The trial protocol was approved by the Research Ethics Committee of the Islamic Azad University Mahabad Branch (IR.IAU.MAHABAD.REC.1404.012). The study was conducted in accordance with the principles of the Declaration of Helsinki [22]. All participants signed an informed consent prior to enrollment in the study, and each patient received a copy of the signed consent form. The sample trial was registered on ClinicalTrials.gov (NCT07302191).

A randomized controlled trial was conducted to evaluate the effects of 12 weeks of exercise interventions, comprising HIIT and CONC training, on health biomarkers in postmenopausal women with overweight or obesity. Participants were randomly allocated to one of the three groups using a stratified block design [23]. An independent statistician generated the randomization sequence, assigning participants to: a) HIIT group, b) CONC group, and c) control group (CG). CG did not receive any exercise program. Randomization followed the Consolidated Standards of Reporting Trials (CONSORT) 2025 guidelines for parallel-group randomized trials (Table S1) [24]. Baseline anthropometric and clinical characteristics of participants were assessed prior to the intervention and are presented in Table 1, ensuring comparability between groups.

The sample size was calculated using G*Power version 3.1, following Faul’s guidelines [25]. A total of 36 participants were required. To account for an anticipated 20% dropout rate, the final sample size was set at 45 participants (15 per group). In addition, a power analysis (1 – β error probability) indicated that a sample size of 36 participants would be sufficient to detect a minimum difference of 1% in lipid biomarkers (TG, LDL-C, or HDL-C). The statistical power was set at 0.95, with an estimated effect size of 0.8.

Each participant underwent two testing sessions under identical conditions, separated by a 12-week period (September 2, 3, and 4, 2025, and November 25, 26, and 27, 2025). Tests were standardized to begin at 9:00 am to minimize circadian variability, with procedures conducted following the same protocol, sequence, and timing. Anthropometric biomarkers, physical parameters, oxidative stress status, and cardiac metabolic indicators were assessed. Five days prior to the start of the study, participants received detailed instructions regarding the exercise protocols and the type of tests to be performed, ensuring appropriate preparation and progression.

Recruitment was conducted in July and August 2025 using convenience sampling through community advertisements in health centers, sports centers, and community networks in Mahabad (Iran). Postmenopausal women with overweight or obesity who responded to the invitations were contacted by phone call and provided with detailed information about the study, including its objectives, methodology, potential benefits, and associated risks.

A total of 45 postmenopausal women (aged 50–65 years) were included in the study. Eligibility criteria comprised BMI > 25 and ≤ 40 kg/m², sedentary behavior defined as less than 150 min/week of moderate physical activity according to World Health Organization (WHO) guidelines [26], being in generally good health and able to safely engage in exercise protocols and having at least one year since their last menstrual period, in accordance with standard clinical criteria for menopause. Exclusion criteria included medical contraindications to strenuous physical activity, joint pain, current use of hormone replacement therapy, missing more than three exercise sessions during intervention, or illness during the study period. Additionally, women taking drugs or supplements known to significantly affect weight, metabolism, or exercise response, as well as those with substance use disorders, were also excluded.

Accordingly, 45 women were selected and allocated in a 1:1:1 ratio to the HIIT group (n = 15), CONC group (n = 15), and CG (n = 15). However, 2 women from each training group (HIIT and CONC) were excluded due to excessive absenteeism, and one participant from the CG did not complete the post-test assessment (Figure 1). Absenteeism withdrawals were unrelated to exercise adverse events. Four participants discontinued the intervention due to scheduling constraints, preventing them from completing 3 consecutive training sessions. Randomization was performed using a computer-generated sequence created using the Research Randomizer software (version 4.0) [27] (random number generator software), concealed in sequentially numbered, sealed, opaque envelopes (SNOSE method), and kept by the principal researcher. In addition, due to the nature of the intervention, participant and trainer blinding was not possible. However, outcome data were coded, and the statistician remained blinded to group assignment throughout the analysis.

Both the HIIT and CONC groups followed a supervised 12-week exercise intervention, performed 3 times per week on alternate days. All training sessions were conducted in the Exercise Physiology Laboratory of the Islamic Azad University (Mahabad Branch) and adjacent training facilities under standardized environmental conditions (temperature 20–22°C; humidity 50–60%). All sessions were supervised by certified exercise specialists to ensure participant safety and protocol fidelity. Each session consisted of a 10-min warm-up (light aerobic activity and dynamic stretching), the main training component, and a 10-min cool-down (low-intensity walking and stretching). Attendance, heart rate responses, and perceived exertion were recorded in each session to verify compliance. Participants missing > 20% of sessions (i.e., more than seven of 36) were classified as non-completers. The exercise mode consisted of treadmill or outdoor walking/running, selected according to participant preference to minimize impact on the lower limbs. In this context, joint load was carefully considered in the design of all selected exercises, particularly in this last exercise (running/walking). Nevertheless, the possibility of adverse effects was systematically monitored at each session, and no joint-related issues were reported during intervention.

The HIIT group training program included 60 seconds of bouts at maximal intensity [90–95% of individual maximum heart rate (HRmax)], separated by 60 seconds at lower intensity (55–60% HRmax) [28]. HRmax was estimated individually using the Tanaka et al. equation [29] (HRmax = 208 − 0.7 × age), which has been validated in middle-aged women. Session duration ranged from approximately 25 to 35 min (excluding warm-up and cool-down).

Progression was achieved by gradually increasing the number of high-intensity bouts across the intervention period: i) week 1: 6 intervals; ii) week 2: 8 intervals; iii) weeks 3–8: 10 intervals; iv) weeks 9–12: 12 intervals. This progression model was adapted from Afrasyabi et al. [30], who applied a similar HIIT framework in adults with metabolic impairments.

To monitor exercise intensity during HIIT sessions, the women wore heart rate monitors (Polar Vantage M, POLAR, Oulu, Finland) throughout all training sessions. Heart rate data were recorded in an online personal training diary (https://flow.polar.com/diary) and used to assess adherence and intensity. Data were downloaded and reviewed weekly to ensure that the correct prescriptions of intensity thresholds were maintained.

To ensure consistent exertion, the Borg 6–20 rating of perceived exertion (RPE) scale was recorded at the end of each interval and maintained between 15 and 17 (“hard” to “very hard”), confirming alignment with the prescribed HR range.

This HR-RPE correspondence followed the intensity prescription criteria outlined by the American College of Sports Medicine [31] for older adult populations.

The CONC group participated in a combined resistance and endurance training program. Resistance training involved 3 sets of selected exercises at designated stations, with one-min rest intervals between sets. During the first week, CONC participants completed 10–15 repetitions at 55% of their one-repetition maximum (1RM). By week 12, participants progressed to 8–12 repetitions at 75% of 1RM. To ensure appropriate training load, 1RM was reassessed every three weeks and adjusted accordingly. Perceived exertion was recorded (Borg 6–20 scale), maintaining RPE 13–15 (“somewhat hard”) to confirm compliance with relative intensity.

The aerobic component of the CONC program consisted of 20 min of running. Intensity was gradually increased from 55% HRmax in the first week to 75% HRmax by the final week, following the protocol described by Kargarfard et al. [32]. The average total duration of each CONC session, including warm-up and cool-down, was approximately 50–60 min.

To maintain comparable total training volume across groups, energy expenditure per session (kcal/session) was estimated from heart rate recordings and adjusted weekly to equalize cumulative training load between HIIT and CONC.

The exercise mode consisted of treadmill or outdoor walking/running, selected according to participant preference to minimize impact on the lower limbs. In this context, joint load was carefully considered in the design of all selected exercises, particularly in this last exercise (running/walking). Nevertheless, the possibility of adverse effects was systematically monitored at each session, and no joint-related issues were reported during intervention.

Finally, participants in CG continued their usual daily activities without any follow-up during the 12-week intervention period. To minimize differences in participant-researcher contact, the CG received biweekly phone calls similar in frequency to the intervention groups. During these calls, participants completed a brief standardized checklist regarding daily activity, health status, and medication changes. Sedentary postmenopausal women do not routinely receive structured exercise programs, and maintaining their usual lifestyle does not expose them to additional risk. All participants were informed during consent that the CG would not receive supervised training during the study but would be offered individualized exercise guidance after completion. This approach was approved by the institutional Ethics Committee to avoid contamination between groups and preserve the internal validity of the trial.

Body composition measurements were taken before and after the 12-week intervention, under fasting conditions. A stadiometer (SECA 213, Hamburg, Germany) was used to assess the heights and weights of participants, wearing minimal clothing. BMI was calculated according to the formula: weight (kg)/height² (m²). WC, hip circumference (HC), and waist-hip ratio (WHR) were measured using a flexible measuring tape. To determine %BF using the Jackson et al. equation [33], subcutaneous fat thickness was measured at the triceps, suprailiac, and thigh sites using a skinfold caliper (Yagami, Japan; accuracy: 1 mm). All measurements were taken on the right side of the body three times, with 20-second intervals between each measurement. These values, along with age, were used to calculate body density, which was subsequently converted to %BF using the Siri equation.

Blood samples were collected in a laboratory setting at the beginning and at the end of the 12-week training program. Following a 12-h fast, venous blood was drawn from the left arm while participants were seated. After centrifugation (12 min at 3,000 rpm in a Hettich centrifuge, Germany), the resulting serum was aliquoted, frozen, and stored at –70°C for subsequent analysis.

Lipid profile indices, including total cholesterol (TC), TG, and HDL-C, were measured with an enzymatic method (Pars Azmun, Iran). LDL-C was calculated using the Friedwald equation [34], which estimates LDL-C by subtracting HDL-C and one-fifth of TG from TC. The atherogenic index of plasma (AIP) was determined using the logarithmic ratio of TG to HDL-C (log[TG/HDL-C]) [35].

Fasting glucose levels were measured using the Pars Azmun glucose kit (Pars Azmun, Iran), based on the glucose oxidase method, with a sensitivity of 5 mg/mL. Serum insulin concentrations were determined using an ELISA kit (Monobind, United States), with a sensitivity of 0.4 IU/mL. The HOMA-IR was calculated using the following formula with glycemia expressed in mmol/L and insulin levels in mIU/L [36]: HOMA-IR = (glucose × insulin)/22.5.

Participants were instructed to maintain their habitual diet throughout the intervention. Habitual diet was not strictly controlled, but participants were advised to avoid any significant dietary changes during the study period. No specific nutritional counseling was provided, as the primary objective of this study was to isolate the effects of exercise training. To minimize potential confounding, participants were asked to replicate their dietary intake during the 24 h preceding each testing session, ensuring consistency. Additionally, to reduce the impact of dietary factors on oxidative stress and inflammatory biomarkers, participants were instructed to avoid antioxidant supplementation (e.g., vitamin C, E, polyphenols) and alcohol consumption for at least 48 h before each testing session. These instructions aim to minimize potential interference with biomarker measurements sensitive to dietary or exogenous antioxidant fluctuations. Compliance with these dietary instructions was verbally confirmed before each testing session and recorded in standardized pre-assessment checklists. Baseline dietary patterns (energy and macronutrient distribution) were qualitatively similar across groups according to the initial screening questionnaire, minimizing potential bias from habitual diet differences.

Statistical analyses were performed using SPSS software version 24.0 (SPSS, Inc., Chicago, IL, USA). Data are presented as means ± standard deviations. A P-value < 0.05 was considered statistically significant. Kolmogorov-Smirnov and Levene’s tests were used to evaluate the normal distribution of the data and homogeneity of variances, respectively. Analyses were conducted following a per-protocol approach. Participants who missed 3 consecutive training sessions, according to the predefined withdrawal criteria, were excluded from the final analysis. An intention-to-treat analysis was not performed. Additionally, sample size was estimated a priori, calculated using G*Power software for repeated measures analysis of variance (ANOVA) (within-between interaction), based on an expected medium effect size, α = 0.05, and power (1 – β) = 0.80.

At baseline, group characteristics were compared using one-way ANOVA to ensure that initial outcomes did not differ significantly across groups. A two-way repeated measures ANOVA was performed to examine the main effects of group and time, as well as their interaction. When significant effects were observed, pairwise comparisons were conducted using the Bonferroni post hoc test. Data are presented as mean (SD). Effect sizes were calculated from the ANOVA output by converting partial eta-squared (η²p) to Cohen’s d. Cohen’s d effect size was calculated using the formula: d = (mean difference between pre- and post-intervention)/(pooled standard deviation). According to Cohen’s conventions, effect sizes were interpreted as small (Cohen’s d = 0.2), medium (d = 0.5), and large (d = 0.8) [37]. For η²p, effect sizes were interpreted as small (η²p = 0.01), medium (η²p = 0.06), and large (η²p = 0.14) [38]. In addition to effect size calculations, the percentage of changes in each outcome was calculated and reported. Relationships between changes in study outcomes (pre-post differences) were examined using Pearson’s correlation coefficient.

Baseline characteristics of participants across groups are presented in Table 1. No significant differences were observed among the three groups in baseline values and characteristics, except for BMI. BMI was significantly higher in the CG compared to the HIIT group (P = 0.011).

ANOVA showed significant time effects for all anthropometric parameters (P < 0.001 for all, d = 1.06 to 1.28). There were significant time-group effects for body weight (F2, 37 = 24.18, P = 0.001, η²p = 0.567, d = 1.12), BMI (F2, 37 = 24.54, P = 0.001, η²p = 0.570, d = 1.13), %BF (F2, 37 = 11.87, P = 0.001, η²p = 0.391, d = 0.79), WC (F2, 37 = 14.30, P = 0.001, η²p = 0.436, d = 0.86) and WHR (F2, 37 = 11.92, P = 0.001, η²p = 0.392, d = 0.79). Neither the time effect (P = 0.085) nor the time-group effect (P = 0.458) was significant for HC (data not shown in the table).

Body weight values in the HIIT group (–2.97%, P = 0.001, d = 0.40) and CONC group (–4.74%, P = 0.001, d = 0.65) were reduced significantly from baseline at the end of intervention, while no significant change was observed in the CG (+0.37%, P = 0.160, d = 0.04). The weight loss rate in the CONC group was significantly higher than in the HIIT (P = 0.014) and CG (P = 0.001) groups. Weight loss in the HIIT group was also significantly higher than in the CG (P = 0.004).

BMI also decreased significantly in the HIIT group (–2.99%, P = 0.001, d = 1.32) and CONC group (–4.71%, P = 0.001, d = 1.45) compared to baseline values. The change of BMI in CG did not show a significant difference (+0.38%, P = 0.132, d = 0.10). In the CONC group, BMI loss was significantly greater than in the HIIT group and CG (P = 0.02). HIIT also lost more BMI than CG (P = 0.004). %BF in the HIIT group (–6.37%, P = 0.001, d = 0.45) and CONC group (–9.81%, P = 0.001, d = 1.21) was reduced significantly in comparison to baseline values, while no significant change was observed in CG (–0.16%, P = 0.565, d = 0.02).

The %BF loss rate in the CONC group was significantly higher than in HIIT (P = 0.048) and CG (P = 0.001). There was also a significant reduction in the %BF in the HIIT group compared to the CG (P = 0.004).

WC decreased significantly in the HIIT group (–2.88%, P = 0.001, d = 0.63) and CONC group (–2.60%, P = 0.001, d = 0.75) compared to baseline values. No significant difference was observed in CG (–0.09%, P = 0.278, d = 0.002). There was a higher reduction in WC in the CONC (P = 0.001) and HIIT (P = 0.001) groups than in the CG. However, no significant difference was observed between the CONC and HIIT groups (P = 0.990).

No significant differences were observed for HC in the three groups: HIIT group (–0.13%, P = 0.147, d = 0.03), CONC group (–0.22%, P = 0.231, d = 0.02), and CG (–0.01%, P = 0.278, d = 0.002) compared to baseline.

WHR decreased significantly in the HIIT group (–3.12%, P = 0.001, d = 1.17) and CONC group (–2.10%, P = 0.001, d = 0.66) compared to baseline values. At the same time, no significant difference was observed in CG (–0.10%, P = 0.104, d = 0.003). In the CONC (P = 0.001) and HIIT (P = 0.001) groups, WHR was reduced compared to the CG. In contrast, there was no significant difference between the CONC and HIIT groups (P = 0.997) (see Table 2 and Figure 2).

The percentage changes in body composition variables are illustrated in Figure 2.

ANOVA showed significant time effects for all lipid profile parameters (P < 0.001 for all, d = 0.65 to 1.03) and time-group effects for TG (F2, 37 = 3.32, P = 0.047, η²p = 0.153, d = 0.41), TC (F2, 37 = 13.31, P = 0.001, η²p = 0.418, d = 0.83), HDL-C (F2, 37 = 8.42, P = 0.001, η²p = 0.313, d = 0.66), LDL-C (F2, 37 = 12.48, P = 0.001, η²p = 0.403, d = 0.81) and AIP (F2, 37 = 6.25, P = 0.005, η²p = 0.253, d = 0.57).

Serum TG concentrations were significantly reduced in the HIIT group (–5.76%, P = 0.011, d = 0.33) and CONC group (–6.47%, P = 0.001, d = 0.44) compared with baseline. No significant differences were observed in CG (–0.93%, P = 0.516, d = 0.05). Despite a significant decrease in serum TG concentration in the CONC and HIIT groups, no significant difference was observed between pre-post-test differences in serum TG concentration before and after the test comparing the three groups (P = 0.250).

Serum TC concentrations were significantly reduced in the HIIT group (–7.53%, P = 0.001, d = 0.53) and CONC group (–7.42%, P = 0.001, d = 0.60) compared with baseline values. No significant differences were observed in CG (+0.59%, P = 0.407, d = 0.03). The serum TC reduction in the CONC (P = 0.001) and HIIT (P = 0.005) groups is significant compared with the CG. However, no significant difference was observed between the CONC and HIIT groups (P = 0.891).

Serum LDL-C concentrations were significantly reduced in the HIIT group (–14.02%, P = 0.001, d = 0.52) and CONC group (–8.31%, P = 0.001, d = 0.63) compared with baseline values. No significant differences were observed in CG (+1.48%, P = 0.252, d = 0.05). Compared to CG, CONC (P = 0.004) and HIIT (P = 0.005) showed reduced serum LDL-C. However, no significant differences were observed between the CONC and HIIT groups (P = 0.998).

Serum HDL-C concentration increased significantly in the HIIT group (+7.74%, P = 0.001, d = 0.45) and CONC group (+6.21%, P = 0.001, d = 0.43) at the end of intervention. No significant differences were observed in the CG (–1.65%, P = 0.110, d = 0.11). HDL-C presented higher values in the CONC (P = 0.029) and HIIT (P = 0.002) groups than in the CG. However, no significant difference was observed between the CONC and HIIT groups (P = 0.589).

The AIP index values in the HIIT group (–11.88%, P = 0.001, d = 0.40) and CONC group (–12.18%, P = 0.001, d = 0.48) decreased significantly compared to baseline values. In contrast, it increased significantly in the CG (+1.00%, P = 0.001, d = 0.02). AIP index values were significantly reduced in CONC (P = 0.040) and HIIT (P = 0.045) groups compared to the CG. However, no significant difference was observed between the CONC and HIIT groups (P = 0.655) (see Table 3).

ANOVA showed significant time effects for all glycemic biomarker parameters (P < 0.001 for all, d = 0.935–1.17). Significant time-group effects were observed for insulin (F2, 37 = 9.70, P = 0.001, η²p = 0.344, d = 0.69), glucose (F2, 37 = 5.36, P = 0.009, η²p = 0.225, d = 0.53) and HOMA-IR (F2, 37 = 11.68, P = 0.001, η²p = 0.387, d = 0.78).

Serum insulin concentrations were significantly reduced in the HIIT group (–18.25%, P = 0.001, d = 0.85) and CONC group (–19.27%, P = 0.002, d = 1.18) compared to baseline values. No significant differences were observed in CG (–0.14%, P = 0.840, d = 0.007). The insulin concentration reduction in the CONC (P = 0.001) and HIIT (P = 0.003) groups was significantly lower than in the CG. However, no significant difference was observed between the CONC and HIIT groups (P = 0.778).

Serum glucose concentrations were significantly reduced in the HIIT group (–11.96%, P = 0.001, d = 0.66) and CONC group (–4.16%, P = 0.002, d = 0.58) compared to baseline values. No significant difference was observed in CG (–0.89%, P = 0.840, d = 0.11). No significant difference was found in glucose changes comparing all study groups. HOMA-IR index concentrations were significantly reduced in the HIIT group (–23.14%, P = 0.001, d = 1.34) and CONC group (–23.88%, P = 0.001, d = 1.24) compared to baseline values. No significant difference was observed in CG (+0.61%, P = 0.366, d = 0.05). The HOMA-IR index reduction in the CONC (P = 0.001) and HIIT (P = 0.001) groups is significant compared to the CG. However, no significant difference was observed between the CONC and HIIT groups (P = 0.970) (see Table 4).

The percentage changes in metabolic biomarkers are presented in Figure 3.

The results of the Pearson correlation analysis showed that weight loss was significantly associated with reductions in BMI (r = 0.992, P = 0.001), %BF (r = 0.695, P = 0.001), WC (r = 0.652, P = 0.001), WHR (r = 0.651, P = 0.001), LDL-C (r = 0.548, P = 0.001), insulin (r = 0.468, P = 0.001), and HOMA-IR (r = 0.420, P = 0.002), and with an increase in HDL-C (r = 0.447, P = 0.001).

BMI reduction had a significant positive relationship with a decrease in %BF (r = 0.703, P = 0.001), WC (r = 0.651, P = 0.001), WHR (r = 0.644, P = 0.001), LDL-C (r = 0.561, P = 0.001), insulin (r = 0.480, P = 0.002), HOMA-IR (r = 0.436, P = 0.005) and with an increase in HDL-C (r = 0.475, P = 0.002).

Decreased %BF also presented a significant positive relationship with decreased WC (r = 0.805, P = 0.001), WHR (r = 0.746, P = 0.001), TC (r = 0.678, P = 0.001), LDL-C (r = 0.667, P = 0.001), insulin (r = 0.462, P = 0.003), HOMA-IR index (r = 0.491, P = 0.001) and with increased HDL-C (r = 0.611, P = 0.001) (see Table 5).

No adverse events, musculoskeletal injuries, or exercise-related complications were reported throughout the 12-week intervention, confirming the safety and feasibility of both exercise modalities in sedentary postmenopausal women with overweight or obesity.

This study presents several limitations. First, the sample size was small and recruited from a single center. Future research involving larger multicentric cohort samples would provide more robust and generalizable results. Second, it is also important to consider that postmenopausal women with overweight or obesity, who voluntarily enroll in CONC or HIIT exercise programs, may be more motivated and physically capable than the broader population of postmenopausal women with overweight or obesity, potentially contributing to higher adherence rates. Consequently, our study’s findings may not be generalizable to the general population, as the study sample represents a specific and self-selected subgroup. Third, the absence of dietary monitoring precludes accurate assessment of participants’ nutritional intake, limiting the ability to control for dietary influences on metabolic outcomes. Fourth, the duration of the study, because the lack of a follow-up period further restricts the evaluation of long-term effects. Five, the study did not incorporate complementary biochemical analyses, particularly those assessing inflammatory biomarkers and antioxidant status. These unmeasured parameters could act as confounding variables. Including such parameters in future studies will be essential to provide a more comprehensive understanding of the physiological effects of exercise interventions. Finally, the idea for the intervention study arose from a clinical observation. This led to a lack of coordination and a delay in registering the study, which was ultimately completed at the time of manuscript submission.

The findings of the present study support the implementation of HIIT and CONC programs as effective tools to improve body composition, circulating lipid profile, insulin sensitivity, and carbohydrate metabolism in postmenopausal women with overweight or obesity. These modalities of exercise can be integrated into clinical and community-based health strategies to reduce cardiometabolic risk, enhance metabolic adaptation, and promote long-term adherence to physical activity. Tailoring exercise prescriptions to individual capabilities and monitoring metabolic responses may further optimize outcomes in this vulnerable population.

Our findings suggest that 12 weeks of CONC and HIIT exercises may serve as effective interventions for obesity management, particularly in sedentary postmenopausal women. Both exercise modalities were associated with improvements in body composition, reductions in BF, and favorable changes in lipid and glycemic biomarker profiles. Based on these outcomes, CONC and HIIT exercises are recommended as viable strategies to enhance metabolic health and body composition in individuals with obesity, particularly by improving blood lipid and glucose regulation.