ABSTRACT

The current study aimed to determine whether word generation performance on individual within-task 20-second time intervals predicted conversion to Mild Cognitive Impairment (MCI) status. Longitudinal data (Mean follow-up=2.95±1.64 years) was collected from cognitively-healthy community-dwelling older adults (N=344; %female=56.1). Performance on letter and semantic fluency tasks was divided into three consecutive within-task 20-second intervals. Incident MCI status (n=50) was determined via established diagnostic case conference. Fully adjusted Cox proportional-hazards regression models revealed that greater word production on semantic fluency across all time intervals significantly predicted a reduced risk of incident MCI [0-20 seconds (HR=0.906, p=0.002), 21-40 seconds (HR=0.904, p=0.02), and 41-60 seconds (HR=0.892, p=0.017)]. Conversely, on letter fluency, greater word production within the 41-60 second time interval only was significantly associated with reduced risk of incident MCI (HR=0.886, p=0.002). Overall, the clinical use of within-interval performance is supported given evidence of predictive sensitivity and ease of administration.

摘要

目前的研究旨在确定单个任务内 20 秒时间间隔的单词生成性能是否可以预测转换为轻度认知障碍 (MCI) 状态。纵向数据（平均随访时间=2.95±1.64 年）收集自认知健康的社区居住老年人（N = 344；女性百分比 = 56.1）。字母和语义流畅性任务的表现分为三个连续的任务内 20 秒间隔。事件 MCI 状态 ( n = 50) 是通过已建立的诊断病例​​会议确定的。完全调整的 Cox 比例风险回归模型显示，在所有时间间隔内，在语义流畅度方面的单词产生量更大，显着预测 MCI 事件的风险降低 [0-20 秒 (HR=0.906,  p=0.002)、21-40 秒 (HR=0.904,  p =0.02) 和 41-60 秒 (HR=0.892,  p =0.017)]。相反，在字母流利度方面，仅在 41-60 秒的时间间隔内产生更多的单词与降低 MCI 事件的风险显着相关（HR = 0.886，  p = 0.002）。总体而言，鉴于预测敏感性和易于给药的证据，支持区间内性能的临床使用。