Volunteers were eight male obese adolescents’ beginners in a long term multidisciplinary obesity intervention program conducted in the outpatient clinic of the University of Pernambuco/Brazil. Adolescents had an average body mass index (BMI) of 33.06 ± 4.78 kg/m2, an average age of 15.44 ± 2.06 y, and pubertal stage 3 or 4 using the Tanner scale . The study was approved by the ethical committee of the University of Pernambuco (#154/08). Informed consent was obtained from all volunteers and/or their parents.
Volunteers completed a maximal ergoespirometric test and three experimental (crossover - on seven separate days) sessions:1) control, volunteers remained seated for 2 h; 2) low intensity exercise (LIE), volunteers exercised on a treadmill in an intensity corresponding to 10% below ventilatory threshold (VT); 3) high intensity exercise (HIE), volunteers exercised on a treadmill at 10% above VT. For the LIE and HIE sessions, energy expenditure (EE) was set at 350 kcal, estimated by indirect calorimetry . Each participant served as his own control. Experimental procedures were initiated after three familiarization sessions on treadmill.
During the first visit to the laboratory, anthropometric profiles and body composition were measured, peak oxygen uptake (VO2peak) and VT were determined. Obese adolescents were weighed wearing light clothing and no shoes on a Filizola scale (Model 160/300, Brazil) to the nearest 0.1 kg. Height was measured to nearest 0.5 cm by using a wall-mounted stadiometer Filizola (Model 160/300, Brazil). Body composition was determined by bioelectrical impedance (Byodinamics A-310 body composition analyzer) .
Oxygen uptake (VO2) was measured directly in a continuous incremental protocol on a treadmill (Inbrasport Super ATL, Brazil) as previously described . The inclination was set at 1.0%, the initial speed was 4.0 km/h (four minutes). Thereafter, speed was increased to 1.0 km/h every minute. The termination criteria were the following: volitional fatigue, Borg scale and gas exchange ratio higher than 18 and 1.15, respectively. The greatest VO2 obtained before test interruption was considered as VO2peak. VO2 and carbon dioxide production (VCO2) were displayed every 15 seconds using an open circuit respiratory metabolic system (Cortex Biophysik Metalyzer IIB, Germany). Ventilatory threshold was determined by two independent researchers, as the point at which there was a systematic increase in VE/VO2 without a corresponding increase in VE/VCO2.
On the other visits, volunteers arrived at the laboratory around 7 AM after an overnight fast, and were given a standard snack 350 kcal (composed of 61.7% carbohydrates, 13.44% proteins and 24.86% lipids), to minimize potential differences in the thermic effect of food (TEF) between subjects. At 7:30 AM, subjects performed the LIE, HIE or control session. After each trial they remained seated for 2 h. The trials were conducted in a controlled temperature room (21–23°C) and in a randomized order, at the same time of the day in order to avoid any circadian variations. Participants were asked to avoid vigorous exercise for 48 h before the sessions.
Spielberger State-Trait Anxiety Inventory (STAI), translated into Portuguese and validated for the Brazilian population  was used to measure anxiety. STAI consists of a self rated questionnaire divided in two parts: anxiety-trait (referring to personality traits) and anxiety-state (referring to systemic aspects of the context). Each part has 20 statements. Responses are in a 1–4 scale. Anxiety- state refers to how individuals feel ‘at the moment’, and anxiety-trait to how they ‘generally feel’. Each part varies from 20 to 80 points, and the scores indicate low (0–30), medium (31–49) or high (50 or more) anxiety levels .
Profile of Mood States (POMS), translated into Portuguese and validated for the Brazilian population was used to measure mood. Internal consistency for the Brazilian version was reported at 0.62 to 0.91 Cronbach alpha rating . POMS evaluates six mood factors or transitory affective states: Anxiety-Tension, Depression, Anger-Hostility, Fatigue, Vigor, and Mental Confusion. The scores derive from the ratings given to 65 adjectives. Each of the 65 items is scored from 0 (not at all) to 4 (extremely), to detect changes in tension/anxiety, depression, anger, vigor, fatigue and confusion. The final score (Total Mood Disorder) is obtained by subtracting the total of the five factors associated with negative emotions from the only factor of positive emotions (Vigor) .
Participants were asked to complete visual analogue scales (VAS) to rate hunger. The questionnaire consisted of six visual analogue scales to rate hunger, fullness, desire to eat, how much you can eat now, urge to eat and preoccupation with thoughts of food. The VAS consisted of horizontal lines of 100 millimeters with defined marks at each extremity that portrayed the expressed desire to eat excessively in the right extremity and have and lack the desire to eat or not hungry in the left extremity . VAS was measured by hand, from left (minimum score of 0 mm) to right (maximum score of 100 mm). In the present study we used only hunger ratings.
Data are presented as median (v.min – v.max). Trials were compared using Kruskal-Wallis test, while differences between basal and acute time points were assessed by Wilcoxon rank test. The relationships between anxiety, mood and hunger feelings were analyzed by Spearman Rank test. Significance was set at p ≤ 0.05. For all analysis, statistical power was calculated a posteriori, and in all cases power was greater than 0.80. All analysis was performed in STATISTICA® version 8 for Windows® software.