Loading ...

A lightweight wearable robotic device that provides adaptive resistance training enabled children with spinal muscular atrophy (SMA) type II to stand up unassisted after just 6 weeks of training, with many of the improvements persisting even after therapy ended.

The results suggest that targeted robotic resistance training may promote lasting neuromuscular recovery in this patient population.

“Watching a child who couldn’t stand unsupported eventually rise from a chair on their own and keep doing it is genuinely moving,” study investigator Yanggang Feng, PhD, with the School of Mechanical Engineering and Automation, Beihang University, Beijing, China, told Medscape Medical News.

The research was published online on May 20 in Nature.

Gains in Strength, Muscle, and Mobility

SMA is a rare inherited neuromuscular disorder caused by mutations in the SMN1 gene, leading to degeneration of motor neurons and progressive muscle weakness. Children with SMA type II can generally sit independently but are unable to walk unaided.

“Over the past decade, a major breakthrough in wearable robotics has been the ability to significantly reduce the energetic cost of walking by assisting movement,” Feng said.

“However, for patient populations requiring neural plasticity repair, such as those with SMA, excessive assistance may actually deprive the neuromuscular system of necessary exercise and potentially accelerate disuse atrophy,” he added.

Feng noted that low-activation conventional training may merely maintain existing musculature, whereas high-activation isokinetic training can yield significant strength improvements and hypertrophy.

The device developed by Feng and colleagues is a wearable knee robot weighing just 0.96 kg. It uses a back-drivable motor and a variable stiffness mechanism to deliver customized resistance training safely at home.

They tested the device in a clinical trial involving six nonambulatory children, aged 6-10 years, with SMA type II who had previously been unable to rise from a seated position without support.

Participants first completed a 6-week control period during which they continued their usual physiotherapy routines but showed no measurable improvement in standing ability. This was followed by 6 weeks of intensive training with the device, using it 5 times a week, followed by 6 additional weeks of lower-intensity training and more than 30 days of follow-up after the robotic therapy ended.

Significant, Durable Improvement

Following the training, all six children were able to rise from a chair without external support, although self-assist using their hands on their knees was allowed.

During the intervention phase, the children performed repeated knee extension exercises while the robot maintained a constant angular velocity. A gamified mobile application provided real-time feedback on knee angle, torque, and speed to encourage participation.

The improvements were substantial. Across the group, peak knee torque increased by 130%, range of motion improved by 51%, and mechanical work generated during knee extension rose by 97%.

Imaging also revealed substantial muscle growth: quadriceps muscle volume increased by 19%, anatomic cross-sectional area by 12%, and physiologic cross-sectional area by 21%.

In addition to muscular changes, the researchers documented neurologic improvements. Femoral nerve conduction, measured through compound muscle action potential, improved by an average of 19%, suggesting enhanced motor neuron recruitment and better neuromuscular signaling. Bilateral coordination between the legs improved by 35%.

“Parents described tangible functional improvements that don’t show up on any formal motor scale: standing more safely with reduced fall risk, independent leg lifting to climb onto a bed from standing, improved lower‑limb coordination for rolling, and more stable squatting and sustained static standing endurance,” said Feng. “One parent even stated that this was the most effective rehabilitation therapy their child had ever received,” he added.

Functional gains in strength, range of motion, muscle size, and nerve activity were largely maintained during follow-up periods in which participants either performed lower-intensity exercise or returned to conventional physiotherapy.

“Importantly, while the robotic system acts as the catalyst for high-intensity training, sustained daily low-intensity exercise remains necessary to preserve these functional gains after the intervention concludes,” Feng told Medscape Medical News.

Limitations of the study include the small sample size and lack of a randomized control group because SMA is a rare disease and families were reluctant to place children in a nontreatment group.

Feng noted that this research has been primarily self-funded up to this point.

“In the near future, our primary goal is to secure sufficient funding to make this technology accessible worldwide. However, regardless of the funding situation, we are fully committed to doing our absolute best to expand its reach to as many patients as possible,” Feng told Medscape Medical News.

The team is also exploring whether the approach could help patients with other neuromuscular diseases.

“Our next focus is SOD1-linked amyotrophic lateral sclerosis (ALS). With disease-modifying medications now available for this genetic condition, we are highly interested in exploring whether pairing these pharmacological treatments with our high-intensity robotic training can further maximize and promote neuromotor recovery,” said Feng.

The authors declared having no conflicts of interest.