Cognitive correlates of Antisaccade Task performance in bipolar disorder

The Antisaccade Task is often used to measure inhibitory control. While pooled effects in meta-analyses demonstrate antisaccade impairments in bipolar disorder (BD), individual studies show considerable variability in effect sizes and case-control significance, particularly in euthymic samples. This variability may reflect cognitive heterogeneity across BD samples. In this exploratory study we investigated whether antisaccade performance correlated with other cognitive domains in BD, and whether impairments were confined to specific cognitive subgroups.

152 individuals (44 BD outpatients and 108 healthy controls [HC]) completed the Antisaccade Task. Participants also completed traditional cognitive measures assessing working memory, immediate and sustained attention, processing speed, cognitive flexibility, and inhibitory control. Correlations between antisaccade performance (error rate, latency) and these cognitive measures were examined. BD participants were also stratified into cognitively equivalent versus cognitively reduced subgroups based on their cognitive performance relative to HCs, and antisaccade performance compared between these subgroups and HCs.

The BD group did not differ significantly from HCs on antisaccade error rate or latency. Antisaccade latency in BD correlated with working memory, while error rate correlated with immediate attention and processing speed. When stratified by correlated cognitive measures, cognitively reduced BD showed significantly worse performance than cognitively equivalent BD and HCs. These group differences were attenuated when BD was stratified by global cognitive performance.

Our findings suggest that antisaccade task underperformance in BD may be confined to individuals with reductions in other specific cognitive domains; with error rate associated with immediate attention and processing speed, and latency with working memory.