In-Person, Online, and Hybrid Healthy-Aging Education for Pre-Frailty and Frailty in Older Adults

Frailty is a potentially modifiable state of reduced physiological reserve that increases vulnerability to falls, disability, hospitalization, loss of independence, and higher healthcare use. Multidomain programs addressing physical activity, nutrition, cognitive and emotional health, and social participation can improve frailty-related outcomes. However, it remains uncertain how in-person, online, and hybrid delivery differ in effectiveness, reach, implementation, resource use, and cost-effectiveness. The trial is not designed to establish equivalence or non-inferiority among the three active modalities. FRAGSALUD-ON builds on the previous FRAGSALUD program by comparing three delivery modalities within the same randomized trial and by integrating clinical, behavioral, biological, implementation, and economic evaluations.

This is a 12-month, four-arm, parallel-group, household-cluster randomized controlled trial conducted in the province of Cadiz, Spain. The household is the unit of randomization: a participant without another enrolled cohabitant constitutes a cluster of one, and all enrolled cohabitants are assigned to the same study arm. Households, rather than individual participants, will be allocated in a 1:1:1:1 ratio to in-person education, online education, hybrid education, or usual care. The target enrollment is 276 participants; participant numbers per arm are expected to be approximately, but not necessarily exactly, equal. Randomization will be performed by an independent statistician who is not involved in recruitment or outcome assessment. Participants and intervention providers cannot be masked because of the nature of the programs. Outcome assessors will remain unaware of group allocation whenever feasible. Assessments will be conducted at baseline, at the end of the 6-month intervention, and 6 months after the intervention ends (12 months after baseline). Blood and stool samples will be obtained at baseline and 6 months.

The three active interventions share four core domains: physical activity, nutrition, cognitive and emotional well-being, and social connection. The in-person group will attend four weekly group sessions during the first month and receive six reinforcement telephone calls during the following 5 months. The hybrid group will attend the same four group sessions plus a digital-platform onboarding session and will use the online platform for reinforcement during the following 5 months instead of receiving the telephone calls. The online group will receive an onboarding session and will use the platform as the main educational delivery method throughout the 6-month intervention.

Educational materials were developed before participant recruitment with artificial-intelligence assistance and were subsequently reviewed, edited, and approved by the research team. Artificial intelligence was used solely to support the development of educational materials before participant recruitment. It will not receive participant data, interact with participants, personalize intervention content, monitor adherence, or provide clinical advice during the trial. The platform will provide interactive educational modules, knowledge activities, navigation support, and automated reminders without artificial-intelligence processing of participant data.

The usual-care group will continue its routine healthcare and health routines and will not receive the FRAGSALUD-ON educational program during the 12-month follow-up. After completing the 12-month assessment, participants in this group will be offered access to the online educational platform, in accordance with the ethics-approved protocol.

The primary objective is to determine whether the average effect of the three active delivery modalities is superior to usual care in reducing the prespecified 40-item Frailty (40-FI) Index from baseline to 6 months. The primary analysis will follow the intention-to-treat principle and will use a linear mixed-effects analysis of covariance for the 6-month 40-FI, with fixed effects for study arm, baseline 40-FI, and the household-level baseline frailty stratum used in randomization (all enrolled household members prefrail versus at least one enrolled household member frail), and a random intercept for household. The confirmatory contrast will compare the equally weighted mean of the three active arms with usual care. Missing 6-month 40-FI values among surviving participants will be addressed using multiple imputation under a missing-at-random assumption, with prespecified missing-not-at-random sensitivity analyses. Death and other intercurrent events will be handled according to the estimand specified in SAP.

If the primary joint contrast is statistically significant, each active modality will be compared with usual care using Dunnett adjustment to control the family-wise error rate. Direct comparisons among the three active modalities will be treated as secondary or exploratory and will be interpreted with confidence intervals rather than as tests of equivalence or non-inferiority.

The target sample size is 276 participants, with 69 participants per group. It was selected to provide approximately 80% power for the primary joint contrast, assuming an adjusted between-group difference of 0.030 in the Frailty Index, a residual standard deviation of 0.070, a two-sided alpha level of 0.05, and 15% missing primary-outcome data at 6 months. Because enrolment of more than one participant from the same household is expected to be uncommon based on prior recruitment experience, the initial target assumes a negligible loss of efficiency from household clustering. During recruitment, an independent statistician blinded to treatment outcomes will conduct a prespecified review of the observed household cluster-size distribution. The enrollment target may be increased if the projected household design effect is not negligible.

Secondary objectives are to assess whether effects persist at 12 months and to compare changes in physical frailty, physical performance, strength, balance, mobility, independence in daily activities, physical activity and 24-hour movement behavior, nutritional status and diet, cognition, mood, sleep, social support, health-related quality of life, falls, and other clinically relevant outcomes.

Blood biomarkers and gut microbiota will be assessed at baseline and 6 months to characterize biological changes and examine within-participant associations between changes in these measures and changes in frailty, physical function, cognition, inflammation, diet, and gastrointestinal function. These longitudinal association analyses will not by themselves be interpreted as proving causation or mediation.

A prospective within-trial economic evaluation over 12 months will compare the four alternatives using costs and quality-adjusted life-years. The primary economic analysis will adopt a societal perspective, with a healthcare-system perspective examined in sensitivity analyses. Incremental costs, incremental effects, cost-effectiveness, net monetary benefit, and decision uncertainty will be reported. Adherence, satisfaction, barriers, facilitators, and implementation experience will also be evaluated to help determine how the delivery modalities could be used in practice.

• Change From Baseline in the Prespecified 40-Item Frailty Index Score at 12 Months [Time frame: Baseline and 12 months]
• Change From Baseline in the Fried Frailty Phenotype Score at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Maximum Dominant-Hand Grip Strength at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Short Physical Performance Battery Total Score at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Best Eyes-Open 30-Second Single-Leg Stance Time at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Fast-Pace 4-Meter Gait Speed at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Estimated Relative Sit-to-Stand Mean Power at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Barthel Index Total Score at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Lawton Instrumental Activities of Daily Living Total Score at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Accelerometer-Measured Moderate-to-Vigorous Physical Activity at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Full Mini Nutritional Assessment Total Score at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]
• Change From Baseline in Montreal Cognitive Assessment Total Score at 6 and 12 Months [Time frame: Baseline, 6 months, and 12 months]