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

A prespecified 40-item deficit-accumulation Frailty Index (40-FI) will be calculated as the sum of observed deficit scores divided by the number of evaluable deficits. The 40-FI ranges from 0 to 1, with higher values indicating greater frailty. A score will be calculated when at least 32 of the 40 deficits (80%) are evaluable. Change will be calculated as the 6-month score minus the baseline score; negative values indicate improvement. The fixed item list, item-level coding rules, permitted intermediate values, denominator, and missing-data rules are defined in the version-controlled 40-FI Specification included as an annex to Protocol and SAP and finalized before enrolment of the first participant.

A prespecified 40-item deficit-accumulation Frailty Index (40-FI) will be calculated as the sum of observed deficit scores divided by the number of evaluable deficits. The 40-FI ranges from 0 to 1, with higher values indicating greater frailty. A score will be calculated when at least 32 of the 40 deficits (80%) are evaluable. Change will be calculated as the 12-month score minus the baseline score; negative values indicate improvement. The fixed item list, item-level coding rules, permitted intermediate values, denominator, and missing-data rules are defined in the version-controlled 40-FI Specification included as an annex to Protocol and SAP and finalized before enrolment of the first participant.

The Fried frailty phenotype will be scored at baseline, 6 months, and 12 months as the number of criteria present (0-5): unintentional weight loss, exhaustion, weakness, slowness, and low physical activity; higher scores indicate greater frailty. At baseline, weight loss is self-reported unintentional loss >4.5 kg in the previous 12 months; at 6 and 12 months, it is measured loss ≥5% from baseline reported as unintentional. Exhaustion is present when either original CES-D item is reported ≥3 days in the previous week. Weakness uses the highest of three dominant-hand grip-strength trials and original sex- and BMI-specific cutoffs. Slowness uses usual-paced 4.57-m walk time and original sex- and height-specific cutoffs. Low activity is VREM <1,250 MET-min/14 days. Each criterion scores 0 or 1. Separate changes are 6-month minus baseline and 12-month minus baseline (range, -5 to 5); negative values indicate improvement. All five criteria are required.

Maximum dominant-hand grip strength will be measured in kilograms with a calibrated Jamar hydraulic dynamometer at handle position II. Higher values indicate greater strength. Participants will be seated with shoulder adducted and neutrally rotated, elbow flexed 90°, forearm neutral, wrist at 0-30° extension and 0-15° ulnar deviation, and feet supported. Three maximal contractions of 3-5 seconds will be performed with 60 seconds of rest and standardized encouragement. The highest valid value will be analyzed; if fewer than three trials are valid, the highest available value will be used. At least one valid trial is required. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate improvement and negative values worsening. The baseline dominant hand and same device, handle setting, and protocol will be used at all visits. Values and reasons for invalid or missing trials will be recorded.

The SPPB assesses lower-extremity performance using standing balance (side-by-side, semi-tandem, and tandem stances), usual-paced walking over 4 meters, and five chair rises performed as quickly as possible with arms folded. Each component is scored 0-4 according to standard rules, producing a total score from 0 to 12; higher scores indicate better performance. The faster valid time from two usual-paced 4-meter walks determines the gait score; if only one trial is valid, it is used. Participants unable to complete a component for physical or safety reasons receive 0 points. A usual walking aid may be used; its type will be recorded and kept constant when clinically appropriate. Separate changes are the 6-month total minus baseline and the 12-month total minus baseline (range, -12 to 12); positive values indicate improvement and negative values worsening. The same equipment, instructions, sequence, stopping rules, and scoring will be used at all assessments.

Eyes-open single-leg stance will be timed in seconds (0-30) on a firm nonslip surface; higher values indicate better balance. Barefoot participants will stand on the preferred stance limb selected at baseline, with arms crossed over the chest and eyes fixed on a target. Stable support may be used to assume the position but not during timing. Timing starts when the nonstance foot leaves the floor and stops if the arms uncross, the raised foot touches the stance limb or floor, the stance foot moves, support or assistance is required, or 30 seconds elapse. Two trials will be performed with 60 seconds of rest; the better valid time will be analyzed. If only one trial is valid, it will be used. Inability to release support and safely assume the position scores 0 seconds. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate improvement and negative values worsening. The same limb and protocol will be used at all visits.

Fast gait speed will be measured over 4 meters from a static start and reported in m/s. In walking shoes, participants will stand with toes behind the start line and, on command, walk as quickly as safely possible without running, continuing past the finish line. Timing starts when the first foot crosses the start line and stops when the first foot crosses the 4-meter line; speed is calculated as 4/time in seconds. Two timed trials 60 seconds apart will be performed; the faster valid speed will be analyzed. If only one trial is valid, it will be used. The usual walking aid is permitted, recorded, and kept constant whenever feasible. A trial requiring physical assistance or involving running is invalid. Higher values indicate better performance. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate improvement and negative values worsening. The same protocol will be used at all visits.

Estimated relative mean lower-limb muscle power during the five-repetition sit-to-stand test will be calculated in W/kg using the prespecified Alcazar equation: [0.9 × 9.81 × (0.5 × body height [m] - chair height [m])] / (5STS time [s] × 0.1). Baseline standing height and the measured 0.43-m chair height will be used at all visits; 5STS time will be the valid completion time obtained during the SPPB chair-stand test. Higher values indicate greater estimated relative power. Power will be calculated only for valid completed tests and will be missing when no valid completion time is available. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate improvement and negative values worsening. The same equation, units, chair, and data-processing rules will be applied at all visits.

The validated Spanish 10-item Barthel Index will assess actual performance in basic activities covering feeding, personal care, dressing, continence, toilet use, transfers, mobility, and stairs. A trained assessor will administer it face-to-face using participant report, an informed proxy when needed, and observation when available. Standard weighted categories produce a total score from 0 to 100; higher scores indicate greater independence. Scoring reflects what the participant actually does rather than potential capacity; supervision, assistance, and aids will be classified using standard Barthel rules. The total will be missing if any item is unclassifiable; scores will not be prorated. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline (range, -100 to 100); positive values indicate improvement and negative values worsening. The same version and scoring rules will be used at all visits.

The validated Spanish eight-item Lawton IADL Scale will assess telephone use, shopping, food preparation, housekeeping, laundry, transportation, medication management, and financial management. It will be administered face-to-face by a trained assessor. Items are coded 1 (able) or 0 (unable or partially able), producing a total score from 0 to 8; higher scores indicate greater independence. All eight items will be assessed in women and men. An activity not routinely performed solely because of customary roles will not automatically be scored 0; functional ability will be established through standardized probing and, when needed, an informed proxy. The total will be missing if any item is unclassifiable; scores will not be prorated. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline (range, -8 to 8); positive values indicate improvement and negative values worsening. The same version and scoring rules will be used at all visits.

Free-living moderate-to-vigorous physical activity (MVPA) will be measured using a triaxial GENEActiv accelerometer worn on the non-dominant wrist continuously (24 h/day) for 7 consecutive days at 60 Hz at baseline, 6 months, and 12 months. Raw acceleration will be processed using the same prespecified GGIR-compatible ENMOa pipeline at all assessments. Waking epochs with acceleration ≥98 mg will be classified as MVPA using the age- and wear-site-specific threshold validated for GENEActiv in older adults. The outcome is mean MVPA minutes/day across valid days; all qualifying epochs will count, without a minimum bout duration. A valid assessment requires ≥4 days with ≥10 h/day of waking wear, including ≥1 weekend day. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline. Positive values indicate increased MVPA.

The official Spanish-language 18-item Full Mini Nutritional Assessment (MNA) will be administered to every participant at baseline, 6 months, and 12 months. All screening and assessment items will be completed regardless of the screening subsection score, with no skip pattern or use of the MNA-SF alone. The MNA combines anthropometric, general, dietary, and subjective assessments, including measured weight, height/BMI, mid-upper-arm circumference, and calf circumference. Items are scored according to the standard manual and summed to a total score from 0 to 30; higher scores indicate better nutritional status. Scores <17 indicate malnutrition, 17-23.5 risk of malnutrition, and 24-30 normal nutritional status. Changes will be calculated separately as the 6-month total minus baseline and the 12-month total minus baseline (range, -30 to 30); positive values indicate improvement. The same version, equipment, procedures, and scoring rules will be used at all visits.

Global cognitive performance will be assessed using the official Spanish (Spain) MoCA Full version 8.1, administered face-to-face at baseline, 6 months, and 12 months. The same version will be used at all assessments to preserve linguistic and cultural equivalence. The MoCA assesses visuospatial/executive functions, naming, attention, language, abstraction, delayed recall, and orientation. The total score ranges from 0 to 30; higher scores indicate better cognitive performance. One point will be added for participants with 12 or fewer years of formal education, without exceeding 30 points. Changes will be calculated separately as the 6-month score minus baseline and the 12-month score minus baseline. Positive values indicate improvement and negative values worsening. Official administration and scoring instructions will be followed.

Sleep quality will be assessed using the validated Spanish version of the Pittsburgh Sleep Quality Index (PSQI), referring to the preceding month. Nineteen self-reported items are used to derive seven component scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction. Each component ranges from 0 to 3 and the components are summed to produce a global score from 0 to 21; higher scores indicate poorer sleep quality. Changes will be calculated separately as the 6-month score minus baseline and the 12-month score minus baseline. Negative values indicate improvement and positive values indicate worsening. The same Spanish version, reference period, administration procedure, and scoring rules will be used at all assessments.

Depressive symptoms will be assessed using the validated Spanish 15-item Geriatric Depression Scale (GDS-15; Yesavage). The scale contains 15 yes-or-no items referring to how the participant has felt during the preceding week. Each response indicative of a depressive symptom scores 1 point, producing a total score from 0 to 15; higher scores indicate more depressive symptoms. The GDS-15 is a screening measure and will not be used to establish a clinical diagnosis of depression. Changes will be calculated separately as the 6-month score minus baseline and the 12-month score minus baseline. Negative values indicate improvement and positive values indicate worsening. The same Spanish version, administration procedure, reference period, and scoring rules will be used at all assessments.

Health-related quality of life will be assessed using the official Spanish version of the EQ-5D-5L. Its descriptive system assesses mobility, self-care, usual activities, pain/discomfort, and anxiety/depression, with five severity levels per dimension. Each health-state profile will be converted into an index value using the Spanish EQ-5D-5L value set (Ramos-Goñi hybrid model, 2018). Index values range from -0.416 to 1, where 1 represents full health, 0 represents a health state equivalent to being dead, and negative values represent states valued as worse than dead; higher values indicate better health-related quality of life. Changes will be calculated separately as the 6-month value minus baseline and the 12-month value minus baseline. Positive values indicate improvement and negative values worsening. An index value will not be calculated if any dimension is missing.

Self-rated overall health will be assessed using the EQ Visual Analogue Scale (EQ VAS) included in the official Spanish EQ-5D-5L. Participants rate their health on the day of assessment on a vertical scale from 0, representing the worst health they can imagine, to 100, representing the best health they can imagine; higher scores indicate better perceived health. Changes will be calculated separately as the 6-month score minus baseline and the 12-month score minus baseline. Positive values indicate improvement and negative values indicate worsening. The same official Spanish version, instructions, and administration procedures will be used at all assessments.

A fall will be defined as an unexpected event in which the participant comes to rest on the ground, floor, or a lower level. Falls will be recorded prospectively from randomization through 12 months using daily fall calendars returned monthly. Missing or unclear calendars will be followed up using standardized telephone contact applied equally across study groups, without medical-record review. For each fall, the date, circumstances, injury, mobility-aid use, and participant- or caregiver-reported healthcare use, including emergency attendance and hospitalization, will be recorded. The outcome is the total number of falls divided by observed follow-up time and expressed as falls per person-year. Separate falls experienced by the same participant will be retained as recurrent events. The principal fall comparison covers months 0-12; rates during months 0-6 and 6-12 will be supportive analyses.

Perceived functional social support will be assessed using the validated Spanish 11-item Duke-UNC Functional Social Support Questionnaire (Duke-UNC-11). The questionnaire assesses affective and confidant support. Each item is rated from 1, indicating much less support than desired, to 5, indicating as much support as desired. Item scores are summed to produce a total score ranging from 11 to 55; higher scores indicate greater perceived functional social support. Changes will be calculated separately as the 6-month total score minus baseline and the 12-month total score minus baseline. Positive values indicate improvement and negative values indicate worsening. The same Spanish version, administration method, item coding, and scoring procedures will be used at all assessments.

Total costs accrued per participant during the 12-month study period will be estimated from a societal perspective and expressed in euros for the prespecified study price year. Costs will include intervention delivery, personnel time, training, technology, facilities, materials and support; healthcare use; participant and caregiver time; and travel. Research-only costs that would not be required for routine implementation will be excluded. Resource use will be collected prospectively at prespecified intervals and valued using official Spanish unit-cost sources. Because the time horizon is 12 months, costs will not be discounted. A healthcare-system perspective will be examined in a sensitivity analysis.

Quality-adjusted life-years (QALYs) accrued during the 12-month study period will be estimated from EQ-5D-5L index values measured at baseline, 6 months, and 12 months using the Spanish value set. QALYs will be calculated using the area-under-the-curve method with linear interpolation between consecutive assessments. Higher values indicate more time lived in better health. Between-group comparisons of QALYs will be adjusted for baseline EQ-5D-5L utility. Missing data and deaths, if applicable, will be handled according to the prespecified health economic analysis plan.

Gut microbial alpha diversity will be assessed in stool samples using the Shannon diversity index derived from amplicon sequence variant profiles obtained through 16S rRNA gene sequencing. The Shannon index is unitless and incorporates both richness and evenness; higher values indicate greater within-sample microbial diversity but do not necessarily represent better health. Change will be calculated as the 6-month index minus baseline; positive values indicate increased diversity and negative values decreased diversity. DNA extraction, sequencing controls, minimum sequencing depth, filtering, normalization, batch handling, and the bioinformatics pipeline will be prespecified and applied consistently. Antibiotic exposure and relevant medication changes will be recorded for sensitivity analyses.

Overall gut microbial community composition will be assessed in quality-controlled stool samples collected at baseline and 6 months using 16S rRNA gene sequencing. The primary analysis will use centered-log-ratio-transformed abundance profiles and Aitchison distance. The intervention effect will be estimated as the group-by-time interaction in a prespecified repeated-measures permutational multivariate model. The permutation scheme will preserve paired observations within participants and treatment assignment at household level, with adjustment for the household-level baseline frailty stratum used in randomization. Bray-Curtis distance and one prespecified phylogenetic distance will be examined in sensitivity analyses. Results will be reported as global multivariate test statistics and effect-size measures. Taxon filtering, zero replacement, sequencing-depth, batch, covariate, and multiplicity rules will be frozen before analysis and treatment-code unblinding.

Systemic low-grade inflammation will be assessed using serum high-sensitivity C-reactive protein (hs-CRP), measured with a validated high-sensitivity laboratory assay and reported in mg/L. Higher concentrations indicate greater systemic inflammation but are not specific to frailty. Change will be calculated as the 6-month concentration minus baseline; negative values indicate reduced inflammation. Morning blood samples will be collected after an 8-12-hour overnight fast using standardized procedures. Acute infection, recent vaccination, vigorous exercise, and relevant medication use will be recorded. Paired baseline and 6-month samples will be allocated to the same analytical batch whenever possible, with study groups balanced across assay plates. Handling of values below the quantification limit and values above 10 mg/L will be prespecified, including sensitivity analyses excluding concentrations potentially reflecting acute inflammation.

The five binary criteria contributing to the Fried frailty phenotype will be evaluated separately: unintentional weight loss, exhaustion, weakness, slowness, and low physical activity. Definitions and cutoffs will be identical to those specified for the Fried phenotype total-score outcome. For each criterion, transitions between absence and presence and the proportion of participants meeting the criterion will be reported at baseline, 6 months, and 12 months. Improvement is defined as transition from present to absent; worsening is transition from absent to present.

Usual-paced gait speed will be obtained from the Short Physical Performance Battery 4-meter walk and reported in meters per second. Two trials will be performed; the faster valid time will be used, or the single valid trial when only one is available. Speed will be calculated as 4 divided by time in seconds. The usual walking aid is permitted and recorded. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate faster gait.

Five-repetition sit-to-stand completion time will be obtained from the chair-rise component of the Short Physical Performance Battery and reported in seconds. Participants will rise five times as quickly as possible from the standardized chair with arms folded. The outcome will be available only for a valid completed test. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; negative values indicate faster performance.

Self-reported habitual physical activity will be assessed with the validated Spanish short version of the Minnesota Leisure Time Physical Activity Questionnaire (VREM). Energy expenditure will be calculated from activity type, frequency, and duration and expressed as MET-minutes per 14 days using the prespecified scoring algorithm. Higher values indicate greater self-reported physical activity. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline.

Overall free-living movement volume will be summarized as average daily Euclidean Norm Minus One acceleration, expressed in milligravity units, from the non-dominant-wrist GENEActiv recording. Raw 60-Hz data will be processed using the same prespecified GGIR-compatible calibration, non-wear, and 24-hour normalization procedures at all assessments. Higher values indicate greater overall movement volume. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline.

The 24-hour composition of accelerometer-derived sleep, waking inactivity, light-intensity physical activity, and moderate-to-vigorous physical activity will be expressed as average minutes per day and normalized to 1,440 minutes. Waking inactivity will be used rather than sedentary time because wrist acceleration does not distinguish sitting from motionless standing. Components will be analyzed jointly using prespecified compositional-data methods and isometric log-ratio coordinates. No single direction of compositional change will be classified a priori as improvement.

Average nightly sleep duration will be estimated in minutes from continuous non-dominant-wrist GENEActiv data using the prespecified GGIR sleep-detection pipeline without a sleep diary. Sleep duration will be calculated within the algorithmically detected main sleep period time window after excluding detected wake. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline. Positive values indicate longer sleep duration but will not automatically be interpreted as improvement.

Wake after sleep onset will be estimated in minutes per night from continuous non-dominant-wrist GENEActiv data using the prespecified GGIR sleep-detection pipeline without a sleep diary. It represents time classified as awake between estimated sleep onset and the end of the main sleep period. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; negative values indicate less wake time during the sleep period.

Adherence to a Mediterranean dietary pattern will be assessed using the 14-item Mediterranean Diet Adherence Screener with the alcohol item excluded. The remaining 13 criteria will each score 0 or 1, producing a prespecified score from 0 to 13; higher scores indicate greater adherence. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate increased adherence.

Habitual dietary intake will be estimated using the validated 19-food-group exchange-based Food Groups Frequency Questionnaire. Prespecified dietary measures will include total energy, protein, carbohydrate, total fat, prespecified fatty-acid categories, fiber, and cholesterol when directly generated by the finalized scoring procedure for the administered questionnaire. Each dietary measure will be analyzed separately. Change will be calculated as the 6-month value minus baseline.

Five dietary components of the Full Mini Nutritional Assessment will be evaluated separately using the official response categories: recent decline in food intake or loss of appetite, number of full meals per day, consumption of protein-rich foods, consumption of fruit and vegetables, and daily fluid intake. Category changes will be reported at 6 and 12 months using the same administration and coding rules as the Full Mini Nutritional Assessment total-score outcome.

Maximum calf circumference will be measured in centimeters using the standardized Full Mini Nutritional Assessment anthropometric protocol. The same side, participant position, non-stretch tape, and measurement procedure will be used at all visits. Calf circumference will be analyzed as a continuous anthropometric measure and will not be labeled as muscle mass or sarcopenia. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate larger circumference.

Participants will complete a 7-day bowel diary at baseline and 6 months. The outcome is the number of spontaneous bowel movements during the 7-day recording period, defined as bowel movements not preceded by rescue laxative use according to the prespecified diary rule. Laxative use will be recorded for descriptive and sensitivity analyses. Change will be calculated as the 6-month weekly frequency minus baseline; positive values indicate more frequent spontaneous bowel movements.

Stool form will be recorded for every bowel movement during each 7-day bowel diary using the Bristol Stool Form Scale. The outcome is the percentage of recorded bowel movements classified as type 3 or type 4, calculated only when at least one bowel movement is recorded. Types 1-2 and 5-7 will be summarized as complementary categories. Change will be calculated as the 6-month percentage minus baseline; positive values indicate a greater proportion of stools in the prespecified type 3-4 category.

Urinary and fecal continence will be assessed separately using the bladder and bowel items of the validated Spanish Barthel Index. Each item is scored 0, 5, or 10 according to the standard categories; higher scores indicate greater continence and independence. For each item, change will be calculated separately as the 6-month score minus baseline and the 12-month score minus baseline. Transitions among the three response categories will be supportive analyses. The same administration and scoring rules will be used at all assessments.

The Montreal Cognitive Assessment Memory Index Score will be derived from delayed recall performance using free recall, category-cued recall, and multiple-choice recognition according to the published scoring algorithm. Scores range from 0 to 15; higher scores indicate better memory performance. Item-level learning and recall responses will be retained to permit exact calculation. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate improvement.

Published Montreal Cognitive Assessment cognitive domain indices will be calculated and analyzed separately: Executive Index Score, 0-13; Visuospatial Index Score, 0-7; Language Index Score, 0-6; Attention Index Score, 0-18; and Orientation Index Score, 0-6. Higher scores indicate better domain performance. The indices share some tasks and will not be treated as statistically independent. For each index, changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate improvement.

Phonemic verbal fluency will be quantified as the number of valid words produced during the 60-second letter-fluency task within the official Spanish (Spain) Montreal Cognitive Assessment version 8.1. Repetitions, proper nouns, and rule violations will be excluded according to official instructions. The raw word count will be retained rather than only the dichotomous Montreal Cognitive Assessment item score. Changes will be calculated separately as 6-month minus baseline and 12-month minus baseline; positive values indicate more valid words.

Participants will be classified at each assessment into six mutually exclusive states defined by robust, prefrail, or frail Fried phenotype status crossed with normal or impaired cognition. Impaired cognition will be defined as a Spanish age- and education-adjusted normalized Montreal Cognitive Assessment scalar score of 7 or lower, derived from the raw total using the published Spanish normative tables. The official education point will not be added before this normative conversion, avoiding double adjustment. Severe dementia, including severe Alzheimer disease, is an enrollment exclusion; participants with any documented dementia will not be classified as cognitively frail. Cognitive frailty is Fried score 3 or higher plus impaired cognition and no dementia. Follow-up transitions will be classified as incident, remitted, persistent, or absent, and the full six-state transition matrix will also be reported.

All-cause hospital admissions will be recorded prospectively from randomization through 12 months using participant or caregiver report collected at prespecified contacts. For each admission, the date and available reason will be recorded. The outcome is the total number of all-cause hospital admissions per participant during observed follow-up; a rate per person-year will be used when follow-up time differs. Repeated admissions for the same participant will be retained as recurrent events.

Cost-utility will be assessed by combining adjusted mean societal costs and QALYs accumulated over 12 months. Incremental cost-effectiveness ratios will be calculated as the difference in mean costs divided by the difference in mean QALYs for each active intervention versus usual care and through a complete incremental comparison of the four alternatives, accounting for dominance and extended dominance. Uncertainty will be characterized using 10,000 nonparametric bootstrap replications, cost-effectiveness planes, cost-effectiveness acceptability curves, and incremental net monetary benefit across willingness-to-pay thresholds relevant to Spain. Sensitivity analyses will examine alternative costing assumptions and the healthcare-system perspective.

Adherence in each active arm will equal 100 multiplied by completed-equivalent units divided by 10 prescribed units. In-person: 4 core sessions plus 6 follow-up calls; hybrid: 4 core sessions, 1 platform-setup session, and 5 monthly digital reinforcement periods; online: 1 setup session, 4 core digital modules, and 5 monthly reinforcement periods. Units score 1 if fully completed, 0.5 if partially completed, and 0 otherwise. Full completion requires at least 75% of a live session, completion of the call checklist, or at least 80% of required digital content; 25% to below these thresholds is partial. First-month units are accepted through day 42; monthly units through 14 days after their target window. Documented study-platform failures do not penalize participants; completion is rescheduled or, if delivery is impossible, the unit is removed from numerator and denominator for all affected participants. Usual care receives no score.

Participants in the three active intervention groups will complete the 8-item Client Satisfaction Questionnaire at the end of the intervention. Each item is scored from 1 to 4 and item scores are summed to produce a total score from 8 to 32; higher scores indicate greater satisfaction with the intervention. The same validated Spanish version, administration method, and scoring rules will be used across active intervention groups.

Qualitative data will be collected from participants in the active intervention groups at the end of the intervention through brief semi-structured interviews or focus groups. Topics will include the most useful elements, main difficulties, factors that facilitated participation, satisfaction, and suggestions for improving delivery. Data will be captured using the prespecified procedure, with participant validation of structured notes when applicable, and analyzed using thematic content analysis. Themes and supporting frequencies where appropriate will be reported by intervention modality.

Mobile-device proficiency will be assessed using the validated Spanish adaptation of the 16-item Mobile Device Proficiency Questionnaire. Items are rated from 1, never tried, to 5, very easily. Within each of eight domains, item ratings are averaged; the eight domain scores are summed to produce a total score from 8 to 40. Higher scores indicate greater mobile-device proficiency. Change will be calculated as the 6-month total minus baseline; positive values indicate increased proficiency. The measure will be administered to all study groups.

Tumor necrosis factor alpha will be measured in a blood-derived specimen using a validated laboratory assay and reported in the assay's validated concentration units. Morning fasting samples will be collected using standardized procedures, and paired specimens will be analyzed together whenever feasible. Change will be calculated as the 6-month concentration minus baseline; negative values indicate reduced concentration. Matrix, manufacturer, platform, quantification limits, batch allocation, acute-inflammation rules, medication handling, and treatment of values outside assay limits will be documented in a dated laboratory manual before testing and treatment-code unblinding and reported in the publication.

Total antioxidant status will be measured in a blood-derived specimen using a validated laboratory assay and reported in the assay's validated units. Higher values indicate greater measured antioxidant capacity but will not automatically be interpreted as better clinical status. Morning fasting samples will be collected using standardized procedures, and paired specimens will be analyzed together whenever feasible. Change will be calculated as the 6-month value minus baseline. Matrix, manufacturer, platform, reporting units, quantification limits, batch allocation, and treatment of values outside assay limits will be documented in a dated laboratory manual before testing and treatment-code unblinding and reported in the publication.

Fasting glucose, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, and triglycerides will be measured in milligrams per deciliter using routine validated laboratory methods. Each analyte will be analyzed separately. Morning samples will be collected after an 8-12-hour overnight fast using standardized procedures. For each analyte, change will be calculated as the 6-month concentration minus baseline. Direction of change will be interpreted according to the specific analyte rather than assigning a single common direction of benefit.

Interleukin-6 will be measured in a blood-derived specimen using a validated laboratory assay and reported in the assay's validated concentration units. Morning fasting samples will be collected using standardized procedures, and paired specimens will be analyzed together whenever feasible. Change will be calculated as the 6-month concentration minus baseline; negative values indicate reduced concentration. Matrix, manufacturer, platform, quantification limits, batch allocation, acute-inflammation rules, medication handling, and treatment of values outside assay limits will be documented in a dated laboratory manual before testing and treatment-code unblinding and reported in the publication.

Gut microbial richness will be assessed as the number of observed amplicon sequence variants in quality-controlled stool samples analyzed using 16S rRNA gene sequencing. The same DNA extraction, sequencing, quality-control, filtering, minimum-depth, batch-handling, and bioinformatics procedures will be applied at baseline and 6 months. Change will be calculated as the 6-month value minus baseline; positive values indicate increased richness and negative values indicate decreased richness. Increased richness will not automatically be interpreted as improved health. Antibiotic exposure and relevant medication changes will be recorded for prespecified sensitivity analyses.

Baseline-to-6-month changes in the relative abundance of a hypothesis-driven panel of gut bacterial taxa will be assessed using quality-controlled 16S rRNA gene sequencing and prespecified compositional repeated-measures differential-abundance methods. The panel will include Faecalibacterium, Roseburia, Anaerostipes, Agathobacter, Akkermansia, Bifidobacterium, Prevotella, Blautia, Coprococcus, Dorea, Collinsella, Eggerthella, Escherichia-Shigella, Ruminococcus sensu stricto, and Mediterraneibacter gnavus. Taxa will be evaluated at genus level and at species level only when the finalized 16S workflow supports reliable taxonomic assignment. The same laboratory and bioinformatics workflow will be applied to all samples. Direction of change will not be classified a priori as beneficial or harmful. False-discovery-rate control will be applied across the prespecified taxon panel.