Introduction: The risk of dementia increases with increasing levels of hemoglobin A1c (HbA1c), a measure of average blood glucose. However the association between glucose peaks and dementia in persons with diabetes is unclear. 1,5-anhydroglucitol (1,5-AG) is a novel biomarker of glucose peaks, with low levels reflecting blood glucose concentrations that exceed the renal filtration threshold (~180 mg/dl) over the prior 7-10 days.
Hypothesis: Glucose peaks, indicated by low levels of 1,5-AG, will be associated with higher risk of dementia, independent of HbA1c and other dementia risk factors.
Methods: We conducted a prospective analysis of 1568 participants with diabetes from visit 2 (1990-1992) of the ARIC Study. Diabetes was defined based on self-reported physician diagnosis, diabetes medication use, or HbA1c ≥ 6.5%. Community surveillance of hospitalizations was used to identify incident dementia. Cases were defined as the first occurrence of a hospitalization with an ICD-9 code for dementia appearing in any position in the hospital discharge record. We used Cox regression to examine the association between 1,5-AG (modeled linearly) and incident dementia, adjusted for demographic and clinical risk factors, including HbA1c.
Results: At baseline, the mean age was 58 years, 58% were female, 54% were white, and mean HbA1c was 8.1%. Over a median follow-up of 18 years, 158 participants had a hospitalization with an ICD-9 code of dementia. Each unit decrease in 1,5-AG increased the risk of dementia by 4% (HR=1.04, p-value=0.03, Figure). Associations were similar when 1,5-AG was modeled using linear splines.
Conclusions: Short-term glucose peaks were associated with higher risk of dementia among adults with diabetes, independent of HbA1c and other dementia risk factors. More studies are needed to determine if targeting glucose peaks, in addition to average glucose, among persons with diabetes can reduce the risk of dementia.