Keith Krehbiel lived with Parkinson’s disease for nearly 25 years before agreeing to try a brain implant that might alleviate his symptoms. He had long been reluctant to submit to the surgery. “It was a big move,” he says. But by 2020, his symptoms had become so severe that he grudgingly agreed to go ahead.
Deep-brain stimulation involves inserting thin wires through two small holes in the skull into a region of the brain associated with movement. The hope is that by delivering electrical pulses to the region, the implant can normalize aberrant brain activity and reduce symptoms. Since the devices were first approved almost three decades ago, some 200,000 people have had them fitted to help calm the tremors and rigidity caused by Parkinson’s disease. But about 40,000 of those who received devices made after 2020 got them with a special feature that has largely not yet been turned on. The devices can read brain waves and then adapt and tailor the rhythm of their output, in much the same way as a pacemaker monitors and corrects the heart’s electrical rhythms, says Helen Bronte-Stewart, a neurologist at Stanford University in California.
Bronte-Stewart received approval to start a clinical trial of this new technology, known as adaptive deep-brain stimulation (aDBS), the same week that Krehbiel was preparing for surgery. He recalls the phone call in which she asked him if he wanted to be her first participant: “I said, ‘Boy, do I!’”
On supporting science journalism
If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Five years on, the results of this 68-person trial, called ADAPT-PD, are under review for publication. Although the exact details are still under wraps, they were convincing enough to earn approval for the technology earlier this year from both US and European regulators.
The results of this study could be a boon for the estimated one million people in the United States and 1.2 million people in Europe who now live with Parkinson’s disease — and for Medtronic, the health-care technology company based in Minneapolis, Minnesota, that manufactures the implants. The seeming success also clears a path for other companies around the world that are racing for approval on advanced devices, says Martijn Beudel, a neurologist at Amsterdam University Medical Center who worked on the trial.
The approaching wave of therapies promises to level up DBS for Parkinson’s and other motor conditions. The technology might even help to treat neurological disorders such as Tourette’s syndrome and psychiatric conditions, including obsessive–compulsive disorder (OCD) and depression. Several clinicians suggest that the new technology could have even more profound health implications — that is, if funding for US brain-implant research isn’t cut.
Deep trade-offs
Since DBS was first approved in Europe and the United States in the late-1990s, the vast majority of devices have been given to people with Parkinson’s disease. Parkinson’s is a progressive disorder, typified by the death of neurons that produce the neurotransmitter dopamine, which is key to controlling movements.
Existing drugs that aim to increase dopamine levels can only manage the symptoms. They can’t match the constant dopamine production of a healthy brain. “No matter how clever we are with it, we have never been able to exactly mimic the way the brain supplies it,” says Bronte-Stewart. This means that symptoms vary throughout the day — from the unwanted involuntary movements induced by the morning flood of dopamine-mimicking drugs to increased rigidity later in the day as the drugs wear off. The medication also comes with other side effects that vary from person to person. For Krehbiel, it was nausea so severe he had to lie down multiple times a day.
When these become overwhelming, a neurologist might recommend DBS. Implants typically dispense pulses of electrical energy deep into the brain 24 hours a day to regulate aberrant brain signals associated with uncontrollable body movements. Before this year, this always-on approach was the only commercially available form of the therapy.
But continuous DBS can sometimes amplify the drugs’ effects — or generate new symptoms. Some of these are harmless: one man with OCD developed a passion for the music of Johnny Cash when his stimulator was turned on, but was uninterested in the artist when it was off. Other symptoms are cause for more concern, including sudden-onset gambling disorders and other temporary changes in impulse control. More frequently, the addition of stimulation can induce speech impairments, such as slurring, raise the risk of falling and cause some involuntary movements.
A clinician can try to balance the system by adjusting the intensity of stimulation, but there are limits to how precisely it can be calibrated.
Krehbiel was aware of these shortcomings, and it contributed to his sense that DBS was a treatment of last resort. In the late 2010s, his neurologist hinted that big improvements were on the way. “So I thought, wait for the new technology.”
Catching waves
Brain-wave activity is different in people with and without Parkinson’s disease. In people with Parkinson’s, there are noticeable differences at one range of frequencies, known as β-oscillations (between about 13 and 30 hertz), in a region deep in the brain called the basal ganglia. This region processes sensorimotor, cognitive and mood information.
β-oscillations have become an important marker of motor state. In the early 2000s, researchers at University College London found increasing evidence that people with Parkinson’s have intense bursts of activity in this range. When drug treatments are working, these bursts are less exaggerated. The same is true for DBS. The more that stimulation normalizes β-oscillations, the better the relief of some symptoms, says Bronte-Stewart.
Aberrant oscillations were dubbed oscillopathies, and in the 2000s, Medtronic started focusing on developing a device that could both read and correct these rhythms, says Tim Denison, a biomedical engineer at the University of Oxford, UK, who was working at the company at the time. “Just like you can build a radio to tune in to an audio channel, can we build a circuit that will tune in to these oscillopathies and help to guide how to adjust the stimulator?” he asks.
By 2006, Denison and his colleagues had built a ‘brain radio’, a sensing chip that could tune in to the different frequency bands in which the electrode sits. The next challenge was finding out how changes in particular bands correspond to specific movement problems. That was “a huge part of the first eight to ten years of the research with the investigational hardware”, says Bronte-Stewart. She and other researchers, including Philip Starr at the University of California, San Francisco, used a succession of new prototype devices to map these oscillopathies and adjust to them.
For example, when β-oscillation intensity begins to dip after a dose of medication, aDBS automatically reduces stimulation, keeping β-power in a healthy range. As the medication wears off, it does the opposite (see ‘Fine-tuned stimulation’). In 2019, Bronte-Stewart developed one of the algorithms that would underpin aDBS. When she tested it on 13 people with Parkinson’s, it improved the halting movements, called bradykinesia, that are associated with the disease. It also helped to reduce the inability to take steps, known as freezing of gait, in a study last year. In a separate study, Starr found that aDBS shortened the duration of volunteers’ most bothersome motor disturbances, but without aggravating side effects.
Other studies have suggested that aDBS reduces speech problems, such as slurring, that were another possible side effect of continuous DBS. “It only suppressed the pathological brain activity,” without suppressing normal speech, says Beudel.
Since 2013, small trials such as these have demonstrated these effects in around 400 people, estimates Robert Raike, the director of neuromodulation research and technology at Medtronic. What was missing was a way to validate these findings in real-world settings: in people’s homes and workplaces over a long period of time. Researchers needed a big trial.
A personalized device
Any Medtronic DBS implant manufactured after 2020 has the ability to be switched into adaptive stimulation mode. If people were enrolled in a clinical trial after 2020, their implant’s experimental capabilities could be activated by a firmware update, “a software unlock, like your iPhone”, says Raike. This approach opened up a large pool of possible trial participants. And the capability could be turned back off at the end of the trial.
After two months with continuous DBS, Bronte-Stewart unlocked Krehbiel’s device. It continued to keep his tremor at bay. He needed fewer drugs.
Other trial volunteers have reported similar improvements, along with a reduction of symptoms associated with continuous stimulation. Although she is not permitted to discuss the results, which are still pending publication, Bronte-Stewart points to data presented at a 2024 conference. Of 45 volunteers in the trial who were given the choice to revert to the continuous DBS or to retain the new adaptive functionality for a further long-term follow-up, 44 chose to stay on aDBS, Krehbiel among them. “I would not have considered for more than 30 seconds reverting back,” he says. “I was feeling good and didn’t much care why.”
Beudel saw a broadly similar trend among his participants. “It’s no secret that the results were positive,” he says. “We now see patients from all over the country coming to our centre saying that they want the aDBS.”
Since the new system was approved earlier this year, the upgrade has been available to anyone with one of the post-2020 devices. Beyond straightforward symptom relief, these users might see beneficial effects that go beyond controlling motor symptoms.
For example, Parkinson’s disease notoriously interferes with sleep, and as the medicine wears off at night, problems emerge that range from insomnia to hallucinations. Sleep deprivation, in turn, worsens the symptoms. “It is a vicious circle,” says Beudel.
Adaptive DBS could reduce sleep disturbances by automatically adjusting to the sleep-induced changes in β-oscillations.
And better sleep might, in turn, protect the brain. If it does, says Denison, aDBS could shed light on the tantalizing but controversial hypothesis that DBS protects the brain when implanted earlier in Parkinson’s disease progression.
Beyond Parkinson’s
It’s not just people with Parkinson’s who might benefit from the new therapy. Slightly more than one-quarter of the estimated 230,000 people with a DBS implant use it to manage other diseases, including dystonia — a movement disorder that causes muscles to contract — essential tremor and OCD.
Researchers are working to identify the associated oscillopathies to allow them to extend aDBS to these populations, and to those whose conditions have not yet been approved for any form of DBS, including Tourette’s syndrome. Beudel is exploring oscillopathies that precede the onset of tremor. Michael Okun, a neuroscientist at the University of Florida in Gainesville, has identified oscillations that could be quashed to dial down tics in Tourette’s syndrome.
And although the evidence for psychiatric conditions is not yet as strong, says Starr, “they might have oscillopathies that could be DBS targets.” OCD is especially promising, says Damiaan Denys, a psychiatrist at the University of Amsterdam. In a soon-to-be-published study, he and his team found clear associations between compulsions and specific brain signatures. “We are near to finding some of these neurological imprints,” he says.
Adaptive DBS has also raised hopes for treatment-resistant depression, which would represent a much larger market than Parkinson’s disease. Although DBS is not approved for this condition anywhere in the world, a few hundred people have had experimental implants.
Helen Mayberg, a neurologist at Icahn School of Medicine at Mount Sinai in New York City, ran two of the largest trials for depression so far. Both failed to meet their primary endpoint. Mayberg got one of Medtronic’s early prototype brain radios to examine the potential role of oscillations in depression. The condition is complicated. Like Parkinson’s, it has many symptoms, but unlike Parkinson’s, no specific abnormal oscillations have been associated with it. “If you round up ten people with depression,” says Alik Widge, a psychiatrist and biomedical engineer at the University of Minnesota, “you would not see the same oscillopathy.”
But Mayberg might be zeroing in on a signal that is associated with individuals getting better: this brain-oscillation pattern emerges as symptoms recede. A month after the brain signal disappeared in one participant with depression, they relapsed.
This work is in its early days, but Mayberg thinks the sensing capabilities of modern devices might one day provide a ‘check engine’ warning light for possible relapse.
Overcomplicating the problem?
As Medtronic and other companies continue to enhance their DBS systems, the number of electrodes and their sophistication has risen swiftly. Some researchers, including Denison and Bronte-Stewart, say that the line is blurring between DBS and brain–computer interfaces.
The enhanced sophistication puts pressure on the clinicians in charge of managing the complex device settings. “Who is going to program these?” asks Okun. He worries that the proliferation of smarter devices could make them, paradoxically, less accessible to people owing to the already-huge demands on clinicians’ time.
Medtronic is working on more-automatic programming to save time: one new technology, approved this year, streamlines the process by which the clinician programs the device. Artificial intelligence might refine settings further; the US Food and Drug Administration is developing new standards for automation. Any further advances, however, will require more large studies, such as ADAPT-PD and the smaller studies that led up to it. These are not cheap. “Every patient can cost a million bucks or more,” says Okun. But the future of a major funding source for DBS is in doubt. Even before the new US administration began slashing funds for medical research, last year Congress made a 40% cut to the BRAIN Initiative, the US National Institutes of Health’s neurotechnology innovation accelerator that has provided instrumental support for DBS research. “I think we’re all worried about where the funding will come from to develop these types of studies,” says Bronte-Stewart, although even now, other countries are taking up the slack.
All the while, the number of people with Parkinson’s is expected to nearly double globally by 2050: to 25 million.
To meet that need, the goal is to make the process as accessible as it was for Krehbiel. “I had the opportunity to get the secret sauce,” he says, “so why not go for it?”
This article is reproduced with permission and was first published on July 15, 2025.