Depression symptoms are associated with affective neural processing during sleep and rest

Abstract Sleep supports offline information processing and is essential for cognitive and emotional functioning. Abnormal sleep patterns are a hallmark of affective disorders. We hypothesized that affective symptoms occur with maladaptive neural processing during offline periods. To test this idea, we used multivariate EEG decoding with cross-state classification. A model trained on EEG data acquired while participants ( N = 52) viewed emotional images was used to classify stimulus valence. Applying this model to data collected during a nap revealed the re-emergence of affective neural patterns. Critically, offline reinstatement of patterns reflecting negatively valenced processing predicted greater depressive symptoms across participants. These associations reflected both cognitive-affective and somatic-performance depression subscales, and they generalized across sleep stages and wakeful rest. The finding that offline neural information processing is linked with emotional well-being supports a model whereby maladaptive negative biases can be perpetuated during rest, potentially shaping the progression of affective disorders.