Salivary amyloid β42 levels in mild cognitive impairment among aged diabetics

Key summary pointsAimThe aim of this study was to explore the link between salivary amyloid beta 42 level (as a surrogate marker for Alzheimer's disease pathology) and mild cognitive impairment among old diabetic patients.FindingsThere is an increased salivary amyloid beta 42 level among aged diabetics with mild cognitive impairment without other central nervous system disorders. Its level is affected by diabetes control but not the duration of diabetes.MessagesThe elevated salivary amyloid beta 42 among diabetic patients with mild cognitive impairment may offer emerging evidence suggesting a possible contribution of Alzheimer's pathology. However, linking it to amyloid burden using amyloid positron emission tomography tracer is highly needed. AbstractPurposeThe underlying pathology for cognitive decline in diabetic patients is uncertain. It was originally linked to vascular causes; however, possible contribution of Alzheimer's pathology was debated. This study explored the link between salivary amyloid beta 42 level (as a surrogate marker for Alzheimer's pathology) and mild cognitive impairment (MCI) among old diabetic patients.MethodsA case-control study included 90 diabetic participants,>= 60years of age, divided into 45 cases with MCI and 45 controls. Patients with history of head trauma, any central nervous system pathology, depression, dementia or delirium, those who received anticholinergic drugs, or refused to participate in the study were excluded. Assessment of the relationship between salivary A beta 42 level and neuropsychological performance was done using a battery consisting of the logical memory test, forward and backward digit span tests, category fluency test, go/no go test, stick design test, and second-order belief.ResultsSalivary A beta 42 levels were higher in MCI diabetics versus controls (P=0.014), it predicted MCI among aged diabetics, even after adjustment for confounding vascular risk factors. Salivary A beta 42 had moderate accuracy to identify MCI (area under curve=0.654, P=0.008). At cut-off >= 47.5pg/ml, sensitivity, specificity, positive predictive value and negative predictive value were 80%, 47%, 60% and 70%, respectively.ConclusionCurrent data support that MCI in diabetics, without CNS disorders, is associated with a surrogate marker of Alzheimer's pathology.