Efficacy of a mobile health with tailored personalized exercise rehabilitation program for ADHD children: randomized controlled trial

Abstract

Background

Interventions delivered via mobile health (mHealth) could increase access to guided exercise interventions. We aimed to investigate effectiveness of a self-developed mHealth app-supported tailored exercise rehabilitation program on core symptoms (i.e., inattention and hyperactivity/impulsivity) and executive functions (EFs, specifically inhibitory control, working memory, and cognitive flexibility) in children with attention-deficit/hyperactivity disorder (ADHD).

Methods

Children with ADHD aged 6–12 (n = 144) were randomly assigned to one of three groups and were offered interventions 3 times a week over a period of 12 weeks, and each session lasted 45–60 min. Experimental group 1 (EG1, n = 53) received exercise interventions through videos from an exercise library that were recommended based on each participant's assessment results via mHealth used by both children and their parents. Experimental group 2 (EG2, n = 45) received integrated merging of physically and cognitively demanding exercise interventions implemented offline by coaches, and control group (CG, n = 46) received minimal intervention with physical activity education. Outcomes were parent-rated symptoms of inattention and hyperactivity/impulsivity, which were assessed using the Swanson, Nolan, and Pelham-version IV rating scale, and three EFs via the Stroop Color and Word Test for inhibitory control, the Rey-Osterrieth Complex Figure Test for working memory, and the Trail Making Test for cognitive flexibility. A linear mixed effects model was used to investigate the intervention effect, and the dependent variable was changed from baseline to the endpoint. Potential confounding variables (i.e., age, sex, intelligence quotient, ADHD subtype, and medicine intake) were used as covariates.

Results

EG1 (β = −3.54, SE 0.96, p < 0.001) and EG2 (β = −3.98, SE 1.00, p < 0.001) significantly reduced inattention scores compared to CG, and EG1 (β = −2.88, SE 0.81, p < 0.001) and EG2 (β = −3.50, SE 0.84, p < 0.001) also significantly reduced hyperactivity/impulsivity scores compared with CG. EG1 (β = −8.48, SE 3.27, p = 0.016) and EG2 (β = −9.41, SE 3.46, p = 0.016) significantly improved inhibitory control compared with CG. EG1 (β = 4.00, SE 1.07, p < 0.001) and EG2 (β = 5.72, SE 1.14, p < 0.001) significantly improved delayed details of working memory compared with CG. For cognitive flexibility, significant group differences were observed in EG1 versus CG (β = −17.9, SE 8.57, p < 0.039), EG2 versus CG (β = −39.80, SE 9.55, p < 0.001), and EG1 versus EG2 (β = 21.9, SE 9.12, p = 0.027).

Limitation

Lack of posttraining follow-up.

Conclusions

mHealth with a tailored personalized exercise rehabilitation program had favorable effects on core symptoms and EFs in children with ADHD, and demonstrated comparable effects to offline face-to-face guidance in most cases.