Sweets Shrinking Your Brain?Written By: Pauline Anderson
September 5, 2012 — Even blood glucose levels in the range of high normal appear to be associated with brain atrophy, a new study shows.
In a sample of randomly selected older middle-aged people, high normal levels of fasting plasma glucose were significantly associated with hippocampal and amygdalar atrophy over 4 years.
These results should spur discussion of whether the range for blood glucose that is currently considered normal should be changed, researchers say.
“This is just one study, so first we have to replicate it, but if we find the same results in other samples, we may need to tighten somewhat the definition of what normal glucose levels should be,” said author Nicolas Cherbuin, PhD, director of the Neuroimaging and Brain Laboratory, Centre for Research on Ageing, Health and Wellbeing, at Australian National University, in Canberra, Australia.
The study is published in the September 4 issue of Neurology.The World Health Organization uses 6.1 mmol/L as the cutoff for normal blood glucose, and the American Diabetes Association uses the more stringent threshold of 5.6 mmol/L. In this study, atrophy was seen in participants with blood glucose levels less than 6.1 mmol/L and less than 5.6 mmol/L.
The study included 249 participants 60 to 64 years of age residing in Canberra and Queanbeyan, in Australia. None of the participants had a neurological disorder, were taking antidepressants, or had diabetes. All had a blood glucose level below 6.1 mmol/L (mean, 4.93 mmol/L).
Each participant underwent magnetic resonance imaging (MRI) at 2 intervals (wave 1 and wave 2). Researchers estimated hippocampal and amygdalar volumes using a manual method of tracing MRI scans one slice at a time.
After a mean follow-up of 4 years, and with control for age, sex, intracranial volume differences from wave 1 to wave 2, APOE4 genotype, hypertension, alcohol intake, body mass index (BMI), smoking, and other factors, researchers found that higher plasma glucose levels were negatively associated with left and right hippocampal and amygdalar volumes.
Glucose level as a predictor of hippocampal volume was -131 (P = .001) for left and -148 (P = .001) for right after control for baseline volume, meaning that higher glucose was associated with greater atrophy, said Dr. Cherbuin.
The rate of hippocampal atrophy was 2% per annum, and plasma glucose effects accounted for 6% to 10% of atrophy, the authors note.
Researchers undertook several sensitivity analyses to further eliminate the possibility that other factors contributed to the association. One analysis used the stricter threshold of fasting plasma glucose less than 5.6 mmol/L. This analysis came up with almost identical results.
Findings of another analysis that included participants with type 2 diabetes, or glucose levels higher than 6.1 mmol/L, followed a similar pattern.
A third sensitivity analysis, which excluded participants with BMI greater than 25, had essentially the same — even slightly stronger — findings. This suggests that effects were due to glucose levels, not to fat mass.
“So these were very stringent benchmarks,” commented Dr. Cherbuin. “Even after doing all these analyses, the results were still as significant and as strong.”
The study results, he added, are unlikely to be attributed to preclinical states, and were not due solely to preexisting individual differences in brain structure, because participants with neurological disorders, dementia, and mild cognitive impairment were excluded.
The authors conclude that “short of a randomized controlled trial,” which would be unethical in this context, these extensive analyses show a “convincing link between plasma glucose and medial temporal lobe atrophy.”
High blood glucose levels can affect the brain through various mechanisms, perhaps the most likely being fallout from inflammatory processes, according to Dr. Cherbuin. “Higher glucose levels trigger an anti-inflammatory cascade. We know this is not good for the brain, and particularly [is] not good for the hippocampus, which is sensitive to environmental stresses, whether inside the body or coming from outside.”
Another possible mechanism is abnormal blood coagulation. High glucose levels may lead to abnormal coagulation, thereby increasing the likelihood of thrombosis, microemboli, and clinical and subclinical strokes, which are risk factors for brain aging.
Psychological stress could be another culprit. Stress is associated with increased activation of the hypothalamic-pituitary-adrenal (HPA) axis, and this activation has been linked to increased glucose levels.
A limitation of the study is that it lacked ethnic diversity.
Researchers have now completed a third wave of the study, which followed study participants to 8 years. These data are in the process of being analyzed, said Dr. Cherbuin.
In addition to testing the association between glucose levels and brain changes in a larger sample, another next step for researchers is to see whether glucose levels affect areas of the brain other than the hippocampus.
To do this, Dr. Cherbuin and colleagues plan to use MRI techniques other than the time-consuming and relatively expensive manual tracing method used in the current study — including a technique that uses a semi-automated measure to extract brain volumes.
“Using these techniques, we may be able to say that particular areas of the brain are associated with glucose levels, and to extract the values of these regions to see if they relate to cognition,” said Dr. Cherbuin. “It’s still a work in progress, but preliminary evidence suggests that glucose may impact other areas of the brain to affect cognitive performance.”
Dr. Cherbuin and the research team are also keen to look at the role of diet. They will investigate how eating foods that are more readily transformed into glucose — for example, refined foods and foods loaded with carbohydrates, fats, and refined sugars — affects the brain.
They plan to investigate how high blood glucose interacts with other risk factors for diabetes, such as lack of exercise and depression, to produce changes in the brain.
“We want to determine which risk factors — diet, exercise, stress, obesity — are more strongly associated with higher blood sugar levels,” he added.
The study was supported by grants from the National Health and Medical Research Council of Australia, and by an Australian Rotary Health Research Fund grant. The authors have disclosed no relevant financial relationships.
Neurology. 2012;79:1019-1026. Abstract