Do deep brain stimulation for psychiatric disorders have enough long-term follow-up evidence?

What does the strongest long-term study show for depression?

The best long-term evidence for DBS in a psychiatric disorder comes from a study of 25 patients with treatment-resistant major depression who received DBS targeting the ventral anterior limb of the internal capsule (vALIC). They were followed naturalistically for 6 to 9 years (average 7.7 years). The response rate—meaning a 50% or greater drop in depression severity—held steady at about 40% from year 3 through the last follow-up, and was 44% at the final visit [1]. That means nearly half of these very ill patients maintained a substantial benefit for years.

Importantly, the study also measured quality of life using the WHOQOL-BREF scale. Scores on the general, physical, psychological, and environmental domains all improved during the initial DBS optimization period and then stayed stable over the long term. The social domain did not show significant change. The study also included a randomized crossover phase after the first year where patients were switched to sham (inactive) stimulation—during that phase, patients scored significantly higher on active DBS for psychological, social, and environmental quality of life, confirming the effect was truly due to the stimulation [1]. This is the only study in this set that combines a long follow-up with a randomized sham control, making it the strongest single piece of evidence.

How does the evidence stack up for other psychiatric disorders?

For conditions other than depression, the long-term data is much weaker. A 2025 review of DBS across a range of treatment-resistant psychiatric disorders—including obsessive-compulsive disorder (OCD), bipolar disorder, schizophrenia, addictions, Tourette's syndrome, and anorexia nervosa—found that while short-term response rates can be promising (e.g., about 48% response in depression and OCD, and over two-thirds of patients with Tourette's or aggression in autism), the literature is 'constrained by inconsistent study designs, a paucity of randomized controlled trials, heterogeneity in DBS targets and stimulation parameters, and limited long-term and quality-of-life outcomes' [7]. In plain terms: we don't have enough long-term data to be confident these benefits last.

For schizophrenia, one small study followed 8 patients with treatment-resistant schizophrenia for 3 years after DBS of the nucleus accumbens or subgenual cingulate gyrus. It found a trend toward improvement in positive and general symptoms—75% of the nucleus accumbens group improved on positive symptoms, compared to 50% in the subgenual cingulate group—but the difference was not statistically significant (p = 0.06) [3]. The sample is tiny, and the authors themselves say the results only 'showed a trend' and call for larger trials. So for schizophrenia, the long-term evidence is essentially preliminary.

For catatonia, a 5-year follow-up of psychiatric patients with and without catatonia found that those with catatonia had significantly worse outcomes—25.8% were treatment-resistant versus 5.9% in the non-catatonic group, and they had more hospital readmissions and lower functioning at the end of follow-up [4]. But this study did not involve DBS at all—it simply describes the natural history of catatonia. It underscores that catatonia is a severe condition that might warrant aggressive treatment, but it does not provide evidence for or against DBS specifically.

Why is long-term follow-up so important—and what can go wrong without it?

A 2023 commentary in The Lancet Global Health makes a sobering point using a blood-pressure trial: a lifestyle intervention that looked great at 1 year (a 5 mm Hg reduction in systolic blood pressure) actually reversed by 5 years, becoming a 5 mm Hg increase—meaning the intervention may have done more harm than good over the long term [2]. The authors note that only 15% of 34 completed randomized trials in that area had follow-up beyond 1 year, and they warn that 'this is letting patients down.' The same principle applies to DBS: early benefits can fade, and rare harms may only emerge after many years.

This is not just theoretical for DBS. A 10-year follow-up of DBS for Parkinson's disease—a neurological condition, not a psychiatric one, but the same technology—found that while motor improvements (especially for tremor and rigidity) were sustained, non-motor symptoms like emotion, cognition, and quality of life improved at 3 years but then returned to baseline or declined by 10 years [5]. The authors describe the first 3 years as a 'DBS honeymoon.' This pattern—early peak, then waning—could also apply to psychiatric DBS, but we simply don't have the data to know. The depression study [1] showed stable effects out to 7-9 years, but that's only one study in one condition.

A review of DBS for pediatric dystonia, with follow-up averaging 8.5 years and ranging up to 22 years, found that motor benefits were generally sustained, but the authors caution that 'data regarding long-term outcomes is somewhat limited, particularly with regards to neuropsychiatric outcomes and adverse events' [6]. Late adverse events tended to be hardware-related, but the point is that even in a field with decades of DBS use, long-term psychiatric data is scarce.

About These Sources

This answer is built on 7 peer-reviewed studies — published from 2022 to 2026, 3 from 2024 or later, 2 in Q1 journals, collectively cited 70 times — selected as the most relevant from 10 studies that passed quality screening, drawn from 70 papers retrieved from a database of over 500 million.

Sources used in this answer

Related research questions