Bibliographical reference
Garfield V, Farmaki AE, Eastwood SV, Mathur R, Rentsch CT, Bhaskaran K, Smeeth L, Chaturvedi N. HbA1c and brain health across the entire glycaemic spectrum. Diabetes Obes Metab. 2021;1 – 10.
At a glance
Both prediabetes and diabetes are harmful to brain health in terms of cognitive decline and dementia, particularly vascular dementia. Conversely, a blood glucose at the low normal limits seems to have a protective effect on brain health.
What is already known
Diabetes has a negative effect on neurocognitive function, as it increases the risk of dementia and generates neurotoxic effects on structures involved in long-term memory, such as the hippocampus. As such, diabetes alters the speed of cognitive processing and significantly worsens attention, concentration and executive functions. It is hypothesized that neurocognitive function is negatively affected not only by known diabetes but also by non-diabetic alterations of the blood glucose spectrum. To date, there is a lack of large-scale studies to prove this link convincingly. Ensuring operational data from the general population through a biobank allows to study the aforementioned link at the epidemiological level, with intuitive advantages for the classification and prevention of neurological damage due to hyperglycemia.
Study characteristics
This is a cross-sectional and longitudinal study on a sample of the UK population included in the UK Biobank. Through individual clinical data and levels of glycated haemoglobin (HbA1c), the authors identified five predefined blood glucose groups: subjects with normal-low HbA1c (HbA1c <35 mmol/mol), normoglycaemia (HbA1c ≥35 and <42 mmol/mol), prediabetes (HbA1c 42≤48 mmol/mol), unknown diabetes (HbA1c ≥48 mmol/mol) and known diabetes. Supplemented with the results of neurocognitive tests, administrative anamnestic data and neuroradiological examinations, this subgrouping provided the ground for measuring the impact of blood glucose spectrum on the incidence of dementia, cognitive function and decline, hippocampus volume and hyperintensity of the white cerebral matter.
Results achieved
The population sample used for the study comprised 449,973 individuals equally divided by sex, aged 40–69y. Multivariate regression models were used taking the normoglycemic group as control. Study results suggest an increased risk of dementia from all causes and Alzheimer’s dementia for those with known diabetes. Although this latter condition tripled the incidence of vascular dementia, an increase of 54% was also observed in persons with prediabetes. In longitudinal analysis, the risk of cognitive decline was increased by 42% in the prediabetes group and by 39% in those with known diabetes. In the neuroradiological study, an increased volume of white cerebral matter hyperintensity and a reduction in hippocampal volume was seen in the group with prediabetes and unknown diabetes, as well as in those with known diabetes. Conversely, a decreased white cerebral matter hyperintensity and an increased hippocampal volume was documented in subjects with normal-low HbA1c. In multivariate model, a key factor responsible for the excess risk of vascular dementia white cerebral matter hyperintensity was constituted by existing antihypertensive therapy, thereby suggesting a potential driving effect operated by pre-existing arterial hypertension.
Limits of the study
Validated algorithms to define diabetes and dementia were used for the study, however a completely accurate diagnosis of dementia remains debated. In addition, the visual memory tests used longitudinally to define cognitive decline have not shown, according to the authors, good reliability. Finally, the (relatively) low response rate for the UK Bank may have led to preselection problems, i.e. participants may have been less likely to have cognitive problems at the start of the study.
What’s new
Data thus far published on the association between non-diabetic hyperglycaemia and neurocognitive damage have been generally obtained in limited population samples and in small comparative analyses on different blood glucose groups. This extensive study has been carried out on data from a prestigious biobank. Therefore, it offers authoritative evidence that it is the entire hyperglycemic spectrum rather than just diabetes diagnosed to increase the risk of cognitive decline, dementia, and brain alterations involved in long-term memory.
What are the prospectives
Both prediabetes and known diabetes can prove harmful in terms of vascular dementia and risk of cognitive decline. On the other hand, the presence of weaker dissociations with dementia from all causes and Alzheimer’s dementia indicates the existence of mechanisms that likely go beyond cardiovascular risk factors, suggesting different but possibly correlated neurological and vascular pathways. In addition, the results of association between HbA1c normal-low and lesser neuroradiological alterations suggest a protective effect warranting further investigations.
Author: Paolo Marzullo
