Association Of Cardiovascular Risk Score With Cognitive Decline And Dementia Risk: A Population-based Cohort Study

Objective:The results of the association of cardiovascular risk burden score with structural brain changes and cognitive phenotypes from cognitive decline to mild cognitive impairment（MCI）and dementia are inconsistent.Moreover,the mechanisms underline this association remain unclear.We aimed to investigate the association of cardiovascular risk burden score with regional brain volumes,cognitive decline,MCI and dementia,and further to explore the relationship between cardiovascular risk burden score and brain pathologies.Methods:Within the Memory and Aging Project,an ongoing longitudinal study,a total of 2,155 participants were recruited and followed for up to 21 years（median duration=5.2 years）.Information of sociodemographic characteristics,lifestyle,health status,and Apolipoproteinε4 alleles（APOEε4）was collected at study entry.All participants underwent comprehensive assessment of cognitive function and ascertainment of MCI,dementia and Alzheimer’s disease at study entry and annual follow-ups.A subsample（n=401）of the study population underwent magnetic resonance imaging（MRI）scans.During follow-up,842 died and underwent autopsies for neuropathological evaluation.At study entry,Framingham General Cardiovascular Risk Score（FGCRS）was calculated based on age,sex,smoking status,blood lipid level,systolic blood pressure,medication for hypertension and diabetes,and was further categorized into tertiles（as highest,middle and lowest）.Of the total population,115 persons with prevalent dementia,311 with missing FGCRS at study entry and 282with missing data on cognitive function during the follow-up were excluded,leaving1,588 dementia-free participants in the study sample,of whom 1,203 participants with cognitive intact and 385 participants with MCI.During the follow-up,503 participants developed MCI,378 participants developed dementia（including 343 with Alzheimer’s disease）.In data analyses,Linear mixed-effects models were used to estimateβ_mix coefficient[95%confidence intervals（CIs）]for the associations between FGCRS/brain volumes and annual change in global cognitive function and five cognitive domains（including episodic memory,semantic memory,working memory,visuospatial ability,and perceptual speed）.An interaction term between FGCRS and APOEε4 status was also included in the mixed-effects models;Linear regressions were used to estimateβcoefficient（95%CIs）for the the relationship between FGCRS and brain volumes;Cox regression models were used to estimate hazard ratios（HRs）（95%CIs）for MCI,dementia and Alzheimer’s disease associated with FGCRS.We also analyzed the effect of the interaction of APOEε4 status and FGCRS on the onset of MCI and dementia;Logistic regressions were used to estimate odds ratios（ORs）（95%CIs）for the relationship between FGCRS and burdens of brain pathologies.Results:（1）Assocciation of cardiovascular risk burden score with cognitive function and structural brain volumes:In the multi-adjusted mixed-effects model,the highest FGCRS was associated with faster rate of decline in global cognition(β_mix:-0.019,95%CI:-0.035 to-0.003),episodic memory(β_mix:-0.023,95%CI:-0.041 to-0.004),working memory(β_mix:-0.021,95%CI:-0.035 to-0.007),and perceptual speed(β_mix:-0.027,95%CI:-0.042 to-0.011).We found no statistical interaction between FGCRS and APOEε4 with respect to cognitive decline.In MRI data analyses,higher FGCRS（as continues variable）was significantly related to smaller volumes of the hippocampus（β:-0.033,95%CI:-0.054 to-0.012）,gray matter（β:-2.492,95%CI:-3.742 to-1.241）,and total brain（β:-2.620,95%CI:-4.167 to-1.074）,and greater volume of white matter hyperintensities（WMH）（β:0.011,95%CI:0.000 to 0.023）after adjustment for potential confounders.Moreover,lower volumes of hippocampus,total gray matter,and total brain as well as higher volume of WMH were associated with accelerated decline in global cognitive function and five cognitive domains during the follow-up（all P values<0.05）.（2）Assocciation between cardiovascular risk burden score and risk of MCI:FGCRS was dose-dependently associated with MCI risk（HR:1.05,95%CI:1.03 to 1.08）after adjustment for education,body mass index（BMI）,stroke,heart disease,alcohol consumption,physical activity engagement and APOEε4.The multi-adjusted HR（95%CI）of MCI related to the highest FGCRS were 1.61（1.27 to 2.05）compared with those with the lowest FGCRS.We did not find a significant interaction between cardiovascular risk burden score and APOEε4 status on risk of MCI.（3）Association of cardiovascular risk burden score with risk of dementia and brain pathologies: The multi-adjusted HRs（95% CIs）of FGCRS（as continuous variable）were 1.06（1.02 to 1.08）for dementia and 1.06（1.03 to 1.10）for Alzheimer’s disease.Compared to those with the lowest FGCRS,the multi-adjusted HRs（95% CIs）were 1.67（1.26 to 2.23）for dementia and 1.72（1.27 to 2.33）for Alzheimer’s disease.We also did not find a significant interaction between cardiovascular risk burden score and APOE ε4 status on dementia/ Alzheimer’s disease.In brain pathological data analysis,FGCRS was dose-dependently associated with higher global Alzheimer’s disease pathology（OR: 1.05,95% CI: 1.00 to 1.11）,gross chronic cerebral infarctions（OR: 1.06,95% CI: 1.00 to 1.12）,and cerebral atherosclerosis（OR: 1.09,95% CI: 1.01 to 1.16）.Conclusions: Higher FGCRS was associated with accelerated decline in global cognition and specific cognitive domains（episodic memory,working memory and perceptual speed）,an increased risk of MCI,dementia and Alzheimer’s disease.Imaging and postmortem neuropathological evaluation showed that cardiovascular risk burden score was related to both neurodegeneration and vascular lesions in the brain,which may underlie the association of vascular risk burden with cognitive decline and dementia.