To explore the functional exercise methods of children with autism spectrum disorder (ASD) after a non-displaced supracondylar fracture of the humerus.

Inclusion criteria: three to seven years old children treated with plaster external fixation who were followed up for not less than six months. Exclusion criteria: combined with other fractures, failure to adopt a conservative treatment, previous ipsilateral upper limb activity limitation, symptomatic ASD children, other congenital diseases. A total of 22 children diagnosed as ASD at the Children Development and Behavior Center in the Third Affiliated Hospital of Sun Yat-Sen University from January 2015 to December 2018 were included and then divided into functional exercise ASD group (15 cases), non-functional exercise ASD group (seven cases) . Normal developmental children (44 cases) with the same type of fracture were set as non-ASD children group. Chi-square test was used to compare the passive range of motion (PROM) of the elbow joint and the ankle joint at two months and six months after fracture in the three groups. For the changes in self-care ability of children with ASD, the rank sum test was used for comparison.

In the non-functional exercise ASD group, the flexion(F=5.974, P<0.05) and extension(F=26.968, P<0.05) of the elbow joint, and the rotation of the radio-ulnar joint (pronation F=5.162, P<0.05; supination F=30.432, P<0.05) were significantly worse than those of the other two groups at two months after the fracture. The elbow joint extension (F=12.318, P<0.05) and the radioulnar joint external rotation(F=24.504, P<0.05) at six months after the injury were still worse in the non-functional exercise group. The self-care ability of the non-functional exercise ASD group showed significantly retrogression at six months (Z=-2.225, P<0.05).

Functional training of fracture should be carried out in conservative treatment of non-displaced humeral supracondylar fractures in ASD children, which can help improve joint mobility and prevent self-care ability from regressing.