Higher segregation of functional networks in the brain has been associated with better cognitive abilities in aging (Chan, PNAS, 2014) and higher cognitive resilience against Alzheimer's disease (Ewers, Brain, in press). Here, we elucidated for the first time the genetic and environmental (i.e. cardiovascular) determinants of system segregation (SyS) in two large population-based cohorts.We included 16,635 UK Biobank (UKB) participants (45-81y, discovery-sample) and 2,414 Rotterdam-Study participants (52-90y, replication-sample). Resting-state-fMRI SyS was computed as the ratio of between-network to within-network connectivity, where networks (N=55) were defined by independent component analysis. Genome-wide association study (GWAS) of SyS was performed in UKB, controlling for twenty principal components, age, sex, genotype-array and assessment center. For out-of-sample prediction, a polygenic risk score (PRS) of SyS was tested in Rotterdam-Study participants. In both cohorts, we determined the effect of cardiovascular health (adherence to Life's simple 7) on SyS. We estimated age-dependent and -independent effects of SyS on cognition (multi-domain factor score) in both cohorts and, in a subsample of 2,113 UKB participants, on cognitive decline over time. To explore causal effects, Mendelian Randomization (MR) analyses were performed.GWAS of SyS in UKB yielded 659 genome-wide significant single-nucleotide polymorphisms (SNPs; P<5e-08, h2 =0.144, Fig. 1). Nine independent risk loci were detected implicated 59 genes. Lead SNPs with highest likelihood to have deleterious consequences were located near the INPP5A and PLCE1 genes. In Rotterdam-Study participants, PRS explained 1.4% of variance in SyS. We found overall better cardiovascular health to be associated with higher SyS, while higher blood pressure alone was a significant predictor of lower SyS (UKB:βstd =-0.059(0.007), P<0.001; Rotterdam-Study:βstd (SE)=-0.060(0.020), P=0.003; Fig. 2). MR analysis in UKB confirmed that genetically elevated levels of blood pressure were associated with lower SyS (β(CI95%)=-0.003(-0.002,-0.001), P=0.002). We found higher SyS to be associated with better cognition across all ages in UKB (βstd (SE)=-0.031(0.012), P=0.005) and in older, but not younger Rotterdam-Study participants (βstd (SE)=0.038(0.016), P=0.0026, Fig. 3). In MR analysis, genetically elevated levels of SyS were associated with better cognition (β(CI95)=0.104(0.03-0.18), P=0.008).The current study highlights the importance of cardiovascular health for maintaining segregated brain systems what in turn was shown to benefit cognitive functions during aging.
Abstract The functional architecture of the brain is composed of distinct networks, where higher system segregation, i.e. greater differentiation of such functional networks, is associated with better cognitive performance. Aging and many neurological diseases have been associated with reduced system segregation and thus cognitive impairment. The genetic basis and risk factors of system segregation are largely unknown. Here, we present the first genome-wide association study of fMRI-assessed system segregation in 16,635 UK Biobank participants, identifying nine independent genomic loci. The 66 implicated genes were significantly downregulated in brain tissue and upregulated in vascular tissue. Of major vascular risk factors (Life’s Simple 7), blood pressure showed a robust genetic correlation with system segregation. Observational and Mendelian randomization analyses confirmed a unfavourable effect of higher blood pressure on system segregation and of lower system segregation on cognition. Replication analyses in 2,414 Rotterdam Study participants supported these conclusions.
To determine the long-term association of hemoglobin levels and anemia with risk of dementia, and explore underlying substrates on brain MRI in the general population.
Methods
Serum hemoglobin was measured in 12,305 participants without dementia of the population-based Rotterdam Study (mean age 64.6 years, 57.7% women). We determined risk of dementia and Alzheimer disease (AD) (until 2016) in relation to hemoglobin and anemia. Among 5,267 participants without dementia with brain MRI, we assessed hemoglobin in relation to vascular brain disease, structural connectivity, and global cerebral perfusion.
Results
During a mean follow-up of 12.1 years, 1,520 individuals developed dementia, 1,194 of whom had AD. We observed a U-shaped association between hemoglobin levels and dementia (p = 0.005), such that both low and high hemoglobin levels were associated with increased dementia risk (hazard ratio [95% confidence interval (CI)], lowest vs middle quintile 1.29 [1.09–1.52]; highest vs middle quintile 1.20 [1.00–1.44]). Overall prevalence of anemia was 6.1%, and anemia was associated with a 34% increased risk of dementia (95% CI 11%–62%) and 41% (15%–74%) for AD. Among individuals without dementia with brain MRI, similar U-shaped associations were seen of hemoglobin with white matter hyperintensity volume (p = 0.03), and structural connectivity (for mean diffusivity, p < 0.0001), but not with presence of cortical and lacunar infarcts. Cerebral microbleeds were more common with anemia. Hemoglobin levels inversely correlated to cerebral perfusion (p < 0.0001).
Conclusion
Low and high levels of hemoglobin are associated with an increased risk of dementia, including AD, which may relate to differences in white matter integrity and cerebral perfusion.
The relation of grey matter atrophy with cognitive decline is well known. However, global measurements of brain atrophy do not account for all variability in brain morphology. A voxel based study may provide additional information to investigate the relation between grey matter and cognition. 4480 stroke-free and non-demented persons from the population-based Rotterdam Study, underwent brain MRI including high-resolution 3D T1-weighting imaging. Cognition was assessed using an extensive cognitive test battery and grouped into four independent domains (fine motor speed, executive functioning, memory and information processing speed). Voxel-based morphometry (VBM) was performed to investigate the association between local grey matter density and cognitive function, according to the optimized VBM protocol. For each voxel in GM linear regression was performed, with GM density as dependent variable, and age, sex, education and cognitive test score as independent variables. The p-value threshold, for FWE correction was calculated from 10000 permutation tests. The mean age was 64.7 ( ± 10.7) years and 55% subjects were women. We found that memory domain significantly associated with more tissue in both hippocampi and thalamus (Figure 1). Interestingly, that detailed mapping of association showed that executive functioning associated with less and more grey matter tissue in different parts of putamen (Figure 2). Additionally, we found lower of grey matter density in Brocca’s area associated with worse executing functional domain. Fine motor speed domain was related to more tissue in the left and right caudate nucleus, right putamen and with less tissue in the cingulate gyrus. Information processing speed domain has significant association with less tissue in remainder of parietal lobe left. Table 1 provides detailed information about significant clusters distribution in grey matter. Medial (A), lateral (B), posterior (C) view of the hippocampus (green), thalamus (blue). Red areas represent clusters of significant voxels of the association between more grey matter tissue with higher score in the memory domain. Clusters formed based on p-value threshold 3x10-7 estimated from permutations. Axial slice of the brain showing the clusters of significant voxels assosiated with executive functioning domain. Red voxels represent assosiation of more grey matter tissue with higher score of cognitive tests from executive functioning domain, blue with less tissue. In this largest to our knowledge VBM population-based study, we found that grey matter density relates differently to specific cognitive domains. Voxel-based method more sensitive for local grey matter changes, therefore our analysis provides more detail information on the association of cognition with brain morphology. Such high-resolution whole brain maps could help to understand the etiology of diseases related to both cognitive decline and grey matter atrophy.
Cerebral blood flow is an important process for brain functioning and its dysregulation is implicated in multiple neurological disorders. While environmental risk factors have been identified, it remains unclear to what extent the flow is regulated by genetics. Here we performed heritability and genome-wide association analyses of cerebral blood flow in a population-based cohort study. We included 4472 persons free of cortical infarcts who underwent genotyping and phase-contrast magnetic resonance flow imaging (mean age 64.8 ± 10.8 years). The flow rate, cross-sectional area of the vessel, and flow velocity through the vessel were measured in the basilar artery and bilateral carotids. We found that the flow rate of the basilar artery is most heritable (h2 (SE) = 24.1 (9.8), p-value = 0.0056), and this increased over age. The association studies revealed two significant loci for the right carotid artery area (rs12546630, p-value = 2.0 × 10 −8 ) and velocity (rs2971609, p-value = 1.4 × 10 −8 ), with the latter showing a concordant effect in an independent sample (N = 1350, p-value = 0.057, meta-analyzed p-value = 2.5 × 10 −9 ). These loci were also associated with other cerebral blood flow parameters below genome-wide significance, and rs2971609 lies in a known migraine locus. These findings establish that cerebral blood flow is under genetic control with potential relevance for neurological diseases.
Sleep problems increase with ageing. Increasing evidence suggests that sleep problems are not only a consequence of age-related processes, but may independently contribute to developing vascular or neurodegenerative brain disease. Yet, it remains unclear what mechanisms underlie the impact sleep problems may have on brain health in the general middle-aged and elderly population. Here, we studied sleep's relation to brain functioning in 621 participants (median age 62 years, 55% women) from the population-based Rotterdam Study. We investigated cross-sectional associations of polysomnographic and subjectively measured aspects of sleep with intrinsic neural activity measured with resting-state functional magnetic resonance imaging on a different day. We investigated both functional connectivity between regions and brain activity (blood-oxygen-level-dependent signal amplitude) within regions, hierarchically towards smaller topographical levels. We found that longer polysomnographic total sleep time is associated with lower blood-oxygen-level-dependent signal amplitude in (pre)frontal regions. No objective or subjective sleep parameters were associated with functional connectivity between or within resting-state networks. The findings may indicate a pathway through which sleep, in a ‘real-life’ population setting, impacts brain activity or regional brain activity determines to
Cerebral hypoperfusion is associated with an increased risk of dementia, which may reflect a direct consequence of neuronal hypoxia. Haemoglobin is the sole transport molecule for tissue oxygenation, but the relation of haemoglobin with cerebral perfusion and dementia is uncertain. Between 1989 and 1993, we determined serum haemoglobin levels in 6309 non-demented participants of the population-based Rotterdam Study (mean age 67.9 years, 59.7% female). We determined the risk of dementia and Alzheimer's disease (until 2014) in relation to haemoglobin, as well as anaemia, using a Cox regression model. Analyses were adjusted for age, sex, educational attainment, cardiovascular risk factors, dietary intake, relevant medications, and APOE genotype. Finally, we determined the association between haemoglobin levels and cerebral perfusion (mL/100mL/min) in a subset of participants who underwent 2D phase contrast MRI. During a median follow-up of 14.6 years (interquartile range 7.8–20.7), 1232 individuals developed dementia, of whom 976 had Alzheimer's disease. We observed a quadratic association between haemoglobin and dementia (p<0.0001), such that both low and high haemoglobin levels were associated with increased risk of dementia (adjusted HR, 95% confidence interval– lowest versus middle quintile 1.19, 1.00–1.42; highest versus middle quintile 1.30, 1.06–1.59). Overall prevalence of anaemia in the population was 2.4%. Anaemia was associated with a 1.66-fold increased risk of dementia (aHR 1.66, 1.16–2.39). Estimates were slightly higher for Alzheimer's disease only, and similar across the 25-year follow-up period. In the subset of 1033 individuals with brain MRI, haemoglobin levels were linearly associated with cerebral perfusion (change in perfusion,95%CI, per mmol/L decrease: 4.0, 3.1–4.9; p<0.0001). In the general population, both low and high levels of haemoglobin are associated with increased risk of dementia, including Alzheimer's disease. Anaemia is associated with a 66% increased risk of dementia. The link between haemoglobin and cerebral blood flow warrants further investigation.
Background The role of subtle disturbances of brain perfusion in the risk of transient ischemic attack ( TIA) or ischemic stroke remains unknown. We examined the association between global brain perfusion and risk of TIA and ischemic stroke in the general population. Methods and Results Between 2005 and 2015, 5289 stroke‐free participants (mean age, 64.3 years; 55.6% women) from the Rotterdam Study underwent phase‐contrast brain magnetic resonance imaging at baseline to assess global brain perfusion. These participants were followed for incident TIA or ischemic stroke until January 1, 2016. We investigated associations between global brain perfusion (mL of blood flow/100 mL of brain/min) and risk of TIA and ischemic stroke using Cox regression models with adjustment for age, sex, and cardiovascular risk factors. Additionally, we investigated whether associations were modified by retinal vessel calibers, small and large vessel disease, blood pressure, and heart rate. During a median follow‐up of 7.2 years (36 103 person‐years), 137 participants suffered a TIA and another 108 an ischemic stroke. We found that lower global brain perfusion was associated with a higher risk of TIA , but not with the risk of ischemic stroke (adjusted hazard ratio, 95% CI, per standard deviation decrease of global brain perfusion: 1.29, 1.07–1.55 for TIA and adjusted hazard ratio of 1.06, 0.87–1.30 for ischemic stroke). Across strata of wider arteriolar retinal calibers, lower brain perfusion was more prominently associated with TIA , but not with ischemic stroke. Conclusions In a community‐dwelling population, impaired global brain perfusion increased the risk of TIA , but not of ischemic stroke.
Abstract Introduction Head motion during magnetic resonance imaging is heritable. Further, it shares phenotypical and genetic variance with body mass index (BMI) and impulsivity. Yet, to what extent this trait is related to single genetic variants and physiological or behavioral features is unknown. We investigated the genetic basis of head motion in a meta-analysis of genome-wide association studies. Further, we tested whether physiological or psychological measures, such as respiratory rate or impulsivity, mediated the relationship between BMI and head motion. Methods We conducted a genome-wide association meta-analysis for mean and maximal framewise head displacement (FD) in seven population neuroimaging cohorts (UK Biobank, LIFE-Adult, Rotterdam Study cohort 1-3, Austrian Stroke Prevention Family Study, Study of Health in Pomerania; total N = 35.109). We performed a pre-registered analysis to test whether respiratory rate, respiratory volume, self-reported impulsivity and heart rate mediated the relationship between BMI and mean FD in LIFE-Adult. Results No variant reached genome-wide significance for neither mean nor maximal FD. Neither physiological nor psychological measures mediated the relationship between BMI and head motion. Conclusion Based on these findings from a large meta-GWAS and pre-registered follow-up study, we conclude that the previously reported genetic correlation between BMI and head motion relies on polygenic variation, and that neither psychological nor simple physiological parameters explain a substantial amount of variance in the association of BMI and head motion. Future imaging studies should thus rigorously control for head motion at acquisition and during preprocessing.
