Hoogendoorn Claire J, Hernandez Raymond, Schneider Stefan, Peters Anne, Hawkins Meredith, Pyatak Elizabeth A, Gonzalez Jeffrey S
Ferkauf Graduate School of Psychology, Yeshiva University, Bronx, NY.
Division of Endocrinology and Diabetes, Department of Medicine, Albert Einstein College of Medicine, Bronx, NY.
Diabetes Care. 2025 May 1;48(5):799-806. doi: 10.2337/dc24-2431.
Despite established relationships between glycemia and cognition, few studies have evaluated within-person changes over time. We paired continuous glucose monitoring (CGM) with ambulatory cognitive testing to examine bidirectional associations among adults with type 1 diabetes (T1D).
Participants wore blinded CGM and completed ambulatory tests of perceptual speed and sustained attention five or six times daily for 14 days. CGM metrics were calculated over 3-h periods (mean glucose, %time in range [70-180 mg/dL], %time in low [<70 mg/dL], %time in high [181-250 mg/dL], %time in very high [>250 mg/dL], and coefficient of variation). Immediate glucose values within 15 min of cognitive assessments were also examined. Dynamic structural equation models evaluated bidirectional relationships over sequential 3-h periods.
Among 182 diverse adults with T1D (age 40 ± 14 years, 46% male, 41% Latino, 29% White, 15% Black), more time in low glucose over 3 h was associated with slower perceptual speed at the end of that interval (P < 0.05) but not 3 h later. More time in high glucose (>250 mg/dL) was associated with faster perceptual speed initially but slower speed 3 h later (P < 0.05). Physical activity partially mediated the effect of high glucose on slower perceptual speed. Glycemia did not predict attention scores within persons. Lower attention and higher perceptual speed predicted higher mean glucose and more time in very high glucose over the following 3 h (P < 0.05).
These novel observations of significant bidirectional association between glycemia and cognitive performance over the course of the day among adults with T1D emphasize the importance of examining within-person longitudinal effects over different time frames.
尽管血糖与认知之间的关系已得到确认，但很少有研究评估个体随时间的变化情况。我们将连续血糖监测（CGM）与动态认知测试相结合，以研究1型糖尿病（T1D）成人患者中的双向关联。
参与者佩戴盲法CGM，并在14天内每天进行5或6次感知速度和持续注意力的动态测试。CGM指标是在3小时时间段内计算得出的（平均血糖、血糖在正常范围[70 - 180 mg/dL]的时间百分比、低血糖[<70 mg/dL]时间百分比、高血糖[181 - 250 mg/dL]时间百分比、极高血糖[>250 mg/dL]时间百分比以及变异系数）。还检查了认知评估前15分钟内的即时血糖值。动态结构方程模型评估了连续3小时时间段内的双向关系。
在182名不同的T1D成人患者中（年龄40±14岁，男性占46%，拉丁裔占41%，白人占29%，黑人占15%），3小时内低血糖时间越长，该时间段结束时的感知速度越慢（P<0.05），但3小时后并非如此。高血糖（>250 mg/dL）时间越长，最初的感知速度越快，但3小时后的速度较慢（P<0.05）。身体活动部分介导了高血糖对感知速度减慢的影响。血糖水平无法预测个体的注意力得分。注意力较低和感知速度较高可预测在接下来的3小时内平均血糖水平较高以及极高血糖时间更多（P<0.05）。
这些关于T1D成人患者一天中血糖与认知表现之间显著双向关联的新观察结果强调了在不同时间框架内检查个体纵向效应的重要性。
