A meta-analysis of data-driven cognitive subgroups in bipolar disorder

Area of Science:

• Neuroscience
• Psychiatry
• Cognitive Psychology

Background:

• Bipolar disorder (BD) is a complex mood disorder with significant heterogeneity.
• Identifying distinct cognitive subgroups within BD may reveal biologically valid subtypes.
• Previous research suggests cognitive impairments are common in BD, but subgroupings require further investigation.

Purpose of the Study:

• To conduct a meta-analysis of studies using data-driven clustering to identify cognitive subgroups in bipolar disorder.
• To determine the prevalence and characteristics of these identified cognitive subgroups.
• To compare cognitive functioning across subgroups and with healthy controls.

Main Methods:

• Systematic search of PubMed, Scopus, and Web of Science for peer-reviewed studies on cognitive subgroups of BD.
• Inclusion of 14 cross-sectional studies with euthymic or mildly symptomatic BD patients.
• Random-effects meta-analysis performed using R software to pool data.

Main Results:

• Consistent support for a 3-cluster solution: severe-impairment (23.1%), moderate-impairment (42.5%), and good-functioning (33.5%).
• Severe and moderate impairment groups exhibited significant deficits across six cognitive domains compared to controls.
• The severe-impairment group showed greater functional impairment, hospital admissions, and higher rates of BD Type I and antipsychotic use.

Conclusions:

• Findings support the existence of three distinct cognitive subgroups in bipolar disorder.
• The severe-impairment and moderate-impairment subgroups are associated with a more severe illness course.
• These cognitive subgroups may represent clinically and biologically meaningful subtypes of BD.