A review on the impact of cognitive dysfunction on social, occupational, and general functional outcomes in bipolar disorder

Abstract

Objectives: Bipolar disorder (BD) is associated with significant impairment in cognitive performance across multiple domains of function that often persist after clinical recovery. It remains unclear, however, as to whether this process is related to the clinical status of BD being depressed, manic/hypomanic, or euthymic. In this review, we examine the literature on the cross-sectional and longitudinal relationships between cognitive function and general function depending on the clinical phase of BD. Methods: A systematic review of original research that studied both cognitive function and general function in adults (18-60 years), restricted to BD, was conducted in a total of 18 studies meeting inclusion/exclusion criteria. Results: Results show cross-sectional and prospective relationships between cognitive function and general function in patients with BD in both symptomatic and euthymic patients with BD. While studies using general measures of function (e.g., Global Assessment of Function scale) show more inconsistent associations with cognitive function, those employing assessments of domain specific function, suggest a consistent relationship between social and occupational function and cognitive performance. Executive function is commonly affected by cognitive deficits in these patients, but in addition a variety of domains show associations with functional outcomes (e.g., social function, occupational function). Notably, the emerging evidence suggests that cognitive function may be a better predictor of future general function than affective symptom severity. Conclusions: Despite some inconsistencies, in sum the literature on the relationship between cognitive function and general function in BD implicates both cross-sectional and longitudinal associations, both in symptomatic and euthymic patients with BD. And in terms of capturing these changes functional scales in particular domain-specific measures seem superior to general measures.