Search

Background: Adults with cardiovascular co-morbidities and risk factors are at greater risk of severe COVID-19. These same risk factors may also be associated with an attenuated antibody response to COVID-19 vaccines, although studies in diverse, U.S. population-based cohorts have been limited. Methods: The Collaborative Cohort of Cohorts for COVID-19 Research (C4R) conducted a serosurvey for SARS-CoV-2 antibodies via dried blood spot (DBS) in 14 U.S. cohorts. IgG antibodies to SARS-CoV-2 spike subunit 1 (S1) and nucleocapsid (N) were measured from DBS using a semi-quantitative microsphere immunoassay and reported as median fluorescence intensity (MFI). Multivariable adjusted linear models regressed log-transformed anti-S1 MFI on age, sex, race/ethnicity, education attainment, self-reported diabetes, hypertension, cardiovascular disease (CVD), chronic kidney disease, smoking history, body mass index (BMI), asthma, obstructive lung diseases, DBS batch, anti-N MFI, vaccine type, time between vaccine and DBS, and vaccine dose at time of DBS collection. Results are presented as the percent difference in anti-S1 MFI compared with a reference group. Results: There were 6614 vaccinated participants prior to booster regimens and DBS collection (April 2021-July 2022) with 50%, 48%, and 2% of the cohort who received BNT162b2, mRNA-1273, or other vaccines, respectively. The mean (SD) time between vaccination and DBS was 3.8 (1.8) months. Over 10% of the cohort had self-reported a history of diabetes, 55% had hypertension, and 74% had a BMI>25 kg/m 2 . Anti-S1 MFI decreased as the time between vaccine dose and DBS collection increased. Diabetes was associated with a 16.1% lower anti-S1 MFI (95%CI:-22.4,-9.5) whereas neither hypertension (-3.8%;95%CI:-9.3,2.1), nor cardiovascular disease history (-5.3%;95%CI:-16.3,7.4) were associated with anti-S1 MFI. Former and current smoking history were each associated with a lower anti-S1 MFI: (-6.6%;95%CI:-12.1,-0.8) and (-16.1%;95%CI:-24.7,-6.6), respectively. Participants with a BMI 25-29.9 kg/m 2 had a 7.6% higher anti-S1 MFI (95%CI:0.3,15.4) whereas those with a BMI of 30-35 kg/m 2 and >35 kg/m 2 had 6.2% (95%CI:-2.4,15.5) higher and 8.9% lower (95%CI:-17.6,0.7) MFI levels, respectively. Older age and male sex were each associated with a lower anti-S1 MFI and mRNA-1273 vaccine, Asian subgroup, higher anti-N titer, and prior COVID-19 infection were each associated with higher anti-S1 MFI. Chronic kidney disease, education attainment, and lung disease were not associated with anti-S1 MFI. Conclusions: Several traditional cardiovascular disease risk factors were associated with diminished humoral responses to the initial COVID-19 vaccine regimens in a diverse U.S. population-based cohort and may have implications on strategies to improve vaccine responses.

Introduction: Long-term neurological consequences (eg, stroke, impaired cognition) have been linked to SARS-CoV-2 infection and severity. There are limited data from studies with racial, ethnic, socioeconomic, and geographic diversity. C4R is a prospective study of adults from 14 cohorts that aims to link pre-COVID phenotyping, including cognition (8 studies), to COVID related illness and sequelae. We aimed to conduct pre-statistical harmonization of cognitive tests administered in English and Spanish from 8 cohorts: ARIC, CARDIA, FHS, HCHS/SOL-INCA, MESA, NOMAS, REGARDS, and SHS (Table). Methods: We conducted extensive item-level review of administration, scoring, and coding procedures and score distributions for 84 tests administered in English (all studies) and Spanish (NOMAS, MESA, HCHS/SOL-INCA). Results: Orientation to time and 3-word registration and recall spanned all studies and both languages. Word list recall and verbal fluency (animal; letter) spanned 7 studies (Table). There was variability in the structure, content, administration, scoring, and data coding procedures for items across cohorts and between Spanish and English. Word lists varied by number of words (9-16) and learning trials (3-5). Animal naming varied by time (30 vs. 60 seconds), animal type (4-legged vs. any animal), and scoring (allowing mythical/imaginary animals). Letter fluency varied by whether both Spanish and English words were permitted. Other tests differed by version, study-specific adaptations, prompts/cues, and specificity of scoring rules across cohorts. Conclusions: Cognitive test harmonization requires detailed review of administration, scoring, coding, translation, and procedural differences. Accounting for this variability is essential to cognitive data interpretation. Our pre-statistical harmonization will inform data augmentation and formal harmonization to yield harmonized measures of cognition to clarify population-level differences in cognitive outcomes linked to SARS-CoV-2 infection.

Introduction: Larger brain arterial diameters, which may be a consequence of connective tissue disorders or hypertension, are associated with mortality, cognitive decline and risk of dementia. Whether there may be genetic determinants of large brain arterial diameters is unknown, but identifying such traits may shed light into possible mechanistic links between large brain arterial diameters and the observed outcomes. Methods: We studied 3654 participants from four ethnically diverse population-based cohorts: The Atherosclerosis Risk in Communities study (n=1565: 74.3% white, 25.7% black), The Northern Manhattan Study (n=1092: 13.8% white, 16.6% black, 69.6% Hispanic), Epidemiology of Dementia in Singapore (n=647: 31.7% Chinese, 35.7% Malay, 32.6% Indian), and Memory Clinic Cohort in Singapore (n=350: 84.9% Chinese, 7.8% Malay, 7.3% Indian). We measured brain arterial diameters for 13 segments and normalized the averaged diameter per participant to obtain global brain arterial diameter as dependent variable. Genome-wide association studies (GWAS) was performed in each study using linear regression, adjusted for sex, age, head size, number of missing arteries and population substructure. Meta-analysis of association statistics was performed using the METAL to generate combined trait-specific effect estimates. We also performed trans-ethnic meta-regression by MR-MEGA to account for the heterogeneity in allelic effect that is correlated with ancestry. Results: Genome-wide significant (P-8 ) associations were observed for global brain arterial diameter with three SNPs: rs7921574(P=1.2x10 -8 ) mapping to an intron of CNNM2; rs2075858(P=2.47x10 -8 ) and rs2075857 (P=2.37x10 -8 ) mapping to introns of MUC5B. MR-MEGA analysis showed suggestive evidence of association (P=5.05x10 -8 ) for rs7921574. These loci comprise genes involved in response to magnesium transport, neuronal transmission, hypertension, mucociliary clearance and host defense. Conclusion: Our study reveals two novel risk loci for large brain arterial diameters. We have begun to replicate these associations and to identify functional causal variants in independent studies. Identifying these loci may help us to elucidate underlying disease mechanisms.

Introduction: Age-related cognitive impairment is driven by the complex interplay of neurovascular and neurodegenerative disease. There is a strong relationship between cerebral microbleeds (CMBs), cerebral amyloid angiopathy (CAA), and the cognitive decline observed in conditions such as Alzheimer’s disease. In the early, preclinical phase of cognitive impairment, the extent to which CMBs and underlying CAA impact volumetric changed in the brain related to neurodegenerative disease remains unclear. Methods: We performed cross-sectional analyses from 3 large cohorts: The Northern Manhattan Study (NOMAS), Alzheimer’s Disease Neuroimaging Initiative (ADNI), and the Epidemiology of Dementia in Singapore study (EDIS). We conducted a confirmatory analysis of 82 autopsied cases from the Brain Arterial Remodeling Study (BARS). We implemented multivariate regression analyses to study the association between two related markers of cerebrovascular disease – MRI-based CMBs and autopsy-based CAA, as independent variables, and volumetric markers of neurodegeneration, as dependent variables. Results: We included 2657 participants with available MRI data and 82 autopsy cases from the BARS. In a meta-analysis of NOMAS, ADNI and EDIS, superficial CMBs were associated with larger gray (β=4.49 ± 1.13, P=0.04) and white (β=4.72 ± 2.1, P=0.03) matter volumes. The association between superficial CMBs and larger white matter volume was more evident in participants with one CMB (B=5.17 ± 2.47, P=0.04) than in those with ≥ 2 CMBs (β=1.97 ± 3.41, P=0.56). In the BARS, CAA was associated with increased cortical thickness (β = 6.5 ± 2.3, P=0.016) but not with increased brain weight (β=1.54 ± 1.29, P=0.26). Discussion: Superficial CMBs are associated with larger morphometric brain measures. This association is strongest in brains with fewer CMBs, suggesting that the CMBs/CAA contribution to neurodegeneration may not relate to tissue loss, at least in early stages of disease.

Introduction: Hospital-Acquired Infections (HAIs) are contracted during hospitalization for another disorder. We hypothesized that older adults with incident stroke and HAI experience faster subsequent cognitive decline than those having stroke without HAI and those without stroke. Methods: The Cardiovascular Health Study (CHS) is a cohort of Medicare-eligible participants aged ≥ 65 years. Incident ischemic stroke was ascertained from medical records. HAIs were assessed by hospital discharge ICD-9 codes. Global cognitive function was measured annually by modified mini-mental state examination (3MSE) and executive function by digit symbol substitution test (DSST). We used linear mixed models to estimate decline in mean 3MSE and DSST scores over a mean of seven years by incident stroke and HAI status among participants free of stroke at baseline. Results: In the 3MSE analysis, 393 participants had incident stroke and HAI, 766 had stroke only, and 4,284 had no stroke. With no stroke ( Fig 1A, black line), cognitive decline accelerated with age: mean declines of 0.27 points/year at age 70, 1.10 points/year at age 80, and 2.47 points/year at age 90. After stroke with HAI (red line), mean decline was 0.29 points/year faster than with no stroke (95% CI: 0.17, 0.40; p < 0.001). In contrast, after stroke without HAI (blue line), mean decline was similar to that with no stroke (0.09 points/year faster; 95% CI: -0.13, 0.32; p = 0.410). Mean decline was 0.19 points/year faster after stroke with HAI than after stroke without HAI, though these differences were not statistically significant (95% CI: -0.05, 0.44; p = 0.123). Mean decline in DSST score ( Fig 1B ) after stroke with HAI was similar to stroke without HAI (0.09 points/year faster decline after stroke with HAI; 95% CI: -0.08, 0.27; p = 0.297). Conclusions: Mean cognitive decline after stroke with HAI was faster than with no stroke but may be similar to that after stroke without HAI. As a limitation, we did not have data on stroke severity.

Introduction: Apart from stroke, cognitive change after incident myocardial infarction (MI) is unclear. We aim to determine associations between incident MI and acute and long-term changes in cognitive function, controlling for risk factors and pre-MI cognitive trajectories, censoring for stroke. Methods: We performed a pooled analysis of 31,377 participants, ≥18 years, free of MI and dementia, from 6 longitudinal cohort studies (1971-2017): Atherosclerosis Risk in Communities Study, Coronary Artery Risk Development in Young Adults Study, Cardiovascular Health Study, Framingham Offspring Study, Multi-Ethnic Study of Atherosclerosis, and Northern Manhattan Study. Outcomes were changes in global cognition (primary), memory, and executive function. Linear mixed-effects models were used to estimate the association between incident MI and acute cognitive decline at the time of MI (Model A), and decline in cognition over the years after MI (Model B). Results: Median study follow up was 6.4 years (IQR 4.9-19.7 years) and 1,047 participants had incident MI. Incident MI was associated with significant acute decline in global cognition and executive function, but not memory, after the MI event (Models A, Table 1). After including change in cognitive function (slope) after incident MI in the model, the effect estimates indicating acute declines in global cognition and executive function were not significant (Models B, Table 1). However, participants with incident MI demonstrated significantly faster declines in global cognition (-0.15 points/year faster [95% CI, -0.21 to -0.10]), memory (-0.13 points/year faster [95% CI, -0.23 to -0.04]), and executive function (-0.14 points/year faster [95% CI, -0.20 to -0.08]) compared to those without MI (Models B, Table 1). Conclusions: Incident MI is associated with faster declines in global cognition, memory, and executive function over years following the MI event.

Introduction: We previously showed that patients with cancer and acute ischemic stroke (AIS) have higher markers of coagulation, platelet, and endothelial activation, and more circulating microemboli than patients with AIS and no cancer. Herein, we evaluate whether these markers are associated with clinical outcomes in patients with cancer and AIS. Methods: In the MOST-Cancer prospective study, we enrolled 50 adults with active solid-tumor cancer and AIS at 2 New York hospitals from 2016-2020. We collected blood from participants 72-120 hours after stroke onset. We performed a 30-minute transcranial Doppler (TCD) microemboli detection study a median of 4 days after stroke onset. The exposure variables were hematological markers of coagulation (D-dimer, thrombin-antithrombin), platelet (P-selectin), and endothelial activation (thrombomodulin, soluble intercellular adhesion molecule-1 [sICAM-1], soluble vascular cell adhesion molecule-1 [sVCAM-1]), and the presence of microemboli on TCD. The primary outcome was a composite of recurrent arterial or venous thromboembolism or death. The secondary outcome was recurrent AIS. We used Cox regression to evaluate individual associations between biomarkers and subsequent outcomes. Results: Median age was 69 years (IQR, 60-76) and 24 participants (48%) were women. The most common cancers were lung (28%) and pancreatic (22%); 86% of participants had distant metastases. During a mean follow-up of 278 days (SD, 367), 43 (86%) participants had the primary outcome of recurrent thromboembolism or death, including 28 (56%) who had recurrent thromboembolism, 13 who had recurrent AIS (26%), and 30 (60%) who died. D-dimer, P-selectin, sICAM-1, sVCAM-1, and microemboli were associated with the composite outcome (Table). Similar results were seen for recurrent AIS. Conclusion: Markers of hypercoagulability and embolic disease were associated with adverse clinical outcomes in cancer-related stroke.

Background: Recent data suggest biomarkers of atrial cardiopathy (ACP) can potentially identify a subset of patients with cryptogenic stroke who may benefit from anticoagulation. However, the specificity of these markers for atrial disease and their independence from measures of more global cardiac dysfunction remain unclear. We hypothesized that biomarkers of ACP collectively reflect left atrial pathology and are not strongly affected by coexisting left ventricular dysfunction. Methods: ARCADIA is an ongoing multicenter trial comparing standard dose apixaban to aspirin therapy in patients with Embolic Stroke of Undetermined Source (ESUS) with ACP, defined as either N-terminal pro-brain natriuretic peptide [NT-proBNP] > 250 pg/mL; P wave terminal force velocity on ECG lead V1 [PTFV1] > 5000 μV*ms; or echocardiographic left atrial diameter index [LADI] ≥ 3 cm/m 2 . Pearson correlation coefficients and descriptive statistics were used to examine the relationship between left ventricular ejection fraction (LVEF) and ACP biomarkers. Results: Among 2,616 ESUS patients screened, 1,046 met ≥ 1 ACP criterion (665 BNP, 531 PTFV1, 14 LADI). Patients with ACP were older (68 ± 11 vs 64 ± 10 years) and 54.5% were women. The great majority (84%) of subjects with ACP had normal LVEF (≥55%). Among them there was no correlation between LVEF and NT-proBNP (r= -0.01, P= 0.5) or PTFV1 (r=0.002, P=0.9). In patients with abnormal LVEF, there was a strong inverse correlation of LVEF with NT-proBNP (r=-0.4, P250 pg/mL and measured LVEF (n=435), only 19% had abnormal LVEF. Compared to patients with abnormal NT-proBNP alone, those with both abnormal NT-proBNP and abnormal LVEF were younger (66.4 ± 11.0 vs. 71.6 ± 11.0, P Conclusion: Left ventricular dysfunction (EF ≤ 55%) is uncommon in patients with ACP as defined in ARCADIA. These findings suggest that, among patients with ESUS, atrial cardiopathy may represent a separate entity from LV dysfunction.

Background: The presence of cardiovascular risk factors and associated diseases, including heart failure (HF), has been associated with worse outcomes following stroke. It is unclear however whether the association between HF and poor outcomes persists in younger cohorts, in whom HF is less often a consequence of long-standing cardiovascular disease. We hypothesize prevalent heart failure has differential associations with post-stroke functional outcomes across age groups. Methods: We performed a retrospective analysis of patients admitted with acute ischemic stroke at Columbia University Medical Center between 1990-2020 for whom adjudicated data was available at the time of analyses. Demographic, clinical variables including NIHSS and TPA administration, and functional outcomes were extracted. Good functional outcome was defined as a discharge modified Rankin score ≤ 2. Multivariable regression models were used to determine adjusted odds ratios (aORs) and 95% confidence intervals (95% CI). Results: Of the 855 patients hospitalized with ischemic stroke, 47% were women, 32% were ≤ 60 years, and 139 (16%) had pre-existing heart failure. Good functional outcome was seen in 456 (53%), in patients who had a mean age of 67.2 (16.4) years. Age as a continuous measure modified the effect of prevalent heart failure on the odds of good functional outcome (P = 0.001 for interaction). Among patients ≤ 60 years, after controlling for cardiovascular risk factors, NIHSS and TPA use, prevalent heart failure was associated with greater odds of good functional outcome (aOR: 1.33, 95% CI: 1.08-13.14) compared to young adults without HF. Among older stroke patients, HF was non-significantly associated with lower odds of good functional outcome (aOR: 0.86, 95% CI: 0.50-1.50). Conclusion: Premature heart failure (often due to congenital defects, hereditary cardiomyopathies, infections or other non-vascular etiologies) in younger stroke patients may be a marker of lower cardiovascular disease burden compared to young adults with stroke secondary to traditional stroke risk factors. These factors may play a larger role in post-stroke outcomes and recovery.

Introduction: Coronavirus Disease 2019 (COVID-19) is associated with an increased risk of stroke and worse stroke outcomes. A clinical score that can identify high-risk patients could enable closer monitoring and targeted preventative strategies. Methods: We used data from the AHA’s COVID-19 CVD Registry to create a clinical score to predict the risk of stroke among patients hospitalized with COVID-19. We included patients aged >18 years who were hospitalized with COVID-19 at 122 centers from March 2020-March 2021. To build our score, we used demographics, preexisting comorbidities, home medications, and vital sign and lab values at admission. The outcome was a cerebrovascular event, defined as any ischemic or hemorrhagic stroke, TIA, or cerebral vein thrombosis. We used two separate analytical approaches to build the score. First, we used Cox regression with cross validation techniques to identify factors associated with the outcome in both univariable (p Results: Among 21,420 patients hospitalized with COVID-19 (mean age 61 years, 54% men), 312 (1.5%) had a cerebrovascular event. Using traditional Cox regression, we created and internally validated a risk stratification score (CANDLE) (Fig) with a C-statistic of 0.66 (95% CI, 0.60-0.72). The machine learning estimator had similar discriminatory performance, with a C-statistic of 0.69 (95% CI, 0.65-0.72). For ischemic stroke or TIA, CANDLE’s C-statistic was 0.67 (95% 0.59-0.76). Conclusion: We developed an easy-to-use clinical score, with similar performance to a machine learning estimator, to help stratify stroke risk among patients hospitalized with COVID-19.

Introduction: Stroke is one of the leading causes of death worldwide. Our study aimed to investigate clinical variables that influence long-term mortality. Methods: A retrospective study was done in patients admitted with stroke to Columbia University Irving Medical Center - New York Presbyterian Hospital between 1990 and 2020 from whom data adjudication and long-term outcome were available at the time of analyses. Stroke was classified as cardioembolic, large vessel, small vessel and cryptogenic. We established 5 models which were progressively adjusted for cardiovascular risk factors and treatment modalities. Data was analyzed using cox proportional regression analysis to obtain hazard ratios (HR) with 95% confidence intervals (CI). Results: The study included 775 patients with a mean age of 67±16 years; 47% were men, 32% had diabetes, 39% had dyslipidemia and 22% had a history of atrial fibrillation (AF). On discharge, 84% received statin therapy. After 8.9±5.9 years (0-17 years) of follow-up, 257 (33.2%) of the initial cohort died. In univariate analysis, stroke due to AF was the only predictor of long-term mortality compared to other mechanisms (HR 1.9, 95% CI 1.5-2.3), but the effect size and statistical significance were progressively attenuated after adjusting for confounders (HR 1.2, 95% CI 0.9-1.6). In the fully adjusted model, statin therapy on discharge (HR 0.6, 95% CI 0.5-0.9) was associated with decreased long-term mortality whereas higher NIHSS score on admission, older age and a history of diabetes were associated with higher long-term mortality (Table 1). Conclusion: Comorbidities, and not stroke mechanism per se, are determinants of long-term mortality after stroke. Advanced age, diabetes mellitus, high NIHSS score are associated with higher long-term mortality whereas statin prescription upon discharge is associated with lower mortality. Post-stroke care strategies need to be targeted to reduce the risk of mortality in high-risk patients.

Introduction: Neurologic deterioration (ND) occurs in a quarter of acute lacunar infarcts and may lead to severe disability, but underlying pathophysiology of ND remains poorly understood. We sought to identify risk factors and clinical characteristics associated with ND. Methods: This retrospective observational study included consecutive patients with acute lacunar stroke admitted to New York University Langone Health and Brown University (NYU/Brown). Lacunar stroke was defined as a subcortical infarct 50%) in the parent vessel and no cardioembolic source. We defined ND as any neurologic deterioration referable to the lacunar stroke and not related to a medical or non-cerebrovascular neurological complication. We used logistic regression analyses to determine associations between ND and luminal stenosis ( Results: The NYU/Brown sample included 205 patients, of whom 41 (20%) had ND. In adjusted models, we found no definite association between luminal stenosis and ND (aOR 1.74, 95% CI 0.73-4.14). From CU, 361 total patients were included, of whom 59 (16%) had ND. In adjusted models, we found an association between luminal stenosis and ND (aOR 2.28 95% CI 1.15-4.50). Meta-analysis of both cohorts found luminal stenosis associated with ND (RR [95% CI]: 1.75 [1.23-2.49]). Conclusions: In this multi-center study, luminal stenosis (

Introduction: Associations between post-stroke vascular risk factors levels with post-stroke cognitive decline are uncertain. We determined associations between systolic blood pressure (BP), glucose, and low-density lipoprotein (LDL) cholesterol levels with post-stroke cognitive decline. Hypothesis: Higher post-stroke systolic BP, glucose, and LDL cholesterol levels are associated with faster post-stroke cognitive decline. Methods: This cohort study pooled data from 832 participants 18 and older with incident stroke and free of dementia before stroke (50% women and 32% Black) from 4 cohort studies from 1971-2019: Atherosclerosis Risk in Communities Study, Cardiovascular Health Study, Framingham Offspring Study, and REasons for Geographic and Racial Differences in Stroke Study. Linear mixed-effects models were used to estimate longitudinal changes in cognition after incident stroke. The primary outcome was change in global cognition. Global cognition was harmonized across studies and standardized as T-scores (mean [SD], 50 [10]); a 1-point difference represents a 0.1-SD difference in cognition. Median (IQR) follow-up was 5.5 (3.3-8.3) years. Results: In separate models, higher post-stroke glucose and LDL levels, but not systolic BP, were associated with faster declines in global cognition (Models M1a, M1b, and M1c, Table 1). In a model with all three vascular risk factors, higher post-stroke mean glucose and LDL levels were associated with faster global cognition decline, though not to traditional levels of statistical significance; however, post-stroke mean systolic BP was not associated with global cognition decline (Model M2, Table 1). Older age was associated with faster cognitive decline in all models. Conclusions: Our results suggest that higher post-stroke glucose and LDL levels are associated with faster post-stroke cognitive decline. We found no evidence that post-stroke BP levels are associated with post-stroke cognitive decline.

Importance: Many cardiac arrest (CA) survivors develop posttraumatic stress disorder (PTSD) symptoms (i.e., CA-induced PTSD). The association of CA-induced PTSD with secondary mortality and cardiovascular risk is unknown. Methods: A prospective, observational, cohort study of adults with return of spontaneous circulation after in-hospital or out-of-hospital CA at a tertiary-care center between September 2015 and September 2017. A consecutive sample of survivors with sufficient mental status to self-report CA-induced PTSD symptoms at hospital discharge were included. The combined primary end point was all-cause mortality (ACM) or a major adverse cardiovascular event (MACE)—hospitalization for nonfatal MI, unstable angina, congestive heart failure (CHF), urgent/emergency coronary revascularization procedures, or urgent implantable defibrillators/pacemaker (ICDs/PPM) placements within 12 months of discharge. In-person assessments for CA-induced PTSD were performed within 24 hours of hospital discharge. PTSD symptomatology was assessed via the PTSD Checklist - Specific (PCL-S) scale; a suggested diagnostic cut-off of 36 for specialized medical settings was adopted. Outcomes for patients meeting (vs. not meeting) this cutoff were compared using Cox regression with adjustment for demographic and clinical covariates. Results: Of 114 included patients, 36 (31.6%) screened positive for CA-induced PTSD at discharge (median 21 days post-CA; interquartile range 11-36). During the follow-up period (median = 12.4 months, range 10.2-13.5), 10 (8.8%) died and 29 (25.4%) experienced a recurrent MACE: rehospitalizations due to MI (n=4, 13.8%), UA (n=8, 27.6%), CHF exacerbations (n=4, 13.8%), emergency revascularizations (n=5, 17.2%), and ICDs/PPM placements (n=8, 27.6%). CA-induced PTSD was associated with ACM/MACE in univariate (Hazard Ratio [HR] =3.19; 95% confidence interval [CI], 1.7-6.0) and in models adjusted for age, sex, charlson comorbidity index and nonshockable initial rhythms (HR=3.2; CI, 1.7-6.1). Conclusions: PTSD is common after CA, and survivors with CA-induced PTSD had significantly higher risk of death and cardiovascular events. Further inquiry into underlying mechanisms is warranted.

Background and Purpose: Recent drug trials have challenged the “HDL-cholesterol anti-atherosclerotic hypothesis”, suggesting that total levels of HDL-C may not be best target for intervention. HDL-cholesterol (HDL-C) subfractions may be better markers of vascular risk than total levels of HDL-C. The objective of this cross-sectional study was to investigate the relation between HDL2-C and HDL3-C fractions and carotid intima-media thickness (cIMT) in the population-based Northern Manhattan Study. Methods: We evaluated 988 stroke-free participants (mean age 66±8 years; 60% women; 66% Hispanic, and 34% Non-Hispanic) with available data on HDL-C subfractions using precipitation method and cIMT assessed by a high-resolution carotid ultrasound. The associations between HDL-C subfractions and cIMT were analyzed by multiple linear regression models. Results: The mean HDL2-C was 14 ± 8 mg/dl, HDL3-C 32 ± 8 mg/dl, and the mean total HDL-C was 46 ±14 mg/dl. The mean cIMT was 0.90 ± 0.08 mm. After controlling for demographics and vascular risk factors, HDL2-C and total HDL-C were inversely associated with cIMT (per 2SDs, beta= -0.012, p=0.03 and beta= -0.012, p=0.04, respectively). The same inverse association was more pronounced among those with diabetes (per 2SDs, HDL2-C: beta= -0.036, p=0.04; HDL-C: beta= -0.032, p=0.04). HDL3-C was inversely associated with cIMT but the association was not significant (per 2SDs, beta= -0.008, p=0.13) after adjustment. Conclusions: HDL2-C had greater effect on cIMT then HDL3-C in this urban population. This effect was especially pronounced among individuals with diabetes. More research in needed to determine anti-atherosclerotic effects of HDL-C subfractions and clinical relevance of raising their levels.

Introduction: The fastest growing segment of the population is those over age 80, in whom there is the highest incidence of stroke. Few clinical trials, however, have targeted this population, and management of risk factors is complicated by higher rates of side effects and polypharmacy. We aimed to characterize whether the risk factors for stroke are different in those age >80 versus Methods: The study sample derives from the Northern Manhattan Study (n=3298), a prospective cohort study among an urban dwelling multi-ethnic population. At enrollment risk factors were assessed by self-report, physical examination, and laboratory data. Leisure-time physical activity (LTPA) was assessed by validated questionnaire and categorized as none versus any. Participants are followed annually for incident stroke which is adjudicated by study neurologists after review of all records and imaging. We fit Cox proportional hazard models to calculate hazard ratios (HR) and 95% confidence intervals (95% CI) for the association of cardiovascular disease risk factors (LTPA, reported alcohol use, fasting HDL-C and LDL-C, history of diabetes, hypertension, heart disease, and tobacco use) with risk of stroke. Multi-variable models also included two-way interaction terms between the primary exposures and age ( Results: The mean age was 69 years and 562 (17%) were age ≥80 years. Physical inactivity was common (41%) and did not differ by age. Subjects were followed for a median of 14 years, and there were 391 total strokes. We found a statistically significant interaction of age ≥80 on the risk of stroke with physical inactivity (p=0.02), diabetes (p=0.04), hypertension (p=0.005), and LDL-C (p=0.03). In stratified models no LTPA versus any (adjusted HR 1.61, 95%CI 1.06-2.43) was associated with an increased risk of stroke among those age ≥80 but not in those younger; LDL-C, hypertension and diabetes were mainly associated with risk of stroke in those below age 80. Conclusion: Lack of leisure-time physical activity is a potentially modifiable stroke risk factor and was more notable than hypertension, diabetes or dyslipidemia among those older than age 80. Improving activity may reduce the risk of stroke in this segment of the population.

Background: There is growing evidence that an increase in dietary sodium increases morbidity and mortality from cardiovascular disease (CVD) at least in part via an increase in blood pressure (BP). Several researchers have identified an association of higher dietary potassium, seen with increased intake of fruits and vegetables, with lower BP. Moreover, there may be a potential interaction of sodium and potassium on CVD health, such that the relative levels of these variables may be more important than their absolute levels. The goal of this study was to determine the association of dietary sodium to potassium ratio (Na:K) with the risk of stroke and CVD in a multi-ethnic urban dwelling population. Methods: Stroke-free participants from the Northern Manhattan Study, a population-based cohort study of stroke incidence, were followed prospectively to assess incident stroke, myocardial infarction and vascular death. Baseline food-frequency questionnaires were analyzed for Sodium and Potassium intake based on self-reported food consumption using DIETSYS software (Block Dietary Data System: Dietsys+ analysis software, version 59, 1999). We examined the association of Na:K with incident stroke (ischemic and overall) as well as combined vascular events (CVE)(stroke, myocardial infarction and vascular death) using Cox-proportional Hazards models adjusting for age, sex, race/ethnicity, total daily calories, adherence to a Mediterranean-style diet, smoking, physical activity, BMI, diabetes, hypertension, and hypercholesterolemia. Results: 2570 participants had dietary data, (mean age 69 ± 10 years, 64% women, 21% white, 53% Hispanic, 24% black); the mean of the Na:K ratio intake was 1.22 with a standard deviation of 0.43. Over a mean follow-up time of 11.9 years (SD=5.4), 274 participants developed stroke (N=232 ischemic) and 745 had a vascular event overall. In adjusted models the Na:K intake ratio was associated with an increased risk for ischemic Stroke (HR=1.5, 95% CI: 1.2-2.0), all stroke (HR=1.5, 95% CI: 1.2-1.9) and CVE (HR=1.2, 95% CI: 1.0-1.4). These results were unchanged in models adjusting for sociodemographic variables only. Conclusions: Our findings suggest that Na:K intake may be an important predictor of the risk of ischemic stroke and overall CVE.

Background: Chronic infections and hemodynamic dysfunction are risk factors for stroke. We hypothesized that blood biomarkers of infection (procalcitonin, or PCT) and hemodynamic stress (midregional pro-atrial natriuretic peptide, or MRproANP) would be associated with subclinical cerebrovascular damage (SCVD), including silent brain infarcts (SBI) and white matter hyperintensity volume (WMHV). Methods: The Northern Manhattan Study was designed to assess risk factors for incident vascular disease in a multiethnic cohort. A subsample underwent brain MRI and had blood samples available for biomarker measurement (n=1178). We used logistic regression models to estimate the odds ratios and 95% confidence intervals (OR, 95%CI) for the association of these biomarkers with SBI, and linear regression for associations with logWMHV, after adjusting for demographic, behavioral, and medical risk factors. Results: Mean age was 70 ±9 years; 60% were women, 66% hispanic, 17% black, 15% white and 2% belonged to other ethnicities. After fully adjusting for risk factors, subjects with PCT or MRproANP in the top quartile, compared to the lowest quartile, were associated with greater odds of having SBI (adjusted OR for PCT 2.2, 95%CI 1.3-3.7; for MRproANP 3.3, 95%CI 1.7-6.3) and increased WMHV (adjusted mean change in logWMHV for PCT 0.29, 95%CI 0.13-0.44; for MRproANP 0.18,95%CI 0.003-0.36). Conclusion: Higher concentrations of PCT, a marker of infection, and MRproANP, a marker of hemodynamic stress, are independently associated with SCVD. Future biomarker- guided primary prevention studies may lead to new approaches to prevent cerebrovascular damage.

Background/Objective: Carotid artery intima media thickness (cIMT) may be a marker of cerebral atherosclerotic disease and therefore related to cognitive status in the elderly. We hypothesized that those with greater cIMT would exhibit worse cognition at baseline and have greater cognitive decline over time. Methods: A sample of 1166 stroke-free community participants from the Northern Manhattan Study (NOMAS) underwent carotid ultrasound and repeated neuropsychological (NP) testing. cIMT was imaged with a standardized B-mode ultrasound protocol and analyzed by a trained sonographer. We used multivariable linear regression to examine cIMT as a correlate of domain-specific NP Z-scores cross-sectionally and over time; we investigated possible effect modification by APOE ε4 allele, adjusting for demographics and vascular risk factors. Results: The mean age of participants was 71±9 years, 60% were women, 15% white, 18% black, 67% Hispanic, and mean cIMT was 0.93 ± 0.09 mm. NP testing was performed among 1166 participants at baseline and among 826 participants with NP follow-up at a mean of 6.2 years apart. Participants with greater cIMT exhibited worse episodic memory at baseline after adjustment for demographics and vascular risk factors (beta=-0.60, P=0.04). APOE ε4 allele presence was a significant effect modifier, and after stratification, APOE ε4 carriers with greater cIMT exhibited worse episodic memory (beta=-1.26, P=0.04), semantic memory (beta=-1.35, P=0.01), and processing speed (beta=-1.22, P=0.02) at baseline. In longitudinal analysis, participants with greater cIMT at baseline did not exhibit cognitive decline after adjustment for demographics and vascular risk factors, but APOE ε4 allele presence was a significant effect modifier. After stratification, APOE ε4 non-carriers with greater cIMT at baseline exhibited greater declines in executive function (beta=-1.03, P=0.05). Conclusions: cIMT may be associated with worse cognition and greater cognitive decline in multiple domains among those at high risk for Alzheimer disease, and with worsening executive function in APOE ε4 non-carriers. Interventions that target early stages of atherosclerosis may modify the course of cognitive aging and prevent cognitive decline.

Background: Prior studies have shown an increased risk of ischemic stroke (IS) after myocardial infarction (MI), particularly in the first few days after the event. There is limited evidence, however, on the long-term risk and whether it is directly related to cardiac injury. We hypothesized that the risk of IS after acute coronary syndrome (ACS) is significantly higher when there is evidence of cardiac injury such as ST elevation MI (STEMI) or non-STEMI (NSTEMI) than when there is no evidence of cardiac injury such as Unstable Angina (UA). Methods: Administrative claims data were obtained from all emergency department encounters and hospitalizations at California’s nonfederal acute care hospitals between 2008 and 2011. Patients with STEMI, NSTEMI, and UA were identified using appropriate ICD-9 codes. Age and ICD-9 codes for sex, race-ethnicity, hypertension, diabetes, hyperlipidemia, congestive heart failure, and atrial fibrillation were analyzed. The outcome was IS during 2 years follow-up. Unadjusted and adjusted Cox proportional hazards models were used to determine the association between ACS subtype and ischemic stroke risk. Results: We identified 74,962 ACS patients: 35.9% STEMI, 54.8% NSTEMI, and 9.3% UA. The mean age in years was 66.7 ± 14.4 and 61.5% females; 3.2% of patients had IS during the two year follow-up. When compared to UA, the long term risk of IS was higher in patients with STEMI (adjusted HR 2.07 95% CI 1.67-2.56) and NSTEMI (adjusted HR 1.87, 95% CI 1.51-2.30), even after adjusting for stroke risk factors at baseline and incident atrial fibrillation (Figure). Other stroke risk factors after ACS are shown in the Figure. Conclusion: Analysis from a large administrative dataset revealed that both NSTEMI and STEMI confer an increased risk of ischemic stroke, independent of risk factors, which may be related to cardiac injury. Studies exploring ischemic stroke mechanisms in cardiac patients are needed to improve and tailor stroke prevention strategies.

Introduction: Small perivascular spaces (SPVS) are gaining momentum as imaging biomarkers of cerebrovascular health. Hypothesis: SPVS confer vascular risks and the coexistence of SPVS with lacunar infarcts (LI) heightens these risks. Methods: Stroke-free participants in the population-based Northern Manhattan Study were followed for incident stroke (ischemic and hemorrhagic), MI, all death, vascular death, and any vascular event. Lesions with diameter of 3 mm or less and absence of FLAIR rim were classified as SPVS on a semi-quantitative scale (range 0 to 28). We defined “high SPVS burden” as the upper quintile and compared the rate of vascular events in this group to individuals in the lower 4 quintiles combined. LI were defined as lesions greater than 3 mm with associated FLAIR rim, round shape, and typical location. Cox models were used to calculate risks of outcomes after adjusting for confounders. Results: This analysis includes 1208 NOMAS participants (40% male, 65% Hispanic; mean age 71 ± 9 years at time of MRI) followed a mean of 6 ± 2 years. SPVS were present in 91% of the sample (median SPVS scale score 5). Compared to participants with a lesser burden of SPVS, participants with a high SPVS burden had a higher incidence rate per 1000 person-years of death (48 vs 34), vascular death (20 vs 11), ischemic stroke (12 vs 7) and any vascular event (37 vs 24). After adjusting for demographics and vascular risk factors, participants with a high burden of SPVS had a higher risk of death (HR 1.35, 1.00-1.78), vascular death (HR 1.55, 0.96-2.51), any stroke (HR 1.53, 0.91-2.57), MI (HR 1.29, 0.69-2.41), and any vascular event (HR 1.50, 1.07-2.11). The presence of lacunar infarcts was an effect modifier such that those with LI and a high SPVS burden had a greater risk of vascular death (B=0.63, P=0.03), any stroke (B=0.72, P=0.03) and any vascular event (B=0.54, P=0.02) compared to those without LI. Conclusions: In this multi-ethnic, population-based study, participants with a high burden of SPVS had increased incidence rates of vascular events. Furthermore, the joint presence of SPVS and LI heighten the risk of vascular death, any stroke and any vascular event. The presence of SPVS may help select subjects for randomized trials to assess intervention strategies.

Introduction: Despite theoretical concerns regarding hypoperfusion in patients with large artery occlusive disease, strict blood pressure (BP) control has become adopted as a safe strategy for stroke risk reduction. We examined the relationship between BP control, flow, and risk of subsequent stroke in the prospective Vertebrobasilar Flow Evaluation and Risk of Transient Ischemic Attack and Stroke (VERiTAS) Study. Hypothesis: Strict BP control may be less efficacious and safe in flow compromised patients with symptomatic vertebrobasilar (VB) disease. Methods: VERiTAS enrolled patients with recent VB TIA or stroke and ≥50% atherosclerotic stenosis or occlusion of vertebral or basilar arteries. Hemodynamic status using large vessel flow in the VB territory was designated as low or normal flow using quantitative MRA. Patients underwent standard medical management and follow-up for primary outcome event of VB territory stroke. Mean SBP during follow-up ( Results: The 72 enrolled subjects had an average of 3.8±1.2 BP recordings over the 20 ±8 months of follow-up; 40 (56%) had mean SBP of Conclusion: Among a subgroup of patients with VB disease and low flow, strict BP control (SBP

Background: Slower gait speed is associated with adverse outcomes of aging, such as falls and mortality. We hypothesized that physical inactivity, which is prevalent in the elderly, is a major risk factor for slowed gait speed independent of white matter hyperintensity volume (WMHV) and silent brain infarcts (SBI) on magnetic resonance imaging (MRI). Methods: In the Northern Manhattan MRI sub-study (n = 1290), 702 participants had physical performance measures a mean of 5 years after initial MRI. Assessment of gait speed was via a timed 5 meter walk test, averaged over 3 trials. Physical inactivity was assessed using a validated questionnaire at initial enrollment. Multi-variable linear and logistic regression models were fit to examine the association between physical inactivity and gait speed as a continuous outcome and binary outcome cut off at the lowest quartile. Models were adjusted for sociodemographics, hypertension, diabetes, waist circumference, osteoarthritis, and grip. Further mediation analyses were performed adjusted for WMHV and SBI. Results: The mean age at gait assessment was 74(±8.5) years, 61% women and 70% Hispanic; mean gait speed was 1.0±0.3 meter/second. In multi-variable models physical inactivity (54% of cohort), compared to physically activity, was associated with mean 0.04 m/s slower gait speed (p=0.03), and a greater odds of gait speed Conclusion: Physical inactivity is associated with slower gait speed independently of osteoarthritis, markers of sarcopenia, and subclinical ischemic brain injury. Modifying sedentary behavior poses a target for interventions aimed at reducing physical aging.

Introduction: Infection has been described as a trigger for acute ischemic stroke, but the relationship between postoperative infection and the risk of postoperative stroke is unclear. We investigated the association between postoperative infection and stroke using the American College of Surgeons National Surgical Quality Initiative Program (NSQIP) database. Hypothesis: Postoperative infection is associated with an increased risk of postoperative stroke. Methods: We used the NSQIP database to identify all patients who underwent surgery between the years of 2000 and 2010 and developed a postoperative stroke within 30 days of surgery. The group was further stratified according to the presence of infection preceding stroke. Using a logistic regression model adjusted for age, race, sex, medical comorbidities, surgical type, and dichotomized functional status, we compared the risk of stroke in patients with and without preceding infections, and investigated the risk of infection following stroke. Results: 729,886 surgical patients were identified, of whom 2,703 (0.3%) developed postoperative stroke. 848 (0.12%) patients developed both postoperative stroke and infection. Among patients who had postoperative stroke, 100 (3.7%) had developed an infection prior to developing a stroke. Patients with infection prior to stroke had a lower risk of stroke than patients who did not develop infection prior to stroke (adjusted odds ratio [OR] 0.25, 95%CI 0.20-0.32). 748 patients (0.1%) developed an infection after having a postoperative stroke. These patients had a higher risk of infection (incidence rate ratio 2.76, 95%CI 2.57-2.97) and a higher odds of infection (adjusted odds ratio [OR] 3.47, 95%CI 3.18-3.78) than patients who did not have a stroke. Conclusions: We found that the presence of a preceding infection was associated with a low risk of postoperative stroke in a large surgical inpatient sample. Although the total number of strokes may have been under-reported, these results conflict with other studies that report that infection is a trigger for ischemic stroke. Further analyses using more granular data are needed to investigate the relationship between postoperative infection and the risk of postoperative stroke.

Background: Increased P-wave terminal force in lead V1 (PTFV1) of a standard 12-lead electrocardiogram (EKG), a marker of left atrial dilatation and possibly fibrosis, has been associated with stroke risk in the absence of atrial fibrillation (AF), and with subclinical infarcts in some cohorts. We hypothesized that PTFV1 would be associated with an increased prevalence of subclinical infarcts, especially cortical ones, and leukoaraiosis in a population-based, multi-ethnic cohort. Methods: PTFV1 was collected manually from baseline EKGs of participants in the population-based, prospective Northern Manhattan Study (NOMAS) who had remained clinically stroke-free and undergone brain MRI (n=1,290). MRIs were read for superficial and deep infarcts and white matter hyperintensity volume adjusted for head size (WMHV). Logistic regression models were used for the association of PTFV1 with all subclinical infarcts and with cortical infarcts, and linear regression models with logWMHV. Models were adjusted for demographics and risk factors. Results: Among the 1174 participants with PTFV1, mean age was 70 + 9 SD years at the time of MRI, 40.3% were male, and 14.4% were white, 17.6% black, and 65.8% Hispanic. Hypertension was present in 68.0%. Mean PTFV1 was 3587.35 ± 2315.62 μV-ms. MRIs were performed a mean of 6.0 + 3.4 years after EKG. Subclinical infarcts were present in 170 (15.1%) participants, and were cortical in 40 (3.6%). PTFV1 >5000 μV-ms was associated with greater WMHV even after adjusting for demographics and risk factors, including baseline AF (mean difference in logWMHV 0.14, 95% CL 0.01-0.28). There was a trend toward an association of PTFV1 with cortical (unadjusted OR per SD change logPTFV1 1.30, 95% CI 0.94-1.81) but not with all subclinical infarcts (unadjusted OR 1.00, 95% CI 0.85-1.18). Conclusion: EKG evidence of left atrial abnormality was associated with leukoaraiosis, and possibly with subclinical cortical infarcts, though the limited number of outcomes did not permit us to confirm this finding. Left atrial cardiopathy may be a source of emboli, but may also cause cerebral hypoperfusion-related injury. Further studies in large cohorts are needed to determine the relationship of PTFV1 to risk of subclinical cerebrovascular disease.

Introduction: Cancer patients who develop acute ischemic stroke (AIS) are at high risk for recurrent thromboembolism. No risk stratification model exists to predict recurrent events in this population. The Khorana score is a validated risk score for predicting venous thromboembolism in newly diagnosed cancer patients. Hypothesis: The Khorana score can effectively classify the risk of recurrent thromboembolism (RTE) in cancer patients with AIS. Methods: We retrospectively identified all adults with active systemic cancer diagnosed with AIS by MRI at a tertiary-care cancer center from 2005 through 2009. Two neurologists independently reviewed all available medical records through July 31, 2012. The Khorana score at the time of index stroke was calculated for each patient. Points were assigned for specific cancer sites (2 points for very high-risk sites: stomach or pancreas; 1 point for high-risk sites: lung, lymphoma, gynecologic, bladder, testicular, or renal), platelet count ≥350,000/mcL, hemoglobin Results: Among 263 study patients, 90 (34%) were diagnosed with RTE, including 36 (14%) with recurrent AIS. The median Khorana score was 2 (IQR 1-2, range 0-5). None of the individual components of the score were independently associated with RTE, although there was a nonsignificant trend for high-risk cancer sites (OR 1.56, 95% CI 0.88-2.77). The rate of RTE was 28% among patients with a score of 0, 36% among patients with a score of 1-2, and 32% among patients with a score of 3-6. The c-statistic was 0.52 (95% CI 0.45-0.58) for predicting RTE and 0.49 (95% CI 0.38-0.59) for predicting recurrent AIS. Discussion: The Khorana score has poor discriminatory ability for predicting RTE in cancer patients with AIS, probably because these patients represent an especially high-risk group. Future research is needed to identify better methods for predicting RTE in this high-risk population.

Introduction: Pregnancy-associated stroke (PAS) remains among the leading causes of maternal morbidity and mortality. Up to 20% of women with PAS have preeclampsia or eclampsia (PEC). However, PEC is relatively common (∼3.4% of pregnancies), while PAS is rare (∼20 per 100,000 deliveries). It is not clear what risk factors differentiate women with PEC who develop PAS. Hypothesis: Identifying risk factors for stroke in women with PEC could help identify women at highest risk for PAS. Methods: We used the 2008-2012 New York State Department of Health Statewide Planning and Research Cooperative System (SPARCS) inpatient sample database to identify pregnant or postpartum women (APR MDC 14) aged 12-50 with mild to severe preeclampsia (642.4x-642.5x) or eclampsia (642.6x). We compared non-stroke PEC patients (controls) to those with ischemic (ICD-9 433.xx, 434.xx, 436), ICH (ICD-9 431), SAH (ICD-9 430), and non-specified pregnancy-related strokes, including post-partum stroke (ICD-9 671.7, 674.0-674.4). Risk factors included delivery type (C-section vs vaginal), age, multigravida, primary hypercoagulable state, pre-existing diabetes, obesity and migraine. The associations between risk factors and stroke were evaluated using generalized estimating equation logistic regression adjusting for patients’ re-hospitalizations to calculate odds ratios and 95% confidence intervals (OR, 95%CI) for stroke . Results: There were 85 stroke cases and 41,634 PEC controls without stroke. A majority of strokes in PEC patients occurred postpartum (n=76, 89.4%). Stroke patients were more likely to have eclampsia (OR: 24.7, 95%CI: 15.6-39.3) or severe pre-eclampsia (OR: 1.6, 95%CI: 1.0-2.5) compared to non-stroke patients. In multivariable analyses, being multigravida (OR: 1.4, 95%CI 1.3-1.6), primary hypercoagulable state (OR:2.1, 95%CI 1.8-2.4) and migraine (OR 2.3, 95%CI 2.1-2.5) were also risk factors for stroke. Conclusions: Women with eclampsia were at at the highest risk for pregnancy-associated stroke. Several other risk factors for stroke in women with preeclampsia may also help identify women who warrant closer monitoring for neurological symptoms in the peripartum period.

Background: There remains lack of consensus on who among minor stroke or transient ischemic attack (MS-TIA) patients constitutes a high-risk group that may be targeted by clinical trials to improve early recurrence risk. We aimed to determine predictors of early recurrent cerebrovascular endpoints (RCVE) among patients with MS-TIA. Methods: We retrospectively analyzed data from consecutive patients who presented to the emergency room of Columbia University between Jan 1 st 2010 - Dec 31 st 2014, with MS-TIA (NIHSS 0-3) diagnosed by a neurologist and within 12 hours from onset. The outcome (adjudicated by 3 stroke neurologists) was RCVE (neurological deterioration defined as worsening deficits not attributed to fever or other medical condition or recurrent ischemic stroke or TIA within hospitalization). Our primary predictor was large vessel disease (LVD) stroke subtype. Covariates included demographics, stroke risk factors, symptoms, and imaging parameters. We confirmed our findings in an independent cohort of consecutive MS-TIA patients (NIHSS 0-3) evaluated at Tulane University during the same time period. Results: The cohort at Columbia University Medical Center included 505 patients; mean length of hospital stay (LOS) 3.2 days, 31 had RCVE (6.1%). The Tulane cohort consisted of 753 patients (mean LOS 4.5 days), 40 had RCVE (5.3%); RCVE predictors on univariate and multivariate analyses in both cohorts were LVD subtype and to a lesser extent positive neuroimaging (Tables 1 and 2). There was an increase in percentage of patients with RCVE with both predictors combined as compared to separate use. When neither predictor was present, the rate of RCVE was very low (Figure). RCVE was associated with poor discharge outcome (mRS≥2). Conclusion: In patients with MS-TIA, LVD stroke subtype and perhaps positive neuroimaging predict RCVE in two independent cohorts.RCVE was associated with poor functional outcome. Prospective studies are needed to confirm our findings.

Introduction: Post-operative stroke (POS) is associated with vascular and cardiac surgery, but this finding has mainly been reported among populations receiving vascular and cardiac procedures. We investigated the association between type of surgery and risk of POS in a large, generalizable inpatient cohort from the American College of Surgeons National Surgical Quality Initiative Program (NSQIP) database. Hypothesis: Cardiac and vascular procedures are associated with an increased risk of POS. Methods: We identified patients that underwent surgery between the years of 2000 and 2010. Our primary outcome was POS within 30 days of surgery. Using a hierarchical model adjusted for age, race, sex, medical comorbidities and dichotomized functional status, we assessed for clustering between type of surgery and POS. We then determined risk factors for POS while adjusting for clustering. Each surgical type was compared against all other surgical types. Results: We identified 729,886 patients, of whom 2,703 (0.3%) developed POS. In the hierarchical analysis, cardiac surgery (incidence rate ratio (IRR) 6.38, 95%CI 5.37-7.55), vascular surgery (IRR 4.41, 95%CI 4.08-4.76), and neurosurgery (IRR 2.05, 95%CI 1.69-2.48) were associated with increased risk of POS. The only patient-level factor associated with surgery type was poor preoperative functional status. Accounting for clustering, patients with poor pre-operative functional status (OR 4.11, 95%CI 3.60-4.69), history of stroke (OR 2.35 95%CI 2.06-2.69), history of transient ischemic attack (OR 2.49 95%CI 2.19-2.83), active smoking (OR 1.20, 95%CI 1.08-1.32), and COPD (OR 1.39 95%CI 1.21-1.59) were at higher risk of POS. There was no interaction between preoperative functional status and type of surgery. Conclusions: In a large cohort of surgical inpatients, we found that the risk of POS was significantly associated with cardiac, vascular, and neurosurgical procedures. Certain patient populations, such as those with a dependent pre-operative functional status, may be at a higher risk of POS and may be more likely to undergo cardiac, vascular, or neurosurgical procedures. Further studies are needed to elucidate the relationship between pre-surgical functional status and type of surgery.

Introduction: The role of regional hypoperfusion as an important contributor to stroke risk in atherosclerotic vertebrobasilar (VB) disease has recently been confirmed by the observational Vertebrobasilar Flow Evaluation and Risk of Transient Ischemic Attack and Stroke (VERiTAS) Study. We examined the stability of hemodynamic status over time and relationship to stroke risk in patients followed in this prospective cohort. Methods: VERiTAS enrolled patients with recent VB TIA or stroke and ≥50% atherosclerotic stenosis or occlusion of vertebral and/or basilar arteries. Large vessel flow in the vertebrobasilar territory was assessed using quantitative MRA, and patients were designated as low flow or normal flow based on distal territory regional flow, incorporating collateral capacity. Patients underwent standard medical management and follow-up for primary outcome event of VB territory stroke. QMRA imaging was repeated at 6, 12 and 24 months. Flow status over time was examined relative to baseline, and relative to subsequent stroke risk using survival analysis. Results: Of 72 enrolled subjects, 50, 45 and 28 had follow-up QMRA at 6, 12 and 24 months respectively. 18 (25%) were characterized as low flow at baseline, 10% at 6 m, 20 % at 12m and 11% at 24m. Over 18±7months of follow-up, 7 of the 12 low flow patients (58%) with follow-up imaging had improved their flow status (low to normal) while only 3 of 44 normal flow patients (7%) had worsened. Baseline flow status has previously been demonstrated to be an independent stroke risk predictor; flow status at latest follow-up and prior to event remained a strong predictor of subsequent stroke (HR 4.9, CI 1.4 to16.9, p=0.01). Conclusions: There is potential both for improvement and worsening of hemodynamics in patients with atherosclerotic VB disease; improvement, likely attributable to development of collateral supply, was evident more frequently than worsening, presumably due to progression of disease. Flow status, both at baseline and over time, remains a robust risk factor for subsequent stroke, thus serving as an important prognostic marker.

Background: The incidence of cerebrovascular events is rising in the HIV population but little is known about the mechanisms of vascular injury among patients with HIV population. Methods: Large cerebral arteries were obtained from autopsies in 80 HIV+ cases and 80 HIV- controls, matched for sex and age ± 3 years, range 30-84. Arterial size, atherosclerosis phenotype, luminal stenosis, and media thickness were obtained in each artery. The average pixel intensities (AI) of metalloproteinase (MMP)-2, MMP-3, MMP-9, tissue inhibitor of MMP (TIMP)-1 and TIMP-2 was assessed with immunohistochemistry defining high intensity as an AI in the upper tertile. Inflammation was assessed with CD68 staining. All estimates were adjusted for age, sex, ethnicity, vascular risk factors and cocaine use. Results: Arterial inflammation was associated with greater staining intensity of MMP3 (Beta=0.36, P=0.003) and TIMP2 (Beta=0.29, P=0.002) in HIV+ cases, but not in HIV- controls. The association of HIV with MMP3 was independent of whether the inflammation was located in the media, intima, or adventitia, while for TIMP-2 the association was only significant with intima inflammation (Beta=-0.36, P=0.03). Cerebral arteries in HIV + cases with high intensity of MMP3 staining had thinner media (Beta=-2.3, P= 50 viral copies/mL (B=0.62, P=0.04), Nucleoside Reverse Transcriptase Inhibitor (B=1.08, P=0.009) or protease inhibitor use at the time of death (B=-0.64, P=0.05), vascular risk factors (B=0.51, P= Conclusions: Arterial inflammation in HIV+ cases is associated with greater expression of MMP3, and arteries with higher expression of MMP3 have a thinner media, despite thicker walls, and greater degrees of luminal narrowing. Through thinning of the media, MMP3 may play a role in HIV-associated dolichoectasia; through secondary effects, it may contribute to paradoxical luminal stenosis.

Background: Leisure time physical activity (LTPA) has been associated with a lower risk of dementia, but whether the effects are specific to particular cognitive abilities or improvement in functional status remains unclear. We examined LTPA in relation to domain-specific neurocognitive (NC) performance and change over time in a diverse community sample. Methods: Data on LTPA was collected during enrollment (1993-2001) into the Northern Manhattan Study (NOMAS), a prospective cohort study of risk factors for stroke and cognitive decline, using a validated in-person questionnaire, and two waves of NC assessments were done on a subcohort of participants undergoing brain MRI an mean of six and 12 years later (NC1: 2003-2008; NC2: 2008-2014). Baseline LTPA was defined in two manners: (1) maximum intensity of all activities performed categorized as moderate-heavy, light, and none; (2) total summarized as a continuous variable with the metabolic equivalent (MET) score, a composite of total reported intensity and time. Factor analysis-derived construct-relevant cognitive domains, including memory (MEM), executive function (EXEC), processing speed (PS), and language ability (LA), were computed by averaging z-transformed NC test scores. We used multivariable linear regression to examine LTPA in relation to baseline domain-specific NC performance, and change in performance over time, adjusting for socio-demographics, vascular risk factors, and MRI markers of cerebrovascular injury (white matter hyperintensity volume and total cerebral volume, both adjusted for total intracranial volume, and silent brain infarcts). Results: There were 1236 participants (mean age=64 years, 61% women, 67% Latino, 18% black, 15% white) with LTPA and NC data, and 879 with a second NC assessment. Moderate-to-heavy activity was associated with higher baseline LA (p Conclusions: Leisure time physical activity is independently protective against a decline in processing speed and memory, and was partly mediated by MRI markers. Leisure-time physical activity may protect against dementia through preventing cerebrovascular correlates of brain injury.

Background: Up to 1/3 of patients with primary intracerebral hemorrhage (ICH) have acute ischemic lesions remote from the hematoma visualized on diffusion-weighted MRI (DWI). Prior reports associated these lesions with large blood pressure fluctuations in the acute hospital setting. Here, the Ethnic/Racial variations of Intracerebral Hemorrhage (ERICH) study explored predictors of ischemic lesions in primary ICH and their impact on long-term outcomes. Methods: ERICH is a multi-center, prospective study of ICH among white, black, and Hispanic patients. At least every 5th patient undergoes a study MRI interpreted by a central core blinded to clinical data. We calculated change in mean arterial pressure (delta MAP) by taking the difference in highest and lowest blood pressure prior to MRI. Results: For the 601 cases with a DWI MRI, mean±SD age was 61±14 years, mean ICH volume was 16±18 cc, 55% were male, 79% had hypertension and 26% had ≥1 DWI lesions. The frequency of DWI lesions differed by ethnicity (blacks 33%, Hispanics 24%, whites 20%, overall p=0.005). The multivariable logistic regression analysis identified lower age (10 year OR=0.68, p Conclusions: The ERICH study confirms that large fluctuations in blood pressure during the acute hospitalization period predict DWI lesions, and that DWI lesions predict poor outcomes. We hypothesize that substantial reductions in blood pressure precipitate acute small vessel ischemia in those ICH patients with a more severe underlying diseased vasculature. Further studies are needed to determine whether there is a level of acute blood pressure reduction at which the risks outweigh the benefits.

Introduction: Many pathological mechanisms associated with smoking have been described, but studies on cigarette smoking in relation to carotid plaque morphology are lacking. Given that certain plaque compositions are more prone to cause cerebrovascular events, we sought to determine the association of carotid plaque density and cigarette smoking in an urban multi-ethnic cohort. Methods: This cross-sectional study analyzed 1743 stroke-free participants from the population-based Northern Manhattan Study using an ultrasonographic index of plaque echodensity, the Grey Scale Median (GSM). Echolucent plaque (low GSM) is more soft and fibrous, while echodense (high GSM) is more calcified. For individuals with multiple plaques, the mean GSM weighted by plaque area for each plaque was calculated. Quintiles of GSM were calculated among those with plaque and compared to those with no plaque as the reference. We used multinomial logistic regression models to assess the associations of current cigarette smoking with GSM, adjusting for demographics and vascular risk factors. Results: The mean age was 65.5±8.9 years, 60% were women, 18% white, 63% Hispanic, 19% black, 14% current smokers, 38% former smokers, and 48% never smokers. Carotid plaque was present in 58% of subjects. Current smokers were at a two-fold increased risk of being in GSM quintiles 1, 2, 4, and 5 as compared to never smokers (Table). Conclusions: A U-shaped relationship between current cigarette smoking and plaque echodensity is observed. Smoking is associated with both echodense calcified plaque as well as with echolucent, soft plaque that are more prone to rupture. Research is needed to determine if these vulnerable plaque phenotypes may mediate an association between cigarette smoking and clinical vascular events.

Background: Stroke may contribute to long-term functional decline apart from its acute effects on neurological function. Few studies have compared long-term disability trajectories before and after vascular events or considered the natural history of aging-related decline. We hypothesized that the increase in long-term disability would be steeper post-event than pre-event for stroke but not myocardial infarction (MI). Methods: In the Cardiovascular Health Study, 5888 Medicare eligible participants were followed for a mean of 13 years for vascular events and had annual disability assessments with an activities of daily living (ADL) and instrumental ADL scale, modified from the National Center for Health Statistics Supplement on Aging (range 0-12, scored by number of ADLs and IADLs which the participant could not perform, analyzed continuously). During follow-up, 382 participants had ischemic stroke and 395 had MI with >=1 post-event disability assessment. Generalized estimating equations models adjusted for baseline demographics, vascular risk factors, arthritis, cognition, and social support and included a test for different slopes of disability before and after event. Results: Participants had a mean of 4 disability assessments each pre- and post-stroke and MI. Stroke (0.88, 95% CI 0.57-1.20, p Conclusions: In this large population-based study, a trajectory of increasing disability became significantly steeper after stroke but not after MI. This worsening trajectory could be due to delayed cell dysfunction in the brain surrounding stroke, long-term inflammatory profile changes, progressive cardiovascular impairment, or silent recurrent infarcts. Stroke may be considered an ongoing, chronic condition with effects on function instead of a single discrete event.

Introduction: Cancer patients can develop hypercoagulability, leading to stroke. The pathophysiology of cancer-mediated hypercoagulability, however, is poorly understood and biomarkers associated with stroke in cancer patients are uncertain. Methods: We performed a pilot cross-sectional study at a tertiary-care cancer center to evaluate potential hematological biomarkers of stroke in cancer patients. From October 2013 through May 2014, we prospectively identified 15 inpatients with active systemic cancer and acute ischemic stroke and 15 inpatients with active systemic cancer and no stroke. Using collected plasma, we compared six potential biomarkers of stroke, selected in advance based on expert opinion and prior data, which included markers of coagulation (thrombin-antithrombin complex and D-dimer), platelet function (P-selectin), and endothelial activation state (sICAM-1, vICAM-1, and thrombomodulin). We used the Wilcoxon rank-sum test to compare biomarker levels between groups. Results: Each group comprised 14 patients with solid cancer and one patient with hematological cancer. All patients in the cancer and stroke group had stage 4 disease, while only 60% of the cancer and no stroke group had stage 4 disease. Age and demographics were similar between groups. Median levels of thrombin-antithrombin complex, D-dimer, P-selectin, sICAM-1, and vICAM-1 were significantly increased in cancer patients with stroke, but thrombomodulin was not (Table). Our results were substantially unchanged in a sensitivity analysis limited to patients with only stage 4 cancer. Conclusion: These preliminary data suggest that thrombin-antithrombin complex, D-dimer, P-selectin, sICAM-1, and vICAM-1 could serve as the focus of future prospective studies to elucidate the potential predictive, diagnostic, or etiologic role of hematological biomarkers for stroke in the cancer population.

Background: Fever on presentation has been demonstrated to be an independent predictor of poor-outcome after ICH. However, prospective data with post-discharge outcomes are limited. Methods: Associations between fever (≥38°C) on presentation and 3-month outcomes were sought among patients in ERICH, a multicenter, triethnic case-control study of ICH. Patients with missing CT or Glasgow Coma Score (GCS) and those lost to 3-month follow-up were excluded. Rates of fever on presentation were compared across admission characteristics, hospital complications, and 3-month outcomes. Additionally, a multivariate analysis was performed to determine factors associated with fever on presentation. Results: A total of 1644 patients out of 2276 enrolled by 12/31/2013 met inclusion criteria for the analysis. Fever on presentation with ICH was present in 2.4% of patients. In univariate analyses, fever was associated with (all data are % febrile in each group): ICH volume (60: 2.9%, p=0.03), higher ED GCS (3-4: 1.8%, 5-12: 3.0% 13-15: 6.6%, p=0.004), and WBC > 10 (yes, 3.4% vs. no, 1.7% p=0.03). Additionally, borderline significance was seen with race (white 3.1%, black 2.9%, Hispanic 1.1%, p=0.06) and ICH Location (lobar 3.9%, primary IVH 3.8%, Deep 1.6%, Brainstem 2.2%, cerebellum 1.7%, p=0.06). Fever on presentation was not associated with 3-month outcomes (mRS 0-2: 2.1%;3-5:1.9%; 6: 3.2%, p=0.45). In a multivariate analysis, fever was associated with GCS (OR 0.9, 95% CI 0.83-0.97) and white or black race (reference Hispanic; white OR 3.0 95% CI 1.14-7.88; Black OR 2.68, 95% CI 1.04-6.96). Conclusion: Fever is relatively uncommon at admission and does not appear to be associated with poor outcomes.

Background: Inflammatory biomarkers have been previously associated with stroke and mortality, but inflammation may also have detrimental effects beyond acute events. The association of these biomarkers with functional status is not well defined. We hypothesized that serum levels of high-sensitivity C-reactive protein (CRP), lipoprotein-associated phospholipase A2 (LpPLA2), and tumor necrosis factor receptor-1 (TNFR1) predict long-term functional status independently of vascular risk factors and stroke and myocardial infarction (MI) occurring during follow-up. Methods: In the prospective, multiethnic Northern Manhattan Study, stroke-free individuals in northern Manhattan aged >=40 years had annual assessments of disability with the Barthel index (BI), for a median of 13 years. BI was analyzed as a continuous variable (range 0-100). Baseline demographics, risk factors, and laboratory studies were collected, including CRP (n=2240), LpPLA2 activity (n=1912), and TNFR1 (n=1863). Separate generalized estimating equation models estimated standardized associations between each biomarker and 1) baseline functional status and 2) change in function over time, adjusting for demographics, vascular risk factors, social variables, cognition, and depression measured at baseline, and stroke and MI occurring during follow-up. Results: Mean age was 69 (SD 10) years, 36% were male, 54% Hispanic, 74% had hypertension, 22% diabetes; 337 MIs and 369 first strokes occurred during follow-up. CRP (-0.41, 95% CI -0.82 to 0.002) and LpPLA2 (-0.40, 95% CI -0.75 to -0.04) were associated with baseline BI but not change over time. TNFR1 was associated with baseline BI (-0.93, 95% CI -1.59 to -0.26) and change over time (-0.36 BI points per year, 95% CI -0.69 to -0.03). Conclusions: In this large population-based study, higher serum inflammatory biomarker levels were associated with disability, even when adjusting for baseline covariates and stroke and MI occurring during follow-up. Elevated TNFR1 predicted greater disability over time, suggesting that systemic inflammation may contribute to long-term functional decline and disability.

Background: While designated stroke centers (DSC) improve quality of care and outcomes for ischemic stroke patients, less is known regarding the benefits of DSCs for patients with intracerebral (ICH) and subarachnoid hemorrhage (SAH). Hypothesis: Compared to non-DSCs, hospitals with DSC status will have lower in-hospital mortality for hemorrhagic stroke patients independently of other hospital-level characteristics, including hospital size, urban location, and teaching status. Methods: We evaluated ICH (ICD-9 431) and SAH (ICD-9 430) hospitalizations that were recorded in the 2008-2012 New York State Department of Health’s Statewide Planning and Research Cooperative System (SPARCS) inpatient sample database. The relationship between DSC status and in-hospital mortality was evaluated using generalized estimating equation logistic regression to calculate OR’s and 95% confidence intervals; analyses were adjusted for clustering of patients within facilities, other hospital characteristics, and individual level characteristics. In secondary analyses we explored if other hospital characteristics were associated with in-hospital mortality. Results: Of 6501 ICH and 3440 SAH patients, in-hospital mortality was higher among those with ICH compared to SAH (23.7% vs. 18.5%). Unadjusted analyses revealed DSC was related with reduced mortality for both ICH (OR: 0.7; 0.6-0.8) and SAH patients (OR: 0.4; 0.3-0.7). DSC remained a significant predictor of in-hospital mortality for SAH patients (OR: 0.6; 0.4-0.9) but not for ICH patients (OR: 0.9; 0.6-1.2) after adjusting for patient demographic characteristics, comorbidities, hospital size, teaching status and location. Conclusions: Admission to a DSC was independently associated with reduced in-hospital mortality for SAH patients but not for those suffering from ICH. Other patient and hospital characteristics may explain the apparent benefit of PSC status on outcomes after ICH. Addressing gaps in stroke care through assessment of hospital-level factors indicates a potential target for quality of care improvement.

INTRODUCTION: Elevated fibroblast growth factor 23 (FGF23) regulates phosphate homeostasis and has been linked with mortality, cardiovascular events, and stroke. However, the role of FGF23 as a risk factor for subclinical cerebrovascular damage (SCVD) is not known. METHODS: We used multivariable linear and logistic regression to evaluate associations between FGF23, continuously and by quartiles, with white matter hyperintensity volume (WMHV), expressed as percent intracranial volume (%ICV), and subclinical brain infarction (SBI) in a race/ethnically diverse communitybased sample. RESULTS: There were 1,170 strokefree NOMAS participants with FGF23 levels and quantitative MRI data on WMHV and SBI. Participants with FGF23 levels in the top quartile (range=84.98 to1424.59 RU/mL) had greater WMHV (β=0.19 %ICV, 95% CI=0.04 to 0.33 %ICV, p=0.01) compared to those in the lowest quartile (range=15.41 to 48.96 RU/mL), adjusted for demographics, vascularrisk factors, and estimated glomerular filtration rate (eGFR). These findings remained significant in those without evidence of chronic kidney disease (CKD). The highest quartile of FGF23 also associated with greater odds of SBI, but this effect was not significant after adjusting for demographic factors and eGFR. However, sex was an effect modifier ofFGF23 in relation to odds of SBI (p for interaction=0.03). Stratified by sex, FGF23 was associated with greater odds of SBI in men (OR=1.7, 95% CI=1.0 to 2.6, p=0.03) after adjusting for demographics, vascular risk factors, and eGFR, but not among women. CONCLUSIONS: These crosssectional communitybased data from a diverse urban sample suggest elevated FGF23 may be a risk factor for small vessel disease, as well as MRIdefined brain infarction in men, independent of CKD. Prospective data on elevated FGF23 and SCVD are needed.

Background: Microbleeds have been reported in up to 60% of patients with primary intracerebral hemorrhage (ICH) and are an important marker of a progressive, small vessel cerebral vasculopathy. The Ethnic/Racial Variations of Intracerebral Hemorrhage (ERICH) study explored the characteristics of microbleeds and their impact on long-term functional outcome. Methods: ERICH is a multi-center prospective study designed to recruit white, black, and Hispanic cases of ICH. At least every 5th ICH patient enrolled undergoes an MRI. All MRIs are interpreted by a core laboratory blinded to clinical data. Results: Among 642 patients with a gradient echo (GRE) sequence, 49% had ≥1 microbleeds (mean 13, median 4). Blacks had the highest rate of microbleeds at 54% compared to 48% for Hispanics, and 42% for whites (overall p=0.046; blacks vs. white p=0.018). There was a significant racial/ethnic difference in the overall rate of microbleeds in lobar locations (79% in whites, 73.5% in blacks, 61% in Hispanics, p3 /μL OR=1.06, p=0.012), and race/ethnicity (overall p value 0.111; compared to white, black OR=1.53, p=0.056; Hispanic OR=1.49, p=0.078). The presence of microbleeds was associated with poor outcome (6 month mRS 4-6, OR=2.21, p=0.001) in a logistic regression analysis that included ICH volume (1 cc OR=1.04, p Conclusions: This study demonstrates substantial differences in microbleed rates across race/ethnicities. This is the first study to report an intermediate microbleed rate among Hispanics relative to white and black ICH cases. In addition to hypertension and leukoaraiosis, our model suggests that inflammation may be an important factor contributing to microbleeds. As a biomarker of progressive vasculopathy and poor outcome, microbleeds may provide a valuable surrogate measure in future studies of therapies targeting optimal approaches to risk factor control.

Background: Blood pressure (BP) control in hypertensive patients is crucial to reducing the risk of future cardiovascular events and heart failure. Left ventricular (LV) global longitudinal strain (GLS) can detect early subclinical cardiac dysfunction, even when LV ejection fraction (EF) is normal, and is an important prognostic indicator. It is not known whether poor BP control is associated with subclinical dysfunction by GLS, and whether ambulatory BP (ABP) is superior to office BP measurements in this regard. Methods: Two-dimensional speckle tracking echocardiography and 24-hour ABP monitoring were performed in 394 treated hypertensive subjects (mean age 72 ± 9 years; 63% women) with LVEF ≥50% from the Cardiovascular Abnormalities and Brain Lesions (CABL) study. Uncontrolled office BP was defined as systolic BP ≥140 mmHg and/or diastolic BP ≥90 mmHg; uncontrolled ABP as mean 24-hour systolic BP ≥130 mmHg and/or diastolic BP ≥80 mmHg. Abnormal GLS was defined as a value above the 95% percentile of healthy normotensive subjects from the CABL cohort. Results: Office BP was uncontrolled in 188 subjects (47.7%), and ABP in 162 (41.1%). Agreement between these two categorization was low [kappa = 0.16, 95% confidence interval (CI), 0.06-0.26; concordant classification in only 58.4% of subjects]. Subjects with uncontrolled ABP showed significantly worse GLS compared to those with controlled ABP (-16.3 ± 3.0% vs -17.7 ± 2.9%, p Conclusions: Uncontrolled 24-hour ABP is strongly associated with LV subclinical dysfunction in hypertensive subjects with normal LVEF. Assessing BP control with ABP monitoring may refine the risk stratification of treated hypertensive patients for early LV dysfunction.

Objective: To determine the attributable risks (AR) for vascular events (VE: stroke, myocardial infarction, and vascular death) due to hypertension and diabetes by race/ethnicity, sex and age. Background: Understanding the distribution of risk factors and their contribution to disease incidence is critical for effective allocation of resources for disease prevention. There is little data on the relative contribution of vascular risk factors among multiethnic elderly populations. Methods: The Northern Manhattan Study is a population-based prospective cohort study of incidence, risk factors, and outcomes in a multiethnic urban population. Multi-variable Cox models adjusted for socio-demographic and vascular risk factors were used to calculate hazard ratios, AR and 95% confidence intervals (95%CI) for (1) all stroke and (2) VE. Results: The cohort (n=3,298) had mean age 69.2+/-10 years with 63% women, 52% Hispanic, 73% hypertensive and 21% diabetics. There were 347 strokes and 835 VE during a median follow-up of 12 years. The AR due to hypertension was 30% (95%CI 13-48%) for stroke and 24% (95%CI 13-35%) for VE; AR due to diabetes was 19% (95%CI 12-27%) for stroke and 13% (95%CI 8-17%) for VE.The AR for stroke from hypertension differed by age and race-ethnicity (table), while for diabetes it only differed by age. For VE, the AR of hypertension and diabetes did not differ by age, sex, or race-ethnicity. Conclusions: On a population level diabetes and hypertension have important effects on the risk of VE across all socio-demographic groups. For stroke, these effects are more notable at age< 80. Our results may in part be driven by competing risks for other VE outcomes with stroke, or a differential impact of risk factors on stroke in younger individuals.

Background: The role of migraine headache as a risk factor for intracerebral hemorrhage (ICH) is less well defined than it is for ischemic stroke. Previous studies of migraine and ICH have been limited by the small number of ICH cases analyzed. We hypothesized that a history of migraine headaches is a risk factor for ICH in the ERICH study. Methods: ERICH is a multi-center, prospective, case-control study of a racial-ethnically mixed population with ICH. Controls were identified through random digit dialing to match cases by age (+/- 5 years), sex, race/ethnicity (non-Hispanic white, non-Hispanic black and Hispanic) and geographic area. Records of cases were reviewed for past medical history and data on ICH presentation. All cases and controls underwent a comprehensive interview which includes questions regarding headache history. Based on those responses subjects were categorized as having any headache (more than once a month), migraine headaches and migraine with aura. Results: A total of 740 cases were matched with controls. Mean age of cases and controls was 58±12 years, 55% were male, 44% non-Hispanic black, 31% non-Hispanic white and 25% Hispanic. Cases were more likely to report a history of migraine (17.3 vs. 14.1%; p=0.076), migraine with aura (5.4 vs. 4.2%; p=0.20) and any headache (38.4 vs. 16.0%; p Conclusion: A history of headache was associated with ICH in this preliminary analysis of ERICH data, although we were not able to confirm that migraine is a risk factor for ICH. Analysis of additional cases as ERICH continues recruitment may clarify the role of migraine in ICH. The increased prevalence of any headache reported by cases requires cautious interpretation as it may represent recall bias.

Background: Long-term disability after lacunar stroke is under-studied. We hypothesized that vascular and demographic risk factors predict worsening disability after recovery from lacunar stroke, even in the absence of recurrent stroke. Methods: The Secondary Prevention of Small Subcortical Stroke Study is a clinical trial in lacunar stroke patients with annual assessments of disability with the Older Americans Resources and Survey Instrumental Activities of Daily Living (IADL) scale (range 0-14), measuring 7 IADLs. Generalized estimating equations were used to model the likelihood of disability (IADL score Results: Among 2820 participants, mean age was 63.4 years (SD 10.8), 63% were male, 51% White, 32% Hispanic, 36% had some college education, 36% had diabetes, 90% had hypertension, and 10% had prior stroke. Mean follow-up was 3.5 years. Mean IADL score at 3 months was 12.5 (SD 2.5), and 43% were disabled. In multivariable models, female sex (OR 1.5, 95% CI 1.3-1.8), diabetes (1.52, 1.30-1.75), current smoking (1.28, 1.06-1.54), non-regular alcohol use (1.6, 1.4-1.9), prior stroke (1.61, 1.28-2.00), cognitive assessment screening instrument score (0.98 per point, 0.97-0.98), and depression (1.79, 1.49-2.17) were associated with disability. The youngest age quartile had decreased odds of disability over time (OR 0.56 per year, 95% CI 0.36-0.91), whereas the oldest age quartile had increased odds (1.33, 0.83-2.22). There was heterogeneity by region (p Conclusion: In lacunar stroke patients, older age predicted worsening long-term disability, even without recurrent stroke. Worse long-term function was associated with vascular risk factors and prior stroke, and regional differences may have been due to geographic variations in health care delivery or scale interpretation.

Introduction: Fibroblast growth factor 23 (FGF23), a hormone that regulates phosphate homeostasis, has recently been linked with mortality, cardiovascular events, and stroke. While limited evidence suggests that FGF23 is associated with arterial calcification in chronic kidney disease (CKD), data on the association between FGF23 and carotid plaque are lacking, especially in population-based studies. This study aimed to determine if relative elevations of FGF23 are associated with presence and area of carotid plaque in a race/ethnically diverse community-based sample. Methods: There were 1,512 stroke-free NOMAS participants with FGF23 levels and 2D carotid ultrasound data available (mean age, 68±9; 61% women; 62% Hispanic, 18% black, 18% white, 2% other). We evaluated the association of FGF23, continuously and by quintiles, with presence of carotid plaque using logistic regression models and with plaque area (cubic root transformed) using linear regression models, adjusting for age, sex, race/ethnicity, estimated glomerular filtration rate (eGFR), body mass index, smoking, alcohol use, blood pressure, fasting glucose, lipids, and hypertension, diabetes, and dyslipidemia medication use. Results: Participants with FGF23 levels in the top quintile had greater odds of having carotid plaque (OR=1.5 95% CI=1.0-2.2, p=0.04) and larger plaque area (beta=0.3, 95% CI=0.1-0.5, p=0.006) than those in the lowest quintile, adjusting for eGFR and demographic factors, although the association was attenuated after adjusting for vascular risk factors. Linear regression models also showed that log transformed FGF23 was associated with greater plaque area adjusting for eGFR and sociodemographic factors (beta=0.2, 95% CI=0.03-0.3, p=0.02), and this remained marginally significant after adjusting for vascular risk factors (beta=0.1, 95% CI=0.0-0.3, p=0.05). Conclusions: These data suggest that FGF23 is associated with the likelihood and burden of carotid atherosclerosis in a stroke-free community based sample. Our finding that the association between FGF23 and carotid plaque measures was weaker after adjusting for vascular risk factors suggests possible mediation and requires further study.

Background: Serological evidence of the burden of common infections (IB=infectious burden) has been associated with stroke, carotid atherosclerosis, and global cognitive function in the race/ethnically diverse community-based Northern Manhattan Study (NOMAS) cohort. We hypothesized that IB is associated with worse performance and decline on a battery of neuropsychological tests. Methods: A subsample of stroke-free NOMAS participants underwent a neuropsychological battery, and a subsample of these had repeated testing a mean of six (SD=2) years later. Using factor analysis-derived domain-specific Z scores for memory, language, executive function, and processing speed, we examined associations between a quantitative IB index previously associated with stroke risk and based on five common infections (Chlamydia pneumoniae, Helicobacter pylori, cytomegalovirus, and herpes simplex viruses 1 and 2) and cognitive performance in each domain. We used multivariable linear regression and adjusted for age, sex, education, race/ethnicity, insurance status, diabetes, hypertension, hypercholesterolemia, BMI, moderate alcohol, moderate-heavy physical activity, and smoking. Results: There were 588 NOMAS participants with IB index and cognitive data (mean age=71, 62%women, 14% white, 16% black, 70% Hispanic) and 287 with repeat cognitive testing. After adjusting for demographic variables, IB index was inversely associated with all cognitive domains, although not all associations were significant (executive function, beta=-0.11, p=0.002; language, beta=-0.07, p=0.04; memory, beta=-0.06, p=0.11; processing speed, beta=-0.07, p=0.07). These effects attenuated only slightly after adjusting for vascular risk factors. The IB index associated with cognitive decline across domains, but a trend toward significance existed only for memory, after adjusting for baseline performance, demographic and vascular risk factors (beta=-0.09, p=0.07). Conclusions: A quantitative measure of IB associated with stroke risk was inversely associated with executive function and language, and decline in memory performance. Past exposure to common infections may contribute to vascular cognitive impairment and warrant further study.

Background: Acute triggers of ischemic stroke (IS) remain poorly characterized. Emerging evidence suggests infections may promote short-term pro-inflammatory, pro-coagulant states that precipitate stroke. Objective: We hypothesized that exposure to influenza-like illness (ILI) is associated with increased risk of IS, as time intervals between events decrease. Methods: HCUP/AHRQ administrative claims from all nonfederal hospitals in California (2009) were queried for IS, using a published ICD-9-CM surveillance algorithm, identifying 41,148 unique individuals. A patient identifier was used to link IS patients across hospitalizations with a validated algorithm for ILI using inpatient and emergency department records. For each IS case, ILI events within 15, 30, and 90-day risk periods preceding IS in 2009, were compared to the same time periods one and two calendar years prior (i.e., 2007 and 2008). A 4-day buffer period was used to ensure ILI preceded IS. Conditional logistic regression models, matched on individuals, were used to calculate odds ratios and 95% confidence intervals (OR, 95%CI) for given risk period, after adjusting for monthly prevalence and mean age of ILI hospitalizations. Effects stratified by age are also explored. Results: Median (IQR) age of cases was 74 (62-83) years; 52.4% were female. There were more ILI during the 90-days preceding strokes in 2009 (n=439) than in the same stroke-free calendar period one year prior (n=303) and two years prior (n=81). ILI was associated with IS with decreasing magnitude of effect using windows of 15-days (adjusted OR 6.5, 95%CI, 2.2-19.7), 30-days (adjusted OR=3.7, 95%CI 1.7-8.3), and 90-days (adjusted OR=3.3, 95%CI, 1.9-5.8). The risk associated with ILI was higher among younger patients. Using the 30-day window, the association between ILI and IS was strongest among those 45-to-65-years (n=27,545, adjusted OR=2.53, 95%CI, 0.95-6.8). Conclusion: Influenza-like illness may contribute to a heightened risk of ischemic stroke during short term periods post-infection, especially among younger patients. Presentation with influenza represents an opportunity for targeted prevention of stroke.

Objective: Examine the association between lipid profile components and subclinical cerebrovascular disease on magnetic resonance imaging (MRI). Design/Methods: Fasting plasma lipids were collected on Northern Manhattan Study participants; a subsample underwent brain MRI (n=1256 with lipids available) a median of 6.2 (range=0-14) years after enrolment, when repeat fasting lipids were obtained (n=1029). We used lipid profile components at the time of initial enrollment as categorical variables, as well as change in clinical categories over the 2 measures. We examined the association between lipid profile components and (1)total white matter hyperintensity volume (WMHV) using linear regression, and (2)silent brain infarcts (SBI) using logistic regression. Results: The association between baseline lipids and WMHV was modified by apoE status (chi-squared with 2 degrees of freedom,p=0.03), such that among apoE4 carriers those with total cholesterol (TC)≥200mg/dl ) had a trend towards smaller WMHV than those with TC50 mg/dl for women,>40 mg/dl for men) at baseline to high risk HDL-C at the time of MRI (versus starting and remaining low risk) was associated with greater WMHV (difference in logWMHV 0.34,p-value 0.03). We noted a similar association with transitioning to a TC≥200mg/dl at the time of MRI (difference in logWMHV 0.25,p-value 0.006) There were no associations with SBI. Conclusions: The association of plasma lipid profile components with greater WMHV may depend on apoE genotype and worsening HDL and TC risk levels over time.

INTRODUCTION: Individuals with HIV live longer since the advent of antiretroviral (ARV) drugs. However, aging with HIV is now associated with vascular disease. Mechanisms underlying this association have not been fully elucidated. We explored arterial correlates that might lead the association between HIV and cerebral infarction. Methods: The Brain Arterial Remodeling Study (BARS) is a collection of 336 brains from donors in the US (69%) and Macedonia (31%) from which large brain arteries have been collected, cut, embedded and stained with H&E and van Gieson. Atherosclerosis and stenosis were graded using Glagov’s method. Cerebral infarction was identified in autopsy. For this analysis, HIV cases (n=138) were age- and sex-matched to controls without HIV. Multivariate logistic and linear regressions were used to determine odds ratios and 95% confidence intervals (OR, 95% CI) assessed associations at P< 0.05. Results: Those with HIV (mean age 48.8 ± 9.1 years, 75% men) were more likely to be black (OR 5.0, 2.4-10.4) or Hispanic (OR 2.2, 1.2-3.9), hypertensive (OR 2.2, 1.3-3.5), dyslipidemic (OR 3.5, 1.1-11.0), cocaine users (OR 10.1, 4.6-22.4) and have cardiac disease (OR 2.0, 1.2-3.5). Cerebral infarction was more common with HIV (adjusted OR 2.8, 1.1-6.8). HIV brains were less likely to have atherosclerosis (OR 0.4, 0.3-0.7) or any stenosis > 25% (OR 0.4, 0.1-1.0). Among those with HIV, after adjusting for demographic and vascular risks, cerebral infarction was more prevalent with prior opportunistic infections (5.3, 1.2-22.9), higher CD4 count (OR 1.2, 1.0-1.4 per every 50 CD4 cell increment), and in those with intimal thickening in more than 75% of the lumen circumference (OR 4.0, 1.1-14.5). Predictors of circumferential intimal thickening in the HIV group included ARV at death (Beta 0.17, P=0.007), internal elastic lamina (IEL) gaps, Beta 0.22, P=0.02), and IEL duplication (Beta 0.28, P=0.001). CONCLUSIONS: Individuals with HIV in this study have a greater prevalence of vascular risks and cerebral infarction, yet the increased infarction prevalence is not associated with atherosclerosis. Patients with HIV have a non-atherosclerotic arterial phenotype consisting of circumferential intimal thickening with disruption of the internal elastic lamina.