Abstract: Background　Patients with major depressive disorder (MDD) have certain cognitive function impairment, but there are still research gaps in the heterogeneity of cognitive function impairment and the influencing factors of different types. Objective　To explore the potential cognitive function impairment subtypes in MDD patients and analyze the influencing factors of subtypes. Methods　The subjects were MDD patients (n=209) in the outpatient and inpatient departments of the Department of Psychiatry at the First Hospital of Shanxi Medical University from July 2020 to December 2023, and healthy controls (n=51) recruited in the same period. The demographic data questionnaires, Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), Hamilton Depression Rating Scale (HAMD-17), and Pittsburgh Sleep Quality Index (PSQI) were used to investigate. Latent profile analysis was conducted on cognitive function classification of MDD patients based on RBANS scores in various dimensions, and unordered multinomial Logistic regression analysis was used to explore the influencing factors of potential cognitive function classifications. Results　The latent profile analysis results showed that the cognitive function of MDD patients could be divided into three potential subtypes: good cognitive function type (64.6%, 136/209), poor cognitive function-high perceptual type (27.7%, 57/209), and poor cognitive function-low perceptual and attention type (7.7%, 16/209). There were significant differences (P<0.05) in the total RBANS score and various dimension scores, HAMD total score, PSQI total score, age, and education years between MDD patients with different cognitive function subtypes and healthy control groups.The results of the unordered multinomial logistic regression analysis showed that the total scores of HAMD, PSQI, and years of education were the influencing factors of cognitive function classification in MDD patients (P<0.05). Conclusion　The cognitive function impairment of MDD patients has significant heterogeneity, and the risk factors of further cognitive impairment are high degree of depression, poor sleep quality and low years of education.

Key words: Major depressive disorder, Cognitive function, Heterogeneity, Repeatable battery for the assessment of neuropsychological status, Latent profile analysis, Logistic regression

摘要： 背景　重性抑郁障碍（MDD）患者具有一定的认知功能损害，但对于其认知功能损害的异质性及不同分型的影响因素还存在研究空白。目的　探究MDD患者潜在认知功能损害分型，并分析分型的影响因素。方法　研究对象为2020年7月—2023年12月山西医科大学第一医院精神卫生科门诊及病房的MDD患者（n=209）及同期社会招募的健康对照者（n=51）。使用人口学资料问卷、重复性成套神经心理状态测验（RBANS）、汉密尔顿抑郁量表（HAMD-17）及匹兹堡睡眠质量指数量表（PSQI）进行调查。根据RBANS各维度得分对MDD患者认知功能分型进行潜在剖面分析，采用无序多元Logistic回归分析探究MDD患者潜在认知功能分型的影响因素。结果　潜在剖面分析结果显示，MDD患者的认知功能可分为3个潜在分型：认知良好型（64.6%，136/209）、认知不良-高知觉型（27.7%，57/209）、认知不良-低知觉注意型（7.7%，16/209）。不同认知功能分型的MDD患者及健康对照者间的年龄、受教育年限、RBANS总得分及各维度得分、HAMD-17得分、PSQI得分比较，差异有统计学意义（P<0.05）。无序多元Logistic回归分析结果显示，受教育年限、HAMD-17得分、PSQI得分是MDD患者认知功能分型的影响因素（P<0.05）。结论　MDD患者的认知功能损害存在明显异质性，且抑郁程度高、睡眠质量差、低受教育年限是其认知功能进一步损害的风险因素。

关键词: 重性抑郁障碍, 认知功能, 异质性, 重复性成套神经心理状态测验, 潜在剖面分析, Logistic 回归