Researchers identify five brain activity profiles in major depressive disorder

Overview

Updated 1 hour ago

Depression treatment is mostly trial and error: a patient tries an antidepressant, waits weeks, and often switches when nothing helps. A University of Helsinki team used magnetoencephalography (MEG), which records brain electrical activity with millisecond precision, to investigate why response varies so much. Scanning 263 people with major depressive disorder and 75 healthy controls, they sorted patients into five distinct brain-activity profiles.

The five profiles differ in functional connectivity, how well brain regions fire together. Two show hyperconnectivity, two show hypoconnectivity, and one looks like healthy controls. Each maps to a different symptom pattern: one tracks with substance abuse, one with post-traumatic stress, one with broad severe depression.

Published in Nature Mental Health, this is the first time oscillation-based connectivity has been used to define depression phenotypes. The team is clear about the gap to clinical use: "We're not yet at the point where brain measurements can be used to choose the right treatment for patients," said professor Satu Palva. "But the study does show one possible route."

Why it matters

One in three depression patients doesn't respond to the first antidepressant; brain-based subtypes could end the trial-and-error cycle.

Questions about this story

Free account needed to ask — your question is kept and asked for you right after sign-up. Answers are public.

No questions yet — be the first to ask.

Key Indicators

263

Patients with major depressive disorder scanned

Measured with magnetoencephalography at resting state.

Healthy control participants

Provided the baseline for comparing connectivity patterns.

Depression phenotypes identified

Two hyperconnectivity, two hypoconnectivity, one healthy-like.

Voices

Curated perspectives — historical figures and your fellow readers.

Ever wondered what historical figures would say about today's headlines?

People Involved

Satu Palva

Professor of neuroscience, University of Helsinki; senior author

Leading follow-up validation work

Organizations Involved

University of Helsinki

Research university

Conducted the study

Finland's largest university; the MEG data collection and analysis were based at its Neuroscience Center.

Nature Mental Health

Peer-reviewed journal

Published the study in September 2026

A Nature Research journal covering neuroscience and psychiatric research.

Timeline

January 2017 September 2026

3 events Latest: 2 days ago

September 2026
Nature Mental Health paper identifies five MEG-based depression phenotypes
Latest Publication

MEG scans of 263 patients yield five oscillation-connectivity profiles with distinct symptom patterns; first such use of oscillation-based connectivity.
2 days ago
January 2026
Molecular Psychiatry reports stable rTMS-responsive depression subtypes
Publication

Study of 1,204 patients finds two connectivity-based subtypes; only one shows anhedonia improvement with repetitive transcranial magnetic stimulation.
January 1st, 2026
January 2017
Drysdale et al. publish four fMRI-based depression biotypes
Publication

Columbia team defines four subtypes by resting-state fMRI connectivity, predicting transcranial magnetic stimulation response in a small sub-study.
January 1st, 2017

Scenarios

Five-phenotype structure replicated in an independent cohort

Likely Resolves by End of 2028

Discussed by: The study authors, who call the phenotypes 'putatively generalizable to a wider population' pending validation

A follow-up study collects MEG data from a separate, larger group of patients and asks whether clustering reproduces the same five profiles with the same spectral and spatial patterns. The original team reported the phenotypes were stable under repeated within-cohort subsampling, which supports the chance of replication. Success would strengthen the case that these are stable biological subtypes.

MEG phenotypes enter treatment-stratification trials

Possible Resolves by End of 2029

Discussed by: The researchers, noting 'near-term clinical value… in specialist settings'; precedent from the 2026 Molecular Psychiatry rTMS-subtype study

A clinical study in a specialist center uses MEG-based phenotypes to assign rTMS or antidepressant treatment and measures whether response differs by subtype. The team said near-term value could come where conventional clinical assessment is insufficient. The 2026 Molecular Psychiatry study already reported rTMS anhedonia response differing across connectivity-based subtypes, suggesting this path is viable.

Replication fails; phenotypes don't transfer across cohorts

Possible Resolves by End of 2028

Discussed by: Researchers citing the patchy replication record of fMRI-based biotypes such as Drysdale's 2017 subtypes

Depression biotypes found with one imaging method have historically struggled to replicate across sites with different equipment and samples. If a larger cohort fails to reproduce five clusters, or if subtypes collapse to fewer, the practical value shrinks. This scenario mirrors what happened to some fMRI-based subtype findings.

Historical Context

3 moments from history that rhyme with this story — and how they unfolded.

1980–1985

Dexamethasone suppression test (early 1980s)

In the early 1980s, the dexamethasone suppression test (DST) was widely promoted as the first biological marker for major depression — an endocrine test where patients failing to suppress cortisol after a dexamethasone dose were deemed depressed. Clinics adopted it enthusiastically.

Then

Large-scale studies found poor specificity: many patients without depression had abnormal results, and many with the disorder did not.

Now

By the late 1980s it was largely abandoned as a diagnostic test, a classic biomarker reversal.

Why this matters now

A cautionary tale for any new depression biomarker: promising lab findings can fail in broader clinical use, which is why the MEG phenotypes still need validation.

2001–2006

STAR*D trial (2001–2006)

The Sequenced Treatment Alternatives to Relieve Depression trial followed over 4,000 outpatients through successive steps of treatment. Roughly a third achieved remission with the first antidepressant; success rates dropped with each additional step.

Then

STAR*D quantified the trial-and-error reality of depression care.

Now

It became the standard reference for why better treatment matching matters.

Why this matters now

This is the clinical problem the five-phenotype study aims to solve — choosing the right treatment from the start.

2009

NIMH Research Domain Criteria launch (2009)

In 2009, the National Institute of Mental Health began pushing Research Domain Criteria (RDoC), a framework to classify mental illness by neurobiological dimensions — genes, circuits, behavior — instead of the DSM's symptom checklists. Then-director Thomas Insel argued that psychiatric diagnosis had to be grounded in biology.

Then

RDoC reshaped how psychiatric research is funded and framed, steering it toward biomarkers and circuit-level mechanisms.

Now

It has not yet translated into routine clinical diagnosis, underscoring how hard biological subclassification has been.

Why this matters now

The search for depression phenotypes is a direct extension of the RDoC premise: biological subtypes will only emerge once diagnosis is grounded in measurement.