Telling bipolar depression apart from unipolar depression during a depressive episode is notoriously tricky — but a new study suggests sleep patterns might hold the key. Researchers in Paris developed a non-invasive predictive model using sleep questionnaires, actigraphy (wrist-worn motion tracking), and overnight polysomnography (PSG) to differentiate the two conditions in 159 patients.

The findings revealed distinct sleep signatures for each disorder. Patients with unipolar depression reported worse sleep quality and more severe insomnia, while those with bipolar depression were less physically active overall and spent more time at rest. On PSG, bipolar patients showed more light (N2) sleep and less deep (N3) sleep compared to their unipolar counterparts.

By the Numbers:

• 0.926 — area under the curve for the final predictive model
• 88.2% sensitivity / 89.5% specificity — the model's diagnostic accuracy
• 49.4% of variance explained by six key variables: sleep quality score, insomnia severity, actigraphic sleep efficiency, rest time, activity level, and N2 sleep proportion
• 159 patients studied (43 bipolar, 116 unipolar) between 2022–2025

Why it matters: Misdiagnosing bipolar disorder as unipolar depression is common and can lead to harmful treatment decisions, like prescribing antidepressants without mood stabilizers. A reliable, non-invasive tool to distinguish the two could meaningfully improve patient outcomes.