Fewer Falls, Better Balance: What Real World Data Shows About Adaptive DBS

In the second part of our conversation with Doris Wang, M.D., Ph.D., neurosurgeon at the University of California, San Francisco, she explains what her team's adaptive deep brain stimulation, or aDBS, system is actually correcting and why early results from long-term patient testing are so encouraging.

At the core of the technology is a concept called gait variability. In healthy walking, the length and timing of each step remain remarkably consistent from one stride to the next. In patients with Parkinson's disease, that consistency breaks down. Walking requires more conscious effort, and the timing and length of left and right steps often differ significantly from one another. By detecting these patterns and adjusting stimulation in real time, the adaptive system aims to restore symmetry and rebuild a more stable, cyclical walking pattern, helping patients maintain better balance as they shift their center of gravity with each step.

The results from long-term, real-world testing, published in Nature Medicine, have been striking. Patients in the study reported a decreased number of falls while using the adaptive stimulation. Even more notable, several patients chose to remain on the adaptive setting entirely on their own after the study's blinded phase ended, with one patient continuing on it for more than a year. When the team temporarily switched these patients back to standard, continuous stimulation, the patients reported a sharp increase in freezing episodes and falls, along with a noticeable decline in gait function, strong evidence that the adaptive approach is providing real, measurable benefit.

Wang compares the underlying concept to a heart pacemaker, which responds dynamically to the heart's rhythm rather than delivering a fixed, constant signal. The brain, she notes, is exponentially more complex, involving billions of neurons in constantly shifting states. Her team's system allows for far more nuanced control, adjusting not just stimulation amplitude but also pulse width and targeted brain regions in response to what the brain needs in the moment.

Looking ahead, Wang identifies several key hurdles. Access to research-level devices with these capabilities unlocked remains limited, since most commercially available DBS systems do not yet have these features enabled. Larger, multi-institutional randomized trials are needed to determine which patients and which types of gait symptoms respond best to this approach. And perhaps most importantly, her team hopes to develop more automated methods of biomarker detection, potentially using wearable devices and machine learning to streamline a discovery process that currently takes years of in-clinic and in-home testing to complete.