Participants who receive nutritional intervention and are enrolled in a high-intensity interval training exercise program.

Participants assigned to the HIIT completed 3 weekly sessions, with 1-2 days of rest between sessions, on a cycle ergometer for 12 weeks. The training load (TL) was determined from an incremental test until volitional exhaustion with one-minute work intervals (from 85% to 165% of maximal power output during VO2max test [Wmax] and with 10% increments) separated by 1.5-minute rest periods (at 20% Wmax). The program consisted of a 3 minutes warm-up with the first 2 minutes at 20% and last minute at 40% of the TL, followed by 10 series of 1 min duration at 90% of TL, with 1 minute of rest between sets at 20% of TL and ending with 2 minutes of cool-down at 20% of TL (estimated total time of the session: 25 minutes). A 5% increase in workload was applied when, for two consecutive sessions, the first three intervals failed to exceed 85%, and the last five did not reach 90% of maximal HR. Participants were asked to maintain a cadence above 80RPM.

Participants assigned to the MICT completed 3 weekly sessions, with 1-2 days of rest between sessions, on a cycle ergometer for 12 weeks. The MICT program consisted of 50 minutes of continuous pedaling at an intensity approximately 10 % above the lactate threshold, with cadence maintained between 60-80 rpm. The training load was increased by 10% if these two conditions were met: 1) no increase of at least 2 beats of heart rate from minute 20 to minute 40 of the training session, and 2) lactate levels below 2 mmol/L in minute 48 of the session. The training load for the first session was determined by an adjustment test consisting of an incremental test with five consecutive 10-minute phases at a constant cadence (60-80 RPM) from 90% to 130% of the ventilatory threshold power, calculated from the FatOx test data, with 10% increments in each step. was determined by the phase in which HR increased by ≥ 3 beats between phases and blood lactate levels increased by ≥ 1 mmol/L between phases.

Insulin sensitivity was estimated through the Quantitative Insulin Sensitivity Check Index (QUICKI), calculated as 1 / [log(fasting insulin (µU/mL)) + log(fasting glucose (mg/dL))].

Fasting blood samples (fast 8-10 hours) were centrifuged (1500×g, 15 min, 4 °C) and the resulting plasma was divided into aliquots (500 µL) and stored at -80°C. Exosomes will be isolated from 600 µL of plasma using the miRCURY Exosome Kits (Qiagen, Netherlands). Total RNA was extracted from 200 μL of exosome samples using the miRNeasy Serum/Plasma Advanced Kit (Qiagen). Reverse transcription was performed with the miRCURY LNA RT Kit, incubated at 42°C for 60 min, 95°C for 5 min, and stored at -20°C. Quantitative PCR was carried out using the miRCURY LNA SYBR Green PCR Kit and miRNA-specific assays on a CFX Connect System (Bio-Rad). MiRNA expressions (Cq values) were normalized using reference miRNAs. Relative expression was determined with the 2^-ΔCq method, and fold changes were calculated using the 2^-ΔΔCq method, ensuring accurate quantification of miRNA expressions from exosomal RNA.

Appetite-regulating hormones will be assessed from blood samples collected after an 8-10-hour overnight fast, with participants seated at rest for 15 minutes. Fasting samples will be drawn from the antecubital vein using EDTA tubes for appetite hormones and CAT tubes for serum cortisol. EDTA tubes will be inverted, centrifuged at 1500 × g for 15 minutes at 4 °C, and plasma aliquoted into 500 µL tubes with 5 µL protease inhibitors, then stored at -80 °C. Ghrelin, leptin, GLP-1, and GIP will be measured fasting and 90 minutes post-75 g glucose using the MILLIPLEX® Metabolic Hormone Panel V3 with a Luminex® xMap system; intra- and inter-assay CVs <10% and <20%. Cortisol from CAT tubes allowed to clot 30 minutes, centrifuged at 2000 × g for 10 minutes at 4 °C, aliquoted into 200 µL tubes, stored at -80 °C, and quantified using one-step CMIA on Alinity CI; emitted light inversely reflects cortisol, range 1-59.8 µg/dL, intra- and inter-assay variation <10%.

Emotional reactivity to food images will be measured using a high-density 128-channel EEG system (HydroCel Sensor Net, EGI, Inc.) during both resting state and a visual-attentional cognitive task. EEG caps will be fitted according to standardized landmarks (GND at the forehead, REF at the vertex, lateral electrodes at the eyebrows and top of ears, posterior electrode at the occipital bone), with impedance checked and adjusted by pipetting or gentle repositioning. Participants will be seated 60-70 cm from a 15'' monitor in dim lighting, with eyes aligned to the center of the screen. A 4-minute eyes-open resting-state recording will be conducted to assess baseline alpha activity (8-14 Hz). The experimental task will present images in blocks with breaks, including high- and low-calorie foods, non-food controls, and, in a separate task, images categorized as beneficial or harmful for gut microbiota based on fiber, prebiotic/probiotic content, caloric density, fat, and meat content.

Since appetite feelings modulate nutritional behavior, evaluating appetite among the intervention groups can improve the study's quality from a comprehensive perspective. After an 8-10 hour fasting period, a Visual Analogue Scale (VAS) will be completed to measure the participant's morning appetite. Likewise, they will be evaluated 90 minutes into the Oral Glucose Tolerance Test (OGTT) and at the end of it. The VAS is a valid measurement instrument for assessing the feeling of appetite.

Body weight (kilograms, kg) was measured using a multifrequency bioelectrical impedance device (TANITA-MC780MA, Tanita Corp., Tokyo, Japan) after at least 8 hours of fasting. Participants were instructed the day before testing to maintain their usual lifestyle, refrain from alcohol or caffeine for 24 hours, and avoid vigorous physical activity for 72 hours. Participants wore light clothing and were asked to urinate immediately before the measurement.

Height (meters, m) was measured using a wall-mounted stadiometer (TANITA-LEICESTER HR-001, Tanita Corp., Tokyo, Japan) with participants standing in accordance with the manufacturer's instructions.

Body Mass Index (BMI) in kilograms per square meter (kg/m²) was calculated as body weight (kg) divided by height squared (m²).

Fat mass (kilograms, kg) was estimated using multifrequency bioelectrical impedance with an 8-electrode device (TANITA-MC780MA, Tanita Corp., Tokyo, Japan) after at least 8 hours of fasting. Participants were instructed the day before testing to maintain their usual lifestyle, refrain from alcohol or caffeine for 24 hours, and avoid vigorous physical activity for 72 hours. Participants wore light clothing and were asked to urinate immediately before the measurement.

Fat mass (percentage of total body weight, %) was estimated using multifrequency bioelectrical impedance with an 8-electrode device (TANITA-MC780MA, Tanita Corp., Tokyo, Japan) after at least 8 hours of fasting. Participants were instructed the day before testing to maintain their usual lifestyle, refrain from alcohol or caffeine for 24 hours, and avoid vigorous physical activity for 72 hours. Participants wore light clothing and were asked to urinate immediately before the measurement.

Fat-free mass (kilograms, kg) was estimated using multifrequency bioelectrical impedance with an 8-electrode device (TANITA-MC780MA, Tanita Corp., Tokyo, Japan) after at least 8 hours of fasting. Participants were instructed the day before testing to maintain their usual lifestyle, refrain from alcohol or caffeine for 24 hours, and avoid vigorous physical activity for 72 hours. Participants wore light clothing and were asked to urinate immediately before the measurement.

Fat-free mass (percentage of total body weight, %) was estimated using multifrequency bioelectrical impedance with an 8-electrode device (TANITA-MC780MA, Tanita Corp., Tokyo, Japan) after at least 8 hours of fasting. Participants were instructed the day before testing to maintain their usual lifestyle, refrain from alcohol or caffeine for 24 hours, and avoid vigorous physical activity for 72 hours. Participants wore light clothing and were asked to urinate immediately before the measurement.

Waist circumference (cm) was measured at the midpoint between the lower lateral margin of the 10th rib and the iliac crest, perpendicular to the longitudinal axis of the trunk. Abdominal circumference (cm) was measured at the level of the umbilicus, perpendicular to the longitudinal axis of the trunk. Hip circumference (cm) was measured at the level of the greatest posterior prominence of the buttocks, perpendicular to the longitudinal axis of the trunk. Thigh circumference (cm) was measured at the midpoint between the greater trochanter and the lateral tibial condyle. All measurements were performed with the participant standing upright, and relaxed. Each measurement was taken twice, and a third measurement was performed if the difference exceeded 1%. The mean of the two most similar records was registered.

Waist-to-hip ratio (WHR) was calculated as waist circumference (cm) divided by hip circumference (cm).

Thigh skinfold thickness was measured at two-thirds of the distance from the iliospinale point (the lowest part of the anterior superior iliac spine) to the superolateral border of the patella, over the belly of the vastus lateralis muscle. Measurements were taken on the dominant leg. Two measurements were obtained, and a third was taken if the difference between the first two exceeded 5%. The mean of the two most similar records was recorded.

Nutrient intakes were estimated using the DIAL software for Windows (version 1.19; Ortega et al., 1999), a validated tool for quantitative and qualitative assessment of dietary intake in the Spanish population. The software converted food consumption into daily energy and nutrient values, expressed absolute intake (g/day), percentage of Recommended Intake (RI) coverage, and the energy distribution profile (percentage of total energy from proteins, fats, carbohydrates, and alcohol).

Adherence to the Mediterranean Diet was evaluated using a validated 14-item questionnaire for the Spanish population (3), with scores ranging from 0 (lowest adherence) to 14 (highest adherence). The resulting score was classified into three categories: low adherence (range: 0-5), moderate adherence (range: 6-9), and high adherence to MedDiet (range: 10-14).

Physical activity and inactivity will be assessed using triaxial accelerometers (GENEActiv; ActivInsights Ltd., Kimbolton, UK) placed on the non-dominant wrist for at least six consecutive days (direct method). Participants will be instructed to wear the device at all times, including during sleep and water-based activities, to ensure comprehensive data capture. Valid data will be defined as a minimum wear of 16 hours per day for at least four days (three weekdays and one weekend day) following established protocols (10.1177/07334648231218095). Accelerometers will be set to 60 Hz (10.1111/sms.12795, 10.1249/MSS.0000000000000289). Raw data will be downloaded using GENEActiv software version 3.3 and processed using the GGIR package (version 3.0.9) in R (v4.1.1, R Core Team, Vienna, Austria), which auto-calibrates the data based on local gravity and calculates the Euclidean Norm Minus One (ENMO) to minimize sensor calibration error (https://doi.org/10.1152/japplphysiol.00421.2014).

Physical activity, sedentary time, and sleep will also be estimated through self-reported questionnaires, using the International Physical Activity Questionnaire Short Version (IPAQ-SF), previously validated in Spain. The IPAQ-SF will provide estimates of sedentary time, moderate and vigorous activity in minutes per week.

After at least 8 hours of fasting, resting energy expenditure was estimated by indirect calorimetry. Participants were instructed the day before testing to maintain their usual lifestyle, refrain from alcohol or caffeine for 24 hours, and avoid vigorous physical activity for 72 hours. Participants lay supine for 30 minutes wearing a mask connected to a gas analyzer (open circuit, Jaeger MasterScreen CPX® (CareFusion, San Diego, USA)). The resting rates of fat and carbohydrate oxidation were used to determine the resting energy expenditure in kcal per minute (kcal/min).

Maximal fat oxidation was determined through a gradual test on a cycloergometer (Lode Excalibur, Groningen, Netherlands). Gas exchange (VO2, VCO2, RER) was recorded using indirect calorimetry (Jaeger MasterScreen CPX®), and heart rate was measured continuously over the test with Polar Team 2 (Polar Electro Inc., Lake Success, NY). The test consisted of a starting load of 15W with increments of 15W every 3 minutes until the respiratory quotient reached a stable value of 1. Throughout the test, a cadence of 60 r.p.m. was maintained. Fat oxidation (g/min) values were estimated from oxygen and carbohydrate data averaged over the last minute of each 3-minute stage, using the Frayn stoichiometric equation, and then the higher fat oxidation value was recorded.

The relative intensity of maximal fat oxidation (FatMax) reflects the exercise intensity (% of maximal oxygen consumption (VO2max)) at which fat oxidation reaches its peak. FatMax was determined from the graded maximal fat oxidation (MFO) cycling test in a cyclo-ergometer (Lode Excalibur, Groningen, Netherlands). The test started at 15 W with 15-W increments every 3 min until the respiratory quotient reached 1, maintaining 60 rpm. Gas exchange (VO2, VCO2, RER) was measured via indirect calorimetry (Jaeger MasterScreen CPX®), and heart rate was continuously monitored (Polar Team 2, Polar Electro, NY). After the MFO test, a second incremental test with 15W/min increments until exhaustion from the last load of MFO test was performed to determine VO2max. Fat oxidation rates (g/min) were calculated using the Frayn equation and plotted against relative VO2 (%VO2max). The VO2 corresponding to maximal fat oxidation was identified as FatMax.

Maximal oxygen consumption (VO2max, ml/min) was assessed using an incremental test on a cycle ergometer (Lode Excalibur) following the maximal fat oxidation (MFO) test. Gas exchange (VO2, VCO2, and RER) was recorded using an indirect calorimeter (Jaeger MasterScreen CPX®), and heart rate was measured continuously over the test with Polar Team 2 (Polar Electro Inc). After a 5-min rest following the MFO test, the VO2max phase started at the last workload achieved in the MFO stage and increased by 15 W every minute, maintaining a pedaling cadence of 70-90 rpm until volitional exhaustion. Maximal effort was confirmed when at least three of the following criteria were met: a VO2 plateau (VO2-time slope <0.05 L·min-1 during the last 30 s), ≥90% of age-predicted maximal heart rate, a respiratory exchange ratio (RER) ≥1.10, perceived exertion of 8-10 on the Borg scale, and volitional fatigue. If fewer than three criteria were reached, the VO2 value was recorded as VO2peak.

Self-reported quality of life will be registered by applying the Short Form 36-health survey (SF-36) questionnaire previously validated in Spain. The SF-36 is a generic measure of quality of life and has been evaluated for a wide variety of medical conditions, including diabetes. The SF-36 includes 36 questions that evaluate 8 subscales, scores for each scale range from 0 to 100, with higher scores indicating higher level of function or wellbeing. The physical component summary is derived from the 4 subscales of physical functioning, role physical, bodily pain, and general health; while the mental component summary is derived from the subscales of vitality, social functioning, role emotional, and mental health.

Systolic and diastolic blood pressures will be recorded three times by using the validated Omron HEM 742 blood pressure monitoring device with the participant seated, back supported in chair and feet flat on floor without legs crossed, after they rested 5 minutes. The guidelines of the Hypertension and Cardiology European Societies and the Spanish one will be used to classify blood pressure stages of participants.