Objectives Although cognitive dysfunction is a prominent feature of bipolar disorder (BD), previous research presents limitations in estimating the proportion of euthymic patients experiencing clinically relevant deficits and identifying predictors of cognitive difficulties. We explored the relevance of recommended definitions of clinically significant cognitive impairment for functional outcomes, estimated its prevalence, and identified patient characteristics associated with cognition.

Methods We assessed cognitive performance across four domains in 80 euthymic participants with BD. Participants were categorized based on two criteria for clinically significant cognitive impairment and we assessed the ability of these criteria to differentiate participant performance on established functional outcomes. Variable selection with elastic net regression was used to identify sociodemographic and clinical factors associated with cognitive performance. Selected variables were examined as predictors of clinically significant cognitive impairment with logistic regression.

Results According to the selected criterion, 34% presented with clinically significant cognitive impairment. Poorer current cognitive performance was associated with older age, lower estimated premorbid IQ, more currently prescribed psychotropic medications, fewer previous psychological therapies, and current use of antipsychotics. A model with premorbid IQ, psychotropic medications and previous psychological therapies as predictors of cognitive impairment correctly classified 75% of the participants.

Conclusions This is one of the first studies to use a model selection approach to identify factors associated with cognitive difficulties in BD. Our findings offer the initial steps towards a predictive model for cognitive impairment. This could improve treatment decisions and prioritization for euthymic patients with BD, particularly the implementation of cognitive interventions.

目标虽然认知功能障碍是双相情感障碍 (BD) 的一个突出特征，但先前的研究在估计经历临床相关缺陷的心境正常患者的比例和识别认知困难的预测因素方面存在局限性。我们探讨了临床显着认知障碍的推荐定义与功能结果的相关性，估计了其患病率，并确定了与认知相关的患者特征。

方法 我们评估了 80 名心境正常的 BD 参与者在四个领域的认知表现。参与者根据具有临床意义的认知障碍的两个标准进行分类，我们评估了这些标准在确定的功能结果上区分参与者表现的能力。使用弹性网络回归的变量选择来识别与认知表现相关的社会人口统计学和临床​​因素。选定的变量作为具有临床意义的认知障碍的预测因子进行了逻辑回归检查。

结果 根据选定的标准，34% 的人出现临床上显着的认知障碍。目前较差的认知表现与年龄较大、估计的病前智商较低、目前处方精神药物较多、既往心理治疗较少以及目前使用抗精神病药有关。将病前智商、精神药物和既往心理治疗作为认知障碍预测因素的模型正确分类了 75% 的参与者。

结论 这是最早使用模型选择方法来确定与 BD 中认知困难相关的因素的研究之一。我们的发现为建立认知障碍预测模型提供了初步步骤。这可以改善心境正常的 BD 患者的治疗决策和优先排序，尤其是认知干预的实施。