Search

Cannabis, or marijuana, has potential therapeutic and medicinal properties related to multiple compounds, particularly Δ-9-tetrahydrocannabinol and cannabidiol. Over the past 25 years, attitudes toward cannabis have evolved rapidly, with expanding legalization of medical and recreational use at the state level in the United States and recreational use nationally in Canada and Uruguay. As a result, the consumption of cannabis products is increasing considerably, particularly among youth. Our understanding of the safety and efficacy of cannabis has been limited by decades of worldwide illegality and continues to be limited in the United States by the ongoing classification of cannabis as a Schedule 1 controlled substance. These shifts in cannabis use require clinicians to understand conflicting laws, health implications, and therapeutic possibilities. Cannabis may have therapeutic benefits, but few are cardiovascular in nature. Conversely, many of the concerning health implications of cannabis include cardiovascular diseases, although they may be mediated by mechanisms of delivery. This statement critically reviews the use of medicinal and recreational cannabis from a clinical but also a policy and public health perspective by evaluating its safety and efficacy profile, particularly in relationship to cardiovascular health.

Introduction:Coronary artery calcification (CAC) score, physical activity (PA), and cardiorespiratory fitness (CRF) are all associated with cardiovascular disease (CVD) risk. While a U-shaped relationship between PA and CRF with CAC has been reported, CAC presence among highly fit individuals was suggested to be benign.Objective:To determine interactive associations of PA/CRF and CAC with outcomes in a cohort of middle-aged adults and to evaluate the relationship of PA/CRF with CAC incidence.Methods:CARDIA participants with CT-assessed CAC at year 20 (2005-06) were included (n=3141, mean age 45y, 57% female, 45% Black). Moderate to vigorous intensity PA (MVPA) was assessed by self-report and accelerometer. CRF was estimated with a treadmill test. Incident CVD events were adjudicated, and mortality data were obtained through 2019. CAC was reassessed at year 25 (2010-11). Cox models assessed hazard ratios (HRs) for CVD and mortality in groups defined by CAC and MVPA/CRF. Logistic models assessed associations with CAC incidence.Results:At baseline, more favorable CVD risk was found among participants with higher MVPA, higher CRF (> median sex-specific duration), and absence of CAC. During a mean follow-up of 13 years, 166 CVD events and 171 deaths occurred. After multivariable adjustment, compared with no CAC and higher CRF (ref), the HRs (95% CIs) for CVD were 5.04 (2.49-10.20) for CAC and higher CRF, 2.26 (1.24-4.12) for no CAC and lower CRF, and 4.27 (2.24-8.14) for CAC and lower CRF (Figure, left panel). The respective HRs for mortality were 1.12 (0.45-2.77), 1.54 (0.91-2.60), and 3.23 (1.82-5.72) (Figure, right panel). Similar findings were observed with self-reported or accelerometer MVPA replacing CRF. Higher CRF and accelerometer MVPA were dose-responsively associated with a lower probability of developing CAC in a 5-year follow-up (P < .01).Conclusions:PA and CRF were inversely associated with CAC incidence in middle-aged adults. CAC presence eliminated the cardiovascular risk advantage of high PA or high CRF.

Background:The association between variability in body mass index (BMI) in early adulthood and cardiac structure and function in midlife has not been previously examined.Methods:We examined 2371 Coronary Artery Risk Development in Young Adults (CARDIA) participants who had BMI assessments across 25-years (CARDIA exam year 0 [1985-1986], 2 [1987-1989], 5 [1990-1991], 7 [1992-1993], 10 [1995-1996], 15 [2000-2001], 20 [2005-2006], and 25 [2010-2011]) as well as echocardiography data at the year-25 exam (2010-2011). BMI variability was assessed by standard deviation (SD) across 25 years. Adjusted multivariable linear regression models were used to assess the association between echocardiography variables (dependent variable) and SD of BMI (independent variable). Model 1 was adjusted for standard cardiac risk factors (age, sex, race, education, blood pressure, anti-hypertension medication use, smoking, fasting plasma glucose, alcohol consumption, physical activity, HDL and total cholesterol. Model 2 was additionally adjusted for mean BMI.Results:Among participants included in the analysis, mean [SD] age at the year 25 exam [2010-2011] was 50.4 [3.6] years; 44.5% were men; and 41.3% were black). In model 1, greater SD of BMI was associated with greater left ventricular mass (β 5.18g, p<0.001), left ventricular global longitudinal strain 0.08, p=0.01, and left atrial volume (β 1.60ml, p<0.001). Additional adjustment for mean BMI, attenuated associations (p>.05 for all). Greater SD of BMI was associated with worse diastolic function (E/é) (β 0.11, p<0.001). Observed association between BMI variability and E/é persisted even after accounting for mean BMI (β 0.08, p=0.01).Conclusions:Greater body weight variability in young adulthood was associated with modest unfavorable midlife alterations in diastolic function.

Breast arterial calcification (BAC) detected in mammograms is an emerging risk-enhancing factor for atherosclerotic cardiovascular disease (ASCVD). The predictive value of coronary artery calcification (CAC) for risk of ASCVD above and beyond traditional risk factors is well-established in women. However, there are few studies that have simultaneously evaluated BAC and CAC. The Multiethnic study of breast arterial calcium gradation and cardiovascular disease (MINERVA) cohort, a large, racially and ethnically diverse cohort of postmenopausal women aged 60-69 (n=5,059) was recruited in 2012-15 at Kaiser Permanente of Northern California (KPNC). BAC was assessed using a densitometry method, and presence of BAC was defined as a calcium mass score>0 mg. Searching the electronic health record up to 5 years after baseline, we identified 33 women (39% non-white) who underwent cardiac computed tomography (CT) because of medical indication. BAC was present in 33% (11/33) whereas CAC (Agastston score >0) was present in 61% (20/33). Nine (27%) were BAC and CAC -; 7 (21%) were BAC and CAC+; 4 (12%) were BAC+ and CAC -; and 13 (39%) were BAC - and CAC+. The correlation between log (BAC+1) and log (CAC+1) was 0.45 (p=0.02) in the entire sample (n=33) and was 0.68 (p=0.09) in the 7 women who were BAC and CAC+. The concordance, sensitivity, specificity, PPV, NPV for BAC considering CAC as the gold standard were 0.48, 0.35, 0.69, 0.63 and 0.41, respectively. The unadjusted odds ratio for log (CAC+1)>0 as a function of Log (BAC+1) was 1.21 (95% CI, 0.27-5.38; p=0.80). Although our sample was small, the findings reflect real-world evidence and suggest that BAC and CAC are not entirely overlapping and thus may convey independent predictive information for ASCVD.

Introduction:Measures of atherogenic particle number (apoB and LDL particle number [LDL-P]) are stronger predictors of atherosclerotic cardiovascular disease (ASCVD) risk than measures of cholesterol concentration (LDL-C, non-HDL-C, total cholesterol [TC]) in middle-aged adults. It is unclear if this is true for younger adults.Methods:Among CARDIA participants (ppts), NMR was used to measure apoB and LDL-P. Non-HDL-C and TC were measured using standard assays; LDL-C was calculated using the Friedewald equation. We stratified the ppts into two age windows: age 20-30y (n=1645) and age 30-40y (n=2922). We used adjusted Cox proportional hazards models to assess the associations of 1SD higher apoB, LDL-P, non-HDL-C, LDL-C, or TC with incident ASCVD events. We substituted each measure of atherogenic lipid burden for TC in a modified Pooled Cohort Equation (PCE) model (with and without HDL-C); and model performance (discrimination and reclassification) was evaluated.Results:There were 81 and 163 ASCVD events over (median [IQR]) 31.8y (31.1-32.0y) for the age 20-30 age window and over 26.8y (19.1-27.1y) for the 30-40y age window, respectively. In ppts age 20-30y, a 1SD higher apoB, LDL-P, non-HDL-C, and LDL-C were significantly associated with incident ASCVD in demographic adjusted models. The strengths of associations with ASCVD were not significantly different across these measures. For the 30-40y age window, all measures of atherogenic lipoproteins were significantly associated with ASCVD; the strengths of association were not significantly different across atherogenic lipid measures in all models. There were no significant differences in the C-statistic and no improvement in reclassification when each measure was used to replace TC in the PCE model.Conclusions:ApoB, LDL-P, LDL-C or non-HDL-C may be slightly better markers of long-term ASCVD risk than TC in adults < 30y. However, in adults between 30-40y all measures of atherogenic lipid burden appeared to be equivalent predictors of long-term risk.

Introduction:Hypertension risk increases with higher age and body mass index (BMI). We previously found that the prevalence of diagnosed hypertension among Filipino adults is similar to Black adults and higher than White, Chinese, and South Asian adults aged 45-84y. Among young women, the difference between Filipina and Chinese adults was >2-fold. We now examine disparities in hypertension prevalence across the adult lifespan, accounting for differences in weight status.Methods:This cross-sectional study used electronic health record data for 1,018,159 non-Hispanic White (White), 146,517 Black, 348,715 Hispanic, 128,124 Filipino, 126,321 Chinese, and 80,723 South Asian adults aged 30-79 who were members of a Northern California health plan in 2016. Hypertension prevalence (ICD 9/10 diagnosis in 2016) was examined by race/ethnicity and sex for ages 30-39, 40-49, 50-59, 60-69, 70-79y. Poisson regression was used to estimate prevalence ratios (aPRs) for hypertension compared to White adults in each age decade, adjusting for age, English language, current smoking, and BMI category (underweight, healthy weight, overweight, obesity class 1, 2, 3). Lower BMI thresholds were used for Asian adults.Results:Across all age decades, Filipino adults had a prevalence of hypertension similar to Black adults and higher than White, Chinese, and South Asian adults. In each age decade, the Filipino:White aPRs approached or exceeded the Black:White aPRs and were much higher than South Asian:White and Chinese:White aPRs (Figure). Differences in hypertension risk were greatest for those aged 30-59y.Conclusions:Filipino adults, similar to Black adults, had earlier onset and higher prevalence of hypertension than White, Chinese, and South Asian adults across 5 age decades, a disparity that persisted after adjusting for weight status, smoking, and language. These results support the need for earlier and more aggressive blood pressure screening in high-risk populations.

Introduction:We previously reported that Chinese, Filipino, and South Asian adults have higher prevalence of prediabetes (PreDM) than non-Hispanic White (White) adults, with nearly twice the burden in healthy BMI ranges. This study further examines the differential burden of diabetes (DM) in these Asian subgroups using Asian specific BMI categories.Methods:283,110 White, 33,263 Chinese, 38,766 Filipino, and 17,959 South Asian adults aged 45-64y in a 2016 Northern California health plan with height and weight were examined. PreDM and DM were classified based on lab data, diagnoses, or DM treatment. Modified log Poisson regression was used to determine Asian group vs. White prevalence ratios (PRs) for DM and preDM, adjusted for age and BMI within healthy weight, overweight, and obesity ranges (BMI 18.5 to <25, 23 to <30, ≥30 kg/m2 for Whites; BMI 18.5 to <23, 23 to <27.5, ≥27.5 kg/m2 for Asians).Results:Across all BMI categories, DM prevalence was higher for the Asian groups than White adults. DM prevalence was also higher for Filipino and South Asian vs White than Chinese vs White, especially at healthy BMI levels. PRs for South Asian men/women at healthy BMI were 1.8/2.8 for preDM and 5.9/8.0 for DM (Figure). PRs for Filipino men/women at healthy BMI were 1.8/2.6 for preDM and 5.0/7.5 for DM, respectively. For Chinese men/women at healthy BMI, the PRs for preDM were similar (2.1/2.9) to the other Asian groups, but the PRs for DM were lower (2.1/3.4).Conclusion:The largest difference in DM prevalence relative to White adults was observed in the healthy BMI range for South Asian and Filipino adults. While PreDM risk in these 2 groups was similar to the preDM risk in Chinese compared with White adults at healthy BMI, their DM risk was 5-8 times greater than White adults at healthy BMI. Our data emphasize the disproportionate metabolic risk among middle-aged adults of different Asian groups and underscore the need for DM screening among high-risk Asian subgroups at healthy BMI levels.

Introduction:Obesity represents a major health and economic burden worldwide and is a key target for cardiovascular disease prevention. With emerging therapeutic options to manage obesity, we aimed to quantify the overall expenditures and resource utilization in adults with and without atherosclerotic cardiovascular disease (ASCVD) across obesity categories.Methods:We used the 2018 Medical Expenditure Panel Survey (MEPS), a nationally representative US survey. BMI (calculated from self-reported height and weight) was divided into 4 groups: overweight (BMI 25 to 30), mild (BMI 30 to 35), moderate (BMI 35 to 40), and morbid obesity (BMI ≥40). Cardiovascular risk factors (hypertension, diabetes, hypercholesterolemia, insufficient physical activity) and ASCVD were ascertained by ICD10 codes and/or self-report. Two-part econometric models were used to study cost data; a generalized linear model with gamma distribution and link log was used to assess expenditures.Results:Among the 19,569 sampled MEPS adults (mean age 48 years [SD 14]), 6.2% had morbid obesity (~ 14 million). Among those without ASCVD, morbid obesity was associated with higher odds for prescription medication utilization, ED and outpatient visits when compared to those with overweight (Figure). Even though individuals with ASCVD reported higher prevalence of all metrics of resource utilization, we found no statistically significant differences between BMI categories in this group. The marginal overall per capita expenditure when comparing those with morbid obesity vs overweight was $2,165 (p=0.11) in those with ASCVD, and $1,605 (p=0.03) in those without ASCVD.Conclusions:Morbid obesity was associated with increased healthcare expenditures and resource utilization, especially in those without ASCVD. Our study findings could aid in policy discussions to discern appropriate candidates for emerging novel anti-obesity therapeutics, as well as apt resource allocation.

Introduction:Black adults experience higher rates of cardiovascular disease (CVD) compared with White adults, particularly across young adulthood. To identify intervention targets to reduce disparities, we quantified the contribution of individual- and neighborhood-level determinants across young adulthood to Black-White differences in incidence of premature CVD.Methods:In Black and White adults (baseline age 18-30 years) in the Coronary Artery Risk Development in Young Adults (CARDIA) Study, the associations of clinical, lifestyle, psychosocial, socioeconomic, and neighborhood determinants with racial disparities in incident CVD were evaluated with sex-stratified Cox proportional hazards models adjusted for time-updated risk factors. Percent reduction in the ß estimate (log-hazard ratio [HR]) for race quantified the contribution of each risk factor group, individually and combined, to incident CVD racial disparities.Results:In 1707 Black and 2001 White young adults followed for median 33.9 (IQR 33.7-34.0) years, Black adults had higher risk of incident CVD than White adults (Figure). The fully adjusted model yielded significant attenuation of the disparity: HR 1.22 (0.68-2.21) for Black vs. White women, HR 1.42 (0.91-2.23) for Black vs. White men. In women, clinical risk factors were associated with the largest percent reduction in the ß estimate (83%) for race. In men, socioeconomic, lifestyle and clinical risk factors were each associated with significant percent reductions in ß estimates (45%, 37%, and 28%, respectively).Conclusions:In CARDIA, risk for premature CVD was no longer significantly higher in Black vs. White adults after adjustment for upstream determinants across multiple domains. The largest contribution to racial disparities was from clinical factors in women and socioeconomic factors in men. These findings may inform earlier-life interventions to reduce disparities in premature CVD.

Introduction:The association between body mass index (BMI) and mortality risk has been previously shown to follow a J- shaped pattern, with a greater mortality risk at the lowest and highest BMI levels. However, this association has not been fully evaluated in a population with atherosclerotic cardiovascular disease (ASCVD). We examined the BMI-mortality associations in the general population, and in a population with ASCVD.Methods:We used data from 2006-2014 National Death Index-linked National Health Interview Survey, a US nationally representative survey, for adults ≥ 18 years. Participants were classified based on self-reported ASCVD. BMI was calculated based on self-reported height and weight and was classified as normal/overweight (BMI 18.5-29.9 kg/m2), obesity class 1 (30-34.9 kg/m2), class 2 (35-39.9 kg/m2), and class 3 (≥40 kg/m2); participants who were underweight (<18.5 kg/m2) were excluded. Multivariable cox proportional hazards models were used to examine the risk of all-cause, cardiovascular, and non-cardiovascular mortality.Results:A total of 252,9568 adults, including a weighted 8.1% (or 18.6 million annually) with ASCVD, were included in the analysis. During a median follow-up of 4.5 (IQR: 2.5-6.8) years, (1179 million person-years), mortality rates, both cardiovascular and non-cardiovascular, were higher in the ASCVD group compared to the general population (Fig: panel A). In the general population, those with obesity class 3 had a higher risk of cardiovascular and non-cardiovascular mortality compared to individuals with normal/overweight, whereas individuals with obesity class 1 had a lower non-cardiovascular mortality risk. An overall similar pattern of association was observed in the ASCVD population in non-cardiovascular, but not cardiovascular, mortality (Fig: panel B).Conclusion:Greater BMI categories were associated with higher non-cardiovascular, but not cardiovascular, mortality in the ASCVD population.

Introduction:Adverse pregnancy outcomes (APOs) are emerging as critical risk markers for the development of cardiovascular disease (CVD). Risk of APOs are higher among women with worse prepregnancy cardiovascular health (CVH), but it remains unclear if APOs mediate the relationship between prepregnancy CVH and incident coronary artery calcium (CAC).Methods:We included individuals in the Coronary Artery Risk Development in Young Adults study with at least one singleton birth prior to 2001 and data on prepregnancy CVH (1985-1996) and CAC (2000-2011). Data on APOs were self-reported at each visit (gestational diabetes, hypertensive disorders of pregnancy, preterm birth and small for gestational age). We quantified prepregnancy CVH by summing across the American Heart Association’s Life’s Simple 7 (score 0-14). We used Cox proportional hazards models to obtain hazard ratios (HRs) for a one-point higher (better) prepregnancy CVH score for incident CAC (CAC > 0) adjusted for age, self-reported race, education (years) and parity. We used causal mediation analyses to estimate the proportion of the association between prepregnancy CVH and incident CAC mediated through any APO and each APO subtype.Results:Of the 906 included participants, 46.1% were Black, mean age ± SD was 28.1 ± 4.5 years at baseline and 43.0 ± 4.8 years at follow-up. Among individuals with low (0-7), intermediate (8-11) and high (12-14) prepregnancy CVH, APOs developed in 41.8%, 29.5% and 19.5% respectively, and CAC developed in 38.8%, 14.8% and 11.6% of women, respectively. Adjusted HR (95% CI) for one-point higher (better) prepregnancy CVH on incident CAC was 0.79 (0.72, 0.87). APOs did not significantly mediate the relationship between prepregnancy CVH and CAC (TABLE).Conclusions:Higher or more favorable prepregnancy CVH is associated with lower risk of subclinical CVD in midlife, regardless of APO status, and highlights the importance optimizing CVH early in the life course prior to pregnancy.

Background:Nonalcoholic fatty liver disease (NAFLD) is an obesity-related condition associated with cardiovascular disease (CVD) and its major risk factors. The latter have been linked to accelerated cognitive decline. Whether NAFLD is associated with decreased cognitive performance in midlife remains uncertain.Methods:Participants from the Coronary Artery Risk Development in Young Adults (CARDIA) study with CT-measured NAFLD and cognitive function assessment at year 25 (2010-2011; n=2,835) were included. Cognitive function was reassessed at year 30 (2015-2016; n=2,388). NAFLD was defined according to liver attenuation and categorized into 3 groups: none [>51 Hounsfield Units (HU)], probable (>40-51 HU), and definite (≤40 HU). Cognitive tests, including the Digit Symbol Substitution Test (DSST), Rey Auditory Verbal Learning Test (RAVLT), and Stroop were analyzed with standardized z scores. Linear models were constructed to examine (a) the cross-sectional association between NAFLD and cognitive scores (year 25) and (b) the predictive role of NAFLD in subsequent (year 30) cognitive scores.Results:The mean age at baseline was 50.1 (SD, 3.6) years (58% women; 48% blacks), with 398 (14%) and 286 (10%) having probable and definite NAFLD, respectively. In unadjusted models, an inverse association was shown between NAFLD and all cognitive tests. For DSST (Pinteraction=.047) and RAVLT (Pinteraction=.033), the association was stronger in whites than blacks. The association was attenuated after age and sex adjustment, and practically eliminated after further adjustment for sociodemographic and CVD risk factors (Table). NAFLD was not predictive of subsequent cognitive performance in any of the measures once baseline scores were adjusted for (all P>.10). Using liver attenuation as a continuous variable yielded similar results.Conclusion:Among middle-aged adults, an inverse association of NAFLD with cognitive function -evident in whites only- was attenuated by sociodemographic and CVD risk factor adjustment.

Introduction:Alcohol use is common in the United States. Most studies describe a J-shaped relationship between alcohol consumption and cognitive function: frequent heavy consumption alters brain functions and decreases cognitive performance; regular light and moderate consumption may have protective impact. We are unaware of any reports utilizing longitudinally collected data to derive a cumulative exposure of alcohol use and assess its association with cognitive function.Hypothesis:Cumulative 30-year exposure to alcohol has a J-shaped association with mid-life cognitive function.Methods:We included 3,068 participants aged 18-30 years at the baseline examination (1985-86) and having up to 8 follow-up exams at 2- to 5-year intervals including at 30 years (2015-16). Cumulative “drink-years” exposure to alcohol was estimated at each exam by multiplying the usual number of drinks/day reported by the number of years since last exam, and then summing across exams. We used multivariable adjusted linear regression models to assess the independent associations of number of drink-years with 6 measures of cognitive function studied at Year 30: Rey Auditory Verbal Learning Test (RAVLT), Digit Symbol Substitution Test (DSST), Stroop Test, Montreal Cognitive Assessment (MoCA), Category Fluency Test, and Letter Fluency Test. We used inverse probability of censoring weighting to account for potential informative censoring. Linear regression models estimated standardized mean test scores for the cumulative alcohol use categories (drink-years: >0 and <15, 15-29.9, 30+). Too few participants were available for assessment of higher categories of use.Results:Compared with never drinking, the 3 categories of cumulative exposure were unassociated with any of the cognitive function measures.Conclusions:Cumulative 30-year alcohol use is not associated with cognitive function in middle-aged adults in the CARDIA study.

Introduction:Masked hypertension is defined as having hypertensive blood pressure (BP) outside of the office setting among adults with non-hypertensive BP when measured in the office. Some guidelines recommend defining out-of-office BP using awake measurements while other guidelines recommend using awake and asleep measurements.Hypothesis:We hypothesized that defining masked hypertension using the awake and asleep BP measurements would increase the prevalence of masked hypertension compared to using the awake period alone, and the magnitude of this difference would be greater among non-Hispanic blacks compared with non-Hispanic whites and Hispanics.Methods:We pooled previously collected data from 5 NHLBI-funded population- and community-based studies including the Jackson Heart Study, the Coronary Artery Risk Development in Young Adults Study (total participants: 2,866). All participants had office systolic BP (SBP)<140mmHg and diastolic BP (DBP)<90mmHg and underwent ambulatory BP monitoring (ABPM) for 24 hours. Hypertensive awake BP was defined as SBP ≥135mmHg or DBP ≥85mmHg while awake, hypertensive asleep BP as SBP ≥120mmHg or DBP ≥70mmHg while asleep and hypertensive 24-hour BP as SBP ≥130mmHg or DBP ≥80mmHg over the entire ABPM period.Results:The prevalence of masked hypertension increased from 29% to 43% when defined using awake, asleep, or 24-hour BP versus using awake BP alone (Table). This increase was larger in non-Hispanic blacks (31-54%) compared with non-Hispanic whites (28-37%) and Hispanics (17-26%). The adjusted prevalence ratio (95% confidence interval) for having masked hypertension for non-Hispanic blacks compared with Non-Hispanic whites was higher from 1.20(1.05,1.37) to 1.33(1.20,1.47) when defined using awake, asleep and 24-hour BP versus awake BP only.Conclusions:Including asleep BP to define masked hypertension increased the prevalence of masked hypertension to a larger extent among non-Hispanic blacks compared to non-Hispanic whites and Hispanics.

Introduction:Heart failure (HF)-related hospitalizations are a growing public health burden. We evaluated two published risk calculators for predicting 30-day readmission after HF hospitalizations: 1) using the original coefficients, 2) updating the coefficients 3) developing a new model with additional variables and updated coefficients.Hypothesis:Recalibrating model coefficients and adding variables would improve the performance of existing 30-day readmission risk calculators.Methods:We identified 45,059 adults hospitalized for HF between 2012-2017 within Kaiser Permanente Northern California, an integrated healthcare delivery system. We used split sampling for development and validation testing. The risk calculators tested included: LACE+ Index and Yale CORE. We used logistic regression on our population to derive the recalibrated coefficients. For the model with additional variables, we included all variables used in the original models plus additional variables, including cardiovascular medication use and socioeconomic status. We used gradient boosting with k-fold cross validation to avoid overfitting. We assessed model performance using area under the curve (AUC) and calibration plots.Results:Discrimination (AUC) was poor using original models: LACE+ [0.56 (0.54-0.58)] and Yale CORE [0.55 (0.54-0.57)]. Recalibrating coefficients resulted in small improvements for LACE+ [0.58 (0.57, 0.60)] and Yale CORE [0.58 (0.57, 0.60)]. Adding variables resulted in a modest improvement for the gradient boosting model [0.61 (0.59, 0.62)]. Calibration plots (Figure 1) showed good calibration except for the Yale CORE model with the original coefficients.Conclusions:Recalibrating coefficients and incorporating prior medication and socioeconomic status led to modest, significant improvements in discrimination while maintaining good calibration. However, overall performance improvements are needed to increase the utility of these published risk calculators to predict readmission.

Introduction:Prevalent diabetes mellitus (DM) in mid-life is a stronger risk factor (RF) for cardiovascular disease (CVD) mortality in women than men. However, it is unclear whether (1) changes in CVDRFs before and after incident DM differ by sex, and (2) if such differences are associated with the relationship of incident DM with subsequent CVD events.Hypothesis:Rates of change in CVDRFs after incident DM will be worse in women and incident DM will have a stronger association with CVD events in women than men.Methods:We included 4893 black and white CARDIA participants, ages 18-30 years at enrollment (1985-86). We ascertained incident DM (n=827) and assessed sex differences in annual change in body mass index (BMI); systolic and diastolic blood pressure (SBP/DBP); total, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) cholesterol before and after the ascertainment of DM using piecewise regression, stratified by race. We estimated hazard ratios (HR) for incident CVD over a median of 30.9 years of follow-up (range: 2, 32) and assessed effect modification between sex and DM. All analyses included adjustment for time-varying behavioral and clinical CVDRFs.Results:Age at DM ascertainment was lower for women than men (42.0 vs 45.8 years, sex-difference: 3.8, 95% confidence limits: 2.5, 5.0). Rates of change in CVDRFs before the ascertainment of DM did not differ by sex (Table). However, compared to men, women had greater annual changes in SBP, DBP, total cholesterol, and LDL, but not BMI and HDL, after DM. Median age at CVD did not differ by sex (48.8 years). Incident DM was associated with higher HR for incident overall CVD (HR: 1.45, 95%: 1.07, 1.96) and this association did not differ by sex (p interaction=0.8). Women had lower HR for incident CVD than men (HR: 0.57; 95%: 0.44, 0.74).Conclusion:While CVDRFs worsened more rapidly after DM for women, women had lower absolute and relative risks for incident CVD in mid-life than men. The association of incident DM by mid-life with subsequent CVD events did not differ by sex.

Introduction:In 2011, the number of deaths with an underlying cause of heart disease (HD) reached its lowest level in 56 years. However, there has subsequently been a steady increase in the annual number of total HD deaths, owing to a rapid increase (23%) in the size of the population ≥65 years of age in the U.S. To understand these trends, we sought to characterize differences between 2005-2011 and 2011-2017 in population and mortality by age subgroups among those ≥65 years.Methods:We determined age-specific population size, HD mortality rate, and absolute number of HD deaths in the ≥65 age group for the time periods 2005-2011 and 2011-2017, as well as for the <65 years age group, using the CDC WONDER online data set.Results:Age-specific population growth was greatest among those 65-74 years between 2005-2011 and 2011-2017, representing 76% of the total population growth among those ≥65 years old in both time periods (Table) and 51% of the growth for the entire population from 2011-2017. From 2005-2011, decreases in the mortality rate of 20% or greater in each of the age subgroups (65-74, 75-84, 85+) resulted in a decline in the number of total HD deaths in each of the age groups in spite of substantial population increases in the 65-74 and 85+ age groups. However, subsequent changes in the age-specific mortality rate among those ≥65 years were lower than population increases from 2011-2017, resulting in an increased number of total HD deaths in all >65 years age subgroups. This was most notable among those age 65-74 years in whom the 32% population increase with a 1.5% decline in the mortality rate resulted in a 30% increase in the number of HD deaths, representing 61.3% of the increase in number of HD deaths in the 65+ years age group and 53% of the increase of HD deaths in all age groups from 2011-2017.Conclusions:The rapid growth of the 65-74 years age group (baby boomers) accounts for most of the population growth in the ≥65 years age group and more than half the increase in both total population growth and the total number of HD deaths from 2011-2017.

Background:The prognostic significance of temporal changes in resting heart rate (RHR) in young adults for premature heart failure (HF) and cardiovascular disease (CVD) is not well-studied.Methods:We examined 4343 Coronary Artery Risk Development in Young Adults (CARDIA) Study participants aged 23-35 years(y) at Y5 (1990-1991) who had a contemporaneous assessment of RHR, echocardiograms, and long-term follow-up for outcomes. Cox proportional hazard models using single-occasion RHR (Y5) and temporal change in RHR (Y0 to Y5) as explanatory variables assessed associations with incident HF and CVD. Multivariable linear regression models were used to assess relations with CVD risk factors and cardiac structure and function.Results:Mean age was 29.9?3.6 years (Y5), 49% were Men, and 45% were African-American. There were 268 CVD events observed over a median follow-up of 26 years and 107991 person-years at risk. In Cox models, there was a near linear relation between temporal change in RHR and clinical outcomes. For each 1SD increment, longitudinal increases in RHR were associated with a higher risk of HF (HR=1.39 95% CI [1.09-1.78]) and CVD (HR=1.21 95% CI [1.06-1.37]). Temporal decreases in RHR were also associated with a lower risk of both HF and CVD. Single-occasion RHR was associated with CVD, but not with HF. Greater longitudinal increases in RHR were associated with higher alcohol consumption (?=0.03, p<0.001), lower physical activity (?=0.002, p=001), current smoking (? =1.58, p<0.001), Men (p<0.001), African-Americans (p<0.001), impaired left ventricular (LV) relaxation (e??= -0.13, p=0.002), and worse LV diastolic function (E/E? ? =0.1, p=0.01).Conclusions:Temporal changes in RHR in young adults were associated with incident HF and CVD in middle age and provided additional information to a single occasion RHR. Contributory factors were relations between secular changes in RHR and early suboptimal CVD risk factors and subsequent cardiac dysfunction.

Introduction:The rate of atherosclerotic cardiovascular disease (ASCVD) events is known to vary substantially across ethnic subgroups, particularly among the Asian population. We previously showed that the ACC/AHA Pooled Cohort ASCVD Risk Estimator (ACC/AHA PCE) overestimates the actual ASCVD risk among Asians within a large, community-based population. We recalibrated the ACC/AHA PCE for Chinese, South Asian, and Filipino subgroups receiving care within a large, integrated healthcare delivery system.Methods:Within Kaiser Permanente Northern California, we identified self-reported Chinese (N=17,567), South Asian (N=4,564), and Filipino (N=13,402) members aged 40-79 years in 2008 who had LDL-C 70 to 189 mg/dL, no prior known ASCVD and no lipid-lowering therapy before entry or during follow-up. ASCVD events through 2015 were ascertained from comprehensive electronic health records and death certificates. Using 70/30 split-sample derivation and validation, we developed models using the same variables used for the ACC/AHA PCE (age, treated and untreated systolic blood pressure, HDL-C, total cholesterol, diabetes and smoking), and recalibrated the equation for each ethnicity and gender subgroup to predict the 7-year risk of ASCVD.Results:In validation cohorts, the ethnicity-specific equations displayed excellent calibration compared to the ACC/AHA PCE, especially in Chinese and Filipino subgroups (Figure 1). The Asian ethnicity-specific equations also had better discrimination compared with ACC/AHA PCE, with improvements in AUC in Chinese (0.81 to 0.84) and South Asian (0.68 to 0.74) patients, with similar AUC in Filipino patients (0.79 for both).Conclusions:In a large, community-based Asian cohort, we developed and internally validated ethnicity-specific equations that were more accurate than the ACC/AHA PCE. These new equations can help tailor primary prevention shared decision-making for Chinese, South Asian and Filipino patients.

Introduction:The degree to which healthy eating, physical ativity and not smoking are protective for coronary heart disease (CHD) by levels of genetic background remains unsettled.Hypothesis:Favorable lifestyle will blunt the effect of high genetic load and favorable genetic background will offset a detrimental lifestyle.Methods:We utilized the Genetic Epidemiology Resource in Adult Health and Aging (GERA) Caucasian cohort of 51,892 (mean age 59; 58% female) Kaiser Permanente of Northern California (KPNC) members and categorized participants at baseline in 2003-07 by smoking status (never, former, current), by Mediterranean-pattern diet (MedDiet) and meeting or not current AHA physical activity recommendations (PAR). We stratified the cohort into three groups of CHD genetic load (low: quintile 1, intermediate: quintiles 2, 3 and 4 combined and high: quintile 5) using a validated 51-SNP weighted polygenic risk score. Incident CHD was primary inpatient codes for angina pectoris, myocardial infarction, revascularization procedures or CHD death (n=2,606) through 12/31/2014; mean follow-up=9.5 years.Results:Age-adjusted CHD rates per 10,000 person years were estimated by Poisson regression in each level of the behavioral factors and, within each level, in the three genetic load groups ( Figure). Genetic background provided further CHD risk stratification within each lifestyle factor considered individually. Favorable lifestyle (never smoking, MedDiet, meeting PAR) blunted the effect of high genetic load by 42% (age-adjusted rates per 10,000 person-years 63 vs. 109). In turn, low genetic load offset the combination of the three detrimental lifestyles by 37% (age-adjusted rates per 10,000 person-years 69 vs 109).Conclusions:Our results are consistent with an additive model of genetic load and lifestyle as determinants of CHD. Lifestyle appears to compensate more for unfavorable genetic background than low genetic load for unfavorable lifestyle.