journal articles
LONG-TERM FASTING INSULIN VARIABILITY AND COGNITIVE FUNCTION: INSIGHTS FROM THE CARDIA STUDY
Bo-Shui Huang, Zuo-Yu Huang, Yu-Hong Zeng, Kun-Hao Bai, Jing-Bin Guo, Jun Weng, Ze-Hua Li, Qing-Yun Hao
BACKGROUND AND AIM: Fasting insulin variability has emerged as a potential marker of metabolic dysregulation, but its long-term implications for cognitive function remain unclear. This study aimed to clarify the role of long-term fasting insulin variability in predicting individual cognitive function risk.
METHODS: We analyzed data from CARDIA study participants who underwent cognitive testing and had at least three insulin measurements. Fasting insulin was measured at 7 timepoints over 30 years. Intra-individual insulin variability was assessed using standard deviation (SD), coefficient of variation (CV), and average real variability (ARV). Cognitive function was evaluated using the Digit Symbol Substitution Test (DSST), Stroop Test, and Rey Auditory Verbal Learning Test (RAVLT), with results standardized to z-scores and combined into a global cognitive z-score. Multivariable linear models were used to assess associations with cognitive performance.
RESULTS: In the 25-year analysis (n = 2712), higher long-term insulin variability was significantly associated with poorer global cognitive performance at year 25 after adjustment for demographic, lifestyle, and cardiometabolic covariates (CV-insulin: β=–0.719; 95% CI: –1.161 to –0.276; P < 0.01; SD-insulin: β=–0.019; 95% CI: –0.036 to –0.002; P < 0.05). These associations remained significant after additional adjustment for either concurrent insulin at year 25 or mean insulin levels over 25 years. Domain-specific analyses showed that higher insulin variability was associated with lower DSST z-scores (worse attention) and higher Stroop interference z-scores (worse executive function). Extended analyses over 30 years (n = 2069) yielded consistent results: higher CV-insulin was inversely associated with global cognitive z-scores (β=–0.837; 95% CI: –1.347 to –0.327), as well as with DSST (β=–0.347; 95% CI: –0.581 to –0.112) and RAVLT z-scores (β=–0.276; 95% CI: –0.522 to –0.031). These associations persisted after full adjustment for year 30 covariates and time-varying confounders across the follow-up, supporting the temporal robustness and clinical relevance of insulin variability as an independent predictor of cognitive function.
CONCLUSIONS: Greater long-term insulin variability is independently associated with poorer midlife cognitive performance. These findings highlight insulin variability as a potential marker of cognitive health risk.
CITATION:
Bo-Shui Huang ; Zuo-Yu Huang ; Yu-Hong Zeng ; Kun-Hao Bai ; Jing-Bin Guo ; Jun Weng ; Ze-Hua Li ; Qing-Yun Hao (2026): Long-term fasting insulin variability and cognitive function: Insights from the CARDIA study. The Journal of Prevention of Alzheimer’s Disease (JPAD). https://doi.org/10.1016/j.tjpad.2026.100487