An adaptive deep brain stimulation (aDBS) system, personalized to sync with the walking cycle, improved gait stability and reduced falls in patients with Parkinson’s disease (PD), a small feasibility study showed.

Unlike conventional DBS, which delivers continuous stimulation, the aDBS system adjusts stimulation within fractions of a second based on patient-specific neural signals associated with each gait phase. 

In laboratory testing, the system improved gait symmetry and consistency, while a blinded home crossover study showed a reduction in falls without compromising overall motor control.

“Walking is one of the most dynamic human behaviors,” senior investigator Doris D. Wang, MD, PhD, associate professor of neurological surgery at the University of California, San Francisco (UCSF), told Medscape Medical News. 

“Our findings suggest that a ‘one-size-fits-all’ stimulation approach may not be optimal for gait. In fact, continuous stimulation may inadvertently interfere with some of the natural brain activity that supports normal walking. By adjusting stimulation in real time based on what a patient is actually doing — in this case, the timing of each leg during walking — we may be able to better restore physiological brain network activity and improve gait function,” Wang said.

The study was published online on June 15 in Nature Medicine.

Dynamic vs Fixed Stimulation

DBS has become an established treatment for tremor, rigidity, and bradykinesia, yet many patients with PD continue to experience gait impairment, freezing of gait, and falls despite otherwise successful therapy. These symptoms continue to significantly contribute to disability, hospitalization, and loss of independence.

The investigators hypothesized that conventional DBS has limited effects on walking because gait is constantly changing. Every step requires quick coordination between the brain, spinal cord, and muscles, yet continuous DBS delivers the same stimulation regardless of whether a patient is standing, walking, turning, or navigating obstacles.

Unlike prior aDBS systems that respond to slower disease-related neural signals, the UCSF aDBS system adapts in real time to movement-related brain activity, modulating stimulation based on behavior rather than disease state.

For the study, investigators used additional motor cortical recording electrodes to identify individualized neural signals associated with left- and right-leg movements and specific gait phases.

The study enrolled five patients with PD (age, 53-68 years; four men and one woman) who already had DBS electrodes implanted in the globus pallidus internus — a brain region involved in movement control. In addition to their therapeutic DBS leads, participants had electrodes placed over motor cortical areas, allowing investigators to record brain activity associated with left- and right-leg movements during walking. 

Using machine-learning algorithms, investigators identified individualized neural biomarkers linked to different phases of gait. These biomarkers were then programmed into an implanted bidirectional neurostimulator, allowing the device to independently adjust stimulation in each hemisphere on a subsecond timescale without external devices.

Fewer Falls, Persistent Freezing

During laboratory testing, aDBS significantly improved step-length asymmetry (P < .001), step-time asymmetry (P < .05), and reduced gait variability compared with clinically optimized continuous DBS. These findings are consistent with more stable and efficient walking.

During the blinded home crossover phase, participants experienced significantly fewer falls during adaptive stimulation than during conventional DBS (odds ratio, 4.35; 95% CI, 1.07-20.22; P = .047). Overall, PD motor symptom control was maintained, and participants tolerated the rapid stimulation adjustments well, with no serious adverse events reported.

However, aDBS did not significantly improve freezing of gait.

This finding reflects important biological differences between the two symptoms, Wang noted.

“Falls and freezing of gait are related, but they are not the same problem,” she said. “Our aDBS approach improved key aspects of gait such as symmetry and step-to-step consistency, which likely made walking more stable and reduced fall risk.”

Freezing of gait is influenced not only by motor circuits but also by cognitive load, emotional state, anxiety, attention, and visual processing.

“Future adaptive stimulation approaches may need to incorporate additional brain signals that capture these nonmotor contributors to freezing,” Wang said.

The study had several limitations, including its small sample size, short follow-up period, and the use of additional research electrodes not currently incorporated into standard DBS systems.

Impressive Feasibility With Caveats

The findings suggest a potential new approach to treating gait impairment but should be viewed as an encouraging feasibility study rather than evidence of clinical efficacy, said Alberto J. Espay, MD, FAAN, professor and endowed chair of the James J. and Joan A. Gardner Family Center for Parkinson’s Disease and Movement Disorders at the University of Cincinnati in Cincinnati.

“The study is technically impressive and conceptually important, but I would be careful not to overstate its clinical implications at this stage, particularly given that it is five patients overall and only three completing the blinded crossover phase,” Espay, who was not involved in the study, told Medscape Medical News.

Despite these limitations, he said the findings suggest that personalized, real-time adaptive protocols may represent a viable alternative to conventional, fixed stimulation.

“This moves adaptive DBS beyond simply responding to broad disease states, such as beta oscillations associated with bradykinesia, and toward responding to specific behaviors in real time,” he said. 

Espay further noted that the study provides an important proof of principle that neuromodulation can be tailored to dynamic motor functions instead of relying on continuous stimulation.

The next step should be a larger, multicenter randomized trial powered for clinically meaningful outcomes, including falls, freezing of gait, mobility, quality of life, and durability of benefit over time, he said.

“Ultimately, the key question is not whether the technology can work, but whether it produces meaningful and sustained improvements in patients’ everyday lives,” he said.

The study was supported by The Michael J. Fox Foundation for Parkinson’s Research, the Burroughs Wellcome Fund Career Awards for Medical Scientists, the National Institute of Neurological Disorders and Stroke, and UCSF Catalyst Grants. Disclosure information for study authors is available in the original study publication. Espay reported no relevant financial disclosures.