The researchers compared each participant's baseline scan with a scan taken one or two weeks after treatment began. Machine learning models then searched for three things: brain-wide changes shared across all groups, changes tied to symptom improvement in the placebo group, and patterns specific to the two antidepressants.

A shared signature of treatment

One finding stood out as universal. Across all participants — regardless of whether they received a drug or a placebo, and regardless of whether their symptoms improved — connectivity increased within a system spanning the visual processing areas, the precuneus (a region toward the back of the brain) and the thalamus (a relay center deep in the brain).

In the placebo group, symptom improvement was linked to a different set of circuits: the striatum, a region involved in motivation and reward, along with networks that govern attention.

Effects specific to antidepressants

The antidepressants left their own distinct fingerprint, but only in a subset of patients. These drug-specific changes involved the amygdala (which processes emotionally significant information), the midcingulate and orbitofrontal cortices (areas of the brain's outer layer), and the cerebellum (a region known mainly for its role in movement but increasingly linked to other functions).

Notably, the patients who did not show these drug-specific changes could be identified using the placebo response prediction model. In other words, the brains of some antidepressant-treated patients behaved like placebo responders.

"These findings reveal generalizable antidepressant-induced early FC changes, parse these changes into constituent placebo and drug-specific effects, and offer mechanistic insights into antidepressant action, supporting the development of interactive treatment optimization for MDD," the authors wrote.

What it means — and what it doesn't yet

The results could help explain why patients respond so differently to available depression treatments. In the longer term, scans taken shortly after a patient starts a new medication might help clinicians decide whether to continue or switch treatment, shortening the trial-and-error process that many patients currently endure.

That prospect remains distant, however. The findings describe patterns observed in two cohorts of patients; they do not yet prove that these brain signatures can reliably guide individual treatment decisions in a clinic. The study examined only two SSRIs, and it remains unclear whether the same patterns apply to other classes of antidepressants. More research is needed before any of the identified signatures can inform clinical practice.

Still, the study marks a step toward decoding how antidepressants and placebos act on the brain in those critical first weeks — a window that researchers have largely been unable to see into until now.

The paper, "Early brain functional connectivity changes induced by antidepressants and placebo," appears in Nature Mental Health (DOI: 10.1038/s44220-026-00729-y).