Bringing doctors together

or

A free, private network for verified physicians

Leading voices

Dr. Nishant Shah

Preventive Cardiology · Duke Hospital

View

Dr. Amit Khera

Preventive Cardiology · UT Southwestern

View

Dr. Martha Gulati

Preventive Cardiology · Houston Methodist

View

Dr. Pradeep Natarajan

Preventive Cardiology · MGH

View

Dr. Christopher Cannon

Preventive Cardiology · BWH

View

Dr. Nicholas Marston

Preventive Cardiology · BWH

View

Dr. Dorairaj Prabhakaran

General Cardiology

View

Dr. Pugazhendhi Vijayaraman

EP · Geisinger

View

Dr. Robert Schaller

EP · HUP

View

Dr. Trey Woods

General Cardiology · Emory Healthcare

View
  • Discover your community

  • Stay current

  • A profile that represents you

  • Find your people

About RoonPrivacyTerms

    Overview

    Circulating Ketone Bodies and Incident Cardiovascular Outcomes and Mortality: Insights From the UK Biobank.

    Feb 20, 2026DOI: 10.1161/jaha.125.042582Journal ArticleAHA

    Written by: Parag Anilkumar ChevliSaeid MirzaiRichard KazibweJeff KingsleyAlexis C WoodJoseph YeboahLeandro SlipczukAnurag MehtaHarpreet S BhatiaAmbarish PandeyMichael D Shapiro

    Find these authors on Roon

    AM
    Dr. Anurag Mehta
    Preventive Cardiology · Emory University

    Methods

    Access to UKB (UK Biobank) resources is available upon reasonable request from the authors and can also be obtained through formal application via the UKB website (https://www.ukbiobank.ac.uk/).The UKB study received ethical approval from the North West Multicenter Research Ethics Committee (Reference: 11/NW/0382), as detailed at https://www.ukbiobank.ac.uk/learn‐more‐about‐uk‐biobank/about‐us/ethics. All participants provided written informed consent. The study was conducted in accordance with the principles of the Declaration of Helsinki.The UKB comprises a prospective, observational, population‐based cohort of >500 000 adult residents in the United Kingdom, aged between 37 and 73 during recruitment from 2006 to 2010. Details of the UKB study, including the rationale, study design, and protocol, have been previously described.11 This analysis involved the UKB subjects with measured KB at enrollment, complete covariate data, and no preexisting atherosclerotic CVD (ASCVD) or HF at baseline (Figure 1). Data analysis was performed under UKB application no. 97439.ASCVD indicates atherosclerotic cardiovascular disease; HDL, high‐density lipoprotein; and HF, heart failure.The primary study exposure was total KB. In epidemiological analyses, plasma levels of the 3 KB are conventionally summed as plasma total KB. EDTA plasma samples of ∼118 000 randomly selected participants that were collected at baseline were used to measure β‐OHB, acetoacetate, and acetone using the Nightingale Health nuclear magnetic resonance platform.12 EDTA plasma samples from aliquot 3 were prepared in 96‐well plates by the UKB Laboratory (Stockport, UK). The plasma samples were shipped to Nightingale Health laboratories in Finland in 96‐well plates on dry ice in batches of 5000 to 20 000 samples. The Nightingale Health laboratories promptly froze samples at −80 °C after their arrival from the UKB laboratory. Before preparation, frozen samples were thawed gradually at +4 °C overnight, mixed gently, and centrifuged at 3400 Η g for 3 minutes to eliminate any potential precipitate. These samples are a random subset of the entire cohort, as no selection criteria were implemented during the sampling process.12, 13, 14A comprehensive battery of baseline assessments was conducted on the participants, which comprised touchscreen questionnaires encompassing sociodemographic information, family history, lifestyle, and medical history. The participants’ reported medical histories were validated by trained study nurses. During the physical assessments, including vital signs and anthropometric data measurements, participants also provided blood and urine samples for laboratory analysis.
    Participants self‐reported their race and ethnicity, and we categorized them as Asian, Black, White, and Mixed/Other. The Townsend area deprivation index was acquired based on the residential postcode and is calculated using combined data on unemployment, automobile and home ownership, and household overcrowding. At the time of enrollment, participants reported their history of tobacco use (current, former, or never), current intake of alcohol, and medication use. Blood pressure, height, and weight were recorded by trained study personnel. Body mass index (BMI) was calculated using weight (in kilograms) divided by height (in meters) squared. Physical activity assessment was conducted using modified questions derived from the validated short International Physical Activity Questionnaire.15 These questions encompassed the frequency, intensity, and duration of walking and vigorous activities. Time spent in these activities was used to calculate metabolic equivalent of task min/week of physical activity. The Oxford WebQ, a web‐based 24‐hour recall questionnaire designed for large population studies, was used to gather detailed dietary information.16 The questionnaire can be accessed at www.ceu.ox.ac.uk/research/oxford‐webq. Hypertension and type 2 diabetes were defined using the relevant International Classification of Diseases, Tenth Revision (ICD‐10) codes. The methods for obtaining and analyzing baseline blood and urine samples have been described in detail.17 Biochemistry measures were performed at a central laboratory between 2014 and 2017, including serum total cholesterol, high‐density lipoprotein‐cholesterol, and triglycerides. Serum creatinine was measured on a Beckman Coulter AU5800 instrument using an enzymatic isotope dilution mass spectrometry‐traceable method.
    The primary outcomes were ASCVD, HF, and all‐cause death. Secondary outcomes were myocardial infarction (MI), ischemic stroke, and CVD death. ASCVD was defined based on ICD, Ninth Revision (ICD‐9) and ICD‐10 codes indicating a diagnosis of coronary artery disease (MI and its acute complications, coronary artery bypass graft surgery, or percutaneous angioplasty/stent placement), ischemic stroke (cerebral infarction due to thrombosis, cerebral atherosclerosis, or cerebrovascular syndromes), peripheral artery disease, and CVD death. Affected individuals were identified using self‐reported data confirmed by a trained health care professionals, hospitalization records, and national procedural and death registries, as previously described.18 The diagnosis date was determined as the earliest day the diagnosis was confirmed. Participants were followed prospectively through linkage to national hospital admission and mortality registries, resulting in minimal loss to follow‐up. Participants were followed from enrollment in the study until they experienced an incident event, death, or the end of available follow‐up. Participants with prevalent ASCVD and HF were excluded from the analysis. Additional details are provided in the Table S1–S5.
    The study’s participants’ baseline characteristics were reported across total KB tertiles as counts (percentage) for categorical variables and mean±SD or median (p25–p75, if skewed distribution) for continuous variables. The participant characteristics were compared across the KB tertiles using ANOVA for normally distributed data, the Kruskal–Wallis test for skewed distributed data, and the chi‐square test for categorical variables. Absolute incidence rates for study outcomes were calculated per 1000 person‐years of follow‐up.
    Total KB, β‐OHB, acetone, and acetoacetate were log‐transformed with base 10 to achieve an approximately normal distribution and enabling the interpretation of data concerning a 10‐fold increase in both total and individual KB. The cumulative incidence function was employed to estimate the incidence rates of various outcomes over time, accounting for competing risks. Non‐CVD death served as the competing risk event in assessing the association of total KB with ASCVD and CVD death. All‐cause death was considered a competing risk when analyzing associations with other outcomes. Additionally, we used a restricted cubic spline model adjusted for variables in model 3 to depict the nonlinear relationship between baseline log‐total KB (continuous) and each CVD outcome. A linearity test was also conducted to examine the relationship between total KB and each CVD outcome. The association between total KB and cardiovascular events was assessed using Cox proportional hazards regression models with following adjustments: Model 1: unadjusted; Model 2: adjusted for age, sex, race, ethnicity, and Townsend deprivation index; Model 3: Model 2 + BMI, smoking, systolic blood pressure, diastolic blood pressure, diabetes, total cholesterol, high‐density lipoprotein‐cholesterol, triglycerides, hemoglobin A1c, statin use, creatinine, and alcohol consumption. As a sensitivity analysis, we also examined the association between individual KB and different outcomes using the aforementioned approach. The proportional hazards assumption was tested using Schoenfeld residuals and the Kolmogorov‐type Supremum test; no violations were detected. In secondary analyses, we used Fine–Gray subdistribution hazard models to account for noncardiovascular death as a competing event when assessing the risk of incident ASCVD, HF, MI, and stroke. Results are reported as subdistribution hazard ratios (HRs) with 95% CIs.
    Due to a considerable amount of missing data regarding physical activity (N=21 584) and dietary intake (N=100 116), we opted not to include them as confounding variables in our primary analyses. However, we integrated them as covariates in sensitivity analyses. Missing values of physical activity and dietary intake in the data set were handled using the multiple imputation method. We generated 20 imputed data sets to account for the uncertainty introduced by the imputation process. For each imputed data set, Cox proportional hazards regression analysis was performed. Subsequently, the parameter estimates obtained from each imputed data set were combined. This step accounted for variability within and between imputed data sets to obtain robust estimates and SEs.19
    Finally, to evaluate how sex, diabetes, BMI, and hypertension may modify the relationship between circulating KB and cardiovascular outcomes, we investigated interactions with these variables, considering a significance threshold of P<0.10. In all other analyses, a 2‐sided P value of <0.05 was considered statistically significant, and all statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).

    Results

    The baseline characteristics of the study participants are presented as tertiles of total KB in Table 1. A total of 90 987 participants were included in the study after applying exclusion criteria (Figure S1–S5). The mean±SD age at the baseline examination was 56.4 (±8.1) years, 49 795 participants were women (54.7%), 85 761 (94.3%) were White, 1454 (1.6%) were Black, and 2065 (2.3%) were Asian. Participants with higher total KB were more likely to be current or former smokers; have higher systolic and diastolic blood pressure; have higher low‐density lipoprotein‐cholesterol, high‐density lipoprotein‐cholesterol, and total cholesterol; and have higher prevalence of diabetes. In addition, participants in the higher tertiles of total KB levels reported lower intake of fat, carbohydrate, protein, and total energy. Figure S1–S5 demonstrates histograms of total KB, β‐OHB, acetoacetate, and acetone levels among UKB participants.
    Baseline Characteristics by Ketone Bodies Tertiles
    T1=total KB tertile 1, T2=total KB tertile 2, T3=total KB tertile 3.
    The 90 987 participants included.
    Indicates P value < 0.05.
    The 90 987 participants included in this analysis were followed for cardiovascular events for a median duration of 13.4 (interquartile range, 12.7–14.0) years. ASCVD was diagnosed in 1480 participants (4.9%) with total KB tertile 1 (T1), 2109 (6.9%) with T2, and 2242 (7.4%) with T3. HF was diagnosed in 698 (2.3%) with total KB T1, 1065 (3.5%) with T2, and 1178 (3.9%) with T3 (Table 2). Higher KB levels were also associated with an increased risk of all‐cause death, as evidenced by the elevated cumulative incidence and incidence rates across KB tertiles. Similar trends were observed for MI, stroke, and CVD death. Figure 2 display the cumulative incidence curves demonstrating that individuals in the higher total KB tertiles had a higher incidence of primary (ASCVD, HF) and secondary (MI, stroke, and CVD death) outcomes.
    Cumulative Incidence and Incidence Rates for All Outcomes
    IR indicates incidence rate; and KB, ketone bodies.
    ASCVD indicates atherosclerotic cardiovascular disease; and KB, Ketone bodies.
    The analysis using restricted cubic splines, which assessed the relationship between total KB across a continuous distribution and various outcomes, revealed a notably higher risk of ASCVD, HF, all‐cause death, stroke, and CVD death at elevated levels of total KB. However, no significant association was found with MI. Furthermore, the test for nonlinearity yielded statistically significant results for ASCVD (P=0.001), HF (P=0.005), all‐cause mortality (P<0.001), and CVD death (P=0.001), indicating a nonlinear relationship between total KB levels and these outcomes (Figure 3).
    Model adjusted for age, sex, race, ethnicity, Townsend deprivation index, body mass index, smoking, systolic blood pressure, diastolic blood pressure, diabetes, total cholesterol, high‐density lipoprotein cholesterol, triglycerides, statin use, hemoglobin A1c, creatinine, and alcohol consumption. The 3 knots were placed at the 5th, 50th, and 95th percentiles.
    In time‐to‐event models adjusted for age, sex, race, ethnicity, and Townsend deprivation index, every 10‐fold increase in total plasma KB at baseline was associated with increased risk of ASCVD (HR, 1.35 [95% CI, 1.22–1.49]; P<0.001), HF (HR, 1.49 [95% CI, 1.32–1.69]; P<0.001), and all‐cause death (HR, 1.55 [95% CI, 1.43–1.68]; P<0.001) (Table 3, Model 2). This association remained significant after adjusting for BMI, smoking, systolic and diastolic blood pressure, diabetes, total cholesterol, high‐density lipoprotein‐cholesterol, triglycerides, hemoglobin A1c, statin use, creatinine, and alcohol consumption (HR, 1.31 [95% CI, 1.18–1.46], 1.44 [95% CI, 1.24–1.67], and 1.51 [95% CI, 1.38–1.66], per 10‐fold increase in total KB, respectively) (Table 3, Model 3). We also observed a statistically significant association between higher total KB level and secondary outcomes of stroke (HR, 1.37 [95% CI, 1.11–1.69]) and CVD death (HR, 1.69 [95% CI, 1.43–2.00]) in the fully adjusted model (Table 3, Model 3). No significant association was observed between total KB and incident MI. In competing‐risk analyses, the Fine–Gray subdistribution HRs were consistent with the cause‐specific Cox models across all cardiovascular outcomes, indicating that associations between higher KB levels and incident events were robust after accounting for noncardiovascular death as a competing event (Table 3).
    Association Between Total Ketone Bodies and Incident Events
    Model 1: unadjusted. Model 2: adjusted for age, sex, race, ethnicity, Townsend deprivation index. Model 3: Model 2 + body mass index, smoking, systolic blood pressure, diastolic blood pressure, diabetes, total cholesterol, high‐density lipoprotein cholesterol, triglycerides, hemoglobin A1c, statin use, creatinine, and alcohol consumption. CVD indicates cardiovascular disease; HR, hazard ratio; KB, ketone bodies; and sHR, subdistribution hazard ratio.
    P value <0.05.
    We individually investigated the relationship between specific KB, namely β‐OHB, acetone, and acetoacetate. Compared with total KB, the association between individual KB and primary and secondary outcomes appeared consistent. Elevated levels of β‐OHB, acetone, and acetoacetate were associated with ASCVD, HF, all‐cause death, stroke, and CVD death, but no significant association was observed with MI (Table 4).
    Association Between Individual Ketone Bodies and Incident Events
    Model 1: unadjusted. Model 2: adjusted for age, sex, race, ethnicity, Townsend deprivation index. Model 3: Model 2 + body mass index, smoking, systolic blood pressure, diastolic blood pressure, diabetes, total cholesterol, high‐density lipoprotein cholesterol, triglycerides, hemoglobin A1c, statin use, creatinine, and alcohol consumption.
    β‐OHB indicates beta‐hydroxybutyrate; CVD, cardiovascular disease; and HR, hazard ratio.
    P‐value <0.05.
    In a sensitivity analysis, including physical activity, total daily energy intake, total daily carbohydrate intake, total daily fat intake, and total daily protein intake in the fully adjusted model, we found a similar association between elevated total KB and cardiovascular outcomes as well as all‐cause death (Table S2). To assess whether exclusion due to missing covariates introduced bias, we compared total KB concentrations and event rates between participants excluded due to missing data (n=22 071) and those included in the analytic sample (n=90 987). Both groups demonstrated nearly identical KB distributions and outcome rates (Table S3), supporting the assumption of noninformative missingness. In cross‐sectional analyses, higher total KB concentrations were significantly associated with higher odds of prevalent heart failure but not with prevalent MI or stroke (Table S4).
    Finally, there was no evidence of significant effect modification between variables of interest (sex, diabetes, BMI, or hypertension) and total KB levels across the various outcomes (P>0.10 for all).

    Discussion

    In a large population‐based cohort of individuals without established ASCVD or HF, we found that elevated circulating total KB were associated with increased risk of adverse cardiovascular events and mortality. Specifically, higher total KB levels were associated with incident ASCVD, HF, and all‐cause death. We also found a significant association of total KB with stroke and CVD death but not with MI. These findings suggest that the presence of higher levels of circulating KB may be detrimental to cardiovascular health and associated with an increased risk of cardiovascular events and mortality. The study also revealed a nonlinear relationship between circulating KB and incident cardiovascular outcomes, suggesting that there may be a threshold effect, where the risk of cardiovascular events and mortality increases once a certain concentration of circulating KB is reached.
    The underlying mechanisms through which circulating KB are associated with cardiovascular outcomes and mortality are not fully understood. However, several potential mechanisms could explain these findings. KB play crucial roles in various signaling pathways, posttranslational modification processes, and gene transcription mechanisms. These actions often influence endothelial function, inflammation, cell proliferation, oxidative stress, and cardiac remodeling.4 A metabolic shift toward using KB as a fuel source has been proposed as a potential mechanism underlying the beneficial effects of SGLT2 (sodium glucose transporter 2) inhibitors.20 In patients with HF, there is an elevation in the production and concentrations of KB. The failing heart undergoes a shift in energy metabolism, increasingly using KB as an alternative or emergency fuel source. This indicates an adaptive response to stress within the cardiovascular system.21 Exogenous ketone administration in patients with HF with reduced ejection fraction caused significant improvement in left ventricular ejection fraction and cardiac output.5 Even among healthy subjects, ketone ester ingestion increased systolic blood pressure, heart rate, biventricular function, and left ventricular and left atrial strain.22 While some animal studies have raised concerns about the impact of sustained KB elevation on insulin sensitivity and myocardial contractility, others suggest that context and exposure pattern may be key determinants of effect.23, 24 For example, intermittent fasting, SGLT2 inhibitors, and KB supplementation have demonstrated cardioprotective effects in mice and porcine models of cardiovascular disease.25, 26 These findings underscore the complex and possibly context‐dependent role of KBs in cardiometabolic health. One possible explanation for the observed difference in our results may relate to exposure time, as immediate alterations in KB levels might lead to distinct effects compared with those linked with chronic or lifelong exposure. Despite the lack of mechanistic studies, a growing body of evidence suggests that the metabolic transition toward KB signifies an adaptive response in individuals with or susceptible to CVD. Further research is needed to investigate these mechanisms and their implications for CVD comprehensively.
    Our results corroborate prior work suggesting that total KB are independently associated with CVD and all‐cause mortality and extend these findings to include ASCVD subtypes.6, 7, 10, 27 An observational study, which included 1758 participants from the cardiovascular health study, demonstrated that higher concentrations of KB were associated with incident heart failure and all‐cause mortality.10 A recent study of 87 899 UKB participants demonstrated a higher risk of incident CKD and all‐cause mortality among those with elevated total KB.27 Interestingly, patients experiencing an ST‐elevation MI exhibited notably elevated levels of circulating KB.8 The strong association between elevated KB levels and all‐cause mortality may also reflect noncardiovascular processes. For example, cancer‐related cachexia and altered tumor metabolism are known to influence KB production and may partially explain the heightened mortality risk in individuals with elevated KB. Moreover, increased concentrations of these KB were associated with left ventricular dysfunction 4 months following the MI in these individuals. Despite experimental studies suggesting the potential neuroprotective effect of KB, we found that total KB were associated with a higher risk of stroke.28, 29 Among 4662 participants of China Kadoorie Biobank study, plasma KB were associated with MI and stroke.30 Interestingly, we observed no significant association between KB and MI in fully adjusted models. This may reflect tissue‐specific metabolic responses, as recent data suggest the myocardium is capable of endogenous KB production under certain stress conditions, including in animal models of HF with preserved ejection fraction.31 These findings raise the possibility that myocardial KB metabolism could differ from systemic KB dynamics and merit further investigation. Although we observed robust associations between elevated circulating KB and adverse cardiovascular outcomes, it is likely that KB reflect a state of metabolic stress, insulin resistance, or subclinical CVD rather than serving as a direct causal factor. This interpretation is consistent with emerging evidence that therapies like SGLT2 inhibitors may exert cardioprotective effects by mimicking a nutrient deprivation signature, in which elevated KBs are a downstream consequence of altered energy metabolism rather than a pathogenic driver.32
    The heart and vasculature can readily metabolize β‐OHB and acetoacetate, and myocardial ketone oxidation is proportional to the circulating concentrations.33 Meanwhile, acetone is a metabolic byproduct of acetoacetate and remains inert. Consequently, acetone levels may offer the most accurate indication of ketogenesis, as fluctuations in the metabolism rates of the other KB can also influence their levels.34 In addition to analyzing total KB, our study also delved into the associations between individual KB and cardiovascular outcomes. This approach allowed for a more nuanced understanding of the potential impact of each specific KB on cardiovascular health and mortality. The findings revealed consistent associations between individual KB and incident cardiovascular outcomes, albeit with varying magnitudes of association. Elevated acetone levels exhibited the strongest association with ASCVD, HF, and mortality compared with β‐OHB and acetoacetate. Few studies have examined the association between individual KB and cardiovascular outcomes and have mainly used β‐OHB and examined their role in HF. A prospective population‐based cohort study including 6134 Dutch participants demonstrated that elevated levels of β‐OHB were associated with an increased risk of HF, particularly in women.7 A retrospective analysis involving 405 stable patients on hemodialysis demonstrated an independent association between elevated serum β‐OHB levels and cardiovascular events and all‐cause mortality.9 Our study highlights the importance of individual KB in assessing cardiovascular risk and emphasizes the need for future research to elucidate their distinct contributions to cardiovascular pathophysiology. Understanding the specific effects of individual KB on CVD is crucial to unraveling the underlying mechanisms.
    Despite the significant insights gained from this study on the associations between circulating KB and cardiovascular outcomes and mortality, it is important to acknowledge the limitations that warrant consideration. As with all observational studies, our findings are susceptible to residual confounding and cannot establish causality. Although we adjusted for numerous clinical, behavioral, and laboratory covariates, unmeasured confounders may still influence the observed associations. Therefore, circulating KB should currently be viewed as a marker of risk rather than a direct causal factor in CVD. The study relies on self‐reported data for certain variables, such as smoking status and dietary intake. Self‐reporting introduces the potential for recall bias and may not accurately capture the true dynamics of these factors. Given that the UKB cohort consists primarily of individuals of European ancestry and that KB were measured using a specific nuclear magnetic resonance‐based platform, our findings may not be directly generalizable to other populations or settings. Although KB levels can be measured using nuclear magnetic resonance or mass spectrometry platforms, these methods are not yet routinely used in clinical practice. Point‐of‐care assays for β‐OHB do exist (eg, in diabetes care), but further work is needed to standardize and validate KB measurement for routine cardiovascular risk assessment. The plasma samples in the UKB were collected under nonfasting conditions, and fasting duration was not recorded. Although dietary intake was assessed and included in sensitivity analyses, we lacked information on adherence to specific dietary patterns (eg, ketogenic diets) or medications that may influence KB levels (eg, SGLT2 inhibitors, insulin, or exogenous ketone supplements). Therefore, KB levels may reflect a combination of metabolic status, subclinical illness, and unmeasured exposures. Although restricted cubic spline models indicated statistically significant tests for nonlinearity, the associations appeared approximately monotonic. Therefore, for interpretability and consistency with prior literature, we summarized results using linear models. Finally, our study used a single baseline measurement of KB and did not include serial sampling. Thus, we were unable to assess intraindividual variability in KB levels over time or whether temporal changes in KB levels correspond to changes in cardiovascular risk. Future studies incorporating repeated KB measurements are needed to assess whether dynamic KB trajectories improve risk stratification.

    Conclusions

    In conclusion, our study provides robust evidence for the association between circulating KB and adverse cardiovascular outcomes and mortality in individuals without established ASCVD or HF. These results suggest that circulating KB levels may serve as valuable biomarkers for identifying individuals at higher risk of cardiovascular events and mortality. Future research should aim to elucidate the underlying mechanisms linking KB metabolism to cardiovascular outcomes and explore the clinical utility of monitoring KB levels as part of cardiovascular risk assessment strategies.

    Sources of Funding

    The UK Biobank is supported by the Wellcome Trust, the Medical Research Council, the Department of Health, the Scottish Government, and the Northwest Regional Development Agency, with additional funding from the Welsh Assembly Government and the British Heart Foundation. The authors conducted all components of the research independently of these funding sources.

    Disclosures

    Dr Pandey reported receiving funding support from the National Institute on Minority Health and Disparities (R01MD017529), National Institute on Aging GEMSSTAR Grant (1R03AG067960–01), and Gilead Sciences and has received consulting fees from Tricog Health Inc and Cytokinetics. Dr Shapiro reported receiving funding support from the National Institute Health, Amgen, Novartis, and Regeneron and served on scientific advisory boards for Amgen, Novartis, Novo Nordisk, Esperion, Regeneron, and Ionis. The remaining authors have no disclosures to report.

    Supporting information

    Tables S1–S5
    Figure S1
    STROBE_Checklist
    AMDr. Anurag Mehta

    Dr. Anurag Mehta

    Cardiology · Preventive Cardiology
    Emory University
    anuragm