Department of Biomechanics and Movement Science Program, University of Delaware, Newark, DE.
Department of Biostatistics Core Facility, University of Delaware, Newark, DE.
Arch Phys Med Rehabil. 2021 Oct;102(10):1880-1887.e1. doi: 10.1016/j.apmr.2021.03.023. Epub 2021 Apr 21.
To identify homogeneous subsets of survivors of chronic stroke who share similar characteristics across several domains and test if these groups differ in real-world walking activity. We hypothesized that variables representing the domains of walking ability, psychosocial, environment, and cognition would be important contributors in differentiating real-world walking activity in survivors of chronic stroke.
Cross-sectional, secondary data analysis.
University/laboratory.
A total of 283 individuals with chronic (≥6mo) stroke (N=238).
Not applicable.
Thirteen variables representing 5 domains were included: (1) walking ability: 6-minute walk test (6MWT), self-selected speed (SSS) of gait; (2) psychosocial: Patient Health Questionnaire-9, Activities-specific Balance Confidence (ABC) scale; (3) physical health: low-density lipoprotein cholesterol, body mass index, Charlson Comorbidity Index (CCI); (4) cognition: Montreal Cognitive Assessment (MoCA); and (5) environment: living situation and marital status, work status, Area Deprivation Index (ADI), Walk Score. Mixture modeling was used to identify latent classes of survivors of stroke. After identifying the latent classes, walking activity, measured as steps per day (SPD), was included as a distal outcome to understand if classes were meaningfully different in their real-world walking RESULTS: A model with 3 latent classes was selected. The 6MWT, SSS, ABC scale, and Walk Score were significantly different among all 3 classes. Differences were also seen for the MoCA, ADI, and CCI between 2 of the 3 classes. Importantly, the distal outcome of SPD was significantly different in all classes, indicating that real-world walking activity differs among the groups identified by the mixture model.
Survivors of stroke with lower walking ability, lower self-efficacy, lower cognitive abilities, and greater area deprivation had lower SPD. These results demonstrate that the physical and social environment (including socioeconomic factors) and cognitive function should also be considered when developing interventions to improve real-world walking activity after stroke.
确定慢性中风幸存者在多个领域具有相似特征的同质亚组，并检验这些组在现实世界中的行走活动是否存在差异。我们假设，代表行走能力、心理社会、环境和认知领域的变量将是区分慢性中风幸存者现实世界行走活动的重要因素。
横断面、二次数据分析。
大学/实验室。
共有 283 名慢性（≥6 个月）中风患者（N=238）参与。
不适用。
共纳入 13 个代表 5 个领域的变量：（1）行走能力：6 分钟步行测试（6MWT）、自主选择速度（SSS）；（2）心理社会：患者健康问卷-9 项、活动特异性平衡信心量表（ABC 量表）；（3）身体健康：低密度脂蛋白胆固醇、体重指数、Charlson 合并症指数（CCI）；（4）认知：蒙特利尔认知评估（MoCA）；（5）环境：居住状况和婚姻状况、工作状况、区域剥夺指数（ADI）、步行评分。混合模型用于识别中风幸存者的潜在类别。在确定潜在类别后，将日常行走步数（SPD）作为远端结果纳入，以了解不同类别在现实世界行走中的差异。
选择了一个具有 3 个潜在类别的模型。6MWT、SSS、ABC 量表和步行评分在所有 3 个类别之间均有显著差异。2 个潜在类别之间的 MoCA、ADI 和 CCI 也存在差异。重要的是，所有类别的远端结果 SPD 均有显著差异，表明通过混合模型确定的组在现实世界中的行走活动存在差异。
行走能力较低、自我效能较低、认知能力较低、区域剥夺程度较高的中风幸存者日常行走步数较少。这些结果表明，在开发改善中风后现实世界行走活动的干预措施时，还应考虑身体和社会环境（包括社会经济因素）以及认知功能。
