Stroke Risk Prediction Research Articles

Abstract 63: Ultra-Processed Food Intake Predicts Stroke Risk in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study

Introduction: Higher intake of ultra-processed foods (UPFs) is linked to cardiometabolic diseases and adverse neurological outcomes, such as stroke and cognitive decline. However, it is controversial whether food processing confers neurological risk independent of nutritive information. We explored associations between UPFs and incident stroke in the longitudinal REasons for Geographic and Racial Differences in Stroke (REGARDS) study. Methods: The NOVA system was used to classify items from a baseline food frequency questionnaire (FFQ) according to level of processing. Intake in grams for each NOVA category was normalized to total grams consumed. Scores quantifying adherence to other diets considered neuroprotective—including a Mediterranean, DASH, and Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) diet—were also derived from FFQ data. Multivariable Cox models were used to evaluate associations between diet patterns and stroke. Results: The cohort included 15,656 participants with reliable dietary intake data and without a previous stroke. Greater intake of UPFs (NOVA4) was associated with increased stroke risk (HR per SD: 1.10, 95% CI [1.03-1.18], p=4.77x10 -3 ) in a model adjusted for demographic factors (age, race, sex, age-by-race interaction) and vascular risk factors (smoking, atrial fibrillation, hypertension, diabetes mellitus, cardiovascular disease, left ventricular hypertrophy). Intake of unprocessed or minimally processed foods (NOVA1) was associated with decreased stroke risk (HR: 0.87, 95% CI [0.81-0.93], p=5.56x10 -5 ) as was adherence to the Mediterranean, DASH, and MIND diets. Adding intake of NOVA1 or NOVA4 items to multivariable models that included other diet patterns resulted in improved model fit (greater log-partial-likelihood). In contrast, adding other dietary patterns to models that included NOVA1 or NOVA4 intake did not improve model fit. While associations between NOVA1 and NOVA4 intake and stroke persisted in models that included the Mediterranean, DASH, or MIND diets, associations between other diet patterns and stroke were not significant. Conclusion: Intake of differentially processed foods is associated with stroke risk independent of adherence to other dietary patterns.

Abstract WP259: Headache and Hypertension and Early Postpartum State: A Proposed Clinical Triad for Maternal Stroke Risk Prediction

Background: Stroke is a major cause of maternal morbidity and mortality, but delays in diagnosis are common. We hypothesize that in individuals who are pregnant or within 1 year postpartum (PP), the triad of early PP state (first 6 weeks after delivery), elevated blood pressure (BP), and headache is associated with higher stroke risk. Methods: We conducted a single-institution case-control study. Cases had a maternal stroke (ischemic, hemorrhagic, subarachnoid hemorrhage, or cerebral venous thrombosis) during pregnancy or PP, identified from a stroke registry. Controls were pregnant or PP but without any stroke and admitted for any reason, identified from a cohort of individuals with a high proportion of hypertensive disorders of pregnancy (HDP), a known maternal stroke risk factor. We created multivariable logistic regression models with stepwise backward elimination to evaluate the association between maternal stroke and the triad of (1) early PP; (2) systolic BP ≥140 or diastolic ≥90 mmHg; and (3) headache. Results: We identified 60 cases and 113 controls. Population characteristics are shown in the Table. Cases had a higher proportion of preexisting headache disorders and controls a had higher proportion of HDP. Adjusting for race, HDP, and headache disorder history, stroke risk was higher in those presenting to acute care with the proposed triad than in those without these features (adjusted odds ratio 13.6, 95% confidence interval 3.8-48.5, p<0.0001). In a predictive model, the triad exhibited excellent discrimination between cases and controls (AUC 0.82, sensitivity 30%, specificity 98%, positive predictive value 90%). Conclusions: In a high risk, diverse maternal population delivering at an urban academic medical center, the triad of early PP state, elevated BP, and headache was associated with a more than 13-fold higher risk of stroke. Patients presenting for acute care with this triad could potentially benefit from a comprehensive stroke evaluation.

Abstract WP266: Prediction of Stroke Incidence Using Machine Learning: The Suita Study

Background: This population-based study investigated the potential of machine learning algorithms to predict stroke incidence and identify important risk factors. This study aimed to evaluate the accuracy of these algorithms in constructing a stroke prediction model. Methods: Participants from the Suita study were included, and baseline measurements were used to predict stroke outcomes over a 15-year follow-up period. In total, 7,389 participants and 51 variables were investigated, including demographics, medical history, medical imaging, laboratory data, and lifestyle habits. Initially, unsupervised K-prototype clustering was used to group participants based on their stroke risk. Subsequently, five supervised models (logistic regression, random forest, support vector machine, extreme gradient boosting, and light gradient boosted machine) were applied to predict the stroke outcomes. The Shapley Additive Explanations (SHAP) method determined the most critical variables. Results: Unsupervised clustering revealed significant differences in stroke incidence among the three identified risk clusters (9.1%, 6.6%, and 3.2%). These clusters were categorized into high-, medium-, and low-risk groups. Among the supervised models, the random forest algorithm demonstrated the best performance. The top ten most important variables for predicting stroke incidence were identified using the SHAP, with age being the most influential variable. Other significant risk markers included systolic blood pressure, hypertension, estimated glomerular filtration rate, metabolic syndrome, and blood sugar level. Additionally, elbow joint thickness and fructosamine, hemoglobin, and calcium levels were found to be potential predictors of stroke risk. Notably, the variables identified by the SHAP were consistent with those obtained from the unsupervised clustering approach in the high-risk group. Conclusion: Machine learning algorithms provide accurate predictions of stroke incidence and offer valuable insights into subclinical markers without the need for prior assumptions of causality. This study presents a data-driven machine-learning framework for stroke risk prediction and biomarker identification.

Abstract WP236: Harmonization of Stroke Risk Prediction Variables Using Natural Language Processing

Introduction: Early identification of stroke risk can profoundly influence an individual's chance of survival. While machine learning-based stroke risk prediction models perform better than traditional counterparts, they require integrating diverse open data sources. Variable discrepancies - where the same concept is described differently - pose challenges for data integration, requiring time-intensive manual harmonization. To address this, we developed an automated harmonization technique using natural language models. Methods: We utilized data from the Atherosclerosis Risk in Communities (ARIC), Multi-Ethnic Study of Atherosclerosis (MESA), Framingham Offspring Study, and Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohorts. Our approach involved training a two-layer fully connected neural network (FCN) to predict a harmonized ‘concept variable’ based on variable descriptions. These predictions employed embeddings from Bidirectional Encoder Representations from Transformers for Biomedical Text Mining (BioBERT). Further refining was achieved through paired sentence classification tasks to determine if description pairs shared identical concepts. We experimented with both separate cohort training and combined cohort training to evaluate context dependency. We used cosine similarity between the embeddings of the pre-trained BioBERT encoder as a prediction score for our baseline method. Results: The harmonization methods using natural language models outperformed the baseline method. The FCN model improved the area under the receiver operating characteristic from 0.787 (baseline) to 0.985 for the paired classification task. For harmonization concept prediction, the top-2 accuracy - correct prediction in the top two answers - improved from 11.9% (baseline) to 54.6%, while the top-5 accuracy improved from 16.1% (baseline) to 67.6%. Conclusions: Utilizing natural language processing for data harmonization provides a scalable approach that improves accuracy and efficiency of prediction. This facilitates the inclusion of diverse cohorts, broadening the sample size and range of indicators available, which in turn advances the versatility and the generalizability of stroke risk prediction models.

The Role of Transcranial-Color-Doppler in the Management of Basilar Artery Stenosis: Case Series and Systematic Review of Literature

Managing vertebrobasilar (VB) stroke, particularly basilar artery occlusion (BAO), presents challenges due to diverse clinical presentations and intricate diagnostics, risking delays in acute-phase reperfusion therapies. The diagnostic complexities of VB stroke prompt questions about whether presentation tempo influences outcomes. For non-urgent cases of basilar artery stenosis (BAS) lacking guidelines, clinical management becomes case-dependent. Current options include aggressive medical therapy (AMT) and percutaneous transluminal angioplasty and/or stenting (PTAS). However, the choice of PTAS remains debated, with recent trials observing higher-than-expected recurrence rates, particularly in intracranial stenosis. Understanding hemodynamic status is crucial in predicting stroke risk, especially in atherosclerotic VB disease. Recent studies highlighted the role of distal flow status in predicting stroke risk, emphasizing the importance of hemodynamic assessment beyond anatomic measures. Neurosonology, especially transcranial color-coded duplex sonography (TCCS), emerges as a valuable tool for assessing hemodynamics. Despite TCCS being operator-dependent and technically challenging for BAS evaluation, it effectively detects significant hemodynamic changes, providing real-time information on collateral flow. This review explores the potential role of TCCS in managing BA hemodynamic failure in VB stroke, with particular regard to the selection of BAS patients who may benefit from PTAS.

Triglyceride-glucose index prediction of stroke incidence risk in low-income Chinese population: a 10-year prospective cohort study.

The Triglyceride-Glucose (TyG) index, an indicator of insulin resistance, has been proposed as a predictor of cardiovascular diseases. However, its role in predicting stroke risk, particularly in low-income populations, is not well understood. This study aimed to investigate the predictive value of the TyG index for stroke incidence in a low-income Chinese population, with a focus on gender and age-specific differences. This 10-year prospective cohort study included 3,534 participants aged ≥45 years from rural areas in northern China. Baseline data on demographic characteristics, lifestyle factors, and clinical measurements were collected. Participants were followed for stroke incidence, categorized into ischemic and hemorrhagic stroke. Multivariate logistic regression models were used to assess the association between the TyG index and stroke incidence, adjusting for potential confounders. During the follow-up period, 368 participants (10.4%) experienced a stroke, with 327 ischemic and 31 hemorrhagic strokes. TyG index was significantly associated with total and ischemic stroke incidence but not hemorrhagic stroke. After adjusting for confounding factors, for every one standard deviation increase in TyG index, the risk of stroke increased by 32% for overall stroke (RR: 1.32; 95% CI: 1.08-1.61; P=0.006) and 39% for ischemic stroke (RR: 1.39; 95% CI: 1.12-1.73; P=0.003). The risk of stroke in the highest TyG tertile levels (tertile 3) increased by 49% (RR: 1.49; 95% CI 1.11-1.99; P=0.007) for overall stroke, compared to those in the lowest tertile levels (tertile 1). For ischemic stroke, the risk of stroke increased by 53% (RR: 1.53; 95% CI 1.12-2.11; P=0.008) in the highest TyG tertile levels (tertile 3) compared to those in the lowest tertile levels (tertile 1). This 10-year prospective cohort study has established the TyG index as an independent predictor of both total and ischemic stroke incidence in a low-income Chinese population. The findings indicate that the TyG index is particularly effective in predicting stroke risk among women and older adults (≥60 years), but not for hemorrhagic stroke. These insights are crucial for improving clinical practice and stroke prevention strategies.

Comparison of Data Mining Techniques on Stroke Clinical Dataset

Stroke is a significant cause of mortality and morbidity worldwide, making it essential to identify individuals at risk of experiencing a stroke. The aim of this research article is to develop a predictive model to determine the stroke risk of individuals based on their medical history and compare the effectiveness of preprocessing techniques on the model's performance. The methodology involves two streams of analysis - with and without data preprocessing - utilizing classification models to predict stroke risk (K-Nearest Neighbor, Decision Tree and Support Vector Machine). The results indicate that data preprocessing improves the performance of all models, with KNN and SVM showing high precision and recall values, making them effective models for predicting strokes. Conversely, the decision tree model performs well with data preprocessing despite slightly lower accuracy and recall values. These findings suggest that preprocessing is a crucial stage in machine learning and can enhance the performance of classification models in predicting stroke risk.

Trans-Balance: Reducing demographic disparity for prediction models in the presence of class imbalance

Introduction:Risk prediction, including early disease detection, prevention, and intervention, is essential to precision medicine. However, systematic bias in risk estimation caused by heterogeneity across different demographic groups can lead to inappropriate or misinformed treatment decisions. In addition, low incidence (class-imbalance) outcomes negatively impact the classification performance of many standard learning algorithms which further exacerbates the racial disparity issues. Therefore, it is crucial to improve the performance of statistical and machine learning models in underrepresented populations in the presence of heavy class imbalance. Method:To address demographic disparity in the presence of class imbalance, we develop a novel framework, Trans-Balance, by leveraging recent advances in imbalance learning, transfer learning, and federated learning. We consider a practical setting where data from multiple sites are stored locally under privacy constraints. Results:We show that the proposed Trans-Balance framework improves upon existing approaches by explicitly accounting for heterogeneity across demographic subgroups and cohorts. We demonstrate the feasibility and validity of our methods through numerical experiments and a real application to a multi-cohort study with data from participants of four large, NIH-funded cohorts for stroke risk prediction. Conclusion:Our findings indicate that the Trans-Balance approach significantly improves predictive performance, especially in scenarios marked by severe class imbalance and demographic disparity. Given its versatility and effectiveness, Trans-Balance offers a valuable contribution to enhancing risk prediction in biomedical research and related fields.

Komparasi Algoritma KNN dan SVM dalam Memprediksi Penyakit Stroke

Stroke is a serious medical condition that affects many people around the world. The ability to predict a person's stroke risk can help in effective prevention, treatment and care. In this study, a comparison between the K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) algorithms was conducted to predict stroke risk. The KNN algorithm is a method that searches for the nearest neighbors among the data points to be predicted and assigns the most common label among its neighbors. Experimental results show that both KNN and SVM can provide fairly accurate stroke predictions. However, from an operational point of view, SVM consistently performed better than KNN in terms of accuracy and precision. This research provides insight into the differences between KNN and SVM algorithms in the context of stroke prediction. The results can provide guidance for researchers and practitioners in choosing the right algorithm to predict stroke risk based on the characteristics of the available datasets.

A-296G variant of THBS1 gene (rs1478605) is associated with a lower frequency of stroke in a Brazilian population with sickle cell anemia

ObjectivesStroke is a devastating clinical outcome that significantly contributes to the morbidity and mortality of sickle cell anemia (SCA) patients. Despite its advantages in predicting stroke risk, transcranial Doppler screening has limitations that restrict its applicability, highlighting the need for emerging prognostic tools. Thrombospondin-1 plays a crucial role in endothelial injury, platelet adhesion, and nitric oxide metabolism and may be implicated in stroke pathophysiology. Here, we aimed to evaluate the association of THBS1 genetic variations with the occurrence of stroke in SCA patients Materials and methodsBy real-time PCR, 512 SCA patients were fully genotyped for THBS1 A-296G (rs1478605) polymorphism ResultsTHBS1 GG genotype was associated with a lower risk for stroke occurrence [odds ratio (OR): 0.30; 95% confidence interval (CI): 0.11–0.78; P = 0.011], although these findings were not consistent with multivariate logistic regression analysis (OR: 0.73, 95% CI: 0.12 – 4.37; P = 0.736). In agreement, the cumulative incidence of stroke for patients with AG/AA genotypes was higher when compared to the GG genotype (P = 0.018). However, the association was not maintained in the multivariate proportional hazards model (hazard ratio: 0.67, 95% CI: 0.12–3.61; P = 0.643) ConclusionsIn summary, the present study shows that the THBS1 A-296G (rs1478605) polymorphism may be a potential modifier for stroke in SCA.

Short- and long-term outcomes of patients with minor stroke and nonvalvular atrial fibrillation

Background and purposeNonvalvular atrial fibrillation (NVAF) is a risk factor for stroke. This study was undertaken to determine the influence of NVAF on the mortality and recurrent stroke after a minor stroke event.MethodsData were derived from the Third China National Stroke Registry (CNSR-III) which enrolled 15,166 subjects during August 2015 through March 2018 in China. Patients with minor stroke (NIHSS ≤ 5) within 24 h after onset were included. Clinical outcomes including all-cause mortality, cardiovascular death, recurrent ischemic stroke, and recurrent hemorrhagic stroke were collected. The Cox proportional hazards models were used to determine the association between NVAF and clinical outcomes.ResultsA total of 4,753 patients were included in our study. Of them, 222 patients had NVAF (4.7%) (mean age, 71.1 years) and 4,531 patients were without AF (95.3%) (mean age, 61.4 years). NVAF was associated with 12-month cardiovascular mortality in both univariate (hazards ratio [HR], 4.13; 95% confidence interval [CI], 1.84 to 9.31; P < 0.001) and multivariate analyses (HR, 4.66; 95% CI, 1.79 to 12.15; P = 0.001). There was no difference in the in-hospital ischemic stroke recurrence rate between the two groups (HR, 0.45 [95% CI, 0.19 to 1.05] P = 0.07 at discharge). However, patients with NVAF had a lower rate of recurrent ischemic stroke at medium- (3 months and 6 months) and long-term (12 months) follow-up (HR, 0.33 [95% CI, 0.16 to 0.68] P = 0.003 at 3 months; 0.49 [95% CI, 0.27 to 0.89] P = 0.02 at 6 months; 0.55 [95% CI, 0.32 to 0.94] P = 0.03 at 12 months, respectively) compared with those without. There was no difference in all-cause mortality and hemorrhagic stroke between the two groups during follow-up.ConclusionsMinor stroke patients with NVAF were at higher risk of cardiovascular death but had a lower rate of recurrent ischemic stroke compared to those without during the subsequent year after stroke event. A more accurate stroke risk prediction model for NVAF is warranted for optimal patient care strategies.

Comparison of Data Mining Classification Algorithms for Stroke Disease Prediction Using the SMOTE Upsampling Method

Stroke is a circulation disorder in the brain that can cause symptoms and signs related to the affected part of the brain and is the leading cause of death and disability in Indonesia. Everyone is at risk of experiencing a stroke, and it is important to recognize and manage risk factors. Data Mining techniques can help in the extraction and prediction of information, as well as finding hidden patterns in stroke medical data. The dataset used in this research comes from Kaggle and is imbalanced, so the SMOTE Upsampling technique is used to address this imbalance issue. The results of the study conclude that the use of SMOTE technique in the C4.5, NB, and KNN algorithms can increase precision, recall, and AUC. The C4.5 algorithm and SMOTE technique as the best performing algorithm were selected for testing new data, and the results show that the model created can predict stroke risk more accurately than the C4.5 model without SMOTE. However, it should be noted that based on the author's interview with one of the medical practitioners, the model cannot be directly used in medical practice because the observations in the medical field to determine factors related to stroke are highly complex. Thus, a new understanding revealed that predicting stroke in a practical setting is highly complex. While data mining can be used as a predictive tool in the initial stage for predictions in the general population, it is strongly recommended to undergo direct examination by doctors in a hospital to obtain more accurate and comprehensive medical evaluations.

Blood pressure-cerebral oxygen coupling model: A new approach for stroke risk prediction.

Stroke is a major cause of death and disability worldwide, but predicting its risk remains challenging. This study aimed to evaluate the cerebral blood flow autoregulation function of subjects with different stroke risk levels and predict their stroke risk. The coupling strength between cerebral oxygen and blood pressure signals was calculated by wavelet analysis and dynamic Bayesian inference and used as a quantitative index of cerebral blood flow autoregulation. A stroke prediction model based on the extreme random tree was constructed using the coupling strength and other data as input features. The results showed that the coupling strength was significantly higher in the high-risk group than the other groups. Moreover, the prediction model achieved an average accuracy of 0.80 across the three groups. The coupling strength of cerebral oxygen and blood pressure can be used as an objective index to predict stroke risk, which has implications for stroke prevention and intervention.

Circulating metabolites are biomarker candidates for stroke risk prediction as assed in a europe-wide population cohort (biomarcare)

Abstract Importance Risk stratification for stroke remains difficult. Classic risk factors seem to only capture a fraction of the complex causative mechanisms. Metabolomic profiling offers the potential to detect new biomarkers and improve risk assessment of incident stroke. Objective and Methods To evaluate the association between circulating metabolites and incident stroke in a large European population cohort (BiomarCaRE). This study utilized the BiomarCaRE case-cohort to measure circulating metabolites. Metabolites were measured in serum samples from 10 299 individuals without a history of stroke. The cohort consisted of a weighted, random subcohort of the original cohort of more than 70 000 individuals. The case-cohort design was applied to 6 European cohorts. Associations with time to stroke onset were assessed individually by applying weighted and adjusted Cox proportional hazard models. The association of metabolites with incident stroke was examined using Cox regressions and C statistics. Results In 10 299 individuals (4069 women (39.5%); median (interquartile range) age, 56.8 (49.5, 62.4), 1516 incident stroke (14.7%) occurred over a median (interquartile range) follow-up time of 8.9 (4.4, 14.7) years. 141 metabolites were analyzed and 26 were significantly associated with incident stroke at a nominal P value of .05, including phosphatidylcholines (PCs), lysoPCs, amino acids, and sphingolipids. Six metabolites (2 phosphatidylcholines, 2 lysophosphatidylcholines, 1 hydroxysphingomyeline and glutamic acid) were significantly associated with incident stroke: diacyl-PC C32:1 (hazard ratio (HR), 1.26 [95% CI, 1.15, 1.38]), SM (OH) C14:1 (HR, 0.77 [95% CI, 0.68, 0.87]), lysoPC a C18:2 (HR, 0.74 [95% CI, 0.64, 0.86]), glutamic acid (HR, 1.28 [95% CI, 1.13, 1.44]), lysoPC a C17:0 (HR, 0.77 [95% CI, 0.68, 0.88]) and diacyl-PC C34:1 (HR, 1.24 [95% CI, 1.11, 1.39]). The strength of the associations competes with those of classic risk factors (C statistics: diacyl-PC C32:1, 0.795 [95% CI, 0.762, 0.828], SM (OH) C14:1, 0.792 [95% CI, 0.759, 0.825], lysoPC a C18:2, 0.791 [95% CI, 0.757, 0.824], Glu, 0.794 [95% CI, 0.760, 0.827]), lysoPC a C17:0, 0.791 [95% CI, 0.758, 0.824], diacyl-PC C34:1, 0.793 [95% CI, 0.760, 0.826]). Adding metabolites to a base risk model including classic risk factors, high-sensitivity C-reactive protein and high-sensitivity troponin I did not improve discrimination by C statistics. Conclusions and Relevance This is the first study assessing &amp;gt;100 metabolites in a large population cohort of &amp;gt;10,000 individuals for their association with stroke risk prediction. Six metabolites were significantly associated with risk of incident stroke and showed comparable discrimination with individual classic risk factors. Although these metabolites may not improve stroke risk assessment beyond that of classic risk factors, these findings hold promise for an improved understanding of the early pathophysiology of stroke.

Incidence of fatal stroke 24 years after acute coronary syndrome: the value of baseline albuminuria and hypertension. The ABC study on acute coronary syndrome

Abstract Background Albuminuria and hypertension are known independent predictors of stroke risk. Purpose To investigate the additive effect of baseline albuminuria and hypertension on the very long-term risk prediction of fatal stroke (FS) in patients with acute coronary syndrome (ACS). Methods The present analysis includes 589 patients admitted with ACS to 3 Italian hospitals and discharged alive. Patients were followed prospectively for 24 years or until death. Baseline clinical and laboratory data were recorded within the first 7 days of hospitalization. Results All but three patients completed the follow-up representing 7065 person-years and 44 patients (7%) developed FS. They shared most of the baseline clinical characteristics with the patients who did not have FS, except for being significantly older (mean age was 72±8 vs. 66±12 years, p=0.0008). Patients with FS also had albuminuria more frequently (66% vs. 49%, p=0.03) with significantly higher values of 3rd-day albumin–creatinine ratio (ACR), (p=0.009). At Cox regression analysis; patients who had hypertension or albuminuria had a higher risk of developing FS, hazard ratios were (HR=1.8; 95%CI=1.0–3.2; p=0.05 and HR=2.9; 95%CI=1.5–5.3; p=0.001 respectively. Patients with both hypertension and albuminuria showed an independent risk association for developing FS than patients with neither HR=4.2; 95%CI=1.8–9.7; p=0.001. Results kept true with a model adjusted for age, gender and presence of atrial fibrillation. The same results were obtained with the competing risk analysis (HR=2.6; 95%CI=1.1–6.0, p=0.02). Conclusions the results showed that the combination of hypertension and albuminuria is associated with an independent greater risk of developing a fatal stroke long after ACS.

Carotid plaque score for stroke and cardiovascular risk prediction in a middle-aged cohort from the general population: the akershus cardiac examination 1950 study

Abstract Background Assessment of carotid atherosclerosis with ultrasound may have an unfilled potential in predicting stroke and cardiovascular events. Purpose We aimed to explore the predictive value of the carotid plaque score compared to the Systematic COronary Risk Evaluation (SCORE) 2 risk prediction algorithm, on incident stroke and major adverse cardiovascular events (MACE), and establish a prognostic cut-off for the carotid plaque score. Methods In the prospective Akershus Cardiac Eexamination (ACE) 1950 cohort study, carotid plaque score was calculated with ultrasound at inclusion in 2012-2015. The largest plaque diameter in each extra cranial segment of the carotid artery both sides was measured and scored from 0-3 points. The sum of each segment provided the carotid plaque score. Incident stroke and composite endpoint MACE (nonfatal stroke, nonfatal myocardial infarction, and cardiovascular death) was assessed by linkage to national registries throughout 2020. Results Carotid plaque score was available in 3650 (98.5%) of the participants, with mean age at inclusion was 63.9±0.64 years. Only 462 (12.7%) were free of plaque and the median carotid plaque score was 2 (IQR 1-4). By the end of 2020, 42 (1.2%) subjects had experienced a stroke and 145 (4.0%) a MACE. Carotid plaque score was a predictor of incident stroke (HR 1.25, 95%CI 1.15-1.36) and MACE (OR 1.21, 95%CI 1.14-1.27) after adjustment for SCORE2, and outperformed SCORE2 in predicting stroke (p=0.001). The best cut-off value for carotid plaque score determined by receiver operator characteristics area under the curve was ≥4 with a positive predictive value 2.5% and negative predictive value 99.3%. Conclusion Carotid plaque score is a strong predictor of incident stroke and MACE and outperforms SCORE2 for risk prediction in a middle-aged cohort recruited from the general population. A cut-off score of ≥4 seems to be suitable to identify high-risk subjects.Graphical AbstractCarotid plaque score distribution

The association between HbA1c variability and strokes in patients receiving anticoagulants

Abstract Background Diabetes, higher glycaemic level, is associated with an increased risk of stroke in patients with atrial fibrillation (AF) and is integrated into the CHA2DS2-VASc score for stroke risk prediction. However, the association between glycaemic variability and stroke risk in AF patients remains unclear. Objective To evaluate the association between glycaemic variability and stroke among diabetic and non-diabetic patients with nonvalvular AF receiving anticoagulants. Methods Newly diagnosed AF from 2013-2018 (n=8,790) were divided by their diabetes status. Variability in glycaemic, represented by coefficient of variation (CV) on HbA1c (haemoglobin A1c), was assessed by at least three HbA1c measurements following enrollment to the end of the follow-up period. Participants with ischaemic stroke (IS)/systemic embolism (SE), haemorrhagic stroke (HS) or all-cause mortality at baseline or within 1 year of enrollment were excluded. Multivariate Cox regression was performed to evaluate the association between variability in HbA1c and subsequent incident IS/SE, HS and all-cause mortality. Sensitivity analyses included defining HbA1c variability using average successive variability (ASV) and variability independent of the mean (VIM). Results Diabetes status significantly modified the relationship between HbA1c variability and risk of IS/SE (Pinteraction = 0.36)and all-mortality (Pinteraction &amp;lt;0.01). Greater incidence rates were observed among individuals with diabetes, compared to those without diabetes (3.74 vs. 2.41 for IS/SE, 0.54 vs. 0.33 for HS and 4.89 vs. 2.42 for all-cause mortality per 100 person years). Among those with diabetes, greater HbA1c variability was significantly associated with increased risk of IS/SE (hazard ratio [HR] = 1.21, 95% confidence interval [CI]: 1.11-1.31), all-cause mortality (HR = 1.13, 95%CI: 1.05-1.20) and HS (HR = 1.10, 95%CI: 1.02-1.23). Similar patterns were found in individuals without diabetes (IS/SE: HR = 1.28, 95%CI: 1.08-1.52; all-cause mortality: HR = 1.50, 95%CI: 1.29-1.75). However, no significant association was observed between glycemia variability and HS among those without diabetes. Results were consistent with alternative measures of HbA1c variability. Conclusion Variability in HbA1c was independently associated with the risk of IS/SE and all-cause mortality among patients with AF, regardless of diabetes status.

Abstract 17169: Beyond CHA2DS2-VASc: Independent Predictors of Ischemic Stroke in Patients With Atrial Fibrillation

Introduction: Factors outside of the CHA 2 DS 2 -VASc score also contribute to stroke risk, and patients prescribed anticoagulants still face a residual stroke risk. Methods: A multivariable model to predict stroke risk was built using data from patients treated for atrial fibrillation (AF) at Duke between 2009 and 2018. The model was then refined in patients presenting with incident AF at 7 hospitals within the larger PCORnet database. Results: Of 43,896 patients with AF included in our multivariable model creation data set, median age was 71.6 years, 44.2% of patients were female, 17.4% were black, and median CHA 2 DS 2 -VASc was 3. DOAC was prescribed in 24.2%, warfarin in 28.7%, and 50.4% were not prescribed either anticoagulant. The 5-year incidence of ischemic stroke was 3.5%. In a multivariable model, patients who were black, had prior major bleeding, or who were prescribed aspirin or an SSRI were at increased risk for ischemic stroke regardless of type of anticoagulation, whereas lower eGFR did not increase the risk. The model was refined in the PCORnet database using 244,213 patients with incident AF, with additional identification of lower risk for patients with BMI &gt;30 or recent cardiac surgery, and use of P2Y12 as a risk factor. Multivariable models showed substantial improvement in stroke prediction within both the Duke and PCORnet treatment cohorts, with improvement in C-indices to 69.6-75.7%, from baseline C-indices of 63.5%-68.7% for CHA 2 DS 2 -VASc alone among patients on DOAC, warfarin, or who were untreated. Conclusions: Lower BMI, prior major bleeding, antiplatelet (ASA or P2Y12) use, and black race were significantly associated with increased risk for ischemic stroke, independent of the commonly used CHA 2 DS 2 -VASc score. Future AF research should attempt to integrate these emerging risk factors with the existing CHA 2 DS 2 -VASc framework to improve the identification of patients at high risk of stroke and in greatest need of anticoagulation.

Abstract 19134: Association Between CHA 2 DS 2 -VASc and LA Volume Index: Insights From DECAAF II

Introduction: LA Volume is strongly associated with atrial fibrillation (AF) and underlying atrial myopathy. The CHA 2 DS 2 -VASc score predicts stroke risk in patients with AF. The aim of this study is to investigate the relationship between the two. Methods: In this subanalysis of the DECAAF II trial, patients with persistent AF undergoing first-time catheter ablation underwent a baseline late gadolinium enhancement magnetic resonance imaging (LGE-MRI) within one month prior to ablation. LA Volume Index (LAVI) was defined as LA Volume on the MRI prior to ablation divided by body surface area ( mL/m2 ). The CHA2 DS2 -VASc score was also calculated at baseline prior to ablation. Analysis of Variance (ANOVA) test and linear regression were used to study the association between the two. Results: 815 patients were included, (78.9% males, mean age 62.7 ± 9.1 years) (table). The mean LAVI was 62.2 ± 18.9 mL/m2, and mean CHA2 DS2 -VASc score was 1.7 ± 1.27. The ANOVA test showed significant differences between group (F = 7.71, p &lt;0.001). For every point increase in CHA2 DS2 -VASc score, LAVI increases by 3 (LAVI = 2.99 * CHA2 DS2 -VASc + 57; p &lt;0.001) (figure). Conclusions: Our data suggests that structural remodeling in patients with AF is likely driven by multiple factors. There was a clear association between CHA2 DS2 -VASc score and higher LAVI. Serial analysis of LA Volume at different time points is needed to further understand the impact of individual comorbidities on LA volume dilation.

Abstract 18629: Discriminative Accuracy of Cha 2 Ds 2 Vasc Score, and Development of Predictive Accuracy Model Using Machine Learning for Ischemic Stroke in Cardiac Amyloidosis

Background: Cardiac amyloidosis (CA) in conjunction with atrial fibrillation (AF) presents unique management challenges. CHA2DS2VASc score in these patients is believed to underestimate the risk of ischemic stroke, necessitating a better predictive model in these patients. Methods: Data was obtained from the National Readmission Database (NRD). Outcomes between CA-AF and no-CA-AF were compared using multivariate regression analysis to calculate adjusted odds ratios (aOR). AutoScore; an interpretable machine learning framework, was used to develop a stroke risk prediction model, the predictive accuracy of which was evaluated with an area under the curve (AUC) using the receiver operating characteristic analysis. Results: A total of 11,860,804 (CA-AF 22,687 [0.19%] and no-CA-AF 11,838,117) patients were identified from 2015-2019. The adjusted odds of mortality (aOR 1.41 and 1.29), stroke (aOR 1.78 and 1.74), non-intracranial hemorrhage (aOR 2.10 and aOR 1.85), and intracranial hemorrhage (aOR 14.4 and aOR 4.26) were significantly higher in CA-AF compared with non-CA-AF at both index admission and 30-days, respectively. The CHA 2 DS 2 VASc score had a poor discriminative accuracy for stroke at 30-days in CA-AF (AUC 49%, 95%CI 47%-51%, p=0.54). The machine learning autoscore integrative model revealed that the predictive ability of our newly proposed E-CHADS score (end-stage renal disease (ESRD), congestive heart failure, hypertension, active cancer, dementia, and diabetes mellitus) for 30-day risk of ischemic stroke in CA-AF was excellent (for a cutoff of 52 points random forest score) with an AUC of 80% (95%CI 74%-86%) Conclusion: Cardiac amyloidosis carries a high risk of ischemic stroke that is not accurately predicted by the CHA 2 DS 2 VASc score. Our proposed model (E-CHADS) identifies 3 new variables (ESRD, dementia, and cancer) that have higher discriminative accuracy for ischemic stroke in these patients.