Deep brain stimulation: an established treatment and a research frontier

Deep brain stimulation (DBS) is a treatment in which a surgeon places a thin electrode in a small target area deep in the brain. A battery-powered device implanted in the body sends small electrical pulses through the electrode. DBS is an established treatment for some movement disorders, mainly Parkinson's disease, and it is being studied for other conditions, including depression. A newer version, adaptive DBS, listens to a brain signal and changes the stimulation when that signal changes. It matters because it could mean less stimulation time, and possibly better control for people whose symptoms come and go. The evidence for that is still small.

This article is general information, not medical advice.

What is established in routine use

DBS for Parkinson's disease is the best-established use. NHS Scotland's patient information describes it as a treatment for Parkinson's disease and explains how batteries are monitored and replaced, as set out below (NHS Scotland).

One important early study targeted the subthalamic nucleus. In 1998, Limousin and colleagues reported in the New England Journal of Medicine on 24 people with advanced Parkinson's disease who had electrodes placed on both sides of this area, 20 of whom were followed for at least 12 months. After a year, motor and daily-living scores off medication had improved by 60%, and the average dose of dopaminergic drugs had been halved. One person had paralysis and loss of speech after bleeding in the brain during the operation (Limousin et al., 1998).

What trials show

Trial sizes matter here, so each result is given with its conditions.

Adaptive DBS in Parkinson's disease

• Little and colleagues, 2013. Eight people with advanced Parkinson's disease were tested in a short, acute study in the days after their electrodes were implanted, while the leads were still brought out through the scalp. Stimulation was switched on only when a beta-band rhythm (13 to 35 Hz) rose above a set threshold. Blinded motor scores improved by 49.7% with adaptive stimulation compared with 30.5% with continuous stimulation, and stimulation was on for 44.2% of the time, a 56% reduction in stimulation time. The authors describe it as a proof-of-principle study (Little et al., 2013).
• Oehrn and colleagues, 2024. Four men with Parkinson's disease, six brain hemispheres in total, took part in a blinded, randomised crossover pilot trial. Each setting ran for at least a month at home, in randomised blocks of 2 to 7 days. Adaptive stimulation cut the share of awake time spent with each person's most bothersome symptom by 16.3 percentage points (p below 0.001). The opposite symptom was not significantly worse (p = 0.26). The authors say the results "provide the foundation for further, larger, clinical trials" (Oehrn et al., 2024).
• ADAPT-PD, 2025. It is the largest of the adaptive trials cited here, with 68 people with Parkinson's disease who had been stable on standard DBS. It was an open-label, non-randomised trial overall. People who tolerated both adaptive modes were randomised, without being told which was which, to 30 days in each of the two adaptive modes. The aim was to show that adaptive stimulation was comparable to standard stimulation, not better: success meant that good-quality "on" time (time with symptoms well controlled and without troublesome involuntary movements) fell by no more than a set amount compared with standard DBS. The authors revised this threshold after the fact, to no more than 2 hours a day, because the original analysis broke the assumptions of the statistical test, so the P values are nominal, as the authors state. Against the original threshold, 95% of people on the dual-threshold mode and 87% on the single-threshold mode met the goal; against the revised one, 91% and 79%. Of 52 people who entered setup, 45 (87%) tolerated at least one adaptive mode. Average adaptive exposure was 17.5 months. The authors report no unanticipated serious adverse device events. Medtronic funded the trial and was involved in its design, conduct and analysis (Bronte-Stewart et al., 2025).

Regulatory status of adaptive DBS

The US Food and Drug Administration (FDA) database lists a decision date of 20 February 2025 for a premarket approval supplement (P960009/S478) covering "a new optional programming feature called adaptive deep brain stimulation (aDBS)" on Medtronic's Activa, Percept and SenSight DBS system. The listed indication is Parkinson's disease: adjunctive therapy in people with levodopa-responsive Parkinson's disease of at least four years' duration whose symptoms are not adequately controlled with medication (FDA PMA database). In an announcement dated 3 March 2025, Medtronic said the approval covered its BrainSense Adaptive DBS feature and that it "quickly followed our CE Mark". Other statements in the announcement are the company's own claims and have not been independently verified (News-Medical, republishing Medtronic's announcement, 2025).

DBS for depression (still experimental)

• Holtzheimer and colleagues, 2017. In a multisite, sham-controlled trial, 90 people with treatment-resistant depression were randomised (60 active, 30 sham). Response, averaged over months 4 to 6 of the six-month double-blind phase, was 20% with active stimulation and 17% with sham. The difference was not statistically significant. The authors concluded that the study confirmed the safety and feasibility of the approach. 28 people had 40 serious adverse events, and eight of these events, in seven people, were judged related to the device or surgery. The abstract names Abbott (previously St Jude Medical) as funder (Holtzheimer et al., 2017).
• Scangos and colleagues, 2021. This was a single-patient case. A brain signal was used to trigger stimulation in one person with treatment-resistant depression, and the authors report "rapid and sustained improvement in depression". They add that future work is needed to see whether the approach applies more widely (Scangos et al., 2021).

Limits and open problems

Surgery. NHS Scotland's patient information gives these risks for the initial operation: stroke from bleeding below 0.5% (1 in 200), death from bleeding below 0.2% (1 in 500), infection below 5% (1 in 20), and seizure below 1%. Lead breakage and parts of the system eroding through the skin are also listed (NHS Scotland). In the Oehrn trial, one person had a hardware infection that needed surgical removal, followed by a new device (Oehrn et al., 2024).

Batteries and replacement. Non-rechargeable batteries often last around 3 to 5 years. Rechargeable versions are expected to last many years, which the page puts at over 15. Replacement is a minor procedure under general anaesthetic or sedation, and teams plan it by monitoring battery life. Battery failure is described as rare (NHS Scotland). Each replacement is a further procedure, so the choice of battery type matters to the person.

Small and sponsor-linked evidence. The adaptive trials have very few participants. All three adaptive studies cited here are in Parkinson's disease, and the largest was open-label, non-randomised against standard DBS, and manufacturer-funded. The depression evidence is weak: the sham-controlled trial was not positive, and the closed-loop report describes one person.

Needs a suitable signal. Adaptive control only works if a brain signal tracks symptoms in each person. In the Oehrn trial, the commonly used beta rhythm did not track symptoms at home in five of six hemispheres, and the team used a different signal chosen through an individual, multi-step process (Oehrn et al., 2024). In ADAPT-PD, 45 of 52 people tolerated adaptive programming (Bronte-Stewart et al., 2025).

Cost and access. The sources used here do not give figures for device or surgery costs, or for how easily people can reach a specialist centre.

Other conditions. DBS for depression and other conditions remains in research. Results from one or a few people should not be read as proof that a treatment works.

What to watch next

• How widely adaptive DBS is used in practice since the February 2025 FDA approval, and whether independent results follow.
• Larger adaptive DBS trials beyond ADAPT-PD and the four-person Oehrn study, including any that compare adaptive with standard stimulation under blinding.
• Whether the single-patient closed-loop depression result is tested in more people.
• Long-term data on battery life, hardware failure and the cost of replacement.