Metabolic and microvascular risks linked to brain health in type 1 diabetes

Adults living with type 1 diabetes for decades show subtle dips in memory and processing speed, but a new study suggests their midlife metabolic health may not shape brain outcomes the way researchers expected.

Longstanding type 1 diabetes has long been suspected to affect the brain, but most of the evidence connecting diabetes-related risk factors to cognitive decline comes from studies of type 2 diabetes. Now, a team at the University of Michigan has taken a closer look at a group of adults who have lived with type 1 diabetes for an average of 35 years, asking whether the metabolic factors and microvascular complications that predict cognitive decline in type 2 diabetes also matter for this population. Their findings, published in Annals of Clinical and Translational Neurology, offer a nuanced, and at times surprising, picture of brain health in midlife and older adulthood among people with longstanding type 1 diabetes.

The investigation drew participants from four prior University of Michigan cohorts of individuals with type 1 diabetes. Between February 2021 and August 2024, the researchers enrolled 51 adults, all of whom completed metabolic testing and a comprehensive cognitive assessment, while 20 of them also underwent magnetic resonance imaging (MRI) of the brain. The cohort’s mean age was 56.2 years, 45 percent were women, roughly half were classified as having obesity, and 90 percent held at least a bachelor’s degree. On average, participants had lived with type 1 diabetes for 35.1 years, with a mean hemoglobin A1c, the standard measure of long-term blood sugar control, of 7.2 percent.

The cognitive battery used was the NIH Toolbox Cognitive Battery, a validated computer-based testing platform that yields two composite scores. The fluid cognition composite captures abilities involved in learning, reasoning, memory, and processing speed, drawing on tests of attention and executive function, episodic memory, working memory, executive function, and processing speed. The crystallized cognition composite reflects accumulated knowledge and verbal skills, such as vocabulary and reading recognition, and is often used as a proxy for premorbid intellectual ability or cognitive reserve. Scores are standardized against normative data adjusted for age, sex, race/ethnicity, and education, with a population mean of 50 and a standard deviation of 10.

The results revealed a consistent pattern of subtle cognitive reductions. Fluid cognition scores averaged 47 among participants, statistically lower than the normative mean, and similar reductions were seen in the episodic memory and processing speed subtests, which each averaged 47. At the same time, participants scored significantly above the normative mean on crystallized cognition, at 55, and on oral reading recognition, at 54. The researchers emphasize that although these fluid cognition differences reached statistical significance, they remained less than one standard deviation below the population mean, meaning cognitive performance generally stayed within the normal range. In other words, the reductions may not signal clinically meaningful impairment, a point the authors stress in interpreting their findings.

The metabolic picture was more unexpected. Higher body mass index was associated with better fluid cognition and faster processing speed in linear regression models, and obesity itself was linked to quicker processing speed in a simple comparison, with participants without obesity averaging 42 on the processing speed test compared with 51 among those with obesity. Elevated systolic blood pressure was the only metabolic factor associated with any MRI-derived metric, correlating with greater subcortical gray matter volume. These results run counter to prior research, including findings from the landmark Diabetes Control and Complications Trial (DCCT) and its long-term follow-up, the Epidemiology of Diabetes Interventions and Complications (EDIC) study, which associated severe hypoglycemic episodes, higher HbA1c, higher body mass index, elevated systolic blood pressure, cholesterol, and chronic kidney disease with poorer fluid cognition in middle-aged and older adults with longstanding type 1 diabetes.

The team also examined microvascular complications, the small-vessel damage that affects nerves, eyes, and kidneys in diabetes. Diabetic peripheral neuropathy was diagnosed using the Toronto Consensus definition of probable neuropathy, retinopathy was determined by the presence of proliferative retinopathy in the worst eye based on fundus photographs, cardiovascular autonomic neuropathy was measured through the expiration-to-inspiration heart rate ratio, and chronic kidney disease was defined as an estimated glomerular filtration rate below 60 mL/min/1.73 m². In this cohort, 63 percent had diabetic peripheral neuropathy, 32 percent had retinopathy, and 14 percent had chronic kidney disease. No complication was associated with overall fluid cognition, but diabetic peripheral neuropathy and retinopathy were each linked to slower processing speed. Unexpectedly, diabetic peripheral neuropathy was associated with increased cortical thickness, whereas earlier studies had reported thinner cortices in people with this complication.

The cognitive tests also correlated with brain structure in expected directions. Higher fluid cognition scores were associated with a thicker cortex, greater gray matter volume, and greater subcortical gray matter volume, and better performance on the dimensional change card sort test, a measure of executive function, was associated with greater gray matter volume. These brain-cognition links align with earlier neuroimaging findings in both type 1 and type 2 diabetes populations.

The authors are careful to frame the work as exploratory. Because of the modest sample size and the large number of statistical comparisons, they applied the Benjamini-Hochberg procedure to control the false discovery rate, and none of the findings survived that correction. Effect sizes from the regression analyses were generally small, with Cohen’s f² values below 0.15, indicating modest associations with limited clinical impact. The paradoxical associations, such as the link between obesity and faster processing speed, may therefore reflect false positive results, though the researchers note that similar counterintuitive findings have emerged in type 2 diabetes, where obesity has in some studies correlated with better cognition, hinting that the relationship between body weight and brain health is more complex than often assumed.

Several explanations are offered for the absence of expected associations. Diabetes duration and HbA1c were not consistently linked to brain outcomes, possibly because the cohort, composed of people actively involved in clinical research, had relatively well-controlled disease with a mean HbA1c of 7.2 percent, reducing detectable effects. Limited statistical power is another possibility. The authors also highlight the role of cognitive reserve: with 90 percent of participants holding at least a college degree, high educational attainment may have buffered the negative impacts of decades of diabetes on cognitive performance during this age period. Higher education has been shown to protect against cognitive decline in type 2 diabetes, and the authors call for future work in populations with lower education levels to determine whether longstanding diabetes exerts differential effects across education strata. They further note that they did not account for comprehensive midlife risk factors such as physical activity, depression, hearing loss, or smoking status.

The study carries important limitations beyond sample size. The cohort was self-selected and predominantly white, the MRI subsample was small, there was no matched control group without diabetes, and dementia risk factor assessment was incomplete. The authors also caution that after correction for multiple comparisons, none of the results remained statistically significant, so the findings should be treated as hypothesis-generating rather than definitive.

Still, the work fills a notable gap. People with type 1 diabetes are now living longer than ever, reaching ages where cognitive impairment becomes clinically significant, yet relatively few studies have examined cognition in midlife and older adults with this condition. By testing whether the associations found in DCCT-EDIC, the only prior study to comprehensively assess how metabolic factors and complications influence midlife brain outcomes in type 1 diabetes, replicate in an independent cohort, this research helps clarify which determinants of brain health hold up outside that landmark trial. The answer, so far, is that the picture is less consistent than expected: metabolic factors did not show robust links to brain outcomes, and some microvascular complications produced unexpected patterns.

The broader implication is that the trajectory from decades of type 1 diabetes to brain health in midlife may be more resilient, or at least more variable, than assumed, potentially thanks to modern glucose control and protective factors such as education. But the researchers are clear that longitudinal studies in larger and more diverse populations are needed to determine how midlife metabolic and microvascular factors ultimately influence brain aging in this growing population. As the cohort of people living for decades with type 1 diabetes continues to expand, understanding these relationships will become increasingly important for protecting cognitive health in aging patients.