Li · The Journal of clinical endocrinology and metabolism 2023 · Cross-sectional observational study · n=29,251

Associations of Glucose Metabolism Status with Brain Macrostructure and Microstructure: Findings from the UK Biobank.

Cited 5 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional observational analysis within a population cohort

PubMed 37497611 · doi:10.1210/clinem/dgad442 · record verified 2026-08-29

What was done

The authors conducted a cross-sectional study of 29,251 UK Biobank participants to examine associations between glucose metabolism status and MRI-derived brain macrostructure and microstructure. Glycemic status was categorized into normal glucose metabolism (NGM), prediabetes, type 2 diabetes (T2D) with HbA1c < 7%, and T2D with HbA1c ≥ 7%. Outcomes included volumes of total and subcortical gray matter, white matter, white matter hyperintensities (WMH), cerebrospinal fluid, brain stem, as well as fractional anisotropy (FA) and mean diffusivity in white matter tracts. Multivariable linear regression adjusted for demographics, lifestyle, medical history, and total intracranial volume.

What was found

Prediabetes was associated with smaller total and subcortical gray matter volumes compared to NGM, with atrophy progressively more pronounced in T2D (all P trend < .05). Participants with T2D had larger white matter and WMH volumes (both P trend < .05). For microstructure, prediabetes was associated with lower FA in commissural fibers (β -0.04; 95% CI -0.08, -0.003). T2D, especially with HbA1c ≥ 7%, showed global and tract-specific microstructural abnormalities (all P trend < .05), except for FA in projection fibers. Exact effect sizes for volumetric outcomes were not reported in the abstract.

Why it matters

This study demonstrates that structural and microstructural brain changes are already present in prediabetes and worsen with uncontrolled diabetes. It highlights the potential importance of early glycemic management to protect brain health.

Limits

The cross-sectional design cannot establish causality or determine longitudinal rates of brain atrophy. The UK Biobank cohort is subject to healthy-volunteer selection bias. Exact effect estimates and confidence intervals for most volumetric metrics were not reported in the abstract.

Cited by