Participants who meet the eligibility criteria will be considered eligible for randomization and will have a baseline evaluation scheduled. To characterize the sample, the initial assessment records information, such as sex, age, anthropometric data, family history, medication use, educational level, occupation, and comorbidities.

All study participants will have to attend 2 face-to-face visits: one at baseline (week –1) and one at the final assessment (week 6). In addition, participants will be assessed by a nutritionist and receive nutritional guidance and diet plans (week 0). The treatment will last 5 weeks. A detailed schedule is available in Table 1.

Initial information (lifestyle and inclusion criteria) will be collected online using a screening form link for men and women. Body mass and height measurements will be taken using a calibrated scale. Participants will be advised to be barefoot and wear light clothing. Height is measured with the participant standing and using a stadiometer. A member of the research team (ACAMS) is responsible for carrying out all study assessments as well as screening and recruiting participants, minimizing bias.

Ultrasound will be used as the main end point to assess abdominal subcutaneous adipose tissue (SAT) [4]. Intrarater reliability was assessed in previous studies [24]. A GE HealthCare Venue 40 ultrasound (model NZCART; GE Medical Systems, Ltd) is used. Subcutaneous fat, defined as the distance in centimeters between the dermis and the external surface of the fascia of the abdominal muscles, is measured using a linear transducer with a frequency of 12 MHz placed transversely 1 cm above and 1 cm below the umbilical scar [25,26]. Participants will be examined in the supine position without any artifact compressing the abdominal region, with their knees bent, and instructed to breathe in and out. A total of 3 measurements will be taken, each at the end of an exhalation, with minimal pressure on the abdominal cavity for greater measurement precision [27,28].

To determine total body mass, fat mass percentage per region, and visceral fat, we use dual-energy x-ray absorptiometry (DXA; Hologic Discovery A). This assessment is considered the gold standard for evaluating body composition in individuals with overweight and obesity [29]. In addition, DXA is used to classify participants as overweight or obese (grade I and II) [30]. Considering that hormonal variation throughout the menstrual cycle can influence body composition results, especially through fluid retention and water redistribution, all female participants will be assessed during the follicular phase of the menstrual cycle, a phase in which there is greater hormonal stability and a reduction in extracellular fluids, providing greater accuracy in body composition measurements [31].

Participants must fast for 4 hours, refrain from vigorous exercise for at least 12 hours, avoid caffeine and alcohol for the previous 24 hours, and consume a normal meal the night before the test. They should not wear any metal objects or accessories during the test to prevent interference with the total body composition results. The participant is placed in a supine position with internal rotation of the thighs, legs, and feet and instructed not to speak or move during the examination. The results obtained will be transmitted to a computer connected to the equipment, where the report shows the data in grams and percentages and which will be used to analyze total lean mass (kg), total bone mass (percentage), total fat mass (g), fat mass index (kg/m²), fat mass in the android and gynoid regions (g), and visceral fat (g) and, through calculations, deduce subcutaneous fat [32,33].

Venous blood samples will be collected in the cubital fossa by a specialized professional after the participants have fasted for 3 hours following a standardized meal. Blood samples will be taken 3 times: before and immediately after the TUS+Aussie current session and immediately after HIIT. Each tube of blood collected will be left to stand at room temperature for 30 minutes. They will then be centrifuged at 1450 × g for 10 minutes (Sorvall ST 8 Benchtop Centrifuge; Thermo Scientific). The serum is divided into 3 aliquots in 1.5-mL microtubes using 5-mL disposable pipettes and stored immediately at −80 °C until further analysis. The serum samples will then be thawed at the UFSCar Department of Chemistry and sent for analysis using the hydrogen nuclear magnetic resonance technique. Serum glycerol, total cholesterol, very low–density lipoprotein, low-density lipoprotein (LDL), high-density lipoprotein, and triglycerides will be assessed. These will be measured using wet chemistry (except for LDL, which will be calculated using the Friedwald equation; ADVIA 1800; Siemens). The experimental procedures for this analysis will follow those described in the work by De Favari Signini et al [34] and Castro et al [35].

To ensure the quality of the electrocardiogram (ECG) signal collection, the recommendations suggested by Catai et al [36] will be followed. All participants will be instructed to have a regular night’s sleep the night before, avoid alcoholic and caffeinated drinks, refrain from strenuous physical exercise at least 24 hours before and on the day of the assessment, and avoid heavy meals up to 2 hours before the assessment. All procedures will be explained beforehand to familiarize participants with the equipment and the assessor. The experiment will be carried out in the afternoon in a quiet room with the minimum number of people in the Cardiovascular Physical Therapy Laboratory, with controlled temperature and relative humidity (22-23 °C and 40%-60%, respectively) [36]. Before starting data collection, the participants will rest for 10 to 15 minutes to stabilize the signals [36]. After this, they will remain in the supine position for 15 minutes and then perform active postural change, remaining in the orthostatic position for 15 minutes. Participants will be instructed to breathe spontaneously and not to move or speak during the experiment.

RR intervals will be captured via ECG (Bio Amp FE132; ADInstruments) from the modified shunt in the fifth left intercostal space lead. In addition, respiratory movements will be captured using a respiratory belt positioned around the participants’ chest (Marazza) to obtain the respiratory rate during the test. Arterial pressure (AP) will be measured continuously beat to beat using a photoplethysmograph (Finometer PRO; Finapres Medical Systems) on the middle finger of the right hand. RR interval signals, AP, and respiratory movement will be sampled at 1 kHz using a commercial acquisition device (PowerLab 8/35; ADInstruments).

Classic linear indexes of HR variability will be calculated in the time and frequency domains. The indexes calculated in the time domain will be the SD of the NN interval, which represents sympathetic and parasympathetic modulation together, and the square root of the mean of successive NN intervals, which indicates parasympathetic modulation [37,38]. After detecting the QRS complex on the ECG, the apex of the R wave will be identified using parabolic interpolation. The heart period (HP) will be calculated as the temporal distance between 2 consecutive parabolic apexes. A stable sequence of 256 beats (for systolic AP [SAP] and HP) will be chosen in the supine and orthostatic positions, and if isolated ectopic beats are present, they will be linearly interpolated [39]. From these 256 points, the mean and variance of the HP and the mean and variance of the SAP will be calculated in the time domain.

The variability parameters of the HP and SAP will be evaluated according to the autoregressive model [39]. The spectral components will be broken down into LF (from 0.04 to 0.15 Hz) and high frequency (>0.15 to 0.4 Hz) reported in absolute units. The power of the spectral components will be expressed in absolute units or normalized units [38].

Symbolic analysis will be used to evaluate the HP and SAP series as defined in the study by Porta et al [40]. This technique is based on the 6-level uniform quantization procedure applied to HP and SAP series, which transforms series into sequences of symbols (from 0 to 5) from which patterns of 3 consecutive symbols will be constructed [41]. All possible patterns will be grouped without loss of information into families according to the number and type of variations between subsequent symbols: (1) 0V pattern—all symbols are the same; (2) 1V pattern—2 underlying symbols are the same, and the rest are different; (3) 2LV pattern—the 2 variations between adjacent symbols have the same sign; and (4) 2UV pattern—the 2 variations between adjacent symbols have opposite signs. Percentages of patterns within each experimental session were evaluated and denoted using 0V%, 1V%, 2LV%, and 2UV%, respectively. The 0V% index will be understood as a marker of sympathetic modulation when calculated in the HP and SAP series, and the 2UV% index will be understood as an index of vagal modulation when calculated in the HP series [40,42,43].

The baroreflex system will be evaluated through cross-spectral analysis using a bivariate autoregressive parametric approach [44]. The time-series relationships between AP and HP will be represented as coherence, phase, and transfer function gain. The coherence function shows the degree of association between the HP and SAP variabilities [44,45]. Its values range from 0 to 1, where values closer to 1 represent better coupling between the signals. The phase function represents the temporal relationship between the series (ie, the delay between the change in the SAP signal resulting from the heartbeat) and will be evaluated in high frequency and LF [44].

Sleep quality is assessed using the Pittsburgh Sleep Quality Index for sleep quality, adapted for Brazilians by Bertolazi et al [46]. It is a reliable instrument (intraclass correlation coefficient=0.65) [47] that assesses 7 components of sleep: subjective quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, medication use, and daily dysfunction. For each component, the score ranges from 0 to 3, and the sum yields a maximum score of 21. Scores of >5 indicate poor sleep quality.

Anxiety and depression will be assessed using the Hospital Anxiety and Depression Scale validated for Brazilians [48]. It consists of 14 items divided into 2 domains (depression and anxiety) with 7 items each. The items are scored on a 4-point Likert scale (from 0 to 3), resulting in total scores ranging from 0 to 21 for anxiety and depression. The cutoff points indicating moderate to severe symptoms are ≥8 for the anxiety domain and ≥9 for the depression domain.

Stress is assessed using the Depression, Anxiety, and Stress Scale, which measures the levels of these disorders based on behaviors in the previous 7 days. The items use a 21-question Likert scale, which has been translated into Brazilian Portuguese, to rate the frequency or severity of the participants’ experiences on a 4-point scale [48].

The Body Shape Questionnaire, validated for the Brazilian population [49], is a self-administered questionnaire consisting of 34 items designed to measure satisfaction and concerns about body shape. It is organized on a 6-point scale: 1=never, 2=rarely, 3=sometimes, 4=frequently, 5=very frequently, and 6=always. The score is the sum of the answers, where ≤110 points indicates no concern, ≥111 to ≤138 points indicates mild concern, ≥139 to ≤167 points indicates moderate concern, and >168 points indicates serious concern [49]. This questionnaire helps assess potential changes in satisfaction or concern about the body after the proposed treatment.

Participants will be referred to a specialized clinic for blood collection. They will be instructed to fast for 10 hours before the test. The following will be collected: complete blood count, glycated hemoglobin (through high-performance liquid chromatography), fasting glycemia, insulin resistance (through enzymatic and chemiluminescent immunoassay), evaluation of the homeostatic model, C-reactive protein (through immunoturbidimetry), homocysteine (through chemiluminescence), total cholesterol, high-density lipoproteins, and LDLs and triglycerides (through the enzymatic method). The tests will be clinically relevant in the context of obesity and abdominal adiposity and may contribute to future studies to determine whether any changes occur in these variables after the intervention. The results will be reported by the unit’s physician and sent to the researcher.

The waist-to-hip ratio is a calculation based on waist and hip circumference measurements [50]. It is often used to assess the risk of developing diseases such as high cholesterol, diabetes, and high blood pressure [50]. The presence of these cardiovascular risk factors, associated with increased abdominal fat, leads to a high risk of diseases such as myocardial infarction, stroke, and hepatic steatosis and mortality. The values considered normal for male and female individuals are <0.95 and <0.80 cm, respectively [50]. The waist-to-height ratio, on the other hand, is an index that assesses health risks by indicating the proportionality between waist circumference (a possible indicator of central fat accumulation) and height, suggesting a risk of cardiovascular or metabolic diseases [51]. Values of <0.5 cm indicate a lower risk of diseases, whereas values of >0.5 cm indicate a higher risk [51]. The measurements will be taken with the participant standing upright wearing light clothing. The waist circumference is measured as the average distance between the last floating rib and the anterior superior iliac crest [52]. The hip circumference is measured around the widest part of the buttocks at the level of the greater trochanter of the femur [52]. All measurements will be taken twice using an inextensible tape measure, and if there is a difference of 3 cm, a third measurement is taken for both waist and hips [52]. The ratio between the measurements is calculated as the waist circumference divided by the hip circumference.

Neck circumference is measured using an inextensible tape measure perpendicular to the longitudinal axis of the neck over the thyroid cartilage [53]. Participants will be in an anatomical position, either standing or sitting, with their head in the Frankfurt plane and their shoulders relaxed and in inspiratory apnea [53]. The cutoff points for overweight and obesity will be 37 cm for men and 34 cm for women [54]. The values above these cutoffs indicate increased cardiovascular risk factors [54].

All participants will be characterized by their level of physical activity at a single point in time (before the intervention). The long version of the International Physical Activity Questionnaire is used. The International Physical Activity Questionnaire is a validated instrument for the Brazilian population that assesses the level of physical activity. It has 27 questions related to physical activities performed in a week at different intensities—light, moderate, and vigorous—lasting at least 10 minutes continuously. The questions cover 4 dimensions: work, transportation, domestic activities, and leisure, as well as the time spent seated [55]. The score is calculated by adding the workload for each subitem separately [55]. This questionnaire helps classify the level of physical activity before the proposed interventions.

Before the intervention, a treadmill ergometric test is conducted using the Ellestad protocol [56] for clinical cardiological assessment and to assist in prescribing HIIT. The test is interrupted according to the criteria recommended by the third version of the Sociedade Brasileira de Cardiologia ergometry guidelines [57]. The original Borg scale is used to assess the participants’ perception of dyspnea, lower-limb fatigue, and angina during exercise. This scale ranges from 0 to 10, where 0 is no dyspnea and 10 is the most intense dyspnea, fatigue, and angina [58]. The following determines the presence of maximum or peak effort: peak HR (HR_peak)≥85% of that predicted for age (220 – age) and Borg perceived exertion of 18 on a scale from 6 to 20 (patient exhausted) [59].

The nutritional intervention will be conducted by an experienced nutritionist. All participants receive nutritional guidance 30 days before the combined therapy and HIIT. This initial intervention aims to adjust macronutrient intake, optimize energy consumption, and improve diet quality by reducing the consumption of ultraprocessed foods and increasing the consumption of minimally processed foods. Macronutrient adjustments follow the recommendations of the dietary reference intakes, with carbohydrates comprising 45% to 65%, proteins comprising 10% to 35%, and lipids comprising 20% to 30% of the total caloric intake [60].

Following this phase, participants undergo a new series of assessments and then begin 5 weeks of combined therapy and HIIT. During these 5 weeks, dietary control is monitored by the nutritionist to ensure consistency in dietary patterns throughout the study. Daily energy intake is considered stable if variations remain at <10% during the protocol.

Food intake will be assessed weekly using a 24-hour recall, a questionnaire that quantifies all food and drink consumed in the previous 24 hours. Participants will receive an electronic link once a week during the 5-week intervention. The data collected will be analyzed by a nutritionist to assess consistency and adherence in total energy and macronutrient intake. Variations of >10% in total calorie intake will be discouraged to ensure dietary stability throughout the protocol. No additional methods of validating adherence to the diet will be used as the weekly recalls provide a feasible and easy monitoring strategy for the participants. Total energy intake and macronutrient composition will be calculated using the Dietbox program [61]. Energy requirements will be estimated based on the dietary reference intakes [62]. For qualitative analysis, foods will be categorized according to their processing level—natural, minimally processed, culinary ingredients, processed, and ultraprocessed—as defined by the 2014 Food Guide for the Brazilian Population [63].

Before the first session, the application time for the combined therapy will be individually determined. Measurements will be conducted by the same blinded assessor using an inelastic tape measure with the participant in a standing position. Specifically, the horizontal distance between the last floating ribs (x) and the vertical distance to the midpoint between the anteroinferior iliac crests (y) will be measured. The abdominal area (cm²) will be calculated by multiplying these 2 values (x × y) and then dividing the result by 18. This divisor corresponds to the total effective radiating area of the ultrasound head used in the therapy, which consists of 3 spheres, each with an effective radiating area of 6 cm². Dividing the abdominal area by 18 allows us to determine the time required to apply the therapy to each participant according to the following formula: application time = (x × y)/18.

For example, if a participant has a horizontal distance (x) of 21 cm and a vertical distance (y) of 26 cm, the area will be calculated as follows: 21 × 26 = 546 cm². Dividing this area by 18 results in approximately 30 minutes of application time (546/18 = 30.33 minutes).

Before each training session, participants will be instructed to (1) wear comfortable clothing and exercise-appropriate footwear, (2) ensure adequate sleep the previous night, (3) hydrate with water 2 hours before the session and bring a personal water bottle, and (4) avoid strenuous physical activity and alcoholic consumption within 24 hours before the session.

Food consumption will also be recommended 24 hours before the first and last sessions, as described in the Nutritional Intervention section.

To minimize data loss, participants receive reminders about the date and time of each session. Physiotherapists responsible for interventions may also call participants if necessary to confirm appointments. For evaluations and re-evaluations, the blinded evaluator contacts the participants via telephone. At the end of the 10 treatment sessions, the re-evaluation is scheduled between 3 and 15 days after the last session.

The primary outcomes involve a comparison (between groups) of the effect of 10 sessions of combined therapy (TUS+Aussie current) alone versus in combination with HIIT, specifically on abdominal SAT thickness, body composition, metabolic profile, and serum glycerol release.

Secondary outcomes assess the impact on CAM, sleep quality, anxiety, depression, body image concerns, and blood biochemical variables (Table 2).

The primary outcome, percentage of body fat, was used to calculate the required sample size using the RStudio software (version 12.1; Posit PBC) assuming a moderate expected effect size of ƒ≥0.15 and a significance level of 5% and to promote a statistical power of 80%. A total of 18 participants per group is necessary to achieve 80% statistical power with an α level of 5%. Considering a dropout of 10%, the target sample size was increased to 20 participants per group, with the total sample size being 60 individuals. Appropriate statistical corrections (eg, Bonferroni or false discovery rate) will be applied to the P values, recognizing the necessity of conducting multiple tests due to the analysis of additional results.

A CONSORT (Consolidated Standards of Reporting Trials)-style flowchart [68] will present the number of patients selected and all reasons for exclusion before randomization. Demographic and socioeconomic information, including age, sex, race, socioeconomic status, and educational level, as well as baseline clinical characteristics such as BMI, glycated hemoglobin, smoking status, alcohol consumption, and comorbidities, among other variables, will be described by study arm and overall. Categorical variables will be reported as absolute numbers and percentages. Continuous variables with a reasonably symmetrical distribution will be summarized as means and SDs. For highly skewed continuous variables, medians and IQRs will be used.

Data analysis will be conducted by a researcher blinded to the participants’ allocation to the groups. The data will be entered into a Microsoft Excel spreadsheet for subsequent analysis.

For quantitative variables, the assumption of normality of data distribution and homogeneity of variances will be checked using the Shapiro-Wilk test and Levene test, respectively. To compare participant characteristics at baseline between groups, a 1-way ANOVA or Kruskal-Wallis test will be applied when appropriate. For comparisons between groups involving categorical variables, the chi-square test or Fisher exact test will be used. To compare primary and secondary outcomes between and within groups, linear mixed models estimated using the maximum likelihood method with an appropriate covariance matrix structure will be used assuming group (AG, PG, and HG), moment (before and after the intervention) or time (baseline, after TUS+Aussie current, and after HIIT), and interactions (group-moment, group-time, moment-time, and group-moment-time) as fixed factors and participants as a random factor [66]. In addition, moment and time will be assumed as a repeated-measure effect. The model that fits the data will be the one with the minimum Akaike information criterion. When appropriate, the data may be subjected to a Box-Cox transformation to better adhere to a normal distribution [69,70]. Whenever a significant F value is obtained in the ANOVA or linear mixed models, a Sidak post hoc adjustment for the purpose of pairwise multiple comparisons will be applied. The significance level will be set at 5% (P<.05). All analyses will be conducted using the PASW Statistics software (version 25.0; SPSS).