Impaired Cerebral Hemodynamics and Frailty in Patients with Cardiovascular Disease

Impaired Cerebral Hemodynamics and Frailty in Patients with Cardiovascular Disease Abstract Background Recent studies suggest that impaired cerebrovascular reactivity (CVR), a marker of cerebral microvascular damage, is associated with a higher risk of stroke, cognitive decline, and mortality. We tested whether abnormal cerebrovascular status is associated with late-life frailty among men with pre-existing cardiovascular disease. Methods A subset of 327 men (mean age at baseline 56.7 ± 6.5 years) who previously participated in the Bezafibrate Infarction Prevention (BIP) trial (1990–1997) and then in the BIP Neurocognitive Study underwent a neurovascular evaluation 14.6 ± 1.9 years after baseline (T1) and were evaluated for frailty 19.9 ± 1.0 years after baseline (T2). CVR was measured at T1 using the breath-holding index and carotid large-vessel disease using ultrasound. Frailty status was measured at T2 according to the physical phenotype developed by Fried. Patients were categorized into CVR tertiles with cutoff points at ≤0.57, 0.58–0.94, and ≥0.95 and also as normal or impaired (<0.69) CVR. We assessed the change in the odds of being in the advanced rank of frailty status (normal, prefrail, and frail) using ordered logistic regression. Results After adjustment, the estimated OR (95% confidence intervals) for increasing frailty in the lower tertile was 1.94 (1.09–3.46) and in the middle tertile 1.24 (0.70–2.19), compared with the higher CVR tertile. The estimated OR for increasing frailty for patients with impaired vs. normal CVR was 1.76 (1.11–2.80). Conclusions These findings provide support that cerebral microvascular dysfunction among patients with pre-existing cardiovascular disease is related to prefrailty and frailty and suggest an added value of assessing the cerebral vascular functional status for identifying patients at-risk of developing frailty. Cerebrovascular reactivity, Cerebral hemodynamics, Cardiovascular disease, Frailty, Small vessel disease Frailty is a geriatric syndrome used to define older adults with impaired resistance to stressors due to a decline in physiological reserve. Frailty is characterized by a nonspecific state of vulnerability, and multisystem dysregulation, leading to a higher risk for cognitive decline, disability, and death (1). Prevalence of frailty is estimated at 11 per cent among adults aged 65 and older and reaching over 26 per cent in those aged ≥85 years (2) and higher in people with cardiovascular disease (CVD) (3). Some studies suggest that frailty is a risk factor for CVD (4), whereas others suggest that CVD precedes frailty (5). Recent evidence suggests that in comparison with nonfrail older adults, those who exhibit the phenotype of frailty have more extensive vascular damage (6, 7). Indeed, cerebrovascular disease is present among both cognitively impaired and frail persons, since these conditions share pathophysiological mechanisms, via cerebrovascular damage (7–9). It is speculated that cerebral microvascular dysfunction occurs due to arteriosclerosis and decrease of elasticity of the cerebral small-vessel walls. There is growing evidence that cerebral microvascular dysfunction is associated with accelerated cognitive decline (10, 11), stroke (12), and mortality (13). Nonetheless, very few studies (6–8, 14) have assessed carotid intima-media thickness (cIMT), a marker of cerebral macrovascular disease (15), and impaired cerebrovascular reactivity (CVR), a marker of cerebral microvascular dysfunction (16), in frail older adults. We used data from the Bezafibrate Infarction Prevention (BIP) Neurocognitive study to address the hypothesis that abnormal cerebrovascular status (impaired CVR and abnormal cIMT) may be associated with late-life frailty among men with pre-existing CVD. Examining this relationship may be important both for studying pathways through which vascular disease and frailty are interrelated and for identifying high-risk patients prone to frailty. Material and Methods Patients The population for the current study includes patients from eight central hospitals (n = 1,232) who reside in the central region of Israel and who previously participated in the BIP clinical trial of lipid modification and then in the BIP Neurocognitive Study. The study design and procedures of the BIP trial have been previously described in detail (17). In brief, the BIP study was a placebo-controlled randomized clinical trial investigating the efficacy of bezafibrate 400 mg daily, a fibric derivative, in secondary prevention among patients with established stable coronary heart disease. The lipid profile of patients at inclusion was as follows: serum total cholesterol 180 to 250 mg/dL, low-density lipoprotein–cholesterol ≤ 180 mg/dL (≤160 mg/dL for people <50 years), high-density lipoprotein–cholesterol ≤ 45 mg/dL, and triglycerides ≤ 300 mg/dL. Other exclusion criteria were insulin-dependent diabetes, hepatic or renal failure, and disabling stroke. Patients included in the BIP Neurocognitive Study have undergone two follow-up evaluations (T1 and T2). The first follow-up evaluation (T1, n = 558) was performed during 2004–2008, an average of 14.6 ± 1.9 years after recruitment, assessing neurovascular and cognitive function. Patients were re-examined during 2011–2013 (T2, n = 351), 19.9 ± 1.0 years after recruitment, assessing frailty and re-assessing cognitive function (Figure 1). The mean time interval between the T1 and T2 assessments was 4.8 ± 1.3 years. The second late-life evaluation took place at a central research center (the Sagol Neuroscience Center), or if patients were unable or unwilling to attend the research center, the assessments were completed at their residence. The study was approved by the local institutional review board, and informed consent was obtained from all patients. Baseline Measurements Methods for assessment of vascular risk factors at baseline of the BIP study are described elsewhere (18). Blood samples were drawn from each study patient at baseline of the BIP trial (1990–1992). Samples were collected after at least 12 hours of fasting, with the use of standardized equipment and procedures and transferred to a central study laboratory. Education, occupation, place of birth, and comorbidity were collected through a baseline questionnaire. We included data on birth place, which is an important key indicator of socioeconomic status disparities and possibly of genetic predisposition (19). Neurovascular Evaluation (T1) CVR was evaluated by transcranial Doppler (TCD) using the breath-holding index (BHI), previously described by Muller and co-workers (20). All exams were performed by a qualified technician using a Trans-Link 9900 TCD device (Rimed, Raanana, Israel) equipped with a 2 MHz pulsed handheld probe. Before proceeding to the definitive recording, patients were trained to perform the procedure correctly. After normal breathing of room air for approximately 4 minutes, patients were asked to hold their breath for 30 seconds or as long as possible, after a normal inspiration. Patients unable to hold their breath for at least 15 seconds were excluded. During the maneuver, the middle cerebral artery (MCA) mean blood flow velocity was recorded continuously. The length time of apnea was measured in seconds. Mean flow velocity at rest was obtained by continuous recording of a 1 minute period of normal breathing. As the CVR measure, the BHI in the MCA was calculated in a standardized manner, as the percent increase in MCA mean blood velocity recorded by breath-holding divided by seconds of breath-holding ([Vbh − Vr/Vr] × 100 × s−1), where Vbh is MCA mean blood velocity at the end of breath-holding, Vr is the MCA mean blood velocity at rest, and s-1 is per second of breath-holding. Carotid ultrasound examination was performed to identify the presence of carotid plaques, following a standard protocol (21), using a HDI 5000 SonoCT system device (Philips, Eindhoven, The Netherlands) with a linear array multifrequency transducer (4 to 7 MHz). The presence of bilateral carotid plaque was regarded as evidence of carotid large-vessel disease. In addition, cIMT was measured at the far wall of both common carotid arteries at 1.0 cm proximal to the carotid bifurcation in length of 1.0 cm using high resolution B-mode ultrasound. Assessment followed a standard protocol (22) that included acquiring images of both carotid arteries and measurement of the distance between the media-adventitia interface and the lumen-intima interface, employing automatic edge detection (METRIS, France). Frailty Evaluation (T2) Frailty was assessed using the phenotype developed by Fried and co-workers (1). It comprises five individual criteria: weight loss, weakness (grip strength), slowness (walking speed), self-reported exhaustion, and low physical activity. Weight loss was categorized as positive if the patient reported loss greater than or equal to 4.0 kg unintentionally in the prior year. Grip strength was defined as isometric dominant handgrip strength assessed using a hydraulic Jamar dynamometer (Sammons & Preston, Bolingbrook, IL). Low grip strength was denoted as ≤29 kg [for patients with body mass index (BMI) ≤24 kg/m2], ≤30 kg (for BMI 24.1–28.0 kg/m2), ≤32 kg (for BMI >28.0 kg/m2). The test was carried out twice and the maximum score was used. Gait speed was measured by gait time in seconds using a 5 m timed walk test. Usual gait speed of less than 1 m/second identifies persons at high risk of health-related outcomes in well-functioning older people (23). We used height-adjusted time as the cutoff. Slowness was denoted as ≥6 seconds (for height ≤173 cm) and ≥ 5 seconds (for height >173 cm) (1). The test instructions were as follows: “On the word ‘go’ you will start walking at your regular pace to the line on the floor.” Exhaustion was defined by two self-report questions (“I felt everything I did was an effort” and “I could not get going”) from the Center for Epidemiological Studies Depression Scale (24). Patients answering “moderate amount of time (3–4 days) or most of the time (5–7 days) last week to these self-report questions were categorized were categorized as frail by the exhaustion criterion. Physical activity was assessed by the Physical Activity Scale for the Elderly scale (25). Patients who scored in the lowest quintile were categorized as positive for the low physical activity criterion. Frailty was defined as the presence of three or more of these five criteria, and those with one or two of these criteria were considered prefrail. Additional Assessments In both evaluations (T1 and T2), data were collected systematically regarding new comorbidities and hospitalizations, medication use, smoking status, physical activity, and anthropometric measurements. In addition, systolic blood pressure (SBP), diastolic blood pressure (DBP), and BMI were measured. Incident stroke during follow-up was assessed by reviewing records from a hospital or emergency department discharge, a primary care physician, or a neurologist. Depressive symptoms were assessed by the short version Geriatric Depression Scale (GDS) using a cutoff of ≥5 (26). Patients completed the NeuroTrax computerized cognitive testing (NeuroTrax Corp., Bellaire, 175 TX). A description of the tests included has been published elsewhere (27). Dementia and incident stroke during follow-up were determined by an adjudication committee composed of three investigators, two of which are experienced board certified neurologists. Dementia was determined based on the sum of cognitive evaluation, clinical interview, and data collected and in accordance with the Diagnostic and Statistical Manual of Mental Disorders 4th Edition (DSM-IV) criteria and stroke according to World Health Organization criteria. For the purpose of the current analysis, patients with dementia before T1 evaluation were excluded (n = 45), in order to assure compliance with and reliable measurement of CVR. Statistical Analysis We have examined the differences in the distribution between right and left CVR and cIMT. Since no differences were found, we have decided to use the average values. Patients were categorized into normal (≥0.69) or impaired (<0.69) CVR based on the mean BHI of both MCAs according to the previously established standard parameters (28) and also into tertiles of CVR, with cut-off points at ≤0.57, 0.58–0.94, and ≥0.95. Patients were categorized into tertiles of mean (both sides) cIMT with cut-off points at ≤0.87, 0.88–1.02, and ≥1.03 mm and also into categories of cIMT ≥ 0.93 mm and/or bilateral carotid plaques vs. others (cIMT < 0.93 and without bilateral carotid plaques) according to the previously established standard parameters (15). The clinical characteristics of the patients are expressed as percentages and as mean ± SD, except when the distribution was strongly skewed, in which case, median and interquartile range are provided. The chi-square test for trend was used to determine the significance of differences between CVR tertiles in the distribution of categorical data. The clinical characteristics of the continuous variables were compared between CVR tertiles by analysis of variance or Kruskal–Wallis test for variables not normally distributed. We used ordered logistic regression, due to the ranked nature of our outcome variable (normal, prefrail, and frail), to estimate odds ratios (OR) and 95% confidence intervals (CI). The appropriateness of these models was assessed using a proportional odds assumption chi-square test, wherein a nonsignificant (p > .05) value suggests that the ordered logistic regression is acceptable for describing the associations between the predictors and outcome variable (29). Parameters resulting from these models were converted to ORs using an ex transformation. For instance, the resulting ORs represent the increased odds of being diagnosed at a more advanced rank of frailty corresponding to impaired CVR compared to normal. We first adjusted for age at T1 and for education, then for BMI, height, place of birth, DBP, smoking, previous myocardial infarction (MI), diabetes, depressive symptoms (GDS ≥ 5), physical activity, and subsequently for global cognitive score at T1. Patients with stroke and/or dementia are prone to develop frailty. To check the association between CVR and frailty (without the effect of dementia and stroke), we repeated the analysis excluding participants with stroke and dementia (n = 55) at T2. Because of loss to follow-up of eligible patients who had either died or refused participation, we estimated the probability of every individual to actually reach the frailty assessment and calculated the inverse probability weights (IPW) (30). We compared the results of weighted analysis with nonweighted analysis. Data were analyzed using SPSS version 21 (SPSS, Chicago, IL). Results Of the 1,232 patients eligible for the late-life evaluation, 214 had died, 259 refused to participate, 102 could not be contacted, 45 were unable to participate due to dementia, language incompatibility, vision or hearing defects or physical disability, and 54 were excluded for missing data. This resulted in 558 people available for the first late-life assessment (T1) (Figure 1). Figure 1. View largeDownload slide Study flow chart [the present analysis is based on patients from eight hospitals in the center of Israel participating in the BIP trial. Patients included in the BIP Neurocognitive Study have undergone two follow-up evaluations (T1 and T2)]. Figure 1. View largeDownload slide Study flow chart [the present analysis is based on patients from eight hospitals in the center of Israel participating in the BIP trial. Patients included in the BIP Neurocognitive Study have undergone two follow-up evaluations (T1 and T2)]. Characteristics of patients who were included in late-life assessments compared with those not included Compared with patients who did not participate in the T1 evaluation, those included were younger and more educated. In addition, patients included in T1 had a lower systolic blood pressure and DBP, but had higher levels of total cholesterol and creatinine (Supplementary Table 1). Of the 558 patients, complete information on CVR was available for 482 patients. Women (n = 12) were not included due to the small sample. A total of 351 patients were reassessed at the second late-life evaluation (T2) with a response rate of 86 per cent, and 327 of them had CVR measurements as well as frailty evaluation (Figure 1). The attrition between the T1 and T2 evaluations was mainly due to interim death (n = 114), and in addition, 58 refused to participate, 4 could not be contacted, and 19 were unable to participate. The mean age of study patients was 56.7 ± 6.5 years at baseline, 71.8 ± 6.4 years at T1 evaluation, and 77.1 ± 6.4 years at T2. Characteristics of study patients at baseline by tertiles of CVR Supplementary Table 2 presents characteristics of study patients according to CVR tertiles. In general, patients in the higher CVR tertile had more often an academic degree, lower total cholesterol and glucose at baseline, and lower DBP and BMI at T1 compared with patients in the lower or middle tertiles. Also, individuals in the higher and middle CVR tertiles compared with those in the lower tertile had higher global cognitive scores, were taller, and were more physically active. Among the 327 patients, 112 (34.3 per cent) were classified as nonfrail and non-pre-frail, 122 (37.3 per cent) as pre-frail, and 93 (28.4 per cent) as frail. Association between CVR and frailty status The distribution of frailty status by CVR tertiles is depicted in Figure 2. Frailty was found among 43.0 per cent of patients in the lower tertile, 30.1 per cent in the middle tertile, and 26.9 per cent in the higher tertile of CVR (p for trend = .004). Thus, higher CVR was associated with nonfrail/non-prefrail status, whereas lower CVR was associated with frailty. Adjusting for age, education, BMI, height, DBP, previous MI, diabetes, depressive symptoms, and physical activity, the estimated adjusted OR (95% CI) for increasing frailty for patients in the lower tertile was 1.84 (1.04–3.26) and for those in the middle tertile 1.21 (0.69–2.12), compared with the higher CVR tertile. Additional adjustment for global cognitive score did not materially alter the results (Table 1). Applying inverse probability weighting the estimated adjusted OR (95% CI) for increasing frailty for patients in the lower vs. higher CVR tertile was 2.00 (1.32–3.01). Other variables associated with increasing frailty were lower education and height, physical inactivity, diabetes, and depressive symptoms. Comparing impaired with normal CVR, the estimated OR (95% CI) for increasing frailty was 1.76 (1.11–2.80; Table 1). After excluding patients with stroke and dementia at the time of the T2 assessment, this association remained statistically significant (Table 1). Table 1. OR (95% CI) for Frailty and Prefrailty by Tertiles of CVR and by CVR Groups (Impaired vs. Normal)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n =36) and dementia at T2 (n=19); Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Table 1. OR (95% CI) for Frailty and Prefrailty by Tertiles of CVR and by CVR Groups (Impaired vs. Normal)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n =36) and dementia at T2 (n=19); Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Figure 2. View largeDownload slide Distribution of frailty status by CVR tertiles. Figure 2. View largeDownload slide Distribution of frailty status by CVR tertiles. In subgroup analyses (Table 2), a significant interaction (p = .004) was noted between age and CVR status with frailty. Impaired CVR was significantly associated with increasing frailty among relatively younger individuals (≤77 years old) with an adjusted OR of 2.61 (1.36–5.03) while among older individuals the OR was 1.22 (0.61–2.47). Table 2. OR (95% CI) for Increasing Frailty by CVR Groups (Impaired vs. Normal): Subgroup Analysis     Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00      Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00  Notes: †Median age, ‡median global cognitive score at T1. Adjusted for age, education, BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity. View Large Table 2. OR (95% CI) for Increasing Frailty by CVR Groups (Impaired vs. Normal): Subgroup Analysis     Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00      Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00  Notes: †Median age, ‡median global cognitive score at T1. Adjusted for age, education, BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity. View Large Association between cIMT and frailty status The distribution of frailty status by cIMT tertiles is depicted in Figure 3. Frailty was found among 36.7 per cent of patients in the higher tertile, 29.6 per cent in the middle tertile, and 19.1 per cent in the lower tertile of cIMT (p for trend = .002). Adjusting for age and education, the estimated OR (95% CI) for increasing frailty for patients in the higher tertile of cIMT was 1.71 (1.02–2.87) and for those in the middle tertile 1.62 (0.97–2.71), compared with the lower cIMT tertile. The association between cIMT and frailty was, however, no longer significant after adjustment for additional health-related factors and global cognitive score (Table 3). The estimated OR (95% CI) for increasing frailty comparing patients with cIMT ≥ 0.93 mm and/or bilateral carotid plaques to others (cIMT < 0.93 and without bilateral carotid plaques) was 1.63 (0.96–2.76; Table 3). Table 3. OR (95% CI) for Increasing Frailty by Tertiles of cIMT and by cIMT Groups (cIMT ≥ 0.93 mm and/or Bilateral Carotid Plaques vs. Others)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n = 36) and dementia (n = 19) at T2 assessment; Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Table 3. OR (95% CI) for Increasing Frailty by Tertiles of cIMT and by cIMT Groups (cIMT ≥ 0.93 mm and/or Bilateral Carotid Plaques vs. Others)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n = 36) and dementia (n = 19) at T2 assessment; Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Figure 3. View largeDownload slide Distribution of frailty status by cIMT tertiles. Figure 3. View largeDownload slide Distribution of frailty status by cIMT tertiles. Discussion In this study of high-risk vascular patients with pre-existing CVD, impaired CVR, a marker of cerebral microvascular hemodynamic dysfunction, was related to late-life prefrailty and frailty. This observed relationship was independent of traditional vascular risk factors, the global cognitive score, and remained unchanged after excluding those with a history of stroke or dementia at the frailty assessment. Furthermore, a significant interaction was noted with age, with a more pronounced association noted at a relatively younger age. Additionally, higher cIMT tertile was associated with frailty; however, this finding was not statistically significant after adjustment for additional health-related factors. Although frailty rates vary depending on the population and scale used, this study has prevalence rates of frailty and prefrailty about 28 and 37 per cent, respectively, that are in line with other studies of patients with CVD (3, 31). It is important to note that frailty has a prefrail phase which allows an opportunity for prevention and intervention. Indeed, the fact that impaired CVR, that is, less common in younger age, was significantly associated with increasing frailty among relatively younger individuals highlights the potential for prevention of frailty in younger age. Frailty and even more prefrailty are reversible conditions if appropriately treated (1,32). In pilot-randomized clinical trials, physical, nutritional, and cognitive interventional approaches seemed to reduce frailty and prefrailty compared with controls (32, 33), and particularly improved muscle strength and energy among older persons (32). Recently published two early-phase trials of mesenchymal stem cells transplantation in frail older humans showed “remarkable improvements” of frailty deficits. These trials represent a promising and innovative approach for the treatment of frailty in older adults (34). This emphasizes the need to better identify patients prone to prefrailty and frailty. In this regard, cerebral microvascular dysfunction may be a useful biomarker that can be assessed noninvasively and inexpensively (16), and which may assist the prediction of patients prone to frailty. There are several lines of evidence lending further support for the role of cerebrovascular disease in the pathogenesis of physical frailty. Some studies suggest that the extent of markers of vascular subclinical disease as well as infarct-like brain lesions and cerebral microbleeds, another indicator of cerebral small vessel disease, was associated with a frail status (6, 35). Buchman and co-workers found in the Memory and Aging Project that microvascular brain pathology is common in older adults and may represent an under-recognized, independent cause of late-life motor impairment (36). Other studies reported a positive association between subclinical cerebrovascular damages such as the white matter hyperintensities (WMH) and frailty (8, 9, 14). A strong relationship was found between impaired cerebrovascular hemodynamics and loss of cerebral white matter structural integrity in elderly individuals with vascular risk factors (37). In the Rotterdam study, lower CVR was associated with an increased risk of death, independent of cardiovascular risk factors and stroke (13). Our study extends prior findings by showing that macro- and microvascular dysfunctions are related to increasing frailty in a sample of men with pre-existing coronary heart disease. We found that frail patients have greater carotid cIMT than their non-frail counterparts; yet, this association was not significant after adjustments for additional health-related factors and global cognitive score. Unlike our results, in the Three Cities cross-sectional study of community-dwelling participants, cIMT and carotid diameter were associated with frailty (7). These conflicting findings may be attributed to differences in characteristics of the study samples, as our study includes only patients with pre-existing CVD who have considerably higher prevalence of impaired cIMT, or due to other methodological differences. Some longitudinal studies show that frailty is a risk factor for cognitive decline, whereas others suggest that cognitive decline precedes frailty (38). Recently, we found that impaired CVR is associated with poorer cognitive performance, particularly in the executive function domain (10). Balestrini and co-workers found that severe carotid stenosis and reduced ipsilateral CVR significantly influenced the occurrence of cognitive decline in stroke-free patients (11). In our study, the association between CVR and frailty persisted despite adjustment for the global cognitive score, highlighting the importance of CVR as a marker in the pathophysiological mechanisms of frailty. There are some plausible biological pathways through which vascular disease and frailty are interrelated. Frailty and CVD share common biological pathways, and CVD may accelerate the development of frailty. Endothelial dysfunction may be an underlying mechanism for frailty (39). The relationship between impaired CVR and frailty may be explained by cerebral and systemic vascular disease, with decreasing total physiological reserve (13, 14). Impaired CVR may be casually related to frailty or alternatively serve as a marker for systemic microvascular disease (13). Nevertheless, frailty may represent another less recognized manifestation of vascular disease and thus explain, at least in-part, several of their adverse health-related outcomes (40). Our study presents some strengths and limitations. The study includes a unique dataset of extensively studied patients who underwent a comprehensive and validated cognitive evaluation as well as assessments of cIMT and CVR. The study has several limitations. First, the causal effect of vascular abnormality on frailty cannot be evaluated because we did not have a prior assessment of frailty status at the time of CVR assessment. Second, CVR was measured by TCD using breath-holding which is a simple and easy to implement tool; yet, more accurate TCD and MRI-based measures may be available. Furthermore, as expected, TCD measurements failed in approximately 7.5 per cent of our patients, due to absence of an acoustic window. The effect of misclassification of CVR may, however, reduce the strength of our reported associations. Third, the generalizability of the study might be limited to men with pre-existing CVD and specific clinical characteristics as included in the current study. Patients with a high burden of vascular risk factors are, however, at a higher risk of developing frailty. Finally, we have not quantified systematically the degree of carotid stenosis, yet only a small minority had hemodynamically significant carotid stenosis. In summary, we observed a relationship between poorer cerebrovascular status, in-particular impaired CVR, and late-life frailty. Frailty increases the risk of adverse health outcomes including decreased quality of life, disability, recurrent hospitalizations, and death. These findings provide support that cerebral microvascular disease is involved in the pathophysiology of physical frailty and suggest an added value of assessing the cerebral vascular functional status for identifying patients at-risk to develop frailty. Supplementary Material Supplementary data are available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online. Authors’ Contributions All authors contributed to study conception, data analysis, and interpretation. All authors contributed to revising the manuscript for publication. All authors have approved the final version for publication. Conflict of interest None declared. References 1. Fried LP, Tangen CM, Walston Jet al.   Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci . 2001; 56: M146– 156. Google Scholar CrossRef Search ADS PubMed  2. Collard RM, Boter H, Schoevers RA, Oude Voshaar RC. Prevalence of frailty in community-dwelling older persons: a systematic review. J Am Geriatr Soc . 2012; 60: 1487– 1492. doi: 10.1111/j.1532-5415.2012.04054.x Google Scholar CrossRef Search ADS PubMed  3. Afilalo J. Frailty in patients with cardiovascular disease: why, when, and how to measure. Curr Cardiovasc Risk Rep . 2011; 5: 467– 472. doi: 10.1007/s12170-011-0186-0 Google Scholar CrossRef Search ADS PubMed  4. Veronese N, Cereda E, Stubbs Bet al.   Risk of cardiovascular disease morbidity and mortality in frail and pre-frail older adults: Results from a meta-analysis and exploratory meta-regression analysis. Ageing Res Rev . 2017; 28: 63– 73. doi: 10.1016/j.arr.2017.01.003 Google Scholar CrossRef Search ADS   5. Gale CR, Cooper C, Sayer AA. Framingham cardiovascular disease risk scores and incident frailty: the English longitudinal study of ageing. Age (Dordr) . 2014; 36: 9692. doi: 10.1007/s11357-014-9692-6 Google Scholar CrossRef Search ADS PubMed  6. Newman AB, Gottdiener JS, Mcburnie MAet al.  ; Cardiovascular Health Study Research Group. Associations of subclinical cardiovascular disease with frailty. J Gerontol A Biol Sci Med Sci . 2001; 56: M158– M166. Google Scholar CrossRef Search ADS PubMed  7. Avila-Funes JA, Meillon C, González-Colaço Harmand M, Tzourio C, Dartigues JF, Amieva H. Association between frailty and carotid central structure changes: the Three-City Study. J Am Geriatr Soc . 2014; 62: 1906– 1911. doi: 10.1111/jgs.13062 Google Scholar CrossRef Search ADS PubMed  8. Avila-Funes JA, Pelletier A, Meillon Cet al.   Vascular Cerebral Damage in Frail Older Adults: The AMImage Study. J Gerontol A Biol Sci Med Sci . 2017; 72: 971– 977. doi: 10.1093/gerona/glw347 Google Scholar CrossRef Search ADS PubMed  9. Siejka TP, Srikanth VK, Hubbard REet al.   Frailty and cerebral small vessel disease: A cross-sectional analysis of the Tasmanian Study of Cognition and Gait (TASCOG). J Gerontol A Biol Sci Med Sci . 2017. 2018;73:255-260. doi: 10.1093/gerona/glx145 10. Haratz S, Weinstein G, Molshazki Net al.   Impaired cerebral hemodynamics and cognitive performance in patients with atherothrombotic disease. J Alzheimers Dis . 2015; 46: 137– 144. doi: 10.3233/JAD-150052 Google Scholar CrossRef Search ADS PubMed  11. Balestrini S, Perozzi C, Altamura Cet al.   Severe carotid stenosis and impaired cerebral hemodynamics can influence cognitive deterioration. Neurology . 2013; 80: 2145– 2150. doi: 10.1212/WNL.0b013e318295d71a Google Scholar CrossRef Search ADS PubMed  12. Gupta A, Chazen JL, Hartman Met al.   Cerebrovascular reserve and stroke risk in patients with carotid stenosis or occlusion: a systematic review and meta-analysis. Stroke . 2012; 43: 2884– 2891. doi: 10.1161/STROKEAHA.112.663716 Google Scholar CrossRef Search ADS PubMed  13. Portegies ML, de Bruijn RF, Hofman A, Koudstaal PJ, Ikram MA. Cerebral vasomotor reactivity and risk of mortality: the Rotterdam study. Stroke . 2014; 45: 42– 47. doi: 10.1161/STROKEAHA.113.002348 Google Scholar CrossRef Search ADS PubMed  14. Soumare A, Elbaz A, Zhu Yet al.   White matter lesions volume and motor performances in the elderly. Ann Neurol . 2009; 65: 706– 715. doi: 10.1002/ana.21674 Google Scholar CrossRef Search ADS PubMed  15. Lorenz MW, Polak JF, Kavousi Met al.  ; PROG-IMT Study Group. Carotid intima-media thickness progression to predict cardiovascular events in the general population (the PROG-IMT collaborative project): a meta-analysis of individual participant data. Lancet . 2012; 379: 2053– 2062. doi: 10.1016/S0140-6736(12)60441-3 Google Scholar CrossRef Search ADS PubMed  16. Dahl A, Lindegaard KF, Russell Det al.   A comparison of transcranial Doppler and cerebral blood flow studies to assess cerebral vasoreactivity. Stroke . 1992; 23: 15– 19. Google Scholar CrossRef Search ADS PubMed  17. Goldbourt U, Behar S, Reicher-Reiss Het al.   Rationale and design of a secondary prevention trial of increasing serum high-density lipoprotein cholesterol and reducing triglycerides in patients with clinically manifest atherosclerotic heart disease (the Bezafibrate Infarction Prevention Trial). Am J Cardiol . 1993; 71: 909– 915. Google Scholar CrossRef Search ADS PubMed  18. Goldbourt U, Brunner D, Behar S, Reicher-Reiss H. Baseline characteristics of patients participating in the Bezafibrate Infarction Prevention (BIP) Study. Eur Heart J . 1998; 19 Suppl H: H42– H47. Google Scholar PubMed  19. Osman A, Walsemann KM. Ethnic disparities in disability among middle-aged and older Israeli adults: the role of socioeconomic disadvantage and traumatic life events. J Aging Health . 2013; 25: 510– 531. doi: 10.1177/0898264313478653 Google Scholar CrossRef Search ADS PubMed  20. Müller M, Voges M, Piepgras U, Schimrigk K. Assessment of cerebral vasomotor reactivity by transcranial Doppler ultrasound and breath-holding. A comparison with acetazolamide as vasodilatory stimulus. Stroke . 1995; 26: 96– 100. Google Scholar CrossRef Search ADS PubMed  21. Grant EG, Benson CB, Moneta GLet al.   Carotid artery stenosis: grayscale and Doppler ultrasound diagnosis—society of radiologists in ultrasound consensus conference. Ultrasound Q . 2003; 19: 190– 198. Google Scholar CrossRef Search ADS PubMed  22. Touboul PJ, Hennerici MG, Meairs Set al.   Mannheim carotid intima-media thickness consensus (2004–2006). An update on behalf of the advisory board of the 3rd and 4th watching the risk symposium, 13th and 15th European Stroke Conferences, Mannheim, Germany, 2004, and Brussels, Belgium, 2006. Cerebrovasc Dis . 2007; 23: 75– 80. doi: 10.1159/000097034 Google Scholar CrossRef Search ADS PubMed  23. Cesari M, Kritchevsky SB, Penninx BWet al.   Prognostic value of usual gait speed in well-functioning older people–results from the Health, Aging and Body Composition Study. J Am Geriatr Soc . 2005; 53: 1675– 1680. doi: 10.1111/j.1532-5415.2005.53501.x Google Scholar CrossRef Search ADS PubMed  24. Radloff L. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Measure . 1977; 1: 385– 401. Google Scholar CrossRef Search ADS   25. Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol . 1993; 46: 153– 162. Google Scholar CrossRef Search ADS PubMed  26. Almeida OP, Almeida SA. Short versions of the geriatric depression scale: a study of their validity for the diagnosis of a major depressive episode according to ICD-10 and DSM-IV. Int J Geriatr Psychiatry . 1999; 14: 858– 865. Google Scholar CrossRef Search ADS PubMed  27. Dwolatzky T, Whitehead V, Doniger GMet al.   Validity of a novel computerized cognitive battery for mild cognitive impairment. BMC Geriatr . 2003; 3: 4. doi: 10.1186/1471-2318-3-4 Google Scholar CrossRef Search ADS PubMed  28. Silvestrini M, Vernieri F, Pasqualetti Pet al.   Impaired cerebral vasoreactivity and risk of stroke in patients with asymptomatic carotid artery stenosis. JAMA . 2000; 283: 2122– 2127. Google Scholar CrossRef Search ADS PubMed  29. Long JS. Regression Models for Categorical and Limited Dependent Variables . Thousand Oaks, CA: Sage; 1997. 30. Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology . 2000; 11: 550– 560. Google Scholar CrossRef Search ADS PubMed  31. Myers V, Drory Y, Goldbourt U, Gerber Y. Multilevel socioeconomic status and incidence of frailty post myocardial infarction. Int J Cardiol . 2014; 170: 338– 343. doi: 10.1016/j.ijcard.2013.11.009 Google Scholar CrossRef Search ADS PubMed  32. Ng TP, Feng L, Nyunt MSet al.   Nutritional, physical, cognitive, and combination interventions and frailty reversal among older adults: a Randomized Controlled Trial. Am J Med . 2015; 128: 1225– 1236.e1. doi: 10.1016/j.amjmed.2015.06.017 Google Scholar CrossRef Search ADS PubMed  33. Cameron ID, Fairhall N, Langron Cet al.   A multifactorial interdisciplinary intervention reduces frailty in older people: randomized trial. BMC Med . 2013; 11: 65. doi: 10.1186/1741-7015-11-65 Google Scholar CrossRef Search ADS PubMed  34. Le Couteur DG, Anderson RM, Newman AB, de Cabo R. Stem cell transplantation for frailty. J Gerontol A Biol Sci Med Sci . 2017; 72: 1503– 1504. doi: 10.1093/gerona/glx158 Google Scholar CrossRef Search ADS PubMed  35. Chung CP, Chou KH, Chen WTet al.   Cerebral microbleeds are associated with physical frailty: a community-based study. Neurobiol Aging . 2016; 44: 143– 150. doi: 10.1016/j.neurobiolaging.2016.04.025 Google Scholar CrossRef Search ADS PubMed  36. Buchman AS, Yu L, Boyle PAet al.   Microvascular brain pathology and late-life motor impairment. Neurology . 2013; 80: 712– 8. doi: 10.1212/WNL.0b013e3182825116 Google Scholar CrossRef Search ADS PubMed  37. Purkayastha S, Fadar O, Mehregan Aet al.   Impaired cerebrovascular hemodynamics are associated with cerebral white matter damage. J Cereb Blood Flow Metab . 2014; 34: 228– 234. doi: 10.1038/jcbfm.2013.180 Google Scholar CrossRef Search ADS PubMed  38. Searle SD, Rockwood K. Frailty and the risk of cognitive impairment. Alzheimers Res Ther . 2015; 7: 54. doi: 10.1186/s13195-015-0140-3 Google Scholar CrossRef Search ADS PubMed  39. Alonso-Bouzón C, Carcaillon L, García-García FJ, Amor-Andrés MS, El Assar M, Rodríguez-Mañas L. Association between endothelial dysfunction and frailty: the Toledo Study for Healthy Aging. Age (Dordr) . 2014; 36: 495– 505. doi: 10.1007/s11357-013-9576-1 Google Scholar CrossRef Search ADS PubMed  40. Gouw AA, Seewann A, van der Flier WMet al.   Heterogeneity of small vessel disease: a systematic review of MRI and histopathology correlations. J Neurol Neurosurg Psychiatry . 2011; 82: 126– 135. doi: 10.1136/jnnp.2009.204685 Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences Oxford University Press

Loading next page...
 
/lp/ou_press/impaired-cerebral-hemodynamics-and-frailty-in-patients-with-HsBjsWqV1s
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1079-5006
eISSN
1758-535X
D.O.I.
10.1093/gerona/glx253
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Recent studies suggest that impaired cerebrovascular reactivity (CVR), a marker of cerebral microvascular damage, is associated with a higher risk of stroke, cognitive decline, and mortality. We tested whether abnormal cerebrovascular status is associated with late-life frailty among men with pre-existing cardiovascular disease. Methods A subset of 327 men (mean age at baseline 56.7 ± 6.5 years) who previously participated in the Bezafibrate Infarction Prevention (BIP) trial (1990–1997) and then in the BIP Neurocognitive Study underwent a neurovascular evaluation 14.6 ± 1.9 years after baseline (T1) and were evaluated for frailty 19.9 ± 1.0 years after baseline (T2). CVR was measured at T1 using the breath-holding index and carotid large-vessel disease using ultrasound. Frailty status was measured at T2 according to the physical phenotype developed by Fried. Patients were categorized into CVR tertiles with cutoff points at ≤0.57, 0.58–0.94, and ≥0.95 and also as normal or impaired (<0.69) CVR. We assessed the change in the odds of being in the advanced rank of frailty status (normal, prefrail, and frail) using ordered logistic regression. Results After adjustment, the estimated OR (95% confidence intervals) for increasing frailty in the lower tertile was 1.94 (1.09–3.46) and in the middle tertile 1.24 (0.70–2.19), compared with the higher CVR tertile. The estimated OR for increasing frailty for patients with impaired vs. normal CVR was 1.76 (1.11–2.80). Conclusions These findings provide support that cerebral microvascular dysfunction among patients with pre-existing cardiovascular disease is related to prefrailty and frailty and suggest an added value of assessing the cerebral vascular functional status for identifying patients at-risk of developing frailty. Cerebrovascular reactivity, Cerebral hemodynamics, Cardiovascular disease, Frailty, Small vessel disease Frailty is a geriatric syndrome used to define older adults with impaired resistance to stressors due to a decline in physiological reserve. Frailty is characterized by a nonspecific state of vulnerability, and multisystem dysregulation, leading to a higher risk for cognitive decline, disability, and death (1). Prevalence of frailty is estimated at 11 per cent among adults aged 65 and older and reaching over 26 per cent in those aged ≥85 years (2) and higher in people with cardiovascular disease (CVD) (3). Some studies suggest that frailty is a risk factor for CVD (4), whereas others suggest that CVD precedes frailty (5). Recent evidence suggests that in comparison with nonfrail older adults, those who exhibit the phenotype of frailty have more extensive vascular damage (6, 7). Indeed, cerebrovascular disease is present among both cognitively impaired and frail persons, since these conditions share pathophysiological mechanisms, via cerebrovascular damage (7–9). It is speculated that cerebral microvascular dysfunction occurs due to arteriosclerosis and decrease of elasticity of the cerebral small-vessel walls. There is growing evidence that cerebral microvascular dysfunction is associated with accelerated cognitive decline (10, 11), stroke (12), and mortality (13). Nonetheless, very few studies (6–8, 14) have assessed carotid intima-media thickness (cIMT), a marker of cerebral macrovascular disease (15), and impaired cerebrovascular reactivity (CVR), a marker of cerebral microvascular dysfunction (16), in frail older adults. We used data from the Bezafibrate Infarction Prevention (BIP) Neurocognitive study to address the hypothesis that abnormal cerebrovascular status (impaired CVR and abnormal cIMT) may be associated with late-life frailty among men with pre-existing CVD. Examining this relationship may be important both for studying pathways through which vascular disease and frailty are interrelated and for identifying high-risk patients prone to frailty. Material and Methods Patients The population for the current study includes patients from eight central hospitals (n = 1,232) who reside in the central region of Israel and who previously participated in the BIP clinical trial of lipid modification and then in the BIP Neurocognitive Study. The study design and procedures of the BIP trial have been previously described in detail (17). In brief, the BIP study was a placebo-controlled randomized clinical trial investigating the efficacy of bezafibrate 400 mg daily, a fibric derivative, in secondary prevention among patients with established stable coronary heart disease. The lipid profile of patients at inclusion was as follows: serum total cholesterol 180 to 250 mg/dL, low-density lipoprotein–cholesterol ≤ 180 mg/dL (≤160 mg/dL for people <50 years), high-density lipoprotein–cholesterol ≤ 45 mg/dL, and triglycerides ≤ 300 mg/dL. Other exclusion criteria were insulin-dependent diabetes, hepatic or renal failure, and disabling stroke. Patients included in the BIP Neurocognitive Study have undergone two follow-up evaluations (T1 and T2). The first follow-up evaluation (T1, n = 558) was performed during 2004–2008, an average of 14.6 ± 1.9 years after recruitment, assessing neurovascular and cognitive function. Patients were re-examined during 2011–2013 (T2, n = 351), 19.9 ± 1.0 years after recruitment, assessing frailty and re-assessing cognitive function (Figure 1). The mean time interval between the T1 and T2 assessments was 4.8 ± 1.3 years. The second late-life evaluation took place at a central research center (the Sagol Neuroscience Center), or if patients were unable or unwilling to attend the research center, the assessments were completed at their residence. The study was approved by the local institutional review board, and informed consent was obtained from all patients. Baseline Measurements Methods for assessment of vascular risk factors at baseline of the BIP study are described elsewhere (18). Blood samples were drawn from each study patient at baseline of the BIP trial (1990–1992). Samples were collected after at least 12 hours of fasting, with the use of standardized equipment and procedures and transferred to a central study laboratory. Education, occupation, place of birth, and comorbidity were collected through a baseline questionnaire. We included data on birth place, which is an important key indicator of socioeconomic status disparities and possibly of genetic predisposition (19). Neurovascular Evaluation (T1) CVR was evaluated by transcranial Doppler (TCD) using the breath-holding index (BHI), previously described by Muller and co-workers (20). All exams were performed by a qualified technician using a Trans-Link 9900 TCD device (Rimed, Raanana, Israel) equipped with a 2 MHz pulsed handheld probe. Before proceeding to the definitive recording, patients were trained to perform the procedure correctly. After normal breathing of room air for approximately 4 minutes, patients were asked to hold their breath for 30 seconds or as long as possible, after a normal inspiration. Patients unable to hold their breath for at least 15 seconds were excluded. During the maneuver, the middle cerebral artery (MCA) mean blood flow velocity was recorded continuously. The length time of apnea was measured in seconds. Mean flow velocity at rest was obtained by continuous recording of a 1 minute period of normal breathing. As the CVR measure, the BHI in the MCA was calculated in a standardized manner, as the percent increase in MCA mean blood velocity recorded by breath-holding divided by seconds of breath-holding ([Vbh − Vr/Vr] × 100 × s−1), where Vbh is MCA mean blood velocity at the end of breath-holding, Vr is the MCA mean blood velocity at rest, and s-1 is per second of breath-holding. Carotid ultrasound examination was performed to identify the presence of carotid plaques, following a standard protocol (21), using a HDI 5000 SonoCT system device (Philips, Eindhoven, The Netherlands) with a linear array multifrequency transducer (4 to 7 MHz). The presence of bilateral carotid plaque was regarded as evidence of carotid large-vessel disease. In addition, cIMT was measured at the far wall of both common carotid arteries at 1.0 cm proximal to the carotid bifurcation in length of 1.0 cm using high resolution B-mode ultrasound. Assessment followed a standard protocol (22) that included acquiring images of both carotid arteries and measurement of the distance between the media-adventitia interface and the lumen-intima interface, employing automatic edge detection (METRIS, France). Frailty Evaluation (T2) Frailty was assessed using the phenotype developed by Fried and co-workers (1). It comprises five individual criteria: weight loss, weakness (grip strength), slowness (walking speed), self-reported exhaustion, and low physical activity. Weight loss was categorized as positive if the patient reported loss greater than or equal to 4.0 kg unintentionally in the prior year. Grip strength was defined as isometric dominant handgrip strength assessed using a hydraulic Jamar dynamometer (Sammons & Preston, Bolingbrook, IL). Low grip strength was denoted as ≤29 kg [for patients with body mass index (BMI) ≤24 kg/m2], ≤30 kg (for BMI 24.1–28.0 kg/m2), ≤32 kg (for BMI >28.0 kg/m2). The test was carried out twice and the maximum score was used. Gait speed was measured by gait time in seconds using a 5 m timed walk test. Usual gait speed of less than 1 m/second identifies persons at high risk of health-related outcomes in well-functioning older people (23). We used height-adjusted time as the cutoff. Slowness was denoted as ≥6 seconds (for height ≤173 cm) and ≥ 5 seconds (for height >173 cm) (1). The test instructions were as follows: “On the word ‘go’ you will start walking at your regular pace to the line on the floor.” Exhaustion was defined by two self-report questions (“I felt everything I did was an effort” and “I could not get going”) from the Center for Epidemiological Studies Depression Scale (24). Patients answering “moderate amount of time (3–4 days) or most of the time (5–7 days) last week to these self-report questions were categorized were categorized as frail by the exhaustion criterion. Physical activity was assessed by the Physical Activity Scale for the Elderly scale (25). Patients who scored in the lowest quintile were categorized as positive for the low physical activity criterion. Frailty was defined as the presence of three or more of these five criteria, and those with one or two of these criteria were considered prefrail. Additional Assessments In both evaluations (T1 and T2), data were collected systematically regarding new comorbidities and hospitalizations, medication use, smoking status, physical activity, and anthropometric measurements. In addition, systolic blood pressure (SBP), diastolic blood pressure (DBP), and BMI were measured. Incident stroke during follow-up was assessed by reviewing records from a hospital or emergency department discharge, a primary care physician, or a neurologist. Depressive symptoms were assessed by the short version Geriatric Depression Scale (GDS) using a cutoff of ≥5 (26). Patients completed the NeuroTrax computerized cognitive testing (NeuroTrax Corp., Bellaire, 175 TX). A description of the tests included has been published elsewhere (27). Dementia and incident stroke during follow-up were determined by an adjudication committee composed of three investigators, two of which are experienced board certified neurologists. Dementia was determined based on the sum of cognitive evaluation, clinical interview, and data collected and in accordance with the Diagnostic and Statistical Manual of Mental Disorders 4th Edition (DSM-IV) criteria and stroke according to World Health Organization criteria. For the purpose of the current analysis, patients with dementia before T1 evaluation were excluded (n = 45), in order to assure compliance with and reliable measurement of CVR. Statistical Analysis We have examined the differences in the distribution between right and left CVR and cIMT. Since no differences were found, we have decided to use the average values. Patients were categorized into normal (≥0.69) or impaired (<0.69) CVR based on the mean BHI of both MCAs according to the previously established standard parameters (28) and also into tertiles of CVR, with cut-off points at ≤0.57, 0.58–0.94, and ≥0.95. Patients were categorized into tertiles of mean (both sides) cIMT with cut-off points at ≤0.87, 0.88–1.02, and ≥1.03 mm and also into categories of cIMT ≥ 0.93 mm and/or bilateral carotid plaques vs. others (cIMT < 0.93 and without bilateral carotid plaques) according to the previously established standard parameters (15). The clinical characteristics of the patients are expressed as percentages and as mean ± SD, except when the distribution was strongly skewed, in which case, median and interquartile range are provided. The chi-square test for trend was used to determine the significance of differences between CVR tertiles in the distribution of categorical data. The clinical characteristics of the continuous variables were compared between CVR tertiles by analysis of variance or Kruskal–Wallis test for variables not normally distributed. We used ordered logistic regression, due to the ranked nature of our outcome variable (normal, prefrail, and frail), to estimate odds ratios (OR) and 95% confidence intervals (CI). The appropriateness of these models was assessed using a proportional odds assumption chi-square test, wherein a nonsignificant (p > .05) value suggests that the ordered logistic regression is acceptable for describing the associations between the predictors and outcome variable (29). Parameters resulting from these models were converted to ORs using an ex transformation. For instance, the resulting ORs represent the increased odds of being diagnosed at a more advanced rank of frailty corresponding to impaired CVR compared to normal. We first adjusted for age at T1 and for education, then for BMI, height, place of birth, DBP, smoking, previous myocardial infarction (MI), diabetes, depressive symptoms (GDS ≥ 5), physical activity, and subsequently for global cognitive score at T1. Patients with stroke and/or dementia are prone to develop frailty. To check the association between CVR and frailty (without the effect of dementia and stroke), we repeated the analysis excluding participants with stroke and dementia (n = 55) at T2. Because of loss to follow-up of eligible patients who had either died or refused participation, we estimated the probability of every individual to actually reach the frailty assessment and calculated the inverse probability weights (IPW) (30). We compared the results of weighted analysis with nonweighted analysis. Data were analyzed using SPSS version 21 (SPSS, Chicago, IL). Results Of the 1,232 patients eligible for the late-life evaluation, 214 had died, 259 refused to participate, 102 could not be contacted, 45 were unable to participate due to dementia, language incompatibility, vision or hearing defects or physical disability, and 54 were excluded for missing data. This resulted in 558 people available for the first late-life assessment (T1) (Figure 1). Figure 1. View largeDownload slide Study flow chart [the present analysis is based on patients from eight hospitals in the center of Israel participating in the BIP trial. Patients included in the BIP Neurocognitive Study have undergone two follow-up evaluations (T1 and T2)]. Figure 1. View largeDownload slide Study flow chart [the present analysis is based on patients from eight hospitals in the center of Israel participating in the BIP trial. Patients included in the BIP Neurocognitive Study have undergone two follow-up evaluations (T1 and T2)]. Characteristics of patients who were included in late-life assessments compared with those not included Compared with patients who did not participate in the T1 evaluation, those included were younger and more educated. In addition, patients included in T1 had a lower systolic blood pressure and DBP, but had higher levels of total cholesterol and creatinine (Supplementary Table 1). Of the 558 patients, complete information on CVR was available for 482 patients. Women (n = 12) were not included due to the small sample. A total of 351 patients were reassessed at the second late-life evaluation (T2) with a response rate of 86 per cent, and 327 of them had CVR measurements as well as frailty evaluation (Figure 1). The attrition between the T1 and T2 evaluations was mainly due to interim death (n = 114), and in addition, 58 refused to participate, 4 could not be contacted, and 19 were unable to participate. The mean age of study patients was 56.7 ± 6.5 years at baseline, 71.8 ± 6.4 years at T1 evaluation, and 77.1 ± 6.4 years at T2. Characteristics of study patients at baseline by tertiles of CVR Supplementary Table 2 presents characteristics of study patients according to CVR tertiles. In general, patients in the higher CVR tertile had more often an academic degree, lower total cholesterol and glucose at baseline, and lower DBP and BMI at T1 compared with patients in the lower or middle tertiles. Also, individuals in the higher and middle CVR tertiles compared with those in the lower tertile had higher global cognitive scores, were taller, and were more physically active. Among the 327 patients, 112 (34.3 per cent) were classified as nonfrail and non-pre-frail, 122 (37.3 per cent) as pre-frail, and 93 (28.4 per cent) as frail. Association between CVR and frailty status The distribution of frailty status by CVR tertiles is depicted in Figure 2. Frailty was found among 43.0 per cent of patients in the lower tertile, 30.1 per cent in the middle tertile, and 26.9 per cent in the higher tertile of CVR (p for trend = .004). Thus, higher CVR was associated with nonfrail/non-prefrail status, whereas lower CVR was associated with frailty. Adjusting for age, education, BMI, height, DBP, previous MI, diabetes, depressive symptoms, and physical activity, the estimated adjusted OR (95% CI) for increasing frailty for patients in the lower tertile was 1.84 (1.04–3.26) and for those in the middle tertile 1.21 (0.69–2.12), compared with the higher CVR tertile. Additional adjustment for global cognitive score did not materially alter the results (Table 1). Applying inverse probability weighting the estimated adjusted OR (95% CI) for increasing frailty for patients in the lower vs. higher CVR tertile was 2.00 (1.32–3.01). Other variables associated with increasing frailty were lower education and height, physical inactivity, diabetes, and depressive symptoms. Comparing impaired with normal CVR, the estimated OR (95% CI) for increasing frailty was 1.76 (1.11–2.80; Table 1). After excluding patients with stroke and dementia at the time of the T2 assessment, this association remained statistically significant (Table 1). Table 1. OR (95% CI) for Frailty and Prefrailty by Tertiles of CVR and by CVR Groups (Impaired vs. Normal)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n =36) and dementia at T2 (n=19); Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Table 1. OR (95% CI) for Frailty and Prefrailty by Tertiles of CVR and by CVR Groups (Impaired vs. Normal)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Lower (≤0.57)  2.29 (1.34–3.91)  .002  1.84 (1.04–3.26)  .04  1.94 (1.09–3.46)  .02  2.08 (1.11–3.90)  .02  2.00 (1.32–3.01)  <.001  Middle (0.58–0.94)  1.30 (0.77–2.21)  .33  1.21 (0.69–2.12)  .52  1.24 (0.70–2.19)  .46  1.26 (0.68–2.33)  .47  1.12 (0.75–1.67)  .58  Higher (≥0.95)  1.00    1.00    1.00    1.00    1.00    Impaired (CVR<0.69)  1.94 (1.25–3.00)  .003  1.72 (1.09–2.73)  .02  1.76 (1.11–2.80)  .02  1.73 (1.04–2.88)  .03  1.81 (1.30–2.50)  <.001  Normal (CVR≥0.69)  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n =36) and dementia at T2 (n=19); Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Figure 2. View largeDownload slide Distribution of frailty status by CVR tertiles. Figure 2. View largeDownload slide Distribution of frailty status by CVR tertiles. In subgroup analyses (Table 2), a significant interaction (p = .004) was noted between age and CVR status with frailty. Impaired CVR was significantly associated with increasing frailty among relatively younger individuals (≤77 years old) with an adjusted OR of 2.61 (1.36–5.03) while among older individuals the OR was 1.22 (0.61–2.47). Table 2. OR (95% CI) for Increasing Frailty by CVR Groups (Impaired vs. Normal): Subgroup Analysis     Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00      Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00  Notes: †Median age, ‡median global cognitive score at T1. Adjusted for age, education, BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity. View Large Table 2. OR (95% CI) for Increasing Frailty by CVR Groups (Impaired vs. Normal): Subgroup Analysis     Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00      Impaired (CVR<0.69)  Normal (CVR≥0.69)        OR (95% CI)  p  OR (95% CI)  p for interaction  Age groups (years)  ≤77†  2.61 (1.36–5.03)  .004  1.00  .004  >77  1.22 (0.61–2.47)  .58  1.00  Global cognitive score  ≥97‡  2.08 (1.11–3.90)  .04  1.00  .25  <97  1.38 (0.71–2.69)  .35  1.00  Diabetes  Yes  1.69 (0.68–4.17)  .25  1.00  .40  No  1.68 (0.96–2.94)  .07  1.00  Physical activity  Yes  1.33 (0.77–2.32)  .31  1.00  .89  No  3.29 (1.33–8.10)  .01  1.00  Notes: †Median age, ‡median global cognitive score at T1. Adjusted for age, education, BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity. View Large Association between cIMT and frailty status The distribution of frailty status by cIMT tertiles is depicted in Figure 3. Frailty was found among 36.7 per cent of patients in the higher tertile, 29.6 per cent in the middle tertile, and 19.1 per cent in the lower tertile of cIMT (p for trend = .002). Adjusting for age and education, the estimated OR (95% CI) for increasing frailty for patients in the higher tertile of cIMT was 1.71 (1.02–2.87) and for those in the middle tertile 1.62 (0.97–2.71), compared with the lower cIMT tertile. The association between cIMT and frailty was, however, no longer significant after adjustment for additional health-related factors and global cognitive score (Table 3). The estimated OR (95% CI) for increasing frailty comparing patients with cIMT ≥ 0.93 mm and/or bilateral carotid plaques to others (cIMT < 0.93 and without bilateral carotid plaques) was 1.63 (0.96–2.76; Table 3). Table 3. OR (95% CI) for Increasing Frailty by Tertiles of cIMT and by cIMT Groups (cIMT ≥ 0.93 mm and/or Bilateral Carotid Plaques vs. Others)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n = 36) and dementia (n = 19) at T2 assessment; Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Table 3. OR (95% CI) for Increasing Frailty by Tertiles of cIMT and by cIMT Groups (cIMT ≥ 0.93 mm and/or Bilateral Carotid Plaques vs. Others)   Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00      Model 1  Model 2  Model 3  Model 4  Model 5    OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  OR (95% CI)  p  Higher (≥1.03 mm)  1.71 (1.02–2.87)  .04  1.39 (0.81–2.41)  .23  1.33 (0.77–2.32)  .31  1.51 (0.82–2.82)  .19  1.33 (0.91–1.95)  .14  Middle (0.88–1.02 mm)  1.62 (0.97–2.71)  .06  1.48 (0.85–2.60)  .17  1.42 (0.80–2.50)  .23  1.48 (0.79–2.79)  .22  1.18 (0.79–1.75)  .42  Lower (≤0.87 mm)  1.00    1.00    1.00    1.00    1.00    cIMT ≥ 0.93 and/or bilateral carotid plaques  1.69 (1.01–2.80)  .04  1.63 (0.96–2.76)  .07  1.51 (0.89–2.57)  .13  1.33 (0.74–2.39)  .34  1.55 (1.05–2.72)  .03  cIMT < 0.93 and without bilateral carotid plaques  1.00    1.00    1.00    1.00    1.00    Notes: Model 1 = adjusted for age and education; Model 2 = model 1+ BMI, height, place of birth, diastolic BP, smoking, previous myocardial infarction, diabetes, depressive symptoms (GDS ≥ 5), physical activity; Model 3 = model 2 + global cognitive score; Model 4 = Model 3 excluding patients with stroke (n = 36) and dementia (n = 19) at T2 assessment; Model 5 = Model 3 applying IPW method; CI = confidence interval. View Large Figure 3. View largeDownload slide Distribution of frailty status by cIMT tertiles. Figure 3. View largeDownload slide Distribution of frailty status by cIMT tertiles. Discussion In this study of high-risk vascular patients with pre-existing CVD, impaired CVR, a marker of cerebral microvascular hemodynamic dysfunction, was related to late-life prefrailty and frailty. This observed relationship was independent of traditional vascular risk factors, the global cognitive score, and remained unchanged after excluding those with a history of stroke or dementia at the frailty assessment. Furthermore, a significant interaction was noted with age, with a more pronounced association noted at a relatively younger age. Additionally, higher cIMT tertile was associated with frailty; however, this finding was not statistically significant after adjustment for additional health-related factors. Although frailty rates vary depending on the population and scale used, this study has prevalence rates of frailty and prefrailty about 28 and 37 per cent, respectively, that are in line with other studies of patients with CVD (3, 31). It is important to note that frailty has a prefrail phase which allows an opportunity for prevention and intervention. Indeed, the fact that impaired CVR, that is, less common in younger age, was significantly associated with increasing frailty among relatively younger individuals highlights the potential for prevention of frailty in younger age. Frailty and even more prefrailty are reversible conditions if appropriately treated (1,32). In pilot-randomized clinical trials, physical, nutritional, and cognitive interventional approaches seemed to reduce frailty and prefrailty compared with controls (32, 33), and particularly improved muscle strength and energy among older persons (32). Recently published two early-phase trials of mesenchymal stem cells transplantation in frail older humans showed “remarkable improvements” of frailty deficits. These trials represent a promising and innovative approach for the treatment of frailty in older adults (34). This emphasizes the need to better identify patients prone to prefrailty and frailty. In this regard, cerebral microvascular dysfunction may be a useful biomarker that can be assessed noninvasively and inexpensively (16), and which may assist the prediction of patients prone to frailty. There are several lines of evidence lending further support for the role of cerebrovascular disease in the pathogenesis of physical frailty. Some studies suggest that the extent of markers of vascular subclinical disease as well as infarct-like brain lesions and cerebral microbleeds, another indicator of cerebral small vessel disease, was associated with a frail status (6, 35). Buchman and co-workers found in the Memory and Aging Project that microvascular brain pathology is common in older adults and may represent an under-recognized, independent cause of late-life motor impairment (36). Other studies reported a positive association between subclinical cerebrovascular damages such as the white matter hyperintensities (WMH) and frailty (8, 9, 14). A strong relationship was found between impaired cerebrovascular hemodynamics and loss of cerebral white matter structural integrity in elderly individuals with vascular risk factors (37). In the Rotterdam study, lower CVR was associated with an increased risk of death, independent of cardiovascular risk factors and stroke (13). Our study extends prior findings by showing that macro- and microvascular dysfunctions are related to increasing frailty in a sample of men with pre-existing coronary heart disease. We found that frail patients have greater carotid cIMT than their non-frail counterparts; yet, this association was not significant after adjustments for additional health-related factors and global cognitive score. Unlike our results, in the Three Cities cross-sectional study of community-dwelling participants, cIMT and carotid diameter were associated with frailty (7). These conflicting findings may be attributed to differences in characteristics of the study samples, as our study includes only patients with pre-existing CVD who have considerably higher prevalence of impaired cIMT, or due to other methodological differences. Some longitudinal studies show that frailty is a risk factor for cognitive decline, whereas others suggest that cognitive decline precedes frailty (38). Recently, we found that impaired CVR is associated with poorer cognitive performance, particularly in the executive function domain (10). Balestrini and co-workers found that severe carotid stenosis and reduced ipsilateral CVR significantly influenced the occurrence of cognitive decline in stroke-free patients (11). In our study, the association between CVR and frailty persisted despite adjustment for the global cognitive score, highlighting the importance of CVR as a marker in the pathophysiological mechanisms of frailty. There are some plausible biological pathways through which vascular disease and frailty are interrelated. Frailty and CVD share common biological pathways, and CVD may accelerate the development of frailty. Endothelial dysfunction may be an underlying mechanism for frailty (39). The relationship between impaired CVR and frailty may be explained by cerebral and systemic vascular disease, with decreasing total physiological reserve (13, 14). Impaired CVR may be casually related to frailty or alternatively serve as a marker for systemic microvascular disease (13). Nevertheless, frailty may represent another less recognized manifestation of vascular disease and thus explain, at least in-part, several of their adverse health-related outcomes (40). Our study presents some strengths and limitations. The study includes a unique dataset of extensively studied patients who underwent a comprehensive and validated cognitive evaluation as well as assessments of cIMT and CVR. The study has several limitations. First, the causal effect of vascular abnormality on frailty cannot be evaluated because we did not have a prior assessment of frailty status at the time of CVR assessment. Second, CVR was measured by TCD using breath-holding which is a simple and easy to implement tool; yet, more accurate TCD and MRI-based measures may be available. Furthermore, as expected, TCD measurements failed in approximately 7.5 per cent of our patients, due to absence of an acoustic window. The effect of misclassification of CVR may, however, reduce the strength of our reported associations. Third, the generalizability of the study might be limited to men with pre-existing CVD and specific clinical characteristics as included in the current study. Patients with a high burden of vascular risk factors are, however, at a higher risk of developing frailty. Finally, we have not quantified systematically the degree of carotid stenosis, yet only a small minority had hemodynamically significant carotid stenosis. In summary, we observed a relationship between poorer cerebrovascular status, in-particular impaired CVR, and late-life frailty. Frailty increases the risk of adverse health outcomes including decreased quality of life, disability, recurrent hospitalizations, and death. These findings provide support that cerebral microvascular disease is involved in the pathophysiology of physical frailty and suggest an added value of assessing the cerebral vascular functional status for identifying patients at-risk to develop frailty. Supplementary Material Supplementary data are available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online. Authors’ Contributions All authors contributed to study conception, data analysis, and interpretation. All authors contributed to revising the manuscript for publication. All authors have approved the final version for publication. Conflict of interest None declared. References 1. Fried LP, Tangen CM, Walston Jet al.   Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci . 2001; 56: M146– 156. Google Scholar CrossRef Search ADS PubMed  2. Collard RM, Boter H, Schoevers RA, Oude Voshaar RC. Prevalence of frailty in community-dwelling older persons: a systematic review. J Am Geriatr Soc . 2012; 60: 1487– 1492. doi: 10.1111/j.1532-5415.2012.04054.x Google Scholar CrossRef Search ADS PubMed  3. Afilalo J. Frailty in patients with cardiovascular disease: why, when, and how to measure. Curr Cardiovasc Risk Rep . 2011; 5: 467– 472. doi: 10.1007/s12170-011-0186-0 Google Scholar CrossRef Search ADS PubMed  4. Veronese N, Cereda E, Stubbs Bet al.   Risk of cardiovascular disease morbidity and mortality in frail and pre-frail older adults: Results from a meta-analysis and exploratory meta-regression analysis. Ageing Res Rev . 2017; 28: 63– 73. doi: 10.1016/j.arr.2017.01.003 Google Scholar CrossRef Search ADS   5. Gale CR, Cooper C, Sayer AA. Framingham cardiovascular disease risk scores and incident frailty: the English longitudinal study of ageing. Age (Dordr) . 2014; 36: 9692. doi: 10.1007/s11357-014-9692-6 Google Scholar CrossRef Search ADS PubMed  6. Newman AB, Gottdiener JS, Mcburnie MAet al.  ; Cardiovascular Health Study Research Group. Associations of subclinical cardiovascular disease with frailty. J Gerontol A Biol Sci Med Sci . 2001; 56: M158– M166. Google Scholar CrossRef Search ADS PubMed  7. Avila-Funes JA, Meillon C, González-Colaço Harmand M, Tzourio C, Dartigues JF, Amieva H. Association between frailty and carotid central structure changes: the Three-City Study. J Am Geriatr Soc . 2014; 62: 1906– 1911. doi: 10.1111/jgs.13062 Google Scholar CrossRef Search ADS PubMed  8. Avila-Funes JA, Pelletier A, Meillon Cet al.   Vascular Cerebral Damage in Frail Older Adults: The AMImage Study. J Gerontol A Biol Sci Med Sci . 2017; 72: 971– 977. doi: 10.1093/gerona/glw347 Google Scholar CrossRef Search ADS PubMed  9. Siejka TP, Srikanth VK, Hubbard REet al.   Frailty and cerebral small vessel disease: A cross-sectional analysis of the Tasmanian Study of Cognition and Gait (TASCOG). J Gerontol A Biol Sci Med Sci . 2017. 2018;73:255-260. doi: 10.1093/gerona/glx145 10. Haratz S, Weinstein G, Molshazki Net al.   Impaired cerebral hemodynamics and cognitive performance in patients with atherothrombotic disease. J Alzheimers Dis . 2015; 46: 137– 144. doi: 10.3233/JAD-150052 Google Scholar CrossRef Search ADS PubMed  11. Balestrini S, Perozzi C, Altamura Cet al.   Severe carotid stenosis and impaired cerebral hemodynamics can influence cognitive deterioration. Neurology . 2013; 80: 2145– 2150. doi: 10.1212/WNL.0b013e318295d71a Google Scholar CrossRef Search ADS PubMed  12. Gupta A, Chazen JL, Hartman Met al.   Cerebrovascular reserve and stroke risk in patients with carotid stenosis or occlusion: a systematic review and meta-analysis. Stroke . 2012; 43: 2884– 2891. doi: 10.1161/STROKEAHA.112.663716 Google Scholar CrossRef Search ADS PubMed  13. Portegies ML, de Bruijn RF, Hofman A, Koudstaal PJ, Ikram MA. Cerebral vasomotor reactivity and risk of mortality: the Rotterdam study. Stroke . 2014; 45: 42– 47. doi: 10.1161/STROKEAHA.113.002348 Google Scholar CrossRef Search ADS PubMed  14. Soumare A, Elbaz A, Zhu Yet al.   White matter lesions volume and motor performances in the elderly. Ann Neurol . 2009; 65: 706– 715. doi: 10.1002/ana.21674 Google Scholar CrossRef Search ADS PubMed  15. Lorenz MW, Polak JF, Kavousi Met al.  ; PROG-IMT Study Group. Carotid intima-media thickness progression to predict cardiovascular events in the general population (the PROG-IMT collaborative project): a meta-analysis of individual participant data. Lancet . 2012; 379: 2053– 2062. doi: 10.1016/S0140-6736(12)60441-3 Google Scholar CrossRef Search ADS PubMed  16. Dahl A, Lindegaard KF, Russell Det al.   A comparison of transcranial Doppler and cerebral blood flow studies to assess cerebral vasoreactivity. Stroke . 1992; 23: 15– 19. Google Scholar CrossRef Search ADS PubMed  17. Goldbourt U, Behar S, Reicher-Reiss Het al.   Rationale and design of a secondary prevention trial of increasing serum high-density lipoprotein cholesterol and reducing triglycerides in patients with clinically manifest atherosclerotic heart disease (the Bezafibrate Infarction Prevention Trial). Am J Cardiol . 1993; 71: 909– 915. Google Scholar CrossRef Search ADS PubMed  18. Goldbourt U, Brunner D, Behar S, Reicher-Reiss H. Baseline characteristics of patients participating in the Bezafibrate Infarction Prevention (BIP) Study. Eur Heart J . 1998; 19 Suppl H: H42– H47. Google Scholar PubMed  19. Osman A, Walsemann KM. Ethnic disparities in disability among middle-aged and older Israeli adults: the role of socioeconomic disadvantage and traumatic life events. J Aging Health . 2013; 25: 510– 531. doi: 10.1177/0898264313478653 Google Scholar CrossRef Search ADS PubMed  20. Müller M, Voges M, Piepgras U, Schimrigk K. Assessment of cerebral vasomotor reactivity by transcranial Doppler ultrasound and breath-holding. A comparison with acetazolamide as vasodilatory stimulus. Stroke . 1995; 26: 96– 100. Google Scholar CrossRef Search ADS PubMed  21. Grant EG, Benson CB, Moneta GLet al.   Carotid artery stenosis: grayscale and Doppler ultrasound diagnosis—society of radiologists in ultrasound consensus conference. Ultrasound Q . 2003; 19: 190– 198. Google Scholar CrossRef Search ADS PubMed  22. Touboul PJ, Hennerici MG, Meairs Set al.   Mannheim carotid intima-media thickness consensus (2004–2006). An update on behalf of the advisory board of the 3rd and 4th watching the risk symposium, 13th and 15th European Stroke Conferences, Mannheim, Germany, 2004, and Brussels, Belgium, 2006. Cerebrovasc Dis . 2007; 23: 75– 80. doi: 10.1159/000097034 Google Scholar CrossRef Search ADS PubMed  23. Cesari M, Kritchevsky SB, Penninx BWet al.   Prognostic value of usual gait speed in well-functioning older people–results from the Health, Aging and Body Composition Study. J Am Geriatr Soc . 2005; 53: 1675– 1680. doi: 10.1111/j.1532-5415.2005.53501.x Google Scholar CrossRef Search ADS PubMed  24. Radloff L. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Measure . 1977; 1: 385– 401. Google Scholar CrossRef Search ADS   25. Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol . 1993; 46: 153– 162. Google Scholar CrossRef Search ADS PubMed  26. Almeida OP, Almeida SA. Short versions of the geriatric depression scale: a study of their validity for the diagnosis of a major depressive episode according to ICD-10 and DSM-IV. Int J Geriatr Psychiatry . 1999; 14: 858– 865. Google Scholar CrossRef Search ADS PubMed  27. Dwolatzky T, Whitehead V, Doniger GMet al.   Validity of a novel computerized cognitive battery for mild cognitive impairment. BMC Geriatr . 2003; 3: 4. doi: 10.1186/1471-2318-3-4 Google Scholar CrossRef Search ADS PubMed  28. Silvestrini M, Vernieri F, Pasqualetti Pet al.   Impaired cerebral vasoreactivity and risk of stroke in patients with asymptomatic carotid artery stenosis. JAMA . 2000; 283: 2122– 2127. Google Scholar CrossRef Search ADS PubMed  29. Long JS. Regression Models for Categorical and Limited Dependent Variables . Thousand Oaks, CA: Sage; 1997. 30. Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology . 2000; 11: 550– 560. Google Scholar CrossRef Search ADS PubMed  31. Myers V, Drory Y, Goldbourt U, Gerber Y. Multilevel socioeconomic status and incidence of frailty post myocardial infarction. Int J Cardiol . 2014; 170: 338– 343. doi: 10.1016/j.ijcard.2013.11.009 Google Scholar CrossRef Search ADS PubMed  32. Ng TP, Feng L, Nyunt MSet al.   Nutritional, physical, cognitive, and combination interventions and frailty reversal among older adults: a Randomized Controlled Trial. Am J Med . 2015; 128: 1225– 1236.e1. doi: 10.1016/j.amjmed.2015.06.017 Google Scholar CrossRef Search ADS PubMed  33. Cameron ID, Fairhall N, Langron Cet al.   A multifactorial interdisciplinary intervention reduces frailty in older people: randomized trial. BMC Med . 2013; 11: 65. doi: 10.1186/1741-7015-11-65 Google Scholar CrossRef Search ADS PubMed  34. Le Couteur DG, Anderson RM, Newman AB, de Cabo R. Stem cell transplantation for frailty. J Gerontol A Biol Sci Med Sci . 2017; 72: 1503– 1504. doi: 10.1093/gerona/glx158 Google Scholar CrossRef Search ADS PubMed  35. Chung CP, Chou KH, Chen WTet al.   Cerebral microbleeds are associated with physical frailty: a community-based study. Neurobiol Aging . 2016; 44: 143– 150. doi: 10.1016/j.neurobiolaging.2016.04.025 Google Scholar CrossRef Search ADS PubMed  36. Buchman AS, Yu L, Boyle PAet al.   Microvascular brain pathology and late-life motor impairment. Neurology . 2013; 80: 712– 8. doi: 10.1212/WNL.0b013e3182825116 Google Scholar CrossRef Search ADS PubMed  37. Purkayastha S, Fadar O, Mehregan Aet al.   Impaired cerebrovascular hemodynamics are associated with cerebral white matter damage. J Cereb Blood Flow Metab . 2014; 34: 228– 234. doi: 10.1038/jcbfm.2013.180 Google Scholar CrossRef Search ADS PubMed  38. Searle SD, Rockwood K. Frailty and the risk of cognitive impairment. Alzheimers Res Ther . 2015; 7: 54. doi: 10.1186/s13195-015-0140-3 Google Scholar CrossRef Search ADS PubMed  39. Alonso-Bouzón C, Carcaillon L, García-García FJ, Amor-Andrés MS, El Assar M, Rodríguez-Mañas L. Association between endothelial dysfunction and frailty: the Toledo Study for Healthy Aging. Age (Dordr) . 2014; 36: 495– 505. doi: 10.1007/s11357-013-9576-1 Google Scholar CrossRef Search ADS PubMed  40. Gouw AA, Seewann A, van der Flier WMet al.   Heterogeneity of small vessel disease: a systematic review of MRI and histopathology correlations. J Neurol Neurosurg Psychiatry . 2011; 82: 126– 135. doi: 10.1136/jnnp.2009.204685 Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

The Journals of Gerontology Series A: Biomedical Sciences and Medical SciencesOxford University Press

Published: Feb 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off