The abstract is read at the publisher; the summary is JClub's.
In the Marburg-Münster Affective Disorders Cohort Study, predicting 2-year relapse in 564 patients with major depressive disorder achieved only moderate balanced accuracy (up to 64.43%).
- Why it matters: Early identification of high-risk patients could enable personalized relapse prevention, but existing models may not capture multimodal contributors or yield predictions usable in clinical practice.
- What they did: A naturalistic longitudinal prognostic pipeline integrated 9 baseline data modalities—clinical/psychometric, 4 MRI modalities, genetics, neurocognition, blood markers, and gut microbiome—to predict relapse versus no relapse over about 2 years.
- The result: Balanced accuracy ranged from 47.21% to 64.43%, and the best model relied only on few established clinical relapse risk factors, with multimodal integration not improving performance.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.