Atherosclerotic Indices for the Prediction of Cognitive Impairment in a Middle‐Aged to Elderly General Population: Shimanami Health Promoting Program Study

Vascular contribution to the pathogenesis of cognitive decline has been demonstrated in vascular dementia as well as dementia with Alzheimer's pathology.1 Several biomarkers for atherosclerosis are being used clinically to assess aspects of atherosclerosis. Among them, pulse wave velocity (PWV),2 carotid arterial intima-media thickness (IMT),3 and the ankle-brachial index (ABI) 4, 5 are associated with cognitive impairment or dementia in the general population. The augmentation index (AI), the ratio of the reflected pressure wave, is another index of atherosclerosis affecting central pulse pressure and has been postulated as an index for pulsatile load to brain circulation,6 indicating that AI could also be an index for cognitive impairment. One study has compared two indices for predicting lower Mini-Mental State Examination (MMSE) scores (<24) in the general population and demonstrated that ABI but not PWV between the brachium and ankle (baPWV) is an independent risk factor for cognitive impairment;5 no other indices were compared. The present study evaluated four indices for atherosclerosis (baPWV, carotid IMT, radial AI, and ABI) to determine the most-effective marker to predict the presence of cognitive impairment in a middle-aged to elderly general population. One thousand two hundred eighty-eight consecutive participants in the “Anti-aging Doc” medical examination program at Ehime University Hospital (509 men; mean age 66.2 ± 9.0) were analyzed. Cognitive function was evaluated using a touch-panel computerized (TPC) test battery consisting of four tests: immediate recall, delayed recall, arithmetic, and visuospatial.7 A score less than 13 out of 15 has been shown to predict the presence of Alzheimer's disease with 96% sensitivity and 86% specificity.7 IMT of the far wall of the common carotid artery was measured using an SSD-3500SV (Aloka Co., Ltd., Tokyo, Japan) with a 7.5-MHz probe as previously described.8 Ankle and brachial blood pressures (BPs) were measured on both sides using a volume-plethysmographic apparatus (form PWV/ABI; Omron Healthcare Co. Ltd., Kyoto, Japan) with participant in the supine position. ABI was calculated as lower ankle BP divided by higher brachial BP. baPWV was measured simultaneously as PWV/ABI and calculated from the time interval between the wave fronts of the brachial and ankle waveforms and the path length from the brachium to the ankle.8 Radial AI was measured in the left radial artery using an automated tonometric method (HEM-9000AI; Omron Healthcare, Kyoto, Japan) with participants in a seated position after at least 5 minutes of rest. Right brachial BP was measured simultaneously using an oscillometric device incorporated into the HEM-9000AI.9 Radial AI 75 was obtained as correction for a heart rate of 75 beats per minute. Carotid IMT, baPWV, and radial AI75 showed significant negative associations with TPC cognitive test scores (Figure 1), although after adjustment for age and sex, only baPWV (β = –0.08, P = .01) remained statistically significant. baPWV was an independent determinant (β = –0.07, P = .02) for cognitive test scores in stepwise regression analysis with other possible confounding parameters. Receiver operating characteristic analysis revealed that the cutoff value of baPWV (1,460 cm/s) had a sensitivity of 81.2% and a specificity of 41.7% for the presence of a TPC score less than 13. Logistic regression analysis for a cut-off value of a cognitive test score of less than 13 showed that baPWV was an independent risk factor (100-cm/s increase in baPWV: odds ratio (OR) = 1.10, 95% confidence interval (CI) = 1.01–1.19, P = .03; 1,460-cm/s increase in baPWV: OR = 1.91, 95% CI = 1.03–3.72, P = .04) after adjustment for possible confounding parameters, including age, sex, body mass index, systolic and diastolic BP, total cholesterol, high-density lipoprotein cholesterol, triglycerides, fasting glucose, smoking status, use of antihypertensive medication, antidyslipidemic medication, and antidiabetic medication. In the present study, of four atherosclerotic indices, baPWV had the strongest association with the presence of cognitive impairment in middle-aged to elderly population, indicating that arterial stiffness may more directly relate to cognitive decline than other indices. In a previous study,5 an ABI of less than 0.9 was more closely related to lower MMSE score than baPWV tertile. The differences between studies may be because of the background characteristics of the studied population. The prevalence of participants with an ABI less than 0.9 was 10.6% (41/388 participants) in their study, whereas it was only 0.5% (7/1,288 participants) in the present population. These findings indicate that modality-specific assessment is necessary for evaluating the vascular risk for cognitive impairment. Vascular risk assessment using PWV could be appropriate for cognitive impairment in a low-risk general elderly population. Conflict of Interest: No authors have and conflicts of interest in the study. The study was supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan (22659143 and 23390188) and a Research Fund from Suzuken Memorial Foundation. Author Contributions: Kohara K, Tabara Y: Study design. Tabara Y, Igase M: Data collection. Kohara K: Manuscript preparation. Miki T: Supervision. Sponsor's Role: No sponsors were involved in the study.