ASSOCIATION BETWEEN HYPERTENSION AND ALZHEIMER’S DISEASE AND RELATED CAUSES OF DEMENTIA: A PATHWAYS ANALYSIS IN THE MEMENTO COHORT

Hypertension is a risk factor for Alzheimer's Disease and related causes of dementia (ADRD). However, the underlying pathophysiological mechanisms are still to be elucidated. On one hand, hypertension is associated with an increase in white matter lesions (WML) load, considered as a risk factor for ADRD, and on the other hand, hypertension is also associated with worsening of amyloid and tau markers, two main hallmarks of AD. Yet, no study has investigated pathways linking hypertension to cognitive performances, by integrating information on intermediate neuroimaging and CSF biomarkers. Through the MEMENTO study, a French nationwide clinic-based cohort of persons with either isolated cognitive complaints or mild cognitive impairment, 2323 participants were enrolled between 2011 and 2014. Baseline data collection included demographics, clinical examinations, neuropsychological testing, blood sampling, cerebral MRI for all participants. In subsamples, cerebral 18F-fluorodeoxyglucose positron emission tomography (FDG-PET)(N=1379), lumbar puncture (N=413), were performed. To take simultaneously into account hypertension, ADRD biomarkers and potential confounders in relation to cognition, we performed a path analysis using structural equation models (SEM) to test an hypothetical causal model which included both direct and indirect effects of hypertension on cognitive performances (Figure). Mean age of participants was 71 year, 62% were women, 40.2% scored 0 at Clinical Dementia Rating scale and 60% were hypertensive (>140/90 or treated) (17.4% were uncontrolled (>140/90 and treated)). From the SEM, uncontrolled participants for hypertension had significantly lower cognition compared with those free of hypertension (mean overall effect relative to one standard deviation of cognition: beta=−0.216, p=0.001). About half of the effect was indirect as mediated by WML load (beta=−0.018, p=0.021), and neurodegeneration (defined in Figure) (beta=−0.084, p=0.024) but not by CSF biomarkers (Aβ42/Aβ40 ratio and tau). After adjustment for these mediators, the direct path from uncontrolled hypertension to cognition was no more significant (beta=−0.103, p=0.154).