[Diabetes mellitus and cognition: is there a link?].

Abstract

The occurrence of diabetes and dementia is very high in older patients. The fact that both conditions are concurrent raises the question of a possible link between the two. Cognitive functions of non-demented patients with diabetes have been extensively studied. In type 1 diabetes, only a mild decrease of the speed of information processing and of the psychomotor efficiency has been shown. Cognitive decline seems to be related to poor metabolic control and not to hypoglycaemia. In older patients with type 2 diabetes, memory and executive functions have been found impaired. Longitudinal studies of the literature have shown that diabetic patients have a higher chance of developing dementia than non-diabetic patient, with a relative risk (RR) between 1.26 and 2.83. The risk of vascular dementia was increased in 3 out of 5 studies, with a RR ranging between 2 and 2.6. With regard to Alzheimer's disease, the results are conflicting. Half of the studies found an increased risk in diabetic patients (RR: 1.3-2). The possible causal mechanisms of dementia in diabetic patients remain hypothetical. MRI studies showed varying degrees of cortical atrophy, cerebral infarcts and deep white matter lesions. In neuropathological studies, senile plaques and neurofibrillary tangle were not found with higher severity in the brain of diabetic patients than in the brain of age-matched controls. Several hypotheses have been raised to explain the relationship between diabetes and cognitive decline. Micro and macrovascular changes in the brain could induce cerebral hypoxia and ischemic conditions resulting in cellular death or white matter lesions. The occurrence of vascular lesions might reduce the threshold at which dementia will occur in Alzheimer disease. The deposition of advanced glycation end products doesn't spare the brain and they have been found in senile plaques, where they can reduce the solubility of proteins such as the beta amyloid and Tau proteins. Some authors favour the hypothesis of a brain insulin resistance because, in a few small studies, insulin was found to improve memory.