Intervention group performed 3 weekly sessions during 8 weeks in the intra-hospital gymnasium. Each session lasted 50-60 min, and started and ended with warm-up and cool-down periods (10-15 min).

The core session included bench press, leg press, lateral row, leg extension, lateral pull-down, abdominal crunch, low back extension, and push-ups exercises. The participants performed three sets of 8-10 repetitions with resting periods of 1-2 min on the weightlifting machines. Load was gradually increased 5-10 % as the participant strength was adapted. The load started at 70% of 6RM.

Functional exercises (abdominal crunch, low back extension, and push-ups) were performed at the end of the session to strengthen the core musculature consisting of 3 sets of 15s of isometric contractions at the beginning of the program and 30 s at the end. Dynamic contractions were added, starting with 10 repetitions at the beginning to 30 repetitions at the end of the program.

Muscular strength was assessed in the upper and lower body following a standardized strength testing protocol in the same resistance weight machines used in training sessions. The weightlifting training machines that were specifically built for the body size of children and adolescents (Strive Inc, PA, US) were used to carry out the 6RM (maximum repetitions) measured in Kilograms (kg). The lower body strength was assessed with seated leg-press, and upper body strength was assessed with seated bench press and seated lateral row

To measure agility, we used the Time Up and Go (TUG) 3m and 10 m tests, and the Timed Up and Down Stairs (TUDS). These tests are reliable and validated for healthy children and adolescents, and used with other chronic pathologies such as cancer.

Body composition was obtained by collecting weight (kg) and height (m), and BMI was calculated as kg/m2. To obtain the % Fat Mass Heyward (2004) equation (specific for anorexia nervosa patients) was used. The skeletal muscle mass (SMM) was calculated using Poortmans (2005) equation. Fat Free Mass, and Fat mass where calculated based on Heyward and Poortmas equations.

Skinfolds, diameters and perimeters where measured following the International Society for the advancement of the Kinanthropometry (ISAK) guidelines.

Functional capacity was assessed by a graded exercise test on a treadmill in conjunction with an ECG response, and under similar environmental conditions. Treadmill speed began at 3.0 km/h with a grade of 5.0%; both were increased by 0.3 km/h and 0.5% respectively every 30s. The test was terminated upon volitional fatigue of the patient or when they showed loss of ability to maintain the required workload.VO2peak was determined as the higher value obtained after a period of 20s. Ventilatory Threshold (VT) was determined using the O2 equivalent (VE∙VO2-1) and O2 end-tidal pressure (PetO2) without increases in the CO2 equivalent (VE∙VCO2-1).

Patients were required to maintain their normal physical activity while they were using an uni-axial accelerometer (Actigraph MTI, GT1M model, Manufacturing Technology Inc., Fort Walton Beach, FL, USA). At least 7 days were recorded (Monday-Sunday) with a minimum of 10-hour registration per day was set as an inclusion criterion. The time sampling interval (epoch) was set at 15 seconds. Average physical activity intensity was expressed as (sum of counts registered per epoch)/(Total Time Registered) (cpm). Time spent in light, moderate, vigorous, very vigorous and moderate to vigorous physical activity (MVPA) was presented as means per day depending on the metabolic equivalents (MET). Time spent in low intensity physical activity or sedentary time was the total time registered minus the total time active. Time spent in moderate to vigorous physical activity was presented as means of minutes per day