Poor glycaemic status singled out for dementia risk in T2D

Poor glycaemic status may predispose individuals with type 2 diabetes (T2D) to dementia, with the risk especially high in those experiencing significant fluctuations in glycaemic levels, as reported in a retrospective cohort study.

Analysis of large data from Taiwan’s National Health Insurance Database showed that poor glycaemic status was associated with a twofold increase in dementia risk among people with T2D (adjusted hazard ratio [HR], 2.01, 95 percent confidence interval [CI], 1.58–2.68). [Diabetes Res Clin Pract 2025;222:112065]

The association was pronounced among those who were at least 45 years of age and held for all types of dementia including Alzheimer’s disease, vascular dementia, and others, with the most pronounced effect observed for vascular dementia, the investigators noted.

Looking at the type of poor glycaemic status, the risk of dementia was elevated for those with either hyperglycaemia (HR, 1.64, 95 percent CI, 1.30–2.18) or hypoglycaemia (HR, 1.56, 95 percent CI, 1.22–2.08) and was highest for those experiencing both hyperglycaemic and hypoglycaemic episodes (HR, 3.10, 95 percent CI, 2.48–4.12) relative to people with T2D who had better glycaemic status.

A median dementia onset of just 6.18 years was seen in people with T2D patients who had both hyperglycaemic and hypoglycaemic events, the shortest duration observed.

The analysis included 192,977 people with T2D, among whom 57,893 had poor glycaemic status during the follow-up period. Dementia occurred in 4,342 people who had either hypoglycaemic or hyperglycaemic events or both.

Predicting dementia

In a receiver operating characteristic (ROC) analysis for predicting dementia, the optimal cutoff of hyperglycaemic and hypoglycaemic events was 15. At a cutoff of 15 episodes of mixed hypoglycaemia and hyperglycaemia, the model showed a moderate discriminative ability. The corresponding area under the curve (AUC) was 0.638 (95 percent confidence interval, 0.619–0.670; p<0.001), and the prediction sensitivity was 49.1 percent.

Therefore, dementia risk is elevated if poor glycaemic status count reaches or exceeds 15, the investigators said. However, they acknowledged that when predicting the risk of dementia, not only the number of mixed hyperglycaemia and hypoglycaemia occurrences should be considered but also the timeframe of occurrence.

“Frequent occurrences of mixed [hypoglycaemic and hyperglycaemic events] within a short timeframe (ie, 1.6 years) may indicate an elevated risk of dementia compared with the risk in patients having the same number of poor glycaemic status occurrences over a relatively long duration,” the investigators said.

They emphasized that dementia screening tools for people with T2D could benefit from the inclusion of poor glycaemic status event count. But to achieve higher predictive accuracy than the current AUC of 0.638, poor glycaemic status should be combined with other relevant clinical indicators, they added.

Damaging the brain

These findings are consistent with those of other cohort studies and indicate that greater glycaemic variability may be a key risk factor for dementia, the investigators said. [Diabetes Care 2021;44:1556-1563; J Gerontol A Biol Sci Med Sci 2017;72:1110-1116]

“Abnormal fluctuations in glucose levels may damage the blood–brain barrier, leading to neurodegeneration,” they explained. “In addition, such fluctuations can lead to insulin resistance, reducing the efficiency of glucose metabolism in the brain and causing the abnormal accumulation of proteins and amyloid deposits.” [J Pharmacovigil 2014;2:125]

The buildup of proteins and amyloid deposits in the brain results in the inflammation of neural tissues, which, in turn, promote dementia, according to the investigators. [Biochim Biophys Acta Mol Basis Dis 2014;1842:1693-1706; Diabetes Res Clin Pract 2019;158:107911]

Taken together, the evidence highlights the importance of optimal glycaemic management in T2D to mitigate the risk of dementia, they said.