Importance: Poor cardiovascular health is an established risk factor for dementia, but little is known about its association with brain physiology in older adults. Objective: To examine the association of cardiovascular risk factors, measured repeatedly during a 20-year period, with cerebral perfusion at older ages. Design, Setting, and Participants: In this longitudinal cohort study, individuals were selected from the Whitehall II Imaging Substudy. Participants were included if they had no clinical diagnosis of dementia, had no gross brain structural abnormalities on magnetic resonance imaging scans, and had received pseudocontinuous arterial spin labeling magnetic resonance imaging. Cardiovascular risk was measured at 5-year intervals across 5 phases from September 1991 to October 2013. Arterial spin labeling scans were acquired between April 2014 and December 2014. Data analysis was performed from June 2016 to September 2018. Exposures: Framingham Risk Score (FRS) for cardiovascular disease, comprising age, sex, high-density lipoprotein cholesterol level, total cholesterol level, systolic blood pressure, use of antihypertensive medications, cigarette smoking, and diabetes, was assessed at 5 visits. Main Outcomes and Measures: Cerebral blood flow (CBF; in milliliters per 100 g of tissue per minute) was quantified with pseudocontinuous arterial spin labeling magnetic resonance imaging. Results: Of 116 adult participants, 99 (85.3%) were men. At the first examination, mean (SD) age was 47.1 (5.0) years; at the last examination, mean (SD) age was 67.4 (4.9) years. Mean (SD) age at MRI scan was 69.3 (5.0) years. Log-FRS increased with time (B = 0.058; 95% CI, 0.044 to 0.072; P < .001). Higher cumulative FRS over the 20-year period (measured as the integral of the rate of change of log-FRS) was associated with lower gray matter CBF (B = -0.513; 95% CI -0.802 to -0.224; P < .001) after adjustment for age, sex, education, socioeconomic status, cognitive status, arterial transit time, use of statins, and weekly alcohol consumption. Voxelwise analyses revealed that this association was significant in 39.6% of gray matter regions, including the posterior cingulate, precuneus, lateral parietal cortex, occipital cortex, hippocampi, and parahippocampal gyrus. The strength of the association of higher log-FRS with lower CBF decreased progressively from the first examination (R2 = 0.253; B = -10.816; 99% CI -18.375 to -3.257; P < .001) to the last (R2 = 0.188; B = -7.139; 99% CI -14.861 to 0.582; P = .02), such that the most recent FRS measurement at mean (SD) age 67.4 (4.9) years was not significantly associated with CBF with a Bonferroni-corrected P < .01 . Conclusions and Relevance: Cardiovascular risk in midlife was significantly associated with lower gray matter perfusion at older ages, but this association was not significant for cardiovascular risk in later life. This finding could inform the timing of cardiovascular interventions so as to be optimally effective.
JAMA Netw Open