Vascular risk factors in Alzheimer’s disease and Mild Cognitive Impairment: population data from the Zabùt Aging Project

Background: Alzheimer's disease (AD) is the most common cause of dementia in older adults, accounting for about 60% of cases. However, autopsy studies suggested that mixed dementia, with vascular and neurodegenerative AD pathology, accounts for nearly 20% of dementia cases. Aims: Aim of the present study was to evaluate the relationship between isolated or clustered Vascular Risk Factors (VRFs)/diseases and Mild Cognitive Impairment (MCI) or AD. The study was conducted using a Sicilian population-based cohort dataset collected in low-educated, rural subjects, the Zabùt Aging Project (ZAP). The effect-modification by age, sex, education, genetic factor (APOE4 allele carrier), undernutrition, inflammatory status and depressive symptoms was taken into account. Materials and Methods: VRFs, hypertension, coronary heart disease, atrial fibrillation (FA), previous Transient Ischemic Attack (TIA)/stroke, diabetes, dyslipidemia, obesity and Metabolic Syndrome were identified with a semi-structured questionnaire, physical measurements and laboratory analysis. The Framingham general cardiovascular disease (CVD) risk profile, the Framingham stroke risk profile, the Cardiovascular Risk Factors, Aging, and Dementia (CAIDE) risk score and Late-Onset Alzheimer Disease (LOAD) vascular risk score were used as markers of vascular burden. Framingham algorithms are multivariable scores that provide a sex-specific absolute risk of cardiovascular events, and CAIDE and LOAD are dementia risk scores. The importance of each VRF in determining the outcome of each diagnosis was modelled by means of a sequential procedure based on the logistic regression, using subjects with normal cognition [NC] as reference group. The contribution of each VRFs in determining the outcome and the mutual relationship among VRFs in separating the diagnoses were studied using canonical discriminant analysis (with Mahalanobis distance). Results and Discussion: When computed as a crude prevalence, NC accounted for near a half (45.6%) of the total population at follow-up, amnestic MCI (aMCI) and nonamnestic MCI (naMCI) for 27.9% and 11.6%, respectively, AD for 8%, while 6.9% were subjects with other dementias/psychosis. Risk of AD was increased by previous TIA/stroke (OR=2.77, C.I. 95%: 1.21-6.32). Risk of naMCI was increased by atrial fibrillation (OR=2.07, C.I. 95%: 1.00-4.27). AD scored a Framingham general CVD risk profile lower than NC or MCI (p<0.0001), whereas Framingham stroke risk profile showed a tendency to increase from NC or aMCI to naMCI or AD (p=0.0001). CAIDE risk score was higher in AD than NC (p=0.003). The tendency of LOAD vascular risk score to increase among diagnoses were similar to those of Framingham stroke risk profile (p=0.0002). In particular, when contemporarily stratifying for APOE4, inflammation, undernutrition and High level of Depressive Symptoms, CAIDE risk score strongly associated with risk of AD (OR =1.38, C.I. 95%: 1.01-1.88), Framingham stroke risk profile with naMCI (OR =1.49, C.I. 95%: 1.02-2.13) and LOAD vascular risk score with aMCI (OR =1.05, C.I. 95%: 1.01-1.09). Data of this study, collected in low-educated, rural subjects, showed that the presence of multiple VRFs was associated with cognitive decline or AD. This suggests that a cumulative exposure to even mild effects of some VRFs would lead to cognitive decline and dementia more than the presence of an over vascular disease. Furthermore, it is likely that the effect of VRFs on cognition is not attenuated by old age but easier to detect if taking into account multiple risk factors simultaneously. These data need to be extended and confirmed by prospective analyses conducted on the ZAP incident cohort.