Cognitive remediation in people with bipolar disorder and cognitive dysfunction: a systematic review

Cognitive impairment is a frequent feature of bipolar disorder (BD), often persisting during euthymic phases and affecting multiple cognitive domains. These deficits impair functional recovery and quality of life. Pharmacological treatments have shown limited efficacy in ameliorating cognitive dysfunction in BD. Cognitive remediation (CR) has emerged as a promising nonpharmacological alternative. However, reviews of its efficacy in BD have reported inconsistent findings. One major limitation highlighted in past reviews is the heterogeneity of study samples, due to the inclusion of participants without cognitive impairment, potentially introducing a ceiling effect that may obscure treatment benefits. This review addresses this gap by examining only studies that selected participants through objective or structured subjective cognitive screening.

A systematic search of PubMed and Web of Science (finalized 27 January 2026) identified interventional studies of CR in adults with BD requiring objective or subjective cognitive impairment for inclusion. Of 2662 records retrieved, 128 full texts were reviewed, and 6 studies met eligibility criteria; 2 secondary analyses derived from 2 of these studies were also retained for synthesis. Four studies were randomized controlled trials, one was non-randomized with a control group, and one was uncontrolled. Two reviewers independently screened studies, extracted data, and assessed risk of bias. Given methodological variability, results were synthesized narratively.

Six studies met inclusion criteria. Findings were inconsistent: three reported significant cognitive improvements following CR or related interventions, whereas three randomized trials found no significant cognitive effects. Variability in screening methods, intervention formats, and outcome selection precluded meta-analysis.

Evidence is constrained by the small number of eligible studies, methodological heterogeneity, and variable risk of bias.

Current evidence from enriched samples suggests that CR may improve specific cognitive domains in BD, but overall results remain inconclusive. Future trials should employ rigorous cognitive screening, harmonized outcome measures, and more individualized interventions to clarify efficacy. PROSPERO registration ID: CRD42025642423.