Discordance among apoB, non-high-density lipoprotein cholesterol, and triglycerides: implications for cardiovascular prevention.

BACKGROUND AND AIMS: Despite growing evidence that apolipoprotein B (apoB) is the most accurate marker of atherosclerotic cardiovascular disease (ASCVD) risk, its adoption in clinical practice has been low. This investigation sought to determine whether low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol (HDL-C), and triglycerides are sufficient for routine cardiovascular care. METHODS: A sample of 293 876 UK Biobank adults (age: 40-73 years, 42% men), free of cardiovascular disease, with a median follow-up for new-onset ASCVD of 11 years was included. Distribution of apoB at pre-specified levels of LDL-C, non-HDL-C, and triglycerides was examined graphically, and 10-year ASCVD event rates were compared for high vs. low apoB. Residuals of apoB were constructed after regressing apoB on LDL-C, non-HDL-C, and log-transformed triglycerides and used as predictors in a proportional hazards regression model for new-onset ASCVD adjusted for standard risk factors, including HDL-C. RESULTS: ApoB was highly correlated with LDL-C and non-HDL-C (Pearson's r = .96, P < .001 for both) but less so with log triglycerides (r = .42, P < .001). However, apoB ranges necessary to capture 95% of all observations at pre-specified levels of LDL-C, non-HDL-C, or triglycerides were wide, spanning 85.8-108.8 md/dL when LDL-C 130 mg/dL, 88.3-112.4 mg/dL when non-HDL-C 160 mg/dL, and 67.8-147.4 md/dL when triglycerides 115 mg/dL. At these levels (±10 mg/dL), 10-year ASCVD rates for apoB above mean + 1 SD vs. below mean - 1 SD were 7.3 vs. 4.0 for LDL-C, 6.4 vs. 4.6 for non-HDL-C, and 7.0 vs. 4.6 for triglycerides (all P < .001). With 19 982 new-onset ASCVD events on follow-up, in the adjusted model, residual apoB remained statistically significant after accounting for LDL-C and HDL-C (hazard ratio 1.06, 95% confidence interval 1.0-1.07), after accounting for non-HDL-C and HDL-C (hazard ratio 1.04, 95% confidence interval 1.03-1.06), and after accounting for triglycerides and HDL-C (hazard ratio 1.13, 95% confidence interval 1.12-1.15). None of the residuals of LDL-C, non-HDL-C, or of log triglycerides remained significant when apoB was included in the model. CONCLUSIONS: High variability of apoB at individual levels of LDL-C, non-HDL-C, and triglycerides coupled with meaningful differences in 10-year ASCVD rates and significant residual information contained in apoB for prediction of new-onset ASCVD events demonstrate that LDL-C, non-HDL-C, and triglycerides are not adequate proxies for apoB in clinical care.

Development and Validation of the American Heart Association's PREVENT Equations.

BACKGROUND: Multivariable equations are recommended by primary prevention guidelines to assess absolute risk of cardiovascular disease (CVD). However, current equations have several limitations. Therefore, we developed and validated the American Heart Association Predicting Risk of CVD EVENTs (PREVENT) equations among US adults 30 to 79 years of age without known CVD. METHODS: The derivation sample included individual-level participant data from 25 data sets (N=3 281 919) between 1992 and 2017. The primary outcome was CVD (atherosclerotic CVD and heart failure). Predictors included traditional risk factors (smoking status, systolic blood pressure, cholesterol, antihypertensive or statin use, and diabetes) and estimated glomerular filtration rate. Models were sex-specific, race-free, developed on the age scale, and adjusted for competing risk of non-CVD death. Analyses were conducted in each data set and meta-analyzed. Discrimination was assessed using the Harrell C-statistic. Calibration was calculated as the slope of the observed versus predicted risk by decile. Additional equations to predict each CVD subtype (atherosclerotic CVD and heart failure) and include optional predictors (urine albumin-to-creatinine ratio and hemoglobin A1c), and social deprivation index were also developed. External validation was performed in 3 330 085 participants from 21 additional data sets. RESULTS: Among 6 612 004 adults included, mean±SD age was 53±12 years, and 56% were women. Over a mean±SD follow-up of 4.8±3.1 years, there were 211 515 incident total CVD events. The median C-statistics in external validation for CVD were 0.794 (interquartile interval, 0.763-0.809) in female and 0.757 (0.727-0.778) in male participants. The calibration slopes were 1.03 (interquartile interval, 0.81-1.16) and 0.94 (0.81-1.13) among female and male participants, respectively. Similar estimates for discrimination and calibration were observed for atherosclerotic CVD- and heart failure-specific models. The improvement in discrimination was small but statistically significant when urine albumin-to-creatinine ratio, hemoglobin A1c, and social deprivation index were added together to the base model to total CVD (ΔC-statistic [interquartile interval] 0.004 [0.004-0.005] and 0.005 [0.004-0.007] among female and male participants, respectively). Calibration improved significantly when the urine albumin-to-creatinine ratio was added to the base model among those with marked albuminuria (>300 mg/g; 1.05 [0.84-1.20] versus 1.39 [1.14-1.65]; P=0.01). CONCLUSIONS: PREVENT equations accurately and precisely predicted risk for incident CVD and CVD subtypes in a large, diverse, and contemporary sample of US adults by using routinely available clinical variables.

Translating ethical and quality principles for the effective, safe and fair development, deployment and use of artificial intelligence technologies in healthcare.

OBJECTIVE: The complexity and rapid pace of development of algorithmic technologies pose challenges for their regulation and oversight in healthcare settings. We sought to improve our institution's approach to evaluation and governance of algorithmic technologies used in clinical care and operations by creating an Implementation Guide that standardizes evaluation criteria so that local oversight is performed in an objective fashion. MATERIALS AND METHODS: Building on a framework that applies key ethical and quality principles (clinical value and safety, fairness and equity, usability and adoption, transparency and accountability, and regulatory compliance), we created concrete guidelines for evaluating algorithmic technologies at our institution. RESULTS: An Implementation Guide articulates evaluation criteria used during review of algorithmic technologies and details what evidence supports the implementation of ethical and quality principles for trustworthy health AI. Application of the processes described in the Implementation Guide can lead to algorithms that are safer as well as more effective, fair, and equitable upon implementation, as illustrated through 4 examples of technologies at different phases of the algorithmic lifecycle that underwent evaluation at our academic medical center. DISCUSSION: By providing clear descriptions/definitions of evaluation criteria and embedding them within standardized processes, we streamlined oversight processes and educated communities using and developing algorithmic technologies within our institution. CONCLUSIONS: We developed a scalable, adaptable framework for translating principles into evaluation criteria and specific requirements that support trustworthy implementation of algorithmic technologies in patient care and healthcare operations.

Novel Prediction Equations for Absolute Risk Assessment of Total Cardiovascular Disease Incorporating Cardiovascular-Kidney-Metabolic Health: A Scientific Statement From the American Heart Association.

Cardiovascular-kidney-metabolic (CKM) syndrome is a novel construct recently defined by the American Heart Association in response to the high prevalence of metabolic and kidney disease. Epidemiological data demonstrate higher absolute risk of both atherosclerotic cardiovascular disease (CVD) and heart failure as an individual progresses from CKM stage 0 to stage 3, but optimal strategies for risk assessment need to be refined. Absolute risk assessment with the goal to match type and intensity of interventions with predicted risk and expected treatment benefit remains the cornerstone of primary prevention. Given the growing number of therapies in our armamentarium that simultaneously address all 3 CKM axes, novel risk prediction equations are needed that incorporate predictors and outcomes relevant to the CKM context. This should also include social determinants of health, which are key upstream drivers of CVD, to more equitably estimate and address risk. This scientific statement summarizes the background, rationale, and clinical implications for the newly developed sex-specific, race-free risk equations: PREVENT (AHA Predicting Risk of CVD Events). The PREVENT equations enable 10- and 30-year risk estimates for total CVD (composite of atherosclerotic CVD and heart failure), include estimated glomerular filtration rate as a predictor, and adjust for competing risk of non-CVD death among adults 30 to 79 years of age. Additional models accommodate enhanced predictive utility with the addition of CKM factors when clinically indicated for measurement (urine albumin-to-creatinine ratio and hemoglobin A1c) or social determinants of health (social deprivation index) when available. Approaches to implement risk-based prevention using PREVENT across various settings are discussed.

Facilitating Harmonization of Variables in Framingham, MESA, ARIC, and REGARDS Studies Through a Metadata Repository.

BACKGROUND: High-quality research in cardiovascular prevention, as in other fields, requires inclusion of a broad range of data sets from different sources. Integrating and harmonizing different data sources are essential to increase generalizability, sample size, and representation of understudied populations-strengthening the evidence for the scientific questions being addressed. METHODS: Here, we describe an effort to build an open-access repository and interactive online portal for researchers to access the metadata and code harmonizing data from 4 well-known cohort studies-the REGARDS (Reasons for Geographic and Racial Differences in Stroke) study, FHS (Framingham Heart Study), MESA (Multi-Ethnic Study of Atherosclerosis), and ARIC (Atherosclerosis Risk in Communities) study. We introduce a methodology and a framework used for preprocessing and harmonizing variables from multiple studies. RESULTS: We provide a real-case study and step-by-step guidance to demonstrate the practical utility of our repository and interactive web page. In addition to our successful development of such an open-access repository and interactive web page, this exercise in harmonizing data from multiple cohort studies has revealed several key themes. These themes include the importance of careful preprocessing and harmonization of variables, the value of creating an open-access repository to facilitate collaboration and reproducibility, and the potential for using harmonized data to address important scientific questions and disparities in cardiovascular disease research. CONCLUSIONS: By integrating and harmonizing these large-scale cohort studies, such a repository may improve the statistical power and representation of understudied cohorts, enabling development and validation of risk prediction models, identification and investigation of risk factors, and creating a platform for racial disparities research. REGISTRATION: URL: https://precision.heart.org/duke-ninds.

Mitigating Racial And Ethnic Bias And Advancing Health Equity In Clinical Algorithms: A Scoping Review.

In August 2022 the Department of Health and Human Services (HHS) issued a notice of proposed rulemaking prohibiting covered entities, which include health care providers and health plans, from discriminating against individuals when using clinical algorithms in decision making. However, HHS did not provide specific guidelines on how covered entities should prevent discrimination. We conducted a scoping review of literature published during the period 2011-22 to identify health care applications, frameworks, reviews and perspectives, and assessment tools that identify and mitigate bias in clinical algorithms, with a specific focus on racial and ethnic bias. Our scoping review encompassed 109 articles comprising 45 empirical health care applications that included tools tested in health care settings, 16 frameworks, and 48 reviews and perspectives. We identified a wide range of technical, operational, and systemwide bias mitigation strategies for clinical algorithms, but there was no consensus in the literature on a single best practice that covered entities could employ to meet the HHS requirements. Future research should identify optimal bias mitigation methods for various scenarios, depending on factors such as patient population, clinical setting, algorithm design, and types of bias to be addressed.

A Synopsis of the Evidence for the Science and Clinical Management of Cardiovascular-Kidney-Metabolic (CKM) Syndrome: A Scientific Statement From the American Heart Association.

A growing appreciation of the pathophysiological interrelatedness of metabolic risk factors such as obesity and diabetes, chronic kidney disease, and cardiovascular disease has led to the conceptualization of cardiovascular-kidney-metabolic syndrome. The confluence of metabolic risk factors and chronic kidney disease within cardiovascular-kidney-metabolic syndrome is strongly linked to risk for adverse cardiovascular and kidney outcomes. In addition, there are unique management considerations for individuals with established cardiovascular disease and coexisting metabolic risk factors, chronic kidney disease, or both. An extensive body of literature supports our scientific understanding of, and approach to, prevention and management for individuals with cardiovascular-kidney-metabolic syndrome. However, there are critical gaps in knowledge related to cardiovascular-kidney-metabolic syndrome in terms of mechanisms of disease development, heterogeneity within clinical phenotypes, interplay between social determinants of health and biological risk factors, and accurate assessments of disease incidence in the context of competing risks. There are also key limitations in the data supporting the clinical care for cardiovascular-kidney-metabolic syndrome, particularly in terms of early-life prevention, screening for risk factors, interdisciplinary care models, optimal strategies for supporting lifestyle modification and weight loss, targeting of emerging cardioprotective and kidney-protective therapies, management of patients with both cardiovascular disease and chronic kidney disease, and the impact of systematically assessing and addressing social determinants of health. This scientific statement uses a crosswalk of major guidelines, in addition to a review of the scientific literature, to summarize the evidence and fundamental gaps related to the science, screening, prevention, and management of cardiovascular-kidney-metabolic syndrome.

Cardiovascular-Kidney-Metabolic Health: A Presidential Advisory From the American Heart Association.

Cardiovascular-kidney-metabolic health reflects the interplay among metabolic risk factors, chronic kidney disease, and the cardiovascular system and has profound impacts on morbidity and mortality. There are multisystem consequences of poor cardiovascular-kidney-metabolic health, with the most significant clinical impact being the high associated incidence of cardiovascular disease events and cardiovascular mortality. There is a high prevalence of poor cardiovascular-kidney-metabolic health in the population, with a disproportionate burden seen among those with adverse social determinants of health. However, there is also a growing number of therapeutic options that favorably affect metabolic risk factors, kidney function, or both that also have cardioprotective effects. To improve cardiovascular-kidney-metabolic health and related outcomes in the population, there is a critical need for (1) more clarity on the definition of cardiovascular-kidney-metabolic syndrome; (2) an approach to cardiovascular-kidney-metabolic staging that promotes prevention across the life course; (3) prediction algorithms that include the exposures and outcomes most relevant to cardiovascular-kidney-metabolic health; and (4) strategies for the prevention and management of cardiovascular disease in relation to cardiovascular-kidney-metabolic health that reflect harmonization across major subspecialty guidelines and emerging scientific evidence. It is also critical to incorporate considerations of social determinants of health into care models for cardiovascular-kidney-metabolic syndrome and to reduce care fragmentation by facilitating approaches for patient-centered interdisciplinary care. This presidential advisory provides guidance on the definition, staging, prediction paradigms, and holistic approaches to care for patients with cardiovascular-kidney-metabolic syndrome and details a multicomponent vision for effectively and equitably enhancing cardiovascular-kidney-metabolic health in the population.

Concordance/discordance between serum apolipoprotein B, low density lipoprotein cholesterol and non-high density lipoprotein cholesterol in NATPOL 2011 participants - An epidemiological perspective.

BACKGROUND: The study compared the distribution of serum LDL-C, non-HDL-C, and apolipoprotein B (apoB) among participants of the NATPOL 2011 survey and analysed concordance/discordance of results in the context of the risk for atherosclerotic cardiovascular disease (ASCVD). METHODS: Serum levels of apoB, LDL-C, non-HDL-C and small dense LDL-C were measured/calculated in 2067-2098 survey participants. The results were compared between women and men, age groups and in relation to body mass index (BMI), fasting glucose and TG levels, and the presence of CVD. Percentile distribution of lipid levels and concordance/discordance analysis were based on medians and ESC/EAS 2019 target thresholds for ASCVD risk and on comparison of measured apoB levels and levels calculated from linear regression equations with serum LDL- C and non-HDL-C as independent variables. RESULTS: Serum apoB, LDL-C and non-HDL-C were similarly related to sex, age, BMI, visceral obesity, cardiovascular disease, and fasting glucose and triglyceride levels. Serum apoB, LDL-C and non-HDL-C very high- and moderate- target thresholds were exceeded in 83%, 99% and 96.9% and in 41%, 75% and 63.7% of subjects, respectively. The incidence of the discordances between the results depended on the dividing values used and ranged from 0.2% to 45.2% of the respondents. Subjects with high apoB / low LDL-C/non-HDL-C discordance had features of metabolic syndrome. CONCLUSIONS: Diagnostic discordances between apoB and LDL-C/non-HDL-C indicate limitations of serum LDL-C/non-HDL-C in ASCVD risk management. Due to the high apoB/low LDL-C/non-HDL-C discordance, obese/metabolic syndrome patients may benefit from replacing LDL-C/non-HDL-C by apoB in ASCVD risk assessment and lipid-lowering therapy.

Semi-supervised calibration of noisy event risk (SCANER) with electronic health records.

OBJECTIVE: Electronic health records (EHR), containing detailed longitudinal clinical information on a large number of patients and covering broad patient populations, open opportunities for comprehensive predictive modeling of disease progression and treatment response. However, since EHRs were originally constructed for administrative purposes not for research, in the EHR-linked studies, it is often not feasible to capture reliable information for analytical variables, especially in the survival setting, when both accurate event status and event times are needed for model building. For example, progression-free survival (PFS), a commonly used survival outcome for cancer patients, often involves complex information embedded in free-text clinical notes and cannot be extracted reliably. Proxies of PFS time such as time to the first mention of progression in the notes are at best good approximations to the true event time. This leads to difficulty in efficiently estimating event rates for an EHR patient cohort. Estimating survival rates based on error-prone outcome definitions can lead to biased results and hamper the power in the downstream analysis. On the other hand, extracting accurate event time information via manual annotation is time and resource intensive. The objective of this study is to develop a calibrated survival rate estimator using noisy outcomes from EHR data. MATERIALS AND METHODS: In this paper, we propose a two-stage semi-supervised calibration of noisy event rate (SCANER) estimator that can effectively overcome censoring induced dependency and attains more robust performance (i.e., not sensitive to misspecification of the imputation model) by fully utilizing both a small-labeled set of gold-standard survival outcomes annotated via manual chart review and a set of proxy features automatically captured via EHR in the unlabeled set. We validate the SCANER estimator by estimating the PFS rates for a virtual cohort of lung cancer patients from one large tertiary care center and the ICU-free survival rates for COVID patients from two large tertiary care centers. RESULTS: In terms of survival rate estimates, the SCANER had very similar point estimates compared to the complete-case Kaplan Meier estimator. On the other hand, other benchmark methods for comparison, which fail to account for the induced dependency between event time and the censoring time conditioning on surrogate outcomes, produced biased results across all three case studies. In terms of standard errors, the SCANER estimator was more efficient than the KM estimator, with up to 50% efficiency gain. CONCLUSION: The SCANER estimator achieves more efficient, robust, and accurate survival rate estimates compared to existing approaches. This promising new approach can also improve the resolution (i.e., granularity of event time) by using labels conditioning on multiple surrogates, particularly among less common or poorly coded conditions.

Cardiovascular Disease Risk Assessment Using Traditional Risk Factors and Polygenic Risk Scores in the Million Veteran Program.

Importance: Primary prevention of atherosclerotic cardiovascular disease (ASCVD) relies on risk stratification. Genome-wide polygenic risk scores (PRSs) are proposed to improve ASCVD risk estimation. Objective: To determine whether genome-wide PRSs for coronary artery disease (CAD) and acute ischemic stroke improve ASCVD risk estimation with traditional clinical risk factors in an ancestrally diverse midlife population. Design, Setting, and Participants: This was a prognostic analysis of incident events in a retrospectively defined longitudinal cohort conducted from January 1, 2011, to December 31, 2018. Included in the study were adults free of ASCVD and statin naive at baseline from the Million Veteran Program (MVP), a mega biobank with genetic, survey, and electronic health record data from a large US health care system. Data were analyzed from March 15, 2021, to January 5, 2023. Exposures: PRSs for CAD and ischemic stroke derived from cohorts of largely European descent and risk factors, including age, sex, systolic blood pressure, total cholesterol, high-density lipoprotein (HDL) cholesterol, smoking, and diabetes status. Main Outcomes and Measures: Incident nonfatal myocardial infarction (MI), ischemic stroke, ASCVD death, and composite ASCVD events. Results: A total of 79 151 participants (mean [SD] age, 57.8 [13.7] years; 68 503 male [86.5%]) were included in the study. The cohort included participants from the following harmonized genetic ancestry and race and ethnicity categories: 18 505 non-Hispanic Black (23.4%), 6785 Hispanic (8.6%), and 53 861 non-Hispanic White (68.0%) with a median (5th-95th percentile) follow-up of 4.3 (0.7-6.9) years. From 2011 to 2018, 3186 MIs (4.0%), 1933 ischemic strokes (2.4%), 867 ASCVD deaths (1.1%), and 5485 composite ASCVD events (6.9%) were observed. CAD PRS was associated with incident MI in non-Hispanic Black (hazard ratio [HR], 1.10; 95% CI, 1.02-1.19), Hispanic (HR, 1.26; 95% CI, 1.09-1.46), and non-Hispanic White (HR, 1.23; 95% CI, 1.18-1.29) participants. Stroke PRS was associated with incident stroke in non-Hispanic White participants (HR, 1.15; 95% CI, 1.08-1.21). A combined CAD plus stroke PRS was associated with ASCVD deaths among non-Hispanic Black (HR, 1.19; 95% CI, 1.03-1.17) and non-Hispanic (HR, 1.11; 95% CI, 1.03-1.21) participants. The combined PRS was also associated with composite ASCVD across all ancestry groups but greater among non-Hispanic White (HR, 1.20; 95% CI, 1.16-1.24) than non-Hispanic Black (HR, 1.11; 95% CI, 1.05-1.17) and Hispanic (HR, 1.12; 95% CI, 1.00-1.25) participants. Net reclassification improvement from adding PRS to a traditional risk model was modest for the intermediate risk group for composite CVD among men (5-year risk >3.75%, 0.38%; 95% CI, 0.07%-0.68%), among women, (6.79%; 95% CI, 3.01%-10.58%), for age older than 55 years (0.25%; 95% CI, 0.03%-0.47%), and for ages 40 to 55 years (1.61%; 95% CI, -0.07% to 3.30%). Conclusions and Relevance: Study results suggest that PRSs derived predominantly in European samples were statistically significantly associated with ASCVD in the multiancestry midlife and older-age MVP cohort. Overall, modest improvement in discrimination metrics were observed with addition of PRSs to traditional risk factors with greater magnitude in women and younger age groups.

Needles in a Haystack: Finding Qualitative and Quantitative Collaborators in Academic Medical Centers.

Translational research is a data-driven process that involves transforming scientific laboratory- and clinic-based discoveries into products and activities with real-world impact to improve individual and population health. Successful execution of translational research requires collaboration between clinical and translational science researchers, who have expertise in a wide variety of domains across the field of medicine, and qualitative and quantitative scientists, who have specialized methodologic expertise across diverse methodologic domains. While many institutions are working to build networks of these specialists, a formalized process is needed to help researchers navigate the network to find the best match and to track the navigation process to evaluate an institution's unmet collaborative needs. In 2018, a novel analytic resource navigation process was developed at Duke University to connect potential collaborators, leverage resources, and foster a community of researchers and scientists. This analytic resource navigation process can be readily adopted by other academic medical centers. The process relies on navigators with broad qualitative and quantitative methodologic knowledge, strong communication and leadership skills, and extensive collaborative experience. The essential elements of the analytic resource navigation process are as follows: (1) strong institutional knowledge of methodologic expertise and access to analytic resources, (2) deep understanding of research needs and methodologic expertise, (3) education of researchers on the role of qualitative and quantitative scientists in the research project, and (4) ongoing evaluation of the analytic resource navigation process to inform improvements. Navigators help researchers determine the type of expertise needed, search the institution to find potential collaborators with that expertise, and document the process to evaluate unmet needs. Although the navigation process can create a basis for an effective solution, some challenges remain, such as having resources to train navigators, comprehensively identifying all potential collaborators, and keeping updated information about resources as methodologists join and leave the institution.

Predictive Accuracy of Stroke Risk Prediction Models Across Black and White Race, Sex, and Age Groups.

Importance: Stroke is the fifth-highest cause of death in the US and a leading cause of serious long-term disability with particularly high risk in Black individuals. Quality risk prediction algorithms, free of bias, are key for comprehensive prevention strategies. Objective: To compare the performance of stroke-specific algorithms with pooled cohort equations developed for atherosclerotic cardiovascular disease for the prediction of new-onset stroke across different subgroups (race, sex, and age) and to determine the added value of novel machine learning techniques. Design, Setting, and Participants: Retrospective cohort study on combined and harmonized data from Black and White participants of the Framingham Offspring, Atherosclerosis Risk in Communities (ARIC), Multi-Ethnic Study for Atherosclerosis (MESA), and Reasons for Geographical and Racial Differences in Stroke (REGARDS) studies (1983-2019) conducted in the US. The 62 482 participants included at baseline were at least 45 years of age and free of stroke or transient ischemic attack. Exposures: Published stroke-specific algorithms from Framingham and REGARDS (based on self-reported risk factors) as well as pooled cohort equations for atherosclerotic cardiovascular disease plus 2 newly developed machine learning algorithms. Main Outcomes and Measures: Models were designed to estimate the 10-year risk of new-onset stroke (ischemic or hemorrhagic). Discrimination concordance index (C index) and calibration ratios of expected vs observed event rates were assessed at 10 years. Analyses were conducted by race, sex, and age groups. Results: The combined study sample included 62 482 participants (median age, 61 years, 54% women, and 29% Black individuals). Discrimination C indexes were not significantly different for the 2 stroke-specific models (Framingham stroke, 0.72; 95% CI, 0.72-073; REGARDS self-report, 0.73; 95% CI, 0.72-0.74) vs the pooled cohort equations (0.72; 95% CI, 0.71-0.73): differences 0.01 or less (P values >.05) in the combined sample. Significant differences in discrimination were observed by race: the C indexes were 0.76 for all 3 models in White vs 0.69 in Black women (all P values <.001) and between 0.71 and 0.72 in White men and between 0.64 and 0.66 in Black men (all P values ≤.001). When stratified by age, model discrimination was better for younger (<60 years) vs older (≥60 years) adults for both Black and White individuals. The ratios of observed to expected 10-year stroke rates were closest to 1 for the REGARDS self-report model (1.05; 95% CI, 1.00-1.09) and indicated risk overestimation for Framingham stroke (0.86; 95% CI, 0.82-0.89) and pooled cohort equations (0.74; 95% CI, 0.71-0.77). Performance did not significantly improve when novel machine learning algorithms were applied. Conclusions and Relevance: In this analysis of Black and White individuals without stroke or transient ischemic attack among 4 US cohorts, existing stroke-specific risk prediction models and novel machine learning techniques did not significantly improve discriminative accuracy for new-onset stroke compared with the pooled cohort equations, and the REGARDS self-report model had the best calibration. All algorithms exhibited worse discrimination in Black individuals than in White individuals, indicating the need to expand the pool of risk factors and improve modeling techniques to address observed racial disparities and improve model performance.

Interplay of Atherogenic Particle Number and Particle Size and the Risk of Coronary Heart Disease.

BACKGROUND: We examined the interplay of apolipoprotein B (apoB) and LDL particle size, approximated by the LDL-cholesterol (LDL-C)/apoB ratio, on the risk of new-onset coronary heart disease (CHD). METHODS: Participants without cardiovascular disease from the UK Biobank (UKB; n = 308 182), the Women's Health Study (WHS; n = 26 204), and the Framingham Heart Study (FHS; n = 2839) were included. Multivariable Cox models were used to assess the relationship between apoB and LDL-C/apoB ratio and incidence of CHD (14 994 events). Our analyses were adjusted for age, sex (except WHS), HDL-cholesterol (HDL-C), systolic blood pressure, antihypertensive treatment, diabetes, and smoking. RESULTS: In all 3 studies, there was a strong positive correlation between apoB and LDL-C (correlation coefficients r = 0.80 or higher) and a weak inverse correlation of apoB with LDL-C/apoB ratio (-0.28 ≤ r ≤ -0.14). For all 3 cohorts, CHD risk was higher for higher levels of apoB. Upon multivariable adjustment, the association between apoB and new-onset CHD remained robust and statistically significant in all 3 cohorts with hazard ratios per 1 SD (95% CI): 1.24 (1.22-1.27), 1.33 (1.20-1.47), and 1.24 (1.09-1.42) for UKB, WHS, and FHS, respectively. However, the association between LDL-C/apoB and CHD was statistically significant only in the FHS cohort: 0.78 (0.64-0.94). CONCLUSIONS: Our analysis confirms that apoB is a strong risk factor for CHD. However, given the null association in 2 of the 3 studies, we cannot confirm that cholesterol-depleted LDL particles are substantially more atherogenic than cholesterol-replete particles. These results lend further support to routine measurement of apoB in clinical care.

Calibration and Uncertainty in Neural Time-to-Event Modeling.

Models for predicting the time of a future event are crucial for risk assessment, across a diverse range of applications. Existing time-to-event (survival) models have focused primarily on preserving pairwise ordering of estimated event times (i.e., relative risk). We propose neural time-to-event models that account for calibration and uncertainty while predicting accurate absolute event times. Specifically, an adversarial nonparametric model is introduced for estimating matched time-to-event distributions for probabilistically concentrated and accurate predictions. We also consider replacing the discriminator of the adversarial nonparametric model with a survival-function matching estimator that accounts for model calibration. The proposed estimator can be used as a means of estimating and comparing conditional survival distributions while accounting for the predictive uncertainty of probabilistic models. Extensive experiments show that the distribution matching methods outperform existing approaches in terms of both calibration and concentration of time-to-event distributions.

An adverse lipoprotein phenotype-hypertriglyceridaemic hyperapolipoprotein B-and the long-term risk of type 2 diabetes: a prospective, longitudinal, observational cohort study.

BACKGROUND: This study examines the risk of new-onset diabetes in patients with hypertriglyceridaemic hyperapolipoprotein B (high triglycerides, high apolipoprotein B [apoB], low LDL cholesterol to apoB ratio, and low HDL cholesterol). The aim was to establish whether this lipoprotein phenotype identified a substantial group at high risk of developing diabetes over the next 20 years. METHODS: In this prospective, longitudinal, observational cohort study, we used data from the Framingham Offspring cohort (recruited in Framingham, MA, USA). Participants were aged 40-69 years and free of diabetes and cardiovascular disease at a baseline examination done between April, 1987, and November, 1991, and were followed up until March, 2014. Cox proportional hazards regression with hierarchical adjustment for age and sex, waist circumference, and fasting blood glucose were used to model the relationship between each lipid marker and incident diabetes, as well as the relationship between hypertriglyceridaemic hyperapoB (defined as values greater than sample medians of triglycerides and apoB, and less than medians of HDL cholesterol and LDL cholesterol to apoB ratio) and incident diabetes. FINDINGS: Of 3446 individuals aged 40-69 years who completed baseline examination, 2515 participants were eligible and included in all analyses. During median 21·1 years (IQR 11·1-23·1) of follow-up, 402 (16·0%) individuals developed diabetes. Age (p=0·032), waist circumference (p<0·0001), fasting blood glucose (p<0·0001), and natural logarithm-transformed triglycerides (p<0·0001) were associated with new-onset diabetes, as were apoB (p=0·0016), LDL cholesterol to apoB ratio (p=0·0018), and HDL cholesterol (p=0·0016) when added to this model. The age and sex-adjusted incidence of diabetes in the hypertriglyceridaemic hyperapoB group was 32·4% (95% CI 27·8-37·7) versus 5·5% (3·5-8·6) in the optimal lipid phenotype group and 15·5% (13·5-17·7) in the mixed lipid phenotype group. The fully adjusted hazard ratio, including glucose and waist circumference, for individuals with hypertriglyceridaemic hyperapoB was 3·30 (95% CI 2·06-5·30; p=0·0008) and for mixed lipid phenotype was 2·17 (1·38-3·40; p<0·0001) compared with those with the optimal lipid phenotype. INTERPRETATION: Our findings suggest that individuals with hypertriglyceridaemic hyperapoB are at high risk of new-onset diabetes and might benefit from intensive measures to prevent diabetes. The association between this phenotype and incident diabetes is consistent with a pro-diabetic effect due to increased clearance of apoB particles from plasma, which could injure pancreatic islet cells. This mechanism might explain the increased risk of diabetes with statin therapy. FUNDING: Doggone Foundation.

Predictive Utility of a Validated Polygenic Risk Score for Long-Term Risk of Coronary Heart Disease in Young and Middle-Aged Adults.

BACKGROUND: Understanding the predictive utility of previously derived polygenic risk scores (PRSs) for long-term risk of coronary heart disease (CHD) and its additive value beyond traditional risk factors can inform prevention strategies. METHODS: Data from adults 20 to 59 years of age who were free of CHD from the FOS (Framingham Offspring Study) and the ARIC (Atherosclerosis Risk in Communities) study were analyzed. Because the PRS was derived from samples of predominantly European ancestry, individuals who self-reported White race were included. The sample was stratified by age and cohort: young (FOS, 20-39 years [median, 30 years] of age), early midlife (FOS, 40-59 years [median, 43] years of age), and late midlife (ARIC, 45-59 years [median, 52 years] of age). Two previously derived and validated prediction tools were applied: (1) a 30-year traditional risk factor score and (2) a genome-wide PRS comprising >6 million genetic variants. Hazard ratios for the association between each risk estimate and incident CHD were calculated. Predicted and observed rates of CHD were compared to assess discrimination for each model individually and together with the optimism-corrected C index (95% CI). RESULTS: Among 9757 participants, both the traditional risk factor score (hazard ratio per 1 SD, 2.60 [95% CI, 2.08-3.27], 2.09 [95% CI, 1.83-2.40], and 2.11 [95% CI, 1.96-2.28]) and the PRS (hazard ratio, 1.98 [95% CI, 1.70-2.30], 1.64 [95% CI, 1.47-1.84], and 1.22 [95% CI, 1.15-1.30]) were significantly associated with incident CHD in young, early midlife, and late midlife, respectively. Discrimination was similar or better for the traditional risk factor score (C index, 0.74 [95% CI, 0.70-0.78], 0.70 [95% CI, 0.67-0.72], and 0.72 [95% CI, 0.70-0.73]) compared with an age- and sex-adjusted PRS (0.73 [95% CI, 0.69-0.78], 0.66 [95% CI, 0.62-0.69], and 0.66 [95% CI, 0.64-0.67]) in young, early-midlife, and late-midlife participants, respectively. The ΔC index when PRS was added to the traditional risk factor score was 0.03 (95% CI, 0.001-0.05), 0.02 (95% CI, -0.002 to 0.037), and 0.002 (95% CI, -0.002 to 0.006) in young, early-midlife, and late-midlife participants, respectively. CONCLUSIONS: Despite a statistically significant association between PRS and 30-year risk of CHD, the C statistic improved only marginally with the addition of PRS to the traditional risk factor model among young adults and did not improve among midlife adults. PRS, an immutable factor that cannot be directly intervened on, has minimal clinical utility for long-term CHD prediction when added to a traditional risk factor model.