Cognitive function is dynamic in post-stroke patients, however, there are limited studies on cognitive recovery after stroke. The definition of cognitive reverters and neuropsychological characteristics according to various definitions have not yet been investigated.

To investigate the neuropsychological characteristics of cognitive reverters after stroke based on the traditional definition and new definition according to Montreal Cognitive Assessment (MoCA) with a double threshold.

A total of 163 patients hospitalized for first onset acute ischemic stroke were recruited from the Department of Neurology of the Second Hospital of Hebei Medical University from December 2020 to February 2022 as the study subjects. All patients were assessed for cognitive function in the acute period of ischemic stroke using the MoCA at baseline. Cognitive function was assessed again at the 6-month post-stroke follow-up with the detailed neuropsychological tests refined at the same time, including the digit span task (DST) , Stroop color and word test (SCWT) , Chinese version of Boston naming test (BNT) , verbal fluency test (VFT) , clock drawing test (CDT) , and auditory verbal learning test (AVLT) . The enrolled patients were divided into the M1 group (MoCA scores <21) , M2 group (21 ≤MoCA scores≤25) and M3 group (MoCA scores≥26) according to the MoCA with adouble threshold. While the patients were also divided into the reverters 1 (R1) group, nonreverters 1 (NR1) group, R2 group and NR2 group according to the two definitions of reverters.

Patients were followed up after 6 months, of which 28 were lost to follow-up and 135 were finally included in the study. The average MoCA score in the acute period of stroke was (20.1±5.1) in 135 patients. There were 40 cases in the M1 group, 61 cases in the M2 group and 34 cases in the M3 group. The scores of the MoCA, forward DST, reverse DST, total DST, Chinese version of BNT, CDT, numbers of VFT-animal, VFT-fruit and VFT-vegetable, AVLT-immediate scores, AVLT-short delay scores, AVLT-long delay scores and AVLT-recognition scores in the M2 and M3 groups 6 months after stroke were higher than the M1 group (P<0.05) , Stroop C time consuming and numbers of Stroop C errors lower than M1 group (P<0.05) . The scores of the MoCA, reverse DST, total DST, AVLT-short delay score and AVLT-long delay score in the M3 group than the M2 group (P<0.05) . 120 patients of the 135 patients with MoCA scores<26 in the acute period of ischemic stroke were selected as the cognitive reveters study subjects. Compared with the MoCA scores during the acute period of ischemic stroke, 73 patients improved ≥2 points during the follow-up period in the R1 group and 47 patients improved <2 points in the NR1 group, with a recovery rate of 60.8% (73/120) . Patients in the R1 group had lower MoCA scores in the acute period of ischemic stroke than the NR1 group, and higher MoCA scores in the follow-up period than the NR1 group (P< 0.05) . Compared with the double threshold classification of MoCA in the acute period of stroke, 50 cases had increased scores and crossed over to the R2 group and 70 cases did not cross over to the NR2 group at follow-up, with a recovery rate of 41.7% (50/120) . Patients in the R2 group had higher MoCA score in the acute period of stroke, MoCA scores in the follold-up period, reverse DST scores, total DST scores, Chinese version of BNT scores, VFT-animal, VFT-vegetable number, CDT scores, AVLT-immediate scores, AVLT-short delay scores, AVLT-long delay scores, and AVLT-recognition scores during follow-up were higher than the NR2 group, and the Stroop C time consuming was lower than the NR2 group (P<0.05) .

There are varying degrees of impairment in attention, visuospatial function and delayed recall in post-stroke patients. There is no significant difference in neuropsychological characteristics between reverters and nonreverters. However, the scores of reverters are higher based on the definition using double threshold for MoCA scores, which is more relevant for clinical use.

卒中后患者的认知功能是动态变化的，但关于卒中后认知"恢复"的研究较少，且尚无研究对"恢复者"定义及不同定义下的神经心理学特征进行分析比较。

探究基于传统定义与蒙特利尔认知评定量表（MoCA）双阈值新定义下卒中患者认知恢复的神经心理学特征。

选取2020年12月—2022年2月在河北医科大学第二医院神经内科住院的首发急性缺血性卒中患者163例为研究对象。基线时采用MoCA评估患者卒中急性期认知功能。卒中后6个月随访时再次评估认知功能，同时完善详细的神经心理学测试，包括数字广度测验（DST）、Stroop色词测验C（Stroop C）、波士顿命名测试（BNT）中文版、言语流畅性试验（VFT）、画钟试验（CDT）、听觉词语学习测验（AVLT）。一方面依据MoCA双阈值（20/21、25/26）将患者分为3组，即MoCA评分<21分为M1组、MoCA评分≥26分为M3组、其余为M2组。另一方面依据两种"恢复者"定义将患者分为恢复者1组（R1组）与非恢复者1组（NR1组）、R2组与NR2组。

6个月后对患者进行随访，其中28例失访，最终纳入135例为研究对象。135例患者平均卒中急性期MoCA评分（20.1±5.1）分。卒中后6个月时M1组40例、M2组61例、M3组34例。M2组、M3组患者卒中后6个月MoCA评分、正向DST评分、逆向DST评分、DST总分、BNT中文版评分、VFT-动物、VFT-水果、VFT-蔬菜、CDT评分、AVLT-即刻评分、AVLT-短延迟、AVLT-长延迟、AVLT-再认评分均高于M1组，Stroop C耗时、Stroop C错误数少于M1组（P<0.05）；M3组患者卒中后6个月MoCA评分、逆向DST评分、DST总分、CDT评分、AVLT-短延迟、AVLT-长延迟高于M2组（P<0.05）。135例患者中120例患者卒中急性期MoCA评分<26分，作为"恢复者"研究对象。与卒中急性期MoCA评分相比，73例患者随访期提高≥2分者为R1组，47例患者提高<2分为NR1组，恢复率为60.8%（73/120）。R1组患者卒中急性期MoCA评分低于NR1组，随访期MoCA评分高于NR1组（P<0.05）。与卒中急性期MoCA双阈值分类相比，50例随访期评分增加并跨类为R2组，70例未跨类为NR2组，恢复率为41.7%（50/120）。R2组患者卒中急性期MoCA评分、随访期MoCA评分、逆向DST评分、DST总分、BNT中文版评分、VFT-动物、VFT-蔬菜、CDT评分、AVLT-即刻评分、AVLT-短延迟、AVLT-长延迟、AVLT-再认评分高于NR2组，Stroop C耗时低于NR2组（P<0.05）。

卒中后患者注意力、视空间功能及延迟回忆存在不同程度受损。传统定义下，恢复者与非恢复者组评分神经心理学测试差异不大。而基于MoCA双阈值的新定义，恢复者评分较高，更具有临床使用意义。