Memory advantage for positive stimuli is compromised in depression

Andrea Cataldo, PhD

McLean Hospital
Memory advantage for positive stimuli is compromised in depression

Scientific Abstract

Background: Healthy adults show better memory for low-arousing positive versus negative stimuli, but depression compromises this positive memory advantage (Burt et al, 1995; White et al, 2009). Existing studies are limited by small samples, under-controlled designs, or analyses that provide limited insight into psychological mechanisms. Our study addresses these concerns by using diffusion modelling to identify precise mechanisms underlying the positive memory advantage and its disruption in depression.

Methods: A total of 1,358 participants completed the BDI-II (Beck, Steer, Ball, & Ranieri, 1996) and an emotional memory task. At study, participants judged whether positive and negative words either (1) were positive or (2) were self-descriptive. At test, participants viewed an equal mix of studied and unstudied words, and judged whether each one was “old” or “new”; if judged “old”, they indicated the study source ( “Positive?” or “Describes?” question).

Results: We replicate the positive memory advantage and its decrease in depression. Bayesian mixed effects logistic regression models of d’ scores suggest better recognition memory (95% HDI: 0.89:1.90) and source accuracy (95% HDI: 0.17:0.33) for positive words. As depressive symptoms increase, however, these advantages decrease. Specifically, applying the Hierarchical Drift Diffusion Model (HDDM; Wiecki et al, 2013) to these data indicated that higher BDI scores are associated with more efficient evidence accumulation for negative words in both the recognition and source accuracy tasks. By contrast, evidence accumulation for positive words is unaffected by BDI during the recognition task, and becomes less efficient with increased BDI during the source accuracy task.

Conclusion: Using a well-controlled design with a large sample, we find that depression reduces the positive memory advantage. HDDM analyses suggest that this reflects differential effects of depression on evidence accumulation for positive and negative material.

Live Zoom Session – April 21st

research Areas


Andrea M. Cataldo, PhD, Luke Scheuer, BS, Laura Germine, PhD, Daniel G. Dillon, PhD