Eye-movement tests show shared antisaccade deficits in bipolar disorder and depression

A three-level meta-analysis published in BMC Psychology has quantified how two simple eye-movement tasks perform as potential objective markers for distinguishing bipolar disorder from unipolar depression, two conditions whose depressive episodes can look nearly identical in the clinic. The systematic review, reported as accepted research shared ahead of final version-of-record edits, pooled 34 clinically diagnosed datasets covering 2,504 participants and found that antisaccade performance, but not prosaccade performance, separated both patient groups from healthy controls with moderate to large effects.

The antisaccade task requires participants to suppress the reflexive glance toward a sudden visual cue and instead look in the opposite direction, making it a probe of inhibitory control and rule-guided responding. In the meta-analysis, antisaccade accuracy was poorer than controls in bipolar disorder with Hedges’ g of −0.69 (95% CI −0.87 to −0.50) and in unipolar depression with g of −0.52 (95% CI −0.74 to −0.31). Antisaccade latency was prolonged in bipolar disorder (g = 0.85, 95% CI 0.52 to 1.17), while in unipolar depression the latency effect did not meet the conventional significance threshold (g = 0.36, 95% CI −0.0006 to 0.72, p = 0.0503).

Prosaccade trials, in which participants simply look toward the cue, showed smaller and less consistent effects. Bipolar disorder showed a modest prosaccade latency prolongation (g = 0.29, 95% CI 0.03 to 0.56), whereas prosaccade accuracy in bipolar disorder and both prosaccade outcomes in unipolar depression were non-significant. One moderator analysis found that stimulus type influenced prosaccade latency in bipolar disorder, with larger delays for face stimuli than for simple targets (p = 0.031), hinting that orienting to socially relevant images is more context-dependent than automatic responding to basic cues.

The authors searched PubMed, Scopus, Web of Science and PsycINFO from inception through 31 August 2026 under PRISMA 2020 guidance, synthesizing clinically diagnosed datasets with three-level random-effects models and study-level cluster-robust inference. To compare the two disorders indirectly, meta-regressions contrasted the bipolar-versus-control and depression-versus-control effect sets. Only antisaccade latency differed nominally between disorders (β = 0.46, 95% CI 0.06 to 0.86, p = 0.027), and that contrast lost statistical significance after excluding two studies that had enrolled both diagnostic groups (p = 0.067).

Robustness checks softened several results. Sensitivity analyses did not reverse the direction of any effect, but statistical significance varied for antisaccade latency in unipolar depression and for both prosaccade outcomes in bipolar disorder. Certainty in the evidence, evaluated with an adapted GRADE framework, was rated very low for all eight primary outcomes, a caution that applies across the entire set of pooled estimates.

The interpretation offered by the researchers is mechanistic rather than diagnostic. Poorer antisaccade accuracy in both disorders suggests a shared difficulty suppressing reflexive orienting and implementing rule-guided responses, while the larger bipolar latency effect may reflect additional slowing in response selection or preparation. Smaller, less consistent prosaccade effects suggest that relatively automatic orienting is either less affected or more dependent on task context. The authors emphasize that direct prospective comparisons using standardized tasks are required before any individual-level diagnostic or clinical application can be considered.

Subject of Research: Meta-analysis of antisaccade and prosaccade eye-movement performance in bipolar disorder and unipolar depression

Article Title: Antisaccade and prosaccade performance in bipolar disorder and unipolar depression: a three-level meta-analysis

Article References: Abolghasemi, M. M., Khodadad, A., Amiri, P., & Vahabie, A.-H. (2026). Antisaccade and prosaccade performance in bipolar disorder and unipolar depression: a three-level meta-analysis. BMC Psychology. https://doi.org/10.1186/s40359-026-05818-2

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05818-2

Keywords: bipolar disorder, unipolar depression, antisaccade, prosaccade, meta-analysis, eye-tracking, cognitive control, biomarkers, BMC Psychology, GRADE certainty, performance, bipolar