· Translated from the Spanish original

Mental health

University of Sydney finds a circadian subtype of depression in nearly 25% of 10,000 patients

The October 20 study in The British Journal of Psychiatry links that group to a weaker response to SSRIs. It is a genetic analysis, not a clinical trial.

What happened

  • Researchers at the University of Sydney described a subtype of depression tied to the 24-hour body clock, which they call “circadian depression.” The university release is dated October 20, 2026.
  • They analyzed data from more than 10,000 Australians with severe, recurrent or persistent depression, from the Australian Genetics of Depression Study.
  • They classified 2,604 participants, nearly 25%, in the circadian subtype.
  • The paper, published in The British Journal of Psychiatry, is titled “Clinical and genetic correlates of a circadian subtype of depression.” The lead author is Emiliana Tonini; Professor Ian Hickie is listed as co-senior author.
  • According to the university, this group responds less to antidepressants that act on serotonin (SSRIs) and more to melatonin, to antidepressants aimed at it and to light therapy.

Why it matters

  • A quarter of severe cases would be a group with different biology. If confirmed, the “try one drug after another” prescription would get a shortcut: look first at sleep and the clock.
  • For health systems like Chile’s, where access to specialists is slow, a simple sleep-wake rhythm marker would be a triage tool in primary care.
  • The consequence the release does not state: the phrase “responds more to melatonin” comes from correlations in an observational, genetic study. It is not the result of assigning treatments at random.

The number

2,604 of more than 10,000. Participants classified with circadian depression, nearly 25% of the sample.

Context

The sample is people with severe or persistent depression, a group that does not improve easily. The authors do not propose changing clinical guidelines. On mental health and technology, see the study on AI companions and well-being and model bias in pediatric triage.

What’s next

  • No clinical trials announced to test treatments aimed at the subtype.
  • Pending: replicating the finding in populations outside Australia.
  • No announced dates. This note is not medical advice.

Bottom line

Data from large genetic cohorts begin to split diagnoses that looked like one. It is the same movement already seen in Illumina’s multiomic genomics, which looks for subtypes where there used to be a label.

Sources

Edited by Rodrigo Cornejo. How we select and verify: who writes these notes.

Related notes

← All notes