The Pilates Training program consisted of warm-up, work-out and cool-down sections. For the warm-up section, breathing, standing roll down, and spine twist exercises were selected; for the work-out section, the hundred, shoulder bridge, scissors, swimming, and quadruped exercises, and for cool-down section, torso twist, cat-camel, and standing stretches. The Pilates Training progression was achieved by increasing the number of repetitions of the selected exercises and adding some equipment (balls, foam rollers) to challenge postural stability. In the first five sessions, after each exercise was performed for six to eight repetitions, the number of repetitions was increased to 10-12. Moreover, the environment where the training took place was carefully arranged according to the needs of children with autism.

Each training session consisted of a 20-minutes treadmill workout (using the motorized Ultıma Ac 3500 treadmill) followed by 20-minutes bicycle workout (Voit Yellow Collection 112u exercie bike). The intensity of training, which began at 50% to 60% of maximum heart rate (HR), was increased each week so that by week five, the children were at 75% to 80% maximum HR. The maximum HR, as recommended for children, was calculated using the formula 208 - 0.7 x (age) (Mahon et al., 2010). To ensure that the exercise intensity remained in the target HR zone during training, the child's HR was continuously monitored from the used treadmill and exercise bike monitors.

Eighteen children with autism did not meet the inclusion criteria, ten children did not want to participate in the study, and five children were not included in the study due to different reasons. A total of 22 children between the age of 7 and 12 years were participated in this study. The inclusion criteria were (i) being diagnosed with autism, (ii) not having physical disabilities or neurological disorders, (iii) having the ability to follow the instructions given during evaluations and training, and (iv) having the ability to perform the requested training. All parents of participants signed an informed consent form for participation. were divided into two groups including Aerobic Training (n=11) and Pilates Training (n=11) groups by simple randomization method (internet-based). Since one child in the experimental group did not continue Pilates Training, the study was completed with the participation of 21 children with autism.

Evaluations of maximal isometric muscle strength for shoulder flexors, hip flexors and knee extensors were performed with Hand-held dynamometry (Jtech Commander PowerTrack II Muscle Tester). Participants were asked to apply the maximum voluntary force they could exert under certain test conditions. Each test procedure for each muscle groups was repeated 3 times, and the best score was used for analysis. The results were recorded for both the right and left side.

The one-legged standing test was used to evaluate static balance performance. The participants were asked to maintain a one-legged stance for as long as they can with their eyes open and arms at the side of the trunk. The time recorded in seconds from the moment a foot is lifted off the ground to the moment it touches the ground. The test was performed for right and left side. The participants performed 3 trials, and the longest time was used for analysis

The test has two sub-test including locomotor skills tasks such as hopping, sliding, galloping, jumping, running, and the object control skills consists of striking and kicking a stationary ball, dribbling, catching, overhand throwing and underhand rolling. The locomotor skills focus on coordinated movements, while object control skills focus on the child's ability to play with or manipulate balls

The Autism Social Skills Profile Turkish Form (ASSP- T) was used to evaluate social skills of participants. The test consists of items related to social participation and socially inappropriate behaviors that lead to negative peer relationships. Higher score indicates proficiency in social functions, and lower score indicates inadequacy in social functions