The study used a longitudinal design to assess the impact of a maximum fat oxidation training program on the psychological well-being, mental health, central fat, and cardiopulmonary biomarkers of school-aged children with obesity compared to an age- and body composition-matched control group without intervention.

The outcome measures included anthropometrics [age, height, body mass (BM), body mass index (BMI), waist and hip circumferences (WC and HC), and waist-to-hip ratio (WHR)], cardiopulmonary fitness [relative peak maximal consumption (VO₂ peak) and relative maximal fat oxidation (FATmax) rate], and psychological questionnaires [anxiety, depression, and self-esteem]. These assessments were conducted in the school setting before and after the two-month training program (Figure 1).

Thirty-six school-aged children with obesity voluntarily participated in this study (age: 13.1 ± 0.9 years; body mass: 90.6 ± 11.3 kg; height: 1.64 ± 0.07 m; BMI: 33.4 ± 2.3 kg.m⁻²; Pubertal stage: 3.4 ± 1.0 PS). The sample size was calculated according to Faul et al. (19) procedures’ using the G*Power software (version 3.1.9.4; Kiel University, Kiel, Germany). The α value was set at 0.05 and the power (1−β) at 0.85. Based on the results of Ben Ounis et al. (11), the effect size estimated was >0.8 (Large effect). Fifteen participants per group would provide sufficient power to minimize the risk of committing a Type II statistical error. The participants were selected based on the International Obesity Task Force's inclusion criteria, which required a BMI above the 97th percentile (20). To ensure a diverse and representative sample, participants were recruited from both rural and urban settings, with clinical data collected via a standardized questionnaire (21). The study involved 36 children out of 60 who were eligible for the study intervention. Twenty participants were excluded due to respiratory symptoms (n = 7), cardiac disease or acute chronic respiratory diseases (n = 3), chest, abdominal, or nasal surgery (n = 5), and smoking history (n = 5). Participants were randomly assigned to a control group (CONTG, n = 16) and an experimental group (EXPG, n = 20). Randomization was performed with the ‘‘RAND’' formula in Microsoft Excel, and allocation concealment was maintained until baseline measurements were taken to prevent selection bias. Stratification variables such as age, pubertal stage, and geographical location were not utilized due to a small sample size, despite being documented in baseline measures. Only EXPG participated in the training program. Four children from the CONTG missed the second assessment and were excluded from the analysis (Figure 1). The study involved children and their parents being informed about the risks and benefits of participating in experimental procedures before giving their written informed consent. The study protocol was approved by the University's Institutional Ethical Committee and the Ministry of Education and Higher Education and Scientific Research (MEHESR-2005/014; January 5, 2025), following the Declaration of Helsinki (2013).

A medical checkup was conducted, and anthropometric measurements were taken from each participant. Body weight was recorded to the nearest 0.1 kg using a digital scale (OHAUS, Florhman Park, NJ), and height was measured to the nearest 0.1 cm using a standing stadiometer. WC and HC were measured using a ribbon meter with a precision of 0.1 cm. The BMI was determined by dividing body mass in kilograms by the square of height in meters, and the WHR was calculated. All measurements were carried out twice by a single competent technician (ISAK Level 1) with an intra-observer technical error of measurement below 1% in all the variables, according to standard protocols. Pubertal status was confirmed for all participants by the school pediatrician at baseline and the two-month follow-up, based on Tanner classification (22) of pubic hair development.

Two progressive cycle ergometer tests were conducted before and after the 2-month intervention. The children had a familiarization session with the cycle ergometer 48 h before the exercise testing. The tests were conducted in a laboratory setting after approximately 12 h of sleep, with a temperature range of 22–24 °C and a relative humidity of 76%. Using an electromagnetically braked cycle ergometer (Ergoline, Bitz, Germany), participants performed the previously reported exercise methodology (23). The test protocol consists of a 5-stage progressive submaximal exercise corresponding to 20, 30, 40, 50, and 60% of Wmaxth, with a 6-minute exercise at each work rate and a 4-minute active recovery period at 20% of Wmaxth. Peak oxygen uptake (VO₂ peak) and gas exchange (VO₂ and VCO₂) were measured via breath-by-breath sampling of expired gases during testing using a metabolic gas analyzer (ZAN 600; ZAN Meβgeräte, Oberthulba, Germany). The ZAN 600 gas analyzer was calibrated before each test using certified gases (16% O₂, 4% CO₂). Day-to-day coefficient of variation for VO₂ and VCO₂ was <2%, consistent with published validation data.¹ VO₂ peak was defined as the highest 30-s average VO₂, with maximal effort confirmed by HR ≥95% of predicted maximum and RER ≥1.00 (24).

Heart rate was monitored continuously throughout the test using ECG (ZAN ECG 800; ZAN Meβgeräte). Participants’ theoretical maximal oxygen intake (VO₂maxth) and aerobic workload (Wmaxth) were determined using prediction equations considering sex and anthropometric characteristics (25): VO₂maxth (mL/min) = [28.5 × body mass (kg)] + 288.1 and the following equation was used to calculate Wmaxth:={VO₂maxth−10 × [body mass (kg)]}/10.3.

After two months, the children performed the same exercise test protocol at the same incremental workload and were compared at the same percentage of their Wmaxth. Ventilatory parameters were recorded during the last 3 min of each workload to calculate substrate oxidation flow rates (26). Whole-body substrate oxidation was calculated from the respiratory exchange ratio, while fat and carbohydrate oxidation rates were calculated using the non-protein respiratory quotient technique (27), as protein breakdown contributes little to energy metabolism during exercise (28). In this study, we calculated the maximal fat oxidation (FATmax) point to represent the balance between fat and carbohydrate utilization induced by increasing exercise intensity (23). This point reaches maximum fat oxidation induced by workload, followed by a decrease as carbohydrate becomes the predominant fuel, allowing for individualized training.

The study used the French version of Spielberger's State-Trait Anxiety Inventory (STAI-Y), a widely used tool for assessing trait and state anxiety, to diagnose and differentiate anxiety from depressive disorders, with 20 questions for the trait anxiety section (29). State anxiety items include: “I am tense; I am worried” and “I feel calm; I feel secure.” Trait anxiety items include: “I worry too much over something that doesn’t matter” and “I am content; I am a steady person.” All items are rated on a 4-point scale (e.g., from “Almost Never” to “Almost Always”). Higher scores indicate greater anxiety. The STAI is suitable for individuals with at least a sixth-grade reading level.

The study used the French version of the Beck Depression Inventory (BDI), a 21-item self-reported rating tool used to assess the characteristics and symptoms of depression, with each item scored from 0 to 3. The total score, ranging from 0 to 63, provides a quantitative evaluation of the depressive state. The abbreviated formula categorizes depression scores into no depression (0–4), slightly depressed (5–7), moderately depressed (8–15), and severely depressed (16 and over) (30). The BDI is also suitable for individuals with at least a sixth-grade reading level.

The study used the French version of Rosenberg's self-esteem questionnaire (31), a widely used self-assessed measure in social science research, with a score below 15 indicating low self-esteem, based on a scale from 0 to 30. The questions consist of 10 Likert-type scale items ranging from 1 (strongly disagree) to 4 (strongly agree), with five items scored positively and five negatively.

Participants completed a physical activity readiness questionnaire, indicating that they only participated in 1 h of school physical education class per week. The CONTG maintained its usual level of physical activity, while the EXPG participated in a supervised exercise training program for two months consisting of 4 weekly sessions of 90 min each, as described by Ben Ounis et al. (11). Each supervised 90-min session consisted of: 10-min warm-up, 60–65 min of moderate aerobic exercise at FATmax heart rate, and 15–20 min of cool-down. Heart rate was continuously monitored to ensure safety for children with obesity (11, 16). They received a two-page summary from an exercise physiologist detailing the health benefits of regular exercise and recommendations and precautions for exercise. The exercise routine included warming up, jogging, jumping, and ball games. The intensity was adjusted based on the participant's heart rate, which was monitored using a heart rate monitor (Vantage NV; Polar Electro, Kempele, Finland). Briefly, the training session schedule included a 10-min warm-up, two 30–35-min blocks of ludic, group-exercise (e.g., relay team races, obstacle courses, modified sport) performed at the correspondent FATmax's heart rate (±10 bpm), 5 min of active recovery, 30-min training at the FATmax's heart rate, and 10 min of cool down. Although FATmax was assessed via cycle ergometry, dynamic field activities (e.g., relay team competition, obstacle course, modified sport) were chosen for the training intervention. Heart rate was maintained within the FATmax zone (±10 bpm) to ensure metabolic consistency. This approach improved compliance and ecological validity in a pediatric population while ensuring that participants maintained the target heart rate associated with FATmax intensity. A detailed session-by-session protocol is provided is the Supplementary Material.

Data were presented as mean ± SD. Statistical analyses were performed using the SPSS package (SPSS Inc. Chicago. IL. version 26.0). After conducting the Shapiro–Wilk test for normality, Levene's test for homogeneity of variances, and Mauchly's test for sphericity, an independent t-test was performed to assess significant differences in baseline values between groups. Separate mixed-factors ANOVA (time  ×  group) for repeated measurements was used to analyze changes in dependent parameters. The main effect for the group (two levels: EXPG and CONTG) and time (two levels: pre- and post-training) was determined through post-hoc analysis with Bonferroni's correction. post-hoc analysis using Bonferroni's correction was then performed to calculate the main effect for the group (two levels: EXPG and CONTG) and time (two levels: pre- and post-training). Partial eta-squared (η_p²) was quantified for analysis of variance of repeated measures (32). Effect sizes were classified as small (η_p² up to 0.059), medium (between 0.059 and 0.138), and large (greater than 0.138). Between-groups standardized mean differences or effect sizes (ES) were calculated using Cohen's d and corrected by Hedge's g to avoid biased estimation of the population effect size. ES can be classified as small (0 ≤ d ≤ 0.49), medium (0.50 ≤ d ≤ 0.79), and large (d ≥ 0.80) (33). The level of significance was set at p < 0.05.

The mean values of anthropometric characteristics, self-esteem, anxiety, depression scores, and cardiopulmonary fitness in pre-and post-FATmax training are illustrated in Table 1.

No significant differences in age, pubertal stage, or geographical location were observed between the EXPG and CONTG at baseline (p values ranging from 0.75 to 0.99). EXPG showed significant pre-to-post reductions in all anthropometric indices (all p < 0.001), whereas CONTG exhibited small but significant increases in BMI and WHR (p < 0.05).

Significant interactions were found (training  ×  group) for BM, BMI, WC, HC, and WHR (F_(1.34) = 129.4, ηp² = 0.79; F_(1.34) = 134.9, ηp² = 0.80; F_(1.34) = 372.5, ηp² = 0.92; F_(1.34) = 296.7, ηp² = 0.90; F_(1.34) = 7.1, ηp² = 0.17, respectively). post-hoc analysis revealed that body composition indices (BM, BMI, WC, HC, and WHR) improved only in EXPG (p < 005 to p < 0.001).

EXPG's baseline FATmax rate and VO₂ peak were 1.4 mg/kg/min and 21.9 mL/kg/min, respectively. After the 2-month FATmax training, these increased to 2.3 mg/kg/min and 28.4 mL/kg/min, ranging from 23% for VO₂ peak to 66% for FATmax rate. These are significantly higher than CONTG, which showed a 0.2% increase in VO₂ peak and −0.4% decrease in FATmax rate, respectively.

Significant interactions were also found for FATmax rate and VO₂ peak (F_(1.34) = 470.0, η_p² = 0.93; F_(1.34) = 469.3, η_p² = 0.93, respectively). post-hoc analysis revealed that the values of the FATmax rate and VO₂ peak increased in EXPG (p < 0.001).

EXPG's baseline self-esteem, anxiety, and depression scores were 23.7, 31.3, and 13.4, respectively. After the 2-month FATmax training, these increased to 37.5 for self-esteem and dropped to 27.5 and 9.6 for anxiety and depression, respectively, ranging from 12% for anxiety to 59% for self-esteem. These are significantly higher than CONTG, which showed a 2% decrease in self-esteem scores and a 3% and 5% increase in anxiety and depression scores, respectively.

Significant interactions were also found for self-esteem, anxiety, and depression scores (F_(1.34) = 554.4, η_p² = 0.94; F_(1.34) = 120.8, η_p² = 0.78; F_(1.34) = 149.8, η_p² = 0.82, respectively). post-hoc analysis revealed that the values of the self-esteem score increased, and anxiety and depression scores decreased in EXPG (p < 0.01 to p < 0.001). Moreover, post-hoc analysis revealed that the values of anxiety and depression scores increased in CONTG (p < 0.05).

Figure 2 displayed the between-group comparison of body composition data, VO₂ peak, and FATmax rate values as well as anxiety, depression, and self-esteem scores. The improvement in BM [ES: 1.88 (95% CI: 1.34–2.45)], BMI [ES: 1.93 (95% CI: 1.37–2.49)], WC [ES: 6.29 (95% CI: 4.76–7.81)], HC [ES: 2.35 (95% CI: 1.85–3.21)] and WHR [ES: 0.41 (95% CI: 0.21–0.78)] were significantly greater in EXPG compared to CONTG (small-to-large). In addition, the improvements in VO₂ peak [ES: 3.61 (95% CI: 2.69–4.52)] and FATmax rate [ES: 3.64 (95% CI: 2.70–4.53)] values as well as anxiety [ES: 1.62 (95% CI: 1.12–2.12)], depression [ES: 1.82 (95% CI: 1.28–2.36)], and self-esteem scores (ES = 3.92 [95% CI: 2.94–4.90] were significantly greater in EXPG compared to CONTG (moderate-to-large).

The relationship between changes in psychological scores and changes in anthropometric characteristics and physical fitness is summarized in Table 2. When both groups' data were pooled, there were positive and negative relationships between the percentage of changes in anxiety, depression, and self-esteem scores with the percentage of changes in BM, BMI, WC, HC, and WHR (r = −0.16 to r = 0.84; p < 0.05 to p < 0.001). In addition, the percentage of changes in anxiety, depression, and self-esteem scores also showed positive and negative relationships with the percentage of changes in VO₂ peak and FATmax rate values (r = −0.72 to r = 0.91; p < 0.01 to p < 0.001). The percentage of change in VO2 peak showed a positive relationship with the percentage of changes in Fatmax rate (r = 0.94; p < 0.001).

The study has identified some limitations that need to be addressed. First, participants self-reported their mental health status using validated French versions of STAI-Y and BDI, potentially causing memory and response bias. Future research should consider incorporating age-specific or observer-rated psychological instruments. Despite this, standardized questionnaires for measuring children's mental health have achieved satisfactory results in previous epidemiological studies. Second, the lack of dietary control among participants may confound results since they maintained habitual diets, but unmeasured nutritional changes could still occur. Third, this study did not assess physiological and biological mediators like hormonal and inflammatory markers, which could explain psychological adaptation processes from FATmax exercise training. In addition, the randomization procedure did not include stratification by relevant baseline variables such as age, pubertal stage, or geographical location (rural/urban), which could affect group comparability. Finally, the transfer of FATmax from cycling to games may reduce exercise specificity.