The blood–brain barrier in diabetes mellitus: A critical review of clinical and experimental findings

Abstract Although dysfunction of blood-retinal barrier in diabetes is well established, the role of the blood-brain barrier (BBB) in diabetic cerebral pathophysiology is unclear. A critical appraisal of the literature revealed many contradictory findings in both experimental and human studies. There is consistent evidence that the diabetic capillary basement membrane is thickened. However evidence of permeability dysfunction is limited, with the BBB reported as either normal (IgG, sucrose, cytochrome c , Horseradish peroxidase, EDTA, α-AIB) or variably leaky (albumin, inulin, Gd-DTPA). Evidence for dysregulation in blood-brain glucose transport, and the GLUT-1 transporter is inconsistent. Abnormal endothelial transport of choline, basic and neutral amino acids have been reported. Abnormalities have been identified in the tight junctional protein Occludin, but not in ZO-1. A difficulty with much of the experimental work in rodents is that different models have been used, numbers have often been small, methods variable and experimental studies performed after variable periods of diabetes or ‘control’ treatment. No studies have addressed dysfunction in efflux proteins such as P-glycoprotein or the organic cationic or anionic transporter families. Pathophysiological studies in humans, which do not support dysfunction of blood-brain glucose transport, are confounded by age, smoking, hypertension, type and duration of diabetes, quality of glycaemic control, and cerebrovascular disease. Future experimental and clinical studies with well defined controls and recorded co-morbidity, and utilisation of genomic and proteomic methodologies will help clarify the structural and functional changes in the diabetic BBB, and their contribution to neurological and cognitive dysfunction.