Impaired visuospatial working memory but preserved attentional control in bipolar disorder

Catherine Barnes, Lara Rösler, Carmen Schiweck, Benjamin Peters, Silke Matura, Jutta S. Mayer, Sarah Kittel‐Schneider, Michael Schaum, Andreas Reif, Michael Wibral, Robert A. BittnerPsychological 2025 · DOI

TL;DR

While people with bipolar disorder (PBD) consistently show deficits in visuospatial working memory, this study reveals that their "attentional hardware" remains remarkably intact. The research suggests that the cognitive bottleneck in PBD isn't an inability to filter out distractions or follow cues, but rather a failure in later stages of memory processing, such as consolidation.

What Problem Does This Paper Address?

Cognitive impairment is a hallmark of bipolar disorder, persisting even during euthymic phases (periods of stable mood). Working memory (WM)—the ability to hold and manipulate information over short periods—is one of the most severely affected domains.

However, researchers have long debated why this deficit exists. Is it a "storage" problem, or is it an "entry" problem? If you can’t effectively use selective attention to filter out irrelevant distractors (bottom-up attention) or follow goal-directed cues (top-down attention), your working memory will naturally be overloaded with "trash" data. This study investigates whether impaired attentional control is the root cause of WM deficits in PBD, or if the two processes are independent.

Methods

The researchers employed a visuospatial change-detection task with 139 participants (63 with PBD, 76 healthy controls).

The task used "Gabor patches" (striped circles) and manipulated two types of attention:

  1. Top-Down Attention: Participants saw a cue before the memory array. A "predictive" cue told them exactly where the targets would appear, while a "non-predictive" cue provided no info.
  2. Bottom-Up Attention (Salience): Some patches flickered (high salience), while others remained static. Sometimes the flickering patches were the targets (helping attention); other times they were distractors (hindering attention).

To quantify performance, the team used Cowan’s K, a standard formula that estimates the number of items successfully stored in working memory. They analyzed the data using linear mixed-models (LMM) to account for variables like age and clinical symptoms.

Datasets

  • PBD Group: 63 euthymic patients recruited from outpatient facilities in Frankfurt, Germany.
  • Control Group (HCS): 76 healthy individuals.
  • Clinical Metrics: The researchers used the Young Mania Rating Scale (YMRS) and the Montgomery-Åsberg Depression Rating Scale (MADRS) to ensure all PBD participants were in a stable mood state during testing.

Key Insights and Ideas

  • The "Island of Preservation": The most striking finding is that despite a clear deficit in total memory capacity, PBD patients utilized attentional cues and handled distracting salience just as well as healthy controls. Their ability to "prioritize" information is preserved.
  • Divergence from Schizophrenia: Previous studies using this exact paradigm in patients with schizophrenia showed significant deficits in attentional control. This suggests that while PBD and schizophrenia share some cognitive overlaps, the underlying "broken" mechanism in their working memory is likely different.
  • Consolidation is the Likely Culprit: Because the "selection" phase (attention) appears healthy, the authors argue that the impairment must happen later in the pipeline—specifically during WM consolidation, the process of transforming a fleeting sensory trace into a stable mental representation.

Key Results

  • General WM Deficit: Across all conditions, the PBD group stored significantly less information. On average, controls stored ~1.25 items, while the PBD group stored ~1.05 items ($p = 0.005$).
  • Intact Attentional Control: There were no significant group-by-salience or group-by-cue interactions. This means that when given a predictive cue, PBD patients improved their performance by the same margin as healthy controls. Similarly, they weren't disproportionately "captured" by flickering distractors.
  • Age as a Factor: Age was a significant predictor of lower WM capacity across both groups, but the PBD deficit remained significant even after controlling for age.

Limitations and Future Work

  • Indirect Evidence for Consolidation: The study concludes that consolidation is likely impaired by a process of elimination (since attention was fine). Future studies should use paradigms that directly manipulate the time allowed for consolidation to confirm this.
  • Sample Characteristics: The PBD participants had a relatively high mean IQ and were all euthymic. Results might differ in patients with more severe cognitive impairment or those currently experiencing mood episodes.
  • Task Specificity: The study focused on visuospatial attention. It remains to be seen if these "islands of preservation" extend to other forms of attention, such as social or emotional attention.

By pinpointing that the "filter" is working but the "storage" is leaky, this research helps narrow the target for future pro-cognitive treatments and neurostimulation therapies in bipolar disorder.

Summary updated

Abstract

While WM was disturbed in PBD, attentional prioritization of salient targets and distractors, as well as the utilization of external top-down cues, were not compromised. Thus, the control of attentional selection appears to be intact at least for our specific manipulation of this cognitive construct. These findings provide valuable clues for models of WM dysfunction in PBD by suggesting that later stages of WM encoding, such as WM consolidation, are likely primarily impaired, while selective attention is not a main source of impairment.
BibTeX
@inproceedings{barnes2025impaired,
  author = {Catherine Barnes and Lara Rösler and Carmen Schiweck and Benjamin Peters and Silke Matura and Jutta S. Mayer and Sarah Kittel‐Schneider and Michael Schaum and Andreas Reif and Michael Wibral and Robert A. Bittner},
  title = {{Impaired visuospatial working memory but preserved attentional control in bipolar disorder}},
  booktitle = {Psychological},
  year = {2025},
  doi = {10.1017/s0033291725102328},
}
Figure