-supported digital coaching with Samsung Health-based activity monitoring
Participants received approximately 10 weeks of Samsung Health-based activity monitoring combined with research-configured AI-supported digital coaching. The intervention provided adaptive daily step targets, real-time coaching and goal adjustment, weekly progress summaries and insight cards, and recovery or rest-day recommendations based on recent activity. Android participants received the configured functions through the research Samsung Health environment. For iOS participants, activity data were synchronized through the study workflow using Apple HealthKit where necessary, and equivalent coaching was delivered through a research-administered web dashboard. Research staff provided technical support but did not manually alter AI recommendations or provide individualized exercise coaching.
BEHAVIORALStandard Samsung Health activity monitoring
Participants used standard Samsung Health monitoring for approximately 10 weeks, including step counting, activity rings, and generic weekly reports. Research-configured AI-supported coaching was disabled for the control condition. Research staff provided standardized technical support, but no individualized exercise coaching or study-provided exercise program was delivered.
Sponsors
Transilvania University of Brasov
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
SINGLE (Outcomes Assessor)
Masking description
Cooper-test outcome assessors were blinded to randomized group allocation. Participants and the coordinator responsible for digital intervention configuration were not blinded. The statistician responsible for the primary analysis remained blinded to allocation until database lock and finalization of the analysis code.
Intervention model description
Participants were randomized in a 1:1 ratio to two parallel groups: AI-supported digital coaching with activity monitoring or standard Samsung Health monitoring for approximately 10 weeks.
Eligibility
Sex/Gender
MALE
Age
18 Years to 25 Years
Healthy volunteers
Yes
Inclusion criteria
* Male university student. * Average of 5,000-7,499 steps/day over the last 14 valid pre-baseline activity-monitoring days available through the study workflow. * Self-reported moderate-to-vigorous physical activity of 150-299 minutes/week on the International Physical Activity Questionnaire-Short Form (IPAQ-SF). * Completion of the Physical Activity Readiness Questionnaire for Everyone (PAR-Q+); participants with an affirmative response were required to obtain medical clearance before enrollment. * Written informed consent provided before baseline assessment and randomization.
Exclusion criteria
* Did not meet the study-defined semi-active physical-activity criteria, including participants who were already highly active. * Medical contraindication to participation or failure to obtain required medical clearance following safety screening. * Planned absence of more than 2 weeks during the study period.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Estimated maximal oxygen uptake (VO₂max) derived from the Cooper 12-minute test | Approximately 10 weeks after baseline (final assessment) | Cardiorespiratory fitness was assessed using the Cooper 12-minute test on a standard 400-m outdoor track. Participants were instructed to run or walk as far as possible in 12 minutes. Distance was recorded to the nearest 10 m. Estimated VO₂max was calculated as (distance in meters - 504.9) / 44.73 and expressed in mL·kg-¹·min-¹. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Mean daily step count by study week | Daily for 68 days (approximately 10 weeks), summarized by study week. | Daily step counts were collected through the study activity-monitoring workflow using Samsung Health and, for iOS participants where necessary, Apple HealthKit. A monitoring day was considered valid when the device recorded at least 10 hours of wear or usage time, the recorded step count was between 100 and 40,000, and no manual-entry flag or device-error code was present. Only valid monitoring days contributed to weekly summaries. Mean daily step counts were summarized by study week and used to evaluate physical-activity trajectories across the monitoring period. |
Countries
Romania
Contacts
PRINCIPAL_INVESTIGATORStefan Alecu, Ph.D
Universitatea Tranilvania din Brasov
Outcome results
None listed