Summary: Researchers have used brain imaging and machine learning to identify six distinct biological subtypes of depression and anxiety, called biotypes, and linked some of these subtypes to treatments more likely to succeed. This advance in precision psychiatry aims to reduce the current trial-and-error approach to prescribing treatments and improve the odds that patients receive effective therapy sooner.
The study proposes a future in which a brief brain scan could help determine the most promising treatment—medication or psychotherapy—based on each person’s brain activity patterns. This personalized approach could shorten the time to remission and reduce the suffering caused by repeated unsuccessful treatments.
Key Facts:
- Six Depression Biotypes: Researchers identified six distinct brain-activity patterns using fMRI and machine learning; each biotype shows different clinical features and treatment responses.
- Precision Psychiatry: Matching treatments to a patient’s brain signature improves the potential for successful outcomes and represents a move toward personalized mental health care.
- Study Publication: The findings, led by Stanford Medicine investigators, are reported in Nature Medicine and underline the value of biologically informed treatment decisions.
Source: Stanford
In the near future, a clinical depression screening could include a quick brain scan that helps choose the optimal treatment.
A research team at Stanford Medicine reports that combining functional brain imaging with machine learning can reveal distinct subtypes of depression and anxiety. Their analysis, published June 17 in Nature Medicine, classified patients into six “biotypes” and showed that three of them have predictable differences in response to common treatments.
Leanne Williams, PhD, director of Stanford Medicine’s Center for Precision Mental Health and Wellness and the study’s senior author, emphasizes the urgent need to improve how treatments are selected. Current practice often relies on trial and error, which can take months or years and sometimes never leads to full recovery. For roughly 30% of people with depression, multiple treatments fail to bring relief, and up to two-thirds do not reach a healthy level of symptom remission.
Biotypes predict treatment response
To uncover biological differences among patients, the team evaluated 801 people diagnosed with depression or anxiety using functional MRI (fMRI). Participants underwent scans both at rest and while completing tasks designed to probe cognitive and emotional processing. The researchers focused on brain regions and connections already implicated in depression—networks such as the default mode, salience, and frontoparietal attention circuits.
Applying cluster analysis, a form of machine learning, the investigators identified six reproducible patterns of brain activity. Of the 801 participants, 250 were then randomized to receive one of three commonly prescribed antidepressants or behavioral talk therapy, allowing the team to compare treatment responses across biotypes.
Results showed meaningful treatment differences by biotype. One biotype, marked by overactivity in cognitive brain regions, responded particularly well to the antidepressant venlafaxine (Effexor). Another biotype—characterized by higher resting activity across three regions linked to depression and problem solving—showed greater benefit from behavioral talk therapy. Conversely, people with a biotype defined by lower resting activity in attention-related circuits were less likely to improve with talk therapy compared with other groups.
These treatment patterns align with what is known about the brain regions involved. For example, behavioral therapies that teach problem-solving and coping skills may be more effective when the underlying brain circuitry supports learning and engagement. In cases where attention-related circuits are underactive, medications that target brain activity may help patients engage more fully with psychotherapy.
Williams described the findings as evidence that depression is not a single brain disorder but rather a set of different disruptions in brain function. The study provides an objective, brain-based method to inform treatment choices—an important step toward personalized mental health care.
In prior work, the team showed that using fMRI to identify a “cognitive biotype” improved prediction of who would remit with standard antidepressant treatment—raising predictive accuracy from about 36% without imaging to about 63% with imaging. The researchers are now testing new treatments tailored to that cognitive biotype and expanding imaging studies to include more participants and a broader range of interventions, including medications not traditionally used for depression.
Further explorations of depression
The six biotypes also corresponded to differences in symptoms and task performance. For instance, the group with overactive cognitive regions showed higher anhedonia (reduced ability to feel pleasure) and poorer performance on executive function tests. Another biotype that responded well to talk therapy tended to perform well on some cognitive tasks despite making errors on executive tests. One biotype did not show notable differences in the scanned regions compared with people without depression, suggesting there may be biological changes outside the regions examined or subtler dysfunction that current imaging did not capture.
Stanford investigators, including clinician-researchers such as Laura Hack, MD, PhD, are beginning to apply the imaging method in clinical practice under experimental protocols and are working to develop practical standards so other psychiatrists can adopt these tools. The goal is to get patients onto treatments most likely to help them as quickly as possible by using validated brain-function signatures.
Collaborators on the study include researchers from Columbia University, Yale School of Medicine, UCLA, UC San Francisco, the University of Sydney, MD Anderson, and the University of Illinois Chicago. The datasets used were funded in part by the National Institutes of Health and by Brain Resource Ltd.
About this depression research news
Author: Lisa Kim
Source: Stanford
Contact: Lisa Kim – Stanford
Image: The image is credited to Neuroscience News
Original Research: Open access. “Personalized brain circuit scores identify clinically distinct biotypes in depression and anxiety” by Leanne Williams et al., Nature Medicine.
Abstract
Personalized brain circuit scores identify clinically distinct biotypes in depression and anxiety
There is an urgent need for quantitative measures that capture coherent neurobiological dysfunctions—or biotypes—to better stratify patients with depression and anxiety. The investigators used both resting-state and task-evoked fMRI data collected across multiple studies from 801 patients when treatment-free, and from 250 patients after randomization to pharmacotherapy or behavioral therapy.
From these data, the team derived personalized, interpretable scores of brain circuit dysfunction grounded in a theoretical taxonomy. Participants clustered into six biotypes defined by distinct profiles of intrinsic functional connectivity within the default mode, salience, and frontoparietal attention circuits, and by activation and connectivity in frontal and subcortical regions during emotional and cognitive tasks.
The six biotypes were consistent with the theoretical framework and differed in symptom patterns, behavioral test performance, and response to pharmacological and behavioral treatments. These results offer a theory-driven, clinically validated, and interpretable method to parse the biological heterogeneity of depression and anxiety and represent a promising approach to advance precision care in psychiatry.