IPM Take
Depression trials love totals. Patients live in symptoms.
Two people can receive the same major-depression diagnosis and the same overall severity score while one is dominated by anxiety and insomnia and another by mood loss, physical slowing and exhaustion. Yet treatment systems often flatten those differences into a single number and then wonder why response is unpredictable.
A new rTMS analysis identified four distinct symptom-response trajectories among people with treatment-resistant depression. By the end of treatment, remission ranged from 65.2% in the strongest-response group to zero in the minimal-response group. The message is not that clinicians can already predict exactly who will respond. It is that depression is too heterogeneous for the average score to remain the whole story. JAMA Network
Executive Summary
Researchers analysed data from 388 adults with treatment-resistant major depressive disorder who had participated in the THREE-D randomised trial of dorsolateral prefrontal repetitive transcranial magnetic stimulation.
Using longitudinal modelling of mood, anxiety, insomnia and somatic symptom clusters, the investigators identified four distinct treatment-response trajectories. Final remission differed dramatically between groups, from 65.22% in the optimal-response group to 0% in the minimal-response group. JAMA Network
Younger age, benzodiazepine use, lower baseline anxiety and higher baseline mood symptoms were among the characteristics associated with the minimal-response trajectory. The analysis is exploratory and retrospective and requires prospective validation before these factors can be used to assign treatment or deny access to rTMS. JAMA Network
Why it matters
- Treatment-resistant depression is expensive in every meaning of the word. Patients may spend years moving through medications, psychotherapy and specialist services before receiving neuromodulation. If response patterns can eventually be identified earlier, treatment could become less trial-and-error.
- But prediction also carries risk. A model designed to personalise care can become a gatekeeping mechanism if payers or services use imperfect probabilities to decide who deserves expensive treatment. The evidence must therefore improve access decisions, not simply automate exclusion.
Major depression is one diagnosis containing hundreds of possible symptom combinations. That fact is clinically familiar and operationally inconvenient, so health systems often ignore it.
Rating scales make the problem manageable. They convert insomnia, sadness, anxiety, appetite, physical symptoms and motivation into one total. That total is useful for measuring overall severity, but it can hide two patients whose experiences are fundamentally different behind the same number.
The new rTMS analysis asks what happens when those symptoms are allowed to move separately over treatment. Four trajectories emerged, and their outcomes were strikingly different. Some patients moved rapidly toward remission. Others barely responded. JAMA Network
The finding does not yet provide a clinical algorithm. The data came from an earlier trial, and the associations identified retrospectively may not survive prospective testing. In particular, benzodiazepine use or baseline symptom patterns should not be treated as reasons to withhold treatment on the basis of this study.
The more important implication is conceptual. Precision psychiatry may not require waiting for a perfect genetic test. Better personalisation could begin by taking symptom architecture more seriously: which symptoms dominate, which move first, which persist, and whether those patterns map onto different neural circuits and stimulation targets.
The average patient with depression does not exist. Treatment systems increasingly have the tools to acknowledge that. The harder challenge is redesigning care around the person who does.
