Introduction

The main objectives of a weight-management programs are to change food and behavioral habits, to reduce fat mass (FM), to maintain fat-free mass (FFM) and to enhance physical capabilities in order to improve long-term weight regulation [1]. Physical activity promotes negative energy balance and also negative fat balance [2] and appears to be one of the major factors determining success in long-term weight maintenance [3]. Therefore, recommendations for the duration, type and intensity of the physical activity sessions are needed to optimize energy expenditure, fat oxidation and improve physical capacities during the weight-management programs.

Moderate-intensity activity for approximately 45–60 min per day seems adequate to prevent the development of overweight or obesity and 90 min per day is reported to be necessary to reduce body weight in overweight subjects [4]. Walking appears to be a more convenient physical activity for obese adolescents than cycling, because in obese adolescents, the target energy expenditure was attained at lower heart rate (HR), with lower blood lactate concentration and greater fat oxidation [5]. Low-intensity (LI) exercise seems to increase fat oxidation during the exercise (but not at rest) and results in a greater total fat oxidation than moderate- or high-intensity (HI) exercise with similar energy expenditure [6–8]. In addition, high-intensity interval training (HIIT) is reported to be a feasible and efficacious strategy for increasing health-related fitness in overweight and obese youth [9]. HIIT generally consists of short, yet intense bouts of exercise interspersed with rest periods. The main appeal of HIIT is that this type of training can be completed in a short period of time (compared to traditional aerobic training), and it requires no or minimal equipment and is perceived as a more enjoyable exercise compared to other exercise modalities [10]. HIIT produced equal or better cardiometabolic gains in a shorter time period in comparison with traditional endurance training [11, 12] in both healthy [13] and obese people [14], while no effects were described on body composition and on muscular fitness [15]. Similarly, there are no consistent findings supporting the view that moderate-intensity training can increase fat oxidation to a greater extent than HI or HIIT exercise in obese subjects [16]. Discrepancies among different studies might be attributed to the fact that prescription of HIIT consists of manipulating different variables (e.g., work interval intensity and duration, relief interval intensity and duration, exercise modality, number of repetitions, number of series, between-series recovery duration and intensity) [17]. For this reason, the HIIT interventions for overweight and obese children are heterogeneous and further studies are still requested to define the most suitable combination for maximizing health improvements in a better way.

Thus, the aim of the present study was to determine the effects of a 3-week multidisciplinary body weight-management program entailing moderate energy restriction, nutritional education, psychological counseling (common to all subjects) and three different exercise training (subgroup a: low intensity, LI: 40 % V′O2max; subgroup b: high intensity, HI: 70 % V′O2max; subgroup c: high-intensity interval training, HIIT), on body composition, energy expenditure and fat oxidation rate in severely obese adolescents.

Study design and methods

Subjects

Thirty severely obese boys (BMI SDS > 2, [18]) aged 15 to 17 years, with a pubertal stage >3, according to the Tanner pubertal staging [19], were admitted to the study. The subjects were recruited as inpatients from the Division of Auxology, Italian Institute for Auxology, IRCCS, Piancavallo (VB), Italy. All subjects had a full medical history and physical examination, with the routine hematology and biochemistry screens and urine analysis. Body mass (BM) was stable during the previous 2 months (changes less than ± 1 kg). None of the subjects had evidence of significant disease, notably non-insulin-dependent diabetes mellitus or other endocrine disease, and none were taking medications regularly or using any medication known to influence energy metabolism.

All the participants were admitted to a 3-week multidisciplinary body weight reduction program entailing moderate energy restriction, nutritional education and psychological counseling. As far as physical activity is concerned, participants were randomly assigned to a low-intensity subgroup (LI, n. 11, exercising at HR corresponding to 40 % of V′O2max), to a high-intensity subgroup (HI, n. 9, exercising at HR corresponding to 70 % of V′O2max) and to a high-intensity interval training (HIIT, n. 10, consisting of repeated 40-s efforts of high-intensity walking at HR corresponding to 100 % of V′O2max, which were intermixed with 5 min of walking at low intensity corresponding to 40 % of V′O2max).

Study protocol

The study was approved by the Ethics Committee of the Italian Institute for Auxology (Milan) and has been performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. The purpose and objective of the study were explained to each subject and his or her parents, and written informed consent was obtained before beginning the study. The adolescents were hospitalized for a multidisciplinary body weight reduction program. During the first days, subjects had physical examination, routine hematology and biochemistry screens and urine analysis. Thereafter, they followed a 3-week personalized weight-management program entailing moderate energy restriction, nutritional education, psychological counseling (common to all subjects) and three different exercise training (described in detail in the following section). Full testing sessions were conducted just before the beginning (Week 0, W0) and at completion of the 3-week weight-management program (Week 3, W3). The testing session included assessment of anthropometric characteristics, body composition, basal metabolic rate (BMR), energy expenditure (EE) and substrate oxidation rate during submaximal exercise and post-exercise recovery.

Diet and nutritional education

During the 3-week weight-management program, personalized diets were offered on the basis of the initial BMR test and physical activity level for each adolescent. Energy supply was adjusted to be close to 1.2 times the initial BMR, which is about 15–20 % less than the estimated daily EE. The composition of diet was formulated according to the Italian recommended daily allowances [20], and the subjects were instructed to consume all the meals. During the 3-week weight-management program, the consumption of meals always occurred under the supervision of a dietician.

Physical activity

During the 3-week weight-management program, the adolescents participated in a personalized exercise training program, from Monday to Friday, which included two endurance training sessions per day (followed by 5- to 7-min stretching) under heart rate (HR) monitoring and medical supervision. All subjects completed 28 ± 2 sessions of physical training.

Participants were randomly assigned to a low-intensity subgroup (LI, n. 11, exercising at HR corresponding to 40 % of V′O2max), or to a high-intensity subgroup (HI, n. 9, exercising at HR corresponding to 70 % of V′O2max), or to a high-intensity interval training subgroup (HIIT, n. 10, consisting of 6 repeated 40-s efforts of high-intensity walking at HR corresponding to 100 % of V′O2max, which were intermixed with 5 min of walking at low intensity corresponding to 40 % of V′O2max).

The amounts of energy expended during a training session were similar for both groups (20 kJ per kg of fat-free mass, FFM, about 1.4 MJ per session). Each training session lasted 45 ± 6 min for LI program, 31 ± 4 min for the HI training program and 37 ± 3 min for the HIIT training program. In addition, subjects had one hour·day−1 of aerobic leisure activities at the institution on Saturday and Sunday. The research assistant and the physical trainers verified that each subject participated in each training session, performed exercises correctly and completed at least 95 % of the exercise session and program. Additionally, all the subjects considered in the present paper completed more than 97 % of the exercise session and program.

Measurements

Physical characteristics and body composition

The medical history and a physical examination of subjects were taken at the time of admission to hospital. Body mass (BM) was measured to the nearest 0.1 kg with an electronic scale (Selus, Italy) with the subject dressed only in light underwear. Stature was measured to the nearest 0.5 cm on a standardized Harpenden stadiometer (Holtain Ltd, UK). The body mass index (BMI) was calculated as weight (kg) divided by stature (m) squared. Body composition was measured by using a multifrequency tetrapolar impedancemeter (BIA, Human-IM Scan, DS-Medigroup, Milan, Italy) with a delivered current of 800 µA at a frequency of 50 kHz. In order to reduce errors of measurement, attention was paid to the standardization of the variables that affect measurement validity, reproducibility and precision. Measurements were performed according to the method of Lukaski (1987) [21] after 20-min resting in a supine position with arms and legs relaxed and not in contact with other body parts. FFM was calculated using the prediction equation developed by Lazzer (2008) [22], and fat mass (FM) was derived as the difference between BM and FFM.

Basal metabolic rate

BMR was determined in the morning (measurements starting between 0800 and 1000 a.m.), after an overnight fast by means of open-circuit, indirect computerized calorimetry (Vmax 29, Sensor Medics, Yorba Linda, Ca, USA) with a rigid, transparent and ventilated canopy. Before each test, the gas analyzers were calibrated using a reference gas mixture (95.00 % O2 and 5.00 % CO2). BMR of subjects was measured for 45 min. Oxygen consumption (V′O2) and carbon dioxide production (V’CO2), standardized for temperature, barometric pressure and humidity, were recorded at 1-min intervals. Results from the first 5–10 min, which corresponded to adaptation to the procedural environment, were excluded from the analysis. Energy expenditure was derived from the measured oxygen uptake and carbon dioxide output according to the formula of Weir (1949) [23] and averaged over the whole measurement period.

Physical capacities and maximal fat oxidation rate

Peak oxygen uptake and maximal fat oxidation rate were determined by using a graded exercise test on a motorized treadmill (TechnoGym, Gambettola, Italy), under medical supervision. Subjects were asked to avoid strenuous exercise the day before the test and came to the laboratory after a 12-h overnight fast. Before the beginning of the study, subjects were familiarized with the equipment and the procedures. Each test was undertaken in the morning (exercise starting between 0800 and 1000 a.m.) and comprised a 10-min rest period followed by walking in stages of 5-min duration. The rates in m s−1 and incline in  % followed a sequence : 0.6 (0 %), 1.0 (0 %), 1.0 (3 %), 1.3 (3 %), 1.4 (6 %), 1.4 (9 %) and 1.4 (12 %). The workload was progressively increased until a HR of approximately 180 beats min−1 was reached, at which point exercise was concluded in order to avoid any cardiovascular complications associated with maximal effort which would be particularly risky in this kind of population. During the experiment, ventilatory and gas exchange responses were measured continuously by indirect calorimetry (CPX Express, Medical Graphics Corp, MN, USA). The flowmeter and gas analyzers of the system were calibrated using, respectively, a 3-L calibration syringe and calibration gas (16.00 % O2; 4.00 % CO2). During the exercise test, an electrocardiogram was recorded continuously and displayed online for visual monitoring, and heart rate (HR) was measured with a dedicated monitor device (Polar, Finland). The peak oxygen uptake (V′O2peak) was estimated for each subject considering the last 20 s of the graded exercise test.

The substrate oxidation rate was calculated from V′O2 and V’CO2 (CPX Express, Medical Graphics Corp, MN, USA) during the last minute of each workload level, according to the protocol of Achten (2002) [24] and using the following equations [25]:

$$ {\text{Fat}}\,{\text{oxidation}}\,{\text{rate}}\,\left( {{{\text{g}} \mathord{\left/ {\vphantom {{\text{g}} { \hbox{min} }}} \right. \kern-0pt} { \hbox{min} }}} \right) = 1.67\, \times\, {\text{V}}^{\prime}{\text{O}}_{2} \left( {{{\text{l}} \mathord{\left/ {\vphantom {{\text{l}} { \hbox{min} }}} \right. \kern-0pt} { \hbox{min} }}} \right) - 1.67\, \times\, {\text{V}}^{\prime}{\text{CO}}_{2} \left( {{{\text{l}} \mathord{\left/ {\vphantom {{\text{l}} { \hbox{min} }}} \right. \kern-0pt} { \hbox{min} }}} \right) - 0.307 \times {\text{Pox}} $$
$$ {\text{Carbohydrate}}\,{\text{oxidation}}\,{\text{rate }}\left( {{{\text{g}} \mathord{\left/ {\vphantom {{\text{g}} { \hbox{min} }}} \right. \kern-0pt} { \hbox{min} }}} \right) = 4.55 \,\times\, {\text{V}}^{\prime}{\text{CO}}_{2} \left( {{{\text{l}} \mathord{\left/ {\vphantom {{\text{l}} { \hbox{min} }}} \right. \kern-0pt} { \hbox{min} }}} \right) - 3.21{\text{V}}^{\prime}{\text{O}}_{ 2} \left( {{{\text{l}} \mathord{\left/ {\vphantom {{\text{l}} { \hbox{min} }}} \right. \kern-0pt} { \hbox{min} }}} \right) - 0.459 \,\times\, {\text{Pox}} $$

where Pox is the protein oxidation rate. The protein oxidation rate was estimated by assuming that protein oxidation contributed approximately 12 % of resting energy expenditure [25]:

$$ \Pr {\text{otein}}\,{\text{oxidation}}\,{\text{rate }}\left( {{{\text{g}} \mathord{\left/ {\vphantom {{\text{g}} {\hbox{min} }}} \right. \kern-0pt} {\hbox{min} }}} \right) = {{\left[ {{\text{energy expenditure }}\left( {\text{kJ/min}} \right) \times 0.12} \right]} \mathord{\left/ {\vphantom {{\left[ {{\text{energy expenditure }}\left( {\text{kJ/min}} \right) \times 0.12} \right]} {16.74 \, \left( {\text{kJ/g}} \right)}}} \right. \kern-0pt} {16.74 \, \left( {\text{kJ/g}} \right)}} $$

For each subject, the results of the graded exercise test were used to compute the relationship between fat oxidation rate as a function of exercise intensity expressed as %V′O2peak and %HRmax. The best fit was obtained with a polynomial relationship of the second order. The relationship between fat oxidation rate and %V′O2 was used to determine the exercise intensity which corresponded to the highest rate of fat oxidation. The graded exercise test on the motorized treadmill was performed in the same conditions (speed and incline) at W0 and W3.

Energy expenditure and substrate oxidation rate during submaximal exercise and post-exercise recovery

After determining the physical capacities of subjects and the exercise intensity that corresponded to the highest rate of fat oxidation, all the subjects were randomly split into three subgroups: 11 subjects participated in a LI training, 9 subjects in a HI training and 10 subjects in a HIIT program.

The submaximal tests took place two days after the V′O2peak test. Tests were designed in such a way that equal amounts of energy were expended during LI, HI and HIIT exercises on a motorized treadmill (TechnoGym, Gambettola, Italy). Subjects were asked to avoid strenuous exercise the day before the test and came to the laboratory after a 12-h overnight fast.

The LI exercise test comprised a 10-min rest period in a standing position on a treadmill, followed by about 45-min walking at maximal fat oxidation rate intensity previously determined for that subject and then a 60-min post-exercise recovery period. The HI exercise test comprised a 10-min rest period in standing position on a treadmill, followed by about 30-min walking at about 70 % of V′O2peak and then a 60-min post-exercise recovery. The HI exercise corresponded to the maximum exercise intensity that these subjects could maintain for 30 min continuously. The HIIT exercise test comprised a 10-min rest period in standing position on a treadmill, followed by about 37-min walking intermixed with 6 high-intensity intervals at 100 % of V′O2peak and then a 60-min post-exercise recovery.

V′O2 and V′CO2 were measured continuously (CPX Express, Medical Graphics Corp, MN, USA) during the rest, exercise and post-exercise recovery periods. Substrate oxidation rate was calculated over consecutive 5-min periods using the equations of Frayn (1983) [25] as described above. Energy supply (kJ min−1) during exercise and during post-exercise recovery was calculated as the sum of each substrate oxidation rate (g min−1) multiplied by the appropriate conversion factor (carbohydrates and protein = 16.7 kJ g−1; fat = 37.7 kJ g−1). During the exercise tests and the post-exercise recovery period, an electrocardiographic record was performed continuously and displayed online for visual monitoring, and HR was measured with a dedicated monitor device (Polar, Finland).

Statistical analyses

Statistical analyses were performed using PASW Statistic 18 (SPSS Inc., IL, USA) with significance set at p < 0.05. All results were expressed as means and standard deviation (SD). Any associations of the period (W0 vs W3), group (LI vs HI vs HIIT) and interaction (period × group) with physical characteristics, body composition, BMR, energy expenditure and substrate oxidation were tested with general linear model repeated measures. Post hoc comparisons were made using Bonferroni procedure for significant differences.

Results

Physical characteristics of subjects

Before the beginning of the weight-management program (W0), no significant differences were observed in the anthropometric and body composition characteristics between the three subgroups (Table 1). Mean BM, BMI, FFM and %FM were 111.8 ± 11.3 kg, 36.9 ± 3.6 kg m−2, 69.8 ± 6.1 kg and 37.5 ± 1.8 %, respectively. BMR and V′O2peak, expressed in absolute and in relative values, were not significantly different between the subgroups (on average: BMR, 9.21 ± 0.84 MJ d−1 and 0.13 ± 0.01 MJ kg FFM−1 d−1; V′O2peak: 3.44 ± 0.50 L min−1 and 0.049 ± 0.006 L kg FFM−1 min−1).

Table 1 Physical characteristics of adolescents before (Week 0) and at the end (Week 3) of the multidisciplinary weight-management program
Full size table

Between W0 and W3, the decreases in BM and BMI were significantly greater in the LI than in the HI and HIIT subgroups (BM: −8.4 ± 1.5 vs −6.3 ± 1.9 and −4.9 ± 1.3 kg; BMI: −2.8 ± 0.4 vs −2.2 ± 0.6 and −1.6 ± 0.5 kg·m−2, p < 0.05).

FFM and FM decreased significantly in all the three subgroups (p < 0.005), but changes were greater in the LI than in the HI and HIIT subgroup (FFM: −4.2 ± 1.4 vs −3.5 ± 2.4 and −2.2 ± 2.2 kg, p < 0.05; FM: −4.2 ± 1.9 vs −2.8 ± 1.2 and −2.3 ± 1.4 kg, p < 0.05).

BMR expressed both in absolute and in relative values did not change significantly between W0 and W3. V′O2peak expressed both in absolute and in relative values did not change significantly in the LI subgroup, a significant increase (p < 0.005) being found in the HI (by 0.43 ± 0.39 L min−1 and 0.009 ± 0.005 L kg FFM−1 min−1) and HIIT (by 0.36 ± 0.21 L min−1 and 0.007 ± 0.004 L kg FFM−1 min−1) subgroups.

Energy expenditure and substrate oxidation rate during the incremental test

Relationships between EE (J min−1) and substrate oxidation rate during walking at different intensities expressed as a percentage of V′O2peak are reported in Fig. 1. At W0, mean energy expenditure, fat and carbohydrate oxidation rate were not significantly different between the LI, HI and HIIT subgroups at all exercise intensities. Energy expenditure and carbohydrate oxidation rate, expressed as absolute values, increased progressively with exercise intensity, whereas fat oxidation rate increased up to 45 % V′O2peak and then decreased. In fact, with increasing exercise intensity, fat oxidation rate increased by mean to a maximum of 0.44 ± 0.05 g min−1 at 44 ± 4 % V′O2peak (Fig. 1g, h, i). At exercise intensities above 57 ± 5 % V′O2peak, fat oxidation rate decreased markedly and carbohydrates became the predominant source of energy for oxidation. The contribution of fat oxidation to energy supply became negligible above 72 ± 8 % V′O2peak.

Fig. 1
figure 1

Energy expenditure (EE, J·min−1, panels a, b, c), carbohydrate (CHO, g·min−1, panels d, e, f) and fat (g·min−1, panels g, h, i) oxidation rates as a function of exercise intensity expressed as percent of peak oxygen uptake (V′O2peak) in LI (panels a, d, g), HI (panels b, e, h) and HIIT (panels c, f, i) groups, before (Week 0, opened circle) and at the end (Week 3, filled circle) of the multidisciplinary weight-management program

Full size image

At W3, energy expenditure, carbohydrate and fat oxidation rates were not significantly different from those at W0 with all exercise intensities in the LI subgroup (Fig. 1 a, d, g). By contrast, the HI and HIIT subgroups exhibited a greater absolute rate of fat oxidation between 50 and 70 %V′O2peak (Fig. 1 h, i), while the energy expenditure and carbohydrate oxidation rate in W3 were not significantly different from those at W0 with all exercise intensities (Fig. 1 b, e, c, f).

Energy expenditure and substrate oxidation rate during the submaximal exercise and during the post-exercise recovery period

At W0, energy expenditures (EEs, Table 2) during the LI, HI and HIIT exercises were not significantly different (1521 ± 256 vs 1527 ± 102 vs 1494 ± 289 kJ, respectively, p = 0.566). Similarly, energy EEs during the post-exercise recovery period were not significantly different between the subgroups (488 ± 134 vs 527 ± 65 vs 498 ± 113 kJ, respectively, p = 0.304). Consequently, total EEs during the exercises and post-exercises recovery periods were not significantly different between the LI, HI and HIIT subgroups (2009 ± 378 vs 2054 ± 93 vs 1992 ± 391 kJ, p = 0.450).

Table 2 Energy expenditure and substrate oxidized during LI, HI and HIIT exercises and post-exercise recovery period before (Week 0) and at the end (Week 3) of the multidisciplinary body weight-management program
Full size table

However, during the exercise periods, energy from fat and the amount of fat oxidized was significantly greater in the LI group than in the HI and HIIT subgroups (+367 and +347 kJ; +10 and +9 g; respectively; p < 0.001), whereas energy from carbohydrates and the amount of carbohydrates oxidized was significantly lower in the LI group than in the HI and HIIT subgroups (−371 and −312 kJ; −21 and −19 g; p < 0.001) (Table 2). In addition, during the post-exercise recovery period, no significant differences were shown between the LI, HI and the HIIT groups, either in fat or carbohydrate oxidation rates (Table 2).

Between W0 and W3, EEs, fat and carbohydrate oxidation rates did not change significantly in the LI, HI and HIIT groups, during the exercise and the post-exercise recovery period (Table 2).

Discussion

The 3-week multidisciplinary body weight-management program entailing moderate energy restriction, nutritional education, psychological counseling (common to all subjects) and three different exercise training (LI, HI or HIIT) results in: (1) greater decreases in BM, FFM and FM in the LI than in the HI and HIIT subgroups; (2) significant improvement of V′O2peak in the HI and HIIT subgroups; (3) significant increase in fat oxidation rate during the incremental test in the HI and HIIT groups; and (4) greater fat oxidation (in average 44 %) and lower carbohydrate oxidation (in average 70 %) during the submaximal exercises in the LI than in the HI and HIIT subgroups, both at W0 and W3.

The main objectives of the body weight-management programs in specialized institutions are to promote a correct lifestyle, to reduce FM, to maintain FFM and to improve cardiovascular fitness, thus enabling obese adolescents to increase spontaneous physical activities. Previous studies have shown that physical training associated with energy restriction (compared with energy restriction group alone) is a highly effective method in improving indexes of health, determining FM loss and preserving FFM during childhood [1, 26, 27]. In addition, increased physical capacities and optimizing fat oxidation during exercise are essential for promoting the reduction in adipose tissue [28]. In the present study, the LI, HI and HIIT groups had the same balanced diets formulated according to the Italian recommended dietary allowances [20]. Therefore, the greater decrease in FM observed between W0 and W3 in the LI than in the HI and HIIT subgroups could be attributed to the modality of training. This assumption concurs with the finding of greater fat oxidation rate in the LI than in the HI and HIIT groups during the submaximal exercise, as suggested by the results obtained during the incremental test in which the maximal fat oxidation rate was reached at 44 ± 4 % V′O2peak. Therefore, LI exercise, corresponding to approximately 40 % V′O2peak, seems an adequate level of physical activity to optimize fat oxidation rate, as observed in previous studies performed in obese children [29] and adolescents [5]. Otherwise, in children and adolescents, LI exercise is perceived as a boring exercise compared to other exercise modalities [10], which would have implications for participation in adherence to this type of activity, which could favor greater autonomous motivation [30].

At W3, HI and HIIT groups showed a greater increase in V′O2peak than in LI (+12 and +11 %, respectively, p < 0.005), as previously shown [7, 15]. V′O2max is an important marker of cardiometabolic health, and V′O2max levels in youth predict cardiovascular disease in later life [31]. Particularly, several authors have shown that HI and HIIT improve V′O2max favoring central effects (such as increasing maximal cardiac output, total hemoglobin and blood plasma volume) [32] and/or peripheral adaptations with an improved ability to extract and use available oxygen because of increased muscle oxidative potential [33]. In addition, in the HI and HIIT groups, fat oxidation rate, during the incremental test, was significantly greater in W3 than in W0, particularly between 50 and 70 %V′O2peak (Fig. 1h, i). This difference could be related to the significantly greater V′O2peak observed in W3 than in W0. In the present study, V′O2peak increased significantly by mean 12 % between W0 and W3 in the HI and HIIT groups. Consequently, absolute workloads with comparable  %V′O2peak were greater in W3 than in W0, which could explain their greater absolute fat oxidation rate during exercise. The increase in fat oxidation rate (without changes in CHO oxidation rate) during exercise in obese subjects actually represents an interesting result because of improved metabolism and fat loss. Particularly, previous studies [34, 35] reported that obese subjects exhibited an earlier shift from lipid to CHO-derived fuels during exercise, compared to nonobese subjects. The earlier and greater participation of anaerobic glycogenolysis to energy production during exercise could be related to mitochondrial malfunction [36], or inefficient/incomplete b-oxidation [37], as well as to modifications of hormonal interactions [38], higher insulin and leptin levels [39] and reduced growth hormone and catecholamines responses to exercise [40]. Favoring fat oxidation rate during exercise could exert beneficial effects in long-term body weight reduction programs, although further additional studies are requested to confirm these preliminary observations.

Moreover, increasing or maintaining FFM is an important objective of the weight-management program because it is strongly related to physical capacities and BMR [41], which accounts for 60–70 % of daily energy expenditure in obese adolescents [42]. In the present study, the LI, HI and HIIT exercises were able to reduce FM, but not to maintain FFM. The lack of positive effects on FFM is not surprising after short-term training periods as in the present study, these effects being observed only with the inclusion of a strength training component in obese adolescents [43].

With reference to the submaximal exercises, previous studies have shown that exercise intensity also affected energy expenditure and fat oxidation rate during the post-exercise recovery period [44, 45]. However, in the present study, the LI, HI and HIIT equicaloric exercises did not significantly affect energy expenditure, or fat and carbohydrate oxidation rate during the post-exercise recovery period. Although the post-exercise period considered in our study was relatively short (60 min), our results are in agreement with previous studies which considered the effects of equicaloric LI and HI exercises on 24-h energy expenditure and substrate oxidation [46, 47].

Finally, several factors such as the patients’ adherence to exercise contribute to the effects of the weight-management programs. Recent studies have reported that HIIT is perceived as a more enjoyable exercise compared to other exercise modalities and it appears more efficient in improving aerobic capacity and blood pressure, especially as compared to moderate-intensity continuous training [10]. Furthermore, both HI and HIIT programs seem to represent an effective alternative to LI program scheduled in a multidisciplinary body weight-management program, with the advantage of a lower volume of activity and reduced time commitment.

In conclusion, our data show that a 3-week weight-management program induces a greater decrease in BM and FM in the LI than in the HI and HIIT groups, and greater increase in V′O2peak and fat oxidation rate in the HI and HIIT than in the LI group.