Long-Term Glycemic Exposure, Control Status and Cognitive Function in Older Adults: A Longitudinal Study

To evaluate the impact of long-term glycemic exposure and control status on cognitive function in older adults.

Using repeated measurements of fasting blood glucose (FBG) from the Prediction for Atherosclerotic Cardiovascular Disease Risk in China project, we calculated 10-year time-weighted cumulative fasting blood glucose (cumFBG) and glycemic variability, and identified FBG trajectories using group-based trajectory modelling among 26 108 participants. In addition, diabetic patients were categorized as consistently controlled or not according to long-term FBG levels. Mini-Mental State Examination was used to evaluate global and domain-specific cognition, and cognitive impairment was defined according to education-specific cutoffs. Linear and logistic regression models were applied to assess associations between FBG-related indicators and cognition.

CumFBG was nonlinearly associated with global cognition and cognitive impairment (both P _{for nonlinearity} < 0.05). Compared with participants with cumFBG < 900 mg/dL × year, those with cumFBG ≥ 1260 mg/dL × year had a 0.091-point lower z-standardized global cognitive score (95% CI: -0.142, -0.040) and a 3% increased cognitive impairment risk (OR: 1.030; 95% CI: 1.010, 1.050). Among specific domains, attention appeared more susceptible, with declines emerging when cumFBG ≥ 900 mg/dL × year. Notably, among individuals with cumFBG ≥ 1260 mg/dL × year, increasing trajectory had a 21.2% increased cognitive impairment risk (OR: 1.212; 95% CI: 1.122, 1.310). Higher variability was associated with worse cognition. Furthermore, only diabetic participants with consistent control exhibited cognition comparable to those without diabetes.

Elevated cumFBG, especially in an increasing pattern, influenced cognition in older adults and sustained glycemic control appeared to mitigate these adverse impacts.