Photo by christopher lemercier on Unsplash
Why Chronic Depression Is So Hard to Treat, According to Brain Scans
In A Nutshell
- Brain scans of 46 depression patients found that chronic and non-chronic cases show opposite patterns in how two key brain networks communicate as symptoms worsen.
- In shorter-term depression, worsening symptoms were linked to weaker brain network communication; in chronic depression (episodes over two years), the relationship reversed.
- About 30% of people with depression develop a chronic form, which carries higher suicide risk and responds more poorly to standard antidepressants.
- Researchers say the findings are preliminary but suggest illness duration may need to be considered separately from symptom severity in future treatment research.
A new brain imaging study suggests that depression duration may change how symptom severity relates to brain network activity, a finding that could help explain why chronic cases so often resist standard care.
Most people think of depression as a single condition on a sliding scale: the worse you feel, the more disrupted your brain. A study published in Scientific Reports puts pressure on that assumption. Researchers found that one key brain-network relationship looked different in chronic versus non-chronic patients. In people with shorter depressive episodes, worsening symptoms were linked to weaker communication between two major brain networks. In people who had been depressed for more than two years, worsening symptoms were linked to stronger communication between those same networks. Two groups, same diagnosis, showing opposite patterns in this one brain-network measure.
That distinction carries real weight. Roughly 30% of people who have ever experienced depression will develop a chronic form, defined as a depressive episode lasting more than two years. Chronic depression is linked to a higher risk of suicide and a weaker response to antidepressant treatments. Yet most brain research has focused on how severe depression looks inside the brain, largely ignoring how long someone has been sick. This study set out to fill that gap.
Medication-Free Patients Reveal Opposite Brain Signals by Episode Length
Researchers recruited 46 adults diagnosed with major depression from a hospital in São Paulo, Brazil, using data collected as part of a larger clinical trial between October 2013 and July 2016. Of the 46 participants, 20 were classified as having chronic depression (episodes longer than 24 months) and 26 had non-chronic depression. All participants scored at least 17 on the Hamilton Depression Rating Scale, placing everyone in the moderate-to-severe range.
Patients who had been taking antidepressants underwent a washout period before brain scans were taken, allowing researchers to look at the brain’s own activity without the influence of medication.
Participants underwent two types of brain scans. One measured activity while participants rested quietly, tracking how different brain regions naturally communicate even when a person isn’t doing any particular task. The second measured brain structure, specifically the size and density of gray matter, the tissue made up of nerve cell bodies, in different regions.
Chronic Depression Flips the Brain’s Network Relationship With Severity
At the center of the results is the relationship between two large-scale brain networks: the Central Executive Network (CEN) and a part of the Default Mode Network (DMN) called the precuneus.
Put simply, the CEN is the brain’s project manager, kicking into gear when someone needs to focus, solve problems, or make decisions. Its counterpart, the DMN, is more like the brain’s idle mode, most active when the mind wanders, reflects on memories, or thinks about oneself or others. In a healthy brain, these two networks generally work in opposition: when one is busy, the other quiets down. In depression, that balance is frequently disrupted.
Among non-chronic patients, higher severity was associated with weaker communication between the CEN and the DMN’s precuneus region. The sicker they felt, the more those two systems fell out of sync. In patients with chronic depression, the relationship ran the other way. More severe symptoms were linked to stronger communication between those same two networks. The researchers describe this as chronicity “moderating” the impact of severity on brain connectivity.
Brain structure told a somewhat different story. Regardless of whether someone had chronic or non-chronic depression, more severe symptoms were linked to larger tissue volume in two regions involved in mood regulation and cognitive control, both of which overlap with the CEN. Chronicity alone showed no significant relationship with tissue size. That may seem backwards given that many studies have reported tissue loss in depression, but the researchers point out that their participants were medication-free at the time of scanning, and other studies of unmedicated patients have reported the same pattern.
Why the Duration of Depression May Matter for Future Research
One of the more clinically relevant questions the findings raise is whether chronic and non-chronic depression should be studied and, eventually, treated in more tailored ways. If the two groups show opposite patterns in brain network activity, it is worth asking whether treatments developed without separating patients by illness duration are well matched to the full range of patients who receive them.
Brain scans are not ready to guide care based on this study. But the results point to a question future research can test: whether patterns like these could help predict chronicity, recovery, or treatment response. The researchers themselves are careful to frame the findings as preliminary, and given the study’s size, that caution is warranted. With only 46 participants and a single-snapshot design, the data cannot establish whether the altered connectivity preceded chronic depression, followed from it, or reflects some combination of both.
What the study does offer is a clear signal that illness duration shapes the brain in ways that a severity score alone won’t capture. For the millions of people who cycle through treatments that never fully work, that may be one of the more important questions depression researchers have yet to answer.
Disclaimer: This article is based on a single exploratory study with a small sample size and a cross-sectional design. The findings are preliminary and do not establish cause and effect. They should not be used to make personal medical decisions. Readers with questions about depression or its treatment should consult a qualified healthcare provider.
Paper Notes
Limitations
The sample of 46 participants (20 chronic, 26 non-chronic) is relatively small, limiting generalizability and statistical power. Because all participants were required to score at least 17 on the Hamilton Depression Rating Scale, the findings may not apply to people with milder depression. The cross-sectional design means no causal conclusions can be drawn. Potential confounders including treatment resistance, recurrence, early onset, and benzodiazepine use were not fully controlled in all analyses. The study did not include a healthy control group, and the relatively short resting-state scan duration may affect reproducibility of connectivity results.
Funding and Disclosures
The study was supported by FAPESP grants (2012/20911-5), with additional funding from the NIHR Oxford Health Biomedical Research Centre (NIHR203316). Lead author Tamires Zanao received FAPESP grants 2020/03235-2 and 2022/09688-4. Pedro Henrique Rodrigues da Silva received FAPESP grants 22/03266-0 and 23/13893-5. Jacinta O’Shea is a Sir Henry Dale Fellow funded by the Royal Society and the Wellcome Trust (215451/Z/19/Z). The Oxford University Centre for Integrative Neuroimaging received Wellcome Trust support between 2017 and 2025. The ELECT-TDCS trial was funded by the São Paulo Research State Foundation and others (ClinicalTrials.gov NCT01894815). The authors declare no competing interests.
Publication Details
Authors: Tamires Zanao, Piergiorgio Salvan, Lais B. Razza, Pedro Henrique Rodrigues da Silva, Andre R. Brunoni, and Jacinta O’Shea | Affiliations: University of São Paulo Medical School; Oxford University Centre for Integrative Neuroimaging and Oxford Centre for Human Brain Activity, University of Oxford; Ghent University; Massachusetts General Hospital and Harvard Medical School; UT Southwestern Medical Center | Journal: Scientific Reports | Title: “Chronicity moderates the impact of severity on central executive-default mode network functional interactions in depression” | DOI: https://doi.org/10.1038/s41598-026-40364-2 | Published: 2026, Volume 16, Article 10116







