A research team co-led by a BBRF grantee has shed new light on how changes in blood glucose levels in the body correspond with changes in cognition.

Publishing in the journal NPJ Digital Medicine, the team reports on its efforts to do something that new technologies have made possible for the first time: to monitor with great precision how moment-to-moment fluctuations in glucose levels in patients with type I diabetes impact moment-to-moment fluctuations in certain cognitive domains.

Co-led by 2017 BBRF Young Investigator Laura T. Germine, Ph.D. and Naomi S. Chaytor, Ph.D., of, respectively, McLean Hospital/Harvard Medical School and Washington State University, the team enrolled 200 type 1 diabetes patients, 190 of whom formed the group whose results were analyzed. Zoe Hawks, Ph.D., was first author of the team’s paper. Type 1 diabetes, characterized by elevated glucose levels and above-normal variability in blood glucose levels over time, is associated with many adverse health outcomes including mild neurocognitive disorder, dementia, and microvascular complications.

The relationship between glucose and insulin levels in the body and brain, on the one hand, and brain function, on the other hand, have been of increasing interest to neuropsychiatric researchers and clinicians in recent years. For example, the winner of BBRF’s 2023 Colvin Prize for Outstanding Achievement in Mood Disorders Research, Roger S. McIntyre, M.D., FRCPC, of the University of Toronto, has been exploring possible relationships between metabolic disorders such as obesity and diabetes and anomalies in insulin signaling in brain cells in major depression, bipolar disorder and possibly other psychiatric illnesses.

The brain constitutes only a few percent of the body by weight, yet consumes about one-fourth of its sugar, which supplies its extraordinary “fuel” requirements. Insulin helps move sugar into cells. While insulin receptors are extremely common throughout the brain, it is not yet clear whether (and if so, how) insulin, metabolic disorders and psychiatric illnesses may be causally related.

The study led by Drs. Germine and Chaytor was made feasible by combining continuous glucose monitoring (CGM) with cognitive ecological momentary assessment (EMA). CGM is used to monitor blood sugar levels in patients frequently, e.g., every 5 minutes. With outputs readable over common devices such as cellphones, CGM is used to help patients manage their blood sugar levels over the course of a day—what is called by researchers a “naturalistic” context. Cognitive EMA is used in studies in which research subjects are asked to complete very brief cognitive tasks several times a day using smartphone devices. This technology has been shown to accurately assess variations in cognitive skills within single individuals over the course of a day or days, as well as the range of variations between different individuals asked to perform the same cognitive tasks. Both tools came into play in the new study.

Making associations between changing glucose levels and cognitive performance hasn’t been possible previously for technical reasons. But addressing this relationship “is critical to clarify when and for whom glucose fluctuations predict cognitive impairment,” the team notes. Most immediately such knowledge would help improve treatment of individual diabetes patients, aiming to “maximize glycemic control and cognitive performance.” But the results also may help researchers better understand the questions about possible linkages between blood sugar, insulin, metabolic disorders, and psychiatric disorders.

The study cohort in the team’s experiments was on average about 45 years old (the range was 18-84), well-educated, mostly White, and 53 percent female. Average blood glucose levels were equally distributed across a wide range seen in type 1 diabetes.

Analysis revealed that large glucose fluctuations over the intervals measured were associated with slower and less accurate neural processing speed. Processing speed refers to the pace at which one can take in and respond to information. Slight elevations in glucose, relative to an individual’s mean level, were associated with faster processing speed. In the other cognitive domain tested, sustained attention, glucose fluctuations were not observed to affect performance. Sustained attention is the ability to focus on an activity or stimulus over extended periods of time.

A variety of factors pertaining to type 1 diabetes patients, and several that were not specific to diabetes, were found to predict individual differences in cognitive vulnerability to glucose fluctuations. Subjects were more likely to be cognitively vulnerable to glucose fluctuations if they were older, spent longer periods in a hypoglycemic state (i.e., had low blood sugar levels); had microvascular complications; tended to have glucose levels that fluctuated more over a given interval; tended to experience fatigue more often; and, curiously, if they had larger neck circumferences. The latter, unexpected, is thought by the team to relate to other factors related to cognitive impairment including sleep apnea risk and an abundance of fatty tissue in the upper body.

People with these risk factors might benefit from “limiting consequential speed-dependent cognitive tasks at moments when glucose is considerably above or below” their typical level. More broadly, the results support theories suggesting accumulated diabetes-related insults increase cognitive vulnerability to short-term changes in glucose levels, the team said.

Although randomized control trials will be needed to lend empirical support for interventions targeting risks the study identified, the current results, in the team’s view, “provide a powerful model” for understanding how cognitive fluctuations are related to glucose metabolism, and specifically, changes in glucose levels over the course of a day.