Differential Cognitive Decline in Alzheimer's Disease Is Predicted by Changes in Ventricular Size but Moderated by Apolipoprotein E and Pulse Pressure

Genetic architectures of cerebral ventricles and their overlap with neuropsychiatric traits

The cerebral ventricles are recognized as windows into brain development and disease, yet their genetic architectures, underlying neural mechanisms and utility in maintaining brain health remain elusive. Here we aggregated genetic and neuroimaging data from 61,974 participants (age range, 9 to 98 years) in five cohorts to elucidate the genetic basis of ventricular morphology and examined their overlap with neuropsychiatric traits. Genome-wide association analysis in a discovery sample of 31,880 individuals identified 62 unique loci and 785 candidate genes associated with ventricular morphology. We replicated over 80% of loci in a well-matched cohort of lateral ventricular volume. Gene set analysis revealed enrichment of ventricular-trait-associated genes in biological processes and disease pathogenesis during both early brain development and degeneration. We explored the age-dependent genetic associations in cohorts of different age groups to investigate the possible roles of ventricular-trait-associated loci in neurodevelopmental and neurodegenerative processes. We describe the genetic overlap between ventricular and neuropsychiatric traits through comprehensive integrative approaches under correlative and causal assumptions. We propose the volume of the inferior lateral ventricles as a heritable endophenotype to predict the risk of Alzheimer's disease, which might be a consequence of prodromal Alzheimer's disease. Our study provides an advance in understanding the genetics of the cerebral ventricles and demonstrates the potential utility of ventricular measurements in tracking brain disorders and maintaining brain health across the lifespan.

Vascular Risk Predicts Plasma Amyloid β 42/40 Through Cerebral Amyloid Burden in Apolipoprotein E ε4 Carriers

BACKGROUND

Understanding the neurobiological underpinnings between established multimodal dementia risk factors and noninvasive blood-based biomarkers may lead to greater precision and earlier identification of older adults at risk of accelerated decline and dementia. We examined whether key vascular and genetic risk impact the association between cerebral amyloid burden and plasma aβ (amyloid β) 42/40 in nondemented older adults.

METHODS

We used nondemented older adults from the UCD-ADRC (University of California, Davis-Alzheimer's Disease Research Center) study (n=96) and Alzheimer's Disease Neuroimaging Initiative (n=104). Alzheimer's Disease Neuroimaging Initiative was examined as confirmatory study cohort. We followed a cross-sectional design and examined linear regression followed by mediation analyses. Vascular risk score was obtained as the sum of hypertension, diabetes, hyperlipidemia, coronary artery disease, and cerebrovascular disease. Apolipoprotein E (APOE) ε4+ risk was genotyped, and plasma aβ42 and aβ40 were assayed. Cerebral amyloid burden was quantified using Florbetapir-PET scans. Baseline age was included as a covariate in all models.

RESULTS

Vascular risk significantly predicted cerebral amyloid burden in Alzheimer's Disease Neuroimaging Initiative but not in the UCD-ADRC cohort. Cerebral amyloid burden was associated with plasma aβ 42/40 in both cohorts. Higher vascular risk increased cerebral amyloid burden was indirectly associated with reduced plasma aβ 42/40 in Alzheimer's Disease Neuroimaging Initiative but not in UCD-ADRC cohort. However, when stratified by APOE ε4+ risk, we consistently observed this indirect relationship only in APOE ε4+ carriers across both cohorts.

CONCLUSIONS

Vascular risk is indirectly associated with the level of plasma aβ 42/40 via cerebral amyloid burden only in APOE ε4+ carriers. Nondemented older adults with genetic vulnerability to dementia and accelerated decline may benefit from careful monitoring of vascular risk factors directly associated with cerebral amyloid burden and indirectly with plasma aβ 42/40.

Bridging Integrator 1 (BIN1, rs6733839) and Sex Are Moderators of Vascular Health Predictions of Memory Aging Trajectories

Background: A promising risk loci for sporadic Alzheimer’s disease (AD), Bridging Integrator 1 (BIN1), is thought to operate through the tau pathology pathway. Objective: We examine BIN1 risk for a moderating role with vascular health (pulse pressure; PP) and sex in predictions of episodic memory trajectories in asymptomatic aging adults. Methods: The sample included 623 participants (Baseline Mean age = 70.1; 66.8% female) covering a 44-year longitudinal band (53–97 years). With an established memory latent variable arrayed as individualized trajectories, we applied Mplus 8.5 to determine the best fitting longitudinal growth model. Main analyses were conducted in three sequential phases to investigate: 1) memory trajectory prediction by PP, 2) moderation by BIN1 genetic risk, and 3) stratification by sex. Results: We first confirmed that good vascular health (lower PP) was associated with higher memory level and shallower decline and males were more severely affected by worsening PP in both memory performance and longitudinal decline. Second, the PP prediction of memory trajectories was significant for BIN1 C/C and C/T carriers but not for persons with the highest AD risk (T/T homozygotes). Third, when further stratified by sex, the BIN1 moderation of memory prediction by PP was selective for females. Conclusion: We observed a novel interaction whereby BIN1 (linked with tauopathy in AD) and sex sequentially moderated a benchmark PP prediction of differential memory decline in asymptomatic aging. This multi-modal biomarker interaction approach, disaggregated by sex, can be an effective method for enhancing precision of AD genetic risk assessment.