2948-LB: Short-Term Glucose and Dynamic Working Memory in Youth with Type 1 Diabetes: Exploratory Evidence from Ecological Momentary Assessment

Introduction and Objective: Studies show lower cognitive scores in youth with type 1 diabetes (T1D), but most assessed long-term glycemia and cognition from a single in-lab session. Little is known about how short-term glucose variability affects dynamic cognition in daily life. We explored whether glucose variability predicts working memory (WM) in the daily lives of youth with T1D using ecological momentary assessment. Methods: Youth with T1D (age 9-16y) completed cognitive tasks on a smartphone app up to 5 x day for 10-14 days while wearing a continuous glucose monitor (CGM). The primary outcome was performance on a spatial WM task (0 = perfect score). Each person’s baseline glucose was the mean over the first 24 hours of CGM. Acute glucose was the mean in each 2-hour period before a WM task. The primary predictor was glucose deviation (acute-baseline). Secondary predictors were acute standard deviation (SD), coefficient of variation (%CV), and 75th percentile (Q75). We fit linear mixed effects models with a suitably transformed WM score as the dependent variable, and baseline glucose and acute glucose as independent variables, along with a random intercept for each person to account for repeated measures. Results: 856 sessions were analyzed from N=20 youth. Glucose deviation did not predict WM (β = 0.00033, p = 0.2). SD of glucose was marginal, indicating some degree of trend in variability (p = 0.1), but CV (p = 0.4) and Q75 (p = 0.2) were not. Per-participant slope analyses indicated that 18/20 had no significant slope differences. Supplementary analyses showed SD was significant when participants with low WM task adherence were excluded (≥30 sessions: p = 0.03). Conclusion: In this exploratory analysis, short-term glucose variability was not associated with WM in youth with T1D, though SD warrants more testing. Future analyses will include other variables (e.g., mood, age) to identify subgroups with different sensitivity to glucose variability, explore other pre-test periods, and test other cognitive domains. Disclosure H. Zhao: None. A.D. Atkinson: None. B. Rosenberg: None. M. Ray: None. Funding NIH/NIDDK 5K01DK13133903 NIH/NHLBI T32HL130357NIH/NIDA T32 DA007261NIH/NCATS/WashU ICTS KL2TR002346NIH/NIDDK/WashU DRC P30DK020579